MASARYK UNIVERSITY
Faculty of science
Jeffrey Clark Nekola
INTERDISCIPLINARY INVESTIGATIONS INTO THE
NATURE OF ECOLOGY AND EVOLUTION
Postcards from along the road
during my first quarter century of research
Habilitation thesis
Brno
2018
i
JEFFREY CLARK NEKOLA.
Interdisciplinary investigations into nature of ecology and evolution.
ABSTRACT
Why are there so many species of life on Earth? Why are they found where they are? How do
they live together? How have they evolved? Why do they look the way they do? What do they
tell us of the past and what do they suggest about the future? And how can they and humanity
coexist in this finite world? These questions - and others - underlie my research. My primary
focus is to understand the mechanisms which have given rise to our diverse world and the things
we need to do to protect it. This thesis represents a current state of play across my
interdisciplinary foray into the nature of ecology, accumulated during the first 25 years of
research. This work begins with an overview of why I have been active in so many ecological
fields, prefaced on the ancient Indian parable of the Blind Men and the Elephant. Following this,
I have broken my working into seven major categories, ranging from Taxonomy, Faunistics, and
Floristics through Population and Community Ecology, Spatio-temporal Ecology, Biogeography
and Macroecology, Ecological Theory and Modeling, Conservation Biology, and Human
Macroecology and Sustainability. The document ends with a concluding set of synthetic ideas
that have emerged from this intergrated multi-scale series of analyses.
ii
ACKNOWLEDGEMENTS
There have been innumerable people who has assisted my research efforts over the last
quarter century – too many to individually name here. But among them a few rise to the surface.
First, my PhD advisor Peter White has encouraged me from the start of my career to think
broadly and holistically about ecological questions. I am grateful for the encouragement that
Robert Cameron, Douglas Larson, Clifford Kraft, James Brown, and Brian McGill have provided
over the years. I would likely not have stayed in the field without their efforts. I also greatly
appreciate those over the last decade who have financially helped me keep the lights on in my
lab and home, in particular Charles Drost, Ken Hotopp and Ben Hutchens. Lastly, I deeply thank
Michal Horsák, Milan Chytrý and everyone else in the Department of Botany and Zoology at
Masaryk University for offering me a career lifeline and opportunity to continue my research
within the field.
iii
TABLE OF CONTENTS
Page
ABSTRACT ................................................................................................................................... i
ACKNOWLEDGEMENTS ........................................................................................................ ii
INTRODUCTION.........................................................................................................................2
SECTION I: Taxonomy, Floristics, and Faunistics
Overview ............................................................................................................................ 3
Nekola, J.C. & B.F. Coles. 2016. Supraspecific taxonomy in the Vertiginidae
(Gastropoda, Stylommatophora). Journal of Molluscan Studies. 82:208-212............ 6
Nekola, J.C., B.F. Coles & M. Horsák. 2015. Species assignment in Pupilla
(Gastropoda: Pulmonata: Pupillidae): integration of DNA-sequence data and
conchology. Journal of Molluscan Studies. 81:196-216............................................ 11
Nekola, J.C. & G. Rosenberg. 2013. Vertigo marciae (Gastropoda, Vertiginidae),
a new land snail from Jamaica. The Nautilus. 127:107-114...................................... 32
Nekola, J.C. 1998. Butterfly (Lepidoptera: Lycaeniidae, Nymphalidae, and
Satyridae) faunas of three peatland habitat types in the Lake Superior drainage
basin of Wisconsin. Great Lakes Entomologist. 31:27-37......................................... 40
SECTION II: Population and Community Ecology
Overview .......................................................................................................................... 52
Nekola, J.C. 2010. Acidophilic terrestrial gastropod communities of North
America. Journal of Molluscan Studies. 76:144-156................................................. 54
Nekola, J.C. 2004. Vascular plant composition gradients within and between
Iowa fens. Journal of Vegetation Science. 15:771-780.............................................. 67
Nekola, J.C. 2003. Large-scale terrestrial gastropod community composition
patterns in the Great Lakes region of North America. Diversity and Distribution.
9:55-71.......................................................................................................................... 77
Larson, D.W., U. Mattes-Sears, J.A. Gerrath, J.M. Gerrath, J.C. Nekola, G.L.
Walker, S. Porembski, A. Charlton & N.W.K. Larson. 1999. Ancient stunted
forests on cliffs. Nature. 398:382-383......................................................................... 94
SECTION III: Spatio-temporal Ecology
Overview .......................................................................................................................... 96
Pigati, J.S., J.A. Rech & J.C. Nekola. 2010. Radiocarbon dating of small terrestrial
gastropod shells in North America. Quaternary Geochronology. 5:519-532.............. 99
Nekola, J.C. & C.E. Kraft. 2002. Spatial constraint of peatland butterfly
occurrences within a heterogeneous landscape. Oecologia. 130:62-71.................... 113
Nekola, J.C. 1999. Paleorefugia and neorefugia: the influence of colonization
history on community pattern and process. Ecology. 80: 2459-2473....................... 122
iv
SECTION IV: Biogeography and Macroecology
Overview ........................................................................................................................ 137
Nekola, J.C. 2014. North American terrestrial gastropods through either
end of a spyglass. Journal of Molluscan Studies. 80:238-248................................... 139
Nekola, J.C. & P.S. White. 1999. Distance decay of similarity in biogeography
and ecology. Journal of Biogeography. 26:867-878................................................. 150
SECTION V: Ecological Theory and Modeling
Overview ........................................................................................................................ 162
Nekola, J.C. & B. McGill. 2014. Scale dependency in the functional form of
the distance decay relationship. Ecography. 37:309-320.......................................... 164
Nekola, J.C. & J.H. Brown. 2007. The wealth of species: ecological
communities, complex systems, and the legacy of Frank Preston. Ecology
Letters. 10:188-196..................................................................................................... 176
Bossenbroek, J.M., C.E. Kraft & J.C. Nekola. 2001. Prediction of long-distance
dispersal using gravity models: Zebra Mussel invasion of inland lakes.
Ecological Applications. 11:1178-1788...................................................................... 185
SECTION VI: Conservation Biology
Overview ........................................................................................................................ 196
Nekola, J.C. 2012. The impact of utility corridors on terrestrial gastropod
biodiversity. Biodiversity and Conservation. 21:781–795........................................ 197
McMillan, M., J.C. Nekola & D.W. Larson. 2003. Impact of recreational rock
climbing on land snail communities of the Niagara Escarpment, southern Ontario,
Canada. Conservation Biology. 17:616-621............................................................... 212
Nekola, J.C. 2002. Effects of fire management on the richness and abundance
of central North American grassland land snail faunas. Animal Biodiversity
and Conservation. 25:53-66........................................................................................ 218
SECTION VII: Human Macroecology and Sustainability
Overview ........................................................................................................................ 232
Nekola, J.C., C.D. Allen, J.H. Brown, J.R. Burger, A.D. Davidson, T.S. Fristoe,
M.J. Hamilton, S.T. Hammond, A. Kodric-Brown, N. Mercado-Silva & J.G. Okie.
2013. The Malthusian-Darwinian dynamic and the trajectory of civilization.
Trends in Ecology and Evolution. 28:127-130........................................................... 234
Burger, J.R., J.H. Brown, C.D. Allen,, W.R. Burnside, A.D. Davidson, T.S. Fristoe,
M.J. Hamilton, N. Mercado-Silva, J.C. Nekola, J.G. Okie & W. Zuo. 2012. The
macroecology of sustainability. PLoS Biology. 10:e1001345..................................... 238
Brown, J.H., W.R. Burnside, A.D. Davidson, J.P. DeLong, W.C. Dunn, M.J.
Hamilton, J.C. Nekola, J.G. Okie, N. Mercado-Silva, W.H. Woodruff & W. Zuo.
2011. Energetic Limits to Economic Growth. Bioscience. 61:19-26....................... 245
CONCLUSIONS ...................................................................................................................... 253
A group of blind men heard that a strange animal, called an elephant, had been brought to the
town, but none of them were aware of its shape and form. Out of curiosity, they said: "We must
inspect and know it by touch, of which we are capable". So, they sought it out, and when they
found it they groped about it. In the case of the first person, whose hand landed on the trunk,
said "This being is like a thick snake". For another one whose hand reached its ear, it seemed
like a kind of fan. As for another person, whose hand was upon its leg, said, the elephant is a
pillar like a tree-trunk. The blind man who placed his hand upon its side said, "an elephant is a
wall". Another who felt its tail described it as a rope. The last felt its tusk, and stated an elephant
is hard and smooth and like a spear. – Buddhist Udana 6.4
Six blind elephants were discussing what men were like. After arguing they decided to find one
and determine it by direct experience. The first felt the man and declared, 'Men are flat.' After
the other blind elephants felt the man, they agreed. – Mark Hartman
We have to remember that what we observe is not nature in itself, but nature exposed to our
method of questioning. – Werner Heisenberg
1
INTRODUCTION
Why are there so many species of life on Earth? Why are they found where they are? How do
they live together? How have they evolved? Why do they look the way they do? What do they
tell us of the past and what do they suggest about the future? And how can they and humanity
coexist in this finite world? These questions - and others - underlie my research. My primary
focus is to understand the mechanisms which have given rise to our diverse world and the things
we need to do to protect it.
As intimated by the ancient Indian parable of the Blind Men and the Elephant, the ability to
accurately interpret large-scale pattern – and identify potential underlying mechanism – is greatly
enhanced by viewing systems through multiple lenses, preferably at different observational
scales. This is especially important for ecological systems which range from sub-cellular to
global scales and include multiple potential focal entities ranging from molecules to biomes.
They are arranged into nested hierarchies of components, ranging from organic molecules and
cells through populations, species, guilds, and trophic levels. They require the continual input of
energy, material, and information to maintain their highly organized, far-from-equilibrium
thermodynamic states. Even the simplest ecological system contains thousands to billions of
individuals of tens to thousands of different species, ranging from unicellular prokaryotes,
protists, and fungi to multicellular plants and animals. These individuals and species interact
with each other and their extrinsic abiotic environment in inherently non-linear ways. Such
relationships are neither entirely deterministic nor stochastic, include positive and negative
feedbacks, and often possess long-lasting contingent effects across both space and time.
Discovering general processes supporting and maintaining diversification thus by necessity
requires an interdisciplinary focus. While this can certainly be accomplished via collaboration –
and I have participated in many such projects – I have also found it useful to be interdisciplinary
within my own brain. In this way I maintain hands-on, real-word experience across the full
range of biodiversity topics, and can use this knowledge to identify common threads and
understand potential analytical pitfalls. Often this leads to novel insights. While the
specialization of research programs can lead to fundamental discoveries, it is important to
remember that for almost a century biological and ecological research was conducted by
investigators who held expertise across a wide breadth of the natural sciences. 19th Century field
naturalists like Charles Darwin and Alfred Russell Wallace made lasting and often radical
impacts on the field, not the least of which was the development of evolutionary theory. Given
the complexity of ecological systems, some advances may only be possible via the inductive,
holistic, and non-experimental techniques pioneered by these scientists. In combination with the
modern tools of contemporary ecology, the field naturalist model can thus serve as an avenue for
important future advances in ecological theory.
Inspired by the astonishing breadth of knowledge accumulated by investigators such as Iowa’s
Bohumil Shimek, who became a national authority in botany, malacology, ecology,
paleoecology, and glacial geology (as well as being a close personal friend of Tomáš Masaryk
and champion for Czechoslovakian independence), I have organized my research program upon
the field naturalist model to address fundamental issues in the diversification and organization of
ecological communities. As a result, I have vertically integrated my research activities to range
2
6 of the approximately 75 North American Vertigo species
from organism taxonomy through population, community, and spatiotemporal ecology to
biogeography, macroecology, and ecological theory and modeling. To ensure that my outlook
does not become myopic, I have also horizontally integrated this program across three divergent
taxa groups: vascular plants (sessile autotrophs), lepidoptera (mobile heterotrophs), and
terrestrial gastropods (functionally sessile heterotrophs). In addition, I have maintained an active
field research program which spans thousands of study sites on three continents. I am
particularly interested in documenting findings that challenge current paradigms, as it is through
such observations that scientific revolutions are based.
The essential component of such an eclectic research program is to do what the blind men in the
above parable did not: integrate their findings to allow documentation of large-scale, crosssystem
reality. Accomplishing this task requires not only the ability to be expert within multiple
frames of reference, but also the ability to take a cross-system philosophic perspective to identify
potential correspondences and synergies. And like the elephants in the second parable, it is
essential that the limitations of given observational approaches and deductive / inductive
strategies be understood. Thus, as intimated by Heisenberg, by observing nature and biodiversity
through multiple methods of questioning, it may be possible to glimpse the true nature of nature.
Over the course of the last quarter century I have published 75 peer-reviewed works. As of
October 27, 2017 these have generated 5266 citations with an H-index=26, a G-index=72, and
an Age-Weighted Citation Rate=443.2. My research activities can be roughly grouped into
seven categories: (1) Taxonomy, Floristics and Faunistics; (2) Population and Community
Ecology; (3) Spatio-Temporal Ecology; (4) Biogeography and Macroecology; (5) Ecological
Theory and Modeling; (6) Conservation Biology; and, (7) Human Macroecology. Each is
represented in the following document by its own section. Each section leads off with a
summary statement of my research activates in that field, followed by a list of publications, and
then by representative examples of published articles.
Section I: Taxonomy, Floristics, and Faunistics
By being independent evolutionary units, species are one of the basic units underlying ecological
pattern and process. An accurate understanding of taxonomy and the valid taxa known within a
given system is thus a vital foundation for most ecological and biogeographic investigations. For
this reason, I maintain an active interest in taxonomy, floristics, and faunistics.
3
Much of my recent work centers on terrestrial gastropods, where I have named new species,
revised taxonomic concepts, and enumerated the taxa and distribution (both geographic and
ecological) of both regional and local faunas based on nDNA and mtDNA sequence, shell
features, ecological preferences, and biogeography. I have just completed a global phylogenetic
analysis of the micro-mollusk genus Vertigo based on analysis of nDNA (ITS1 and ITS2), and
mtDNA (CytB and 16S) amplicons from 91 putative taxa and 398 individuals. Using a
consensus across both sets of DNA sequence, shell features, ecological preferences, and
biogeography, six subgenera and 87 species or subspecies are empirically validated. This work
demonstrated: (1) a suite of diagnostic shell features usually exists to demarcate each specieslevel
taxon, though these are often not the features historically used; (2) ecological preferences
evolve more slowly as compared to shell features; (3) subgeneric transcontinental ranges are
ubiquitous; (4) 1/3 of all species possess continental or trans-continental ranges, with very few
having range extents <1000 km.; (5) all subgenera and fully 2/3 of the global Vertigo taxa reside
in North America, more than 2½ times the number found in central and eastern Asia, the second
most diverse region. This is similar to several other molluscan groups for which North America
serves as the global biodiversity hotspot. Earlier work also documents that diversification in the
genus appears synchronous with past episodes of rapid global environment change.
Determination of general trends and potential underlying mechanism requires that these analyses
be replicated in other distantly related groups. The Pristilomatidae are an obvious target being
not only in a different superfamily but also small (2-8 mm in diameter) and diverse, representing
the fifth largest family within the North American fauna (64 recognized species). Paravitrea (42
species) is the sixth most speciose genus. Unlike Vertigo many Pristilomatidae have very limited
ranges (often only a few counties in extent) with 100% of Pilsbryna species, 90% of Paravitrea
species and of 73% Pristiloma species being listed as Globally Threatened or higher. Specieslevel
diagnoses have historically relied on the number, shape, and placement of lamellae across
shell ontogeny. This is troubling given the significant incidence of ecophenotypic and
biogeographic plasticity in these features in other genera. As a result it is unclear how many
valid species-level groups actually exist or what their true biodiversity, biogeographic and
evolutionary patterns represent.
Representative Publications
[number of citations as of October 27, 2017]
1. Peer-reviewed Books:
Schlicht, D.W., J.C. Downey & J.C. Nekola. 2007. The Butterflies of Iowa. University of Iowa
Press. [26]
2. Peer-reviewed Articles:
Nekola, J.C., S. Chiba, B.F. Coles, C.A. Drost, T. vonProschwitz & M. Horsák. In press. A
phylogenetic overview of the genus Vertigo O. F. Müller, 1773 (Gastropoda: Pulmonata:
Pupillidae: Vertigininae). Malacologia.
Nekola, J.C. & B.F. Coles. 2016. Supraspecific taxonomy in the Vertiginidae (Gastropoda,
Stylommatophora). Journal of Molluscan Studies. 82:208-212. [4]
4
Nekola, J.C., B.F. Coles & M. Horsák. 2015. Species assignment in Pupilla (Gastropoda:
Pulmonata: Pupillidae): integration of DNA-sequence data and conchology. Journal of
Molluscan Studies. 81:196-216. [12]
Nekola, J.C. & G. Rosenberg. 2013. Vertigo marciae (Gastropoda, Vertiginidae), a new land
snail from Jamaica. The Nautilus. 127:107-114. [2]
Nekola, J.C., A. Jones, G. Martinez, S. Martinez, K. Mondragon, T. Lebeck, J. Slapcinsky & S.
Chiba. 2012. Vertigo shimochii Kuroda & Amano 1960 synonymized with Gastrocopta
servilis (Gould, 1843) based on conchological and DNA sequence data. Zootaxa.
3161:48-52. [3]
Nekola, J.C. & B.F. Coles. 2010. Pupillid land snails of eastern North America. American
Malacological Bulletin. 28:29-57. [38]
Nekola, J.C., B.F. Coles & U. Bergthorsson. 2009. Evolutionary pattern and process in the
Vertigo gouldii (Mollusca: Pulmonata, Pupillidae) group of minute North American land
snails. Molecular Phylogenetics and Evolution. 53:1010-1024. [38]
Coles, B.F. & J.C. Nekola. 2007. Vertigo malleata, a new extreme calcifuge land snail from
the Atlantic and Gulf coastal plains of the U.S.A. (Gastropoda, Vertiginidae). The
Nautilus. 121:17-28. [10]
Nekola, J.C. & M. Barthel. 2002. Morphometric analysis of the genus Carychium in the Great
Lakes region of North America. Journal of Conchology. 37(5):515-531. [18]
Nekola, J.C. & B.F. Coles. 2001. Systematics and ecology of Gastrocopta (Gastrocopta)
rogersensis (Gastropoda: Pupillidae), a new species from the midwest of the United
States of America. The Nautilus. 115:105-114. [16]
Nekola, J.C. 1998. Butterfly (Lepidoptera: Lycaeniidae, Nymphalidae, and Satyridae) faunas
of three peatland habitat types in the Lake Superior drainage basin of Wisconsin. Great
Lakes Entomologist. 31:27-37. [10]
Nekola, J.C. 1994. Environment and vascular flora of northeastern Iowa fen communities.
Rhodora. 96:121-169. [23]
5
RESEARCH NOTE
Supraspecific taxonomy in the Vertiginidae (Gastropoda: Stylommatophora)
Jeffrey C. Nekola1 and Brian F. Coles2
1
Department of Biology, University of New Mexico, Albuquerque, NM 87131, USA; and
2
Mollusca Section, Department of Biodiversity, Amgueddfa Cymru - National Museum Wales, Cathays Park, Cardiff CF10 3NP, UK
Correspondence: J. C. Nekola; e-mail: jnekola@unm.edu
The genus Vertigo Mu¨ ller, 1774 consists of c. 100 species of terrestrial
microsnails c. 1.5–3 mm in length with a rounded aperture
and 0–6 (sometimes more) apertural lamellae at maturity. As
currently defined, the genus is largely Holarctic in distribution
with only a few Neotropical species being known (Pilsbry, 1948;
Nekola & Rosenberg, 2013).
Consensus does not exist concerning supraspecific taxonomy
of the genus. Pilsbry (1919, 1927, 1948) placed Vertigo, Columella
Westerlund, 1878 and Truncatellina Lowe, 1852 in the family
Pupillidae, subfamily Vertigininae. Based largely on anatomy,
he differentiated these from the subfamily Nesopupinae, into
which he placed 11 mostly tropical genera: Bothriopupa Pilsbry,
1898, Campolaemus Pilsbry, 1892, Costigo Boettger, 1891,
Cylindrovertilla Boettger, 1880, Lyropupa Pilsbry, 1900, Nesopupa
Pilsbry, 1900, Pronesopupa Iredale, 1913, Ptychalaea Boettger,
1889, Pupisoma Stoliczka, 1873, Staurodon Lowe, 1854 and Sterkia
Pilsbry, 1898. Bouchet & Rocroi (2005) assigned Vertigo to the
family Vertiginidae, which in their scheme comprised three
subfamilies: Vertigininae, Nesopupinae (both largely as designated
by Pilsbry, 1927, but with Pupisoma being moved to the
Valloniidae) and Gastrocoptinae (including Gastrocopta Wollaston,
1878 and 10 related genera). The Gastrocoptinae have subsequently
been assigned to the Chondrinidae (Pokryszko et al., 2009).
Pilsbry (1919, 1948) recognized four subgenera within Vertigo:
Vertigo s.s. and the monotypic Angustula Sterki, 1888, Vertilla
Moquin-Tandon, 1855 and Vertillaria Pilsbry, 1919. He further
divided Vertigo s.s. into several informal sections and groups. This
treatment was followed until Turgeon et al. (1998) and Sysoev &
Schileyko (2009), respectively, considered the group Nearctula
Sterki, 1892 and the subgenus Vertilla to be of generic rank.
Similar intrageneric division has been prominent within Nesopupa;
e.g. Gittenberger & van Bruggen (2013) recognized nine genera
originally proposed as sections by Pilsbry & Cooke (Pilsbry, 1919):
Afripupa, Cocopupa, Helenopupa, Indopupa, Infranesopupa, Insulipupa,
Nesodagys, Nesopupa and Nesopupilla.
To address these issues empirically, we have assembled and
phylogenetically analysed DNA sequence data from the nuclear
28S ribosomal RNA (28S) and mitochondrial 16S ribosomal
RNA (16S) genes. The species selected for study included the
full conchological, biogeographic and ecological range of
Vertigo, plus representatives of Columella, Gastrocopta, Nearctula
and Truncatellina, and a variety of putative nesopupillids and
other genera within the Pupillidae (sensu Pilsbry, 1927). We also
included a representative sampling across the infraorder
Orthurethra as designated by Wade, Mordan & Clarke (2001).
Among the analysed taxa, the following represent the type
species of their respective genera: Acanthinula aculeata (Mu¨ ller,
1774), Chondrina avenacea (Bruguie` re, 1792), Cochlicopa lubrica
(Mu¨ ller, 1774), Columella edentula (Draparnaud, 1805), Helix
pomatia Linne´ , 1758, Lauria cylindracea (da Costa, 1778), Leiostyla
anglica (Wood, 1828), Nearctula californica (Rowell, 1862),
Planogyra asteriscus (Morse, 1857), Pupilla muscorum (Linne´ , 1758)
Pyramidula rupestris (Draparnaud, 1801), Solatopupa similis
(Bruguie` re, 1792), Sterkia calamitosa (Pilsbry, 1889), Strobilops
labyrinthica (Say, 1817), Vallonia costata (Mu¨ ller, 1774), Vertigo
pusilla (Mu¨ ller, 1774), Vertilla angustior (Jeffreys, 1830) and
Zoogenetes harpa (Say, 1824). Carychium tridentatum, Helix pomatia
and Cornu aspersum were used as comparative outgroups.
Archived GenBank sequences were used for 24 specimens, including
data first reported by Armbruster et al. (2005)
[AY546471], Dinapoli, Zinssmeister & Klussmann-Kolb (2010)
[GU331954], Gaitan-Espitia, Nespolo & Opazo (2013)
[JQ417194], Ketmaier et al. (2010) [jGU046389], Nekola &
Rosenberg (2013) [KF214500, KF214496], Nekola et al. (2012)
[JN941017, JN941032, JN941041, JN941044], Nekola, Coles &
Bergthorsson (2009) [GQ921543], Wade, Mordan & Naggs
(2006) [AY841284, AY841285, AY841286, AY841333], Wade et al.
(2001) [AY014019, AY014020, AY014022, AY014023, AY014024,
AY014025, AY014027, AY014028, AY014028, AY014030,
AY014032, AY014033, AY014040, AY014148] and Weigand
et al. (2013) [KC206171]. Sequences for the remaining 35 specimens
(Table 1) were newly obtained. DNA extraction, purification,
PCR amplification and sequencing were performed using
previously published methods and primers (see Wade &
Mordan, 2000; Nekola & Rosenberg, 2013). The amplified 28S
region (Fig. 1) encompasses ITS-2, a region that cannot be
aligned because of its intergeneric hypervariability. As a result,
all sequence more than 284 bp upstream of the LSU2 primer of
Wade & Mordan (2000) was excluded from analysis. The resultant
analysed 28S amplicon ranged in length from 809 bp
(Acanthinula aculeta) to 827 bp (Columella edentula). The entire 16S
amplicon was used for analysis and ranged in length from
403 bp (Cornu aspersum) to 594 bp (Truncatellina cylindrica).
Sequences were aligned using ClustalX, with adjustment by
eye. Mega v. 5.0 was used to conduct neighbour-joining (NJ),
maximum parsimony (MP) and maximum likelihood (ML)
analyses separately for the 28S and 16S sequences. NJ analysis
was based on maximum composite distance including transitions
# The Author 2015. Published by Oxford University Press on behalf of The Malacological Society of London, all rights reserved
Journal of The Malacological Society of London
Molluscan Studies
Journal of Molluscan Studies (2016) 82: 208–212. doi:10.1093/mollus/eyv034
Advance Access publication date: 20 July 2015
atUniversityofNewMexicoonFebruary15,2016http://mollus.oxfordjournals.org/Downloadedfrom
6
Table 1. Collection information for specimens analysed in this study.
Taxon Country State or equivalent County or equivalent Site Lat./Long. Collection/
accession no.
16S GenBank
acc. no.
28S GenBank
acc. no.
Afripupa bisulcata Ghana Fetish Grove 5.3356 N, 0.0801 W deWinter KT008314 KT008353
Bothriopupa tenuidens Costa Rica Puntarenas Saladera Lodge 8.6988 N, 83.3302 W Tattersfield KT008355
Columella edentula France Normandy Calvados La Vale´e des Vaux 48.9446 N, 0.4678 W BC00001 KT008310 KT008360
Leiostyla anglica UK Cumbria Meathop Cliff Seep 54.2066 N, 2.8729 W JCN KT008308 KT008363
Nearctula californica USA California Monterey Moss Landing Beach 36.8095 N, 121.7880 W JCN13934 KT008315 KT008349
Nearctula rowelli USA California Placer Auburn 38.9072 N, 121.0520 W JCN17232 KT008327 KT008347
Nesopupa newcombi USA Hawaii Honolulu Waianae Forest Reserve 21.5009 N, 158.1680 W UF469195 KT008307 KT008356
Planogyra asteriscus USA Maine Piscataquis Atkinson Mills 45.1421 N, 69.0312 W JCN10494 KT008305 KT008364
Pronesopupa boettgeri USA Hawaii Oahu Kalihi Valley 21.3585 N, 157.8580 W Hayes KT008306 KT008354
Ptychalaea tamagonori Japan Tokyo Ogasawara Chichijima 27.0948 N, 142.2166 E Chiba KT008323 KT008348
Pupilla muscorum Czech Rep. Moravia Brno 49.2509 N, 16.5738 E Horsak KT008313 KT008361
Pupisoma dioscoricola USA Florida Alachua Batram-Carr Woods 29.6436 N, 82.3456 W Slapcinsky KT008304 KT008357
Sterkia calamitosa USA California San Diego Sumner Canyon 32.8734 N, 117.2483 W JCN18169 KT008324 KT008350
Sterkia clementia USA California San Diego San Clemente Island 32.9953 N, 118.5516 W JCN19109 KT008325 KT008351
Sterkia hemphilli USA California San Diego Border Fields State Park 32.5428 N, 117.1061 W JCN19780 KT008331 KT008352
Strobilops labyrinthica USA Michigan Mackinac Brevort Lake 45.9829 N, 84.8618 W JCN17151 KT008309 KT008362
Truncatellina callicratis France Normandy Calvados Pont, N. of Falaise 48.9774 N, 0.0902 W BC00002 KT008302 KT008358
Truncatellina cylindrica Ukraine Crimea Rozovyi 44.6920 N, 34.3190 E Cameron KT008303 KT008359
Vallonia costata USA New Mexico Bernalillo Albuquerque 35.0727 N, 106.6160 W JCN KT008312
Vertigo alabamensis USA Florida Leon Wolf Trap Bay 30.3680 N, 84.5700 W JCN12323 KF214500 KT008339
Vertigo antivertigo UK Wales Anglesey Waun Eraud Fen 53.3008 N, 4.2411 W JCN KT008316 KT008338
Vertigo clappi USA Tennessee Carter Lynn Mountain 36.3545 N, 82.1535 W UF299578 KT008321 KT008337
Vertigo cristata USA Wisconsin Oneida Sugar Camp Bog 45.8499 N, 89.2958 W JCN12213 GQ921543 KT008344
Vertigo gouldii USA West Virginia Lincoln Big Ugly WMA 38.0859 N, 82.0009 W Doursen KT008330 KT008333
Vertigo cf. okinerabuensis Japan Iwate Ichinoseki Sarusawa 38.9869 N, 141.2550 E Chiba JN941044 KT008341
Vertigo malleata USA North Carolina Pender Holly Shelter Game Land 34.5492 N, 77.7817 W UF449308 KT008318 KT008345
Vertigo meramecensis USA Kentucky Bullitt Cave Hollow 37.9379 N, 85.6334 W JCN KT008329 KT008335
Vertigo milium USA Iowa Buchanan Rowley North fen 42.3764 N, 91.8507 W JCN5747 KT008328 KT008346
Vertigo moulinsiana Netherlands South Holland Kaag en Braassem Woubrugge, Widje Aa 52.1710 N, 4.6119 E JCN KT008326 KT008334
Vertigo pseudosubstriata Russia Altai Republic Seminski Pass S 50.9855 N, 85.6817 E JCN KT008317 KT008343
Vertigo pusilla Czech Rep. Moravia Podyji National Park 48.8586 N, 15.8960 E Horsak KF214496 KT008342
Vertigo pygmaea Czech Rep. Moravia Kotrle Fen 49.3779 N, 18.0236 E JCN KT008322 KT008340
Vertigo rugosula USA Arkansas Jefferson Lock & Dam #5 34.4040 N, 92.1020 W UF409059 KT008319 KT008332
Vertilla angustior Czech Rep. Moravia Pozdechov Fen 49.2339 N, 17.9864 E JCN KT008320 KT008336
Zoogenetes harpa Canada Quebec Manicougan Pointe-des-Montes 49.3256 N, 67.3700 W JCN13609 KT008311
Abbreviations: BC, Coles collection in National Museum of Wales, of which the accession numbers are to be preceded by ‘NMW.Z.2015.009’; JCN, Nekola collection in University of New Mexico; UF, material in Florida
Museum of Natural History, Gainesville.
RESEARCHNOTE
209
atUniversityofNewMexicoonFebruary15,2016 http://mollus.oxfordjournals.org/ Downloadedfrom
7
Figure 1. Map of the rRNA gene cluster with primer locations for the analysed segment of 28S in relation to the 5.8S and ITS-2 regions.
Figure 2. Maximum-likelihood phylogenetic tree reconstruction based on nuclear 28S (left) and mitochondrial 16S (right) data. Nodes with strong to
moderate support (!70) across all four phylogenetic reconstruction methods have been labelled to the left of that node by four support values: upper
left (normal font) is for NJ; upper right (bold italic font) is for MP; lower left (bold font) is for Bayesian; lower right (italic font) is for ML.
RESEARCH NOTE
210
atUniversityofNewMexicoonFebruary15,2016http://mollus.oxfordjournals.org/Downloadedfrom
8
and transversions with pairwise gap deletion. MP analysis used
the close-neighbour interchange search option with the random
addition of 10 replicate trees. ML analysis used all sites and was
based on the Tamura-Nei substitution model, a five-category
gamma distribution for substitution rates and the nearest neighbour
interchange ML heuristic method. In all cases support
values were estimated from 1,000 bootstrap replicates. Additionally,
Bayesian trees were generated using MrBayes v. 3.1 (Huelsenbeck
& Ronquist, 2001), with a generalised time reversible substitution
model assuming gamma-shaped rate variation over 1,000,000
generations and a sampling frequency of once each 1,000
generations.
Analysis of these data (Fig. 2) demonstrates poorer resolution
using the mitochondrial 16S as compared with the nuclear 28S
data, presumably due to higher base-pair saturation rates.
Therefore, we used the 28S data to deduce phylogenetic relationships
and the 16S data for quasi-independent corroboration.
The 28S data show a highly-supported clade composed of two
highly-supported sister clades, together including most of the
putative Vertigininae (Clade V) and Nesopupinae (Clade N)
included in this study. This ‘vertiginid’ clade in general corresponds
with the Vertiginidae as outlined by Bouchet & Rocroi
(2005) and Pokryszko et al. (2009), with the marked exceptions
that: (1) Gastrocopta is excluded as it is neither closely related to
Vertigo nor to the chondrinids, but rather represents a distinct
branch within the Orthurethra, and (2) both Columella and
Truncatellina are excluded as both are members of a moderately
supported clade that includes Chondrina. Wade et al. (2001, 2006)
have shown that the chondrinids are actually sister to the
Orthurethra. The analysis of 16S data in general corroborates
these findings, albeit without resolving the sister status of clades
V and N.
The 28S and 16S trees both demonstrate that Clade N consists
only of tropical species historically assigned to the Nesopupinae.
These are known to differ anatomically from the Vertigininae
by possessing a penial appendix and a forked retractor muscle
(Pilsbry, 1919). Clade V includes not only all analysed Vertigo,
but also the putative genera Nearctula and Vertilla and members of
some genera historically assigned to the Nesopupinae (Afripupa,
Ptychalaea and Sterkia). Pilsbry (1919) lacked anatomical data for
any species within these latter three genera and placed them
outside of Vertigo solely on biogeographical (tropical/subtropical
range) and conchological grounds (pustulate shell sculpture,
and/or strength and location of angular lamella). Previous analyses
have shown that these shell features are not reliable for use in
supraspecific taxonomy of pupillids (Coles & Nekola, 2007;
Nekola et al., 2009; Nekola, Coles & Horsa´ k, 2015).
It should also be noted that for both the 28S and 16S trees,
monophyly of Vertigo is only preserved when the taxonomic
concept of the genus is expanded to include all members of
Clade V. Because support levels for this clade are similar to other
apparent genus-level clades (including Cochlicopa, Gastrocopta, Lauria
and Truncatellina), our data suggest that it would be most parsimonious
to assign all species within Clade V to the genus Vertigo. Our
more limited sampling and representation of type species in the
narrowly-constrained Nesopupinae (Clade N) makes it impossible
for us to say how many generic-level entities might be supported in
that group. However it is interesting to note that the genetic divergence
between Bothriopupa and Nesopupa is well within the range
found within Vertigo. It thus remains an open question how many
genera within the Nesopupinae will ultimately be supported by
DNA sequence data.
These data do not support subgeneric classification within
Vertigo, given that no nodes within Clade V possess support
values .70 in either 28S or 16S analyses. However, the weakly
supported monophyly of Nearctula/Sterkia suggests that useful
subgeneric groupings may in fact exist. Consideration of these
issues must await analysis of other amplicons that are more
rapidly evolving (therefore with more taxonomically useful sites)
than 28S, but more slowly evolving (thus with lower base-pair
saturation rates) than 16S.
If retained, the family Vertiginidae represents a highly supported
clade possessing only two branches: Vertigo on one hand
and a subset of nesopupids on the other. But how useful is recognition
of so small a family? We agree with Puillandre et al.
(2015) that such decisions are ultimately matters of taxonomic
opinion that cannot be subjected to empirical criteria. In this we
freely admit to erring on the side of conservatism, as we do not
see the utility of erecting a large number of high-level taxonomic
groups that constitute only single branches or simple two-branch
entities when a more inclusive division containing more taxa is
possible by simply stepping back one level in the tree.
Even though outside our primary focus, these analyses also
hint that other traditional supraspecific concepts within the
Orthurethra may be in need of revision. In particular, the
Valloniidae may be polyphyletic with both Pupisoma and
Acanthinula/Planogyra each representing their own unique
branches within the Orthurethra. The 16S data also suggest that
Zoogenetes may be more closely related to Pupisoma. Pilsbry
(1927: vii) appears to have been prescient when he stated “The
division of the Orthurethra into families seems to be largely a
matter of expediency.” These empirical data validate this statement
and strongly suggest that formal reconsideration of supraspecific
concepts across the entire infraorder are warranted,
based upon DNA sequence data.
ACKNOWLEDGEMENTS
Specimens used in this analysis were generously provided by
Robert Cameron, Satoshi Chiba, Dan Doursen, Ton de Winter,
Ken Hayes, Michal Horsa´ k, John Slapcinsky, John Stanisic and
Peter Tattersfield. Generous personal donations by Michal
Horsa´ k covered laboratory expenses and this work would simply
have been impossible without his assistance.
REFERENCES
ARMBRUSTER, G.F.J., BOEHME, M., BERNHARD, D. &
SCHLEGEL, M. 2005. The H3/H4 histone gene cluster of land
snails (Gastropoda: Stylommatophora): TS/TV ratio, GC3 drive
and signals in Stylommatophoran phylogeny. Journal of Molluscan
Studies, 71: 339–348.
BOUCHET, P. & ROCROI, J.-P. 2005. Classification and
nomenclature of gastropod families. Malacologia, 47: 1–397.
COLES, B.F. & NEKOLA, J.C. 2007. Vertigo malleata, a new extreme
calcifuge land snail from the Atlantic and Gulf coastal plains of the
U.S.A. (Gastropoda, Vertiginidae). Nautilus, 121: 17–28.
DINAPOLI, A., ZINSSMEISTER, C. & KLUSSMANN-KOLB, A.
2010. New insights into the phylogeny of the Pyramidellidae
(Mollusca, Gastropoda). Journal of Molluscan Studies, 7: 1–7.
GAITAN-ESPITIA, J.D., NESPOLO, R.F. & OPAZO, J.C. 2013. The
complete mitochondrial genome of the land snail Cornu aspersum
(Helicidae: Mollusca): intra-specific divergence of protein-coding
genes and phylogenetic considerations within Euthyneura. PLoS One,
8: E67299.
GITTENBERGER, E. & VAN BRUGGEN, A.C. 2013. Land snails of
the islet of Misali, off Pemba Island, Zanzibar, Tanzania. Zoologische
Mededelingen, 87: 235–273.
HUELSENBECK, J.P. & RONQUIST, F. 2001. MRBAYES: Bayesian
inference of phylogeny. Bioinformatics, 17: 754–755.
KETMAIER, V., MANGANELLI, G., TIEDEMANN, R. &
GIUSTI, F. 2010. Prei-Tyrrhenian phylogeography in the land snail
Solatopupa guidoni (Pulmonata). Malacologia, 52: 81–96.
NEKOLA, J.C., COLES, B.F. & BERGTHORSSON, U. 2009.
Evolutionary pattern and process in the Vertigo gouldii (Mollusca:
Pulmonata, Pupillidae) group of minute North American land
snails. Molecular Phylogenetics and Evolution, 53: 1010–1024.
RESEARCH NOTE
211
atUniversityofNewMexicoonFebruary15,2016http://mollus.oxfordjournals.org/Downloadedfrom
9
NEKOLA, J.C., COLES, B.F. & HORSA´ K, M. 2015. Species assignment
in Pupilla (Gastropoda: Pulmonata: Pupillidae): integration of DNAsequence
data and conchology. Journal of Molluscan Studies, 81: 196–216.
NEKOLA, J.C., JONES, A., MARTINEZ, G., MARTINEZ, S.,
MONDRAGON, K., LEBECK, T., SLAPCINSKY, J. & CHIBA,
S. 2012. Vertigo shimochii Kuroda & Amano 1960 synonymized with
Gastrocopta servilis (Gould, 1843) based on conchological and DNA
sequence data. Zootaxa, 3161: 48–52.
NEKOLA, J.C. & ROSENBERG, G. 2013. Vertigo marciae (Gastropoda,
Vertiginidae), a new land snail from Jamaica. Nautilus, 127: 107–114.
PILSBRY, H.A. 1919. Manual of conchology, second series. Vol. 25: Pupillidae
(Vertigininae, Pupillinae). Academy of Natural Sciences, Philadelphia.
PILSBRY, H.A. 1927. Manual of conchology, second series. Vol. 28:
Geographic distribution of Pupillidae. Academy of Natural Sciences,
Philadelphia.
PILSBRY, H.A. 1948. Land Mollusca of North America (north of Mexico).
Academy of Natural Sciences of Philadelphia Monograph, 3: vol. 2(2).
POKRYSZKO, B.M., AUFFENBERG, K., HLAVA´ Cˇ , J. & NAGGS,
F. 2009. Pupilloidea of Pakistan (Gastropoda: Pulmonata):
Truncatellininae, Vertigininae, Gastrocoptinae, Pupillinae (in part).
Annales Zoologici (Warszawa), 59: 423–458.
PUILLANDRE, N., DUDA, T.F., MEYER, C., OLIVERA, B.M. &
BOUCHET, P. 2015. One, four or 100 genera? A new classification
of the cone snails. Journal of Molluscan Studies, 81: 1–23.
SYSOEV, A. & SCHILEYKO, A. 2009. Land snails and slugs of Russia
and adjacent Countries. Pensoft, Sofia.
TURGEON, D.D., QUINN, J.F., JR., BOGAN, A.E., COAN, E.V.,
HOCHBERG, F.G., LYONS, W.G., MIKKELSEN, P., NEVES,
R.J., ROPER, C.F.E., ROSENBERG, G., ROTH, B.,
SCHELTEMA, A., THOMPSON, F.G., VECCHIONE, M. &
WILLIAMS, J.D. 1998. Common and scientific names of aquatic
invertebrates from the United States and Canada: mollusks. Edn 2.
American Fisheries Society Special Publication 26, Bethesda, MD.
WADE, C.M. & MORDAN, P.B. 2000. Evolution within the gastropod
molluscs; using the ribosomal RNA gene-clustre as an indicator of
phylogenetic relationships. Journal of Molluscan Studies, 66: 565–570.
WADE, C.M., MORDAN, P.B. & CLARKE, B. 2001. A phylogeny of
the land snails (Gastropoda: Pulmonata). Proceedings of the Royal
Society B, 268: 413–422.
WADE, C.M., MORDAN, P.B. & NAGGS, F. 2006. Evolutionary
relationships among the Pulmonate land snails and slugs
(Pulmonata, Stylommatophora). Biological Journal of the Linnean
Society, 87: 593–610.
WEIGAND, A.M., JOCHUM, A., SLAPNIK, R., SCHNITZLER, J.,
ZARZA, E. & KLUSSMANN-KOLB, A. 2013. Evolution of
microgastropods (Ellobioidea, Carychiidae): integrating taxonomic,
phylogenetic and evolutionary hypotheses. BMC Evolutionary Biology,
13: 18.
RESEARCH NOTE
212
atUniversityofNewMexicoonFebruary15,2016http://mollus.oxfordjournals.org/Downloadedfrom
10
Species assignment in Pupilla (Gastropoda: Pulmonata: Pupillidae): integration of
DNA-sequence data and conchology
Jeffrey C. Nekola1, Brian F. Coles2 and Michal Horsa´ k3
1
Biology Department, University of New Mexico, Albuquerque, NM 87131, USA;
2
Mollusca Section, Department of Biodiversity, Amgueddfa Cymru – National Museum Wales, Cathays Park, Cardiff CF10 3NP, UK; and
3
Department of Botany and Zoology, Masaryk University, Kotla´rˇska´ 2, Brno CZ-611 37, Czech Republic
Correspondence: J.C. Nekola; e-mail: jnekola@unm.edu
(Received 23 October 2013; accepted 2 September 2014)
ABSTRACT
Using the Pupilla faunas of Europe, North America, the Altai region of central Asia and eastern Asia,
we consider whether the existing taxonomy based primarily on shell apertural characteristics correlates
with relationships established on the basis of mitochondrial and nuclear DNA-sequence data. We
obtained DNA sequence from nuclear ITS1 and ITS2 and mitochondrial COI and CytB from 80 specimens
across 22 putative Pupilla taxa. The sequence data were analysed using maximum likelihood,
maximum parsimony, Bayesian and neighbour-joining phylogenetic tree reconstruction, as well as
base-pair substitution and insertion-deletion analysis. Revised species-level concepts were generated by
identifying reciprocally monophyletic clades that exhibited unique conchological features. These analyses
document that, although many previously described taxa have biological merit, the highly plastic
nature of shell apertural features makes them unreliable indicators of species identity in several independent
lineages. However, shell surface sculpture and architecture appear to provide more reliable
diagnoses. Because of the traditional reliance of species-level taxonomy in Pupilla on plastic apertural
features, too many species-level entities have been described in Europe and the Altai. Also, because
taxonomically useful shell sculpture features have tended to be ignored, too few species have been
described in eastern Asia and North America. As a result, confusion exists about species ranges, ecological
tolerances and interpretation of Quaternary fossils within the genus. Based on these analyses
three new species are described: P. alaskensis, P. hudsonianum and P. hokkaidoensis.
INTRODUCTION
The use of protein markers and DNA sequence data has shown
that conchological traits can be poor indicators of relatedness,
for example in eastern North American Polygyridae (Emberton,
1995), Thailand Gastrocoptinae (Tongkerd et al., 2004) and the
clausiliid subfamily Alopiinae from Greece (Uit de Weerd et al.,
2004). Similarly, our analysis of DNA sequence data in the
genus Vertigo (Pupilloidea) indicates that while shell features do
generally provide accurate assignment of genetically validated
species-level entities, they are too labile to resolve evolutionary
relatedness (Nekola et al., 2009).
We now extend our molecular studies to include the genus
Pupilla, a close relative of Vertigo. Both genera are in the
Orthurethra and have long been considered to be in the same
family, the Pupillidae (Pilsbry, 1948; Hubricht, 1985). Like
Vertigo, Pupilla is distributed across the entire Holarctic and
members of both genera commonly co-occur within the same sites.
Pupilla species possess a minute shell (,5 mm in height) of conserved
cylindrical-ovate form, although shell apertures show considerable
variability, ranging from simple to callused and/or
lamellate. Consequently, apertural features have been much used
as diagnostic species-specific characters (Pilsbry, 1921, 1948).
However, preliminary DNA sequence analysis of putative North
American P. muscorum suggested that this emphasis on shell apertural
characters for species assignment was unsatisfactory (Nekola
et al., 2009). We have thus undertaken the present study to
examine whether the classically used apertural traits used to identify
Pupilla species are able accurately to assess taxonomy supported
by mitochondrial and nuclear DNA sequence data across
the Holarctic range of the genus. Because genitalic structure has
been found to be of only limited utility in making species-level
taxonomic distinctions in this genus (Pokryszko et al., 2009),
DNA sequence data offer the only practical non-shell-based
method for cross-validation of taxonomic concepts in Pupilla.
METHODS
Specimen selection and identification
Specimens used for analysis were primarily obtained from collections
made in 2000–2012 (Nekola, 2005; Horsa´ k et al., 2010,
# The Author 2014. Published by Oxford University Press on behalf of The Malacological Society of London, all rights reserved
Journal of The Malacological Society of London
Molluscan Studies
Journal of Molluscan Studies (2015) 81: 196–216. doi:10.1093/mollus/eyu083
Advance Access publication date: 21 December 2014
atUniversityofNewMexicoonMay4,2015http://mollus.oxfordjournals.org/Downloadedfrom
11
2012; Horsa´ k, Chytry´ & Axmanova´ , 2013; Nekola & Coles,
2010). These include most of the currently recognized Pupilla
taxa from western and central Europe (east to the Ural
Mountains), central Asia (Altai Republic), Japan (Hokkaido)
and North America (Canada and the USA including Alaska),
as established by original descriptions, authoritative accounts of
regional molluscan faunas and monographs (Pilsbry, 1921,
1948; Kerney & Cameron, 1979; Schileyko, 1984; Meng &
Hoffman, 2008; Pokryszko et al., 2009; von Proschwitz et al.,
2009). We were unable to secure tissue samples from only two
species within these target regions: P. seminskii Meng &
Hoffman, 2009 (Altai Republic) and P. sterkiana Pilsbry, 1889
(North America). For each analysed taxon, an attempt was
made to select multiple individuals from across their known geographic
and ecological range. Six specimens also represent topotype
or near-topotype material: AP2 (P. altaica), AP13 (P.
alluvionica), ET7 (P. muscorum xerobia), P1 (P. hebes kaibabensis),
P10 (P. syngenes) and P12 (P. sonorana). Archival museum material
up to 65 years old up to was used to supplement the specimen
set for P. triplicata (specimens H19-21; Table 1).
Each specimen was taxonomically assigned using currently
recognized diagnostic conchological features (Table 2) as
reported by Pilsbry (1921, 1948), Kerney & Cameron (1979),
Schileyko (1984), Meng & Hoffman (2008), Pokryszko et al.
(2009) and von Proschwitz et al. (2009). In these works, apertural
lamellar architecture has been given particular weight,
with little variation being reported in their number, shape or
placement within a given taxonomic concept. Apertural crest
size, callus development and colour are also frequently used as
diagnostic features. Shell sculpture, suture depth and shell apex
shape have been used less frequently to distinguish some entities.
Based on these diagnoses, shells from all analysed individuals
and their respective populations were examined for nine conchological
traits (see below).
DNA extraction, PCR amplification and sequence analysis
Live specimens of Pupilla were preserved in absolute ethanol,
allowed to desiccate at ambient temperature and humidity, or in
some cases were used before death. DNA was extracted using the
Omega BioTek Mollusk DNA Extraction Kit. Because of the inability
of water to displace air within these tightly coiled shells,
shell destruction was required to allow access of proteinase to
mummified tissue. Thus (with few exceptions) specimens were
taken from lots containing multiple examples of each respective
taxon, with the actual specimens used for DNA preparation
being imaged at 15Â magnification prior to shell destruction
using methods described by Nekola, Coles & Bergthorsson
(2009).
The internal transcribed spacers (plus flanking sequence) of
the nuclear ribosomal RNA complex (ITS1 and ITS2), and
mitochondrial cytochrome oxidase subunit I (COI) and cytochrome
b (CytB) were amplified using published methods with
modifications as listed in Table 3. PCR products were sequenced
in both forward and reverse directions using Perkin Elmer ABI
Big Dye termination and standard protocols. COI and CytB
sequences were also obtained from the GenBank database for
data analysed by von Proschwitz et al. (2009) that could be unambiguously
assigned to a single individual (P. muscorum, BadenWu¨
rtemberg, Germany; P. pratensis, Lagmansro, O¨ stergo¨ tland,
Sweden; P. pratensis, Mecklenburg-Vorpommern, Germany)
and for two outgroups (Vertigo pusilla and Gastrocopta cristata) that
were previously analysed by Nekola & Rosenberg (2013).
Phylogenetic analyses
Sequences (excluding primer regions) were aligned using
ClustalX with adjustment by eye for ITS1 and ITS2. COI and
CytB were concatenated, and ITS1 and ITS2 sequences were
analysed as a single construct by omitting 81 invariant base
pairs from the intervening 5.8S region. Mega v. 5.0 was used to
conduct neighbour-joining (NJ), maximum parsimony (MP)
and maximum likelihood (ML) analyses separately for the concatenated
nuclear and mitochondrial DNA sequences. NJ analysis
was based on maximum composite distance including
transitions and transversions with pairwise gap deletion. MP
analysis used the close-neighbour interchange search option
with the random addition of 10 replicate trees. ML analysis used
all sites and was based on the Tamura-Nei substitution model, a
five-category gamma distribution for substitution rates, and
the nearest neighbour interchange ML heuristic method. In
all cases support values were estimated from 1000 bootstrap
replicates. Additionally, Bayesian trees were generated using
MrBayes v. 3.1 (Huelsenbeck & Ronquist, 2001), using a GTR
substitution model assuming gamma-shaped rate variation over
1,000,000 generations with a sampling frequency of once each
1000 generations. Because none of these methods makes full
informatic use of insertions and deletions, we also constructed a
matrix of all variable bases in both the ITS1 and ITS2 regions,
including not only base-pair substitutions, but also insertions
and deletions.
Post-hoc species delimitation and conchology
Identification of potential species-level (and higher) clades
based on DNA sequence data was accomplished by examining
the nDNA and mtDNA trees for highly supported, reciprocallymonophyletic
clades. This approach was of limited value for the
nDNA data, because of low node support due to limited variation
of 90 informative sites across 1500 bp. To help resolve
relationships using these data, we examined the matrix of variable
sites by eye for base-pair substitutions, insertions and deletions
held in common among groups of sequences. Apparent
incongruencies in specimen placement between the nuclear vs
mitochondrial sequences were identified as potential cases of
interspecific mitochondrial introgression or incomplete lineage
sorting.
We have not used any of the various methods for species demarcation
using single-locus analyses of base-pair variation
(e.g. generalized mixed Yule-coalescent functions). Although
we have previously used these methods (Nekola et al., 2009),
they universally require generation of ultrametric trees, which
assume constant evolutionary rates across all clades. As a
result, these methods do not function well when base-pair substitution
rates are clade-specific. Because assumption of rate
homogeneity appeared unjustified within the current Pupilla
dataset, we have instead opted for reciprocal-monophyly as our
decision-rule to identify potential species-level clades based on
genetic data.
We then attempted to verify the biological validity of these
potential genetically-supported species concepts by reanalysing
shell features from the imaged shells as well as additional shells
within each analysed population. The range of expressed shell
variation within each reciprocally-monophyletic species-level
clade was documented for nine conchological traits: height
(mm), width (mm), shell form, apex shape, shell sculpture,
suture depth, aperture shape, apertural crest and callus strength,
and apertural lamellae number and configuration. Potential
species-level clades were considered taxonomically validated
when some subset of the above shell features was found to be
unique to and thus diagnostic of that entity. Based on this
revised taxonomy, we then updated biogeographic and ecological
information for each species based on our extensive community
ecology datasets (e.g. Horsa´ k et al., 2010; Nekola, 2014) in
combination with other published accounts.
TAXONOMY OF PUPILLA
197
atUniversityofNewMexicoonMay4,2015http://mollus.oxfordjournals.org/Downloadedfrom
12
Table1.InformationforspecimensofPupillausedinDNAsequenceanalysis.
Taxon/locationLatitude/longitudeSpecimencodeCollectionandlotno.Namesupportedby
DNAsequencedata
(whendifferent)
GenBankaccessionno.
CO1CytBITS1ITS2
PupillaalluvionicaMeng&Hoffman,2008
Russia
Belyashi,Altai49.26918N,87.98388EAP13JNKM518545KM518468KM518390KM518312
Pupillaalpicola(Charpentier,1837)
Russia
Belyashi,Altai49.52068N,88.01808EAP12JNKM518546KM518469KM518391KM518313
Slovakia
Raksˇa48.87928N,18.89018EH6MHKM518547KM518470KM518392KM518314
Vazˇec49.06858N,19.99948EH11MHKM518548KM518471KM518393KM518315
Raksˇa48.87928N,18.89018EH12MHKM518549KM518472KM518394KM518316
Liptovska´Teplicˇka48.96328N,20.10448EH13MHKM518550KM518473KM518395KM518317
PupillaaltaicaMeng&Hoffman,2008
Russia
Ustˇ-Muny,Altai51.72978N,85.73828EAP1JNP.turcmenicaKM518551KM518474KM518396KM518318
Kurai,Altai50.30808N,87.64858EAP2JNP.turcmenicaKM518552KM518475KM518397KM518319
Kurai,Altai50.23348N,87.78948EAP17MHP.turcmenicaKM518553KM518476KM518398KM518320
Pupillabigranata(Rossma¨ssler,1839)
France
Pont,Calvados48.97748N,0.09038WAP25BC2014.013.00081P.muscorumKM518554KM518477KM518399KM518321
PupillablandiMorse,1865
Canada
Irvine,Alberta49.95958N,110.25118WAP34BC2014.013.00079KM518555KM518478KM518400KM518322
MooseJaw,Saskatchewan50.04348N,105.62468WAP35BC2014.013.00074KM518556KM518479KM518401KM518323
USA
LoganCanyon,Utah41.74268N,111.76038WP7JN18327P.hebesKM518557KM518480KM518402KM518324
LasVegas,NewMexico35.64158N,105.18758WP15JN18417KM518558KM518403KM518325
RubyMountains,Nevada40.77518N,115.33648WP16JN18280P.hebesKM518559KM518481KM518404KM518326
WilsonArch,Utah38.27458N,109.37248WP18JN18237P.hebespithodesKM518560KM518482KM518405KM518327
PupillablandicharlestonensisPilsbry,1948
USA
EastTinticRange,Utah39.96608N,112.04848WP2JN18257P.hebesKM518561KM518483KM518406KM518328
BullionCanyon,Utah38.41718N,112.31268WAP28JN17211P.hebespithodesKM518562KM518484KM518407KM518329
PupillablandipithodesPilsbry&Ferriss,1917
USA
BullionCanyon,Utah38.41718N,112.31268WAP27JN17210KM518563KM518485KM518408KM518330
TusasRidge,NewMexico36.65198N,106.03818WAP38JN13013KM518564KM518486KM518409KM518331
Bland,NewMexico35.75048N,106.45698WP9JN14816KM518565KM518487KM518410KM518332
Pupillahebes(Ancey,1881)
Japan
J. C. NEKOLA ET AL.
198
atUniversityofNewMexicoonMay4,2015http://mollus.oxfordjournals.org/Downloadedfrom
13
Toyokoro,Nakagawa,Hokkaido42.60508N,143.55648EVH29JNP.hokkaidoensisKM518566KM518488KM518411KM518333
USA
KnikIsland,Anchorage,Alaska61.50848N,149.03438WAP29JN15390P.alaskensisKM518567KM518489KM518412KM518334
UteCreekCanyon,Colorado38.58488N,105.96868WAP37JN12898P.blandiKM518568KM518490KM518413KM518335
HappyValley,Alaska69.33558N,148.73028WNS48JN15142P.alaskensisGQ921663KM518491KM518414KM518336
JarbidgeMountains,Nevada41.68678N,115.50618WP5JN18292P.hebespithodesKM518569KM518492KM518415KM518337
LoopeEast,California38.65918N,119.72228WP14JN17254KM518570KM518493KM518416KM518338
PupillahebeskaibabensisPilsbry&Ferriss,1911
U.S.A.
KaibabPlateau,Arizona36.82998N,112.25428WP1JN18400P.hebesKM518571KM518494KM518417KM518339
PupillahebesnefasPilsbry&Ferriss,1910
USA
SantaCatalinaMts.,Arizona32.41968N,110.73118WP6JN14052P.hebespithodesKM518572KM518495KM518418KM518340
Pupillacf.khunjerabicaAuffenberg&Pokryszko,2009
Russia
Chagan-Uzun,Altai50.08698N,88.39418EAP11MHKM518573KM518496KM518419KM518341
Pupillacf.limataSchileyko,1989
Russia
Bestyakh,Yakutia61.36248N,128.84338EAP20MHKM518574KM518497KM518420KM518342
Kapitonovka,Yakutia62.32928N,129.92828EAP39MHKM518575KM518498KM518421KM518343
PupillaloessicaLozˇek,1954
Russia
Belyashi,Altai49.41868N,87.59288EAP5JNKM518576KM518499KM518422KM518344
Kosh-Agach,Altai49.99298N,88.54968EAP6JNKM518577KM518500KM518423KM518345
Belyashi,Altai49.28048N,87.49558EAP7JNKM518578KM518501KM518424KM518346
Kosh-Agach,Altai49.66098N,88.22788EAP8JNKM518579KM518502KM518425KM518347
Belyashi,Altai49.26918N,87.98388EAP9JNKM518580KM518503KM518426KM518348
Belyashi,Altai49.52068N,88.01808EAP10JNKM518581KM518504KM518427KM518349
Ulagan,Altai50.47678N,87.63018EAP19MHKM518582KM518505KM518428KM518350
Pupillamuscorum(Linnaeus,1758)
Europe
CzechRepublic
Brno,Moravia49.25098N,16.57388EmtG-PupMHKM518583KM518506KM518429KM518351
USA(naturalized)
CedarRapids,Iowa41.98668N,91.74008W22JN14592GQ921664
Syracuse,NewYork43.00748N,76.11058WAP26JN13955KM518584KM518507KM518430KM518352
Canada
Cochrane,Alberta51.26428N,114.73268WAP36BC2014.013.00058P.hudsonianumKM518585KM518508KM518431KM518353
Churchill,Manitoba58.70868N,94.12308WP8JN11098P.hudsonianumKM518586KM518509KM518432KM518354
LaGrandePointe,Quebec50.20598N,63.39688WP13BC2014.013.00001P.hudsonianumKM518587KM518510KM518433KM518355
USA
BullionCanyon,Utah38.41718N,112.33038WP17JN17219P.hebesKM518588KM518511KM518434KM518356
Continued
TAXONOMY OF PUPILLA
199
atUniversityofNewMexicoonMay4,2015http://mollus.oxfordjournals.org/Downloadedfrom
14
Table1.Continued
Taxon/locationLatitude/longitudeSpecimencodeCollectionandlotno.Namesupportedby
DNAsequencedata
(whendifferent)
GenBankaccessionno.
CO1CytBITS1ITS2
LakeBemidji,Minnesota47.53288N,94.82478W23JN9054P.hudsonianumGQ921662KM518357
LakeBemidji,Minnesota47.53288N,94.82478WAP33JN9054P.hudsonianumKM518589KM518512KM518435KM518358
PupillamuscorumxerobiaPilsbry,1914
USA
BannonRanch,NewMexico36.91668N,103.78008WET7JN16491P.blandiKM518590KM518513KM518436KM518359
LakeBemidji,Minnesota47.53288N,94.82478WP4JN9054P.blandiKM518591KM518514KM518437KM518360
Pupillapratensis(Clessin,1871)
Russia
Aktash,Altai50.44728N,87.60788EH3MHP.alpicolaKM518592KM518515KM518438KM518361
Ulagan,Atlai50.47678N,87.62188EH4MHP.alpicolaKM518593KM518516KM518439KM518362
Europe
CzechRepublic
Pozdeˇchov,Moravia49.23398N,17.98648EH1MHP.alpicolaKM518594KM518517KM518440KM518363
Vysoke´My´to,Bohemia49.96118N,16.18928EH7MHP.alpicolaKM518518KM518441KM518364
Slovakia
Za´vod48.53318N,16.99638EH5MHP.alpicolaKM518595KM518519KM518442KM518365
Pupillasonorana(Sterki,1899)
USA
SacramentoMountains,NewMexico32.71418N,105.75418WP12BC2005.011.03117KM518596KM518520KM518443KM518366
Pupillasterrii(Forster,1840)
Albania
PeriferiDibre41.81728N,20.50038EAP16MHKM518597KM518521KM518444KM518367
CzechRepublic
Pavlov,Moravia48.87738N,16.66358EH8MHKM518598KM518522KM518445KM518368
Praha-Hlubocˇepy,Bohemia50.04198N,14.37618EH14MHKM518599KM518523KM518446KM518369
Klentnice,Moravia48.84678N,16.64058EAP22MHKM518600KM518524KM518447KM518370
Russia
VerkhneBikberda,Bashkortostan52.32078N,56.80328EAP15MHKM518601KM518525KM518448KM518371
Slovakia
Valaska´Dubova´49.15118N,19.64418EAP21MHKM518602KM518526KM518449KM518372
Pupillasyngenes(Pilsbry,1890)
USA
Mogollon,NewMexico33.39448N,108.80568WAP30BC2005.011.02961KM518603KM518527KM518450KM518373
KaibabPlateau,Arizona36.69188N,112.29898WP10BCKM518604KM518528KM518451KM518374
PupillasyngenesdextroversaPilsbry&Vanatta,1900
USA
KaibabPlateau,Arizona36.69188N,112.29898WP11BCP.syngenesKM518605KM518529KM518452KM518375
Pupillatriplicata(Studer,1820)
Russia
J. C. NEKOLA ET AL.
200
atUniversityofNewMexicoonMay4,2015http://mollus.oxfordjournals.org/Downloadedfrom
15
OzeroKureevo,Altai52.48118N,85.76058EAP32MHKM518606KM518530KM518453KM518376
CzechRepublic
Tocˇnı´k,Bohemia49.89058N,13.88668EH2MHKM518607KM518531KM518454KM518377
Pavlov,Moravia48.87738N,16.66358EH9MHKM518608KM518532KM518455KM518378
Milesˇov,Bohemia50.53448N,14.94378EH10MHKM518609KM518533KM518456KM518379
Kamy´k,Bohemia50.56438N,14.08878EH15MHKM518610KM518534KM518457KM518380
Chroustov,Moravia49.17188N,16.05028EH16MHKM518611KM518535KM518458KM518381
Hracholusky,Bohemia49.99748N,13.79048EH17MHKM518612KM518536KM518459KM518382
Louny,Bohemia(1948)H19VLKM518613KM518537KM518460KM518383
Kamy´k,Bohemia(1970)H20VLKM518614KM518538KM518461KM518384
Pupillatriplicata(Studer,1820)
CzechRepublic
Srdov,Bohemia(1950)H21VLKM518615KM518539KM518462
France
Cahors,Dordogne44.47728N,1.43038EAP31BC2014.013.00082KM518616KM518540KM518463KM518385
Russia
Nugush,Bashkortostan53.00668N,56.53568EAP23MHKM518617KM518541KM518464KM518386
Pupillaturcmenica(O.Boettger,1889)
Russia
Kosh-Agach,Altai50.07298N,88.72018EAP3MHKM518618KM518542KM518465KM518387
Belyashi,Altai49.29558N,87.73448EAP4MHKM518619KM518543KM518466KM518388
Kurai,Atlai50.23348N,87.78948EAP18MHKM518620KM518544KM518467KM518389
AllspecimensfromtheNekolacollection(JN)arecurrentlymaintainedattheUniversityofNewMexicobeforeultimatedepositionattheAcademyofNaturalSciencesatDrexelUniversity(ANSP).Allspecimensfromthe
Colescollection(BC)arehousedattheNationalMuseumofWales;allaccessionnumbersforthismaterialareprecededby‘NMW.Z’.AllspecimensfromtheHorsa´kcollection(MH)arehousedattheDepartmentofBotany
andZoologyatMasarykUniversity,Brno.MaterialfromtheVojenLozˇekcollection(VL)ishousedatCharlesUniversity,Prague.
TAXONOMY OF PUPILLA
201
atUniversityofNewMexicoonMay4,2015http://mollus.oxfordjournals.org/Downloadedfrom
16
Table2.HistoricaltaxonomicandtraditionalconchologicalconceptsforanalysedPupilla.
TaxonHeight
(mm)
Width
(mm)
ShellformApexshapeShellsculptureSuturedepthAperturalcrestAperturalcallusAperturallamellae
alluvionica3.3–4.32.2–2.4WidecylindricalTapered/
domed
Smooth/weakstriaeClearly
evident
Hardlythickened,
clearlywhite
0–1(rareweakparietal)
alpicola2.8–3.31.8–1.9WidecylindricalDomedObviousDeepWeak-absent0–1(rareweakparietal)
altaica2.5–3.21.6–1.8CylindricalovoidDomedFineribswithdermal
edges
PronouncedThick,white2–3(angularpad;palatalrare)
bigranata3.0–3.61.6CylindricalovoidTaperedAlmostsmooth;fine
linesonly
ShallowPresentThick,white2–3(rarecolumellar)
blandi2.5–3.31.5–1.6CylindricalDomedDelicatestriationsDeepPresentYellow/tan3
blandi
charlestonensis
3.01.4CylindricalDomedSlightlystriateDeepWell
developed
Brown3(palatal,parietalstrong,long)
blandipithodes3.1–3.71.7–1.8WidecylindricalDomedDelicatestriaeDeepLow-strongWeak;brown3
hebes3.1–4.01.6–1.9SubcylindricDomedMinutestriaeDeepNone-weakNone0–1(rareparietal)
hebeskaibabensis2.7–2.81.5SubcylindricDomedMinutestriaeDeepNone-weakNone0–1(rareparietal)
hebesnefas3.2–4.21.7–1.9Subcylindric,sinistralDomedMinutestriaeDeepNone-weakNone1(parietal)
cf.khunjerabica3.2–4.01.7–1.9CylindricalovoidTapered/
domed
Almostsmooth,striae
faint,irregular
Moderately
deep
WeakThin-lacking0
cf.limata2.8–3.11.6–1.7CylindricalovoidTapered/
domed
DelicatestriaeModerately
deep
Absent-weakNone0
loessica3.0–3.41.7–
1.75
CylindricalovoidTapered/
domed
Finely,irregularlyribbedDeepAbsent-weakNone0
muscorum3.2–4.01.7CylindricalovoidTaperedAlmostsmooth;fine
linesonly
ShallowStrongThick,white0–2
muscorumxerobia2.3–2.81.4–1.5CylindricalovoidDomedAlmostsmooth;fine
linesonly
ShallowStrongThick,white1(parietal)
pratensis3.5–4.51.9–2.1CylindricalovoidDomedFinestriationDeepWeakWeak,white0–2(weakifpresent);no
depressionoverpalatal
sonorana2.5–3.31.3–1.4CylindricalDomedFinestriaeDeepStrongThick,white3(alllong)
sterrii2.8–3.51.6CylindricalTapered/
domed
CoarsestriaeVerydeepWeak-
moderate
Moderatelythick,
white
2(bothpeg-shaped)
syngenes3.0–4.21.7–1.8Biconic,widestatc.4thwhorl;8
narrowwhorls;sinistral
Tapered/
domed
DelicatestriaeShallowStrong3
syngenes
dextroversa
3.0–4.51.6–1.8Biconic,widestatc.4thwhorl;8
narrowwhorls;dextral
Tapered/
domed
DelicatestriaeShallowStrong3
triplicata2.2–2.8
(4)
1.4CylindricalDomedFine,closestriationDeepModerateDistinct,white3(bladeshaped)
turcmenica3.0–3.21.1–1.4Cylindrical;shellthinnerTapered/
domed
CoarsestriaeVerydeepWeakWeak,white0–2(weakparietaland/or
palatal)
J. C. NEKOLA ET AL.
202
atUniversityofNewMexicoonMay4,2015http://mollus.oxfordjournals.org/Downloadedfrom
17
RESULTS
DNA sequence data
A total of 80 specimens from 22 putative Pupilla taxa underwent
DNA extraction (1–12 individuals/taxon, see Table 1). DNA
sequences were obtained for 79 specimens for COI, 77 specimens
for CytB and 77 specimens for ITS1 and ITS2. All COI and
CytB amplicons consisted of 655 and 377 bp, respectively, and
could be unambiguously aligned. The COI amplicon contained
240 and CytB 137 variable sites. The ITS1 amplicon length was
615–635 bp and the ITS2 amplicon length was 868–874 bp.
All Pupilla ITS sequences could be unambiguously aligned, but
those of the outgroup (Vertigo pusilla and Gastrocopta cristata)
could not. The 5.8S region between the ITS amplicons (based
on 23 Pupilla specimens for which it was determined) consisted
of 81 invariant bases so that the entire contiguous sequence
between the 50
-end of the ITS1 amplicon and the 30
-end of the
ITS2 amplicon is 1569–1593 bp. The total informative sites in
ITS consisted of 56 bp substitutions and 36 bp comprising 12
insertions/deletions.
Phylogenetic reconstructions and supported taxonomic entities
Phylogenetic tree reconstruction and base-pair variation maps
based on concatenated COI þ CytB mtDNA and ITS1 þ ITS2
nDNA sequences support the presence of 17 putative reciprocallymonophyletic
species or subspecies-level taxa (Figs 1 and 2,
Table 4): P. alluvionica, P. alpicola, P. blandi, P. cf. khunjerabica, P.
cf. limata, P. hebes, P. hebes pithodes, P. loessica, P. muscorum, Pupilla
n. sp. (Alaska), Pupilla n. sp. (Hokkaido), Pupilla n. sp.
(Hudsonian), P. sonorana, P. sterrii, P. syngenes, P. triplicata and
P. turcmenica. The three new species identified by these analyses
are formally described below and will be referred to hereafter as
P. alaskensis, P. hokkaidoensis and P. hudsonianum, respectively.
Pupilla alluvionica is a xeric rock outcrop and steppe species
that differs from all other analysed Pupilla by possessing adult
shells .2.1 mm in diameter with a smooth or very weakly
striate shell, a crest and a white callus. The single individual
analysed for DNA sequence possessed an ITS1 þ ITS2 sequence
with four bases different from all other Pupilla (111C and 577A
in ITS1; 333G and 490C in ITS2). However, its COI þ CytB
sequence was part of the same highly supported clade defining
P. turcmenica, which consistently co-occurs with P. alluvionica in
the Altai.
Pupilla alpicola is a wetland species whose shells are up to
2.1 mm wide with a shallow suture and a body whorl often slightly
narrower than the penultimate whorl. This species is defined
by 92C and usually 340C in ITS2. Two subpopulations are
noted, one with AC at 171–172 in ITS1 and the other maintaining
the consensus GT at these positions. While the former subpopulation
is more prevalent in Europe and the latter in central
Asia, individuals characteristic of either occur throughout its
known range. Pupilla pratensis has been traditionally differentiated
from P. alpicola by lacking a depression or flattening on the
palatal wall of the aperture and having a rather pronounced
shell apex (von Poschwitz et al., 2009). However, P. pratensis
shares the same unique ITS2 bases as P. alpicola, with individuals
referable to P. pratensis occurring in both of the unique ITS1 subpopulations.
Additionally, P. pratensis mtDNA occurs throughout
the same highly supported clade that contains all analysed
P. alpicola. If P. pratensis is considered a shell form of P. alpicola,
then P. alpicola is monophyletic for both nDNA and mtDNA.
According to Pilsbry (1948), P. blandi is characterized by a cylindrical
shell with a prominent crest, a yellow to tan callus and
three apertural lamellae. Using these criteria, individuals
conchologically assignable to P. blandi demonstrate polyphyly
both in ITS1 þ ITS2 and COI þ CytB, occurring within three
well-supported species-level clades. However, if these traditional
conchological characteristics are abandoned in favour of surface
sculpture, with P. blandi being differentiated by its irregular,
very weak striae, shiny shell surface and shallow suture, this
species becomes a strongly supported monophyletic entity for
COI þ CytB that uniquely possesses a GAC insertion at 181–
183 in ITS2. Because this entity varies in apertural lamella
number from 0 to 3, and has a crest and callus ranging from
weak to strong, it includes a number of shell forms that were previously
assigned to other taxa including P. hebes and P. muscorum
xerobia.
Pilsbry (1948) characterized P. hebes as possessing a minutely
striate, subcylindric shell with no apertural callus, an absent to
rarely weak apertural crest and absent to rarely weak parietal
lamella. Shells displaying these traits demonstrate polyphyly
among three different species-level clades. However, as with P.
blandi, monophyletic grouping is apparent when a different suite
of shell features is used to diagnose P. hebes, including a
cylindrical-ovoid shell tapered for the upper 1/3-1/4 of the shell
height, a normal to deep suture and possession of numerous sharp
thread-like striae. Shells possessing these features all uniquely
possess 508T in ITS1, while sharing 495A in ITS1 with P. blandi
and P. hebes pithodes. Using the shell characters of Pilsbry (1948)
for identification causes some individuals within this group to be
incorrectly assigned to P. blandi, P. blandi charlestonensis and P. muscorum.
It should be noted that COI þ CytB suggests that P. hebes
exists as two discrete subpopulations, one ranging throughout the
Great Basin from California to north-central Utah (samples P5,
P7, P14 and P16) and the other being restricted to the canyonlands
region of the Colorado Plateau (samples P1, P2 and P17).
This latter subpopulation would equate to P. hebes kaibabensis of
Pilsbry (1948). However, as the shells of this clade completely
overlap with typical material as well as possessing identical
ITS1 þ ITS2 sequence, it seemed best not to formally recognize
this subpopulation at this time.
Pilsbry (1948) characterized P. blandi pithodes as being wider
than P. blandi, with a weak to absent crest and callus. He
hypothesized that it was intermediate between P. blandi and P.
hebes. ITS1 þ ITS2 indicate that in fact this entity is more
Table 3. Primers used for genetic analysis.
Region Direction Anneal Sequence Source
COI f 458C 5′
-ATTCAACGAATCATAAAGATATTGG-3′
Author Design
r 5′
-TATACTTCAGGATGACCAAAAAACCA-3′
Author Design
CytB f 478C 5′
-TGAGGTGCAACAGTNATTAC-3′
Author Design
r 5′
-GCAAATAAAAAGTATCACTCTGG-3′
Author Design
ITS1 f 528C 5′
-TAACAAGGTTTCCGTATGTGAA-3′
Armbruster & Bernhard (2000)
r 5′
-TCACATTAATTCTCGCAGCTAG-3′
Author Design
ITS2 f 528C 5′
-CTAGCTGCGAGAATTAATGTGA-3′
Wade & Mordan (2000)
r 5′
-GGTTTCACGTACTCTTGAAC-3′
Author Design
TAXONOMY OF PUPILLA
203
atUniversityofNewMexicoonMay4,2015http://mollus.oxfordjournals.org/Downloadedfrom
18
Figure 1. Maximum-likelihood phylogenetic tree reconstruction for Pupilla based on concatenated ITS1 þ ITS2 data. Nodes with strong to moderate
support across all four phylogenetic reconstruction methods have been labelled to the left of that node by four support values: upper left (normal font)
is for NJ; upper right (bold italic font) is for MP; lower left (bold font) is for Bayesian; lower right (italic font) is for ML. Branch tip labels represent
initial identifications based on traditional conchological features, whereas labels to the right of brackets represent valid names supported by both
nDNA and mtDNA sequence analysis.
J. C. NEKOLA ET AL.
204
atUniversityofNewMexicoonMay4,2015http://mollus.oxfordjournals.org/Downloadedfrom
19
Figure 2. Maximum-likelihood phylogenetic tree reconstruction for Pupilla based on concatenated COI þ CytB. Labelling conventions as in Figure 1.
Specimens that have a significantly different topological location as compared with the nDNA tree are highlighted in gray.
TAXONOMY OF PUPILLA
205
atUniversityofNewMexicoonMay4,2015http://mollus.oxfordjournals.org/Downloadedfrom
20
Table4.MatrixofvariablebasesinITS1andITS2.
J. C. NEKOLA ET AL.
206
atUniversityofNewMexicoonMay4,2015http://mollus.oxfordjournals.org/Downloadedfrom
21
Forbothamplicons,numbersrefertothebasepairsdownstreamoftheterminationoftheforwardprimer.Basepairsinvariantacrossallsamplesareomitted,whilebasepairsthatdivergefromthegenusconsensusare
highlighted.Dashindicatesbase-pairdeletion.Samplesaresortedbypost-hocsupportedspeciesnames(inbolditalic),whileinitialnamesusedpriortoDNAanalysis,andbasedontraditionalconchologicalfeatures,are
listedintheSampleinformation.Potentialgroupingsofrelatedspeciesarenotedinthefarthest-rightcolumn,labelledSG.Notethattheinsertions:ITS168–71and72–75;andITS2181–183and504–506,are
microsatelliterepeats.
TAXONOMY OF PUPILLA
207
atUniversityofNewMexicoonMay4,2015http://mollus.oxfordjournals.org/Downloadedfrom
22
closely related to P. hebes, lacking the GAC insertion at 181–183
in ITS2. However, it is distinguished from this species by possessing
508G in ITS1. The COI þ CytB of P. hebes pithodes exists as
a highly supported clade. Because we observed a gradation from
barrel-shaped P. pithodes to more cylindrical P. hebes shells, we
have chosen to consider P. pithodes as a subpopulation that
occurs to the east of the main range of P. hebes, ranging from
eastern Arizona and central Utah into Colorado, New Mexico
and Texas. Use of traditional shell characteristics (Pilsbry, 1948)
results in some P. hebes pithodes being incorrectly assigned to P.
blandi charlestonensis. ITS1 þ ITS1 and COI þ CytB both indicate
that the sinistral P. hebes nefas of southeastern Arizona
should be regarded as a shell form of P. hebes pithodes.
DNA analysis also demonstrates that P. alaskensis and P. hokkaidoensis,
which have previously been confused with P. hebes, actually
represent undescribed species related to the central Asian
P. loessica, as all share 207A in ITS2. Both of these new species
uniquely share 177G in ITS2, with P. hokkaidoensis differing from
P. alaskensis based on an 11-bp deletion at 38–48 in ITS1, 111C
in ITS1 and 79C in ITS2. Pupilla alaskensis shells differ from P.
hebes in their widely spaced, coarse, somewhat anastomosing
striae. Pupilla hokkaidoensis shells differ from P. hebes in their
rotund cylindrical shape with shallow suture and anastomosing
coarse striae (see taxonomic descriptions below).
Pupilla cf. khunjerabica is represented in the sample by a single
population from a riparian forest corridor in the Altai. This specimen
possesses a cylindrical-ovoid shell tapered for the upper
half of the shell height, very weak irregular thread-like striae, a
moderately deep suture, a weak to absent callus or crest, and no
apertural lamellae. Although its COI þ CytB sequence fell
within the highly supported clade defining P. alaskensis, it
uniquely possessed 137C in ITS2, and was the only taxon
outside the southwestern North America P. syngenes/sonorana
group to possess 156G in ITS2.
Pupilla cf. limata was sampled from riparian forest in the
Yakutia region of eastern Siberia and is characterized by a
cylindrical-ovoid shell with numerous, sharp, somewhat anastomosing
striae, a weak crest and no callus or apertural lamellae.
While this taxon is highly supported as a monophyletic entity in
COI þ CytB, one of the two analysed individuals possessed
ITS1 þ ITS2 identical with P. hebes pithodes from western North
America. The other individual uniquely possessed 444C in ITS1.
Pupilla loessica is a steppe-tundra species that in modern times
is restricted to central Asia, although Pleistocene fossils extend
west into central Europe (e.g. Horsa´ k et al., 2010). It is distinguished
by its numerous strong rounded anastomosing striae
and lack of callus and apertural lamellae. It uniquely possessed
266C in ITS1 and shares 207A in ITS2 with P. alaskensis and P.
hokkaidoensis. Pupilla loessica exists as a strongly supported clade
in COI þ CytB. While one individual (AP6) possessed COI þ
CytB characteristic of P. alaskensis, this individual had ITS1 þ
ITS2 characteristic of P. loessica.
Pupilla muscorum has traditionally been characterized by its cylindrical
shell with smooth sculpture, shallow suture, strong
crest, thick white callus and from zero to two parietal lamellae
(Pilsbry, 1948; von Proschwitz et al., 2009). This species exists as
a well defined clade in ITS1 þ ITS2 by uniquely possessing
375C and 609A in ITS1 and 336C in ITS2. It also exists as a
strongly supported clade in COI þ CytB. Pupilla bigranata,
which is distinguished from P. muscorum by its three strong apertural
lamellae, was found to have COI þ CytB sequence identical
with Moravian P. muscorum and ITS1 þ ITS2 sequence
identical with New York P. muscorum. Individuals from the High
Plains and Rocky Mountains of the western USA, which have
been previously identified as P. muscorum using the above conchological
characteristics, were shown by DNA sequence analysis
actually to represent P. blandi or P. hebes.
In North America P. muscorum was thought to exist in two disjunct
populations, one ranging from the northeastern Atlantic
seaboard west through the Great Lakes to Iowa, and the other
ranging across the northern taiga (Oughton, 1948; Hubricht,
1985). This latter entity (P. hudsonianum) exists as a well defined
monophyletic group uniquely possessing 56A, 541G and 583C
in ITS2. It also exists as a strongly supported clade in COI þ
CytB, although one individual (P13) possessed sequence characteristic
of P. alaskensis. It is easily differentiated from P. muscorum
by the strong taper over the top half of its shell height and its
sculpture of dense thread-like striae (see taxonomic description
below).
Pupilla syngenes possesses a distinctive shell that is widest in the
top half and has three apertural lamellae including a long,
curved blade-like parietal lamella. This species of wooded, xeric
rock outcrops in the southwestern USA uniquely possessed 83A
in ITS2 and also exists as a strongly supported clade in COI þ
CytB. The dextral individual of this typically sinistral species,
termed P. syngenes dextroversa, had ITS1 þ ITS2 and COI þ
CytB sequences identical to a typical individual within the same
population.
Pupilla sonorana was compared by Pilsbry (1948) with P. triplicata
and distinguished by its small size, columnar shell, strong
crest and three apertural lamellae including a curved blade-like
parietal lamella. It appears closely related to P. syngenes by
sharing 453T in ITS1 and 156G and 233A in ITS2. However, it
differed from that species in possessing 218A and 538C and a
GGCA insertion from 68 to 71 in ITS1 and 83G and 356A in
ITS2. As only one individual was analysed, no species-level
clade can be assigned in COI þ CytB. However, it differed from
P. syngenes by 54 bp at these loci.
Pupilla triplicata of rock outcrops from western Europe to
central Asia has shells that differ from P. sonorana only by their
weak, rounded (rather than sharp) striae. This species uniquely
possesses 295A in ITS2. Most individuals also exhibit 617C in
ITS1. COI þ CytB of this species form a highly supported
clade. While two distinct subpopulations are suggested by 740A
vs 740G in ITS2, there is no correspondence in COI þ CytB or
in any noted conchological features. As such, this grouping
appears to have no taxonomic merit.
Pupilla sterrii of dry calcareous grasslands from central Europe
to the Urals is characterized by its very deep suture and sharp,
coarse, anastomosing striae. It is defined by uniquely possessing
509T in ITS1. The mtDNA of this species is highly variable,
with individuals variously possessing COI þ CytB characteristic
of P. alpicola, P. muscorum, P. triplicata or P. turcmenica. No analysed
individual from Europe was found to possess mtDNA sequence
with the expected topological position as sister to P. turcmenica.
However the Urals specimen did and it may represent the only
individual with both nDNA and mtDNA characteristic of P.
sterrii.
Pupilla turcmenica is a species of xeric calcareous grasslands that
ranges across Asia Minor and central Asia. It is conchologically
distinguished from the similar P. sterrii by its less deep sinus and
more widely spaced striae. It uniquely displays 207G in ITS1,
with its COI þ CytB forming a highly supported clade. Pupilla
altaica has been recently differentiated from this species by its
larger crest, more massive white callus and the presence of an
angular pad on the parietal wall of the aperture (Meng &
Hoffman, 2008). However, this entity did not possess any
unique ITS1 þ ITS2 distinctions from P. turcmenica, being distributed
throughout the same highly supported COI þ CytB
clade encompassing that species. As such it appears to simply represent
the high-calcificiation endpoint within the normal conchological
range of P. turcmenica.
These supported entities could be further associated into nine
groups using ITS1 þ ITS2 data (used to order Table 4). Group
J. C. NEKOLA ET AL.
208
atUniversityofNewMexicoonMay4,2015http://mollus.oxfordjournals.org/Downloadedfrom
23
1 consists of P. blandi, P. hebes, P. hebes pithodes and P. cf. limata
and is characterized by insertion 495A in ITS1. Group 2 is
represented only by P. hudsonianum and is characterized by the
56A insertion, 541G and 583C in ITS2. Group 3 consists of P.
syngenes, P. sonorana and P. cf. khunjerabica and is characterized by
156G in ITS2. Group 4 consists of P. loessica, P. alaskensis and P.
hokkaidoenis and is characterized by 207A in ITS2. Group 5
includes only P. alluvionica and is characterized by 111C and
577A in ITS1 and 333G and 490C in ITS2. Group 6 includes
only P. alpicola and is characterized by 92C and typically 340C
in ITS2. Group 7 consists only of P. muscorum and is characterized
by 375C and 609A in ITS1 and 336C in ITS2. Group 8 is
represented only by P. triplicata and is characterized by 295A in
ITS2. Group 9 consists of P. sterrii and P. turcmenica and is characterized
by 717T in ITS2. Because no variable bases are shared
between groups, however, possible relationships between them
cannot be inferred.
The greater amount of variation within COI þ CytB allows
resolution of deeper relationships. The nine interspecific groups
suggested by ITS1 þ ITS2 are generally validated with high
support by the mtDNA tree topology. The major exception is
Group 1, whose members are spread across two major mtDNA
clades: P. blandi, P. hebes hebes and P. hebes pithodes belong to one
strongly supported clade, while P. cf. limata appears more related
to P. loessica and P. alpicola in the mtDNA tree.
Topological incongruence between mitochondrial and
nuclear DNA phylogenies
Comparison of the mtDNA tree with nDNA tree and base-pair
variation matrix reveals topological incongruence in eleven
specimens, or almost 15% of the total. These are largely
limited to two groups: P. alaskensis and the P. sterrii/turcmenica
clade. The highly supported mtDNA species-level clade containing
both P. alaskensis specimens also harbours individuals
with nDNA characteristic of P. loessica, P. cf. khunjerabica or P.
hudsonianum. Individuals harbouring mtDNA characteristic of
P. turcmenica may possess nDNA characteristic of either P. alluvionica
or P. sterrii. In P. sterrii, specimens possessing nDNA
characteristic of that species may harbour mtDNA characteristic
of P. alpicola, P. triplicata, P. muscorum or P. turcmenica. While
the current analysis is not capable of resolving the cause of
these incongruencies, it does seem likely that mitochondrial
introgression is responsible in the case of P. sterrii as European
individuals variously possess mtDNA from all other known
European species.
Conchological variation: traditional vs other traits
Comparison of conchological features among geneticallyidentified
individuals demonstrates that the size of the apertural
crest, degree of callus deposition, callus colour and number,
shape and placement of apertural lamellae are of little taxonomic
value (Table 5). For instance, in Europe P. alpicola, P. muscorum
and P. triplicata all show variation ranging from zero to
multiple apertural lamellae, and from absent/weak to strong
crest and callus (Fig. 3). This same pattern is repeated in central
Asia with P. turcmenica (Fig. 4) and in North America with
P. blandi and P. hebes (Fig. 5). The reliance of traditional taxonomy
on these traits is thus responsible for oversplitting in some
regions (e.g. P. bigranata and P. pratensis in Europe and P. altaica
in central Asia; Figs 3, 4) and for the abysmal initial sorting of
western North American material (Fig. 5).
However, other conchological traits do accurately reflect
genetic relationships and are capable of accurately sorting individuals
into species-level groups (Table 5; Figs 3–5). The most
important of these are shell sculpture, including not only shape,
strength, density and complexity of shell striae, but also the
lustre of the underlying shell surface. Suture depth and apex
architecture were also found to be valuable for species identification,
as was shell width, which appeared to be relatively independent
of shell height.
SYSTEMATIC DESCRIPTIONS
Pupillidae
Pupilla Leach, in Fleming, 1828
Pupilla alaskensis Nekola & Coles, n. sp.
(Fig. 6A–H, K)
Types: Holotype (Fig. 6A–D, K): ANSP 458632, Happy Valley,
North Slope Borough, Alaska, USA (698200
700
N, 1488430
4800
W).
Paratypes: 10 shells, ANSP 458633, collected with holotype; 100
shells, NMW.Z.2014.013.00013, collected with holotype; 5 shells
ANSP 458634, Sukakpak Mountain, Yukon-Koyukuk Census
Area, Alaska, USA (678350
5500
N, 1498470
400
W); 60 shells
NMW.Z.2014.013.00011, same loc. as preceding; 5 shells ANSP
458635, Livengood East, Yukon-Koyukuk Census Area, Alaska,
USA (658270
5500
N, 1488200
4000
W); 3 shells ANSP 458636, Knik I.,
Matanuska-Susitna Borough, Alaska, USA (618300
3000
N,
149820
300
W).
Zoobank registration: urn:lsid:zoobank.org:act:D98DD52C-BFAE-
4752-A9D4-18F116CE8957
Other material examined: NMW.Z.2005.011.01468, NMW.Z.2014.
013.00002 – 00023 c. 1000 shells from Alaska, USA; 30 lots
(3739 individuals) from Alaska, USA in Nekola collection.
Etymology: Specific name alaskensis refers to region in which
species is known to occur.
Diagnosis: Shell small, cylindrical-ovoid, similar to P. hebes but
differing by its deeper suture and shell sculpture of widelyspaced,
anastomosing radial striae.
GenBank: GQ921663, KM518334, KM518336, KM518412,
KM518414, KM518489, KM518491, KM518567.
Description: Shell 2.6–3.3 mm tall  1.6–1.8 mm wide, opaque to
translucent, yellowish-brown to cinnamon-brown; 6–6.5 whorls;
apical whorls rounded-conical, remainder ovate-cylindrical to cylindrical;
suture typically deep though sometimes of only normal
depth with whorls consequently appearing swollen; shell surface
silky in general appearance, the post-neanic whorls bearing sharp,
irregular, often widely spaced, anastomosing radial striae occasionally
developed into fine lamellae superimposed on a minutely and
irregularly papillate surface (Fig. 6K); aperture 1/4 of shell
height, ranging from slightly wider than tall (Fig. 6A, E, G)
through circular (Fig. 6F) to slightly taller than wide (Fig. 6H), in
profile ascending onto body whorl (Fig. 6B); umbilicus closed by
preceding whorls (Fig. 6C); peristome interrupted by body whorl,
apertural lip flared (Fig. 6B, D), shell slightly contracted behind
(Fig. 6D); crest absent or weakly developed but not thickened or
callused (Fig. 6A–H); apertural lamellae generally absent, though
a vestigial, plate-shaped columellar is occasionally present
(Fig. 6E, G).
Geographical distribution: Currently known from just south of
Arctic Ocean coastal plain in far northern Alaska to Pacific
Coast near Anchorage, Alaska. It seems likely that this species
will be found in adjacent areas of the Yukon and northwestern
British Columbia, Canada. The published records for P. hebes
TAXONOMY OF PUPILLA
209
atUniversityofNewMexicoonMay4,2015http://mollus.oxfordjournals.org/Downloadedfrom
24
Table5.PupillashellfeaturesforthosetaxonomicunitsvalidatedbyDNAsequencedata.
TaxonHeight
(mm)
Width
(mm)
ShellformApexshapeShellsculptureSuturedepthApertureshapeAperturalcrestand
callus
Aperturallamellae
alaskensis2.6–3.31.6–1.8Cylindricalovoidto
cylindrical
TaperedWidelyspaced,sharp,
somewhatanastomosing
coarsestriae;dullshell
Normalto
deep
Slightlywiderthantallto
slightlytallerthan
wide
Absenttoweakcrest;
nocallus
0–1(vestigal,plate-like
columellaronly)
alluvionica3.3–4.32.0–2.4CylindricalovoidTapered/domedLow,rounded,irregular,wide
spacedstriae;shellshiny
Shallow/
normal
TallerthanwideWeaktoverystrong
crestandwhite
callus
0–1(occasionalweak
parietal)
alpicola2.8–4.21.6–2.1Cylindricalovoid;body
whorloftennarrower
thanpenultimate
TaperedFine,rounded,close,
irregularstriae;shellshiny/
satiny
ShallowRoundtoslightlywider
thantall;smallin
proportiontoshell
size
Weaktostrongcrest;
absentto
moderatewhite
callus
0–2(occasionalweak
parietalandpalatal)
mostlypalatal
depression
blandi2.3–3.21.2–1.6CylindricalDomedIrregular,veryweakstriae;
shellshiny
ShallowRoundtoslightlywider
thantall
Weaktostrongcrest,
absenttostrong,
browntowhite
callus
0–3
hebes2.6–3.51.4–1.7CylindricalovoidTaperedfor
upper1/3–1/4of
shellheight
Sharp,numerousthread
striae;shellsilky
Normal/deepRoundtotallerthan
wide
Weaktostrongcrest;
absenttostrong,
browntowhite
callus
0–3
hebespithodes2.9–3.31.6–1.8Cylindricalovoid;body
whorloftennarrower
thanpenultimate
Taperedfor
upper1/3–1/4of
shellheight
Sharp,numerousthread
striae;shellsilky
Normal/deepRoundtowiderthantallWeaktoverystrong
crest;absentto
strongbrownto
whitecallus
1–3
hokkaidoensis3.0–3.11.7–1.8OvoidcylindricalTaperedAnastomosingcoarsestriae;
shellsilkytodull
ShallowRoundtowiderthantallWeakabsentcrest;
nocallus
0
hudsonianum3.3–3.61.7–1.8CylindricalovoidStronglytapered
forupper1/2
Densethreadstriae;shell
silkytodull
Normalto
deep
Tallerthanwide(rarely)
towiderthantall
Weaktostrong;white
callus
0–2(weakparietaland
vestigalcolumellar
occasionallypresent)
cf.khunjerabica2.9–3.51.7–1.8CylindricalovoidTaperedfor
upper1/2of
shellheight
Veryweak,irregularthread
striae;dullshell
Moderately
deep
TallerthanwideThin-lacking0
cf.limata2.7–3.11.6–1.7CylindricalovoidTaperedSharp,numerous,somewhat
anastamosingstriae;dull
shell
ShallowTallerthanwideWeakcrest;nocallus0
loessica2.6–3.61.6–2.0CylindricalovoidTaperedStrongbutrounded
numerous,anastamosing
striae;dullshell
NormalRoundWeaktomoderate
crest;nocallus
0
muscorum2.7–4.01.6–1.8CylindricalovoidTaperedLow,rounded,somewhat
irregularstriae;shellshiny
tosilky
ShallowRoundStrongtoverystrong
crestandwhite
callus
1–2(columellarabsent)
sonorana2.5–3.31.3–1.4Ovoidcylindrical,widest
inupper1/2
DomedWeak,irregularsharpthread
striae;shellsilky
NormalRoundStrongcrest;strong
toverystrongwhite
callus
3(palatalrangingfrompeg
tolongblade)
J. C. NEKOLA ET AL.
210
atUniversityofNewMexicoonMay4,2015http://mollus.oxfordjournals.org/Downloadedfrom
25
from Anchorage, Alaska in Forsyth (2004) refer to material of
the present authors and represent P. alaskensis.
Habitat: This species has been found in upland and lowland
tundra, taiga, fens, herb-rich meadows, coastal grasslands and
riparian forest.
Remarks: Shell reminiscent of P. loessica, but differs from that
species by its widely-spaced, sharp and only slightly anastomosing
striae, deeper suture and weaker (or absent) crest. It differs
from P. hokkaidoensis in its deep suture, darker shell colour and
more regular striae (Table 5).
Pupilla hudsonianum Nekola & Coles, n. sp.
(Fig. 7A–H, K)
Types: holotype (Fig. 7A–D, K): ANSP 458637, Lake Bemidji
State Park, Beltrami County, Minnesota, USA (478310
5800
N,
948490
2800
W). Paratypes: 10 shells, ANSP 458638, collected with
holotype; 50 shells, NMW.Z.2005.011.00835, collected with
holotype; 2 shells, ANSP 458639, highway 40 at Rabbit Hill
Road, east of Benchlands (Calgary), Bighorn #8 Municipal
District, Alberta, Canada (518150
5100
N, 1148430
5700
W); 100
shells, NMW.Z.2014.013.00058, same loc. as preceding; 5 shells,
ANSP 458640, Goose Creek Road, Churchill, Manitoba,
Canada (588420
3000
N, 94870
2200
W); 1 shell, ANSP 458641, La
Grande Pointe, Duplessis District, Quebec, Canada (508120
2100
N,
638230
4800
W); 30 shells, NMW.Z.2014.013.00001, same loc. as
preceding.
Zoobank registration: urn:lsid:zoobank.org:act:B9E21337-42C7-
4BFD-87B9-EFE157D2A5A2.
Other material examined: NMW.Z.2014.01300054-00060, 00066-
00071, 00080; c. 500 shells. Ten lots from Nekola collection (including
one of Pleistocene fossil material; 651 individuals).
Etymology: The specific name hudsonianum refers to Hudson Bay
and to the Hudsonian life zone, which has been used to refer to
the North American taiga, and which defines much of this
species’ range.
Diagnosis: Shell ovoid-cylindrical, similar to P. hebes, but differentiated
by its more ovate shell shape with a surface sculpture
consisting of densely packed radial thread-like striae, giving shell
a silky lustre.
GenBank: GQ921662, KM518353, KM518354, KM518355,
KM518357, KM518358, KM518431, KM518432, KM518433,
KM518435, KM518508, KM518509, KM518510, KM518512,
KM518585, KM518586, KM518587, KM518589.
Description: shell 3.3–3.6 mm tall  1.7–1.8 mm wide; opaque
to translucent yellowish-brown; c. 6.5–7 whorls; apical whorls
rounded-conical in outline, remainder cylindrical; suture moderately
deep; shell surface silky in general appearance, postneanic
whorls bearing irregular, dense, closely-spaced, weakly
anastomosing radial thread-like striae superimposed on a minutely
scaly surface, with minute papillae present between the
scales (Fig. 7K); aperture c. 1
4 of shell height, approximately circular
(Fig. 7F) to wider than tall (Fig. 7A, E, G), rarely taller
than wide (Fig. 7H), in profile ascending onto body whorl
(Fig. 7B); umbilicus closed by preceding whorls (Fig. 7C); peristome
interrupted by body whorl; apertural lip expanded, shell
slightly contracted behind; lip thickened by a weakly to strongly
developed pale callus of shallow depth corresponding to a
weakly to strongly developed crest (Fig. 7A, B, D, E); apertural
sterrii2.6–3.51.5–1.8CylindricalovoidTapered/domedSharp,anastomosingcoarse
striae;shellsilkytodull
VerydeepRoundWeaktostrongcrest;
absenttostrong
whitecallus
1–2(pegshapedpalatal)
syngenes3.0–4.51.6–1.8Biconic,widestinupper
1/3ofshell;8+whorls
Tapered/domedWeakthreadstriae;shelldullShallowTallerthanwideVerystrongcrest;
weaktomoderate
browncallus
3
triplicata2.2–3.1
(rarely
to4)
1.3–1.6Cylindricaltocylindrical
ovoid
TaperedLow,rounded,irregularstriae;
shellshinytosilky
Normalto
deep
RoundModeratetostrong
crest;absentto
strongwhitecallus
0–3(columellaroften
absentorweak)
turcmenica2.7–3.21.5–1.6CylindricalovoidTapered/domedRegularremote,somewhat
anastomisingcoarsestriae;
shelldull
DeepRoundWeaktostrongcrest;
absenttovery
strongwhitecallus
0–3(columellarabsent;
angularpadsometimes
present)
TAXONOMY OF PUPILLA
211
atUniversityofNewMexicoonMay4,2015http://mollus.oxfordjournals.org/Downloadedfrom
26
Figure 3. Pupilla species of primarily European distribution. Names are those supported by DNA sequence analysis. A–D. P. muscorum. A. Cedar
Rapids, Iowa, USA (22). B. Brno, Moravia, Czech Republic (mtG-Pup) C. Syracuse, New York, USA (AP26). D. Pont, Calvados, France (AP25).
E–H. P. alpicola. E. Raksˇ a, Slovakia (H6). F. Belyashi, Altai, Russia (AP12). G. Za´ vod, Slovakia (H5). H. Pozdeˇ chov, Moravia, Czech Republic (H1).
I–M. P. sterrii. I. Verkhne Bikberda, Bashkortostan, Russia (AP15). J. Periferi Dibre, Albania (AP16). K. Klentnice, Moravia, Czech Republic
(AP22). L. Pavlov, Moravia, Czech Republic (H8). M. Valaska´ Dubova´ , Slovakia (AP21). N–Q. P. triplicata. N. Hracholusky, Bohemia, Czech
Republic (H17). O. Ozero Kureevo, Altai, Russia (AP32). P. Pavlov, Moravia, Czech Republic (H9). Q. Cahors, Dordogne, France (AP31).
Figure 4. Pupilla species primarily of Asian/Beringian distribution. Names are those supported by DNA sequence analysis. A, B. P. loessica. A. Belyashi,
Altai, Russia (AP7). B. Belyashi, Altai, Russia (AP9). C. P. alaskensis, Knik Island, Anchorage, Alaska, USA (AP29). D. P. hokkaidoensis, Toyokoro,
Nakagawa, Hokkaido, Japan (VH29). E–H. P. turcmenica. E. Ust’-Muny, Altai, Russia (AP1). F. Kurai, Altai, Russia (AP17). G. Kurai, Altai, Russia
(AP18). H. Kosh-Agach, Altai, Russia (AP3). I. P. alluvionica, Belyashi, Altai, Russia (AP13). J. P. cf. khunjerabica, Chagan-Uzun, Altai, Russia
(AP11). K. P. cf. limata, Kapitonovka, Yakutia, Russia (AP39).
J. C. NEKOLA ET AL.
212
atUniversityofNewMexicoonMay4,2015http://mollus.oxfordjournals.org/Downloadedfrom
27
lamellae generally absent (Fig. 7G, H), but a weak parietal
(Fig. 7E) and vestigial plate-like columellar lamella (Fig. 7A, F)
occasionally present.
Geographical distribution: Currently documented from DNA sequence
data from the foothills of the Rockies in western Alberta,
Canada east through the northern (Churchill, Manitoba,
Canada) and southern (Lake Bemidji, Minnesota, USA) taiga
limits in central North America to the north shore of the Gulf of
St Lawrence in Quebec. Shell lots at the Academy of Natural
Sciences at Drexel University (ANSP 106909, 141759, 141770,
141776, 141783, 150006, 150026), Carnegie Museum (CM
86989, 87010, 62.20823), Museum of Comparative Zoology
(MCZ 048304, 201542), National Museum of Canada (NMC
2892, 69132), Royal Ontario Museum (ROM 21464) and
University of Michigan Museum of Zoology (UMMZ 55951,
109819, 109829, 168485, 180110, 180112) indicate that P. hudsonianum
occurs across the southern shore of Hudson Bay in
Ontario and along the Gulf of St Lawrence shore from the
Gaspe´ and Anticosti Island in Quebec to the west shore of
Newfoundland All reports of P. ‘muscorum’ from Pleistocene
sediments in central North America (Hubricht, 1985) represent
P. hudsonianum (Fig. 6J).
Habitat: This species occurs has been found in mesic taiga, calcareous
fens, dry sandy lakeshores and tundra-like turfs on
shoreline limestone pavements.
Remarks: Pupilla hudsonianum is most readily distinguished from
P. muscorum, with which it has been previously confused, by its
deeper suture, less massive and more yellow apertural callus,
and sharp, fine striae which give the shell a matte luster
(Table 5).
Pupilla hokkaidoensis Nekola, Coles & S. Chiba, n. sp.
(Fig. 8A–K)
Types: holotype (Fig. 8A–D, K): ANSP 458642, Toyokoro,
Nakagawa District, Hokkaido Prefecture, Japan (428360
1800
N,
1438330
2300
E). Paratypes: 10 shells, ANSP 458643, collected
with holotype; 11 shells, NMW.Z.2014.013.00061, collected
with holotype; 5 shells, ANSP 458644, Kushiro Marsh, Kushiro
District, Hokkaido Prefecture, Japan (43820
200
N, 1448230
2400
E);
12 shells, NMW.Z.2014.013.00062, same loc. as preceding; 5
shells, ANSP 458645, Betsukai, Notsuke District, Hokkaido
Prefecture, Japan (438200
5000
N, 1458190
600
E); 5 shells, ANSP
458646, Hama-koshimizu, Shari District, Hokkaido Prefecture,
Japan (438560
100
N, 1448260
3800
E); 25 shells, NMW.Z.2014.
013.00064, same loc. as preceding.
Zoobank registration: urn: lsid:zoobank.org:act:891CB0E8-E4AA-
43BB-9BDB-8F579461E48A.
Other material examined: 6 shells NMW.Z.2005.011.03876, 03878;
4 lots from Nekola collection (176 individuals).
Etymology: The specific name hokkaidoensis refers to the island of
Hokkaido, where all known populations reside.
Diagnosis: Shell small, ovoid-cylindrical, similar to P. hebes
but differentiated by a more ovate shell with shallower sutures
and shell surface sculpture of coarse, anastomosing, radial striae.
GenBank: KM518566, KM518488, KM518411, KM518333.
Description: Shell 3.0–3.1 mm tall  1.7–1.8 mm wide; opaque
to translucent, yellow-brown; c. 6 whorls; apical whorls conical
Figure 5. Pupilla species of North American distribution. Names are those supported by DNA sequence analysis. A–D. P. blandi. A. Ute Creek
Canyon, Colorado, USA (AP37). B. Irvine, Alberta, Canada (AP34). C. Moose Jaw, Saskatchewan, Canada (AP35). D. Bannon Ranch, New
Mexico, USA (image for ET7 was lost, so a similar shell from same population is figured). E–H. P. hebes pithodes. E. Tusas Ridge, New Mexico, USA
(AP38). F. Bullion Canyon, Utah, USA (AP27). G. Bullion Canyon, Utah, USA (AP28). H. Bear Wallow, Arizona, USA (P6). I–M. P. hebes hebes. I.
Loope East, California, USA (P14). J. Bullion Canyon, Utah, USA (P17). K. Ruby Mountains, Nevada, USA (P16). L. Kaibab Plateau, Arizona,
USA (P1). M. East Tintic Range, Utah, USA (P2). N. P. sonorana, Sacramento Mountains, New Mexico, USA (P12). O. P. hudsonianum, Lake Bemidji,
Minnesota, USA (AP33). P, Q. P. syngenes. P. Mogollon, New Mexico, USA (AP30). Q. Kaibab Plateau, Arizona, USA (P11).
TAXONOMY OF PUPILLA
213
atUniversityofNewMexicoonMay4,2015http://mollus.oxfordjournals.org/Downloadedfrom
28
in outline, remainder ovoid-cylindrical giving shell slight barrel
shape; suture shallow; shell surface silky in general appearance,
post-neanic whorls bearing irregular, anastomosing radial striae
most strongly developed on mid whorls, superimposed on a minutely
and irregularly papillate surface (Fig. 8K); aperture c. 1
4
of shell height, ranging in shape from approximately circular
(Fig. 8A, E, H) to taller than wide (Fig. 8F, G), in profile
ascending onto body whorl (Fig. 8B); umbilicus closed by preceding
whorls (Fig. 8C); peristome interrupted by body whorl;
apertural lip flared (Fig. 8B–D), shell slightly contracted
behind; crest absent or weakly developed (Fig. 8D), callus
absent; apertural lamellae absent.
Geographical distribution: Currently known only from the eastern
coast of Hokkaido, Japan.
Habitat: This species was found in beach grasslands, wetland
margins and old fields.
Remarks: Pupilla hokkaidoensis differs from P. loessica in its more
ovate shell, more yellow shell colour, shallower suture and
coarser, more widely spaced and irregular striae. It differs from
P. alaskensis in its more yellow shell colour, shallower suture and
more irregular striae. It differs from P. cf. limata from Yakutia,
Siberia, in its larger size, lighter shell colour and presence of
anastomosting striae.
DISCUSSION
These analyses show that in three widely separated geographic
regions the understanding of species-level taxonomy within the
genus Pupilla has been hampered by the traditional reliance on a
suite of highly plastic shell apertural features that are of little
taxonomic value. As a result, too many species have been
described in Europe and central Asia, and too few species in
North America and eastern Asia, with confusion existing about
actual species ranges and ecological tolerances. However, DNA
sequence analysis also confirms that most previously described
taxa have biological merit, with alternative conchological traits
such as shell sculpture and architecture being able accurately to
distinguish these entities.
Because traditional taxonomic concepts within Pupilla have
been based on unstable shell features, larger patterns regarding
biodiversity, biogeography and ecology must also be reconsidered.
While we cannot deal here with these issues for the entire
genus, the current analysis does allow for reconsideration within
each of our three study regions.
Reassessment of Pupilla biodiversity
In Europe oversplitting has been predominant, with both P.
bigranata and P. pratensis having been differentiated from P.
Figure 6. A–H. Pupilla alaskensis, n. sp. A–D, K. Holotype, ANSP
458632; Happy Valley, North Slope Borough, Alaska, USA. E.
Paratype, ANSP 458633; Happy Valley, North Slope Borough, Alaska,
USA. F. Paratype, ANSP 458635; Livengood East, Yukon-Koyukuk
Census Area, Alaska, USA. G. Paratype, ANSP 458634; Sukakpak
Mountain, Yukon-Koyukuk Census Area, Alaska, USA. H. Paratype,
ANSP 458636; Knik I., Matanuska-Susitna Borough, Alaska, USA. I,
L. P. hebes, JCN 17254; Loope East, Alpine Co., California, USA. J. P.
loessica, Belyashi, Altai Republic, Russia; 498160
800
N, 878590
200
E.
Figure 7. A–H, J, K. Pupilla hudsonianum. A–D, K. Holotype, ANSP
458637; Lake Bemidji State Park, Beltrami County, Minnesota, USA. E.
Paratype, ANSP 458638; Lake Bemidji State Park, Beltrami County,
Minnesota, USA. F. Paratype, ANSP 458641; La Grande Pointe,
Duplessis District, Quebec, Canada. G. Paratype, ANSP 458639; East of
Benchlands, Bighorn #8 Municipal District, Alberta, Canada. H.
Paratype, ANSP 458640; Goose Creek Road, Churchill, Manitoba,
Canada. J. Pleistocene loess fossil, Wenig Road, Cedar Rapids, Linn
Co., Iowa, USA; 42800
800
N, 918400
4000
W; JCN 3650. I, L. P. muscorum;
Syracuse University South Campus, Syracuse, Onondaga Co.,
New York, USA; 43800
2700
N, 76860
3800
W; JCN 13955.
J. C. NEKOLA ET AL.
214
atUniversityofNewMexicoonMay4,2015http://mollus.oxfordjournals.org/Downloadedfrom
29
muscorum and P. alpicola, respectively, based upon unstable shell
apertural characters. It is fortunate that degree of apertural calcification
has never been used to split P. triplicata, as its development
of lamellae can range from three very strong (French
Pyrenees) to absent (basalt talus in northern Bohemia).
In central Asia, oversplitting has also been an issue. While
some recently described taxa are strongly demarcated (e.g. P.
alluvionica and P. cf. khujerabica), others (P. altaica and P. pratensis)
appear to represent high-calcification endpoints in apertural development
within previously described species (P. turcmenica and
P. alpicola, respectively). Perhaps it is not surprising that such
high-calcification shell forms tend to originate from drier, lower
elevation sites, which would have higher calcium availability
due to lower leaching and higher potential evapotranspiration
rates (Lapenis et al., 2008).
In North America, overlumping and ignorance of taxonomically
valid shell traits has led to considerable confusion. First, P.
muscorum is not a native North American species, with all examined
putative native populations representing either P. hebes
(southwestern USA) or P. hudsonianum (central/eastern taiga and
tundra) and with P. muscorum xerobia being a junior synonym of
P. blandi. Pupilla blandi should be distinguished from other North
American species not by apertural features, but rather by its
weak to obsolete striation and shining shell lustre. Pupilla hebes
should be distinguished by its strong thread-like striae and
narrow, columnar shell. Pupilla hebes pithodes is most closely
related to P. hebes, but differs in its wider and more barrel-shaped
shell. Additionally, P. alaskensis has been variously regarded as
P. muscorum or P. hebes in spite of its coarser striation, ovate shell
shape and deeper suture than either of these species.
Reassessment of Pupilla biogeography
In Europe, P. alpicola cannot be considered a central European
endemic with a disjunct set of populations in the Altai (Horsa´ k
et al., 2010). Rather, it extends continuously northwest into
southern Scandinavia and Ireland (as the former P. pratensis)
and east into central Asia. Although demarcation between the
central Asian and European populations is evident in nDNA,
the central Asian subpopulation extends at least as far west as
Bohemia. The amount of mixing of these two populations
during full glacial stages is thus unclear. Pupilla triplicata occurs
as far east as the Altai in central Asia. Pupilla sterrii is the western
sibling of P. turcmenica, with populations extending from central
Europe east to the Urals. Pupilla muscorum is not a Holarctic
species, but is a European endemic with confirmed Pleistocene
fossil occurrences in loess deposits of central Europe.
In the Altai, species status for two putative central Asian
endemics (P. alluvionica and P. cf. khunjerabica) was established.
However, another (P. altaica) was found to be simply a shell
form within P. turcmenica, which ranges from western China and
Tibet to the Iran–Turkmenistan border (Pilsbry, 1921). Pupilla
loessica was shown to be a member of a Beringian group that also
includes P. hokkaidoensis and P. alaskensis.
In North America, the lack of true P. muscorum as a Pleistocene
fossil suggests that it is an exotic species ranging from the western
Great Lakes east to Virginia and north into the Canadian maritime
provinces. The identical COI haplotype of the Brno and
Cedar Rapids P. muscorum specimens suggest that both populations
were sourced from the same pool. This is not surprising
given that extensive immigration from the Czech Republic to
eastern Iowa happened during the mid-1800s. Pupilla hudsonianum,
which has been previously regarded as P. muscorum, extends
west from the north shore of the St Lawrence River in Quebec to
the southern border of Hudson Bay, northwestern Minnesota
and the foothills of the Rockies in Alberta. It also represents the
putative Pleistocene fossil ‘P. muscorum’ reported by Hubricht
(1985) across the central Midwestern USA and Plains. Pupilla
blandi is limited to the Plains (NE New Mexico to NW
Minnesota to southern Saskatchewan and Alberta) and rarely
penetrates west into the Rockies as far as the continental divide.
Pupilla hebes is characteristic of the Great Basin from Arizona and
California to Utah and Idaho, with a well-demarcated subpopulation
from the Colorado Plateau being demonstrated by
mtDNA. While this subpopulation would equate to P. hebes kaibabensis,
its shells and nDNA do not differ in any meaningful
way from typical P. hebes, and we have not chosen to recognize it
here. Pupilla hebes pithodes is found to the south and east of typical
P. hebes, ranging from eastern Arizona and SE Utah to the
eastern foothills of the Rockies in Colorado and New Mexico.
Pupilla alaskensis, formerly confused with P. hebes, is actually a
sibling of the western Beringian P. loessica.
Reassessment of Pupilla ecology
The existence of so much apertural variation within species
across Pupilla begs for an explanation. In particular, how much
of these differences are due to genetic variation and how much
to ecophenotypic response? Little empirical data exist to address
this question. However, we typically noted limited variation in
apertural features within populations. Populations expressing a
poorly developed apertural callus and lamellae tended to be
found in sites with low calcium availability, such as P. blandi
Figure 8. A–H. Pupilla hokkaidoensis. A–D, K. Holotype, ANSP 458642;
Toyokoro, Nakagawa District, Hokkaido Prefecture, Japan. E.
Paratype, ANSP 458643; Toyokoro, Nakagawa District, Hokkaido
Prefecture, Japan. F. Paratype, ANSP 458644; Kushiro Marsh, Kushiro
District, Hokkaido Prefecture, Japan. G. Paratype, ANSP 458646;
Hama-koshimizu, Shari District, Hokkaido Prefecture, Japan. H.
Paratype, ANSP 458645; Betsukai, Notsuke District, Hokkaido
Prefecture, Japan. I. P. hebes, Charleston, Elko Co., Nevada, USA;
418410
1200
N, 1158300
2200
W; JCN 18292. J. P. cf. limata, Kapitonovka,
Yakutia Republic, Russia; 628190
4500
N, 1298550
4200
E.
TAXONOMY OF PUPILLA
215
atUniversityofNewMexicoonMay4,2015http://mollus.oxfordjournals.org/Downloadedfrom
30
(AP37) on acid metamorphic rock in the Colorado Rockies, and
P. triplicata on basalt talus slopes in Bohemia (e.g. H10). In contrast,
the most heavily calcified P. turcmenica in the Altai tend to
be restricted to xeric, low-elevation steppe, often on calcium-rich
metamorphic rock or limestone. Individual age and the season
when maturity is reached may also play important factors.
While these observations suggest that ecophenotypic or developmental
response is responsible for much of the observed variation,
different shell forms have nevertheless been observed in
co-occurring individuals that share identical mtDNA and
nDNA haplotypes (e.g. AP27, AP28), suggesting that multiple
factors may be operating.
ACKNOWLEDGEMENTS
We wish to thank Albuquerque Institute of Math & Science students
Alex Jones, Gabe Martinez, Sarah Martinez and
Khayman Mondragon for assisting with initial PCR and sequencing
of the North American specimens. Japanese field work
would not have been possible without the logistical support of
Satoshi Chiba. Lucie Jurˇ icˇ kova´ (Charles University) kindly provided
historical Pupilla collections by Vojen Lozˇ ek from
Bohemia, Toma´ sˇ Cˇ ejka (Slovak Academy of Sciences) provided
material of Slovakian P. pratensis and Zolta´ n Fehe´ r (Hungarian
Natural History Museum) provided P. sterrii from Albania.
Major funding for field work was provided by the USA National
Science Foundation (EAR-0614963), the Czech Science
Foundation (P504-11-0454) and the Japanese Society for the
Promotion of Science (Kakenhi-24340128).
We initially embarked on this project to discern whether P. triplicata
with vestigal or absent lamellae from the basalt screes of
northern Bohemia warranted erection as a new species. Had our
initial hypothesis been validated, we intended to name this
taxon after Dr Lozˇ ek. While this work ultimately documented
an entirely different story, we would still like to thank Dr Lozˇ ek
for his lifetime of work on Eurasian land snails, including the
initial description of P. loessica.
REFERENCES
ARMBRUSTER, G.F.J. & BERNHARD, D. 2000. Taxonomic
significance of ribosomal ITS-1 sequence markers in self-fertilizing
land snails of Cochlicopa (Stylommatophora, Cochlicopidae).
Zoosystematics & Evolution, 76: 11–18.
EMBERTON, K.C. 1995. When shells do not tell—145-million years of
evolution in North American polygyrid land snails, with a revision
and conservation priorities. Malacologia, 37: 69–109.
FORSYTH, R.G. 2004. Land snails of British Columbia. Royal BC
Museum, Victoria, Canada.
HORSA´ K, M., CHYTRY´ , M. & AXMANOVA´ , I. 2013.
Exceptionally poor land snail fauna of central Yakutia (NE Russia):
climatic and habitat determinants of species richness. Polar Biology,
36: 185–191.
HORSA´ K, M., CHYTRY´ , M., POKRYSZKO, B.M., DANIHELKA,
J., ERMAKOV, N., HA´ JEK, M., HA´ JKOVA´ , P., KINTROVA,
K., KOCˇ I´ , M., KUBESˇ OVA´ , S., LUSTYK, P., OTY´ PKOVA´ , Z.,
PELA´ NKOVA´ , B. & VALACHOVICˇ , M. 2010. Habitats of relict
terrestrial snails in southern Siberia: lessons for the reconstruction of
palaeoenvironments of full-glacial Europe. Journal of Biogeography,
37: 1450–1462.
HORSA´ K, M., HA´ JEK, M., SPITALE, D., HA´ JKOVA´ , P., DI´ TEˇ , D.
& NEKOLA, J. 2012. The age of island-like habitats impacts habitat
specialist species richness. Ecology, 93: 1106–1114.
HUBRICHT, L. 1985. The distributions of the native land mollusks of
the eastern United States. Fieldiana, New Series, 24: 1–191.
HUELSENBECK, J.P. & RONQUIST, F. 2001. MRBAYES: Bayesian
inference of phylogeny. Bioinformatics, 17: 754–755.
KERNEY, M.P. & CAMERON, R.A.D. 1979. Field guide to the land
snails of the British Isles and northwestern Europe. Collins, London.
LAPENIS, A.G., LAWRENCE, G.B., BAILEY, S.W., APARIN, B.F.
& SHIKLOMANOV, A.I. 2008. Climatically driven loss of calcium
in steppe soil as a sink for atmospheric carbon. Global Biogeochemical
Cycles, 22: GB2010, doi:10.1029/2007GB003077.
MENG, S. & HOFFMAN, M.H. 2008. Pupilla altaica n. sp. and Pupilla
alluvionica n. sp., two new species of the Pupillidae (Gastropoda:
Pulmonata: Pupilloidea) from the Russian Altay. Mollusca, 26:
229–234.
MENG, S. & HOFFMAN, M.H. 2009. Pupilla seminskii n. sp.
(Gastropoda: Pulmonata: Pupilloidea) from the Russian Altay.
Mollusca, 27: 83–86.
NEKOLA, J.C. 2005. Latitudinal richness, evenness, and shell size
gradients in eastern North American land snail communities.
Records of the Western Australian Museum, Supplement, 68:
39–51.
NEKOLA, J.C. 2014. North American terrestrial gastropods through
either end of a spyglass. Journal of Molluscan Studies, 80: 238–248.
NEKOLA, J.C. & COLES, B.F. 2010. Pupillid land snails of eastern
North America. American Malacological Bulletin, 28: 29–57.
NEKOLA, J.C., COLES, B.F. & BERGTHORSSON, U. 2009.
Evolutionary pattern and process in the Vertigo gouldii (Mollusca:
Pulmonata, Pupillidae) group of minute North American land
snails. Molecular Phylogenetics and Evolution, 53: 1010–1024.
NEKOLA, J.C. & ROSENBERG, G. 2013. Vertigo marciae (Gastropoda,
Vertiginidae), a new land snail from Jamaica. Nautilus, 127:
107–114.
OUGHTON, J. 1948. A zoogeographical study of the land snails of
Ontario. University of Toronto Studies. Biological Series, 57: 1–126.
PILSBRY, H.A. 1921. Manual of conchology, second series. Vol. 26: Pupillidae
(Vertigininae, Pupillinae). Academy of Natural Sciences, Philadelphia.
PILSBRY, H.A. 1948. Land Mollusca of North America (north of Mexico).
Academy of Natural Sciences of Philadelphia Monograph 3, vol. 2.
POKRYSZKO, B.M., AUFFENBERG, K., HLAVA´ Cˇ , J.Cˇ . &
NAGGS, F. 2009. Pupilloidea of Pakistan (Gastropoda: Pulmonata):
Truncatellininae, Vertigininae, Gastrocoptinae, Pupillinae (in part).
Annales Zoologici (Warszawa), 59: 423–458.
SCHILEYKO, A.A. 1984. Nazemnye molljuski podotryda Pupillina
fauny SSSR (Gastropoda, Pulmonata, Geophila). Fauna SSSR,
Molljuski, 3: 1–399 (In Russian).
TONGKERD, P., LEE, T., PANHA, S., BURCH, J.B. & O´
FOIGHIL, D. 2004. Molecular phylogeny of certain Thai
Gastrocoptinae micro land snails (Stylommatophora: Pupillidae)
inferred from mitochondrial and nuclear ribosomal DNA sequences.
Journal of Molluscan Studies, 70: 139–147.
UIT DE WEERD, D.R., PIEL, W.H. & GITTENBERGER, E. 2004.
Widespread polyphyly among Alopiinae snail genera: when
phylogeny mirrors biogeography more closely than morphology.
Mollecular Phylogenetics and Evolution, 33: 533–548.
VON PROSCHWITZ, T., SCHANDER, C., JUEG, U. &
THORKILDSEN, S. 2009. Morphology, ecology, and
DNA-barcoding distinguish Pupilla pratensis (Clessen, 1871) from
Pupilla muscorum (Linnaeus, 1758) (Pulmonata: Pupillidae). Journal
of Molluscan Studies, 75: 315–322.
WADE, C.M. & MORDAN, P.B. 2000. Evolution within the gastropod
molluscs; using the ribosomal RNA gene-clustre as an indicator of
phylogenetic relationships. Journal of Molluscan Studies, 66: 565–570.
J. C. NEKOLA ET AL.
216
atUniversityofNewMexicoonMay4,2015http://mollus.oxfordjournals.org/Downloadedfrom
31
Vertigo marciae (Gastropoda: Vertiginidae), a new land snail
from Jamaica
Jeffrey C. Nekola
Department of Biology
University of New Mexico
Albuquerque, NM 87131 USA
jnekola@unm.edu
Gary Rosenberg
Department of Malacology
Academy of Natural Sciences of Drexel University
1900 Benjamin Franklin Parkway
Philadelphia, PA 19103 USA
rosenberg@ansp.org
ABSTRACT
Vertigo marciae, a new species of gastropod mollusk (Pupilloidea:
Vertiginidae), is described from Jamaica. This species is known
in the Recent fauna only from John Crow Peak in the Blue
Mountains, but also occurs as a Pleistocene fossil at Red Hills
Road Cave. Vertigo marciae has been confused with Vertigo
gouldii, but differs by its smaller shell size, lack of distinct
shell striation, lack of an angular lamella, and presence of a
flared aperture base. DNA sequence analyses document that
V. marciae possesses unique mtDNA and nDNA sequences
and is most closely allied with Vertigo alabamensis, Vertigo
hebardi, and Vertigo oscariana. This group of species comprises
a highly supported clade whose members are limited
either to the Caribbean or the southeastern USA.
Additional Keywords: Vertigo, biogeography, southeastern USA,
Caribbean, DNA sequence analysis
INTRODUCTION
In their synopsis of the Jamaican land snail fauna,
Rosenberg and Muratov (2006) reported three species
from the genus Vertigo, all of which possess ranges
extending into the eastern half of North America
(Nekola and Coles, 2010): V. gouldii, V. milium, and
V. ovata. However, images of putative Jamaican
‘V. gouldii’ on the Discover Life website (http://pick4
.pick.uga.edu/mp/20q?search=Vertigoþgouldii&guide¼1)
illustrate a shell quite unlike V. gouldii from eastern
North America (Nekola and Coles, 2010) in its lightyellow
shell color, lack of sharp shell striation, and
absence of a basal lamella. Observation of the three
known Jamaican lots for this entity in the Academy of
Natural Sciences of Philadelphia (ANSP) collections
confirmed these differences. Although they looked most
like V. hebardi Vanatta, 1912 of the Florida Keys, these
lots clearly differed from that species by their taller shell,
reduced striation, and absence of an angular lamella.
Thus, rather than representing a population of V. gouldii
isolated by 2000 km from its nearest neighbors (Nekola
and Coles, 2010), these specimens appear to represent
an undescribed new species. Based on shell descriptions,
reports of Vertigo gouldii from the Pleistocene
Red Hills Road Cave deposits of Jamaica (Paul and
Donovan, 2005) also appeared to represent this new species.
This conclusion was subsequently confirmed via
observation of digital images.
We therefore investigated this putative new species
via analyses of both shell morphology and DNA
sequence data, and are now in the unusual situation of
describing and reporting a new species simultaneously
from both living and fossil material, while also being
able to report on its phylogenetic relationships.
MATERIALS AND METHODS
Field Collection: The three known Recent lots of the
putative new Jamaican species (ANSP 402244, 403039,
and 403040) were collected on John Crow Peak in the
Blue Mountains of eastern Jamaica on May 22, 1999,
during the Jamaican Biotic Survey. Collecting methods
were detailed in Rosenberg and Muratov (2006), and
included drying of soil litter samples over a Berlese
Funnel. This encouraged snails to enter aestivation, allowing
them to mummify upon death. For this reason, it was
possible to successfully extract, amplify, and sequence
selected amplicons from both their mitochondrial and
nuclear genomes (Nekola et al., 2009).
Shell Measurements: Measurements were determined
in 0.1 mm increments for adult shells with
aperture facing up, using a dissecting microscope with
a calibrated ocular micrometer. Height and width were
measured as the dimensions of a bounding box with
long axis parallel to the shell axis and sides tangent
to the tip of the protoconch, the base of the lip, the
right-most margin of the aperture and the left-most
margin of the body whorl.
Imaging: Shells were imaged at 20Â magnification
using a digital camera attached to a stereomicroscope.
THE NAUTILUS 127(3):107–114, 2013 Page 107
32
Approximately 14 separate 1388Â1040 pixel images
were made of each specimen with the image focal
lengths positioned at 75 mm increments from the front
to back of the shell. CombineZ5 freeware (http://www
.hadleyweb.pwp.blueyonder.co.uk/CZ5/combinez5.htm)
was used to assemble a final image from the focused
parts of each separate image. Assembled images were
imported into Adobe Photoshop, where brightness and
contrast were optimized and the background made uniformly
black. These images were then compiled into a
single plate.
DNA Sequence Analysis: Mitochondrial cytochrome
b (CytB), 16S ribosomal RNA (16S) and the internal transcribed
spacers 1 (ITS–1) and 2 (ITS-2) of the nuclear
ribosomal RNA gene were investigated to test the distinctiveness
of the putative new Jamacian species and to
resolve its nearest evolutionary neighbors. The mitochondrial
cytochrome oxidase subunit 1 (CO1) was not
analyzed as we were unable to amplify this gene in the
Jamaican species.
Sixteen specimens were chosen for analysis (Table 1).
This set includes not only two specimens of the
Jamaican species but also Vertigo alabamensis, V.
gouldii, V. hebardi, and V. oscariana. We attempted to
maximally spread these individuals across the known
geographic range of each species (see Nekola and
Coles, 2010). Because V. conecuhensis is likely a simple
shell form of V. alabamensis (Nekola and Coles, 2010),
only a single individual of this taxon was included.
Topotype or near-topotype material was selected for
V. conecuhensis (Pond Creek, Covington Co., Alabama,
about 50 km from the type locality in Evergreen,
Alabama) and V. hebardi (Long Key, Florida which is
the type locality). Previously analyzed 16S sequence
data from three Vertigo gouldii specimens analyzed
by Nekola et al. (2009) were retrieved from GeneBank
(Table 1). Vertigo pusilla, the type species of the genus,
was also included, as were Gastrocopta cristata and
Pupilla muscorum for outgroup comparisons.
Genomic DNA was extracted from live, ethanolpreserved,
or mummified material using the OmegaBioTek
Mollusk DNA Extraction Kit. Because shell destruction
was required, all shells were imaged prior to
extraction using the methods detailed above. PCR
amplification and sequencing of CytB, 16S, ITS-1, and
ITS-2 were accomplished using standard methods
(Nekola et al., 2009). Resultant traces were examined,
primer ends removed, and aligned by eye. Because of
their lower evolutionary rates, the nDNA ITS-1 and
ITS-2 regions were concatenated to provide a roughly
similar number of nucleotide differences to either of
the mtDNA amplicons. Mega 5.0 (Tamura et al., 2011)
was used to calculate the average number of nucleotide
differences in the CytB, 16S, and concatenated ITS-1 þ
ITS-2 regions between all putative species-level taxa.
Substitutions included both transitions and transversions
with pairwise gap deletion.
Table 1. Specimen information for material used in DNA sequence analysis.
Taxon
Location Latitude/Longitude
GENBANK Accession number
CytB 16S ITS-1 ITS-2
Vertigo marciae new species
John Crow Peak, St. Andrew Parish, Jamaica 18.1132 N., 76.6685 W. KF214513 KF214503
John Crow Peak, St. Andrew Parish, Jamaica 18.1132 N., 76.6685 W. KF214514 KF214502 KF214489 KF214477
Vertigo alabamensis Clapp, 1915
Johnson Mill Bay, Bladen Co., North Carolina, USA 34.7125 N., 78.5261 W. KF214515 KF214501 KF214490 KF214478
Wolf Trap Bay, Leon Co., Florida, USA 30.3680 N., 84.5700 W. KF214516 KF214500 KF214491 KF214479
Vertigo conecuhensis Clapp, 1915
Pond Creek, Covington Co., Alabama 31.1036 N., 86.5343 W. KF214517 KF214499 KF214492 KF214480
Vertigo gouldii (A. Binney, 1843)
Brush Creek Canyon, Fayette Co., Iowa, USA 42.7796 N., 91.6890 W. KF214508 GQ921506 KF214484 KF214472
Panther Creek, Searcy Co., Arkansas, USA 36.0858 N., 92.5649 W. KF214509 GQ921507 KF214485 KF214473
11-Point River, Oregon Co., Missouri, USA 36.7931 N., 91.3334 W. KF214510 KF214506 KF214486 KF214474
Vertigo hebardi Vanatta, 1912
Long Key, Monroe County, Florida, USA 24.8146 N., 80.8211 W. KF214511 KF214505 KF214487 KF214475
Elliott Key, Miami-Dade County, Florida, USA 25.4553 N., 80.1925 W. KF214512 KF214504 KF214488 KF214476
Vertigo oscariana (Sterki, 1890)
Blanchard Springs, Stone Co., Arkansas, USA 35.9582 N., 92.1778 W. KF214518 KF214498 KF214493 KF214481
Wadboo Creek, Berkeley Co., South Carolina, USA 33.1971 N., 79.9461 W. KF214519 KF214497 KF214494 KF214482
Vertigo pusilla Mu¨ller, 1774
Podyji National Park, Moravia, Czech Republic 48.8586 N., 15.8960 E. KF214520 KF214496 KF214495 KF214483
Gastrocopta cristata (Pilsbry and Vanatta, 1900)
Albuquerque, Bernalillo Co., New Mexico, USA 35.0727 N., 106.6160 W. KF214522 JN941032
Pupilla muscorum (Linnaeus, 1758)
Masaryk University, Brno, Moravia, Czech Republic 49.2509 N., 16.5738 E. KF214521 KF214507
Page 108 THE NAUTILUS, Vol. 127, No. 3
33
Mega 5.0 was used to conduct nearest–neighbor joining
(NNJ), maximum parsimony (MP), and maximum
likelihood (ML) trees for the three focal regions. NNJ
was based on Maximum Composite Distance including
transitions and transversions with pairwise gap deletion.
MP used the close neighbor interchange search
option with the random addition of 10 replicate trees.
ML used all sites and was based on the Tamura–Nei
substitution model, a five–category Gamma Distribution
for substitution rates, and the Nearest Neighbor
Interchange ML heuristic method. In all cases support
values were estimated from 1000 bootstrap replicates.
Additionally, Bayesian trees were generated using
MrBayes 3.1 (Huelsenbeck and Ronquist, 2001) using
a GTR substitution model assuming gamma-shaped rate
variation over 1,000,000 generations with a sampling frequency
of once each 1000 generations.
Nomenclature: Taxonomic names and concepts follow
Nekola and Coles (2010). Apertural lamellae and fold
nomenclature follows that of Pilsbry (1948: 869, fig. 469),
i.e., parietal “teeth” are referred to as “folds” and all
other “teeth” are termed “lamellae”, whatever their form.
Also, we follow Pilsbry (1948) by referring to the raised
riblets on the surface of Vertigo shells as “striae.”
SYSTEMATICS
Class Gastropoda
Subclass Pulmonata
Order Stylommatophora
Family Vertiginidae
Genus Vertigo Mu¨ller, 1773
Vertigo marciae new species
(Figures 1–6; Tables 1, 2)
GenBank Accessions KF214477, KF214489, KF214502,
KF214503, KF214513, KF214514
Diagnosis: Minute shell reminiscent of Vertigo hebardi
but taller, less striate, and lacking angular lamella; shell
surface smooth and glossy with indistinct wrinkles; aperture
flared toward bottom; four lamellae/folds present,
including parietal, columellar, and two palatals.
Description: Shell 1.4–1.6 mm tall  0.8–1.0 mm wide
(holotype 1.5 Â 0.8 mm), columnar-ovoid to ovoid,
approximately 4–4.5 whorls, with moderately shallow
suture and domed apex. Translucent, pale yellow-brown
color. Body whorl approximately 60% of total height
(Figures 1–3, 5, 6). Protoconch and neanic whorls
smooth (Figure 4), with subsequent whorls having
irregular, infrequent, and weak wrinkles. Immediately
behind aperture sculpture takes the form of irregular
low striae (Figure 2). Aperture flared on bottom, making
it taller than wide and approximately 1/3 of shell
height. Lip unthickened and slightly reflexed, sinulus
moderate-weak, sometimes expressed as simple flattening
of palatal wall. Basally aperture abruptly inflates to
form rounded swelling but not crest (Figure 4). Umbilicus
closed (Figure 3). Aperture with four lamellae/folds:
parietal lamella strong, slightly sinuous (Figures 1, 2, 5,
6); columellar lamella downward-sloping, peg-shaped;
two palatal folds with lower being approximately twice as
long as upper and extending approximately 0.2 whorls
into body whorl, lower slightly more immersed than
upper, both highest at mid length (Figures 1, 2, 5, 6).
Apertural end of lower palatal fold coincides with abrupt
inflation of basal aperture. Externally shell only slightly
impressed over palatal folds (Figure 4).
Type Material: Holotype (Figures 1–4), ANSP 450580,
from type locality, May 22, 1999; paratypes (Figures 5–6),
ANSP 402244, 8 shells and Institute of Jamaica, 2 shells,
from type locality; ANSP 403039, Jamaica, St. Andrew
Parish, John Crow Peak, Blue Mountains, elfin forest
with bamboo near summit, 18
050
4500
N, 76
400
08.400
W,
altitude 1755 m, May 22, 1999; sta. JBS4a, 16 shells;
ANSP 403040, Jamaica, St. Andrew Parish, Vinegar Hill
Trail near head of Clyde River, 18
050
N, 76
390
1800
W,
altitude 1520 m, May 22, 1999, sta. JBS 5, 3 shells.
Type Locality: Jamaica, St. Andrew Parish, John
Crow Peak, Blue Mountains; litter sample collected at
base of limestone cap, 18
50
5000
N, 76
400
500
W, altitude
1550 m, May 22, 1999, sta. JBS4c.
Other Material Examined: Digital images of two specimens
from Red Hills Road Cave, Jamaica (Paul and
Donovan, 2005), Paul collection.
Etymology: The specific name marciae refers to
Dr. Marcia Mundle, then of the Jamaica Conservation
and Development Trust. Dr. Mundle arranged for the
vehicle used and park ranger guide that accompanied
the expedition to the type location, and was present
when the species was first collected.
RESULTS AND DISCUSSION
Variation: Vertigo marciae is rather constant in general
appearance in terms of its shape, color, sculpture,
and development of the apertural lamellae. While some
variation in size (especially height) was noted, this was
minor—only 16% difference between the largest and
smallest shells was observed.
Comparison with Other Vertigo Species: Vertigo
marciae differs from all other Vertigo species by its
small (<1.7 mm in height) yellow shell with indistinct
striae/wrinkles, flared aperture base, and lack of angular
and basal lamellae. It is closest in appearance to
V. hebardi of the Florida Keys (Figure 7), with which
it shares a small yellow shell, the lack of a basal lamella,
and preference for accumulations of tropical forest leaf
litter. However, it differs from this species in its taller
shell with less distinct striae and absence of an angular
lamella. It is also reminiscent of Vertigo marki Gulick,
J.C. Nekola and G. Rosenberg, 2013 Page 109
34
Figures 1–12. Stereomicroscope images of Vertigo marciae and related taxa. 1–4. Vertigo marciae, holotype, Jamaica, St. Andrew
Parish, John Crow Peak, Blue Mountains, 18
50
4500
N, 76
400
800
W to 18
50
5000
N, 76
400
500
W, ANSP 450580. 1. Apertural view.
2. Profile. 3. Umbilical view showing parietal and upper palatal lamellae. 4. Apical view showing protochonch and the basal
apertural dilation. 5. Vertigo marciae, ANSP 402244, Paratype, second specimen from the type locality, exhibiting a shorter
shell. 6. Vertigo marciae, ANSP 402244, Paratype, third specimen from the type locality, exhibiting a slightly more worn shell.
7. Vertigo hebardi, Long Key, Monroe County, Florida, 24
480
5200
N, 80
490
1400
W, JCN 17479. 8. Vertigo gouldii (small southern
form), Tellico Gorge, Monroe Co., Tennessee, 35
190
4900
N, 84
100
5900
W, BFC 1332. 9. Vertigo alabamensis, Lanier Quarry, Pender
Co., North Carolina, 34
370
4900
N, 77
400
2700
W, JCN 10781. 10. Vertigo “conecuhensis,” Pond Creek seep, Covington Co., Alabama,
31
60
1200
N, 86
320
300
W, JCN 12364. 11. Vertigo oscariana, Wadboo Creek, Berkeley Co., South Carolina, 33
110
5000
N, 79
560
4600
W,
JCN 10908. 12. Vertigo gouldii (normal form), Deer Creek, Fillmore Co., Minnesota, 43
430
5600
N, 92
200
3900
W, JCN 14646.
Page 110 THE NAUTILUS, Vol. 127, No. 3
35
1904 from Bermuda, with which it shares a smooth,
yellowish shell that lack both basal and angular lamellae
(Pilsbry, 1919). However, V. marciae differs from this
species by its smaller shell height and lack of a callus
on the palatal wall of the aperture. Vertigo oscariana
from the southeastern USA (Figure 11) also has a
small, smooth yellow shell without a basal lamella,
but this species also lacks an upper palatal fold,
has a callus on the outer apertural margin, and a
shell that is widest in the middle, tapering both to
base and apex. Vertigo alabamensis of the southeastern
North American coastal plain (Figures 9, 10) also has a
yellowish shell lacking striae, but this species is larger
(generally >¼1.8 mm tall), possessing both angular and
basal lamellae, a strong crest on the apertural margin,
and a strong sinulus on the palatal wall of the aperture.
Rosenberg and Muratov (2006) identified Vertigo
marciae as Vertigo gouldii (Figures 8, 12) because of
its general shell shape and placement of the parietal
lamella and palatal folds, and also because V. gouldii
was reported from Jamaica by Pilsbry and Cooke
(1919: 99). However, V. gouldii has a larger shell of
brown color, has a duller shell luster from the presence
of abundant strong but irregular striae, and always
possesses a basal lamella. Specimens documenting the
Pilsbry and Cooke record have not been traced, and are
apparently based on personal communication between
Victor Sterki and Pilsbry. Bequaert and Miller (1973: 95)
rejected Antillean records of V. gouldii.
While Vertigo milium—which also occurs in Jamacian
tropical forest leaf litter—possesses a shell <1.7 mm tall;
this species is easily distinguished from V. marciae in
its dark red-brown color, presence of both angular and
basal lamellae, and its long, curved lower palatal fold
which deeply enters the shell.
Geographic Distribution and Ecology: Vertigo
marciae is currently known in the Recent fauna only
from the crest of John Crow Peak and its immediate
vicinity in the Blue Mountains of eastern Jamaica at elevations
of 1520–1755 m. John Crow Peak is capped by
an isolated limestone outlier known to have distinctive
plant communities, including endemic species (Grossman
et al., 1993). (Note that John Crow Peak is not the same
location as the John Crow Mountains, the easternmost
range of Jamaica.) Vertigo marciae is found mainly on or
adjacent to limestone boulders and outcrops in tropical
as well as scrub forest with bamboo, but has also been
found at one site (JBS 5) that lacks exposed limestone.
As a Pleistocene fossil, it is known only from Red Hills
Road Cave, which is about 21 km west of Rosenberg
and Muratov’s sites on John Crow Peak and 1000 m
lower in elevation (520 m, Paul and Donovan, 2005).
Vertigo marciae joins the two Radiodiscus species and
the Punctum reported by Paul and Donovan (2005) on
the basis of personal communication from Rosenberg
as species known in Jamaica only from John Crow Peak
and from the Red Hills Road Cave. Paul and Donovan
interpreted the faunal changes from the Red Hills Road
Cave fauna to the recent fauna around the cave as
suggesting drying of the climate since the Pleistocene.
As Vertigo marciae was found nowhere else in Jamaica
among the hundreds of sites sampled by Rosenberg
and Muratov, its current occurrence might represent a
relict distribution.
Although it must currently be considered a Jamaican
endemic, the general lack of local endemism in the
genus Vertigo (Nekola, 2009) suggests the possibility
that it may occur elsewhere in the adjacent Caribbean,
especially on carbonate substrates in mid- to high-elevation
montane forest.
Table 2. Average number of base-pair differences between all pairwise combinations of Vertigo marciae and related species.
A. CytB amplicon (355 base pairs)
V. oscariana 53.1
V. alabamensis 65.6 54.3
V. hebardi 62.3 52.0 8.3
V. marciae 63.8 46.5 38.7 36.0
V. gouldii V. oscariana V. alabamensis V. hebardi
B. 16S amplicon (443–447 base pairs)
V. oscariana 40.9
V. alabamensis 37.3 44.5
V. hebardi 35.8 44.5 4.0
V. marciae 42.5 42.0 18.0 18.0
V. gouldii V. oscariana V. alabamensis V. hebardi
C. Concatenated ITS-1 and ITS-2 amplicons (1274–1284 base pairs)
V. oscariana 34.8
V. alabamensis 32.3 20.5
V. hebardi 34.3 22.5 5.0
V. marciae 37.3 23.5 9.0 12.0
V. gouldii V. oscariana V. alabamensis V. hebardi
J.C. Nekola and G. Rosenberg, 2013 Page 111
36
Genetic Distinctness and Phylogenetic Relationships:
DNA sequence analysis clearly demonstrates that
V. marciae is distinct at the species level. It possesses
an average of 36 and 39 base pair differences
with V. hebardi and V. alabamensis, respectively,
in the 355 bp CytB amplicon (Table 2). This equates
to a 10–11% variation across the entire amplicon. In
addition, it possessed an average of 46.5 base pair
differences in CytB (13%) with V. oscariana and
63.8 differences (18%) with V. gouldii. In the more
slowly evolving 443 to 447 bp 16S region V. marciae
differed by 18 bases (4.0%) from V. hebardi and
V. alabamensis, 42 bases (9.4%) from V. oscariana,
and 42.5 bases (9.6%) from V. gouldii. In the
concatenated 1274–1284 bp ITS-1 þ ITS-2 nDNA
amplicon, V. marciae possessed an average of 9 basepair
(0.7%) differences with V. alabamensis, 12 (0.9%)
with V. hebardi, 23.5 (1.8%) with V. oscariana, and
37.3 (2.9%) with V. gouldii. These levels of difference
were considerably larger than those exhibited
between the conchologically distinct V. alabamensis
and V. hebardi (Figures 7, 9, 10), which amounted
to an average difference of 8.3 bases (2.3%) in CytB,
4.0 (0.9%) in 16S, and 5.0 (0.4%) in concatenated
ITS-1 and ITS-2.
The topologies across phylogenetic reconstructions
were largely compatible, with the same highly supported
nodes being identified in all cases (Figures 13–15).
These demonstrate that V. marciae is clearly a member
of the genus Vertigo, being within the same highly
supported clade in CytB and 16S that includes
V. pusilla, the type species of the genus. While it was
not possible to root the concatenated ITS-1 þ ITS-2 tree
because of profound differences with both Gastrocopta
cristata and Pupilla muscorum sequences, V. marciae
was easily aligned with all Vertigo sequences.
The phylogenetic reconstructions also demonstrate
that V. marciae is a member of a highly supported
clade that includes both V. alabamensis and V. hebardi.
The existence of a single ancestor to all of these species
implies long distance dispersal from the southeastern
United States to Jamaica, perhaps with migrating birds
as a vector. This is not an unreasonable scenario, given
that much longer feats of long distance dispersal via
migrating birds have been documented in the eastern
Atlantic (Gittenberger et al., 2006).
Figure 13. The phylogenetic relationships of Vertigo marciae as reconstructed by maximum-likelihood analysis from the CytB
amplicon. Nodes with strong to moderate support across all four phylogenetic reconstruction methods have been labeled to the
left of that node by four support values: The upper left (normal font) is for nearest neighbor joining. The upper right (bold italic
font) is for maximum parsimony. The lower left (bold font) is for Bayesian. The lower right (italic font) is for maximum likelihood.
Page 112 THE NAUTILUS, Vol. 127, No. 3
37
Figure 14. The phylogenetic relationships of Vertigo marciae as reconstructed by maximum-likelihood analysis from the 16S
amplicon. Nodes with strong to moderate support across all four phylogenetic reconstruction methods have been labeled to the
left of that node by four support values: The upper left (normal font) is for nearest neighbor joining. The upper right (bold italic
font) is for maximum parsimony. The lower left (bold font) is for Bayesian. The lower right (italic font) is for maximum likelihood.
Figure 15. The phylogenetic relationships of Vertigo marciae as reconstructed by maximum-likelihood analysis from the concatenation
of the ITS-1 and ITS-2 amplicons. Nodes with strong to moderate support across all four phylogenetic reconstruction
methods have been labeled to the left of that node by four support values: The upper left (normal font) is for nearest neighbor
joining. The upper right (bold italic font) is for maximum parsimony. The lower left (bold font) is for Bayesian. The lower right
(italic font) is for maximum likelihood.
J.C. Nekola and G. Rosenberg, 2013 Page 113
38
While resolution of deeper nodes within the genus
Vertigo was not possible in the CytB and 16S mtDNA
amplicons (Figures 13, 14), presumably due to basepair
saturation, the concatenated ITS-1þ ITS-2 data
(Figure 15) demonstrated a very highly supported node
connecting the V. marciae / V. hebardi / V. alabamensis
clade to V. oscariana. This radiation includes some of
the most distinct members of the genus (Pilsbry, 1948).
Additional field work across the Caribbean—especially
in Cuba, Hispaniola, Puerto Rico, and Bermuda—will
be required to determine the actual number of species
contained within this group.
ACKNOWLEDGMENTS
John Slapcinsky of the Florida Museum of Natural History
provided ethanol-preserved tissue samples of Vertigo
hebardi, while Michal Horsak of Masaryk University
in Brno, the Czech Republic, provided live samples
for Vertigo pusilla and Pupilla muscorum. CytB and 16S
sequences for V. pusilla, P. muscorum, and Gastrocopta
cristata were retrieved from full mitochondrial genomic
sequences provided by Jason Marquardt and
Ulfar Bergthorsson of the University of New Mexico.
Christopher Paul, University of Bristol, provided digital
images of the Red Hills Road Cave fossils. Field work
which established the persistence of Vertigo hebardi in
the Florida Keys was supported by The Bailey-Matthews
Shell Museum through an R.T. Abbott Visiting Curatorship
to JCN. Additional funding for PCR and sequencing
analysis was provided by Michal Horsak. The field
work which resulted in the Recent samples of the new
species being collected was supported by NSF Grant
DEB-9870233 to Gary Rosenberg.
LITERATURE CITED
Bequaert, J.C. and W.B. Miller. 1973. The Mollusks of the
Arid Southwest. University of Arizona, Tucson, xviþ271 pp.
Gittenberger, E., D.S.J. Groenenberg, B. Kokshoorn and
R.C. Preece. 2006. Molecular trails from hitch-hiking snails.
Nature 439: 409.
Grossman, D.H., S. Iremonger, and D.M. Muchoney.
1993. Jamaica: map of natural communities and
modified vegetation types. Jamaica: a rapid ecological
assessment. Phase 1: An island-wide characterization
of mapping of natural communities and modified
vegetation types. The Nature Conservancy, Washington,
D.C., USA.
Huelsenbeck, J.P. and F. Ronquist. 2001. MRBAYES:
Bayesian inference of phylogeny. Bioinformatics 17:
754–755.
Nekola, J.C. and B.F. Coles. 2010. Pupillid land snails of
eastern North America. American Malacological Bulletin
28: 29–57.
Nekola, J.C. 2009. Big ranges from small packages: North
American vertiginids more widespread than thought.
The Tentacle 17: 26–27.
Nekola, J.C., B.F. Coles, and U. Bergthorsson. 2009. Evolutionary
pattern and process in the Vertigo gouldii
(Mollusca: Pulmonata, Pupillidae) group of minute North
American land snails. Molecular Phylogenetics and Evolution
53: 1010–1024.
Paul, C.R.C. and S.K. Donovan. 2005. Quaternary and
Recent land snails (Mollusca: Gastropoda) from Red
Hills Road Cave, Jamaica Bulletin of the Mizunami
Fossil Museum 32: 109–144.
Pilsbry, H.A. 1919. American Species of Vertigo. Manual
of Conchology. Second Series (Pulmonata) 25: 74–150,
pls. 6–13.
Pilsbry, H.A. 1948. Land Mollusca of North America (North of
Mexico). Academy of Natural Sciences of Philadelphia
Monographs 3, vol. 2(2): i–xlvii, 521–1113.
Rosenberg, G. and I.V. Muratov. 2006. Status report
on the terrestrial mollusca of Jamaica. Proceedings of
the Academy of Natural Sciences of Philadelphia 155:
117–161.
Tamura, K., D. Peterson, N. Peterson, G. Stecher, M. Nei,
and S. Kumar. 2011. MEGA5: molecular evolutionary
genetics analysis using maximum likelihood, evolutionary
distance, and maximum parsimony methods. Molecular
Biology and Evolution 28: 2731–2739.
Page 114 THE NAUTILUS, Vol. 127, No. 3
39
40
41
42
43
44
45
46
47
48
49
50
51
Section II: Population and Community Ecology
A vital component of understanding ecological systems is determining the factors controlling
species abundance, range and community composition at local scales. I have conducted such
work on vascular plants, lepidoptera, and terrestrial gastropods. The focus for many of these
studies is determination of the local environmental factors that most strongly regulate occurrence
and composition patterns, especially for regionally rare taxa at or near their geographic range
limits. Often these studies have employed multivariate statistical techniques such as
multidimensional scaling ordination and model-based cluster analysis. I have also been involved
in determining the age structure of cliff-face tree populations (including millennium-aged red
and white cedars) along the Niagaran Escarpment in the Great Lakes region. These types of
studies not only help focus future biological surveys, but also help determine optimum
conservation strategies. They also allow for large-scale comparisons of the factors influencing
range limits by allowing for comparison of distribution and abundance drivers across a species
distribution.
Using my community biodiversity database
which currently represents 1823 sites, ~600
species (½ of the continental fauna) and ¾
million individuals ranging from Prudhoe Bay
to San Diego east to Labrador and Key West (all
personally sampled using uniform protocols) I
have begun to investigate continental αdiversity.
Two unexpected patterns emerge:
First, site richness does not monotonically
increase towards the tropics, with a local
maximum generally occurring between 35°-45°
N. Thus, central Appalachian and Ozark sites
tend to support more species than those to their
immediate south. Second, a strong linear and
almost order of magnitude increase is noted
from west to east between 30°-40° N, with a
similar though weaker (roughly 60%) increase
from 40°-50° N. Because regional species pools
are roughly similar in size, it is likely that these
differences are rooted in contemporaneous
ecological process rather than environmental
and/or evolutionary history. The upper limit of
α-diversity in fact does appear to be strongly
influenced by current climate, in particular Aridity Index (AI), peaking at values around 1, which
occurs when annual precipitation = annual potential evaportranspiration. This result can be
related back to land snail physiology, which demands both water to make mucus required for
movement and Ca for shell construction. At low AI, land snails will be stressed by droughtinduced
inactivity, while at high AI they will be stressed by low [Ca] caused by leaching via
excess precipitation. Given that land snail physiology is uniform across the biosphere, similar
relationships must exist elsewhere in the globe. Initial analysis of European and New Zealand
Alpha-diversity (community species richness) vs.
Aridity Index (yearly precipitation/yearly potential
evapotranspiration) for North American land snails.
The upper limit was determined using quantile
regression at tau=0.99. For the domain AI<0.85 the
best fit two parameter model is a modified logit
function; for AI>0.85 it is a power-law function.
52
land snail communities suggests that this pattern does represent a general global expectation.
Additional work is also underway to investigate this relationship in east and southeast Asia,
Australia, Africa, and the Neotropics. This work is being supported in part by grants from the
National Park Service and Texas Parks and Wildlife.
Representative Publications
[number of citations as of October 27, 2017]
Růžička, V., M. Zacharda, L. Němcová, P. Šmilauer &. J.C. Nekola. 2012. Periglacial
microclimate in low altitude scree slopes supports relict biodiversity. Journal of Natural
History. 46:2145-2157. [23]
McClain, C.R., J.C. Nekola, L. Kuhnz, P.N. Slattery & J.P. Barry. 2011. Local-Scale Faunal
Turnover on the Deep Pacific Seafloor. Marine Ecology Progress Series. 422:193-200.
[15]
Nekola, J.C. 2010. Acidophilic terrestrial gastropod communities of North America. Journal
of Molluscan Studies. 76:144-156. [29]
Stanisic, J, R.A.D. Cameron, B.M. Pokryszko & J.C. Nekola. 2007. Forest snail faunas from
S.E. Queensland and N.E. New South Wales (Australia): Patterns of local and regional
richness and differentiation. Malacologia. 49:445-462. [26]
Nekola, J.C. 2004. Vascular plant composition gradients within and between Iowa fens.
Journal of Vegetation Science. 15:771-780. [57]
Nekola, J.C. 2003. Large-scale terrestrial gastropod community composition patterns in the
Great Lakes region of North America. Diversity and Distribution. 9:55-71. [52]
Nekola, J.C. 2001. Distribution and ecology of Vertigo cristata (Sterki, 1919) in the western
Great Lakes region. American Malacological Bulletin. 16:47-52. [8]
Nekola, J.C. & P.A. Massart. 2001. Distribution and ecology of Vertigo nylanderi Sterki, 1909
in the western Great Lakes region. American Malacological Bulletin. 16:53-60. [5]
Larson, D.W., U. Mattes-Sears, J.A. Gerrath, N.W.K. Larson, J.M. Gerrath, J.C. Nekola, G.L.
Walker, S. Porembski & A. Charlton. 2000. Evidence for the widespread occurrence of
ancient forests on cliffs. Journal of Biogeography. 27:319-331. [42]
Nekola, J.C. 1999. Terrestrial gastropod richness of carbonate cliff and associated habitats in
the Great Lakes region of North America. Malacologia. 41:231-252. [36]
Nekola, J.C. & T.A. Smith. 1999. Terrestrial gastropod richness patterns in Wisconsin
carbonate cliff communities. Malacologia. 41:253-269. [55]
Larson, D.W., U. Mattes-Sears, J.A. Gerrath, J.M. Gerrath, J.C. Nekola, G.L. Walker, S.
Porembski, A. Charlton & N.W.K. Larson. 1999. Ancient stunted forests on cliffs.
Nature. 398:382-383. [40]
53
ACIDOPHILIC TERRESTRIAL GASTROPOD COMMUNITIES OF
NORTH AMERICA
JEFFREY C. NEKOLA
Biology Department, University of New Mexico, Albuquerque, NM 87131, USA
Correspondence: J.C. Nekola, e-mail: jnekola@unm.edu
(Received 10 September 2009; accepted 19 October 2009)
ABSTRACT
Habitats with soil pH , 4 and Ca , 100 ppm, such as pocosins, Sphagnum bogs and heathlands,
would appear inimical to land-snail biodiversity. Nevertheless, a survey of 1,356 sites, c. 1/2 million
individuals and over 240 species (c. 1/5 of the continental fauna) across North America shows that c.
10% of species appear to favour such highly or moderately acidic sites, spread from subtropical
forests of the Gulf of Mexico coast to the arctic tundra. Nonmetric multidimensional scaling ordination
of faunas from 292 sites that support at least five co-occurring species documents that the principal
axis of compositional variation in highly and moderately acidic habitats is significantly (P ,
0.0005) correlated with latitude, while the second axis is significantly (P , 0.0005) correlated with
moisture level. Composition was found to vary continuously along both axes, implying that discrete
acidophilic communities are not present. While highly and moderately acidic sites were shown to
have significantly (P , 0.000000005) lower richness and abundance compared with neutral/calcareous
habitats, even the most acidic sites still typically supported 5–10 species. Abundance distributions
in highly acidic habitats were found to be more uneven those of neutral/calcareous sites. The
greater richness of the North American acidophilic land snail fauna compared with that in Europe
has allowed communities to display replacement-driven compositional turnover. These results demonstrate
that it is vitally important for biodiversity surveys of North American land snails not to ignore
acidic habitats, because they harbour an important and surprisingly diverse fauna.
INTRODUCTION
It is well established that terrestrial gastropod communities in
lime-rich habitats such as carbonate bedrock outcrops, fens,
chalk grasslands and rich upland and lowland forests often
support abundant and diverse land-snail communities (e.g.
Boycott, 1934; Burch, 1962; Kerney & Cameron, 1979;
Stanisic, 1997; Barker & Mayhill, 1999; Nekola, 1999;
Schilthuizen, 2004; Horsa´ k, 2005). In fact, one of the principle
global trends identified in land-snail ecology is the strong positive
correlation between individual abundance, species richness
and the pH of the organic litter in which land-snail communities
reside (Valovirta, 1968; Wa¨ reborn, 1970; Walde´ n, 1981;
Outeiro, Agu¨ era & Parejo, 1993; Barker & Mayhill, 1999;
Nekola & Smith, 1999; Hermida, Ondina & Rodreiguez, 2000;
Pokryszko & Cameron, 2005; Horsa´ k, 2006). This is not surprising
given the high metabolic calcium demands of land
snails not only for shell generation but also for egg production
(Wa¨ reborn, 1970; Ga¨ rdenfors, 1992). Across the entire northwestern
European fauna of about 300 species, only five
(Columella aspera, Vertigo ronnebyensis, Vertigo lilljeborgi, Zonitoides
excavatus and Zoogenetes harpa) have been noted to favour acidic
habitats (Boycott, 1934; Kerney & Cameron, 1979). Highly
acidic habitats, such as Sphagnum bogs, heathlands and pine
forest have thus generally been regarded as unfavourable for
high land-snail biodiversity (Karlin, 1961; Kerney &
Cameron, 1979; Horsa´ k & Ha´ jek, 2003).
No strongly acidophilic/calcifugic land-snail fauna had previously
been reported from North America (Pilsbry, 1948;
Hubricht, 1985) when Vertigo malleata was described and shown
to be limited to ombotrophic peatlands of the Atlantic coastal
plain (Coles & Nekola, 2007). These sites possess soils with
organic litter pH ranging from 3.5 to 4.0 and Ca concentrations
of only 25–100 ppm (Woodwell, 1958, Richardson,
1983). An additional 35 co-occurring species, representing
about one-fifth of the entire North American Atlantic coastal
plain fauna, were noted from these and adjacent upland pineflatwood
sites, which have soil pH ranging from 3.5 to 4.5 and
Ca ppm values of 20–500 (Binkley et al., 1992). Some of the
species observed in these habitats, such as Triodopsis soelneri and
Vertigo alabamensis, are thought to be among the continent’s
rarest species (Hubricht, 1985; NatureServe, 2009).
Does this apparent acidophilic land-snail fauna from the
Atlantic seaboard represent an idiosyncratic outlier, or do
similar faunas occur across North America? How many species
should be considered acidophilic/calcifugic within the continental
fauna? What major gradients in community composition
can be identified, and with which environmental
variables do these correlate? How does community richness
and abundance vary between base-rich and base-poor habitats
across the major land-snail habitat types? To provide a better
understanding of the community ecology of acidophilic
land-snail communities within North America these issues are
addressed in the present paper.
MATERIAL AND METHODS
Study sites
Land-snail faunas were analysed from 1,356 sites ranging from
the Alaskan North Slope and southern California coast eastwards
to central Quebec, and the Atlantic seaboard from
Maine to Georgia (Fig. 1). Matrix vegetation varies from
tundra and taiga in the north, to tallgrass prairie in the
Midwest, to mixed deciduous-conifer forest in the southwest
and to mixed boreal-hardwood, deciduous and evergreen subtropical
forest in the east and south (Barbour & Billings,
1988). Using the principal land-snail community types identified
by Nekola (2003), this sample includes 377 rock outcrops,
Journal of Molluscan Studies (2010) 76: 144–156. Advance Access Publication: 17 January 2010 doi:10.1093/mollus/eyp053
# The Author 2010. Published by Oxford University Press on behalf of The Malacological Society of London, all rights reserved
atUniversityofNewMexicoonApril28,2010http://mollus.oxfordjournals.orgDownloadedfrom
54
357 upland forests, 286 lowland forests, 114 upland grasslands
and 222 lowland grasslands.
Field methods
Documentation of terrestrial gastropod faunas from representative
100–1,000 m2
areas within sites was accomplished by
hand collection of larger shells and litter sampling for smaller
taxa. Particular focus was placed on the latter approach as it
provides the most complete assessment of site faunas (Oggier,
Zschokke & Baur, 1998; Cameron & Pokryszko, 2005). As
suggested by Emberton, Pearce & Randalana (1996), collections
were made at places of high micromollusc density such as
loosely compacted leaf litter lying on top of highly compacted
damp mineral soil or humus. Approximately 500 ml of sieved
litter ranging from 0.425 to 9.5 mm in minimum particle size
was retrieved from each site, either by field sieving or collection
of c. 4 l of unsorted leaf litter. These unsorted bagged samples
were slowly and completely dried with the c. 500 ml of sorted
litter extracted from each sample in the laboratory by use of
careful but vigorous water disaggregation. A half litre of sieved
litter was set as the sample goal, as this volume typically captured
10 times the number of individuals as species, which is
advocated by Cameron & Pokryszko (2005) for accurate documentation
of land-snail community composition.
The latitude–longitude position of each site was determined
using either USGS 7.5 min topographic maps or a hand-held
GPS. Each site was scored in one of the five principal
land-snail habitat categories identified by Nekola (2003). Sites
were also scored along a moisture gradient from 1 (dry vertical
rock outcrops and xeric uplands) to 5 (completely saturated
vegetation mats). Scores of 2 represented dry-mesic uplands, 3
represented mesic sites and 4 lowland sites with nearly saturated
soils.
The soil acidity of each site was assessed by observation of
exposed bedrock, vascular plant and bryophyte cover.
Bryophyte composition was particularly noted as this group
serves as one of the best bio-indicators of soil-water solution
acidity (Vitt & Chee, 1990). Sites were categorized as ‘highly
acidic’ when exposed bedrock consisted of silica sand, shale,
quartzite, schist, gneiss or felsic igneous rock, when vascular
plant cover was dominated by pines and/or oaks, the shrub
layer was limited to various heaths, Myrica and/or Ilex, and
when bryophytes were limited to acidophilic genera such as
Leucobryum, Polytrichum and Sphagnum (Crum & Anderson,
1981). Sites were categorized as ‘moderately acidic’ when
exposed bedrock consisted of CaCO3-cemented sandstone or
intermediate to intermediate-felsic igneous rock or when vascular
plant and bryophyte cover included not only acidophilic
taxa but also more base-demanding genera such as Climacium,
Hypnum, Mnium, Rhodobryum, Rhytididelphus and Thuidium (Crum
& Anderson, 1981). Sites were characterized as ‘neutral or calcareous’
when exposed bedrock consisted of carbonates or
mafic/ultramafic igneous rock and/or when acidophilic vascular
plants and bryophytes were absent.
Laboratory procedures
Retained leaf litter fractions were passed through a standard
sieve series [ASTME 3/8" (9.5 mm), #10 (2.0 mm), #20
(0.85 mm) and #40 (0.425 mm) mesh screens], and then hand
picked against a neutral-brown background. All shells and
shell fragments were removed, and assigned to species (or subspecies)
using the author’s reference collection. The total
numbers of shells per species per site were recorded, as were
the number of unassignable immature individuals and fragmentary
shells. Nomenclature is based principally on Hubricht
(1985), with updates by Turgeon et al. (1998), Nekola (2004)
and Nekola & Coles (in press). Authors and dates for each
species are given in Table 1.
Statistical procedures
Occurrences for all species encountered were analysed in terms
of their presence/absence in highly acidic, moderately acidic,
or neutral/calcareous soils. The species limited only to highly
acidic and/or moderately acidic sites were identified, as well as
those species limited to neutral/calcareous sites. Additionally,
species found across the entire base-status spectrum were
enumerated.
The impact of base-status on population size was documented
by first recording the observed population sizes for each
species across all sites. These abundances were then grouped
by their soil acidity category, with statistical difference in
abundances between these groups being calculated using the
nonparametric Kruskal–Wallis test. To provide appropriate
statistical power to mitigate potential type-II errors, tests were
limited to the 128 species with at least 10 total occurrences and
30 total individuals. A Bonferroni correction was used to
modify the significance threshold of these tests to 0.05/128 or
0.00039.
Gradients in composition of acidic habitat land-snail communities
were identified via global nonmetric multidimensional
scaling (NMDS) using DECODA (Minchin, 1990). NMDS
was used as it makes no assumptions regarding the underlying
nature of distributions of species along compositional gradients.
As such, NMDS is the most robust form of ordination for
detection of ecological patterns (Minchin, 1987). To ordinate
sites, a matrix of dissimilarity coefficients was calculated
between all pairwise combinations of highly or moderately
acidic sites with five or more species, using the Czekanowski
(Bray–Curtis) index (Faith, Minchin & Belbin, 1987) on
species abundance data which had been doubly standardized,
first to make all species maxima ¼ 1, and then to equalize the
total number of individuals per site. All species (including the
most rarely encountered) were considered. NMDS in one
through four dimensions was then performed, with 200
Figure 1. Location of all 1,356 sampling sites. Black circles represent
highly or moderately acidic habitats; open circles represent neutral/
calcareous stations.
ACIDOPHILIC LAND-SNAIL COMMUNITIES
145
atUniversityofNewMexicoonApril28,2010http://mollus.oxfordjournals.orgDownloadedfrom
55
Table 1. Calcifuge/calcicole status of species, based on their presence
or absence across the soil acidity spectrum.
Occurring only in highly acidic sites:
Daedalochila leporina (Gould, 1848)
Glyphyalinia n. sp.
Mesomphix perlaevis (Pilsbry, 1900)
Neohelix divesta (Gould, 1851)
Neohelix solemi Emberton, 1988
Triodopsis soelneri (J.B. Henderson, 1907)
Vertigo oralis Sterki, 1898
Occurring only in moderately acidic sites:
Lobosculum pustula (Fe´russac, 1822)
Pristiloma arcticum (Lehnert, 1884)
Occurring only in highly or moderately acidic sites:
Glyphyalinia luticola Hubricht, 1966
Paravitrea petrophila (Bland, 1883)
Triodopsis juxtidens (Pilsbry, 1894)
Vertigo alabamensis Clapp, 1915
Vertigo malleata Coles & Nekola, 2007
Occurring in both acidic and neutral/calcareous habitats:
Anguispira alternata (Say, 1817)
Anguispira fergusoni (Bland, 1861)
Carychium exiguum (Say, 1822)
Carychium exile H.C. Lea, 1842
Carychium mexicanum Pilsbry, 1891
Catinella avara (Say, 1824)
Cepaea hortensis (Mu¨ller, 1774)
Cochlicopa lubrica (Mu¨ller, 1774)
Cochlicopa lubricella (Porro, 1838)
Cochlicopa morseana (Doherty, 1878)
Columella columella alticola (Ingersoll, 1875)
Columella simplex (Gould, 1841)
Daedalochila dorfeuilliana (I. Lea, 1838)
Deroceras leave (Mu¨ller, 1774)
Discus catskillensis (Pilsbry, 1898)
Discus cronkhitei (Newcomb, 1865)
Discus patulus (Deshayes, 1830)
Discus patulus edentulus (Pilsbry, 1948)
Discus shimeki (Pilsbry, 1890)
Euchemotrema fraternum (Say, 1824)
Euchemotrema leai leai (A. Binney, 1840)
Euconulus alderi (Gray, 1840)
Euconulus chersinus (Say, 1821)
Euconulus dentatus (Sterki, 1893)
Euconulus fulvus (Mu¨ller, 1774)
Euconulus fulvus alaskensis (Pilsbry, 1899)
Euconulus polygyratus (Pilsbry, 1899)
Euconulus trochulus (Reinhardt, 1883)
Gastrocopta contracta (Say, 1822)
Gastrocopta holzingeri (Sterki, 1889)
Gastrocopta pentodon (Say, 1821)
Gastrocopta riparia Pilsbry, 1948
Gastrocopta rupicola (Say, 1821)
Gastrocopta tappaniana (C.B. Adams, 1842)
Glyphyalinia indentata (Say, 1823)
Glyphyalinia rhoadsi (Pilsbry, 1899)
Glyphyalinia solida (H.B. Baker, 1930)
Continued
Table 1. Continued
Glyphyalinia wheatleyi (Bland, 1883)
Guppya sterkii (Dall, 1888)
Haplotrema concavum (Say, 1821)
Hawaiia minuscula (A. Binney, 1840)
Helicodiscus inermis H.B. Baker, 1929
Helicodiscus notius Hubricht, 1962
Helicodiscus parallelus (Say, 1817)
Helicodiscus shimeki Hubricht, 1962
Hendersonia occulta (Say, 1831)
Inflectarius inflectus (Say, 1821)
Mesodon elevatus (Say, 1821)
Mesodon zaletus (A. Binney, 1837)
Mesomphix capnodes (W.G. Binney, 1857)
Mesomphix friabilis (W.G. Binney, 1857)
Mesomphix globosus (MacMillan, 1940)
Mesomphix subplanus (A. Binney, 1842)
Neohelix albolabris (Say, 1816)
Neohelix alleni (Wetherby in Sampson, 1883)
Nesovitrea binneyana (Morse, 1864)
Nesovitrea electrina (Gould, 1841)
Oxyloma retusa (I. Lea, 1834)
Paravitrea andrewsae (W.G. Binney, 1858)
Paravitrea lamellidens (Pilsbry, 1898)
Paravitrea multidentata (A. Binney, 1840)
Paravitrea significans (Bland, 1866)
Paravitrea simpsoni (Pilsbry, 1889)
Patera binneyanus (Pilsbry, 1899)
Patera perigraptus (Pilsbry, 1894)
Planogyra asteriscus (Morse, 1857)
Pomatiopsis lapidaria (Say, 1817)
Punctum blandianum Pilsbry, 1900
Punctum californicum Pilsbry, 1898
Punctum minutissimum (I. Lea, 1841)
Punctum n. sp.
Punctum randolphi (Dall, 1895)
Punctum vitreum H.B. Baker, 1930
Pupilla hebes (Ancey, 1881)
Pupilla muscorum (Linne´, 1758)
Pupisoma dioscoricola (C.B. Adams, 1845)
Pupisoma macneilli (Clapp, 1918)
Stenotrema barbatum (Clapp, 1904)
Stenotrema labrosum (Bland, 1862)
Stenotrema stenotrema (Pfeiffer, 1819)
Stenotrema unciferum (Pilsbry, 1900)
Striatura exigua (Stimpson, 1847)
Striatura ferrea Morse, 1864
Striatura meridionalis (Pilsbry & Ferriss, 1906)
Striatura milium (Morse, 1859)
Strobilops aenea Pilsbry, 1926
Strobilops affinis Pilsbry, 1893
Strobilops labyrinthica (Say, 1817)
Strobilops texasiana Pilsbry & Ferriss, 1906
Succinea ovalis Say, 1817
Succinea strigata Pfeiffer, 1855
Trichia hispida (Linne´, 1798)
Triodopsis hopetonensis (Shuttelworth, 1852)
Vallonia costata (Mu¨ller, 1774)
Continued
J. C. NEKOLA
146
atUniversityofNewMexicoonApril28,2010http://mollus.oxfordjournals.orgDownloadedfrom
56
Table 1. Continued
Vallonia gracilicosta Reinhardt, 1883
Vallonia pulchella (Mu¨ller, 1774)
Ventridens brittsi Pilsbry, 1892
Ventridens cerinoideus (Anthony, 1865)
Ventridens intertextus (A. Binney, 1841)
Ventridens pilsbryi Hubricht, 1964
Vertigo AK 1
Vertigo AK 2
Vertigo AK 3
Vertigo AK 5
Vertigo arthuri (von Martens, 1884)
Vertigo bollesiana (Morse, 1865)
Vertigo coloradensis (Cockerell, 1891)
Vertigo columbiana Sterki, 1892
Vertigo cristata Sterki, 1919
Vertigo elatior Sterki, 1894
Vertigo aff. genesii (Gredler, 1865)
Vertigo gouldii (A. Binney, 1843)
Vertigo hannai Pilsbry, 1919
Vertigo hubrichti Pilsbry, 1934
Vertigo milium (Gould, 1840)
Vertigo modesta hoppii (Mo¨ller, 1842)
Vertigo modesta modesta (Say, 1824)
Vertigo modesta ultima Pilsbry, 1948
Vertigo nylanderi Sterki, 1909
Vertigo oscariana (Sterki, 1890)
Vertigo oughtoni Pilsbry, 1948
Vertigo ovata Say, 1822
Vertigo paradoxa Sterki, 1900
Vertigo perryi Sterki, 1905
Vertigo pygmaea (Draparnaud, 1801)
Vertigo rugosula Sterki, 1890
Vertigo ventricosa (Morse, 1865)
Vitrina alaskana Dall, 1905
Vitrina limpida Gould, 1850
Vitrinizonites latissimus (Lewis, 1875)
Zonitoides arboreus (Say, 1816)
Zonitoides nitidus (Mu¨ller, 1774)
Zoogenetes harpa (Say, 1824)
Occurring only in neutral/calcareous sites:
Allogona profunda (Say, 1821)
Anguispira kochi (Pfeiffer, 1845)
Anguispira strongylodes (Pfeiffer, 1854)
Angusipira jessica Kutchka, 1938
Appalachina sayana (Pilsbry, 1906)
Carychium clappi Hubricht, 1959
Carychium nannodes Clapp, 1905
Catinella exile (Leonard, 1972)
Catinella gelida (F.C. Baker, 1927)
Catinella vermeta (Say, 1829)
Catinella wandae (Webb, 1953)
Cepaea nemoralis (Linne´, 1798)
Daedalochila bisontes Coles & Walsh, 2006
Daedalochila lithica (Hubricht, 1961)
Daedalochila peregrina (Rehder, 1932)
Discus macclintockii (F.C. Baker, 1928)
Continued
Table 1. Continued
Discus rotundatus (Mu¨ller, 1774)
Euchemotrema hubrichti (Pilsbry, 1940)
Euchemotrema leai aliciae (Pilsbry, 1893)
Gastrocopta abbreviata (Sterki, 1909)
Gastrocopta armifera (Say, 1821)
Gastrocopta corticaria (Say, 1816)
Gastrocopta cristata (Pilsbry & Vanatta, 1900)
Gastrocopta pellucida (Pfeiffer, 1841)
Gastrocopta procera (Gould, 1840)
Gastrocopta rogersensis Nekola & Coles, 2001
Gastrocopta similis (Sterki, 1909)
Gastrocopta sterkiana Pilsbry, 1912
Glyphyalinia caroliniensis (Cockerell, 1890)
Glyphyalinia cumberlandiana (Clapp, 1919)
Hawaiia n. sp.
Helicina orbiculata (Say, 1818)
Helicodiscus n. sp.
Helicodiscus roundyi (Morrison, 1935)
Helicodiscus singleyanus (Pilsbry, 1890)
Inflectarius edentatus (Sampson, 1889)
Inflectarius subpalliatus (Pilsbry, 1893)
Lamellaxis gracilis (Hutton, 1834)
Lobosculum pustuloides (Bland, 1851)
Mesodon clausus clausus (Say, 1821)
Mesodon normalis (Pilsbry, 1900)
Mesodon thyroidus (Say, 1816)
Mesomphix cupreus (Rafinesque, 1831)
Mesomphix inornatus (Say, 1821)
Mesomphix rugeli (W.G. Binney, 1879)
Microphysula cookei (Pilsbry, 1922)
Neohelix dentifera (A. Binney, 1837)
Neohelix major (A. Binney, 1837)
Oxychylus cellarius (Mu¨ller, 1774)
Oxychylus draparnaudi (Beck, 1837)
Oxyloma haydeni (W.G. Binney, 1858)
Oxyloma peoriensis (Wolf, in Walker, 1892)
Oxyloma verrilli Bland, 1865
Paravitrea AR 1
Paravitrea tridens Pilsbry, 1946
Patera clenchi (Rehder, 1932)
Patera pennsylvanicus (Green, 1827)
Pilsbryna quadrilamellata Slapcinsky & Coles, 2002
Polygyra cereolus (Muhlfeld, 1818)
Polygyra semptemvolva Say, 1818
Pomatiopsis cincinnatiensis (I. Lea, 1840)
Punctum conspectum (Bland, 1865)
Punctum smithi Morrison, 1935
Pupoides albilabris (C.B. Adams, 1821)
Rabdotus dealbatus (Say, 1821)
Stenotrema altispira (Pilsbry, 1894)
Stenotrema hirsutum (Say, 1817)
Stenotrema pilsbryi (Ferriss, 1900)
Succinea indiana Pilsbry, 1905
Succinea putris (Linne´, 1798)
Trichia striolata (Pfeiffer, 1828)
Triodopsis cragini Call, 1886
Triodopsis discoidea (Pilsbry, 1904)
Continued
ACIDOPHILIC LAND-SNAIL COMMUNITIES
147
atUniversityofNewMexicoonApril28,2010http://mollus.oxfordjournals.orgDownloadedfrom
57
maximum iterations, a stress ratio stopping value of 0.9999 and
a small stress stopping value of 0.01. Output was scaled in halfchange
units, so that an interpoint distance of 1.0 will correspond,
on average, to a 50% turnover in species composition.
Because a given NMDS run may locate a local (rather than
the global) stress minimum, multiple NMDS runs were conducted
on a given set of data from different initial random
starting points to assess the stability of an individual solution
(Minchin, 1987). In this analysis, DECODA used a total of 10
random starting configurations. Solutions in each of the four
dimensions were compared using a Procrustes transformation
to identify those that were statistically identical. The number
of unique solutions, and number of runs which fell into each,
was then calculated across each dimension. The modal solution
out of 10 runs was identified, and was considered to be a
global optimum when it was achieved in at least 50% of starts.
Compositional variation across the diagram was documented
by dividing it into regions whose boundaries were either 0.5
standard deviation above or below the ordination centroid
along all major axes of variation. Within each of these regions,
all taxa found within at least 25% of included sites were
identified.
Environmental correlations with this diagram were determined
by calculating in DECODA the maximum correlation
vector for site latitude and moisture level. The significance of
each was estimated through Monte-Carlo simulations using
1,000 replications.
Full two-way ANOVAs were used to test for differences in
the central tendency of richness and log-transformed abundance
across the three acidity levels and five major terrestrial
gastropod habitat types. ANOVA was performed as richness
and log-transformed abundance were essentially normally distributed
across all three categories of habitat acidity. Even
though log-transformation greatly alters the appearance of the
species abundance distribution within sites, this technique is
appropriate when comparisons of identically transformed data
are made between sites (Nekola et al., 2008). The patterns
among habitat types and acidity levels were visualized using
box plots.
Abundance distribution variation between highly acidic,
moderately acidic and neutral/calcareous categories within
each major habitat type was displayed using rank–frequency
(dominance–diversity) plots. These were created by calculating
the proportional abundance for each species within a given
site. Abundances were then sorted from most to least abundant.
The median proportional abundance for the most abundant
species was then determined across all sites of a given
base-status within that habitat type. This process was repeated
for the second, third, fourth etc. to the least abundant species.
These abundance values were plotted on a log-axis vs rank position.
This process was repeated for all three base-status levels
within a given habitat type.
RESULTS
Even though almost two-thirds of sites sampled in California
and the Southwest occurred on highly felsic igneous outcrops,
acidophilic vegetation was never noted, with vascular plants
and bryophytes appearing essentially identical across the
bedrock acidity spectrum. In addition, no differences were
noted between the land-snail community compositions of these
sites. This result is almost certainly due to the high rates of
dryfall Ca inputs (Waring & Schlesinger, 1985) which have
been estimated to range between 190 and 200 mg/m2
/year in
southern California (Ellis, Verfaillie & Kummerow, 1983).
Because of low leaching rates, at least 98% of soil Ca in the
Southwest originates from atmospheric deposition (Capo &
Chadwick, 1999), allowing ample base to be present in all
soils, no matter what the parent material. For this reason, all
162 sites sampled in Arizona, California, Colorado and New
Mexico have been eliminated from further analysis, because
none actually represents acidophilic conditions, regardless of
the underlying bedrock type.
Calcifuge/calcicole status
Of the 1,194 analysed sites outside of the Southwestern USA,
land snails were observed from 1,188. Of these, 391 were
observed to be highly or moderately acidic: 91 rock outcrops,
67 upland forests, 124 lowland forests, 22 upland grasslands
and 87 lowland grasslands. In total, 241 species (c. 40% of the
eastern and northern North American fauna and c. 20% of the
entire North American fauna) and 486,153 identified individuals
were encountered across these sites. Essentially all wideranging
eastern and northern taxa were observed, with most of
the unencountered species representing regional endemics,
principally in the Polygyridae and Zonitidae (Hubricht,
1985).
Seven species only occurred in highly acidic habitats, two
only in moderately acidic habitats and a remaining five only in
either moderately or highly acidic habitats (Table 1). While 93
species were restricted to neutral/calcareous habitats, the
remaining 134 were found both in neutral/calcareous and
moderately/highly acidic sites.
Kruskal–Wallis analysis of abundances between highly
acidic, moderately acidic and neutral/calcareous sites for the
128 species with at least 10 occurrences and 30 individuals
(Table 2) documented 10 species with abundances that were
significantly (P , 0.00039) greater in highly acidic sites, and
an additional seven that significantly (P , 0.00039) favoured
moderately acidic sites. One species tended (0.05 , P,
0.00039) to favour moderately acidic sites. A total of 34 species
was shown to display no significant (P . 0.05) changes in
abundance across the acidity spectrum. Of these, 23 had only
between 10 and 30 total occurrences, making it probable that
some of these nonsignificant test scores represent type-II errors.
This is particularly likely for Discus macclintockii, Hawaiia n. sp.
and Oxyloma verrilli, which had their twelve or fewer occurrences
limited to neutral/calcareous habitats. Likewise, it seems
Table 1. Continued
Triodopsis neglecta (Pilsbry, 1899)
Triodopsis obsoleta (Pilsbry, 1894)
Triodopsis tridentata (Say, 1816)
Triodopsis vulgata (Pilsbry, 1948)
Vallonia excentrica Sterki, 1893
Vallonia parvula Sterki, 1892
Vallonia perspectiva Sterki, 1892
Ventridens acerra (Lewis, 1870)
Ventridens ligera (Say, 1821)
Vertigo AK 4
Vertigo binneyana Sterki, 1890
Vertigo clappi Brooks & Hunt, 1936
Vertigo meramecensis Van Devender, 1979
Vertigo morsei Sterki, 1894
Vertigo parvula Sterki, 1890
Vertigo tridentata Wolf, 1870
Webbhelix multilineata (Say, 1821)
Xolotrema denotata (Fe´russac, 1821)
Xolotrema fosteri (F.C. Baker, 1932)
Zonitoides limatulus (W.G. Binney, 1840)
J. C. NEKOLA
148
atUniversityofNewMexicoonApril28,2010http://mollus.oxfordjournals.orgDownloadedfrom
58
Table 2. Base-status preferences based upon Kruskal–Wallis analysis
of site abundance.
Species P-value
Species significantly favouring highly acidic sites:
Euconulus chersinus 0.00000039
Glyphyalinia luticola 0
Glyphyalinia n. sp. 0
Glyphyalinia solida 0
Striatura meridionalis 0.00000009
Strobilops texasiana 0
Ventridens cerinoideus 0.00000002
Vertigo alabamensis 0
Vertigo malleata 0
Vertigo modesta hoppii 0.00004532
Species significantly favouring moderately acidic sites:
Planogyra asteriscus 0.00000001
Striatura exigua 0.00000002
Striatura ferrea 0.00002965
Vertigo cristata 0
Vertigo perryi 0
Vertigo ventricosa 0
Zoogenetes harpa 0
Species tending to favour moderately acidic sites
Vertigo nylanderi 0.02831
Species demonstrating no preference in base status of habitat (30
occurrences):
Cochlicopa morseana 0.1217
Euconulus fulvus alaskensis 0.9992
Glyphyalinia rhoadsi 0.2121
Pupilla hebes 0.08355
Striatura milium 0.06981
Succinea strigata 0.4538
Vertigo AK 2 0.906
Vertigo coloradensis 0.1796
Vertigo hannai 0.5831
Vertigo modesta modesta 0.1135
Vertigo ovata 0.3306
Species demonstrating no preference in base status of habitat (,30
occurrences):
Discus macclintockii 0.08603
Euconulus dentatus 0.404
Euconulus trochulus 0.7056
Glyphyalinia wheatleyi 0.0512
Hawaiia n. sp. 0.08603
Helicodiscus inermis 0.0714
Helicodiscus notius 0.1385
Inflectarius inflectus 0.2089
Neohelix albolabris 0.0578
Oxychylus cellarius 0.08603
Oxyloma verrilli 0.05241
Paravitrea significans 0.1658
Pomatiopsis lapidaria 0.0666
Punctum blandianum 0.1177
Pupilla muscorum 0.3372
Strobilops aenea 0.508
Ventridens brittsi 0.9471
Vertigo AK 1 0.3169
Continued
Table 2. Continued
Species P-value
Vertigo AK 3 0.1944
Vertigo AK 5 0.9488
Vertigo genesii 0.819
Vertigo modesta ultima 0.2797
Vertigo oscariana 0.2958
Species significantly favouring neutral/calcareous sites:
Allogona profunda 0.00001154
Anguispira alternata 0
Carychium exiguum 0
Carychium exile 0
Catinella avara 0.00000003
Cochlicopa lubrica 0
Cochlicopa lubricella 0
Columella simplex 0
Deroceras laeve 0
Discus catskillensis 0.00000001
Discus cronkhitei 0.00000013
Euchemotrema fraternum 0
Euchemotrema leai leai 0
Euconulus alderi 0.0001001
Euconulus fulvus 0.00000592
Euconulus polygyratus 0
Gastrocopta armifera 0
Gastrocopta contracta 0
Gastrocopta corticaria 0
Gastrocopta holzingeri 0
Gastrocopta pentodon 0
Gastrocopta similis 0.00000013
Vallonia perspectiva 0
Vallonia pulchella 0.00000001
Vertigo arthuri 0.00000276
Vertigo bollesiana 0.00000102
Vertigo elatior 0
Vertigo gouldi 0
Gastrocopta tappaniana 0.00000005
Glyphyalinia indentata 0
Guppya sterkii 0.0001567
Haplotrema concavum 0.0001865
Hawaiia minuscula 0
Helicodiscus parallelus 0
Helicodiscus shimeki 0.0001685
Helicodiscus singleyanus 0.0001555
Hendersonia occulta 0
Nesovitrea binneyana 0
Nesovitrea electrina 0
Oxyloma retusa 0.00000399
Paravitrea multidentata 0.00000049
Punctum vitreum 0
Pupoides albilabris 0.00000021
Stenotrema barbatum 0.0001403
Strobilops affinis 0.00008082
Strobilops labyrinthica 0
Succinea ovalis 0.00008805
Vallonia costata 0
Vallonia gracilicosta 0
Vallonia parvula 0.0000716
Continued
ACIDOPHILIC LAND-SNAIL COMMUNITIES
149
atUniversityofNewMexicoonApril28,2010http://mollus.oxfordjournals.orgDownloadedfrom
59
probable that with additional observations Punctum blandianum
will be found strongly to favour acidic sites, as its population
sizes are more than five times greater on highly acidic as compared
to neutral/calcareous sites. The remaining 76 taxa either
significantly (P , 0.00039) favoured (56 species) or tended
(0.05 , P ,0.00039) to favour (20 species) neutral/calcareous
sites.
Compositional gradients
Visual observation of the optimal NMDS result (the twodimensional
solution) using all 305 highly or moderately acidic
sites with at least five observed species indicated a group of 13
outliers along Axis 1 that represented tundra or near-tundra
sites in Alaska and northern Manitoba. To assist gradient
interpretation, these sites were assigned to a unique compositional
group, and then removed from analysis. NMDS of the
remaining 292 sites in one dimension generated a minimum
stress configuration of 0.2603, which was achieved from nine of
10 random starting configurations. Procrustes analysis showed
that these solutions fell into only a single group. NMDS along
two dimensions generated a minimum stress configuration of
0.1934, which was achieved from all 10 random starting configurations.
Procrustes analysis showed that these solutions fell
into three unique categories, of which one was achieved in six
different starts. NMDS along three dimensions generated a
minimum stress configuration of 0.1489, which was achieved
from six of the 10 random starting configurations. Procrustes
analysis showed that these minimum stress solutions fell into
two unique groups, each achieved in three different starting
configurations. NMDS along four dimensions generated a
minimum stress configuration of 0.1220, which was achieved
from nine of 10 random starting configurations. Procrustes
analysis showed that each of these solutions was unique.
Based on these results the modal configuration from the twodimensional
NMDS was chosen as the most stable for depicting
land-snail community composition trends (Fig. 2).
Approximately two complete turnovers in species composition
are indicated along the first axis, with a single complete turnover
being noted along the second. Environmental biplot
analysis (Fig. 3) demonstrates that the first axis strongly (P ,
0.0005) correlates with site latitude, with a maximum observed
correlation of r ¼ 0.8366. The second axis strongly (P ,
0.0005) correlates with water level, with a maximum observed
correlation of r ¼ 0.6753. Ordination region one is found in the
upper left of the diagram, and principally represents low latitude
wetlands such as pocosins and bay-head swamps. Twelve
species were found to occur in at least 25% of these sites
(Table 3). Ordination region two is found at the mid-left side
of the diagram, and principally represents low latitude mesic
sites such as pine forest, bay forest and heathlands. Nine
species were found to occur in at least 25% of these sites.
Ordination region three is found on the lower left side of the
diagram, and principally represents low latitude xeric sites
such as dry bedrock outcrops, talus slopes and heath balds.
Nine species were found to occur in at least 25% of these sites.
Ordination region four is in the upper-mid portion of the
diagram, and principally represents mid-latitude wetlands such
as sedge meadows, and swamps of tamarack, northern white
cedar, willow, red maple and alder. Twenty-three species were
found to occur in at least 25% of these sites. Ordination region
five is found in the centre of the diagram, and principally represents
mid-latitude mesic sites such as pine, hemlock and
yellow birch forests as well as some mesic bedrock outcrops.
Fourteen species were found to occur in at least 25% of these
sites. Ordination region six is on the bottom margin of the
diagram, and principally represents mid-latitude xeric sites
such as exposed bedrock outcrops and dry upland forest.
Fourteen species were found to occur in at least 25% of these
sites. Ordination region seven is on the upper right of the
diagram, and principally represents high-latitude wetlands
such as cattail swamps, sedge meadows and Sphagnum bogs.
Ten species were found to occur in at least 25% of these sites.
Ordination region eight is on the right-centre margin of the
diagram, and principally represents high-latitude mesic sites
such as heathlands, spruce-fir forest and some mesic rock outcrops.
Eleven species were found to occur in at least 25% of
these sites. Ordination region nine is found on the lower right
of the diagram, and principally represents high-latitude xeric
sites such as exposed bedrock outcrops. Nine species were
found in at least 25% of these sites. Lastly, ordination region
10 represents the outlier sites found along the middle of Axis 2
to the right of the ordination diagram. These sites represent
tundra sites across the full extent of the moisture gradient,
ranging from sedge meadows to exposed rock outcrops. Ten
species were found in at least 25% of these sites.
Richness and abundance patterns
ANOVA of richness and abundance patterns between the three
soil acidity classes within the five major land-snail habitat
types demonstrate that soil acidity imposed a highly significant
(P , 0.000000005) negative impact on both site richness and
log-transformed abundance (Table 4). Median richness levels
ranged from 4.5 to 10 species per site in highly acidic soils,
increased to 7–15 in moderately acidic soils, and were 11.5–20
in neutral/calcareous sites (Fig. 4). Median total abundance
Table 2. Continued
Species P-value
Vertigo hubrichti 0
Vertigo milium 0
Vertigo paradoxa 0.00000027
Vertigo pygmaea 0.0001343
Vertigo tridentata 0
Zonitoides arboreus 0
Species tending to favour neutral/calcareous sites:
Catinella exile 0.0005619
Catinella ’gelida’ 0.000435
Catinella ’vermeta’ 0.03188
Columella columella alticola 0.04251
Gastrocopta procera 0.002008
Gastrocopta rogersensis 0.00552
Mesodon clausus clausus 0.0007254
Mesodon thyroidus 0.004289
Neohelix alleni 0.002879
Oxyloma haydeni 0.02484
Punctum minutissimum 0.00181
Punctum n. sp. 0.007416
Triodopsis tridentata 0.0071
Webbhelix multilineata 0.00552
Vertigo meramecensis 0.001558
Vertigo morsei 0.002587
Vertigo oughtoni 0.02095
Vitrina limpida 0.001014
Xolotrema fosteri 0.04088
Zonitoides nitidus 0.001456
The significance threshold has been modified to P  0.00039 based on a
Bonferroni correction.
J. C. NEKOLA
150
atUniversityofNewMexicoonApril28,2010http://mollus.oxfordjournals.orgDownloadedfrom
60
ranged from 50.5 to 119.5 individuals in highly acidic soils to
57–292.5 in moderately acidic soils and 211–440 in neutral/
calcareous sites (Fig. 4). Habitat type also had a highly significant
(P , 0.000000005) impact on site richness, but only a
marginally significant (P ¼ 0.04) impact on log-transformed
total abundance. Bedrock outcrop sites had the highest
observed richness with a median score ranging from 10 to 20
species, followed by upland forests (8–15), lowland forests (6–
15), lowland grasslands (5–12) and upland grasslands (4.5–
11.5). While the interaction between habitat type and soil
acidity significantly impacted site richness (P ¼ 0.001), no
effect on log-transformed abundance was noted (P ¼ 0.69).
The significant interaction appears due to the fact that richness
of highly and moderately acidic bedrock outcrop sites was identical,
as were the richness of moderately acidic and neutral/calcareous
upland forests (Fig. 4).
Rank–abundance plots across all five habitat types demonstrate
a clear tendency for highly acidic sites to have more
uneven abundance distributions than for neutral/calcareous
sites (Fig. 5), with the commonest species in highly acidic habitats
tending to have a higher proportional median abundance
and with scores rapidly decreasing to much smaller values for
a given rank. In most cases, curves from moderately acidic sites
were intermediate between highly acidic and neutral/calcareous
sites. Exceptions to this were noted in lowland forest and
upland grasslands, in which the curves for moderately acidic
and neutral/calcareous sites were largely coincident.
DISCUSSION
The North American acidophilic land-snail fauna
These analyses clearly document that Vertigo malleata is not
an idiosyncratic anomaly within the North American
land-snail fauna. Almost two-thirds of all analysed northern
and eastern North American species were found to occur at
least occasionally in highly to moderately acidic habitats. Of
these, 16 (Daedalochila leporina, Euconulus chersinus, Glyphyalinia
luticola, Glyphyalinia n. sp., Glyphyalinia solida, Mesomphix
perlaevis, Neohelix divesta, Neohelix solemi, Striatura meridionalis,
Strobilops texasiana, Triodopsis soelneri, Ventridens cerinoideus,
Vertigo alabamensis, V. malleata, Vertigo modesta hoppii and Vertigo
oralis) only occurred or had statistically higher abundance in
highly acidic sites. An additional 10 (Lobosculum pustula,
Planogyra asteriscus, Pristiloma arcticum, Striatura exigua, Striatura
ferrea, Vertigo cristata, Vertigo nylanderi, Vertigo perryi, Vertigo
ventricosa and Zoogenetes harpa) only occurred or had statistically
higher abundance in moderately acidic sites. Paravitrea
petrophila and Triodopsis juxtidens were also only seen in moderately
and highly acidic sites. Thus, about 12% of the
sampled fauna exhibits clear signs of acidophilic/calcifugic
affinities. Almost all of these taxa are wide-ranging, with
only one (Triodopsis solneri) being a local endemic. While
other presumed acidophiles (e.g. Catinella oklahomarum,
Inflectarius downieanus, Mesodon normalis, Neohelix dentifera,
Praticolella bakeri, Praticolella jejuna, Praticolella lawae and
Praticolella mobiliana) have been reported previously
(Hubricht, 1985), they have either not yet been encountered
or have too few observed occurrences to draw firm conclusions
regarding their preferred base status. The actual
number of acidophilic/calcifugic land-snail species in
North America is thus certainly higher than is reported here.
It is therefore vitally important for North American
land-snail biodiversity surveys not to ignore acidic habitats,
because they harbour an important and surprisingly diverse
fauna.
Acidophilic tendencies are not limited to a few closely
related groups, with the taxa listed above representing at
least eight families (Euconulidae, Gastrodontidae,
Oxychilidae, Polygyridae, Pristolomatidae, Valloniidae and
Vertiginidae of Bouchet et al., 2005). With additional
sampling, two more (Punctidae and Succineidae) will also
likely be added to this list. Most of this fauna shares little
phylogenetic overlap with the known European acid-tolerant
fauna: only two of the above genera (Vertigo and Zoogenetes)
possess acidophilic taxa in both the Old and New Worlds
(Kerney & Cameron, 1979). Instead, most of the North
American acidophilic fauna represents endemic families or
genera, such as the Polygyridae (representing one-fifth of
identified acidophilic species) and the gastrodontid genus
Striatura (all species demonstrated acidophilic or generalist
tendencies). Biogeographically, acidophilic species were found
to range across the entire continent from semitropical forests
of the Gulf Coast to arctic tundra on the Alaska North
Slope. These factors all strongly suggest that the evolution of
acidophilic traits has occurred in situ multiple times within
North America.
A few clear analogues between the European and North
American acidophilic faunas can be observed, however. For
instance, Vertigo ronnebyensis not only shares a very similar shell
with V. cristata, but also an apparent preference for acidic taiga
soils (Kerney & Cameron, 1979). Additionally, Vertigo lilljeborgi
not only shares roughly similar shell characteristics with V.
perryi, but similar affinities with tussock-side sedge litter
accumulations in acidic wet meadows. The Holarctic Zoogenetes
harpa also appears to demonstrate acidophilic tendencies across
its range. It remains unknown how close the evolutionary
relationships between these taxa and populations may be,
given that at least some supposed Holarctic land-snail species
can be shown to be members of only distantly related clades
based on their mitochondrial DNA sequences (Nekola, Coles &
Bergthorsson, in press).
Figure 2. Optimum two-dimensional NMDS solution, showing
locations of the 10 regions used to determine compositional changes.
Figure 3. Environmental biplot diagram showing the vectors of
maximum correlation for latitude and moisture levels.
ACIDOPHILIC LAND-SNAIL COMMUNITIES
151
atUniversityofNewMexicoonApril28,2010http://mollus.oxfordjournals.orgDownloadedfrom
61
Table 3. Ordered matrix of species frequency within each of the 10 identified NMDS ordination regions, with the reported number representing
the per cent of sites within that region which supported populations of that given species.
Species Region 1 Region 2 Region 3 Region 4 Region 5 Region 6 Region 7 Region 8 Region 9 Region 10
Vertigo malleata 64.3 41.4
Strobilops texasiana 57.1 55.2
Punctum minutissimum 57.1 82.8 35.3 86.5 88.4 65.4
Gastrocopta pentodon 50.0 62.1 70.6 27.9
Vertigo milium 35.7
Striatura milium 35.7 34.5 78.9 95.4 84.6
Striatura meridionalis 35.7 48.3 76.4
Euconulus chersinus 35.7
Gastrocopta tappaniana 35.7 80.8 36.0
Glyphyalinia solida 35.7
Hawaiia minuscula 35.7
Gastrocopta contracta 28.6
Zonitoides arboreus 34.5 58.8 44.2 53.5 88.5 51.7 74.2
Vertigo alabamensis 31.0
Glyphyalinia indentata 27.6 47.1
Columella simplex 29.4 80.8 72.1 36.5 28.0 51.7
Haplotrema concavum 29.4
Helicodiscus parallelus 29.4 27.9
Guppya sterkii 29.4
Nesovitrea electrina 92.3 32.6 80.0 58.6
Striatura exigua 78.9 65.1 51.9 27.6 29.0
Carychium exiguum 71.2
Strobilops labyrinthica 67.3 81.4 42.3
Euconulus fulvus 67.3 79.1 36.5 48.0 93.1 71.0
Euconulus alderi 65.4 32.0
Vertigo cristata 59.6 74.4 67.3 72.4 80.7
Vertigo nylanderi 51.9
Vertigo ventricosa 50.0 28.0
Nesovitrea binneyana 50.0 79.1 71.2 32.0 55.2 74.2
Planogyra asteriscus 46.2 32.6
Striatura ferrea 34.6 41.9 27.6
Discus catskillensis 32.7 65.1 76.9 74.2
Vertigo elatior 32.7
Punctum n. sp. 30.8
Succinea ovalis 30.8
Helicodiscus shimeki 30.8 39.5 36.5
Carychium exile 26.9
Vertigo bollesiana 34.9
Zoogenetes harpa 44.2 48.3 38.7
Vertigo paradoxa 38.5 38.7
Anguispira alternata 26.9
Discus cronkhitei 40.0 51.7 38.7
Vertigo perryi 32.0
Vertigo modesta modesta 28.0 69.2
Vitrina limpida 27.6
Euconulus fulvus alaskensis 76.9
Vertigo modesta hoppii 76.9
Vertigo AK 2 61.5
Vertigo hannai 46.2
Succinea strigata 46.2
Columella columella alticola 38.5
Pupilla hebes 38.5
Vertigo modesta ultima 38.5
Vertigo genesii 30.8
Only species with a frequency of at least 25% within a region are listed.
J. C. NEKOLA
152
atUniversityofNewMexicoonApril28,2010http://mollus.oxfordjournals.orgDownloadedfrom
62
Acidic land-snail habitats
Acidophilic land-snail faunas were observed in a number of
habitat types across the continent. In the south, some of the
most diverse assemblages occurred in ombotrophic pocosin
peatlands, bayhead seepage swamps, Atlantic white cedar
forests, pine-wiregrass savanna, pine barrens, heath balds and
upland oak woods. To the north, other important reservoirs
for acidophilic land-snail diversity included Sphagnum bogs,
swamps of base-poor cattail, sedge, northern white cedar, red
maple, balsam fir and tamarack, wooded igneous rock outcrops
and base-poor upland forests. Acidophilic land snails were also
abundantly found across North America in base-poor taiga
and acidic tundra over base-poor rock such as sandstone,
quartzite and gneiss.
It is astonishing not only that land-snail communities exist on
such sites, but also that many of the included species possess
large populations (up to 2,000/m2
; Coles & Nekola, 2007) and
strongly calcified shells (e.g. S. texasiana, V. alabamensis). Because
of the very small Ca pools in such habitats (Binkley et al., 1992),
it is clear that these populations must be successfully capturing
Ca that is fluxing rapidly through the environment. Ca sources
in these sites are undoubtedly various organically derived Ca
salts, including Ca citrate and Ca oxalate (Wa¨ reborn, 1970),
which are leached from various tree and shrub species such as
aspen, dogwood or cedar (Karlin, 1961; Nation, 2007). Because
of the high densities of individuals observed in their litter
accumulations, it appears that heaths such as Chamaedaphne,
Gaylussacia, Lyonia, Kalmia, Vaccinium and Zenobia, and other
shrubs such as Ilex, Gordonia, Magnolia and Myrica, must also be
important Ca sources. No matter the source, all acidophilic
land snails must have evolved highly efficient Ca-pumps that
are able to move ions counter to very strong osmotic gradients.
If these pumps require low Ca concentrations or pH for proper
functioning, it might provide a metabolic explanation for the
calcifugic behaviour of some species.
Because of the very high soil acidity at many of these sites
(e.g. pH , 4; Woodwell, 1958; Binkley et al., 1992), the shells
of many acidophilic snails are being actively dissolved even on
living individuals. For instance, it is difficult to find living
V. malleata or V. alabamensis individuals that do not possess at
least one shell ulceration. Incipient ulcers can be seen in the
individuals figured by Coles & Nekola (2007) as opaque white
patches on otherwise translucent shells. Shell erosion is so
rapid in these habitats that no adult shells of the seasonally
active V. alabamensis could be found during autumn sampling,
even on sites that supported colonies of .200 individuals/m2
in spring (Nekola & Coles, in press).
Acidophilic land-snail community composition gradients
The dispersion of sites across the optimal two-dimensional
NMDS ordination diagram is continuous along both axes,
implying that there are no hard compositional breaks within
the fauna. This can be readily seen in the continuous variation
exhibited in pineland and ombotrophic bog faunas along
the Atlantic seaboard, with sites in the New Jersey Pine
Barrens being intermediate between eastern Maine raised
bogs and Carolina pocosins (Coles & Nekola, 2007). Similar
transitions can be found between Atlantic coastal plain and
mixed-deciduous forest faunas in central Massachusetts and
southern Maine, between mixed-deciduous and taiga faunas
along the Cote Nord in eastern Quebec and the western shore
of Lake Superior, and between taiga and tundra faunas in the
Alaskan interior, at Churchill, Manitoba, and in central
Quebec.
Analysis of environmental covariates identifies essentially
the same organizing factors as have been previously shown
in calcareous sites in the Great Lakes region (Nekola, 2003):
Table 4. Summary statistics for ANOVA on the impacts of soil acidity
and habitat type on community richness and log-transformed
individual abundance.
Predictor
variable
df Sum-of-
squares
Mean
squares
F-value P
A. Site richness
Habitat type 4 10,181 2,545 86.459 ,0.000000005
Soil acidity 2 11,602 5,801 197.064 ,0.000000005
Habitat type *
soil acidity
1 302 302 10.260 0.001396
B. Log-transformed total abundance
Habitat type 4 17.97 4.5 2.5178 0.03981
Soil acidity 2 451.26 225.6 126.4315 ,0.000000005
Habitat type *
soil acidity
0.28 0.3 0.1596 0.68959
Figure 4. Box-plot diagrams demonstrating the impact of habitat type and soil acidity on site richness (top) and log-transformed site abundance
(bottom).
ACIDOPHILIC LAND-SNAIL COMMUNITIES
153
atUniversityofNewMexicoonApril28,2010http://mollus.oxfordjournals.orgDownloadedfrom
63
latitude was most strongly correlated with the first axis,
while water level was strongly related to the second. As a
result, it appears that at a continental scale acidophilic
faunas respond in similar ways to the same environmental
variables as do calciphilic faunas. Although found to be very
important in base-rich communities, the impact of soil
surface architecture was not documented in the current
analysis, because thick organic litter accumulations (e.g. ‘duff
soils’) were present on essentially all sites. However, acidic
soils did tend towards turf-style architecture with increasing
water levels.
Diversity patterns
While faunas of acidic habitats were both more depauperate
in species and individuals than those of neutral/calcareous
sites, median richness fell by no more than half, while
abundance fell by no more than an order of magnitude.
Median richness of even the most acidic sites ranged up to 10
species. This contrasts with patterns observed in Europe, where
moderately acidic sites may often harbour as many species as
typical base-rich sites (Bishop, 1977, 1980; Tattersfield, 1990;
Cameron & Greenwood, 1991), and the most acidic sites typically
have maximum diversity levels of ,10 (Kerney &
Cameron, 1979).
As in Europe (Cameron & Pokryszko, 2005), faunas were
found to become increasingly uneven in their abundance distributions
with increasing soil acidity. Acidic site faunas thus
tend to be dominated by a relatively few common species, with
the remainder being rare. This pattern is common among a
large number of high-stress communities (Whittaker, 1975).
The steepness of the rank–abundance curve on acidic sites also
indicates that adequate documentation of site faunas may
require extra field collection effort as compared to sites with
higher base status (Cameron & Pokryszko, 2005).
Acidophilic land-snail communities in a global perspective
This documentation of acidophilic land-snail faunas is not
without precedent. Bishop (1977), Tattersfield (1990) and
Cameron & Greenwood (1991) have reported on land-snail
community patterns from acidic sites in southwestern Ireland,
the English Pennines and Scottish Highlands, respectively.
Valovirta (1968), Wa¨ reborn (1970), Walde´ n (1981),
Ga¨ rdenfors (1992) and Ma¨ nd, Ehlvest & Kristaja (2001) have
documented acidophilic land-snail faunas in Scandinavia and
the Baltic Republics. Ago´ csy (1968) and Sˇ tamol (1991) have
documented acidophilic land-snail faunas from eastern Europe,
while Bishop (1980) also demonstrated the existence of such
faunas in Italy. The existence of a low-elevation acidophilic
fauna from the North Island of New Zealand is also suggested
by the work of Barker & Mayhill (1999). The species found on
these sites can make up a not-inconsequential fraction of their
respective regional faunas: Boycott (1934) reported roughly
half of the British land-snail fauna to be ‘indifferent to lime’,
occurring freely in both acidic and calcareous sites, while
Bishop (1977) reported that about 40% of all Irish land snails
occur in acidic woodlands.
What makes the North American fauna outstanding,
however, is the presence of so many clearly acidophilic/calcifugic
species. While these make up an estimated 10% of the
sampled northern and eastern North American fauna, Boycott
(1934) could only identify one such species from Britain. Even
with updating by Kerney & Cameron (1979), it is clear that
Figure 5. Rank–abundance (a.k.a. dominance–diversity) plots for median proportional abundance across all highly acidic (dotted line),
moderately acidic (dashed line) and neutral/calcareous (solid line) sites within a given major habitat type.
J. C. NEKOLA
154
atUniversityofNewMexicoonApril28,2010http://mollus.oxfordjournals.orgDownloadedfrom
64
perhaps 1% or less of the northwestern European land-snail
fauna can be considered acidophilic. This has profound implications
in terms of composition gradients. First, the lack of
acid-preferring species allows for considerable stochastic compositional
variation between adjacent European acidic sites
(Cameron & Pokryszko, 2005). In North America, however,
the presence of acidophilic species allows for much greater consistency
in site composition: the standard deviations of Axis 1
and Axis 2 scores for Carolina pocosins were only 0.36 and
0.25, respectively (implying c. 82-87% similarity), 0.26 along
both Axis 1 and 2 for New England conifer swamps (c. 87%
similarity), and 0.20 and 0.19 (c. 90% similarity) for western
Lake Superior igneous outcrop sites.
Acidic-habitat faunas throughout Europe also tend to be
characterized by a suite of ubiquitous species which occur
across the entire base-status gradient. The major compositional
change with increasing base-status is the continuous addition
of increasingly calciphilic species (Valovirta, 1968; Walde´ n,
1981; Cameron & Greenwood, 1991; Horsa´ k & Ha´ jek, 2003)
until a few taxa drop out at the sites of highest base-status
(Horsa´ k, 2006). Among European studies, only Ago´ csy (1968)
did not recognize this pattern, instead noting that about 60%
of acidic-habitat snails were restricted to these sites, whereas
there was only 20% similarity between the faunas of adjacent
acidic and calcareous habitats. European land-snail community
compositions can thus be generally characterized as being
nested along a base-status axis. Although also suggested from
coastal British Columbia forests by Cameron (1986), this
pattern appears not to be typical for the rest of North
America. Rather, acidophilic faunas are made not only of
species ‘indifferent’ to lime, but also a group of species (c. 20%
of site faunas) that actually favour highly or moderately acidic
sites. As a result, replacement-driven turnover in species occurrence
can be seen along the base-status axis across much of
North America.
An important question that remains unanswered, however, is
why North America and Europe differ so much in their frequency
of acidophilic land-snail species. Given the apparently
frequent presence of acidophiles in land-snail faunas of West
Africa (De Winter & Gittenberger, 1998) and Borneo
(Schilthuizen & Rutjes, 2001), it seems possible that a productive
line of inquiry may be directed at determining why Europe has
an unusually depauperate fauna, as opposed to concentrating on
why that of North America appears strangely enriched.
ACKNOWLEDGEMENTS
Major funding for this project was provided by the National
Science Foundation (EAR-0614963), the Maine Department
of Inland Fisheries and Wildlife, Wildlife Resource Assessment
Section, the Massachusetts Natural Heritage and Endangered
Species Program, and the Minnesota Nongame Wildlife Tax
Checkoff and Minnesota State Park Nature Store Sales
through the Minnesota Department of Natural Resources
Natural Heritage and Nongame Research Program. Brian
Coles provided innumerable insights regarding field sampling
and microhabitat preferences for many of these species, as well
as thoughts regarding potentially analogous species in Europe
and North America. Invaluable suggestions on earlier drafts
were provided by Linda Fey, Robert Cameron and two anonymous
reviewers.
REFERENCES
AGO´ CSY, P. 1968. Data to quantitative conditions in the mollusk
faunas of two different substrates in central Hungary. Acta Zoologica
Academiae Scientiarum Hungaricae, 14: 1–6.
BARBOUR, M.G. & BILLINGS, W.D. 1988. North American terrestrial
vegetation. Cambridge University Press, New York.
BARKER, G.M. & MAYHILL, P.C. 1999. Patterns of diversity and
habitat relationships in terrestrial mollusc communities of the
Pukeamaru Ecological District, northeastern New Zealand. Journal
of Biogeography, 26: 215–238.
BINKLEY, D., RICHTER, D., DAVID, M.B. & CALDWELL, B.
1992. Soil chemistry in a Loblolly/Longleaf Pine forest with
interval burning. Ecological Applications, 2: 157–164.
BISHOP, M.J. 1977. The Mollusca of acid woodland in west Cork
and Kerry. Proceedings of the Royal Irish Academy, Section B, 77:
227–244.
BISHOP, M.J. 1980. The Mollusca of acid woodland in the Italian
province of Novara. Journal of Conchology, 30: 181–188.
BOUCHET, P., ROCROI, J.-P., FRY´ DA, J., HAUSDORF, B.,
PONDER, W., VALDE´ S, A´ . & WARE´ N, A. 2005. Classification
and nomenclator of gastropod families. Malacologia, 47: 1–397.
BOYCOTT, A.E. 1934. The habitats of land Mollusca in Britain.
Journal of Ecology, 22: 1–38.
BURCH, J.B. 1962. The Eastern land snails. W. C. Brown Co.,
Dubuque, Iowa.
CAMERON, R.A.D. 1986. Environment and diversities of forest snail
faunas from coastal British Columbia. Malacologia, 27: 341–355.
CAMERON, R.A.D & GREENWOOD, J.J.D. 1991. Some montane
and forest molluscan faunas from eastern Scotland: effects of
altitude, disturbance and isolation. In: Proceedings of the Tenth
International Molluscan Congress, Tu¨bingen, 1989 (C. Meier-Brook, ed.),
pp. 437–442. Unitas Malacologica, Tu¨ bingen.
CAMERON, R.A.D. & POKRYSZKO, B.M. 2005. Estimating the
species richness and composition of land mollusc communities.
Journal of Conchology, 38: 529–547.
CAPO, R.C. & CHADWICK, O.A. 1999. Sources of strontium and
calcium in desert soil and calcrete. Earth and Planetary Science Letters,
170: 61–72.
COLES, B.F. & NEKOLA, J.C. 2007. Vertigo malleata, a new extreme
calcifuge land snail from the Atlantic and Gulf coastal plains of the
U.S.A. (Gastropoda, Vertiginidae). Nautilus, 121: 17–28.
CRUM, H.A. & ANDERSON, L.E. 1981. Mosses of eastern North
America. Columbia University Press, New York.
DE WINTER, A.J. & GITTENBERGER, E. 1998. The terrestrial
gastropod fauna of a square kilometer patch of rainforest in
southwestern Cameroon: high species richness, low abundance, and
seasonal fluctuations. Malacologia, 40: 231–250.
ELLIS, B.A., VERFAILLIE, J.R. & KUMMEROW, J. 1983. Nutrient
gain from wet and dry atmospheric deposition and rainfall acidity in
southern California chaparral. Oecologica, 60: 118–121.
EMBERTON, K.C., PEARCE, T.A. & RANDALANA, R. 1996.
Quantitatively sampling land-snail species richness in Madagascan
rainforests. Malacologia, 38: 203–212.
FAITH, D.P., MINCHIN, P.R. & BELBIN, L. 1987. Compositional
dissimilarity as a robust measure of ecological distance. Vegetatio,
69: 57–68.
GA¨ RDENFORS, U. 1992. Effects of artificial liming on land snail
populations. Journal of Applied Ecology, 29: 50–54.
HERMIDA, J., ONDINA, M.P. & RODREIGUEZ, T. 2000. The
relative importance of edaphic factors on the distribution of some
terrestrial gastropod species: autecological and synecological
approaches. Acta Zoologica Scientiarum Hungaricae, 46: 265–274.
HORSA´ K, M. 2005. Molluscs. In: Ecology and palaeoecology of spring fens
of the west Carpathians (A. Poulı´cˇ kova, M. Ha´ jek & K. Rybnı´cˇ ek,
eds), pp. 197–208. Olomouc, Czech Republic.
HORSA´ K, M. 2006. Mollusc community patterns and species
response curves along a mineral richness gradient: a case study of
fens. Journal of Biogeography, 33: 98–107.
HORSA´ K, M. & HA´ JEK, M. 2003. Composition and species richness
of molluscan communities in relation to vegetation and water
chemistry in the western Carpathian spring fens: the poor-rich
gradient. Journal of Molluscan Studies, 69: 349–357.
HUBRICHT, L. 1985. The distributions of the native land mollusks of
the eastern United States. Fieldiana, new series, 24: 1–191.
ACIDOPHILIC LAND-SNAIL COMMUNITIES
155
atUniversityofNewMexicoonApril28,2010http://mollus.oxfordjournals.orgDownloadedfrom
65
KARLIN, E.J. 1961. Ecological relationships between vegetation and
the distribution of land snails in Montana, Colorado and New
Mexico. American Midland Naturalist, 65: 60–66.
KERNEY, M.P. & CAMERON, R.A.D. 1979. Field guide to the land
snails of the British Isles and northwestern Europe. Collins Press, London.
MA¨ ND, R., EHLVEST, A. & KRISTAJA, P. 2001. Land snails in an
afforested oil-shale mining area. Proceedings of the Estonian Academy of
Science: Biology and Ecology, 50: 37–41.
MINCHIN, P.R. 1987. An evaluation of the relative robustness of
techniques for ecological ordination. Vegetatio, 69: 89–108.
MINCHIN, P.R. 1990. DECODA software. Australian National
University, Canberra, Australia.
NATION, T.H. 2007. The influence of flowering dogwood (Cornus
florida) on land snail diversity in a southern mixed hardwood forest.
American Midland Naturalist, 157: 137–148.
NATURESERVE. 2009. NatureServe Explorer: an online encyclopedia of
life. Version 7.0. NatureServe, Arlington, Virginia. http://www
.natureserve.org/explorer. (accessed 5 September 2009).
NEKOLA, J.C. 1999. Terrestrial gastropod richness of carbonate cliff
and associated habitats in the Great Lakes region of North
America. Malacologia, 41: 231–252.
NEKOLA, J.C. 2003. Large-scale terrestrial gastropod community
composition patterns in the Great Lakes region of North America.
Diversity and Distribution, 9: 55–71.
NEKOLA, J.C. 2004. Terrestrial gastropod fauna of northeastern
Wisconsin and the southern Upper Peninsula of Michigan. American
Malacological Bulletin, 18: 21–44.
NEKOLA, J.C. & COLES, B.F. In press. Pupillid land snails of
eastern North America. American Malacological Bulletin.
NEKOLA, J.C., COLES, B.F. & BERGTHORSSON, U.
Evolutionary pattern and process in the Vertigo gouldii group of
minute North American land snails. Molecular Phylogenetics and
Evolution, 53: 1010–1024.
NEKOLA, J.C., Sˇ IZLING, A.L., BOYER, A.G. & STORCH, D.
2008. Artifactions in the log-transformation of species abundance
distributions. Folia Geobotanica, 43: 259–468.
NEKOLA, J.C. & SMITH, T.A. 1999. Terrestrial gastropod richness
patterns in Wisconsin carbonate cliff communities. Malacologia, 41:
253–269.
OGGIER, P., ZSCHOKKE, S. & BAUR, B. 1998. A comparison of
three methods for assessing the gastropod community in dry
grasslands. Pedobiologia, 42: 348–357.
OUTEIRO, A., AGU¨ ERA, D. & PAREJO, C. 1993. Use of
ecological profiles and canonical correspondence analysis in a study
of the relationship of terrestrial gastropods and environmental
factors. Journal of Conchology, 34: 365–375.
PILSBRY, H.A. 1948. Land Mollusca of North America (north of Mexico).
Academy of Natural Sciences of Philadelphia Monograph vol. 1(1)
and 2(2).
POKRYSZKO, B.M. & CAMERON, R.A.D. 2005. Geographic
variation in the composition and richness of forest land snail faunas
in northern Europe. Records of the Western Australian Museum,
Supplement, 68: 115–132.
RICHARDSON, C.J. 1983. Pocosins: vanishing wastelands or
valuable wetlands? Bioscience, 33: 626–633.
SCHILTHUIZEN, M. 2004. Land snail conservation in Borneo:
limestone outcrops act as arks. Journal of Conchology, Special
Publication, 3: 149–154.
SCHILTHUIZEN, M. & RUTJES, H.A. 2001. Land-snail diversity
in a square kilometre of tropical forest in Sabah, Malaysian
Borneo. Journal of Molluscan Studies, 67: 417–423.
Sˇ TAMOL, V. 1991. Coenological study of snails (Mollusca:
Gastropoda) in forest phytocoenoses of Medvednica mountain
(NW Croatia, Yugoslavia). Vegetatio, 95: 33–54.
STANISIC, J. 1997. An area of exceptional land snail diversity: the
Macleay Valley, north-eastern New South Wales. Memoirs of the
Museum of Victoria, 56: 441–448.
TATTERSFIELD, P. 1990. Terrestrial mollusc faunas from some
south Pennine woodlands. Journal of Conchology, 33: 355–374.
TURGEON, D.D., QUINN, J.F., JR, BOGAN, A.E., COAN, E.V.,
HOCHBERG, F.G., LYONS, W.G., MIKKELSEN, P., NEVES,
R.J., ROPER, C.F.E., ROSENBERG, G., ROTH, B.,
SCHELTEMA, A., THOMPSON, F.G., VECCHIONE, M. &
WILLIAMS, J.D. 1998. Common and scientific names of aquatic
invertebrates from the United States and Canada: mollusks. Edn. 2.
American Fisheries Society Special Publication vol. 26. Bethesda,
Maryland.
VALOVIRTA, I. 1968. Land molluscs in relation to acidity on
hyperite hills in central Finland. Annales Zoologici Fennici, 5:
245–253.
VITT, D.H. & CHEE, W.L. 1990. The relationships of vegetation to
surface water chemistry and peat chemistry in fens of Alberta,
Canada. Vegetatio, 89: 87–106.
WALDE´ N, H.W. 1981. Communities and diversity of land molluscs in
Scandinavian woodlands. I. high diversity communities in taluses
and boulder slopes in SW Sweden. Journal of Conchology, 30:
351–372.
WA¨ REBORN, I. 1970. Environmental factors influencing the
distribution of land molluscs of an oligotrophic area in southern
Sweden. Oikos, 21: 285–291.
WARING, R.H. & SCHLESINGER, W.H. 1985. Forest ecosystems:
concepts and management. Academic Press, New York.
WHITTAKER, R.H. 1975. Communities and ecosystems. MacMillan
Publishing Co., New York.
WOODWELL, G.M. 1958. Factors controlling growth of Pond Pine
seedlings in organic soils of the Carolinas. Ecological Monographs, 28:
219–236.
J. C. NEKOLA
156
atUniversityofNewMexicoonApril28,2010http://mollus.oxfordjournals.orgDownloadedfrom
66
- Vascular plant compositional gradients within and between Iowa fens - 771
Journal of Vegetation Science 15: 771-780, 2004
© IAVS; Opulus Press Uppsala.
Abstract
Question: What is the nature and relative importance of
compositional gradients within- and between fens?
Location: Iowa, USA.
Methods: 506 0.5 m × 0.5 m quadrats were sampled from 31
fens across a 550 km extent. Presence/absence of all vascular
plant taxa, plus the non-vascular genera Sphagnum and Chara,
and values for 24 environmental variables were noted. Global
Non-Metric Multidimensional Scaling and Monte Carlo tests
were used to describe compositional variation and identify
significant environmental co-variables. Model-based cluster
analysis was used to identify the optimal number of groups
supported by the data, while k-means clustering was used to
assign each quadrat to a group. The number of occurrences
(and frequency) of each species within each group was calculated.
Two-dimensional 95% Gaussian confidence intervals,
ANOVA, correlation coefficient homogeneity tests, log-linear
modelling, and Fisher’s exact tests were used to document
patterns of compositional change.
Results: Two stable axes of variation were identified: the first
being most closely correlated with soil pH, Mg, Ca, P, S,
vegetation height, surface and –10 cm soil temperature, site
area, perimeter, perimeter/area ratio, growing season, and air
temperature, with the second being most correlated to soil
moisture, N, disturbance level, % organic matter, hummock
height, N-S coordinate, and precipitation. Individual sites
harboured between 20-47% of total compositional variation,
with 28% of Axis 1 and 55% of Axis 2 scores being contained
within-sites. Five compositional regions were identified that
differed in the proportion of calciphile and hydrophile species.
Compositional groups differed significantly between geologic
types.
Conclusions: While the principal axis of variation (corresponding
to the rich-poor fen gradient) is present largely
between sites, the second axis (corresponding to water level)
is largely repeated within sites. Documentation and protection
of vegetation patterns and species diversity within Iowa fens
will thus require consideration of multiple sites across the
landscape.
Keywords: Cluster analysis; Community ecology; Gradient
analysis; Landscape pattern; Multidimensional scaling; North
America; Peatland.
Nomenclature: Nekola (1994).
Vascular plant compositional gradients within and between Iowa fens
Nekola, Jeffery C.
Department of Natural and Applied Sciences; University of Wisconsin - Green Bay;
Green Bay, Wisconsin 54311, USA; E-mail nekolaj@uwgb.edu
Introduction
To understand the full range of environmental factors
influencing peatland vegetation, simultaneous documentation
of both within- and between-site gradients
must take place. Unfortunately, previous investigations
have focused either on numerous (> 80) quadrats from
individual sites (van der Valk 1975; Bernard et al. 1983;
Stewart 1987; Singsaas 1989; Gerdol 1990; Glaser et al.
1990; Magnússon & Magnússon 1990; Mooney &
O’Connell 1990; Wassen et al. 1990a; Stoynoff 1993;
McCormac & Schneider 1994; Fojt & Harding 1995;
Bowlesetal.1996;Choesin&Boerner2000;Gunnarsson
et al. 2000; Choesin & Boerner 2002; Southall et al.
2003), or on few (< 10) quadrats sampled from multiple
sites in the landscape (Charman 1993; Johnson &
Leopold 1994; Motzkin 1994; Slack 1994; Bergamini et
al. 2001; Vanderpuye et al. 2002; Stammel et al. 2003).
The following study considers the relative importance
of compositional gradients within- and between-site
fens based on moderate numbers (10-30) of quadrats
sampled from multiple sites (31) across Iowa. These
data will be used to address the following questions: 1.
What are the major compositional gradients? 2. What
environmental factors underlie these trends? 3. How
much compositional variation is captured within vs.
between sites?
Methods
Study region
Fens are peatland habitats with saturated (not inundated)
soils whose source water has been enriched in
nutrients by passage through the ground (Sjörs 1952).
Fewer than 200 fens remain in the northeastern half of
Iowa out of an original ca. 2400 pre-settlement sites
(Nekola 1994). Extant examples are most common on
the Iowan Erosional Surface and the northwestern margin
of the Des Moines Lobe (Fig. 1). The Iowan Erosional
Surface was formed through intense periglacial
erosion during the Wisconsinan, while the Des Moines
Lobe was formed ca. 14000-12000 yr ago during a final
67
772 Nekola, J.C.
surge of Wisconsinan ice sheet following the onset of
climatic warming (Prior 1991).
These fens can be broadly classified into five geological
types, depending upon aquifer type (Thompson
et al. 1992): glacial till, bedrock, eolian sand, fluvial
sand, and former lake basins (pothole or oxbow lakes).
While statistical differences exist in peat chemistry between
these groups (e.g. percent organic matter, pH, P,
Ca, and Mg), no more than 35% of observed variance is
accounted for by them (Nekola 1994).
Study sites
In total 31 fens from 23 counties across a 550 km
extent (Fig. 1) were selected for analysis based on their
natural integrity and geographical location. Investigation
was limited to high-quality, undisturbed sites that
maximized geographic spread across the landscape.
While an attempt was made to maintain at least 40 km
separation, some very high-quality sites were selected
even though they were more closely positioned. Selected
fens represent all five major geologic groups (22
till, 4 basin, 2 bedrock, 2 eolian, and 1 fluvial) and
range in size from 0.1 - 28.8 ha (Table 1).
Field methods
A total of 506 0.5 m × 0.5 m quadrats were sampled
in June and July, 1989. While this quadrat size is
smaller than some previous North American investigations
(i.e. Motzkin 1994), it is consistent with many
other peatland vegetation studies (e.g. Bellamy 1967;
Singaas 1989; Magnússon & Magnússon 1990;
Stoynoff 1993; Bowles et al. 1996). This grain size is
adequate for ordination analysis as it captures an average
of 12 co-occurring species, or ca. 15% of an
individual site flora.
Sites were sampled at two intensity levels. One
group (10 sites) was more intensively sampled at 30
quadrats per site. An effort was made to geographically
stratify these across the landscape. At Mt. Auburn
fen, only 26 quadrats were sampled due to weather
conditions; 10 quadrats per site were recorded from the
second group (21 sites). Only 7 and 5 quadrats were
recorded from Rochester South and Rowley North,
respectively, due to weather conditions.
Each site was divided into five concentric zones of
equal width positioned from the centre to edge. While
zone widths differed between sites, their moisture levels
were similar. Stratified-random sampling was used
to position quadrats within each zone. When this sampling
protocol missed major vegetation types or rare
species, additional quadrats were non-randomly placed
to capture them. Such subjective placement was limited
to less than 10 quadrats (< 2% of total).
All vascular plant species, plus Sphagnum and
Chara spp., were noted from each quadrat. Surface and
subsurface (–10 cm) soil temperature, vegetation height,
and maximum hummock height were also recorded.
Level of disturbance was noted using a subjective
score ranging from 1 (least) to 3 (most). Site locations,
perimeters and areas were calculated through digitization
of USDA Soil Conservation Service county
soil maps.
Soil chemistry
A 100-cm3 soil sample was taken from the centre of
each quadrat. These were transported in water-tight
plastic bags and not allowed to come in contact with
sulfur-rich paper products. Samples were either oven
dried within eight hours of collection, or frozen for
later drying. Following the methodologies of Dahnke
(1988), percent organic matter, pH, NO3-N, extractable
P, exchangeable K, extractable SO4-S, exchangeable
Ca, and exchangeable Mg were determined by
Minnesota Valley Testing Laboratory of Nevada, Iowa.
The percent water mass present in each wet sample
was determined. However, because the water holding
capacity of soils is positively correlated with organic
matter content (Hudson 1994), many waterlogged but
mineral-rich samples had lower water mass percentages
than seemingly drier samples with higher organic
matter levels. To help correct for this, soil moisture
(SM) was calculated for each sample based on the
following:
SM = W * (1 – O) (1)Fig. 1. Landform regions of Iowa with location of sampled fen
sites. Site letter codes correspond to those in Table 1.
68
- Vascular plant compositional gradients within and between Iowa fens - 773
Table 1. List and locations for sampled fen sites, including their size, geologic type, sampled quadrats, and percent compositional
overlap with all other sites.
Size Geologic Quadrats Site
County Site name Location (ha) type sampled % overlap code
Allamakee County Clear Creek 1 91° 25' 28" W., 43° 27' 30" N. 1.3 Fluvial 11 80.4 J
Benton County Mt. Auburn 92° 5' 59" W., 42° 14' 50" N. 28.8 Till 26 22.3 R
Black Hawk County Hammond Road 92° 20' 36" W., 42° 20' 48" N. 1.6 Till 10 23.8 P
Bremer County Brayton-Horsley 92° 6' 27" W., 42° 48' 35" N. 4.7 Till 30 35.7 A
Bushing 92° 27' 1" W., 42° 49' 48" N. 2.0 Till 12 43.5 I
Buchanan County Cutshall Access 92° 4' 17" W, 42° 32' 47" N. 4.0 Basin 10 46.6 C
Rowley 91° 51' 7" W., 42° 22' 27" N. 2.2 Till 30 22.9 D
Rowley North 91° 51' 5" W., 42° 22' 36" N. 0.3 Till 5 70.1 Y
Cedar County Rochester South 91° 8' 12" W., 41° 36' 59" N. 0.3 Eolian 7 23.4 X
Cerro Gordo County Buffalo Slough 93° 11' 54" W., 43° 10' 36" N. 12.6 Basin 30 36.6 O
Neuhring 93° 18' 39" W., 43° 13' 3" N. 1.0 Till 10 45.8 L
Chickasaw County Chickasaw 1 92° 31' 3" W., 43° 0' 33" N. 18.4 Till 10 65.9 K
Kleiss 92° 6' 20" W., 43° 0' 58" N. 1.6 Till 10 28.8 B
Clay County Gillett 95° 0' 52" W., 43° 0' 53" N. 3.1 Till 10 48.6 %
Clayton County Postville 91° 34' 0" W., 43° 2' 3" N. 37.5 Till 10 84.7 M
Clinton County Toronto 2 90° 48' 9" W., 41° 53' 49" N. 1.3 Eolian 30 87.0 V
Delaware County Hawker 91° 32' 38" W., 42° 30' 47" N. 7.3 Till 30 45.6 G
Dickinson County Silver Lake 95° 21' 55" W., 43° 26' 16" N. 2.6 Till 12 84.4 #
Lower Gar Lake 95° 6' 34" W., 43° 20' 48" N. 3.3 Till 10 61.7 $
Emmet County O’Brien 94° 50' 36" W., 43° 26' 16" N. 0.1 Till 10 31.5 *
Fayette County Hunter Creek 91° 56' 50" W., 42° 39' 32" N. 0.9 Till 10 18.3 T
Smithfield Township Hall 91° 47' 18" W., 42° 46' 57" N. 3.5 Till 30 21.2 Q
Franklin County Maynes Creek 93° 1' 49" W., 42° 42' 24" N. 1.0 Bedrock 10 35.1 N
Grundy County New Hartford 92° 38' 22" W., 42° 33' 16" N. 1.4 Till 10 34.7 S
Howard County Staff Creek 92° 30' 34" W., 43° 26' 40" N. 2.4 Till 13 54.4 Z
Linn County Matus 91° 29' 4" W., 42° 7' 57" N. 18.5 Till 30 55.9 E
Western College 91° 37' 2" W., 41° 51' 56" N. 0.6 Till 10 31.9 F
Mitchell County St. Ansgar 92° 55' 27" W, 43° 23' 21" N. 2.0 Bedrock 30 55.2 H
Muscatine County Nichols 91° 16' 54" W., 41° 27' 24" N. 49.3 Basin 30 17.9 W
Palo Alto County Dead Man’s Lake 93° 33' 59" W., 43° 14' 57" N. 1.2 Basin 10 11.3 @
Winneshiek County Jackson Juncton 92° 2' 52" W., 43° 7' 15" N. 1.4 Till 10 46.0 U
where W = percent water mass of wet soil sample; and
O = percent organic mass of dry soil sample. When soil
organic content is low SM remains close in value to the
observed percent water mass. However, as soil organic
matter content increases, SM will fall accordingly.
Meaningful comparisons between sites were made possible
by severe drought conditions that allowed differences
to reflect differential groundwater seepage rates
as opposed to timing of rainfall events.
Water chemistry was not measured for two principal
reasons: First, as standing water was not present in
all quadrats it could have been sampled only through
the digging of pits that will cause significant alterations
to water chemistry (Glaser et al. 1990). Second,
soil chemistry is a much better correlate for vascular
plant composition (Sjörs & Gunnarsson 2002) while
water chemistry principally impacts bryophytes (Vitt
& Chee 1990).
Climate analysis
Principal Components Analysis in conjunction with
a varimax rotation was conducted on 30-yr mean values
for 22 climate variables measured from 96 Iowa
recording stations. Four major gradients, accounting
for over 95% of the total climate variation, were identified:
air temperature (correlated with Average Yearly,
Summer, Fall, and Spring Temperatures, Heat Stress
Days/Year, Heat Stress Degrees/Year, Heating Degree
Days, Cooling Degree Days, Growing Degree Days),
non-summer precipitation (correlated with Average
Yearly, Winter, Spring and Fall Precipitation, Average
Winter Temperature, Maximum 24-hr Precipitation),
growing season (correlated with 32º, 30º, 28º, 26º, and
24º Growing Seasons, number of Freezing Days/Year),
and average summer precipitation (Nekola 1994). PCA
scores along each of these axes were determined for
each recording station, with punctual kriging (Burgess
& Webster 1980) being used to estimate site-specific
values along each axis based upon the nearest 20
stations.
Statistical procedures
Ordination
Species lists from each quadrat were subjected to
global non-metric multidimensional scaling (NMDS)
using the DECODA software program (Minchin 1990).
69
774 Nekola, J.C.
NMDS was used as it is the most robust form of
ordination for detection of ecological pattern (Minchin
1987). Using all species, a dissimilarity matrix was
calculated using the Kulczynski index (Faith et al.
1987) for all pairwise combinations of sites. NMDS in
one through four dimensions was then performed with
200 maximum iterations, a stress ratio stopping value
of 0.999900, and a small stress stopping value of
0.010000. Output was scaled in half-change units.
Because a given NMDS run may locate a local
(rather than the global) stress minimum, multiple
NMDS runs starting from different initial random points
must be compared to determine solution stability
(Minchin 1987). For this ordination, DECODA used a
total of 20 random starting configurations. Solutions in
each of the four dimensions were compared using
Procrustes transformations to identify those that were
statistically distinct. The number of unique solutions,
and runs that fell into each was then calculated across
each of the four dimensions. The modal solution out of
twenty runs was identified and was considered stable
when it was achieved in at least 25% of starts.
Identification of compositional groups
Clustering was performed on the selected ordination
output rather than raw data as the former are less
susceptible to sampling or other inadvertent errors
(Equihua 1990). Model-based cluster analysis (Banfield
& Raftery 1992) was used to identify the number of
groups most supported by the data. A sum-of-squares
model was employed as it generates spherical clusters
that will be of maximal compositional similarity. The
approximate weight of evidence for k clusters (AWEk)
was calculated for k = 1 to n–1 clusters (where n = the
total number of ordinated sites) via the S+ MCLUST
algorithm (Anon. 1995); k-means iterative relocation
(Hartigan 1975) was then used to assign each site to a
cluster as it is also based on sum-of-squares criteria.
Ordination interpretation
The number of occurrences (and frequency) of
each species within each k-means cluster was calculated.
The 20 most frequent taxa, and rare taxa (see
Nekola 1994) reaching modal frequency within each
cluster, were then determined. Two-dimensional 95%
Gaussian confidence intervals (Sokal & Rohlf 1981)
were calculated for each site and used to estimate
compositional overlap between sites. Quadrats from
other sites were considered statistically similar in composition
when they fell within a given site’s 95%
confidence ellipse. Quantification of Axis 1 and Axis 2
stand score variation between sites was accomplished
via ANOVA and represented using box plots. The
significance of differences in the variation accounted
for between sites was estimated using a correlation
coefficient homogeneity test (Sokal & Rohlf 1981).
The maximum correlation vectors for all environmental
variables were calculated by DECODA and
their significance estimated through Monte-Carlo tests.
Variables were assigned into one of two groups: those
that varied within sites (pH, NO3-N, P, K, SO4-S, Ca,
Mg, % organic matter, vegetation height, hummock
height, surface soil temperature, –10 cm soil temperature,
disturbance level, soil moisture), and those that
only varied between sites (N-S Coordinate, E-W Coordinate,
Perimeter, Area, Perimeter/Area Ratio, NonSummer
Precipitation, Summer Precipitation, Growing
Season, Air Temperature).
Significance of differences in frequency of each kmeans
cluster between the five geologic fen types was
estimated via log-linear modelling. As predicted values
were sparse (< 5) in more than one-fifth of cells (Zar
1984), Fisher’s Exact test was used to identify those
compositionalclusterssignificantlyfavouredwithineach
geologic group. Because the total number of quadrats
sampled from glacial till sites exceeded 200, log-linear
modelling was used to estimate significance for this
group. The significance threshold for these analyses
were adjusted to p = 0.01 using a Bonferroni correction.
Results
Site ordination
A total of 217 taxa (including Sphagnum and Chara)
were identified (App. 1). NMDS demonstrated that in
one dimension the most stable solution (minimum stress
configuration = 0.3607) was achieved in 14 starts; in two
dimensions the most stable solution (minimum stress
configuration = 0.2467) was achieved in five starts; in
three dimensions the most stable solution (minimum
stress configuration = 0.1923) was achieved in two starts;
and in four dimensions all starts provided different solutions
(minimum stress configuration = 0.1594). The most
stable two-dimensional solution was chosen for further
analysis because it possessed a relatively low stress level
while also being achieved in at least 25% of starts.
Visual observation of this solution demonstrated
essentially constant variation across both axes (Fig. 2).
One major outlier (from Silver Lake Fen and the lone
quadratlocatedinahypercalcarouswatertrack)ispresent
approximately 0.75 axis units to the right of the general
distribution. This microhabitat is quite rare in Iowa,
being limited to this and two adjacent sites.
70
- Vascular plant compositional gradients within and between Iowa fens - 775
Compositional variation within and between sites
Considerable compositional overlap was noted between
sites, ranging from 11.3% of all other quadrats
(Dead Man’s Lake) to 87% (Toronto 2), with a median
of 43.5% (Table 1). These values are skewed slightly
to the right, with the bulk ranging from 20 - 47%.
ANOVA identified significant (p < 0.0005) partitioning
of both Axis 1 and Axis 2 scores between sites.
However, the amount of variance explained significantly
(p < 0.000005) varies from 72% for Axis 1 to
45% for Axis 2 (Fig. 3).
Identification and description of compositional
clusters
The maximum AWEk score (2426.9) was achieved at
73 compositional clusters. As this represents too many
groups for generalization of compositional trends, AWEk
scores from k = 1 to 15 were calculated, along with the
percent increase in AWEk from k–1 to k clusters. After
six clusters, the percent increase in AWEk fell below
10%, and decreased steadily to the 2.5% range by cluster
15. However, as the sixth k-means cluster contained
only a single outlying quadrat (the Silver Lake fen water
track), it was considered a statistical artifact with low
inferential power. Each quadrat was thus assigned to
one of five clusters.
Cluster A (100 quadrats) is located on the lower left
of the ordination diagram (Fig. 2) and is dominated by
plants such as Campanula aparinoides, Impatiens
capensis, Aster puniceus, Carex prairea, and Lycopus
americanus that demand high moisture levels but can
tolerate non-calcareous conditions (Table 2). In this
cluster 14 rare Iowa fen species reach their modal frequencies
(Table 3) including Aster junciformis, Berula
erecta, Galium labradoricum, Menyanthes trifoliata,
Salix pedicellaris and Solidago uliginosa.
Cluster B (88 quadrats) is located in the upper left of
the ordination diagram (Fig. 2) and is dominated by two
ferns (Thelypteris palustris and Onoclea sensibilis) and
other species such as Polygonatum sagittatum and
Triadenum fraseri that can tolerate somewhat drier,
non-calcareous conditions (Table 2). Four rare Iowa fen
species/genera reach modal frequencies in this cluster
(Table 3) including Drosera rotundifolia, Potentilla
palustris, Rubus pubescens, Sphagnum spp. and
Triadenum fraseri.
Cluster C (85 quadrats) is located in the upper centre
of the ordination diagram (Fig. 2) and is dominated by
typical sedge meadow species such as Carex stricta,
Aster puniceus, Thelypteris palustris, Pycnanthemum
virginianum, and Solidago altissima that require rich,
moderately wet soils (Table 2). Only three rare Iowa fen
plants (Cypripedium parviflorum, Ophioglossum
vulgatum, Viola pallens) reach modal frequencies in
this cluster (Table 3).
Cluster D (169 quadrats) is located to the lower right
of Cluster C (Fig. 2). This cluster demarcates the most
commonly encountered vegetation of Iowa fen mats,
andisdominatedbyAsterpuniceus,Lycopusamericanus,
Muhlenbergia glomerata, Viola nephrophylla and
Lysimachia quadriflora (Table 2). Carex species are
present but do not dominate, with the flora representing
a wide mix of relatively low-statured forbs and
graminoids. Nine rare Iowa fen plants reach modal
frequencies in this cluster (Table 3) including such
characteristicMidwesternUSAfentaxaasBetulapumila,
Carex prairea, Cypripedium candidum, Salix candida,
Solidago riddellii and Valeriana edulis.
ClusterE(64quadrats)islocatedinthecentre-rightof
the ordination diagram (Fig. 2) and is dominated by
species requiring both relatively high moisture levels and
calcareous conditions such as Muhlenbergia glomerata,
Parnassia glauca, Eleocharis elliptica, Rhynchospora
capillacea, and Lobelia kalmii (Table 2). In cluster E 24
rare Iowa fen plants reach their modal frequencies in this
cluster (Table 3) including Carex sterilis, Eleocharis
pauciflora, Juncus alpinus, Platanthera hyperborea,
Rhynchospora capillacea, Scleria triglomerata, Spiranthes
romanzoffiana, Triglochin maritimum, T. palustre
and Utricularia minor.
Fig. 2. NMDS ordination of 506 Iowa fen quadrats. Letters
correspond to the compositional group that each quadrat was
assigned via k-means clustering.
71
776 Nekola, J.C.
Table 3. Rare taxa (as listed in Nekola 1994) reaching modal
frequencies in each compositional cluster.
Cluster A
Angelica atropurpurea
Aster junciformis
Berula erecta
Carex rostrata
Cirsium muticum
Eleocharis smallii
Equisetum fluviatile
Galium labradoricum
Menyanthes trifoliata
Mimulus glabratus
Rumex orbiculatus
Salix pedicellaris
Solidago uliginosa
Cluster B
Drosera rotundifolia
Potentilla palustris
Rubus pubescens
Sphagnum spp.
Triadenum fraseri
Cluster C
Cypripedium parviflorum
Ophioglossum vulgatum
Viola pallens
Cluster D
Betula pumila
Bromus ciliatus
Carex prairea
Cypripedium candidum
Salix candida
Salix × rubella
Salix × clarkei
Solidago riddellii
Valeriana edulis
Cluster E
Carex granularis
Carex sartwellii
Carex tetanica
Carex sterilis
Chara spp.
Eleocharis elliptica
Eleocharis pauciflora
Eriophorum angustifolium
Gentiana crinita
Gentiana procera
Gerardia paupercula
Juncus alpinus
Juncus balticus
Liparis loeselii
Lobelia kalmii
Muhlenbergia glomerata
Parnassia glauca
Platanthera hyperborea
Rhynchospora capillacea
Scleria verticillata
Spiranthes romanzoffiana
Triglochin maritimum
Triglochin palustre
Utricularia minor
Table 4. Contingency table analysis showing the number of
quadrats within each of the five compositional clusters represented
in each of the geologic fen types. Log-linear test
statistic represents significance across entire table.
Cluster Geologic fen type
Till Bedrock Fluvial Basin Eolian
A 41 10 4 37 8
B 40 0 0 37 11
C 62 2 0 5 16
D 144 21 2 1 1
E 51 7 5 0 1
Significance within each geologic type
< 0.00005 0.0017 0.2552 < 0.00005 0.0087
Log-linear model test statistic = 221.03; df = 16; p < 0.0000005.
Analysis of environmental co-variables
All environmental variables except K (p = 0.060)
and Average Summer Precipitation (p = 0.067) demonstrated
significant correlation with the ordination diagram.
Maximumr2 valuesforsignificantvariablesranged
from 0.0246-0.3204. For those varying within sites, pH, -
10 cm soil temperature, Mg, Ca, surface soil temperature,
and S were most positively correlated with Axis 1 scores,
while vegetation height and P were negatively correlated
(Fig. 4). N, disturbance level, % organic matter, and
hummock height were positively correlated with Axis 2
scores, while soil moisture was negatively correlated. Of
the factors varying only between sites, site perimeter/
area ratio was positively correlated with Axis 1 scores,
while site area, air temperature, and site perimeter were
negatively correlated. E-W coordinate location was positively
correlated with Axis 1 but negatively correlated
with Axis 2 scores, while precipitation, growing season,
ClusterA(n=100quadrats)ClusterB(n=88)ClusterC(n=85)ClusterD(n=169)ClusterE(n=64)
SpeciesFreq.SpeciesFreq.SpeciesFreq.SpeciesFreq.SpeciesFreq.
Campanulaaparinoides64.00%Thelypterispalustris93.18%Carexstricta82.35%Asterpuniceus90.30%Muhlenbergiaglomerata87.50%
Impatienscapensis64.00%Impatienscapensis63.64%Asterpuniceus82.35%Lycopusamericanus88.48%Violanephrophylla54.69%
Asterpuniceus61.00%Polygonumsagittatum61.36%Thelypterispalustris80.00%Muhlenbergiaglomerata81.82%Lycopusamericanus51.56%
Carexprairea59.00%Asterpuniceus50.00%Pycnanthemumvirginianum51.77%Violanephrophylla78.79%Parnassiaglauca48.44%
Lycopusamericanus48.00%Onocleasensibilis48.86%Solidagoaltissima50.59%Lysimachiaquadriflora72.73%Carexprairea45.31%
Carexstricta41.00%Triadenumfraseri44.32%Lycopusamericanus45.88%Carexprairea66.67%Eleochariselliptica42.19%
Pileafontana40.00%Carexstricta43.18%Campanulaaparinoides41.18%Carexstricta63.64%Lysimachiaquadriflora39.06%
Muhlenbergiaglomerata39.00%Campanulaaparinoides42.05%Asterumbellatus36.47%Pycnanthemumvirginianum60.61%Scirpusvalidus35.94%
Calthapalustris34.00%Eupatoriummaculatum31.82%Violanephrophylla35.29%Carexinterior44.24%Carexinterior32.81%
Eupatoriummaculatum32.00%Typhalatifolia31.82%Helianthusgrosserratus34.12%Campanulaaparinoides40.00%Rhynchosporacapillacea32.81%
Lysimachiathyrsiflora25.00%Lycopusamericanus26.14%Onocleasensibilis29.41%Thelypterispalustris38.18%Eupatoriumperfoliatum31.25%
Cardaminebulbosa24.00%Scutellariagalericulata23.86%Muhlenbergiaglomerata23.53%Eleochariselliptica35.15%Lobeliakalmii29.69%
Equisetumfluviatile24.00%Pileafontana21.59%Polygonumsagittatum22.35%Eupatoriummaculatum34.55%Eupatoriummaculatum28.12%
Violanephrophylla23.00%Boehmeriacylindrica21.59%Eupatoriummaculatum18.82%Lythrumalatum30.91%Carexhystericina28.12%
Galiumlabradoricum21.00%ImmatureMuhlenbergia20.45%Fragariavirginica17.65%Salixcandida27.88%Asterumbellatus26.56%
Thelypterispalustris19.00%Carexlacustris18.18%Geumaleppicum17.65%Lobeliasiphilitica26.06%Carexstricta26.56%
Epilobiumleptophyllum19.00%Asterumbellatus15.91%Polemoniumreptans16.47%Sphenopholisintermedia24.24%Juncusnodosus26.56%
Leerziaoryzoides19.00%Carexlanuginosa12.50%Poapratensis15.29%Asterumbellatus20.61%Glyceriastriata25.00%
Typhalatifolia18.00%Bidenscoronata12.50%Saxifragapensylvanica15.29%Calthapalustris19.39%Scirpusamericana23.44%
Sphenopholisintermedia17.00%Leerziaoryzoides12.50%Carexinterior14.12%Parnassiaglauca19.39%Equisetumarvense20.31%
Polygonumsagittatum17.00%Calthapalustris14.12%Scleriaverticillata20.31%
Table 2. Most frequent taxa (n = 20 or 21) in each of the five
compositional clusters.
72
- Vascular plant compositional gradients within and between Iowa fens - 777
and N-S coordinate location were negatively correlated
with Axis 1 but positively correlated with Axis 2.
The frequency of the five compositional clusters
was found to significantly (p < 0.0000005) vary between
the five geologic groups (Table 4). Glacial till
sites were over-represented in Cluster D and underrepresented
in Clusters A and B (p < 0.00005). Bedrock
outcrop sites were over-represented in Cluster D and
underrepresented in Clusters B and C (p = 0.0017).
Basin sites were over-represented in Clusters A and B
and underrepresented in Clusters C, D, and E (p <
0.00005). Eolian Sand sites were over-represented in
Clusters B and C and underrepresented in Clusters D
and E (p = 0.0087). No significant patterns (p = 0.2552)
were noted from the lone Fluvial Sand site sampled.
Discussion
Major compositional gradients
Theseanalysesdemonstratetwostrongcompositional
gradients (3 and 2 half-change units) within Iowa fens.
The principal axis varies from vegetation supporting
few to many calciphiles, while the second axis varies
from vegetation supporting few to many hydrophiles.
The compositional clusters created to help interpret
these patterns are not separated in ordination space, with
quadrats from different clusters often being more similar
to each other than they are to other quadrats within
their own group.
The variation along Axis 1 is consistent with the
Fig. 3. Box plot of variation
within each site along NMDS
Axes 1 and 2. Site letter codes
correspond to those in Table 1.
Fig. 4. Environmental variable biplots for the ordination diagram.
73
778 Nekola, J.C.
rich-poor fen gradient noted in most peatland landscapes,
including Estonia (Masing 1982), Poland
(Wassen et al. 1990a), the eastern Carpathians (Hájek
2002), Switzerland (Bergamini 2001), the southern Alps
(Gerdol1990),northwesternEurope(Sjörs1952;Økland
1990; Sjörs & Gunnarsson 2002), Scotland (Charman
1993), boreal Canada (Vitt & Chee 1990), the northern
USA (Glaser 1987; Glaser et al. 1981), northeastern
USA (Motzkin 1994), and New Zealand (McQueen &
Wilson 2000). Similarly, the wet-mesic gradient demonstrated
along the Axis 2 has been previously documented
from Illinois (Bowles et al. 1996), northern
Minnesota (Glaser et al. 1990), Ohio (Stewart 1987),
New York (Bernard et al. 1983), Scotland (Charman
1993), and Norway (Singsaas 1989) fens.
From a conservation perspective, the most important
microsites for rare plants were those that had the highest
soil moisture levels. Fourteen rare Iowa plants reached
their modal frequencies in Cluster A (moderate pH and
cation levels), while 25 rare species reached modal
frequencies in Cluster E (high pH and cation levels).
This pattern is consistent with previous investigations of
Minnesota (Almendinger & Leete 1998a) and Suffolk,
England (Fojt & Harding 1995) fens that have documented
a positive correlation between soil water levels
and rare plant richness.
Environmental co-variables
As for many other fen systems (e.g. Gerdol 1990;
Vitt & Chee 1990; Wassen et al. 1990a; Charman 1993;
Motzkin1994;Fojt&Harding1995;McQueen&Wilson
2000; Bergamini et al. 2001; Sjörs & Gunnarsson 2002)
the principal axis of compositional variation was significantly
correlated with pH, Ca, and Mg. The strong
correlationoftheprincipalaxiswithP,vegetationheight,
and soil temperature is likely due to the limited solubility
of P compounds in areas with high soil/water pH and
Ca (Bedford et al. 1999). This causes reduced primary
production, fertility, and plant height in Minnesota
(Almendinger & Leete 1998a), English (Boyer &
Wheeler 1989), Italian alpine (Gerdol 1990) and Dutch
(Verhoeven et al. 1990) fens. With increased penetration
of solar energy to the ground, areas with less vegetation
growth will also have higher soil temperatures.
As low vegetation mats support the highest plant diversity
and number of rare species in Iowa (twice the
number of rare species occur in Cluster E vs. A) and
other fen systems (Johnson & Leopold 1994; Wheeler &
Shaw 1991; Jensen & Meyer 2001), maintenance of low
available P levels will be important to fen conservation.
Previous research has also documented that soil
moisture, N concentration, and disturbance level will
influence fen vegetation. Water levels are one of the
most commonly documented correlates with peatland
vegetation gradients (Bernard et al. 1983; Stewart 1987;
Økland 1989; Singsaas 1989; Glaser et al. 1990; Wassen
et al. 1990a; Charman 1993; Bowles et al. 1996). Soil N
has been shown to impact vegetation composition in
Swiss montane (Bergamini et al. 2001) and Icelandic
(Magnússon & Magnússon 1990) fens. Disturbance gradients
also underlie some peatland vegetation patterns,
with grazing being more intense on drier soils (Bowles
et al. 1996). In Iceland these environmental factors are
linked as fen sites subjected to higher grazing pressures
alsoexhibitthehighestsoilNconcentrations(Magnússon
& Magnússon 1990).
Climate and geology may also significantly impact
peatland ecology as increased precipitation and temperature
lead to increased leaching rates and decreased
soil Ca levels in boreal North America (Glaser 1992),
Fennoscandia (Økland 1989), and Minnesota (Almendinger
& Leete 1998a) fens. Similarly, in Iowa the
frequency of calcareous microsites increase towards the
northwest where precipitation and temperature are the
lowest. A linkage between fen soil and aquifer chemistry
has been previously demonstrated in Minnesota
(Almendinger & Leete 1998b). Not surprisingly, Iowa
sites originating from the most base-rich aquifers (e.g.
bedrock, till) supported the richest fen vegetation. However,
the linkage between site configuration (area,
perimiter/area ratio) and vegetation has not been previously
documented. While these could be related to edgeeffects,
they are more likely related to differential landscape
positions for large (basins) vs. small (sideslope)
sites, increasing the contribution of cation-rich groundwater
in the latter.
Within- vs. between-site compositional gradients
While most individual sites capture between 25-
50% of total observed compositional variation, this
overlap was less evident along Axis 1 as compared to
Axis 2. This result is to be expected as Axis 1 environmental
co-variables are most strongly related to soil pH,
Ca, Mg, and P, which will vary most between sites due
to differences in aquifer chemistry (Almendinger &
Leete 1998b), while Axis 2 scores are most strongly
related to differences in soil water levels, which are
largely repeated from the dry margins (lags) to wet
centres of each site (Wassen et al. 1990b; Wassen &
Barendregt 1992; Gerdol 1993). Thus, water levels vary
more within vs. between New York fens (Slack 1994),
while in nutrient levels vary more between vs. within
Dutch fens (Verhoeven et al. 1990).
This differential within- and between-site variation
along both axes has important consequences to the study
and conservation of Iowa fens. To observe (and protect)
74
- Vascular plant compositional gradients within and between Iowa fens - 779
theprincipalcompositionalgradientfrompoor-richsites,
multiple locations within the landscape must be considered.
When only single sites are investigated or conserved,
gradients associated with water levels gain importance.
It is thus not surprising that as previous quantitative
vegetation analysis of Midwestern fens have
been limited to single or few sites (e.g. van der Valk
1975; Stewart 1987; Wilcox et al. 1986; Stoynoff 1993;
McCormac & Schneider 1994; Bowles et al. 1996;
Choesin & Boerner 2000) the poor-rich fen gradient has
not been previously reported from this landscape. How-
ever,thesesingle-sitestudieshavegenerallydocumented
the importance of water levels in determining vegetation
pattern. Conservation prioritization simply based
upon the number of rare species per site will likely lead
to an over-representation of calcareous sites, as these
areas support the greatest number of rare species. The
rare species characteristic of poor and intermediate fens,
though fewer in number, may thus be overlooked.
Acknowledgements. Gretchen and the late John Brayton
suggested that I study diversity patterns within eastern Iowa
peat ‘bogs’, and are in large part responsible for my interest in
these sites. Bob Moats played a critical role by introducing me
to northwestern Iowa fens. Dick Baker, Willard Hawker, John
Nehnevaj, Dean Roosa, Arlan Kirchman, Dennis Schlicht and
Dave Wendling, provided me with locality information. Fred
Nekola provided invaluable assistance in field sampling. Herb
Wilson furnished soil survey maps for all counties within the
study region. Richard Carlson of the Department of Agronomy
at Iowa State University kindly provided climatic data for
Iowa. Helge Bruelheide, Paddy Coker, Jim Graves, Robert
Peet, Martin Wassen, Thomas Wentworth, Peter White and
Susan Wiser made useful comments regarding earlier drafts.
The substantial financial support needed for completion of soil
analyses was provided by the McElroy Foundation of Waterloo,
Iowa. Field work was also funded in part through a
National Science Foundation Graduate Fellowship.
References
Anon. 1995. S-PLUS guide to statistical and mathematical
analysis. Version 3.3. Statistical Sciences, Seattle, WA,
US.
Almendinger, J.E. & Leete, J.H. 1998a. Peat characteristics
and groundwater geochemistry of calcareous fens in the
Minnesota River basin, U.S.A. Biogeochemistry 43: 17-
41.
Almendinger, J.E. & Leete, J.H. 1998b. Regional and local
hydrology of calcareous fens in the Minnesota River basin,
USA. Wetlands 18: 184-202.
Banfield, J.D. & Raftery, A.E. 1992. Model-based Gaussian
and non-Gaussian clustering. Biometrics 49: 803-822.
Bedford, B.L, Walbridge, M.R. & Aldous, A. 1999. Patterns
in nutrient availability and plant diversity of temperate
North American wetlands. Ecology 80: 2151-2169.
Bellamy, D.J. 1967. Ecological studies on some European
mires. Ph.D. Dissertation, University of London. London,
UK.
Bergamini, A., Peintinger, M., Schmid, B. & Urmi, E. 2001.
Effects of management and altitude on bryophyte species
diversity and composition in montane calcareous fens.
Flora 196: 180-193.
Bernard, J.M., Sieschab, F.K. & Gauch, H.G. Jr. 1983. Gradient
analysis of the Byron-Bergen swamp, a rich fen in
western New York. Vegetatio 53: 85-91.
Bowles, M., McBride, J., Stoynoff, N. & Johnson, K. 1996.
Temporal changes in vegetation composition and structure
in a fire-managed prairie fen. Nat. Areas J. 16(4):
275-288.
Boyer, M.L.H. & Wheeler, B.D. 1989. Vegetation patterns in
spring-fen calcareous fens: calcite precipitation and constraints
on fertility. J. Ecol. 77: 597-609.
Burgess, T.M. & Webster, R. 1980. Optimal interpolation and
isarithmic mapping of soil properties. II. Block kriging. J.
Soil Sci. 31: 333-341.
Charman, D.J. 1993. Patterned fens in Scotland: evidence
from vegetation and water chemistry. J. Veg. Sci. 4: 543-
552.
Choesin, D.V. & Boerner, R.E.J. 2000. Vegetation and ground
water alkalinity of Betsch Fen, a remnant preiglacial fen in
south central Ohio. Castanea 65: 193-206.
Choesin, D.V. & Boerner, R.E.J. 2002. Vegetation boundary
detection: a comparison of two approaches applied to field
data. Plant Ecol. 158: 85-96.
Dahnke, W.C. 1988. Recommended chemical soil test procedures
for the North-central Region. North Dakota Agricultural
Experiment Station, Bull. No. 499, North Dakota
State University, Fargo, ND, US.
Equihua, M. 1990. Fuzzy clustering of ecological data. J.
Ecol. 78: 519-534.
Faith, D.P., Minchin, P.R. & Belbin, L. 1987. Compositional
dissimilarity as a robust measure of ecological distance.
Vegetatio 69: 57-68.
Fojt, W. & Harding, M. 1995. Thirty years of change in the
vegetation communities of three valley mires in Suffolk,
England. J. Appl. Ecol. 32: 561-577.
Gerdol, R. 1990. Vegetation patterns and nutrient status of two
mixed mires in the southern Alps. J. Veg. Sci. 1: 663-668.
Gerdol, R. 1993. The vegetation of wetlands in the southern
Carnian alps (Italy). Gortania Atti del Museo Friuland di
Storia Naturale 15: 67-107.
Glaser, P.H. 1987. The ecology of patterned boreal peatlands
of northern Minnesota: a community profile. USDI Fish
and Wildlife Service, Biological Report 85 (7.14), Washington,
DC, US.
Glaser, P.H. 1992. Raised bogs in eastern North America –
regional controls for species richness and floristic assemblages.
J. Ecol. 80: 535-554.
Glaser, P.H., Wheeler, G.A., Gorham, E. & Wright, H.E.
1981. The patterned mires of the Red Lake peatland,
northern Minnesota: vegetation, water chemistry, and
landforms. J. Ecol. 69: 575-599.
Glaser, P.H., Janssens, J.A. & Siegel, D.I. 1990. The response
of vegetation to chemical and hydrological gradients in
75
780 Nekola, J.C.
the Lost River peatland, northern Minnesota. J. Ecol. 78:
1021-1048.
Gunnarsson, U., Rydin, H. & Sjörs, H. 2000. Diversity and pH
changes after 50 years on the boreal mire Skattlösbergs
Stormosse, central Sweden. J. Veg. Sci. 11: 277-286.
Hájek, M. 2002. The class Scheuchzerio-Caricetea fuscae in
the western Carpathians: indirect gradient analysis, species
groups and their relation to phytosociological classification.
Biologia (Bratislava) 57: 461-469.
Hartigan, J.A. 1975. Clustering algorithms. John Wiley Press,
New York, NY, US.
Hudson, B. 1994. Soil organic matter and available water
capacity. J. Soil Water Cons. 49: 189-193.
Jensen, K. & Meyer, C. 2001. Effects of light competition and
litter on the performance of Viola palustris and on species
composition and diversity of an abandoned fen meadow.
Plant Ecol. 155: 169-181.
Johnson, A.M. & Leopold, D.J. 1994. Vascular plant species
richness and rarity across a minerotrophic gradient in the
wetlands of St. Lawrence County, New York, USA.
Biodivers. Conserv. 3: 606-627.
Magnússon, B. & Magnússon, S.H. 1990. Studies in the grazing
of a drained lowland fen in Iceland. I. The responses of
the vegetation to livestock grazing. Búvísindi 4: 87-108.
Masing, V. 1982. The plant cover of Estonian bogs: a structural
analysis. in: Masing, V. (ed.) Peatland ecosystems,
pp. 50-92. Academy of Sciences of the Estonian
S.S.R.Valgus, Tallinn, EE.
McCormac, J.S. & Schneider, G.J. 1994. Floristic diversity of
a disturbed western Ohio fen. Rhodora 96: 327-353.
McQueen, A.A.M. & Wilson, J.B. 2000. Vegetation and environment
of a New Zealand raised bog. J. Veg. Sci. 11: 547-
554.
Minchin, P.R. 1987. An evaluation of the relative robustness
of techniques for ecological ordination. Vegetatio 69: 89-
108.
Minchin, P.R. 1990. DECODA software. Australian National
University, Canberra, AU.
Mooney, E.P. & O’Connell, M. 1990. The phytosociology and
ecology of the aquatic and wetland plant communities of
the lower Corrib basin, County Galway. Proc. R. Irish
Acad. 90b: 57-97.
Motzkin, G. 1994. Calcareous fens of western New England
and adjacent New York state. Rhodora 96: 44-68.
Nekola, J.C. 1994. The environment and vascular flora of
northeastern Iowa fen communities. Rhodora 96: 121-
169.
Økland, R.H. 1990. Regional variation in SE Fennoscandian
mire vegetation. Nord. J. Bot. 10: 285-310.
Prior, J.C. 1991. Landforms of Iowa. University of Iowa Press,
Iowa City, IA, US.
Singsaas, S. 1989. Classification and ordination of the mire
vegetation of Stormyra near Tynset, S Norway. Nord. J.
Bot. 9: 413-423.
Sjörs, H. 1952. On the relation between vegetation and electrolytes
in North Swedish mire waters. Oikos 2: 241-258.
Sjörs, H. & Gunnarsson, U. 2002. Calcium and pH in north
and central Swedish mires. J. Ecol. 90: 650-657.
Slack, N.G. 1994. Can one tell the mire type from the
Bryophytes alone? J. Hattori Bot. Lab. 75: 149-159.
Sokal, R.R. & Rohlf, F.J. 1981. Biometry. W.H. Freeman and
Co., San Francisco, CA, US.
Southall, E.J., Dale, M.P. & Kent, M. 2003. Floristic variation
and willow carr development within a southwest England
wetland. Appl. Veg. Sci. 6: 63-72.
Stammel, B., Kiehl, K. & Pfadenhauer, J. 2003. Alternative
management on fens: Response of vegetation to grazing
and mowing. Appl. Veg. Sci. 6: 245-254.
Stewart, D.J. 1987. The vegetation and flora of Gott fen
(Ohio), with a comparison of community classifications.
M.Sc. Thesis, Miami University, Oxford, OH, US.
Stoynoff, N.A. 1993. A quantitative analysis of the vegetation
of Bluff Spring Fen Nature Preserve. Trans. Illinois Acad.
Sci. 86: 93-110.
Thompson, C.A., Bettis, E.A. III & Baker, R.G. 1992. Geology
of Iowa fens. J. Iowa Acad. Sci. 99: 53-59.
van der Valk, A.G. 1975. Floristic composition and structure
of a fen in northwest Iowa. Proc. Iowa Acad. Sci. 82: 113-
118.
Vanderpuye, A.W., Elvebakk, A. & Nilsen, L. 2002. Plant
communities along environmental gradients of high-arctic
mires in Sassendalen, Svalbard. J. Veg. Sci. 13: 875-884.
Verhoeven, J.T.A., Maltby, E. & Schmitz, M.B. 1990. Nitrogen
and Phosphorus mineralization in fens and bogs. J.
Ecol. 78: 713-726.
Vitt, D.H. & Chee, W.L. 1990. The relationships of vegetation
to surface water chemistry and peat chemistry in fens of
Alberta, Canada. Vegetatio 89: 87-106.
Wassen, M.J. & Barendregt, A. 1992. Topographic position
and water chemistry of fens in a Dutch river plain. J. Veg.
Sci. 3: 447-456.
Wassen, M.J., Barendregt, A., Palczynski, A., de Smidt, J.T.
& de Mars, H. 1990a. The relationship between fen vegetation
gradients, groundwater flow and flooding in an
undrained valley mire at Biebrza, Poland. J. Ecol. 78:
1106-1122.
Wassen, M.J., Barendregt, A., Schot, P.P. & Beltman, B.
1990b. Dependency of local mesotrophic fens on a regional
groundwater flow system in a poldered river plain
in the Netherlands. Landscape Ecol. 5: 21-38.
Wheeler, B.D. & Shaw, S.C. 1991. Above-ground crop mass
and species richness of the principal types of herbaceous
rich-fen vegetation of lowland England and Wales. J.
Ecol. 79: 285-301.
Wilcox, D.A., Shedlock, R.J. & Hendrickson, W.H. 1986.
Hydrology, water chemistry, and ecological relations in
the raised mound of Cowles Bog. J. Ecol. 74: 103-117.
Zar, J.H. 1984. Biostatistical analysis. Prentice-Hall, Inc.,
Englewood Cliffs, NJ, US.
Received 17 November 2003;
Accepted 16 August 2004.
Co-ordinating Editor: H. Bruelheide.
For App. 1, see JVS/AVS Electronic Archives;
www.opuluspress.se/pub/archives/index.htm
76
© 2003 Blackwell Publishing Ltd. http://www.blackwellpublishing.com/journals/ddi 55
BIODIVERSITY RESEARCH
Diversity and Distributions (2003) 9, 55–71
Blackwell Science, Ltd
Large-scale terrestrial gastropod community composition
patterns in the Great Lakes region of North America
JEFFREY C. NEKOLA Department of Natural and Applied Sciences, University of Wisconsin —
Green Bay, Green Bay, Wisconsin 54311, U.S.A., E-mail: nekolaj@uwgb.edu
Abstract: Previous ordination studies of land
snail community composition have been limited
to four or fewer habitat types from sites separated
by no more than 300 km. To investigate the
nature of large-scale patterns, North American
land snail assemblages at 421 sites, representing
26 habitat types and covering a 1400 × 800 km
area, were ordinated using global, nonmetric
multi-dimensional scaling (NMDS). These data
were then subjected to model-based cluster
analysis and kmeans clustering to identify the
main compositional groups and most important
environmental covariables. Six primary compositional
groups were identified. Three of these
largely represent upland forest and rock outcrop
sites, while the remaining largely represent
either lowland forest, lowland grassland or
upland grassland habitats. The geographical
location and moisture level of sites also influences
community composition. A strong compositional
difference exists between sites having
duff vs. turf soil surface layers. Only 8% of sites
were improperly classified when soil surface
architecture was used as the sole predictor
variable. Fully 43% of taxa exhibited significant
preferences towards one of these surface types,
while only 15% of relatively common (10 +
occurrence) taxa showed no preferences. Twelve
groups of closely related taxa within the same
genus had members that favoured different surface
types, indicating that differential selection
pressures have existed over evolutionary time
scales. While turf faunas appeared unaffected
by anthropogenic disturbance, duff faunas were
strongly impacted, suggesting that their conservation
will require protection of soil surface
architecture.
Key words. Cluster analysis, community ecology,
conservation biology, landscape pattern, multidimensional
scaling, North America, soil architecture,
terrestrial gastropods.
INTRODUCTION
Land snails are regarded typically as generalist
herbivores, fungivores and detritivores (Burch &
Pearce, 1990) that exhibit weak levels of intraspecific
competition (Cain, 1983; Cowie & Jones,
1987; Smallridge & Kirby, 1988; Barker & Mayhill,
1999). As land snail communities can consume
less than 0.5% of annual litter input per year
(Mason, 1970), some speculate that few resources,
beyond CaCO3 (Boycott, 1934) and appropriate
resting site availability (Pearce, 1997), will limit
distribution. This concept is supported by high
levels of microsympatry in land snail communities,
where 13–35 (representing up to 50% of the
regional fauna) co-occurring taxa can be found
at < 1 m2
grains (Schmid, 1966; Cameron &
Morgan-Huws, 1975; Nekola & Smith, 1999;
Cameron, 2002).
However, at landscape scales, land snail population
size and faunistic composition has been
suggested to vary with habitat and vegetation
types (e.g. Burch, 1956; Wäreborn, 1970; Van Es
& Boag, 1981; Young & Evans, 1991; Stamol,
1991; Stamol, 1993; Ports, 1996; Theler, 1997;
Nekola, 2002). Habitat preferences for individual
species have also been discussed (without
supporting empirical data) at subcontinental scales
77
56 J. C. Nekola
© 2003 Blackwell Publishing Ltd, Diversity and Distributions, 9, 55–71
in both western Europe (Kerney & Cameron, 1979)
and eastern North America (Hubricht, 1985).
Ordination techniques have documented significant
species turnover along local environmental
gradients in western Europe (e.g. Tattersfield,
1990; Magnin et al., 1995; Hermida et al., 2000;
Ondina & Mato, 2001) and Pacific Island (e.g.
Cowie et al., 1995; Barker & Mayhill, 1999) faunas.
Unfortunately, these (and all other published
snail ordination) studies have been conducted
only at limited ecological (< four sampled habitat
types) and geographical (maximum separation of
no more than 300 km) scales.
The following study addresses these concerns
by using global nonmetric multi-dimensional
scaling (NMDS) ordination and model-based
cluster analysis to analyse land snail composition
patterns within 26 habitat types across a 1400-km
extent of central North America. These data will
be used to address: (1) if land snail community
composition predictably varies across this subcontinental
region; and (2) what environmental
and geographical factors underlie any such
patterns. This study represents not only the first
use of NMDS in the analysis of land snail
communities, but also represents the first time
North American faunas have been subjected to
ordination.
METHODS
Study region
Land snail faunas were sampled across a
1400 × 800 km area centred on the western portion
of the Great Lakes basin in eastern and
central North America (Fig. 1). This area covers
a wide range of bedrock, climate and vegetation
types. Both Palaeozoic sedimentary and Precambrian
igneous bedrock outcrops in the region.
One of the more prominent sedimentary exposures
is the Niagaran Escarpment, a band of
Silurian limestones and dolomites extending from
western New York state to north-eastern Iowa.
Outcrops along the western Lake Superior shore
typically represent late-Precambrian mafic igneous
rocks associated with the Keewenawan midcontinental
rift system (Anderson, 1983). Average
Fig. 1 Map of the study region showing the location of the 443 sample sites.
78
Great Lakes land snail community patterns 57
© 2003 Blackwell Publishing Ltd, Diversity and Distributions, 9, 55–71
minimum winter temperatures range from −25 °C
in northern Minnesota to −10 °C in western New
York State. Average maximum summer temperatures
range from 25 °C along the western Lake
Superior shore to 30 °C in south-eastern Iowa.
The average length of the 0 °C growing season
varies from 100 to 110 days in northern Minnesota
and Michigan to 180–190 days in southern
Iowa, southern Ontario, and western New York
state. In areas adjacent to or downwind from
(east of ) the Great Lakes (especially the Lower
Peninsula of Michigan, southern Ontario, and
western New York State), the climate tends to be
buffered over that normally experienced in the
continental interior, being warmer in the winter,
cooler in the summer and having a longer
growing season with more constant precipitation
(Eichenlaub, 1979). Matrix vegetation varies
from tallgrass prairie in the west to deciduous
forest in the east to mixed boreal–hardwood forest
in the north (Barbour & Billings, 1988).
Study sites
Four hundred and forty-three sites (Fig. 1) were
surveyed across the range of habitats known to
support diverse land snail assemblages (Nekola,
1999). The 26 habitats surveyed were broadly
grouped into five major categories: rock outcrops,
upland forests, lowland forests, upland
grasslands and lowland grasslands (Table 1).
While sites generally represent undisturbed examples
of their respective habitats, an effort was
also made to sample some (25 rock outcrop, 12
Table 1 Distribution of samples among surveyed habitat types. Habitat descriptions can be found in
Nekola (1999)
Group Habitat type Sites sampled Geographic range
Rock outcrop Carbonate cliff 129 Illinois, Iowa, Minnesota, Michigan,
Ontario, New York, Wisconsin
Lakeshore carbonate ledge 23 Michigan, Ontario, Wisconsin
Algific talus slope 27 Illinois, Iowa
Igneous cliff 72 Michigan, Minnesota
Sandstone/quartzite cliff 5 Wisconsin
Shale cliff 3 New York
Upland forest Oak–hickory forest 2 Wisconsin
Maple–basswood forest 3 Wisconsin
Hemlock–birch forest 1 Wisconsin
Lakeshore forest 16 Michigan, Wisconsin
Rocky woodland 26 Iowa, Michigan, Ontario, Wisconsin
Lowland forest Floodplain forest 2 Wisconsin
Black ash swamp 6 Wisconsin
Tamarack swamp 33 Minnesota, Michigan, Ontario, Wisconsin
White cedar swamp 16 Michigan, Ontario, Wisconsin
Shrub-carr 2 Wisconsin
Upland grassland Tallgrass prairie 1 Iowa
Sand dune 1 Wisconsin
Bedrock glade 13 Iowa
Alvar 6 Michigan, Wisconsin
Igneous shoreline 4 Michigan
Successional old field 4 Wisconsin
Lowland grassland Sedge meadow 5 Michigan, Wisconsin
Fen 29 Iowa, Michigan, New York, Wisconsin
Calcareous meadow 7 Michigan, Wisconsin
Cobble beach 7 Michigan, Wisconsin
79
58 J. C. Nekola
© 2003 Blackwell Publishing Ltd, Diversity and Distributions, 9, 55–71
upland forest, nine lowland forest, four upland
grassland and eight lowland grassland) that had
been disturbed anthropogenically by grazing, logging,
recreational/urban development or bedrock/
soil removal. Examples of such sites include fieldedge
stone piles, abandoned agricultural fields,
abandoned building foundations, old quarries,
pastures, road verges and exploited forests.
Field methods
Documentation of terrestrial gastropods from
each site was accomplished by hand collection of
larger shells and litter sampling for smaller taxa
from representative 100–1000 m2
areas. Soil litter
sampling was primarily used as it provides the
most complete assessment of site faunas (Oggier
et al., 1998). As suggested by Emberton et al.
(1996), collections were made at places of high
micromollusc density, with a constant volume
of soil litter (approximately 4 L) being gathered
from each site. For woodland sites, sampling was
concentrated: (1) along the base of rocks or trees;
(2) on soil covered bedrock ledges; and/or (3) at
other places found to have an abundance of
shells. For grassland sites, samples consisted of:
(1) small blocks (c. 125 cm3
) of turf; (2) loose soil
and leaf litter accumulations under or adjacent
to shrubs, cobbles, boulders and/or hummocks;
and (3) other locations observed to have an
abundance of shells.
The latitude–longitude location of each sample
was determined using either USGS 7.5 minute
topographic maps or a hand-held GPS. To minimize
bias from use of polar-coordinates, these locations
were converted subsequently to Cartesian UTM
Zone 16 coordinates using .
The presence or absence of anthropogenic
disturbance and soil surface architecture (duff vs.
turf) was also recorded from each site. For purposes
of this study ‘duff’ soils represent sites
where the organic horizon was deep (> 4 cm) and
subtended by a friable upper A horizon consisting
largely of humus and mineral soil. ‘Turf’ soils
represent sites where the organic horizon is thin
(< 4 cm) and immediately subtended by an upper
A horizon firmly bound together by living plant
roots. While many habitats only supported a
single soil architecture type (e.g. all carbonate
cliffs were duff, and all bedrock glades were turf ),
some (such as white cedar swamps) could possess
either turf or duff surface layers, depending upon
individual site conditions. Thus, habitat type
could not be used as a surrogate for soil surface
architecture.
Laboratory procedures
Samples were dried slowly and completely in
either a low-temperature soil oven (c. 80–95 °C)
or in a greenhouse. Dried samples were then
soaked in water for 3–24 h, and subjected to
careful but vigorous water disaggregation through
a standard sieve series (ASTME 3/8′ (9.5 mm),
#10 (2.0 mm), #20 (0.85), and #40 (0.425
mm) mesh screens). These fractions were then
dried and passed again through the same sieve
series, and hand-picked against a neutral-brown
background. All shells and shell fragments were
removed.
All identifiable shells from each site were
assigned to species (or subspecies) using the
author’s reference collection and the Hubricht
Collection at the Field Museum of Natural History
(FMNH). Identification of some additional
specimens representing Holarctic taxa more common
in western Europe were verified by Robert
Cameron of the University of Sheffield, UK. All
specimens have been catalogued and are housed
in the author’s reference collection at the University
of Wisconsin — Green Bay. Nomenclature
generally follows that of Hubricht (1985), with
updates and corrections by Frest (1990, 1991)
and Nekola (2002).
Statistical procedures
Ordination
Species lists were determined for each sample.
Sites with four or fewer taxa were excluded from
further analysis, as such samples can bias results
and obscure compositional trends. The remaining
sites were subjected to global nonmetric multidimensional
scaling (NMDS) using 
(Minchin, 1990). NMDS was used as it makes no
assumptions regarding the underlying nature of
species distributions along compositional gradients.
As such, NMDS is the most robust form of
ordination for detection of ecological patterns
(Minchin, 1987).
To ordinate sites, a matrix of dissimilarity
coefficients was calculated between all pairwise
80
Great Lakes land snail community patterns 59
© 2003 Blackwell Publishing Ltd, Diversity and Distributions, 9, 55–71
combinations of sites using the Czekanowski
(Bray–Curtis) index (Faith et al., 1987). All species
(including the most rarely encountered) were
considered. NMDS in one to four dimensions
was then performed, with 200 maximum iterations,
a stress ratio stopping value of 0.9999, and
a small stress stopping value of 0.01. Output was
scaled in half-change units, so that an interpoint
distance of 1.0 will correspond, on average, to a
50% turnover in species composition.
Because a given NMDS run may locate a local
(rather than the global) stress minimum, multiple
NMDS runs must be conducted on a given set of
data from different initial random starting points
to assess the stability of an individual solution
(Minchin, 1987). For this ordination, 
used a total of 20 random starting configurations.
Solutions in each of the four dimensions
were compared using a Procrustes transformation
to identify those that were statistically identical.
The number of unique solutions, and
number of runs which fell into each, was then
calculated across each of the four dimensions
(Minchin, 1990). The modal solution of 20 runs was
identified, and was considered a global optimum
when it was achieved in at least 50% of starts.
Identification of compositional groups
The chosen optimal NMDS solution was then
subjected to model-based cluster analysis
(Banfield & Raftery, 1992) to identify the number
of compositional groups most supported by the
data. Clustering was performed on the selected
ordination output, rather than raw data, as
ordination results are more robust and less susceptible
to sampling or other inadvertent errors
(Equihua, 1990). A sum-of-squares model was
chosen, as it assumes that clusters will be spherical
in ordination space, making them maximally
compact and similar in composition. The approximate
weight of evidence for k clusters (AWEk)
was calculated via the S + MCLUST algorithm
(Statistical Sciences, 1995) for k = 1 to n − 1 clusters
(where n = the total number of ordinated
sites). The larger the AWEk, the more evidence
exists for that number of clusters. After the optimum
number of clusters was determined, kmeans
iterative relocation (Hartigan, 1975) was used to
assign each site to a cluster. Kmeans clustering
was chosen as it operates under the same sum-ofsquares
criteria used for AWEk calculation.
Ordination interpretation
The number of occurrences (and frequency) of
each species within each kmeans cluster was
calculated. Species frequencies between clusters
were compared using a Spearman’s rank correlation.
The 10 most frequent taxa, taxa reaching
modal frequency and species richness for each
cluster were calculated.
The frequency of the five major habitat groups
between the compositional clusters was analysed
using a contingency table. As predicted values
were sparse (< 5) in more than one-fifth of cells,
log-linear modelling was used to estimate significance
(Zar, 1984).
The maximum correlation vectors for the four
recorded environmental variables (UTM E coordinate,
UTM N coordinate, soil surface type, presence of
anthropogenic disturbance) was calculated by .
The significance of each was estimated through
Monte-Carlo simulations using 1000 replications.
Discriminant analysis was used to help further
describe the impact of soil surface type and
anthropogenic disturbance on site position in
ordination space. Three tests were conducted: (1)
effect of soil surface architecture (duff vs. turf)
and the effect of anthropogenic disturbance
separately for (2) duff and (3) turf soils.
Lastly, the number of duff and turf sites containing
and lacking each species was calculated.
The significance of observed differences in these
ratios between duff and turf sites was estimated
using log-linear modelling. As this test was repeated
for each species, a Bonferroni correction was used
to adjust the significance threshold. This conservative
adjustment was used so that only the
most robust deviations would be used for data
interpretation.
RESULTS
Site ordination
One hundred and eight terrestrial gastropod taxa
were identified from the 443 inventoried sites
(Appendix I). Twenty-two sites were species poor
(four or fewer taxa) and removed from further
analysis. These included eight igneous cliffs, three
lakeshore forests, three tamarack wetlands and
single shale cliff, oak–hickory forest, maple–
basswood forest, floodplain forest, sand dune, old
field, sedge meadow and cobble beach sites.
81
60 J. C. Nekola
© 2003 Blackwell Publishing Ltd, Diversity and Distributions, 9, 55–71
NMDS of the remaining 421 sites demonstrated
that the only stable solution occurred
along two axes of variation, where one was
achieved in 50% of starts. The minimum stress
configuration of this solution was 0.197412.
In other dimensions, the most stable solution(s)
were achieved in five (one dimension), three
(three dimensions) and one (four dimensions)
runs (Table 2).
Identification and description of
compositional clusters
Visual observation of the chosen NMDS ordination
solution demonstrated apparent natural
clustering in faunal composition, with at least
one well-defined group existing in the lowercentre
of the diagram (Fig. 2). AWEk analysis
demonstrated that the maximum score (2131.5)
was achieved at the 53rd cluster. As this result
provides too many groups to be useful for generalization
of faunistic trends, AWEk scores from k =
1–10 were calculated, along with the percentage
increase in AWEk from k to k + 1 clusters
(Table 3). These data demonstrate that over 50%
of maximum AWEk was achieved by the 6th cluster.
The percentage increase in AWEk fell by
almost 50% for cluster 7 (7.3%), and decreased
steadily to the 3.4% range by cluster 10. Based
on this, the optimal number of clusters was set at
six (Fig. 3), even though it does not represent
maximum AWEk.
Contingency table analysis (Table 4) demonstrates
that habitat representation significantly
(P < 0.00005) varies between the six nonoverlapping
kmeans clusters. Clusters A–C were equally
(P = 0.3778) represented by rock outcrop and
upland forest sites, while cluster D was dominated
by lowland forests, cluster E by lowland
grasslands and cluster F by upland grasslands.
The 10 most frequent taxa also varied greatly,
Table 2 NMDS summary statistics from an ordination of 421 sites with five or more taxa
Dimensions Stress level
Runs achieving
minimum stress Unique solutions
Number of runs
in modal solution
1 0.339103 20 11 5
2 0.197412 20 6 10
3 0.147405 18 16 3
4 0.116912 18 18 1
Fig. 2 NMDS ordination of 421 sites with land
snail richness of 5 or more, showing distribution of
the five main habitat groups. Units are scales in halfchange
units, such that a distance of 1 represents a
50% turnover in fauna composition.
Table 3 Approximate weight of evidence for k
clusters (AWEk) in land snail ordination based on a
sum-of-squares model
No. of
clusters AWEk
% Change from
AWEk to AWEk−1
1 0 —
2 298.3 —
3 767.8 157.4
4 919.4 19.7
5 1097.8 19.4
6 1244.0 13.3
7 1335.2 7.3
8 1411.4 5.7
9 1456.7 3.2
10 1508.9 3.6
53 2131.5 —
82
Great Lakes land snail community patterns 61
© 2003 Blackwell Publishing Ltd, Diversity and Distributions, 9, 55–71
with approximately 50% turnover occurring
between even the most similar groups (Table 5).
Spearman’s rank correlations of species occurrence
frequencies indicated that clusters D and E
were most similar (0.831), while clusters A and F
were the most different (0.316). Six species
possessed modal occurrence frequencies in cluster
A, 25 in cluster B, 32 in cluster C, five in cluster
D, 20 in cluster E and 20 in cluster F (Table 6).
Forty-one total taxa were encountered in cluster
A, 78 in cluster B, 87 in cluster C, 58 in cluster D,
60 in cluster E and 60 in cluster F (Appendix I).
Analysis of environmental co-variables
Monte Carlo testing of the maximum correlation
vectors for the four recorded environmental variables
(Fig. 4) demonstrated that all were highly
significant (P < 0.0005), having maximum rFig.
3 NMDS ordination of 421 sites with land
snail richness of five or more. Letters represent each
of the six compositional clusters assigned via
Kmeans Clustering.
Table 4 Contingency table analysis of main habitat groups vs. compositional clusters, with species richness
within each cluster
Compositional
cluster
Habitat group
Total sites Species richness1 2 3 4 5
A 57 6 2 4 0 69 41
B 100 20 2 1 0 123 78
C 86 9 4 3 0 102 87
D 2 2 37 2 11 54 58
E 0 1 10 2 34 47 60
F 5 5 0 15 1 26 60
Comparison Log-likelihood ratio statistic d.f.
P
Entire table 451.898 20 < 0.00005
Clusters A,B,C 8.593 8 0.3778
Clusters D,E,F 112.053 8 < 0.00005
Fig. 4 Environmental biplot for NMDS ordination.
The direction of each vector represents the angle of
maximum correlation, while the length represents
the strength of the correlation.
83
62J.C.Nekola
©2003BlackwellPublishingLtd,DiversityandDistributions,9,55–71
Table 5 Ten most frequent taxa in each of the six compositional clusters
Rank
order Cluster A Cluster B Cluster C Cluster D Cluster E Cluster F
1 Discus catskillensis
(82.61%)
Discus catskillensis
(93.50%)
Punctum vitreum
(89.22%)
Carychium exiguum
(88.89%)
Gastrocopta tappaniana
(93.62%)
Hawaiia minuscula
(57.69%)
2 Nesovitrea binneyana
(82.61%)
Punctum minutissimum
(91.06%)
Gastrocopta contracta
(88.24%)
Striatura milium
(83.33%)
Carychium exiguum
(85.11%)
Vallonia costata
(46.15%)
3 Zonitoides arboreus
(82.61%)
Zonitoides arboreus
(90.24%)
Carychium exile
(88.24%)
Nesovitrea electrina
(81.48%)
Nesovitrea electrina
(85.11%)
Cochlicopa lubrica
(46.15%)
4 Vertigo cristata
(72.46%)
Strobilops labyrinthica
(87.80%)
Vertigo gouldi
(87.25%)
Strobilops labyrinthica
(72.22%)
Vertigo elatior
(74.47%)
Helicodiscus parallelus
(46.15%)
5 Striatura milium
(71.01%)
Anguispira alternata
(84.55%)
Anguispira alternata
(85.29%)
Striatura exigua
(72.22%)
Euconulus alderi
(70.21%)
Gastrocopta contracta
(42.31%)
6 Punctum minutissimum
(62.32%)
Vertigo gouldi
(83.74%)
Strobilops labyrinthica
(79.41%)
Zonitoides arboreus
(68.52%)
Hawaiia minuscula
(59.57%)
Gastrocopta similis
(42.31%)
7 Zoogenetes harpa
(53.62%)
Columella simplex
(83.74%)
Gastrocopta holzingeri
(78.43%)
Vertigo elatior
(59.26%)
Gastrocopta contracta
(57.45%)
Punctum vitreum
(42.31%)
8 Euconulus fulvus
(49.28%)
Striatura milium
(73.98%)
Hawaiia minuscula
(78.43%)
Punctum minutissimum
(59.26%)
Oxylama retusa
(55.32%)
Gastrocopta holzingeri
(38.46%)
9 Vertigo paradoxa
(49.28%)
Helicodiscus shimeki
(66.67%)
Gastrocopta pentodon
(74.51%)
Gastrocopta tappaniana
(53.70%)
Stenotrema leai
(55.32%)
Vallonia pulchella
(38.46%)
10 Striatura exigua
(46.38%)
Euconulus fulvus
(65.85%)
Zonitoides arboreus
(74.51%)
Euconulus alderi
(48.15%)
Deroceras spp.
(55.32%)
Vertigo pygmaea
(38.46%)
84
GreatLakeslandsnailcommunitypatterns63
©2003BlackwellPublishingLtd,DiversityandDistributions,9,55–71
Table 6 Taxa reaching modal frequencies in each compositional cluster
Cluster A Cluster B Cluster C Cluster D Cluster E Cluster F
Nesovitrea binneyana Carychium nannodes Allogona profunda Carychium exiguum Catinella avara Catinella ‘vermeta’
Vertigo cristata Cochlicopa morseana Anguispira alternata Planogyra asteriscus Catinella exile Cepaea nemoralis
Vertigo modesta modesta Columella simplex Carychium exile Striatura exigua Cochlicopa lubricella Cochlicopa lubrica
Vertigo modesta parietalis Discus catskillensis Catinella ‘gelida’ Striatura milium Discus cronkhitei Gastrocopta armifera
Vertigo paradoxa Discus patulus Deroceras spp. Vertigo nylanderi Euconulus alderi Gastrocopta procera
Zoogenetes harpa Euconulus fulvus Discus macclintockii Gastrocopta tappaniana Gastrocopta rogersensis
Euconulus polygyratus Gastrocopta contracta Hawaiia n.sp. Gastrocopta similis
Glyphyalinia rhoadsi Gastrocopta corticaria Helicodiscus n.sp. Glyphyalinia wheatleyi
Helicodiscus shimeki Gastrocopta holzingeri Nesovitrea electrina Helicodiscus inermis
Mesomphix cupreus Gastrocopta pentodon Oxyloma peoriensis Helicodiscus parallelus
Mesomphix inornatus Glyphyalinia indentata Oxyloma retusa Helicodiscus singleyanus
Oxychylus draparnaudi Guppya sterkii Pomatiopsis lapidaria Pupilla muscorum
Paravitrea multidentata Haplotrema concavum Punctum n.sp. Pupoides albilabris
Punctum minutissimum Hawaiia minuscula Stenotrema leai Succinea indiana
Striatura ferrea Hendersonia occulta Strobilops affinis Vallonia costata
Strobilops labyrinthica Mesodon clausus Triodopsis multilineata Vallonia excentrica
Triodopsis albolabris Mesodon pennsylvanicus Vertigo elatior Vallonia parvula
Triodopsis denotata Mesodon thyroidus Vertigo milium Vallonia pulchella
Triodopsis tridentata Oxychylus cellarius Vertigo morsei Vertigo pygmaea
Vallonia gracilicosta Punctum vitreum Vertigo ovata Zonitoides nitidus
Vertigo n.sp. Stenotrema barbatum
Vertigo bollesiana Stenotrema fraternum
Vertigo hubrichti Strobilops aenea
Vitrina limpida Succinea ovalis
Zonitoides arboreus Trichia striolata
Triodopsis alleni
Triodopsis fosteri
Vallonia perspectiva
Vertigo gouldi
Vertigo meramecensis
Vertigo tridentata
Zonitoides limatulus
85
64 J. C. Nekola
© 2003 Blackwell Publishing Ltd, Diversity and Distributions, 9, 55–71
values ranging from 0.2111 to 0.7909 (Table 7).
In duff sites, only UTM N coordinate and anthropogenic
disturbance were found to correlate
significantly (P < 0.0005) with the ordination
diagram (Table 7). Northern sites tended to occur
in the lower left of this group (maximum r =
0.8409), while disturbed sites tended to occur
further to the right (maximum r = 0.4812;
Fig. 5). In turf sites, only UTM N and E coordinates
were found to correlate significantly
(P < 0.0005) with the ordination diagram (Table 7).
In this group, more northern (maximum r =
0.6690) and eastern (maximum r = 0.5067) sites
tended to occur to the lower left (Fig. 5).
Discriminant analysis demonstrated that the
location of duff and turf sites in ordination space
differs significantly (P < 0.0005) (Table 8), with
duff sites being essentially limited to the upper
left half of the diagram, and turf sites being
found largely in the lower right (Fig. 6). The
classification summary for this analysis indicates
that only 33 of the 421 sites (7.8%) were classified
improperly when soil surface type was used as
the sole predictor variable.
Discriminant analyses conducted separately on
duff and turf sites demonstrated that disturbed
duff sites were significantly (P < 0.0005) shifted
to the right of undisturbed ones (Fig. 7). The
Table 7 Two-dimensional correlation statistics for environmental variables in land snail ordination space
Variable Maximum r Angle to first axis P
Entire ordination
UTM E-W coordinate 0.2111 129.4 < 0.0005
UTM N-S coordinate 0.7909 148.9 < 0.0005
Soil surface type 0.7843 52.1 < 0.0005
Disturbance presence 0.2829 1.7 < 0.0005
Duff soil sites only
UTM E-W coordinate 0.1020 100.4 0.190
UTM N-S coordinate 0.8409 134.7 < 0.0005
Disturbance presence
Turf soil sites only 0.4812 31.9 < 0.0005
UTM E-W coordinate 0.5067 150.4 < 0.0005
UTM N-S coordinate 0.6690 162.0 < 0.0005
Disturbance presence 0.0671 1.3 0.820
Fig. 5 Environmental biplots for duff and turf sites. The direction of each vector represents the angle of
maximum correlation, while the length represents the strength of the correlation.
86
Great Lakes land snail community patterns 65
© 2003 Blackwell Publishing Ltd, Diversity and Distributions, 9, 55–71
classification summary for this analysis indicates
that only 50 of the 295 duff sites (16.9%) were
improperly classified when anthropogenic disturbance
was used as the sole predictor variable.
However, no significant differences (P = 0.758)
were noted in the location of disturbed vs. undisturbed
turf sites (Table 8).
Faunistic turnover between duff and
turf soils
As differences in occurrence frequency between
duff and turf sites were analysed for all 108 species,
the significance threshold was lowered using
a Bonferroni correction to P = 0.000463. Species
with P-values ranging from 0.05 to 0.000463 were
considered to possess statistically nonsignificant
trends in their response to soil surface architecture.
Species with P-values exceeding 0.05 were
considered generalists.
Thirty-six species demonstrated no significant
differences in their occurrence frequencies
between duff and turf soils (Appendix I). Sixteen
of these (Cochlicopa lubricella, Deroceras spp.,
Discus cronkhitei, Gastrocopta armifera, Gastrocopta
contracta, Haplotrema concavum, Hawaiia
Table 8 Summary statistics for discriminant analysis of substrate and disturbance comparisons in land snail
community ordination
Factor
Comparison
Duff vs. turf
(all sites)
Disturbed vs. pristine
(duff sites only)
Disturbed vs. pristine
(turf sites only)
Canonical correlation 0.784 0.482 0.067
Eigenvalue 1.596 0.302 0.005
Likelihood ratio 0.385 0.768 0.996
Approximate F 322.74 43.915 0.278
Number d.f. 2 2 2
Density d.f. 417 291 123
P < 0.0005 < 0.0005 0.758
Fig. 7 Location of disturbed and undisturbed sites in the ordination diagram for duff and turf sites.
Fig. 6 Location of duff and turf sites within the
ordination diagram.
87
66 J. C. Nekola
© 2003 Blackwell Publishing Ltd, Diversity and Distributions, 9, 55–71
minuscula, Helicodiscus parallelus, Helicodiscus
singleyanus, Planogyra asteriscus, Striatura ferrea,
Striatura milium, Triodopsis multilineata,
Vallonia costata, Vertigo pygmaea, Vitrina limpida)
were found in 10 or more sites, and clearly
represent generalists. However, the remaining 20
(Carychium nannodes, Cepaea nemoralis, Discus
patulus, Glyphyalinia wheatleyi, Helicodiscus
inermis, Mesodon pennsylvanicus, Mesomphix
cupreus, Mesomphix inornatus, Oxychylus cellarius,
Oxychylus draparnaudi, Oxyloma peoriensis,
Pomatiopsis lapidaria, Pupilla muscorum, Succinea
indiana, Trichia striolata, Triodopsis denotata,
Triodopsis fosteri, Vallonia excentrica,
Vertigo modesta parietalis, Zonitoides limatulus)
are known from fewer locations, making Type 2
errors a significant concern. Additional data will
be needed to adequately assess the response of
these species to duff vs. turf soils.
Eighteen species (Cochlicopa lubrica, Cochlicopa
morseana, Discus macclintockii, Mesodon
clausus clausus, Mesodon thyroidus, Punctum
minutissimum, Punctum vitreum, Stenotrema barbatum,
Strobilops aenea, Strobilops labyrinthica,
Succinea ovalis, Triodopsis albolabris, Triodopsis
alleni, Triodopsis tridentata, Vallonia perspectiva,
Vertigo meramecensis, Vertigo modesta modesta,
Vertigo tridentata) nonsignificantly favoured duff
sites. Another eight (Gastrocopta procera, Gastrocopta
rogersensis, Gastrocopta similis, Hawaiia
n.sp., Helicodiscus n.sp., Striatura exigua, Vallonia
pulchella, Zonitoides nitidus) nonsignificantly
favoured turf sites. While a number of these (e.g.
Discus macclintockii, Gastrocopta procera, Gastrocopta
rogersensis, Vertigo meramecensis) demonstrated
very strong absolute preferences, their
few total occurrences in combination with the
conservative Bonferroni correction prevented them
from exhibiting significant responses.
The remaining 46 species demonstrated clear
and significant soil surface preferences. Twentyeight
species (Allogona profunda, Anguispira
alternata, Carychium exile, Catinella ‘gelida’, Columella
simplex, Discus catskillensis, Euconulus
fulvus, Euconulus polygyratus, Gastrocopta corticaria,
Gastrocopta holzingeri, Gastrocopta pentodon,
Glyphyalinia indentata, Glyphyalinia rhoadsi,
Guppya sterkii, Helicodiscus shimeki, Hendersonia
occulta, Nesovitrea binneyana, Paravitrea multidentata,
Stenotrema fraternum, Vallonia gracilicosta,
Vertigo bollesiana, Vertigo cristata, Vertigo
gouldi, Vertigo hubrichti, Vertigo n.sp., Vertigo
paradoxa, Zonitoides arboreus, Zoogenetes harpa)
favoured duff soils, while another 18 (Carychium
exiguum, Catinella avara, Catinella exile, Catinella
‘vermeta’, Euconulus alderi, Gastrocopta
tappaniana, Nesovitrea electrina, Oxyloma retusa,
Punctum n.sp., Pupoides albilabris, Stenotrema
leai leai, Strobilops affinis, Vallonia parvula,
Vertigo elatior, Vertigo milium, Vertigo morsei,
Vertigo nylanderi, Vertigo ovata) favoured turf
soils.
DISCUSSION
These data demonstrate clearly that at large environmental
and spatial scales most land snail species
possess pronounced ecological preferences.
They thus represent a paradox, being generalists
at small scales, yet responding to specific environmental
factors at larger ones. At large scales,
species tend to congregate into six major compositional
clusters related to habitat type, soil surface
architecture, geography, moisture levels and
presence of anthropogenic disturbance.
Habitat type
The six compositional clusters significantly differ
in their habitat representations. Clusters A–D
generally consist of forested sites while clusters
E–F generally consist of grasslands. These results
are in agreement with previous studies from
other regions, including north-western Spain
(Ondina & Mato, 2001), southern France (Magnin
et al., 1995), western Switzerland (Baur et al.,
1996), Croatia (Stamol, 1991, 1993), Hungary (Bába,
1989) and north-eastern Nevada (Ports, 1996).
The contrast between open-ground and forest
faunas is not limited to terrestrial gastropods.
Other soil invertebrate groups that demonstrate
this pattern include fungus-eating microarthropods
(Branquart et al., 1995), carabid beetles
(McCracken, 1994), terrestrial amphipods
(Taylor et al., 1995) and collembola (Greenslade,
1997). In an ordination of global earthworm
communities, Lavelle et al. (1995) demonstrated
that open-ground and forest assemblages were
distinct from the warm-tropics to the arctic. The
distinction between forest and grassland faunas
thus appears to be a general driving factor in soil
biota community composition.
88
Great Lakes land snail community patterns 67
© 2003 Blackwell Publishing Ltd, Diversity and Distributions, 9, 55–71
Soil surface architecture
The greater similarity of most lowland forest faunas
to lowland grasslands, as opposed to upland
forests and rock outcrops (Fig. 2), suggests additional
factors underlie observed land snail composition
patterns. The potential importance of
soil surface architecture is implied as many lowland
forest sites (e.g. tamarack, white cedar and
most black ash swamp forests), and all lowland
grasslands, possess turf soils. Only 8% of sites
were improperly classified when soil surface type
was used as the sole predictor variable (Table 7).
Even this rate may be exaggerated, as most
misclassifications were limited to two specific
instances. First, even though having turf soils,
igneous shoreline habitats had faunas almost
identical to surrounding rock outcrop sites.
Snails in this habitat, however, were largely
restricted to friable accumulations of organic
matter under stunted white cedar trees. Secondly,
almost all duff sites with faunas similar to
upland grasslands had experienced severe levels
of anthropogenic disturbance.
Striking differences exist between the species
of duff and turf sites: 43% of taxa significantly
favoured one soil surface type over the other
(even with use of a conservative Bonferronicorrected
significance threshold), while only 15%
of frequent taxa (10 + occurrences) showed no
preference. Inspection of these data indicate that
for eight groups of closely related taxa within the
same genus (24 total), one or more significantly
favour duff soils, while the other(s) significantly
favour turf (Table 9). In another four groups (10
additional taxa), one or more taxa significantly
favour one of these soil types, while the other(s)
exhibit a nonsignificant preference (Table 9).
These 12 groups represent a wide variety of
phylogenetic stocks (representing nine families:
Carychiidae, Helicarionidae, Polygyridae, Punctidae,
Pupillidae, Strobilopsidae, Succineidae, Valloniidae,
Zonitidae), shell shapes (five wider than
tall, five taller than wide and two equally tall as
wide), and maximum shell dimensions (0.8 mm–
12 mm). The presence of so many pairs of closely
related duff- and turf-specialist taxa across such
a wide range of phylogenies, shell shapes and
dimensions suggests that very strong selective
pressures between these soil surface types have
extended over evolutionary time scales.
Table9Closelyrelatedintergenericspeciespairsthatdemonstratesignificantdifferencesand/ornonsignificanttrendsintheirsoilsurfacepreferences.A
significancelevelof**representsthosespecieswhichpossessP<0.000463;*representthosespecieswhere0.000463<P<0.05
Duff-specialistSign.levelTurf-specialistSign.levelFamily
Shell
shapeShellsize(mm)
Carychiumexile**Carychiumexiguum**CarychiidaeTall1.5–2
Catinella‘gelida’**Catinellaavara,C.exile,C.‘vermeta’**SuccineidaeTall4–6
Euconulusfulvus**Euconulusalderi**HelicarionidaeEqual3–4
Gastrocoptapentodon**Gastrocoptatappaniana**PupillidaeTall2
Nesovitreabinneyana**Nesovitreaelectrina**ZonitidaeWide4–7
Punctumminutissimum,P.vitreum*Punctumn.sp.**PunctidaeWide0.8–1.5
Stenotremafraternum**Stenotremaleai**PolygyridaeWide8–12
Strobilopslabyrinthica*Strobilopsaffinis**StrobilopsidaeEqual2–3
Valloniaperspectiva*Valloniaparvula**ValloniidaeWide2
Vertigogouldi‘group’**Vertigoovata‘group’**PupillidaeTall1.5–3
(V.bollesiana,V.cristata,V.gouldi)(V.elatior,V.morsei,V.ovata)
Vertigohubrichti,V.paradoxa,V.n.sp.**Vertigonylanderi**PupillidaeTall1.5–2
Zonitoidesarboreus**Zonitoidesnitidus*ZonitidaeWide4–6
89
68 J. C. Nekola
© 2003 Blackwell Publishing Ltd, Diversity and Distributions, 9, 55–71
It is not possible via the current analyses to
definitively identify what such factors might be.
They must be limited to the detritusphere (Coleman
& Crossley, 1996), as almost 90% of snails
live within 5 cm of the soil surface (Hawkins
et al., 1998). One possible mechanism is increased
competition with living plant roots in turf soils
for inorganic nutrients (Lavelle et al., 1995).
Another may be the greater organic litter thickness
in duff soils, as the abundance (Berry, 1973),
diversity (Cain, 1983; Locasciulli & Boag, 1987)
and composition (Cameron & Morgan-Huws,
1975; Baur et al., 1996; Barker & Mayhill, 1999)
of land snail communities often correlates positively
with litter depth. The architecture of
organic litter (Burch, 1956; Cameron, 1986;
Young & Evans, 1991; Alvarez & Willig, 1993),
and the underlying soil (Cameron, 1982; Hermida
et al., 2000) may also have strong impacts on land
snail community structure.
Similarly, the composition and abundance of
other soil taxa communities can be influenced by
organic litter depth and architecture, including
amphipods (Taylor et al., 1995), microarthropods
(Borcard & Matthey, 1995; Branquart et al.,
1995; Kay et al., 1999; Whitford & Sobhy, 1999),
collembola (Kovac & Miklisova, 1997) and ground
beetles (McCracken, 1994). Thus, like habitat type,
upper soil layer architecture appears to be another
vital factor driving soil biota composition.
Geography
The geographical location of sites also influences
community composition, particularly in duff
soils (Fig. 5). Each of the three duff clusters
have an unique geographical affiliation, with
cluster A being largely restricted to the most
northern sites, cluster B to sites in the northern
half of the Lake Michigan–Huron basin, and
cluster C to sites in Iowa, Illinois, and southern
Wisconsin.
While a significant correlation with both latitude
and longitude was also observed in turf
sites, this result is almost certainly an artefact
of the limitation of upland grassland sites to
the south-west of the study region. Fens and
lowland forests, found throughout, exhibited
little geographical trends inside of the ordination
diagram. Geographical location was presumably
less important for these sites due to the
overriding importance of habitat type and soil
moisture.
Soil moisture and temperature
Soil moisture and sunlight levels also appear to
influence land snail community composition in
turf sites, with the driest and sunniest habitats
(upland grasslands) being most different in
composition from wet, shaded lowland forests.
However, temperature and relative humidity, not
sunlight, are probably the important driving
factors (Suominen, 1999), as in both duff and turf
sites the coolest and wettest habitats (northern
cliff, upland forest and lowland forest) were most
different in composition from the hottest and driest
sites (southern cliff, upland forest and upland
grassland).
Disturbance and conservation
Anthropogenic disturbance influences snail composition
differentially between duff and turf sites.
While turf faunas were not impacted, the most
disturbed duff sites had faunas more characteristic
of upland grasslands. Typical species found
on such disturbed sites include Cochlicopa
lubrica, Pupilla muscorum, Vallonia costata,
Vallonia excentrica, Vallonia pulchella and Vertigo
pygmaea. These faunistic differences may be
related to differential changes in soil surface
architecture with disturbance. Because undisturbed
turf soils usually have thinner and less
structurally complex organic litter layers, they
may be less susceptible to soil compaction (and
changes in composition) as compared to duff
sites.
As anthropogenic soil compaction negatively
impacts soil invertebrates more severely than
plants in the same communities (Duffey, 1975),
conservation of duff-specialist land snails will
likely require protection of the soil litter layer
architecture, perhaps by limiting forestry and recreation
activities in duff soil sites of conservation
importance. While turf sites appear to be more
tolerant of human disturbance, this should not
indicate that their land snail communities are
immune to human activity. For instance, heavy
grazing can negatively impact grassland snails
(Cameron & Morgan-Huws, 1975), while the use
of fire-management can lead to significant reduc-
90
Great Lakes land snail community patterns 69
© 2003 Blackwell Publishing Ltd, Diversity and Distributions, 9, 55–71
tions in both species richness and abundance
(Nekola, 2002b).
ACKNOWLEDGMENTS
Robert Cameron and Peter White provided valuable
comments on earlier drafts. Matt Barthel,
Tracy Kuklinski, Pete Massart, Chela Moore,
Eric North, J.J. Schiefelbein and Tamara Smith
processed many soil litter samples and assisted
in field collection. Assistance in litter sample
processing was also provided by students participating
in the Land Snail Ecology Practicum at
the University of Wisconsin — Green Bay. Funding
was provided by the Door County Office of
the Wisconsin Chapter of The Nature Conservancy,
a Louis Almon grant (administered by the
Wisconsin Academy of Sciences, Arts and Letters),
three Cofrin Arboretum grants (administered
by the Cofrin Arboretum Committee at the
University of Wisconsin — Green Bay), the U.S.
Fish and Wildlife Service and the Small Grants
Program of the Michigan Department of Natural
Resources. Funding for the survey of Minnesota
sites was received from the Minnesota Nongame
Wildlife Tax Checkoff and Minnesota State Park
Nature Store Sales through the Minnesota
Department of Natural Resources Natural Heritage
and Nongame Research Program.
SUPPLEMENTARY MATERIAL
The following material is available from http://
www.blackwellpublishing.com/products/journals/
suppmat/DDI/DDI165/DDI165sm.htm
Appendix 1
Table S1 Species occurrences and frequencies
within the six main ordination clusters, and
within sites with duff or turf organic horizons.
REFERENCES
Alvarez, J. & Willig, M.R. (1993) Effects of treefall
gaps on the density of land snails in the Luquillo
Experimental Forest of Puerto Rico. Biotropica
25, 100–110.
Anderson, W.I. (1983) Geology of Iowa, p. 268. Iowa
State University Press, Ames, Iowa.
Bába, K. (1989) Zoogeographical conditions of snails
living on grass-associations of two Hungarian
lowland regions. Tiscia 24, 59–67.
Banfield, J.D. & Raftery, A.E. (1992) Model-based
Gaussian and non-Gaussian clustering. Biometrics
49, 803–822.
Barbour, M.G. & Billings, W.D. (1988) North American
terrestrial vegetation, p. 434. Cambridge University
Press, New York.
Barker, G.M. & Mayhill, P.C. (1999) Patterns of
diversity and habitat relationships in terrestrial
mollusc communities of the Pukeamaru Ecological
District, northeastern New Zealand. Journal of
Biogeography 26, 215–238.
Baur, B., Joshi, J., Schmid, B., Hänggi, A., Borcard, D.,
Stary, J., Pedroli-Christen, A., Thommen, G.H.,
Luka, H., Rusterholz, H.P., Oggier, P., Ledergerber, S.
& Erhardt, A. (1996) Variation in species richness
of plants and diverse groups of invertebrates in
three calcareous grasslands of the Swiss Jura
mountains. Revue Suisse de Zooligie 103, 801–833.
Berry, F.G. (1973) Patterns of snail distribution at
Ham Street Woods National Nature Reserve, East
Kent. Journal of Conchology 28, 23–35.
Borcard, D. & Matthey, W. (1995) Effect of a controlled
trampling of Sphagnum mosses on their
Oribatid mite assemblages (Acari, Oribatei).
Pedobiologia 39, 219–230.
Boycott, A.E. (1934) The habitats of land mollusca
in Britain. Journal of Ecology 22, 1–38.
Branquart, E., Kime, R.D., Dufrêne, M., Tavernier, J.
& Wauthy, G. (1995) Macroarthropod-habitat
relationships in oak forests in South Belgium. 1.
Environments and communities. Pedobiologia 39,
243–263.
Burch, J.B. (1956) Distribution of land snails in
plant associations in eastern Virginia. Nautilus 70,
60–64.
Burch, J.B. & Pearce, T.A. (1990) Terrestrial Gastropoda.
Soil Biology Guide (ed. by D.L. Dindal),
pp. 201–309. John Wiley & Sons, New York.
Cain, A.J. (1983) Ecology and ecogenetics of terrestrial
molluscan populations. The Mollusca, Vol. 6.
Ecology (ed. by W.D. Russell-Hunter), pp. 597–
647. Academic Press, New York.
Cameron, R.A.D. (1982) Life histories, density, and
biomass in a woodland snail community. Journal
of Molluscan Studies 48, 159–166.
Cameron, R.A.D. (1986) Environment and diversities
of forest snail faunas from coastal British
Columbia. Malacologia 27, 341–355.
Cameron, R.A.D. (2002) Some species/area relationships
in the British land mollusc fauna and their
implications. Journal of Conchology 37, 337–348.
Cameron, R.A.D. & Morgan-Huws, D.I. (1975) Snail
faunas in the early stages of a chalk grassland
succession. Biological Journal of the Linnean
Society 7, 215–229.
Coleman, D.C. & Crossley, D.A. Jr (1996) Fundamentals
of soil ecology. Academic Press, New
York.
Cowie, R.H. & Jones, J.S. (1987) Ecological interactions
between Cepaea nemoralis and Cepaea hortensis:
91
70 J. C. Nekola
© 2003 Blackwell Publishing Ltd, Diversity and Distributions, 9, 55–71
competition, invasion, but no niche displacement.
Functional Ecology 1, 91–97.
Cowie, R.H., Nishida, G.M., Basset, Y. & Gon, S.M. III
(1995) Patterns of terrestrial gastropod distribution
in a montane habitat on the island of Hawaii.
Malacologia 36, 155–169.
Duffey, E. (1975) The effects of human trampling
on the fauna of grassland litter. Biological Conservation
7, 255–274.
Eichenlaub, V.L. (1979) Weather and climate of the
Great Lakes region. University of Notre Dame
Press, Notre Dame, Indiana.
Emberton, K.C., Pearce, T.A. & Randalana, R.
(1996) Quantitatively sampling land-snail species
richness in Madagascan rainforests. Malacologia
38, 203–212.
Equihua, M. (1990) Fuzzy clustering of ecological
data. Journal of Ecology 78, 519–534.
Faith, D.P., Minchin, P.R. & Belbin, L. (1987) Compositional
dissimilarity as a robust measure of
ecological distance. Vegetatio 69, 57–68.
Frest, T.J. (1990) Final report, field survey of Iowa
spring fens. Contract no. 65–2454. Iowa Department
of Natural Resources, Des Moines.
Frest, T.J. (1991) Summary status reports on eight
species of candidate land snails from the Driftless
Area (Paleozoic Plateau), Upper Midwest. Final
Report, contract no. 301–01366, USFWS Region
3. Ft. Snelling, Minnesota.
Greenslade, P. (1997) Are Collembola useful as
indicators of the conservation value of native
grasslands? Pedobiologia 41, 215–220.
Hartigan, J.A. (1975) Clustering Algorithms. John
Wiley Press, New York.
Hawkins, J.W., Lankester, M.W. & Nelson, R.R.A.
(1998) Sampling terrestrial gastropods using cardboard
sheets. Malacologia 39, 1–9.
Hermida, J., Ondina, M.P. & Rodriguez, T. (2000)
The relative importance of edaphic factors on the
distribution of some terrestrial gastropod species:
autecological and synecological approaches. Acta
Zoologica Academiae Scientiarum Hungaricae 46,
265–274.
Hubricht, L. (1985) The distributions of the native
land mollusks of the eastern United States. Fieldiana
N.S. 24, 1–191.
Kay, F.R., Sobhy, H.M. & Whitford, W.G. (1999)
Soil microarthopods as indicators of exposure to
environmental stress in Chihuahuan Desert rangelands.
Biology and Fertility of Soils 28, 121–128.
Kerney, M.P. & Cameron, R.A.D. (1979) Field guide to
the land snails of the British Isles and northwestern
Europe, p. 288. Collins Press, London.
Kovac, L. & Miklisova, D. (1997) Collembolan communities
(Hexapoda, Collembola) in arable soils
of East Slovakia. Pedobiologia 41, 62–68.
Lavelle, P., Lattaud, C., Trigo, D. & Barois, I. (1995)
Mutualism and biodiversity in soils. Plant and Soil
170, 23–33.
Locasciulli, O. & Boag, D.A. (1987) Microdistribution
of terrestrial snails (Stylommatophora) in forest
litter. Canadian Field Naturalist 101, 76–81.
Magnin, F., Tatoni, T., Roche, P. & Baudry, J. (1995)
Gastropod communities, vegetation dynamics and
landscape changes along an old-field succession in
Provence, France. Landscape and Urban Planning
31, 249–257.
Mason, C.F. (1970) Snail populations, beech litter
production, and the role of snails in litter decomposition.
Oecologia 5, 215–239.
McCracken, D.I. (1994) A fuzzy classification of
moorland ground beetle (Coleoptera: Carabidae)
and plant communities. Pedobiologia 38, 12–
27.
Minchin, P.R. (1987) An evaluation of the relative
robustness of techniques for ecological ordination.
Vegetatio 69, 89–108.
Minchin, P.R. (1990) DECODA Software. Australian
National University, Canberra.
Nekola, J.C. (1999) Terrestrial gastropod richness
of carbonate cliff and associated habitats in the
Great Lakes region of North America. Malacologia
41, 231–252.
Nekola, J.C. (In press) Terrestrial gastropod fauna
of northeastern Wisconsin and the southern
Upper Peninsula of Michigan. American Malacological
Bulletin, in press.
Nekola, J.C. (2002) Effects of fire management on
the richness and abundance of central North
American grassland land snail faunas. Animal
Biodiversity and Conservation 25: 53–66.
Nekola, J.C. & Smith, T.A. (1999) Terrestrial gastropod
richness patterns in Wisconsin carbonate cliff
communities. Malacologia 41, 253–269.
Oggier, P., Zschokke, S. & Baur, B. (1998) A comparison
of three methods for assessing the gastropod
community in dry grasslands. Pedobiologia
42, 348–357.
Ondina, P. & Mato, S. (2001) Influence of vegetation
type on the constitution of terrestrial gastropod
communities in northwest Spain. Veliger 44,
8–19.
Pearce, T.A. (1997) Interference and resource competition
in two land snails: adults inhibit conspecific
juvenile growth in field and laboratory.
Journal of Molluscan Studies 63, 389–399.
Ports, M.A. (1996) Habitat affinities and distributions
of land gastropods from the Ruby Mountains and
East Humboldt Range of northeastern Nevada.
Veliger 39, 335–341.
Schmid, G. (1966) Die mollusken des Spitzbergs. Der
Spitzberg bei Tübingen, Natur und Landschaftsschutzgebiete,
Baden-Würtemburg 3, 596–701.
Smallridge, M.A. & Kirby, G.C. (1988) Competitive
interactions between the terrestrial gastropods
Thebia pisana (Müller) and Cernuella virgata
(DaCosta) from South Australia. Journal of Molluscan
Studies 54, 251–258.
92
Great Lakes land snail community patterns 71
© 2003 Blackwell Publishing Ltd, Diversity and Distributions, 9, 55–71
Stamol, V. (1991) Coenological study of snails
(Mollusca: Gastropoda) in forest phytocoenoses of
Medvednica mountain (NW Croatia, Yugoslavia).
Vegetatio 95, 33–54.
Stamol, V. (1993) The influence of ecological characteristics
of phytocoenoses on the percentage
proportions of zoogeographical elements in the
malacocoenoses of land snails (Mollusca: Gastropoda
terrestrial). Vegetatio 109, 71–80.
Statistical Sciences (1995) S-PLUS Guide to Statistical
and Mathematical Analysis, Version 3.3. Seattle,
Washington.
Suominen, O. (1999) Impact of cervid browsing and
grazing on the terrestrial gastropod fauna in the
boreal forests of Fennoscandia. Ecography 22,
651–658.
Tattersfield, P. (1990) Terrestrial mollusc faunas
from some South Pennine woodlands. Journal of
Conchology 33, 355–374.
Taylor, R.J., Walsh, A.M. & Brereton, R.N. (1995)
Local distributions of terrestrial amphipods within
highland wet forests in central Tasmania. Pedobiologia
39, 78–85.
Theler, J.L. (1997) The modern terrestrial gastropod
(land snail) fauna of western Wisconsin’s hill
prairies. Nautilus 110, 111–121.
Van Es, J. & Boag, D.A. (1981) Terrestrial molluscs
of central Alberta. Canadian Field-Naturalist 95,
76–79.
Wäreborn, I. (1970) Environmental factors influencing
the distribution of land molluscs in an
oligotrophic area in southern Sweden. Oikos 21,
285–291.
Whitford, W.G. & Sobhy, H.M. (1999) Effects of
repeated drought on soil microarthropod communities
in the northern Chihuahuan Desert. Biology
and Fertility of Soils 28, 117–120.
Young, M.S. & Evans, J.G. (1991) Modern land
mollusc communities from Flat Holm, South
Glamorgan. Journal of Conchology 34, 63–70.
Zar, J.H. (1984) Biostatistical Analysis. Prentice
Hall, Inc., Englewood Cliffs, New Jersey.
93
© 1999 Macmillan Magazines Ltd
diameter) per treatment. The tubes were
then placed in a water bath with a temperature
gradient from top to bottom of 20 to
10 °C. We used a spectrophometric assay
procedure10
to establish the TMA concentration
in 10 litres of the medium that had
previously contained four fish (ides, Leuciscus
idus, 10 cm long), and recorded the
resulting vertical migration behaviour of
daphnids exposed to this medium. TMA
concentration in this ‘fish’ water was
between 10 and 25 ȖM.
We found that even low concentrations
of TMA induce Daphnia to migrate to
deeper waters during the day in our test
system. At night, they migrate back to the
surface when exposed to small amounts of
TMA, and stay near the bottom when TMA
levels are high (more than 100 ȖM). When
the lowest TMA concentration that induces
vertical migration is compared with the
activity in the fish water, it is clear that the
reaction of Daphnia to fish water is
stronger than just to TMA alone. This indicates
that, although TMA is an active component
of the ‘fish factor’, it is likely to be
part of a cocktail of substances that deter
Daphnia. The other substances probably
help to reduce the chemical threshold that
induces migration. It is not clear whether
the TMA concentrations used in this study
represent realistic values for aquatic systems,
as we could find no published details
of TMA concentrations in aquatic systems.
It has been shown9
that the fish kairomone
is broken down by bacteria. We
therefore compared the average day depth
of daphnids exposed to 75 ȖM TMA in
autoclaved water, and with the antibiotic
ampicillin added and the TMA dissolved in
non-autoclaved lake water. The migration
activity of the Daphnia in non-sterile water
slowly decreased over time (Fig. 2). After 72
hours, the average day depth of the animals
in non-sterile conditions was no longer significantly
different from the average day
depth of animals in sterile control medium.
To investigate whether simply adding
any substance to the water induces vertical
migration, we added the same amount of
TMAO and triethylamine to sterile water
(both at 75 ȖM). As with the control Daphnia,
there was no significant increase in day
depth. We therefore conclude that the reaction
to TMA is a specific one, and not simply
a response to a change in conductivity
or ionic strength of the medium.
Hinnerk Boriss, Maarten Boersma,
Karen H. Wiltshire
Max-Planck-Institut für Limnologie,
Postfach 165, 24302 Plön, Germany
e-mail: boriss@mpil-ploen.mpg.de
1. Stich, H. B. & Lampert, W. Nature 293, 396–398 (1981).
2. Gliwicz, M. Z. Nature 320, 746–748 (1986).
3. Ringelberg, J. J. Mar. Biol. Assoc. UK 75, 15–25 (1995).
4. Larsson, P. & Dodson, S. Arch. Hydrobiol. 129, 129–155 (1993).
5. Lange, R. & Fugelli, K. Comp. Biochem. Physiol. 25, 283–292
(1965).
6. Yancey, P. H. & Somero, G. N. Biochem. J. 183, 317–323 (1979).
7. Hill, R. W. Comparative Physiology of Animals (Harper & Row,
New York, 1976).
8. Gram, L., Wedell Neergaard, C. & Huss, H. H. Int. J. Food
Microbiol. 10, 303–316 (1990).
9. Loose, C. J., von Elert, E. & Dawidowicz, P. Arch. Hydrobiol.
126, 329–337 (1993).
10.Kakác, B. & Vejdelek, Z. J. Handbuch der Photometrischen
Analyse Organischer Verbindungen (Chemie, Weinheim, 1974).
scientific correspondence
382 NATURE | VOL 398 | 1 APRIL 1999 | www.nature.com
Trimethylamine induces
migration of waterfleas
It has long been known that the waterflea
Daphnia hyalina exhibits diel vertical
migration in the water column, but the
chemical that triggers this behaviour has
not been identified. We find that trimethylamine
(TMA), which is a major component
of the odour produced by decaying fish,
induces Daphnia to migrate to greater
depths during the day, presumably to avoid
predation by fish1,2
. We observed a gradual
increase in average depth of Daphnia with
increasing TMA concentration. Changes in
light intensity are known to trigger migra-
tion3
, and chemicals produced by their
predators must also be present4
. Because
migration has demographic and physiological
costs, this chemical cue ensures that
zooplankton migration occurs only when
fish are present.
It has been shown that, whatever the
migration trigger substance may be, it is
produced only by fish in the presence of
bacteria. This led to the hypothesis that
trimethylamine-N-oxide (TMAO), which is
important for cell volume regulation5
and
protein stabilization6
in marine and freshwater
fish, could be a precursor of the
kairomone. Surplus TMAO is deposited in
fish scales7
, and bacteria of the skin mucus
layer are known to reduce TMAO to TMA8
.
We tested the hypothesis that TMA
induces vertical migration in Daphnia by
using a range of different concentrations,
from 0 to 500 ȖM TMA, in autoclaved lake
water9
(Fig. 1). Five individuals of a single
Daphnia hyalina clone were put into each of
four Perspex tubes (1 m long and 1.5 cm in
0
10
20
30
40
50
60
70
80
90
100
0
10
20
30
40
50
60
70
80
90
100
fish water
TMA concentration
(µM)
Temperature
(°C)
10 10 12 14 16 18100
Daytimedepth(cm)
Depth(cm)
a b
Sterile
Non-
sterile
Control
20
40
60
80
100
0 20 40 60 80 100
Time (hours)
Daytimedepth(cm)
FFiigguurree 11 Behavioural responses of Daphnia hyalina
to trimethylamine. a, Average day depth (DŽstandard
error) of five daphnids (four replicates per treatment)
on three consecutive days of different TMA concentrations.
The day:night cycle was 16:8 hours. The
reaction of the daphnids is plotted against the concentration
of TMA in fish water. Bacterial decay of
TMA was prevented by the addition of ampicillin.
Migration depth increased in all tubes from day 1 to
day 3. b, Temperature profile of the tubes. As TMA
concentration increases, daphnids migrate to
greater depths and colder water during the day.
FFiigguurree 22 Average day depth (DŽs.e.) of five daphnids.
There were eight replicates per treatment. Symbols
with different colours indicate significant differences
between the treatments (Tukey post hoc comparison
after analysis of variance, with treatment as a
factor and time as a repeated measurement factor).
Sterile and non-sterile treatments are not significantly
different from each other, but were significantly different
from the control for the first 60 hours; after 72
hours, daphnids kept under sterile conditions are
significantly deeper than the control animals and
those kept under non-sterile conditions.
Ancient stunted trees
on cliffs
An undisturbed ancient woodland, dominated
by tiny, slow-growing and widely
spaced trees, grows on vertical cliffs of the
Niagara escarpment in southern Canada1
.
To investigate whether this woodland is
unusual or is part of a previously undetected
global pattern, we sampled ages and radial
growth rates for trees on cliffs in the United
States and in western Europe. We find that
vertical cliffs often support populations of
widely spaced trees that are exceptionally
old, deformed and slow growing. Some of
the most ancient and least-disturbed wooded
habitat types on Earth are found on cliffs,
even at sites close to heavy agricultural and
industrial development.
We selected 21 cliffs for sampling in 15
eastern states of the United States, and 25
cliffs in eastern and southern Germany,
eastern and southern France, central and
northern England and Wales. All areas have
been heavily developed for industry and
agriculture over a long time. The cliffs visited
were at least 500 metres long and
between 20 and 1,500 metres high. We
extracted increment cores for analysis from
near the base of 224 mature trees. Some
individual living trees grew as thin stemstrips
on large amounts of dead wood. We
94
© 1999 Macmillan Magazines Ltd
sometimes took cross-sections from the
dead parts of these trees to allow for more
accurate dating of core samples extracted
from the living portions.
We recorded the diameter of each tree
and the location of the pith for more accurate
age estimations when core samples did
not include the pith2
. We then used the distance
from the cambium to the youngest
ring to calculate radial growth rate,
expressed as millimetres of new wood per
year. All sample cores or sections were
mounted and sanded3
. Results were plotted
as frequency distributions for both age estimate
and growth rate, and US and European
results were compared using
non-parametric tests4,5
. For age estimates,
we expected to find evidence of a negative
exponential frequency distribution if the
mature trees on the cliffs were part of a selfmaintaining
natural woodland6
. For growth
rate, we expected to find a tight clustering
of very low values if the mature trees were
highly constrained in their productivity7
.
Most of the US trees sampled are
between 160 and 400 years old, with two
trees being more than 1,000 years old and
very few less than 100 years old. The trees
were mainly Thuja occidentalis in the northern
states and Juniperus virginiana in the
rest. Despite our non-random sampling,
the portion of the curve above 80 years of
age has the shape of a negative exponential
(Fig. 1a). European trees also have a negative
exponential age distribution when the
youngest age category is excluded (Fig. 1b),
and some J. phoenicea and Taxus baccata
trees in France and the United Kingdom,
respectively, were more than 1,000 years
old. We found unevenly aged living tree
populations on all cliffs, and observed
coarse woody debris in a variety of decay
states. The findings indicate that the populations
on both continents have remained
undisturbed for long periods, and have had
continuous, rather than pulsed, recruitment
over the past millennium.
Radial growth rates for trees in the
United States (Fig. 1c) and Europe (Fig. 1d)
were mostly less than 1 millimetre per year,
similar to the rate for trees growing on the
Niagara escarpment1
. They must be some
of the slowest-growing woody plants on
Earth. All the trees were grossly deformed
as well as stunted. A 1.5-metre-tall, 1,140year-old
J. phoenicea tree, for example,
from the Verdon Gorge, France, had a
radius of nearly 8 cm and an annual radial
growth rate slightly more than 0.06 mm. Its
axial morphology was typical of cliff trees
in exhibiting partial cambial mortality and
an inverted morphology. All core samples
from cliff-face trees showed radial growth
rates that were the same in the sapling and
the mature stages of development.
The ages of the trees may indicate the
ages and growth rates of the entire plant
communities on the cliffs. Cliffs across the
world may support ancient, slow-growing,
open woodland communities that have
escaped major human disturbance, even
when they are situated close to agricultural
and industrial activity, which has destroyed
or altered most other natural habitats 8–10
.
D. W. Larson*, U. Matthes*, J. A. Gerrath*,
J. M. Gerrath†, J. C. Nekola‡, G. L. Walker§,
S. Porembski¶, A. Charlton#,
N. W. K. Larson*
*Cliff Ecology Research Group,
University of Guelph,
Guelph, Ontario, N1G 2W1, Canada
e-mail: dwlarson@uoguelph.ca
†Department of Biology,
University of Northern Iowa,
Cedar Falls, Iowa 50614, USA
‡Department of Natural and Applied Sciences,
University of Wisconsin,
Green Bay, Wisconsin 54311, USA
§Department of Biology,
Appalachian State University,
Boone, North Carolina 28608, USA
¶Botanisches Institut, University of Bonn,
53115 Bonn, Germany
#Department of Biology, University of Manchester,
Manchester M13 9PT, UK
1. Larson, D. W. & Kelly, P. E. Can. J. Bot. 69, 1628–1636 (1991).
2. Kelly, P. E. & Larson, D. W. J. Ecol. 85, 467–478 (1997).
3. Kelly, P. E., Cook, E. R. & Larson, D. W. Can. J. Forest. Res. 24,
1049–1057 (1994).
4. Zar, J. H. Biostatistical Analysis (Prentice-Hall, Englewood
Cliffs, NJ, 1974).
5. Conover, W. J. Practical Non-parametric Statistics 2nd edn
(Wiley, New York, 1971).
6. Hett, J. M. & Loucks, O. L. J. Ecol. 64, 1029–1044 (1976).
7. Kelly, P. E., Cook, E. R. & Larson, D. W. Int. J. Plant Sci. 153,
117–127 (1992).
8. Vitousek, P. M. Ecology 75, 1861–1876 (1994).
9. Spencer, J. W. & Kirby, K. J. Biol. Conserv. 62, 77–93 (1992).
10.Peterken, G. F. Natural Woodland (Cambridge Univ. Press. 1996).
scientific correspondence
NATURE | VOL 398 | 1 APRIL 1999 | www.nature.com 383
20
15
10
5
0
Numberoftrees
Age estimate (years)
0
160
320
480
640
800
960
20
10
0
Radial growth rate (mm per year)
0
0.34
0.67
1.00
1.34
1.67
.00
2.34
2.67
>3.00
30
40
Numberoftrees
20
10
0
Radial growth rate (mm per year)
0
0.34
0.67
1.00
1.34
1.67
2.00
2.34
2.67
>3.00
30
40
Numberoftrees
20
10
0
Numberoftrees
Age estimate (years)
0
160
320
480
640
800
960
30
a b
c d
FFiigguurree 11 Age estimates and growth rates of trees on cliffs. Frequency distributions of estimated tree ages (a,
b) and radial growth rates (c, d) for mature trees growing on cliffs in the United States (a, c, nǃ97) and
Europe (b, d, nǃ127).
Molecular basis of
triclosan activity
Triclosan (5-chloro-2-(2,4-dichlorophenoxy)
phenol) has been used for more than 30
years as a general antibacterial and antifungal
agent, and is found in formulations as
diverse as toothpastes, cosmetics, antiseptic
soaps, carpets, plastic kitchenware and toys.
It has recently been suggested that triclosan
blocks lipid biosynthesis by specifically
inhibiting the enzyme enoyl–acyl carrier
protein reductase (ENR)1
. We have carried
out a structural analysis and inhibition
experiments on a complex of ENR from the
bacterium Escherichia coli with triclosan and
NAD+
. We find that triclosan acts as a sitedirected,
very potent inhibitor of the
enzyme by mimicking its natural substrate.
ENR catalyses the final, regulatory step
in the fatty-acid synthase cycle: the reduction
of a carbon–carbon double bond in an
enoyl moiety that is covalently linked to an
acyl carrier protein. ENR has been identified
as a target for the diazaborines2
, a family
of antibacterial agents, and many triclosanresistant
strains of E. coli have been found
to be resistant to diazaborines1
.
The structure of the E.coli ENR complexed
with NAD+
and triclosan (triclosan
was provided by Ciba Speciality Chemicals
and Coalite Chemicals) was determined to
2.2 Å resolution. The electron density map
of the complex is of high quality (Fig. 1a)
and reveals the mode of binding of the triclosan
adjacent to the nicotinamide ring of
the nucleotide cofactor in the enzyme’s
active site. The phenol ring of triclosan
forms a face-to-face interaction with the
95
Theoretical relation between δ18
O of land snail shell
carbonate and latitude. The red horizontal lines marked
A-E denote our five proposed transects to calibrate land
snail δ18
Oshell systematics
Section III: Spatio-Temporal Ecology
Ecological patterns are not solely based upon
the local, contemporaneous environment. Due to
the lags inherent in complex systems, these
patterns are also often impacted by drivers
separated by distance in either space or time. As
a result, simply knowing local conditions is
often not enough to accurately predict what
patterns will develop. Understanding how
ecological systems change over space and time,
and how spatial and temporal contingencies
effect current patterns has been a long-standing
focus of my research efforts. These
investigations help determine the nature of
spatial and temporal heterogeneities in
ecological systems, and the extent to which these systems are in equilibrium with the local
environment.
Additionally, there is no better way to determine the potential impacts of global environmental
change than to empirically study how past changes have impacted ecological systems. Land
snails are excellent for this purpose as they are found across all of Earth’s terrestrial biomes and
represent one of the most diverse groups of multicellular life. Their shells also represent one of
the most common fossils in the temperate terrestrial geological record. As a result, this group
has the potential to be as important in generating paleoclimate proxies for terrestrial systems as
foraminifera are in the deep sea. Along with Jeffery Pigati of the US Geological Survey and
Jason Rech of Miami University, our team first documented through a series of highly cited
papers that many land snail phylogenetic groups have shells that are in equilibrium with
environmental 14
C and thus can be
used to accurately date terrestrial
sediments that lack traditional
organic carbon sources (e.g., wood,
peat). We have now turned our
efforts to O-isotopes. Our current
NSF grant (running through 2018)
will document climatic,
environmental and biological
controls on the oxygen isotopic
composition of land snail shells.
We anticipate that this work will
lead to a subsequent proposal
which, if funded, will establish
through field surveys across North
America the fidelity of shell Oisotopes
ratios in the
documentation of modern and prior
96
regional climatic conditions. This work has the potential to revolutionize terrestrial
paleoclimatic proxies given that the same shells can theoretically provide not only useful
chronometric but also environmental data.
Additionally, along with Michal Horsák and Jan Divíšek I am exploring macroevolutionary
process in Holarctic land snails using modern macroecology of these species in combination with
their current molecular phylogenetics. These data are being applied to Late Glacial Maximum
climate models to identify reasons for allopatric replacement within the boreal land snail fauna
between three general regions: Europe, Beringia, and North America. Preliminary results
suggest that geographic isolation of these regions during LGM times (climatic between Europe
and Beringia; ice sheet between Beringia and North America) may well explain modern
observed turnover patterns.
Representative Publications
[number of citations as of October 27, 2017]
Chytrý, M., M. Horsák, V. Syrovátka, J. Danihelka, N. Ermakov, D.A. German, M. Hájek, O.
Hájek, P. Hájková, V. Horsáková, M. Kočí, S. Kubešová, P. Lustyk, J.C. Nekola, Z.
Preislerová, P. Resl & M. Valachovi. 2017. Refugial ecosystems in central Asia as
indicators of biodiversity change during the Pleistocene–Holocene transition. Ecological
Indicators. 76:357-367. [4]
Bajc, A.F., P.F. Karrow, C.H. Yansa, B.B. Curry, J.C. Nekola, K. Seymour & G.L. Mackie.
2015. Geology and paleoecology of a Middle Wisconsin fossil occurrence in Zorra
Township, southwestern Ontario, Canada. Canadian Journal of Earth Sciences. 52:386–
404. [4]
Horsák, M., M. Chytrý, P. Hájková, M. Hájek, J. Danihelka, V. Horsáková, N. Ermakov, D.A.
German, M. Kočí, P. Lustyk, J.C. Nekola, Z. Preislerová & M. Valachovič. 2015.
European glacial relict snails and plants: environmental context of their modern refugial
occurrence in southern Siberia. Boreas. 44:638-657. [14]
Pigati, J.S., J.P. McGeehin, D.R. Muhs, D.C. Grimley & J.C. Nekola. 2015. Radiocarbon
dating loess deposits in the Mississippi Valley using terrestrial gastropod shells
(Polygyridae, Helicinidae, Discidae). Aeolian Research. 16:25-33. [5]
Rakovan, M.T., J.A. Rech, J.S. Pigati, J.C. Nekola & G.C. Wiles. 2013. An evaluation of
Mesodon and other large terrestrial gastropod shells for dating late Holocene and historic
alluvium in the midwestern USA. Geomorphology. 193:47-56. [11]
Horsák, M., M. Hájek, D. Spitale, P. Hájková, D. Dítě & J.C. Nekola. 2012. The age of islandlike
habitats impacts habitat specialist species richness. Ecology. 93:1106-1114. [46]
Rech, J.A., J.C. Nekola & J.S. Pigati. 2012. Radiocarbon ages of terrestrial gastropods extend
duration of ice-free conditions at the Two Creeks forest bed, Wisconsin, USA.
Quaternary Research. 77:289-292. [11]
Pigati, J.S., D.M. Miller, S.A. Mahan, J.A. Bright, J.C. Nekola & J.B. Paces. 2011.
Chronology, sedimentology, and microfauna of ground-water discharge deposits in the
central Mojave Desert, Valley Wells, California. Geological Society of America Bulletin.
123:2224-2239. [24]
97
Rech, J.A., J.S. Pigati, S.B. Lehmann, C.N. McGimpsey, D.A. Grimley & J.C. Nekola. 2011.
Assessing open-system behavior of 14-C in terrestrial gastropod shells. Radiocarbon.
53:325-335. [17]
Pigati, J.S., J.A. Rech & J.C. Nekola. 2010. Radiocarbon dating of small terrestrial gastropod
shells in North America. Quaternary Geochronology. 5:519-532. [94]
Nekola, J.C. & C.E. Kraft. 2002. Spatial constraint of peatland butterfly occurrences within a
heterogeneous landscape. Oecologia. 130:62-71. [21]
Kraft, C.E., P.J. Sullivan, A.Y. Karatayev, L.E. Burlakova, J.C. Nekola, L.E. Johnson & D.K.
Padilla. 2002. Landscape patterns of an aquatic invader: assessing dispersal extent from
spatial distributions. Ecological Applications. 12:749-759. [68]
Nekola, J.C. 1999. Paleorefugia and neorefugia: the influence of colonization history on
community pattern and process. Ecology. 80: 2459-2473. [121]
98
Research Paper
Radiocarbon dating of small terrestrial gastropod shells in North America
Jeffrey S. Pigati a,*, Jason A. Rech b
, Jeffrey C. Nekola c
a
U.S. Geological Survey, Denver Federal Center, Box 25046, MS-980, Denver CO 80225, USA
b
Department of Geology, Miami University, Oxford, OH 45056, USA
c
Department of Biology, University of New Mexico, Albuquerque, NM 87131, USA
a r t i c l e i n f o
Article history:
Received 26 May 2009
Received in revised form
20 January 2010
Accepted 21 January 2010
Available online 29 January 2010
Keywords:
Radiocarbon
Land snails
Limestone effect
Chronology
Quaternary
a b s t r a c t
Fossil shells of small terrestrial gastropods are commonly preserved in wetland, alluvial, loess, and glacial
deposits, as well as in sediments at many archeological sites. These shells are composed largely of
aragonite (CaCO3) and potentially could be used for radiocarbon dating, but they must meet two criteria
before their 14
C ages can be considered to be reliable: (1) when gastropods are alive, the 14
C activity of
their shells must be in equilibrium with the 14
C activity of the atmosphere, and (2) after burial, their
shells must behave as closed systems with respect to carbon. To evaluate the first criterion, we conducted
a comprehensive examination of the 14
C content of the most common small terrestrial gastropods in
North America, including 247 AMS measurements of modern shell material (3749 individual shells) from
46 different species. The modern gastropods that we analyzed were all collected from habitats on
carbonate terrain and, therefore, the data presented here represent worst-case scenarios. In sum, w78% of
the shell aliquots that we analyzed did not contain dead carbon from limestone or other carbonate rocks
even though it was readily available at all sites, 12% of the aliquots contained between 5 and 10% dead
carbon, and a few (3% of the total) contained more than 10%. These results are significantly lower than
the 20–30% dead carbon that has been reported previously for larger taxa living in carbonate terrain. For
the second criterion, we report a case study from the American Midwest in which we analyzed fossil
shells of small terrestrial gastropods (7 taxa; 18 AMS measurements; 173 individual shells) recovered
from late-Pleistocene sediments. The fossil shells yielded 14
C ages that were statistically indistinguishable
from 14
C ages of well-preserved plant macrofossils from the same stratum. Although just one site,
these results suggest that small terrestrial gastropod shells may behave as closed systems with respect to
carbon over geologic timescales. More work on this subject is needed, but if our case study site is
representative of other sites, then fossil shells of some small terrestrial gastropods, including at least five
common genera, Catinella, Columella, Discus, Gastrocopta, and Succinea, should yield reliable 14
C ages,
regardless of the local geologic substrate.
Published by Elsevier B.V.
1. Introduction
Gastropods are one of the most successful animal groups on
Earth, with at least 70,000 extant species occupying terrestrial,
marine, and freshwater habitats. Globally, terrestrial gastropods
encompass at least 35,000 species (Barker, 2001), span 4–5 orders
of magnitude in shell volume, and represent a variety of trophic
levels, including polyphagous detritivores, herbivores, omnivores,
and carnivores (Kerney and Cameron, 1979; Burch and Pearce,
1990). They are so exceptionally diverse in their appearance,
ecology, and physiology that determining their phylogenetic relationships
from conchological and/or anatomical characteristics
remains difficult and controversial (e.g., Ponder and Lindberg, 1997
and references therein). It is clear, however, that the preference for
terrestrial habitats of North American gastropods developed independently
in three of six basal clades (Neritomorpha, Caenogastropoda,
and Heterobranchia), with the informal group Pulmonata
representing more than 99% of the continental fauna. Of the Pulmonata,
the most common size class1
in both modern and fossil
assemblages are individuals with adult shells that are <10 mm in
maximum dimension (Nekola, 2005) (Fig. 1).
* Corresponding author. Tel.: þ1 303 236 7870; fax: þ1 303 236 5349.
E-mail address: jpigati@usgs.gov (J.S. Pigati).
1
Size classes of gastropods are categorized by the maximum shell dimension
(length or diameter) as follows: large (>20 mm), medium (10–20 mm), small (5–
10 mm), minute (2–5 mm), and micro (<2 mm). Although the size of the gastropods
targeted in this study range from small to micro, for simplicity, we refer to
them collectively as ‘‘small’’.
Contents lists available at ScienceDirect
Quaternary Geochronology
journal homepage: www.elsevier.com/locate/quageo
1871-1014/$ – see front matter Published by Elsevier B.V.
doi:10.1016/j.quageo.2010.01.001
Quaternary Geochronology 5 (2010) 519–532
99
Today, small terrestrial gastropods occupy and thrive in
incredibly diverse habitats, from marshes, wet meadows, and
grasslands to upland forests and tundra. Species are known from all
continents, save Antarctica, and occupy almost all climate regimes
except hyperarid deserts and the high Arctic. Their distribution in
the fossil record is equally diverse. Gastropod shells are commonly
preserved in wetland, alluvial, loess, and glacial deposits, as well as
within sediments at archeological sites worldwide (e.g., Evans,
1972). But even though their distribution is widespread and their
aragonitic shells contain w12% by weight carbon, terrestrial
gastropods are often avoided for 14
C dating because many taxa
incorporate 14
C-deficient (or ‘‘dead’’) carbon from limestone and
other carbonate rocks when building their shells. This phenomenon,
referred to as the ‘‘Limestone Problem’’ by Goodfriend and
Stipp (1983), can cause 14
C ages of gastropod shells to be as much
as w3000 yrs too old.
Despite the Limestone Problem, geochronologists have continued
to investigate the possibility of using terrestrial gastropod shells for
14
C dating because of their widespread occurrence and potential for
dating Quaternary sediments. Most 14
C studies of gastropod shells
have found that gastropods consistently incorporate dead carbon
from limestone in their shells when it is available (Frye and Willman,
1960; Leighton, 1960; Rubin et al., 1963; Tamers, 1970; Evin et al.,
1980; Goodfriend and Hood, 1983; Goodfriend and Stipp, 1983;
Goslar and Pazdur, 1985; Yates, 1986; Goodfriend, 1987; Zhou et al.,
1999; Quarta et al., 2007; Romaniello et al., 2008). These studies,
however, were generally limited to a few individual gastropods
collected from a small number of sites, and were biased toward large
taxa and warm climates. Brennan and Quade (1997) analyzed
a number of small terrestrial gastropod taxa and found that small
shells generally yielded reliable 14
C ages for late-Pleistocene paleowetland
deposits in the American Southwest. Pigati et al. (2004)
followed by measuring the 14
C activities of a suite of small gastropods
living in alluvium dominated by Paleozoic carbonate rocks in
Arizona and Nevada and found that while some of the small gastropods
did incorporate dead carbon from limestone when building
their shells, others did not.
Based in part on these initial positive results, small terrestrial
gastropod shells have been used recently to date Quaternary
wetland and lacustrine deposits in the Americas (e.g., Pedone and
Rivera, 2003; Placzek et al., 2006; Pigati et al., 2009). However, it
is unclear if the results obtained from modern gastropods collected
from a limited number of sites in the American Southwest can be
extrapolated to all geologic, ecologic, and climatic environments.
Moreover, it is not known if results for one taxonomic level (family,
genus, or species) can be extrapolated to other members of the
Fig. 1. Photographs of select small terrestrial gastropods included in this study (1 mm bar in each panel for scale). (a) Cochlicopa lubricella, (b) Columella columella alticola, (c) Discus
macclintockii, (d) Euconulus fulvus, (e) Hawaiia miniscula, (f) Hendersonia occulta, (g) Punctum minutissimum, (h) Pupilla muscorum, (i) Strobilops labyrinthica, (j) Vallonia gracilicosta,
(k) Vertigo elatior, and (l) Zonitoides arboreus.
J.S. Pigati et al. / Quaternary Geochronology 5 (2010) 519–532520
100
same level living elsewhere, or even between individuals living
within the same population.
Here we report the results of a comprehensive analysis of the
Limestone Problem for small terrestrial gastropods from 163
localities in North America (Fig. 2). All samples that we analyzed
were collected from habitats on carbonate terrain and, therefore,
the data reported here represent worst-case scenarios. In addition,
we measured the 14
C activities of a number of fossil shells recovered
from well-dated sediments at a late-Pleistocene site in the
American Midwest as a case study to determine if the shells remain
closed systems with respect to carbon over geologic timescales.
Positive results for both tests for a particular taxon would allow us
to consider 14
C ages derived from fossil shells of that taxon to be
reliable, regardless of the local geologic substrate.
2. Shell carbonate and 14
C dating
All materials (organic and inorganic) that yield reliable 14
C ages
have two common characteristics. First, the initial 14
C activity of the
material – a plant, for example – was in equilibrium with
atmospheric 14
C at the time that it was alive. In other words, the 14
C
activity of a plant that lived T yrs ago was the same as the 14
C
activity of the atmosphere T yrs ago (after accounting for isotopic
fractionation). Second, after death, the material behaved as a closed
system; carbon was neither added to nor removed from the sample
material. If both of these criteria are met, then the measured 14
C
activity is a function of only two parameters: the initial 14
C activity
of the atmosphere and the amount of time elapsed since the death
of the organism.
The measured 14
C activity and the 14
C age of the material are
related by the familiar decay equation
A ¼ AoeÀlt
(1)
where A and Ao are the measured and initial 14
C activities of the
material, respectively, l is the decay constant, and t is the time
elapsed since the death of the organism. Conventional radiocarbon
ages assume that the atmospheric 14
C activity is invariant through
time (i.e., Ao ¼ 1). Radiocarbon ages can be converted to calendar
year ages to account for temporal variations in the 14
C activity of the
atmosphere (Reimer et al., 2009).
Fig. 2. Locations of modern localities (red dots) and the fossil locality at the Oxford East outcrops in southwestern Ohio (star). Modern localities and collection information are listed
in Tables S1 and S2, respectively. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article).
J.S. Pigati et al. / Quaternary Geochronology 5 (2010) 519–532 521
101
2.1. Sources of shell carbon
In order to evaluate the validity of a 14
C age of a given sample,
the 14
C contents of the original sources of the carbon and their
contribution to the total carbon content must be known. Carbon in
gastropod shell carbonate originates from as many as four different
sources: atmospheric CO2, food, water, and carbonate rocks.
Gastropods incorporate atmospheric CO2 in their shell
carbonate via respiration. Respired CO2 is introduced to the bicarbonate
pool in the gastropod’s hemolymph and passed along to the
extrapallial fluid, from which the shell carbonate is ultimately
precipitated (Wilbur, 1972). Estimates of the contribution of
atmospheric CO2 to gastropod shell carbonate vary between
negligible (Stott, 2002), 16–48% (Romaniello et al., 2008), and 30–
60% (Goodfriend and Hood, 1983).
Carbon from food sources (e.g., living plants, fungi, organic
detritus) is incorporated into the extrapallial fluid through two
mechanisms: direct digestion and breakdown of urea. When
gastropods consume and digest food, carbon is introduced to the
hemolymph and passed along to the extrapallial fluid in the same
manner as atmospheric CO2. There it mixes with atmospheric
carbon before becoming incorporated in the shell carbonate
(Wilbur, 1972). Carbon derived from urea takes a more indirect
pathway. Urea that is not expelled by the gastropod breaks down
into CO2 and NH3 via a urease reaction (Stott, 2002). The resulting
CO2 is then reintroduced directly to the extrapallial fluid and ultimately
incorporated into the shell carbonate. Estimates of the
amount of carbon derived from plants, either directly or indirectly
through urea, vary between 25 and 40% (Goodfriend and Hood,
1983), 36–73% (Romaniello et al., 2008), and w100% (Stott, 2002).
Terrestrial gastropods ingest water from multiple sources,
including dew, soil moisture, standing water, and precipitation, all
of which contain some amount of dissolved inorganic carbon (DIC).
Water is taken up through the foot of the gastropod by contact
rehydration (Balakrishnan and Yapp, 2004) and introduced to the
hemolymph before being passed on to the extrapallial fluid. Pigati
et al. (2004) found that aqueous carbon sources account for
w10% of the shell carbon for one species of Catinella, but it is not
known if this value is constant across the entire Succineidae family.
To our knowledge, data for other terrestrial taxa do not exist.
Finally, some terrestrial gastropods are able to scrape carbonate
rocks (limestone, dolomite, soil carbonate), and ingest the powder
or granules which then dissolve in their stomach acid to produce
CO2. As before, the dead carbon from the rocks is introduced to the
hemolymph, passed on to the extrapallial fluid, and ultimately
incorporated in the shell carbonate. Dead carbon from limestone
can account for up to w30% of the total carbon in shells of large
terrestrial gastropods. (Goodfriend and Stipp, 1983).
2.2. Effects of carbon sources on 14
C ages of shell carbonate
In most environments, the 14
C activities of live plants are in
equilibrium with atmospheric carbon. Gastropods that obtain their
shell carbon from live plants and the air, therefore, should yield
reliable 14
C ages, assuming they behave as closed systems after
burial (Fig. 3). Gastropods that consume organic detritus (i.e.,
decaying plant litter) typically do not pose a significant problem for
14
C dating because the time between plant death, its incorporation
into decomposition products, and consumption by gastropods is
usually quite short, on the order of a few yrs.
The 14
C activity of water that is available for consumption by
terrestrial gastropods (e.g., dew, standing water, precipitation) is at
or near equilibrium with atmospheric 14
C and, therefore, water is
unlikely to introduce a significant error to 14
C ages of terrestrial
gastropod shells. Exceptions include gastropods living directly
adjacent to springs that discharge waters from deeply-circulating
carbonate aquifers, lakes or rivers with significant hard water
effects, or in active volcanic areas where 14
C-deficient CO2 in
surface waters may be abundant (e.g., Riggs, 1984; Grosjean, 1994).
14
C ages of gastropods living in such areas should be evaluated
carefully.
The incorporation of 14
C from limestone and other carbonate
rocks can present a significant problem for 14
C dating of terrestrial
gastropod shells. The 14
C activity of atmospheric carbon, plants, and
water consumed by gastropods is essentially the same, w100%
modern carbon (pMC). In contrast, because most carbonate rocks
are of pre-Quaternary age, their 14
C activity is typically 0 pMC. Thus,
for 14
C dating, the magnitude of the potential error introduced by
carbonate rocks is a direct function of the amount of carbon from
rocks that is incorporated in the gastropod shell (Fig. 3). Unfortunately,
a simple correction that accounts for the incorporation of
14
C-deficient carbon in gastropod shells is not possible because we
cannot know a priori how much of the shell carbon was derived
from carbonate rocks versus other sources. Thus, to be confident in
14
C ages derived from terrestrial gastropod shells, it is imperative
that we identify and avoid taxa that incorporate dead carbon from
rocks altogether.
2.3. Open-system behavior
Even if some terrestrial gastropods consistently manage to avoid
the Limestone Problem regardless of the local geologic substrate or
environmental conditions, there is another hurdle that must be
overcome before we can confidently use their shells for 14
C dating.
That is, gastropod shells must remain closed systems with respect
to carbon after burial. For reliable 14
C dating, the pool of carbon
atoms measured during the 14
C dating process must consist solely
of carbon atoms that originally resided in the shell. Thus, following
burial, shells must resist the addition or exchange of 14
C atoms with
the local environment. Shells that exhibit open-system behavior
typically yield 14
C ages that are too young, and the magnitude of the
error depends upon the degree of such behavior (Fig. 3).
Previous work has suggested that 14
C ages from small terrestrial
gastropod shells recovered from fossil deposits in arid environments
may be reliable back to at least w30,000 14
C yrs B.P., but
a small degree of open-system behavior appears to compromise 14
C
dates obtained from shells older than this (Pigati et al., 2009). In
-8
-7
-6
-5
-4
-3
-2
-1
0
1
2
0 5 10 15 20 25 30 35 40
True 14
C age (ka)
eurtmorfnoitaiveD41
)ak(egaC
Fig. 3. Modeled deviation from the true 14
C age for four scenarios: (1) closed-system
behavior and no dead carbon (thick solid line), (2) closed-system behavior and 10%
dead carbon (dashed line), (3) open-system behavior equivalent to 1% modern carbon
contamination and 10% dead carbon (thin solid line), and (4) open-system behavior
equivalent to 1% modern carbon contamination and no dead carbon (dotted line).
J.S. Pigati et al. / Quaternary Geochronology 5 (2010) 519–532522
102
this study, we analyzed fossil shells from the American Midwest
because of their abundance in Quaternary deposits in the region,
the presence of multiple calcareous substrates (Paleozoic limestone,
calcareous till, and loess), and the humid climate (annual
precipitation in southwestern Ohio is w100 cm yrÀ1
). If fossil shells
exhibit even a small degree of open-system behavior in arid environments,
it may be exacerbated and, therefore, more easily
detected in humid environments where interaction between shells
and DIC in ground water is more prevalent.
3. Methods
3.1. Live gastropods
Previous ecological sampling by one of us (JCN) has resulted in
an extensive collection of modern terrestrial gastropods from North
America, constituting w250 taxa and over 470,000 individuals
from more than 1000 modern environments. Gastropods were
collected at each site from a representative 100–1000 m2
area by
hand collection of larger taxa and litter sampling of smaller taxa,
which provides the most complete assessment of site faunas
(Oggier et al., 1998; Cameron and Pokryszko, 2005). Collections
were made at places of high mollusc density, such as loosely
compacted leaf litter lying on top of highly compacted damp soil or
humus (Emberton et al., 1996). Litter was removed by hand and
sieved by shaking, tapping, or other agitation in the field using
a shallow sieve (ASTME #10; 2.0 mm mesh) nested inside a second
sieve (ASTME #30; 0.6 mm mesh). The process was continued for
15–60 min during which time 50–500 mg of material was collected
and retained.
Gastropods and detritus were dried at room temperature in the
laboratory and then hand-picked against a neutral background. All
shells, shell fragments, and slug plates were removed and identifiable
material was assigned to species using JCN’s reference
collection. Nomenclature generally follows that of Hubricht (1985)
with updates and corrections by Nekola (2004).
We selected 247 aliquots of shell material (3749 individual
shells) from 163 sites across the United States and southern Canada
for 14
C analysis. Nearly all of the specimens that we chose for
analysis were collected live, but at a few sites, only recently-dead
gastropods were available, which were identified by a translucent
appearance or the retention of color in the shells. Shells of small
gastropods that are dead for more than a year or so become
increasingly white and opaque with time (J. Nekola, unpublished
data), and were excluded from our study.
In the fossil record, species-level identification of fossil shells is
possible for most small terrestrial gastropods and, therefore, the
results of our investigation of modern gastropods can be applied
directly to the fossil record. An exception is the Succineidae family,
which is composed of three genera (Catinella, Oxyloma, and Succinea)
that are difficult to differentiate in modern faunas, let alone the
geologic record (Fig. 4). Their simple shells exhibit few diagnostic
characteristics and, therefore, species-level identification is based
on soft-body reproductive organ morphology, which is rarely
preserved in the fossil record. This presents a significant problem for
geochronologists; that is, can we be confident in 14
C ages derived
from shells from any taxon within the Succineidae family, or do we
need to target a specific genus or species? To address this issue, we
measured the 14
C activity of 100 aliquots of gastropod shell material
(802 individual shells) for twelve species of the Succineidae family to
determine the level of identification (i.e., family, genus, or species)
required to apply our results to the fossil record.
We prepared aliquots of modern shell material for 14
C analysis at
the University of Arizona Desert Laboratory (JSP) and Miami
University (JAR). We selected multiple shells at random for X-ray
diffraction (XRD) analysis using a Siemens Model D-500 diffractometer
to verify that only shell aragonite remained prior to
preparation for 14
C analysis. There was no evidence of primary or
secondary calcite in any of the shells that we analyzed. When
possible, shells were broken, the adhering soft parts were removed
using forceps, and the shells were treated with 6% NaOCl for 18–
24 h at room temperature to remove all remnants of organic matter.
Shells were not powdered during pretreatment to minimize the
potential for adsorption of atmospheric 14
C (Samos, 1949). We
selectively dissolved some of the shells by briefly introducing dilute
HCl to remove secondary carbonate (dust) from primary shell
material. Shells were washed repeatedly in ASTM Type 1, 18.2 MU
(hereafter ‘‘ultrapure’’) water, sonicated for a few seconds to
remove adhered solution, washed again with ultrapure water, and
dried in a vacuum oven overnight at w70 C.
Shell aragonite was converted to CO2 using 100% H3PO4 under
vacuum at either 50 or 75 C until the reaction was visibly complete
(w1 h). Water, SOx, NOx, and halide species were removed using
passive Cu and Ag traps held at w600 C and the resulting CO2 was
split into two aliquots. One aliquot was converted to graphite by
catalytic reduction of CO (modified after Slota et al., 1987) and
submitted to the Arizona-NSF Accelerator Mass Spectrometry
(AMS) facility for 14
C analysis. The second aliquot was submitted for
d13
C analysis in order to correct the measured 14
C activity of the
shell carbonate for isotopic fractionation.
14
C data for modern gastropods are presented as D14
C values in
per mil (Stuiver and Polach, 1977; Reimer et al., 2004) and analytical
uncertainties are reported at the 2s (95%) confidence level. d13
C
values are given in the usual delta (d) notation as the per mil
deviation from the VPDB standard. Analytical uncertainties for d13
C
measurements are less than 0.1& based on repeated measurements
of standards.
We also measured the 14
C content of several gastropod bodies
for comparison to their corresponding shells. The bodies were
treated with 10% HCl for 15–30 min to remove any remnants of the
carbonate shell, rinsed repeatedly, and dried in a vacuum oven at
Fig. 4. Photographs of the three genera of the Succineidae family: Catinella (left
panels), Oxyloma (middle panels), and Succinea (right panels); all are w10 mm in
length. The simple shells of the three genera contain few diagnostic characteristics
and, therefore, species-level identification is based on soft-body reproductive organ
morphology, which is rarely preserved in the fossil record.
J.S. Pigati et al. / Quaternary Geochronology 5 (2010) 519–532 523
103
w70 C. The dried bodies were placed in 6 mm quartz tubes along
with w100 mg of cupric oxide (CuO) and a small piece
(1 mm  5 mm) of silver foil, all of which were pre-combusted at
900 C for 4–6 h. The tube was evacuated, sealed with a glassblower’s
torch, and combusted offline at 900 C. The resulting CO2
gas was isolated and converted to graphite as above.
3.2. Modeling of 14
C values of gastropod diets
To quantify the amount of carbon in a gastropod shell that was
derived from limestone or other carbonate rocks, it is necessary to
compare the measured D14
C values of shells with D14
C values of the
gastropod diet. D14
C values of live plants consumed by gastropods
are identical to D14
C values of the atmosphere (Fig. 5a), which were
calculated using 14
C data averaged over the Northern Hemisphere
(Hua, 2004). We assigned a 5& uncertainty to the atmospheric
values to account for short-term, regional variations (Hsueh et al.,
2007) and small changes in atmospheric 14
C values that occur at
a given site during the short (usually annual) lifespan of the small
gastropods (Manning et al., 1990; Meijer et al., 1995).
Plant detritus from previous years’ primary production has
a higher 14
C activity than live plants because of the 14
C ‘‘bomb-spike’’
(Hua, 2004). Simply comparing the D14
C values of shells with the
D14
C of the atmosphere or live plants during the year that the
gastropod was alive, therefore, would ignore the potential impact of
detritus in the gastropod diet. We estimated the amount of detritus
from a given year that was available for consumption using a wide
range of carbon turnover rates (0.2–0.002 yrÀ1
; Fig. 5b), which are
applicable to O horizons and the upper few centimeters of A horizons
in which small terrestrial gastropods typically live (Guadinski et al.,
2000; Torn et al., 2005; Brovkin et al., 2008). We intentionally
chose a wide range of carbon turnover rates, which are equivalent to
mean residence times for carbon of 5–500 yrs, to encompass the
potentially wide range of detritus ages at the 163 localities included
in our study. We then used Monte Carlo simulation to generate
10,000 values for the D14
C of the gastropod diets for each of the past
13 yrs (the time span of ourcollections), which allowed two factors to
varyrandomly:thecarbonturnover rate(0.2–0.002yrÀ1
) andtheage
of the detritus fraction included in the diet (range of 0–1000 yrs). We
also ran simulations in which we let the age of the detritus vary up to
10,000 yrs, but the results did not change significantly.
We took the average and standard deviation of the 10,000
generated values as the gastropod diet D14
C value for each year of
collection between 1996 and 2008 (Fig. 5c). All uncertainties are
reported at the 2s (95%) confidence level. As expected, the modeled
D14
C value of the gastropod diet for a given year is slightly higher
than the atmospheric D14
C value for the same year. Uncertainties
associated with the modeled values are relatively large, on the
order of w35%, because of the large range of detritus D14
C values
that could be present at a given site.
Shells with D14
C values that are lower than the modeled D14
C
value of the gastropod diet during the year in which the gastropod
was alive indicate the presence of dead carbon from limestone or
other carbonate rocks. When applicable, the difference between the
D14
C values was converted into 14
C yrs to estimate the ‘‘limestone
effect’’, which represents the potential error introduced by the
incorporation of dead carbon in the shells. The magnitude of the
limestone effect should be considered a maximum value because
the calculation assumes that all of the dead carbon came from
carbonate rocks, rather than older (but not infinitely-aged) organic
matter. Because of the uncertainties associated with modeling the
D14
C of the gastropod diet, we are unable to discern limestone
effects smaller than w300 14
C yrs.
3.3. Fossil gastropods
Fossil gastropod shells were collected from glacial deposits at
the Oxford East glacial outcrops in southwestern Ohio as a case
study to determine if the shells remain closed systems with respect
to carbon over geologic timescales. These outcrops contain a series
of glacial diamictons from the Miami Lobe of the Laurentide Ice
0
50
100
150
200
250
1980 1985 1990 1995 2000 2005 2010
Year
41
)‰(erehpsomtafoC
0.0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1.0
0 200 400 600 800 1000
Time after plant death (years)
elbaliavasutirtedfonoitcarF
noitpmusnocrof 0.002 yr-1
0.2 yr-1
0.05 yr-1
0.02 yr-1
0.01 yr-1
0
50
100
150
200
250
1993 1998 2003 2008
Date of collection
41
)‰(teiddoportsagfoCΔΔ
a
b
c
Fig. 5. Modeling results for (a) the 14
C activity of the gastropod diet using measured
atmospheric values for the northern hemisphere (after Hua, 2004), (b) carbon turnover
rates (CTRs) ranging from 0.2 to 0.002 yrÀ1
, and (c) Monte Carlo simulation to generate
estimates of the D14
C values of gastropod diets for each of the past 13 yrs.
J.S. Pigati et al. / Quaternary Geochronology 5 (2010) 519–532524
104
Sheet that are separated by thin (3–5 cm) units of calcareous
organic-rich silt that contain gastropods, plant macrofossils, and
rooted tree stumps. AMS 14
C dating of plant macrofossils and in situ
tree stumps has shown that the age of this unit is between w20,100
and 21,400 14
C yrs B.P. (Eckberg et al., 1993; Lowell, 1995).
Gastropod taxa identified previously from this unit include Columella
columella, Discus cronkhitei, Euconulus fulvus, Hendersonia
occulta, Pupilla muscorum, Vertigo elatior, and multiple Succineidae
taxa (Dell, 1991).
Gastropod-bearing sediment was collected and placed in
deionized water with a deflocculant for several days to soften the
sediment enough to pass through a 0.5 mm sieve. A few samples
were placed in an ultrasonic bath for w1 h to further disaggregate
the sediment. Fossil shells were hand-picked from the retained
fraction, placed in a beaker of ultrapure water, subjected to an
ultrasonic bath for a few seconds, and then repeatedly dunked in
a second beaker of ultrapure water to remove sediment that
adhered to the shell surface or was lodged within the shell itself.
The recovered shells were broken and examined under a dissecting
microscope to ensure that the interior whorls were free of
secondary carbonate and detritus. Fossil shells that were free of
detritus were then processed for 14
C in the same manner as the
modern specimens, including random selection of shells for XRD
analysis. None of the fossil shells that we analyzed contained
measurable quantities of either primary or secondary calcite.
Organic samples, which included bark, charcoal, plant fragments,
and wood, were subjected to a standard acid–base–acid
(ABA) chemical pretreatment with 1N HCl (1 h at 60 C), 1N NaOH
(18–24 h at 60 C), and 1N HCl again (2–4 h at 60 C) before
combustion at 900 C in the presence of cupric oxide and silver foil.
The resulting CO2 was purified and converted to graphite in the
same manner as above.
Conventional radiocarbon ages are reported in 14
C yrs and, after
calibration, in calendar yrs. For calibration, we used the IntCal09.14C
dataset (CALIB 6.0.0, Stuiver and Reimer, 1993; Reimer et al., 2009).
4. Results
4.1. Modern shells
A few aliquots (11 of 247, or 4.5% of the total) yielded D14
C values
that were higher than the modeled dietary D14
C values of the
corresponding year of collection, which indicates these individuals
consumed unusually high amounts of bomb-spike carbon (Table S1,
Fig. 6). Data from these shells were excluded from further analysis.
For the remaining 236 aliquots of gastropod shells from the 46
different species that we analyzed, w78% did not contain any dead
carbon from limestone or other carbonate rocks even though it was
readily available at all sites, w12% contained between 5 and 10%
dead carbon, and a few (3% of the total) contained more than 10%
(Table S1, Fig. 6). D14
C values for all taxa ranged from À97.5 to
158.4&, and limestone effects averaged only w180 14
C yrs.
Dead carbon was not detected in the shells of at least 23
different species, including (number of shells in parentheses) Catinella
avara (99), Catinella gelida (66), Catinella vermeta (39),
Cochlicopa lubricella (17), Cochlicopa morseana (16), Columella
columella (94), Discus catskillensis (40), Discus cronkhitei (47), Discus
macclintockii (5), Euconulus alderi (71), Euconulus polygyratus (57),
Gastrocopta pentodon (131), Nesovitrea binneyana (46), Punctum
minutissimum (542), Pupilla hebes (7), Strobilops affinis (36), Succinea
bakeri (27), Succinea grosvernori (1), Succinea n. sp. ‘Minnesota
A’ (1), Succinea ovalis (50), Succinea strigata (14), Vertigo hubrichti
(144), and Vertigo modesta (73).
D14
C values were most negative (i.e., contained the most dead
carbon) for shells from the Pupilla and Vallonia genera. Maximum
limestone effects for these genera ranged from 780 Æ 310 14
C yrs for
P. muscorum to 1590 Æ 280 14
C yrs for P. sonorana, and from
1010 Æ 380 14
C yrs for Vallonia perspectiva to 1500 Æ 270 14
C yrs for
V. cyclophorella, respectively (Table 1). The only other species that
exhibited a limestone effect that was greater than 1000 14
C yrs was
H. occulta (1210 Æ 250 14
C yrs).
These results can be applied directly to the fossil record if it is
possible to identify the taxa to the species-level based on shell
morphology. For the Succineidae family, the data must be evaluated
at the family or genus level to be applicable. Taking the Succineidae
family as a whole, 85% of the shell aliquots that we analyzed did not
contain measurable amounts of dead carbon; the remaining aliquots
contained an average of 5.2% dead carbon. Within the family, D14
C
values for the genus Catinella ranged from 63.7 to 147.0&, Oxyloma
values ranged from À0.8 to 135.0&, and Succineavalues ranged from
16.1 to 147.8&. Members of the Catinella genus incorporate little, if
any, dead carbon from limestone or other carbonate rocks in their
shells; 32 of 33 aliquots (97%) of Catinella shells did not contain
measurable amounts of dead carbon. The remaining sample
(DL-170) contained only a very small amount of dead carbon,
equivalent to a limestone effect of 320 Æ 310 14
C yrs. Similarly, 36 of
39 aliquots (92%) of Succinea shells did not contain dead carbon. The
remaining three aliquots were all Succinea indiana; limestone effects
for these shells ranged from 430 Æ 370 to 610 Æ 380 14
C yrs. For
Oxyloma, 15 of 24 aliquots (63%) did not contain measurable
amounts of dead carbon. Limestone effects for the remaining
samples ranged between 250 Æ 210 and 670 Æ 240 14
C yrs.
4.2. Modern gastropod bodies
We also measured the 14
C activity of the bodies of eight gastropods
to determine the magnitude of the offset between the body
carbon and shell carbonate (Table 2). Ideally, we would have
preferred to measure the 14
C content of the extrapallial fluid to
comparewith the shell carbonate todetermine if carbon isotopes are
fractionated when the shells are formed, but this was not feasible
because the gastropod bodies were simply too small. Regardless, the
measured D14
C values of gastropod body carbon ranged from 42 to
101&, similar to the D14
C values of shell carbonate from the same
sites, which ranged from 43 to 133&. However, we did not observe
a clear relation between the 14
C activity of gastropod body carbon
and shell carbonate. D14
C values of bodies of Oxyloma retusa
collected from Maquokata River Mounds, Iowa were indistinguishable
from the D14
C values of their corresponding shells
(D14
Cbody ¼ 42 Æ 6&, D14
Cshell ¼ 43 Æ 10&), as were values for S.
ovalis from Dave Pepin Homestead, Minnesota (D14
Cbody ¼ 101 Æ8&,
D14
Cshell ¼ 110 Æ 8&). In contrast, D14
C values for bodies of S. ovalis
were significantly lower than their shells from Brewer Boat Ramp,
Maine (D14
Cbody ¼ 55 Æ 6&, D14
Cshell ¼ 82 Æ 5&) and Zippel Bay
State Park, Minnesota (D14
Cbody ¼ 84 Æ 6&, D14
Cshell ¼ 133 Æ 12&).
The reason(s) for this difference is unclear.
4.3. Fossil shells
Well-preserved fossil organic material (bark, plant macrofossils,
and wood) recovered from sediments at the Oxford East outcrops
yielded calibrated ages that ranged from 24.60 Æ 0.40 to
25.28 Æ 0.55 ka, with an average of 24.93 Æ 0.30 ka (n ¼ 5; Table 3,
Fig. 7). Gastropod shells recovered from the same stratigraphic unit
yielded ages that ranged from 23.92 Æ 0.66 to 25.81 Æ 0.94 ka, and
averaged 24.73 Æ 0.44 ka (n ¼ 18). Average ages of six of the seven
fossil taxa were indistinguishable from the average age of the
organic matter: Discus shimeki (24.80 Æ 0.17 ka; n ¼ 3), P. muscorum
(24.34 Æ 0.32 ka; n ¼ 3), Vallonia gracilicosta (24.20 Æ 0.40 ka;
n ¼ 2), Vertigo hannai (24.90 Æ 0.77 ka; n ¼ 1), V. modesta
J.S. Pigati et al. / Quaternary Geochronology 5 (2010) 519–532 525
105
(24.86 Æ 0.32 ka; n ¼ 3), and Succineidae (25.29 Æ 0.54 ka; n ¼ 3).
The average age of H. occulta (24.34 Æ 0.15 ka; n ¼ 3) was slightly
younger than the organic ages.
5. Discussion
5.1. Small terrestrial gastropods and the limestone problem
Approximately 78% of the modern shells that we analyzed did
not contain any dead carbon from limestone or other carbonate
rocks even though it was readily available at all sites, w12% of the
aliquots contained between 5 and 10% dead carbon, and a few (3%
of the total) contained more than 10%. Even at the high end, the
amount of dead carbon in the small shells is significantly less than
the 20–30% dead carbon that has been previously reported for
larger taxa (e.g., Goodfriend and Stipp, 1983).
If we extrapolate our results to the fossil record and assume that
the shells behave as closed systems with respect to carbon over
geologic timescales, then small terrestrial gastropod shells should
provide accurate 14
C ages w78% of the time and ages that are
a
Fig. 6. Shell carbonate D14
C values compared to modeled dietary D14
C values for modern gastropods. Data points that fall on the solid black line in each panel represent gastropods
that obtained their carbon from live plants and the atmosphere. Data points that fall below the solid line indicate that dead carbon from limestone or other carbonate rocks was
incorporated during shell construction. The magnitude of this phenomenon, called the ‘‘limestone effect’’, depends upon the amount of shell carbon that was derived from rocks as
shown by the dashed lines.
J.S. Pigati et al. / Quaternary Geochronology 5 (2010) 519–532526
106
within w1000 14
C yrs of the true age w97% of the time. Shells from
at least 23 different species did not contain dead carbon, and
therefore should yield reliable 14
C ages if the modern shell data can
be applied directly to the fossil record.
For the Succineidae family as a whole, 85% of the shell aliquots
that we analyzed did not contain measurable amounts of dead
carbon; the remaining aliquots contained an average of 5.2% dead
carbon, equivalent to a limestone effect of 425 14
C yrs. At the
genus level, shells of the genus Catinella should yield reliable 14
C
ages w97% of the time, again assuming closed-system behavior,
and ages that are within w300 14
C yrs of the true age every time.
Similarly, Succinea shells should yield reliable 14
C ages w92% of
the time and ages that are within w600 14
C yrs of the true age
every time. Results for the genus Oxyloma suggest that some
caution should be used when evaluating 14
C ages derived from
these shells. AMS results for the Oxyloma shells show that nearly 1
in 3 aliquots contained at least some dead carbon. Although the
amount was relatively minor, <7% of the total, dead carbon was
present in Oxyloma shells more frequently than in either Catinella
or Succinea shells.
b
Fig. 6. (continued).
J.S. Pigati et al. / Quaternary Geochronology 5 (2010) 519–532 527
107
5.2. Ca-limiting hypothesis
Large gastropod shells (>20 mm in maximum dimension)
routinely contain 20–30% dead carbon when living in habitats on
carbonate terrain (e.g., Evin et al.,1980; Goodfriend and Stipp,1983),
whereas the small shells measured in this study rarely contained
more than w10%. We speculate that calcium may hold the clues to
determining the reasons for the difference. Gastropod shell carbonate
(aragonite) is composed of three elements – calcium, carbon, and
oxygen. The latter two elements are readily available in the environments
in which gastropods live and, therefore, cannot be considered
as possible limiting factors for shell construction. In most settings,
however, calcium is present inplants and water in lowconcentrations
(typically parts per million). If small terrestrial gastropods can acquire
enough calcium from their‘‘normal’’diet (plants, detritus, and water),
then they may not have to consume carbonate rocks to supplement
theircalcium intakewhen buildingtheir shells.Largertaxamay find it
more difficult to obtain enough calcium from these sources without
turning to carbonate rocks when they are available.
Our results support this hypothesis, but only on a gross scale.
There is clearly a significant difference in the amount of dead
carbon incorporated in the shells of large taxa previously studied
and the small taxa studied here. However, shell size alone is not the
only factor to consider when evaluating the results within the small
body size class. For example, in the present study, we did not
observe a significant correlation between shell size and measured
D14
C values (R2
¼ 0.015). The largest taxon that we included in our
analysis, H. occulta, averaged 15.6 mg per shell and contained
approximately the same amount of dead carbon as P. muscorum and
Vallonia shells, which averaged only 1.4 and 0.7 mg per shell,
respectively. Similarly, we did not find a clear correlation between
shell size and measured D14
C values even within a single family. For
Succineidae, Catinella shells were generally the smallest, averaging
1.1 mg per shell and contained the least amount of dead carbon, and
Table 1
Summary of 14
C results for modern gastropod shells.
Limestone effecta
(14
C yrs)
Family Genus Species Aliquots Shells Negligibleb
Maximumc
Cochlicopidae Cochlicopa Cochlicopa lubrica 3 17 53% 650 Æ 390
Cochlicopa lubricella 3 17 100% –
Cochlicopa morseana 4 16 100% –
Discidae Discus Discus catskillensis 6 40 100% –
Discus cronkhitei 6 47 100% –
Discus macclintockii 3 5 100% –
Discus shimeki 7 33 97% 430 Æ 240
Helicarionidae Euconulus Euconulus alderi 3 71 100% –
Euconulus dentatus 3 34 68% 670 Æ 290
Euconulus fulvus 8 107 57% 730 Æ 300
Euconulus polygyratus 3 57 100% –
Helicinidae Hendersonia Hendersonia occulta 13 13 46% 1210 Æ 250
Punctidae Punctum Punctum minutissimum 3 542 100% –
Pupillidae Columella Columella columella 3 94 100% –
Gastrocopta Gastrocopta pentodon 3 131 100% –
Gastrocopta tappaniana 3 105 92% 540 Æ 360
Pupilla Pupilla blandi 9 90 67% 1000 Æ 270
Pupilla hebes 1 7 100% –
Pupilla muscorum 8 80 38% 780 Æ 310
Pupilla sonorana 2 29 66% 1590 Æ 280
Vertigo Vertigo elatior 3 113 38% 500 Æ 280
Vertigo hannai 3 123 65% 280 Æ 270
Vertigo hubrichti 3 144 100% –
Vertigo modesta 3 73 100% –
Vertigo paradoxa 3 124 74% 380 Æ 240
Strobilopsidae Strobilops Strobilops affinis 3 36 100% –
Strobilops labyrinthica 3 43 65% 520 Æ 320
Succineidae Catinella Catinella avara 9 99 100% –
Catinella exile 15 316 97% 320 Æ 310
Catinella gelida 9 66 100% –
Catinella vermeta 4 39 100% –
Oxyloma Oxyloma retusa 17 136 65% 540 Æ 360
Oxyloma verrilli 7 47 57% 670 Æ 240
Succinea Succinea bakeri 8 27 100% –
Succinea grosvernori 1 1 100% –
Succinea indiana 4 6 17% 610 Æ 380
Succinea n.sp. ‘Minnesota A’ 1 1 100% –
Succinea ovalis 19 50 100% –
Succinea strigata 6 14 100% –
Valloniidae Vallonia Vallonia cyclophorella 6 129 95% 1500 Æ 270
Vallonia gracilicosta 7 155 60% 1370 Æ 270
Vallonia perspectiva 5 221 68% 1010 Æ 380
Zonitidae Hawaiia Hawaiia minuscula 3 152 29% 340 Æ 270
Nesovitrea Nesovitrea binneyana 3 46 100% –
Nesovitrea electrina 4 33 73% 710 Æ 240
Zonitoides Zonitoides arboreus 4 20 85% 530 Æ 240
a
Defined as the theoretical difference between the measured and true 14
C ages for gastropods that incorporate the same amount of dead carbon in their shells as the
aliquots measured here. These values are based on the difference between the modeled diet and shell carbonate D14
C values and converted into 14
C yrs.
b
Percent of shells measured by AMS that did not contain dead carbon from limestone or other carbonate rocks (i.e., the D14
C values for the shells were statistically
indistinguishable from the modeled diet D14
C value).
c
Maximum limestone effect for a given taxon (given in 14
C yrs). Uncertainties are given at the 2s (95%) confidence level.
J.S. Pigati et al. / Quaternary Geochronology 5 (2010) 519–532528
108
Oxyloma shells averaged 2.1 mg per shell and contained the most
dead carbon. Succinea shells were the largest, averaging 5.8 mg per
shell, but were between Catinella and Oxyloma in terms of the
amount of dead carbon in their shells (Table S1).
The results presented here suggest that the Limestone Problem
for small terrestrial gastropods is often negligible and always
much less than the 20–30% dead carbon for larger taxa. However,
there are additional factors that apparently influence the dietary
intake of carbonate rocks of small terrestrial gastropods living
side by side, which may include opportunistic behavior, variations
in microhabitats, and the dietary needs or wants of individual
gastropods.
Table 2
14
C results for modern gastropod bodies and corresponding shells.
Lab # AA # Taxon Sitea
Lat (
N) Long (
W) Mass (mg) d13
C (vpdb) F14
Cb
Shell D14
C Atmos D14
C Diet D14
C Limestone Effectc
(14
C yrs)
Bodies
MU-109 80177 Succinea ovalis 1 44.819 68.723 4.22 À24.9 1.0547 Æ 0.0064 55 Æ 6 70 Æ 5 96 Æ 33 330 Æ 280
MU-114 80181 Succinea ovalis 2 48.410 94.819 8.64 À24.8 1.0996 Æ 0.0082 101 Æ 8 80 Æ 5 107 Æ 36 50 Æ 300
MU-108 80176 Succinea ovalis 3 48.906 96.027 3.79 À25.1 1.0946 Æ 0.0062 96 Æ 6 80 Æ 5 107 Æ 36 90 Æ 300
MU-111 80178 Succinea ovalis 4 47.874 96.422 11.05 À26.0 1.0877 Æ 0.0062 89 Æ 6 80 Æ 5 107 Æ 36 150 Æ 300
MU-117 80183 Oxyloma retusa 5 42.559 90.713 9.04 À26.1 1.0419 Æ 0.0064 42 Æ 6 52 Æ 5 73 Æ 25 250 Æ 210
MU-115 80182 Succinea ovalis 6 50.264 66.411 12.88 À25.0 1.0485 Æ 0.0072 49 Æ 7 57 Æ 5 79 Æ 27 250 Æ 230
MU-112 80179 Succinea strigata 7 64.858 147.862 4.01 À24.2 1.0740 Æ 0.0060 74 Æ 6 54 Æ 5 75 Æ 26 0 Æ 210
MU-113 80180 Succinea ovalis 8 48.866 94.843 6.88 À23.5 1.0833 Æ 0.0060 84 Æ 6 80 Æ 5 107 Æ 36 180 Æ 300
Shells
MU-162 80928 Succinea ovalis 1 44.819 68.723 8.80 À10.0 1.0817 Æ 0.0048 82 Æ 5 70 Æ 5 96 Æ 33 110 Æ 270
MU-128 80194 Succinea ovalis 2 48.410 94.819 11.72 À9.8 1.1090 Æ 0.0084 110 Æ 8 80 Æ 5 107 Æ 36 0 Æ 300
MU-179 80942 Oxyloma retusa 5 42.559 90.713 13.08 À11.0 1.0425 Æ 0.0100 43 Æ 10 52 Æ 5 73 Æ 25 250 Æ 220
MU-141 80910 Succinea ovalis 8 48.866 94.843 8.91 À9.6 1.1316 Æ 0.0116 133 Æ 12 80 Æ 5 107 Æ 36 0 Æ 300
Uncertainties are given at the 2s (95%) confidence level.
a
Key to sites: 1 ¼ Brewer Boat Ramp, Maine; 2 ¼ Dave Pepin Homestead, Minnesota; 3 ¼ Duxby, Minnesota; 4 ¼ Huot Forest WMA, Minnesota; 5 ¼ Maquokata River
Mounds, Iowa; 6 ¼ September Islands, Quebec; 7 ¼ University of Alaska – Fairbanks; 8 ¼ Zippel Bay State Park, Minnesota.
b
F14
C values are derived from the measured 14
C activity, corrected for fractionation, and account for decay that occurred between the time of collection and the AMS
measurement.
c
Defined in Table 1.
Table 3
14
C results for the Oxford East outcrops.
Lab # AA # Taxon Na
Mass (mg) d13
C (vpdb) F14
C 14
C age (ka) Calendar age (ka)b
Pc
Organics
MU-212 82584 bark – 4.47 À24.6 0.0731 Æ 0.0026 21.02 Æ 0.28 25.15 Æ 0.34 1.00
MU-213 82585 bark – 3.28 À25.1 0.0767 Æ 0.0025 20.62 Æ 0.27 24.64 Æ 0.37 1.00
MU-214 82586 twig – 3.80 À24.9 0.0721 Æ 0.0028 21.13 Æ 0.32 25.28 Æ 0.55 1.00
MU-211 82583 wood – 3.82 À22.6 0.0743 Æ 0.0042 20.88 Æ 0.45 24.97 Æ 0.63 1.00
MU-210 82582 wood chip – 4.29 À23.6 0.0772 Æ 0.0025 20.58 Æ 0.26 24.60 Æ 0.40 1.00
Average 24.93 Æ 0.30
Gastropod shells
MU-194 82567 Discus shimeki 2 9.20 À6.4 0.0758 Æ 0.0052 20.72 Æ 0.55 24.73 Æ 0.75 1.00
MU-195 82568 Discus shimeki 2 10.94 À6.9 0.0762 Æ 0.0052 20.68 Æ 0.54 24.67 Æ 0.74 1.00
MU-196 82569 Discus shimeki 2 14.99 À6.4 0.0745 Æ 0.0053 20.87 Æ 0.57 25.00 Æ 0.79 1.00
Average 24.80 Æ 0.17
MU-188 82561 Hendersonia occulta 1 12.41 À6.4 0.0805 Æ 0.0051 20.24 Æ 0.51 24.18 Æ 0.67 1.00
MU-189 82562 Hendersonia occulta 1 10.29 À6.1 0.0787 Æ 0.0055 20.42 Æ 0.56 24.37 Æ 0.68 1.00
MU-190 82563 Hendersonia occulta 1 18.42 À7.1 0.0777 Æ 0.0055 20.52 Æ 0.57 24.47 Æ 0.70 1.00
Average 24.34 Æ 0.15
MU-199 82572 Pupilla muscorum 8 10.86 À6.5 0.0760 Æ 0.0052 20.71 Æ 0.55 24.71 Æ 0.74 1.00
MU-200 82573 Pupilla muscorum 9 11.88 À6.4 0.0805 Æ 0.0052 20.24 Æ 0.52 24.18 Æ 0.68 1.00
MU-201 82574 Pupilla muscorum 10 14.13 À6.4 0.0809 Æ 0.0051 20.20 Æ 0.51 24.13 Æ 0.67 1.00
Average 24.34 Æ 0.32
MU-191 82564 Succineidae 8 10.58 À5.3 0.0712 Æ 0.0052 21.23 Æ 0.59 25.33 Æ 0.80 1.00
MU-192 82565 Succineidae 4 9.32 À5.5 0.0689 Æ 0.0056 21.49 Æ 0.65 25.81 Æ 0.94 1.00
MU-193 82566 Succineidae 15 9.53 À5.8 0.0759 Æ 0.0055 20.71 Æ 0.58 24.73 Æ 0.78 1.00
Average 25.29 Æ 0.54
MU-202 82575 Vallonia gracilicosta 15 10.72 À6.0 0.0776 Æ 0.0058 20.53 Æ 0.60 24.49 Æ 0.76 1.00
MU-203 82576 Vallonia gracilicosta 15 11.08 À6.2 0.0824 Æ 0.0058 20.05 Æ 0.56 23.92 Æ 0.66 0.98
Average 24.20 Æ 0.40
MU-205 82577 Vertigo hannai 30 10.46 À6.5 0.0750 Æ 0.0052 20.81 Æ 0.55 24.90 Æ 0.77 1.00
MU-206 82578 Vertigo modesta 15 11.47 À6.7 0.0751 Æ 0.0052 20.80 Æ 0.55 24.88 Æ 0.77 1.00
MU-207 82579 Vertigo modesta 15 14.14 À7.0 0.0771 Æ 0.0053 20.59 Æ 0.55 24.54 Æ 0.71 1.00
MU-208 82580 Vertigo modesta 20 15.28 À6.7 0.0729 Æ 0.0053 21.04 Æ 0.58 25.18 Æ 0.77 1.00
Average 24.86 Æ 0.32
Uncertainties for the raw and calibrated 14
C ages are given at the 2s (95%) confidence level.
a
Number of shells per aliquot.
b
Calibrated ages were calculated using CALIB v. 6.0.0, IntCal09.14C dataset; limit 50.0 calendar ka B P. Calibrated ages are reported as the midpoint of the calibrated range.
Uncertainties are reported as the difference between the midpoint and either the upper or lower limit of the calibrated age range, whichever is greater. Multiple ages are
reported when the probability of a calibrated age range exceeds 0.05.
c
P ¼ probability of the calibrated age falling within the reported range as calculated by CALIB.
J.S. Pigati et al. / Quaternary Geochronology 5 (2010) 519–532 529
109
5.3. Small terrestrial gastropods and open-system behavior
The present study includes only a single fossil locality, the Oxford
East outcrops of southwestern Ohio, which we present here as
a case study. Average ages of organic materials and fossil gastropod
shells from the Oxford East outcrops were statistically indistinguishable;
24.93 Æ 0.30 ka for the organics and 24.73 Æ 0.44 ka for
the shells. Six of the seven taxa that we analyzed (D. shimeki, P.
muscorum, Succineidae, V. gracilicosta, V. hannai, and V. modesta)
yielded average 14
C ages that are indistinguishable from the organic
ages; the seventh (H. occulta) yielded ages that were slightly
younger than the organic ages. The range of ages of the shell
material, 1.8 ka, is significantly larger than the range of ages of the
organics, 0.8 ka. If the dispersion of shell ages was related to opensystem
behavior, then we would expect the ages to be systematically
younger than the organic matter ages, which they are not. It
may be that the small number of organic samples fails to adequately
capture the full range of time represented by the sampled stratum.
More work needs to be done on this subject, including additional
comparisons of shell and organic ages from other sites, but the
results from the Oxford East outcrops suggest that small terrestrial
gastropod shells may behave as closed systems with respect to
carbon in the American Midwest for at least the past w25 ka.
6. Summary and conclusions
Fossil shells of small terrestrial gastropods are commonly
preserved in Quaternary sediment across North America, including
loess,wetland, glacial,andalluvial deposits, aswellas in sedimentsat
many archeological sites. Their aragonitic shells contain w12% by
weight carbon, and therefore contain sufficient carbon for 14
C dating.
However, terrestrial gastropod shells in carbonate terrains are often
avoided for 14
C dating because large taxa are known to incorporate
dead carbon from limestone or other carbonate rocks when building
their shells, which can cause their 14
C ages to be up to 3000 yrs too
old. Previous studies suggested that small terrestrial gastropod shells
may yield reliable 14
C agesin arid environments, but a systematic and
comprehensive analysis was needed before ages derived from their
shells could be considered reliable outside of the Desert Southwest.
To this end, we measured the 14
C activity of 247 aliquots of
modern shell material (3749 individual shells) from 163 localities
across North America. Approximately 78% of the aliquots did not
contain measurable amounts of dead carbon even though limestone
orother carbonate rocks were readilyavailable at all sites, w12 of the
aliquots contained between 5 and 10% dead carbon, and the
remaining few (3% of the total) contained more than 10%. The
average Limestone Effect for these samples was only w180 14
C yrs,
which is significantly less than the 2000–3000 14
C yrs that previous
researchers found for larger taxa. Assuming that the small gastropod
shells behave as closed systems with respect to carbon after burial,
they should yield reliable 14
C ages w78% of the time, and ages that
are within w1000 yrs of the true age w97% of the time, regardless of
the taxon analyzed, local bedrock type, climate, or environmental
conditions. If fossil shells can be identified to the species level, then
at least 23 different species should yield reliable 14
C ages if the
modern shell data can be applied directly to the fossil record.
The terrestrial gastropod family Succineidae is one of the most
common gastropod taxa in North America. Unlike the other
Fig. 7. Photograph of the section at the Oxford East outcrops and the calibrated ages obtained from the organic material and fossil gastropod shells.
J.S. Pigati et al. / Quaternary Geochronology 5 (2010) 519–532530
110
gastropods studied here, our 14
C data for Succineidae must be
evaluated at the genus or even family level because species-level
identification is based on soft-part morphology, which is rarely
preserved in the fossil record. Based on the data from modern
shells, the Succineidae family as a whole should yield reliable 14
C
ages w85% of the time, and ages that are within w700 14
C yrs every
time. At the genus level, Catinella should yield reliable 14
C ages
w97% of the time, again assuming closed-system behavior, and
ages that are within w300 14
C yrs of the true age every time.
Similarly, Succinea shells should yield reliable 14
C ages w92% of the
time and ages that are within w600 14
C yrs of the true age every
time. Caution should be used when evaluating shells of the genus
Oxyloma, however, as nearly 1 in 3 aliquots contained dead carbon,
equivalent to a limestone effect of up to w700 14
C yrs.
Fossil shells of small terrestrial gastropods recovered from welldated,
late-Pleistocene sediments in the Midwest yielded ages that
were statistically indistinguishable from ages obtained from wellpreserved
plant macrofossils (wood, bark, plant remains).
Although just one site, these results suggest that small terrestrial
gastropod shells may behave as closed systems with respect to
carbon over geologic timescales. More work on this subject is
needed, but if our case study site is representative of other sites,
then fossil shells of some small terrestrial gastropods, including at
least five common genera, Catinella, Columella, Discus, Gastrocopta,
and Succinea, should yield reliable 14
C ages, regardless of the local
geologic substrate. Fossil shells of these and other small terrestrial
gastropods are common in a wide range of Quaternary deposits in
North America and, therefore, our results may have broad chronologic
applications to Quaternary geology and New World
archeology.
Acknowledgments
We thank K. Gotter for converting carbon dioxide samples to
graphite at the JAR’s lab at Miami University and J. Quade and the
University of Arizona Desert Laboratory for access to their vacuum
extraction systems. We also thank J. Rosenbaum and D. Van Sistine
of the U.S. Geological Survey for their help with Figs. 1 and 2,
respectively. This manuscript benefited from constructive reviews
from E. Ellis, G. Hodgins, D. Muhs, M. Reheis, J. Southon, and two
anonymous reviewers. This project was funded by the National
Science Foundation Sedimentary Geology and Paleobiology
Competition, award #EAR 0614840. Additional support was
provided to JSP by the U.S. Geological Survey’s Mendenhall postdoctoral
research program.
Appendix. Supporting information
Supplementary data associated with this article can be found in
the online version at doi:10.1016/j.quageo.2010.01.001.
Editorial Handling by: A. Hogg
References
Balakrishnan, M., Yapp, C.J., 2004. Flux balance models for the oxygen and carbon
isotope compositions of land snail shells. Geochimica et Cosmochimica Acta 68,
2007–2024.
Barker, G.M., 2001. Gastropods on land: phylogeny, diversity, and adaptive
morphology. In: Barker, G.M. (Ed.), Biology of Terrestrial Molluscs. CABI
Publishing, Oxon, U.K, pp. 1–146.
Brennan, R., Quade, J., 1997. Reliable Late-Pleistocene stratigraphic ages and shorter
groundwater travel times from 14
C in fossil snails from the southern Great
Basin. Quaternary Research 47, 329–336.
Brovkin, V., Cherkinsky, A., Goryachkin, S., 2008. Estimating soil carbon turnover
using radiocarbon data: a case study for European Russia. Ecological Modelling
216, 178–187.
Burch, J.B., Pearce, T.A., 1990. Terrestrial Gastropoda. Soil Biology Guide. John Wiley
& Sons, New York.
Cameron, R.A.D., Pokryszko, B.M., 2005. Estimating the species richness and
composition of land mollusc communities. Journal of Conchology 38, 529–547.
Dell, A.M., 1991. Reconstruction of late Pleistocene paleoecology in southwestern
Ohio from nonmarine gastropod assemblages. Unpublished M.S., University of
Cincinnati, 97 p.
Eckberg, M.P., Lowell, T.V., Stuckenrath, R., 1993. Late Wisconsin glacial advance and
retreat patterns in southwestern Ohio, USA. Boreas 22, 189–204.
Emberton, K.C., Pearce, T.A., Randalana, R., 1996. Quantitatively sampling land-snail
species richness in Madagascan rainforests. Malacologia 38, 203–212.
Evans, J.G., 1972. Land Snails in Archaeology. Seminar Press, New York, NY.
Evin, J., Marechal, J., Pachiaudi, C., 1980. Conditions involved in dating terrestrial
shells. Radiocarbon 22, 545–555.
Frye, J.C., Willman, H.B., 1960. Classification of the Wisconsinan stage in the Lake
Michigan glacial lobe. Illinois Geological Survey Circular 285, 1–16.
Goodfriend, G.A., 1987. Radiocarbon age anomalies in shell carbonate of land snails
from semi-arid areas. Radiocarbon 29, 159–167.
Goodfriend, G.A., Hood, D.G., 1983. Carbon isotope analysis of land snail shells: implications
for carbon sources and radiocarbon dating. Radiocarbon 25, 810–830.
Goodfriend, G.A., Stipp, J.J., 1983. Limestone and the problem of radiocarbon dating
of land-snail shell carbonate. Geology 11, 575–577.
Goslar, T., Pazdur, M.F., 1985. Contamination studies on mollusk shell samples.
Radiocarbon 27, 33–42.
Grosjean, M., 1994. Paleohydrology of the Laguna Lejı´a (North Chilean Altiplano)
and climatic implications for late-glacial times. Palaeogeography, Palaeoclimatology,
Palaeoecology 109, 89–100.
Guadinski, J.B., Trumbore, S.E., Davidson, E.E., Zheng, S., 2000. Soil carbon cycling in
a temperate forest: radiocarbon-based estimates of residence times, sequestration
rates, and partitioning of fluxes. Biogeochemistry 51, 33–69.
Hsueh, D.Y., Karkauer, N.Y., Randerson, J.T., Xu, X., Trumbore, S.E., Southon, J.R.,
2007. Regional patterns of radiocarbon and fossil fuel-derived CO2 in surface air
across North America. Geophysical Research Letters 34, L02816. doi:10.1029/
2006GL027032.
Hua, Q., 2004. Review of tropospheric bomb 14
C data for carbon cycle modeling and
age calibration purposes. Radiocarbon 46, 1273–1298.
Hubricht, L., 1985. The distributions of the native land mollusks of the eastern
United States. Fieldiana: Zoology 24, 1–191.
Kerney, M., Cameron, R.A.D., 1979. Field Guide to the Land Snails of Britain and
Northwestern Europe. Collins Press, London.
Leighton, M.M., 1960. The classification of the Wisconsin Glacial stage of north
central United States. Journal of Geology 68, 529–552.
Lowell, T.V., 1995. The Application of Radiocarbon Age Estimates to the Dating of
Glacial Sequences: an Example from the Miami Sublobe, Ohio, USA. Quaternary
Science Reviews 4, 85–99.
Manning, M.R., Lowe, D.C., Melhuish, W.H., Sparks, R.J., Wallace, G.,
Brenninkmeijer, C.A.M., McGill, R.C., 1990. The use of radiocarbon measurements
in atmospheric studies. Radiocarbon 32, 37–58.
Meijer, H.A.J., van der Plicht, J., Gislefoss, J.S., Nydal, R., 1995. Comparing long term
atmospheric 14
C and 3
H records near Groningen, the Netherlands with Fruholmen,
Norway and Izan˜a, Canary Islands 14
C stations. Radiocarbon 37, 39–50.
Nekola, J.C., 2004. Terrestrial gastropod fauna of northeastern Wisconsin and the
southern Upper Peninsula of Michigan. American Malacological Bulletin 18,
21–44.
Nekola, J.C., 2005. Geographic variation in richness and shell size of eastern North
American land snail communities. Records of the Western Australian Museum
Supplement 68, 39–51.
Oggier, P., Zschokke, S., Baur, B., 1998. A comparison of three methods for assessing
the gastropod community in dry grasslands. Pedobiologia 42, 348–357.
Pedone, V., Rivera, K., 2003. Groundwater-discharge deposits in Fenner Wash,
eastern Mojave Desert. Geological Society of America Abstracts with Programs
35, 257.
Pigati, J.S., Bright, J.E., Shanahan, T.M., Mahan, S.A., 2009. Late Pleistocene
Paleohydrology Near the Boundary of the Sonoran and Chihuahuan Deserts,
Southeastern Arizona, USA. Quaternary Science Reviews 28, 286–300.
Pigati, J.S., Quade, J., Shanahan, T.M., Haynes Jr., C.V., 2004. Radiocarbon dating of
minute gastropods and new constraints on the timing of spring-discharge
deposits in southern Arizona, USA. Palaeogeography, Palaeoclimatology,
Palaeoecology 204, 33–45.
Placzek, C., Quade, J., Patchett, P.J., 2006. Geochronology and stratigraphy of late
Pleistocene lake cycles on the southern Bolivian Altiplano: implications for
causes of tropical climate change. Geological Society of America Bulletin 118,
515–532.
Ponder, W.F., Lindberg, D.R., 1997. Towards a phylogeny of gastropod molluscs: an
analysis using morphological characters. Zoological Journal of the Linnean
Society 119, 83–265.
Quarta, G., Romaniello, L., D’Elia, M., Mastronuzzi, G., Calcagnile, L., 2007. Radiocarbon
age anomalies in pre- and post-bomb land snails from the coastal
Mediterranean basin. Radiocarbon 49, 817–826.
Reimer, P., Baillie, M., Bard, E., Bayliss, A., Beck, J., Blackwell, P., Ramsey, C.B., Buck, C.,
Burr, G., Edwards, R., Friedrich, M., Grootes, P., Guilderson, T., Hajdas, I.,
Heaton, T., Hogg, A., Hughen, K., Kaiser, K., Kromer, B., McCormac, F.,
Manning, S., Reimer, R., Richards, D., Southon, J., Talamo, S., Turney, C.,
van der Plicht, J., Weyhenmeyer, C., 2009. IntCal09 and Marine09 radiocarbon
age calibration curves, 0–50,000 years cal B.P. Radiocarbon 51, 1111–1150.
J.S. Pigati et al. / Quaternary Geochronology 5 (2010) 519–532 531
111
Reimer, P.J., Brown, T.A., Reimer, R.W., 2004. Discussion: reporting and calibration of
post-bomb 14
C data. Radiocarbon 46, 1299–1304.
Riggs, A.C., 1984. Major carbon-14 deficiency in modern snail shells from southern
Nevada springs. Science 224, 58–61.
Romaniello, L., Quarta, G., Mastronuzzi, G., D’Elia, M., Calcagnile, L., 2008. 14
C age
anomalies in modern land snails shell carbonate from Southern Italy. Quaternary
Geochronology 3, 68–75.
Rubin, M., Likins, R.C., Berry, E.G., 1963. On the validity of radiocarbon dates from
snail shells. Journal of Geology 71, 84–89.
Samos, G., 1949. Some observations on exchange of CO2 between BaCO3 and CO2
gas. Science 110, 663–665.
Slota, P.J., Jull, A.J.T., Linick, T.W., Toolin, L.J., 1987. Preparation of small samples for
14
C accelerator targets by catalytic reduction of CO. Radiocarbon 29, 303–306.
Stott, L.D., 2002. The influence of diet on the d13
C of shell carbon in the pulmonate
snail Helix aspersa. Earth and Planetary Science Letters 195, 249–259.
Stuiver, M., Polach, H.A., 1977. Reporting of 14
C data. Radiocarbon 19, 355–363.
Stuiver, M., Reimer, P.J., 1993. Extended 14
C database and revised CALIB radiocarbon
calibration program. Radiocarbon 35, 215–230.
Tamers, M.A., 1970. Validity of radiocarbon dates on terrestrial snail shells. American
Antiquity 35, 94–100.
Torn, M.S., Vitousek, P.M., Trumbore, S.E., 2005. The influence of nutrient availability
on soil organic matter turnover estimated by incubations and radiocarbon
modeling. Ecosystems 8, 352–372.
Wilbur, K.M., 1972. Shell formation in mollusks. In: Florkin, M., Scheer, B.T.
(Eds.), Chemical Zoology, Mollusca, vol. 7. Academic Press, New York, pp.
103–145.
Yates, T., 1986. Studies of non-marine mollusks for the selection of shell samples for
radiocarbon dating. Radiocarbon 28, 457–463.
Zhou, W., Head, W.J., Wang, F., Donahue, D.J., Jull, A.J.T., 1999. The reliability of AMS
radiocarbon dating of shells from China. Radiocarbon 41, 17–24.
J.S. Pigati et al. / Quaternary Geochronology 5 (2010) 519–532532
112
Abstract The occurrence of ten butterfly taxa (Clossiana
eunomia dawsonii, Clossiana freija, Clossiana
frigga, Clossiana titania, Coenonympha inornata, Erebia
discoidalis, Incisalia augustinus, Lycaena dorcas,
Lycaena epixanthe, Oeneis jutta) was analyzed within
three acid peatland habitat types from the Lake Superior
drainage basin of northwestern Wisconsin. Both first(nearest-neighbor
spatial analysis) and second-order
(Ripley’s K) spatial point process statistics were used to
identify the extents over which each distribution pattern
significantly deviated from random expectations. Versions
of these tests were used that identified significant
spatial pattern uncorrelated to habitat location and habitat
preference. These analyses documented non-random
occurrence patterns in seven species. Deviations from
random were largely confined to two extents: <50 km
and 70–100+ km. The majority of non-random patterns
at <50 km extents were examples of aggregation, while
the majority of non-random patterns noted at the
70–100+ km scale were examples of segregation. These
results demonstrate that even for winged animals inside a
limited landscape, spatially constrained processes can be
important determinants of distribution. It is likely that
metapopulation dynamics and dispersal limitation help
explain why aggregation is dominant at small scales. The
mechanisms underlying the predominance of segregation
at large scales are less clear, but may be related to migration
history and/or weak environmental gradients.
Keywords Butterflies · Metapopulation ·
Nearest-neighbor spatial analysis · Ripley’s K ·
Spatial pattern
Introduction
Species occurrence patterns have often been described as
a simple reflection of species niche requirements and underlying
environmental conditions (MacArthur 1972;
Tilman 1988). This idea requires the assumption that dispersal
never limits distribution at regional scales, giving
all species access to all potential habitats (Krebs 1985).
Saur (1988) equated this concept with Beijerinck’s Law,
which states that “everything is everywhere but the environment
selects”. In such situations, species occurrences
should not have significant spatial pattern beyond that
exhibited by the physical environment.
However, the spatial relationship of populations, independent
of the physical environment, may also influence
occurrence patterns. Supply-side ecology (Roughgarden
et al. 1987) and the mass effect (Shmida and Ellner 1984)
demonstrate how proximity to potential source populations
can positively influence occurrence frequency. Additionally,
analyses of distance decay (Nekola and White
1999) and competitive co-equivalency (Shmida and Ellner
1984) demonstrate how habitat isolation can negatively
influence population frequency. For these reasons, patch
size and isolation can correlate as significantly to patch
occupancy as does the physical environment (Moilanen
and Hanski 1998). When such spatial factors are at work,
non-random spatial occurrence patterns should be evident
independent of the physical environment. Identification of
such patterns is an important first step in the recognition
of situations where spatially constrained processes are (or
have been) important (Diggle 1983).
How frequently do populations deviate from random
occurrence patterns, independent of their environment?
One way of assessing this question is to compare distributions
across a well-defined taxonomic group within a
particular habitat in a given landscape. This allows for at
least partial control of the environmental, geographic,
temporal, and phylogenetic templates upon which distributions
have developed.
Butterflies of northern Wisconsin acid peatlands represent
such a system. Recent and historical investigaJ.C.
Nekola (✉)
Department of Natural and Applied Sciences,
University of Wisconsin-Green Bay, 2420 Nicolet Drive,
Green Bay, WI 54311, USA
e-mail: nekolaj@gbms01.uwgb.edu
Tel.: +1-920-4652937
C.E. Kraft
Department of Natural Resources, Cornell University, Ithaca, NY,
USA
Oecologia (2002) 130:53–61
DOI 10.1007/s004420100782
Jeffrey C. Nekola · Clifford E. Kraft
Spatial constraint of peatland butterfly occurrences
within a heterogeneous landscape
Received: 21 March 2000 / Accepted: 19 July 2001 / Published online: 1 September 2001
© Springer-Verlag 2001
113
tions have documented ten species [Clossiana eunomia
dawsonii (Nymphalidae), Clossiana freija (Nymphalidae),
Clossiana frigga (Nymphalidae), Clossiana titania
(Nymphalidae), Coenonympha inornata (Satyridae),
Erebia discoidalis (Satyridae), Incisalia augustinus
(Lycaenidae), Lycaena dorcas (Lycaenidae), Lycaena
epixanthe (Lycaenidae), Oeneis jutta (Satyridae)], which
live out their entire life cycles in these habitats (Masters
1971a, b, 1972; Ferge and Kuehn 1976; Kuehn 1983;
Swengel 1995; Nekola 1998). Within the Lake Superior
drainage basin of northwestern Wisconsin, three types of
acid peatland habitats exist: muskeg, kettlehole, and
coastal. Muskeg sites are dominated by open black
spruce-cottongrass-wiregrass savanna, are relatively dry
(except in the proximity of moats or lakes), and usually
have an elevation similar to the surrounding uplands.
Kettlehole peatlands are generally wetter, commonly
contain floating sphagnum-leatherleaf mats, and are typically
found fringing lakes or in kettlehole depressions.
Coastal wetlands are limited to estuaries along the Lake
Superior coast. Considerable similarity exists within the
physical environment, habitat size, and flora of each
peatland type, leading to the presence of very similar
habitats throughout the region. However, subtle environmental
and vegetation differences exist between peatland
types, producing slightly different butterfly faunas within
each (Nekola 1998).
The following paper analyzes the occurrence patterns
of these butterflies within these habitats in order to identify
the frequency of non-random distributions that are uncorrelated
with environment, and the spatial extents at
which the majority of any such non-random patterns occur.
These results identify potentially important spatial
processes, and can be used to guide conservation strategy.
Materials and methods
Site selection
Identification of all high quality peatlands within or adjacent to
the Lake Superior drainage basin (Fig. 1) was accomplished by a
fly-over of the study area on 20 May 1996 in a small aircraft. Approximately
5 h of flight time were required to survey the entire
region at an altitude of 3,000–5,000 m in north-south transects
running approximately 10 km apart. Other high-quality sites were
identified through review of the Biological Conservation Database
and discussions with the staff of the Wisconsin Department of
Natural Resources Bureau of Endangered Resources. While an attempt
was made to inventory all high-quality sites, some (like the
Kakagon Sloughs) were not visited due to difficulty in political
and/or physical access. As few such sites existed, and as they were
spread across the study region, we assume that their absence from
this analysis has not biased results.
Field sampling
The latitude-longitude location of each surveyed site was determined
through digitization of USGS 7.5 min topographic quads
using the ATLAS DRAW software package. The centroids for
these sites were then converted to Zone 15 UTM coordinates using
PC-ARCINFO.
The peatland habitat type represented by each site (muskeg,
kettlehole, or coastal) was also recorded. Muskeg peatlands (37 to-
54
Fig. 1 Distribution of surveyed peatland habitats in northwestern
Wisconsin. The shaded area in the south of the region represents
areas outside of the Lake Superior drainage basin
Fig. 2 A Occurrence pattern of Clossiana eunomia dawsoni within
the study region. Black circles represent occupied sites; open
circles represent sites vacant during flight time. B Black line
shows distance to nearest neighbor for observed distribution of occupied
sites; dashed lines show upper and lower 95% confidence
intervals from 5,000 Monte Carlo NNSA simulations. C Black
line shows K(t) for observed distribution of occupied sites; dashed
lines show upper and lower 95% confidence intervals from 5,000
Monte Carlo Ripley’s K simulations
114
tal inventoried sites) were concentrated along the southern divide
basin boundary in the western and eastern sides of the study area
and were absent from the pitted outwash plain in the central region.
Kettlehole peatlands (23 sites) were generally concentrated
in the central outwash plain that lacked muskeg sites, although a
few were also scattered throughout the eastern half. Coastal peatlands
(10 sites) were limited to the Lake Superior shore in the
Bayfield Peninsula and the Apostle Islands (Fig. 1).
Each site was visited 4–7 times during the 1996 flight season.
Surveys were only conducted during sunny, dry weather when
temperatures were 18°C or higher. Surveys were made at weekly
intervals early in the flight season when a number of closely
spaced adult emergences occur. Later in the season, when species
emergence periods were more temporally separated and individual
populations were in flight for longer periods of time, the survey
interval increased to 3 weeks. Each site visit lasted for 15–90 min,
depending upon site size, and all encountered peatland butterfly
taxa were recorded. Voucher specimens are housed in the first
author’s collection at the University of Wisconsin-Green Bay.
Nomenclature follows that of Miller and Brown (1981).
Cumulative presence-absence species lists for all peatland taxa
from each site were compiled from site visit lists combined with
any additional collections archived at the Milwaukee Public
Museum and element occurrence records maintained by the
Wisconsin Bureau of Endangered Resources.
As weather and logistical constraints prevented visitation of all
sites during a given flight time, only sites inventoried during the
flight duration for a given species were used for statistical analysis.
Flight times for a given species were determined by noting the
earliest and latest dates of adult encounters in the study region
during 1996.
Statistical analyses
Two general approaches exist to analyze spatial point patterns:
first-order statistics consider the mean distances between points,
while second-order statistics consider the variance in these distances.
One of the most commonly used first-order statistics is
nearest-neighbor analysis (Diggle 1983; Manly 1991), while one
of the most commonly used second-order statistics is Ripley’s K
analysis (Ripley 1977; Haase 1995). As little has been written to
suggest whether first- or second-order statistics are more effective
at identifying non-random patterns, both were used.
Both methods, as commonly employed, generate random null
models from the assumption that occurrence is possible anywhere
within the study region (e.g. Diggle 1983; Manly 1991). This is
not appropriate for acid peatland butterflies, which cannot reside
outside of peatland habitats. If peatlands are non-randomly distributed
(as is the case in northwestern Wisconsin), it will be impossible
to determine from such null models whether a non-random occurrence
pattern for a given species is due to habitat distribution or
other spatially constrained processes. Additionally, the physical
environment may lead to significant non-random pattern when differences
exist between habitats (as is also the case for peatland
butterflies in our study). To more clearly test for non-random distributions
uncorrelated to such environmental pattern, the firstand
second-order statistics used in this study constrained random
null models to actual peatland locations and to the actual occurrence
frequencies of each species within each peatland type.
Nearest-neighbor spatial analysis (NNSA; Davis et al. 2000)
was chosen as the first-order statistic. This method is related to the
nearest-neighbor method described by Manly (1991), in which the
average distance between 1st, 2nd, 3rd to n–1 nearest neighbors
(where n equals the total number of occurrences of that species in
the landscape) is calculated and compared to expected values generated
from random distributions. Unlike Manly’s test, the NNSA
method generates random expectations based upon the subsampling
of a larger finite set of points within a landscape. This permits
factoring out spatial pattern caused solely by habitat distribution.
This method also constrains random subsamples to be drawn
at the same frequencies for which taxa were observed in that given
habitat type. This allows any spatial pattern caused by habitat differences
to be factored out (Davis et al. 2000).
NNSA null expectations were based on 5,000 Monte Carlo
simulations. Two-tailed 95% confidence intervals were generated
from this process, and plotted versus observed nearest-neighbor
distances (Figs. 2, 3, 4, 5, 6, 7, 8, 9, 10, 11). From these, the total
55
Fig. 3 A Occurrence pattern of Clossiana freija within the study
region. Black circles represent occupied sites; open circles represent
sites vacant during flight time. B Black line shows distance to
nearest neighbor for observed distribution of occupied sites;
dashed lines show upper and lower 95% confidence intervals from
5,000 Monte Carlo NNSA simulations. C Black line shows K(t)
for observed distribution of occupied sites; dashed lines show upper
and lower 95% confidence intervals from 5,000 Monte Carlo
Ripley’s K simulations
Fig. 4 Occurrence pattern of Clossiana frigga within the study region.
Black circles represent occupied sites; open circles represent
sites vacant during flight time
115
range of distances were determined over which observed distributions
deviated from random.
Ripley’s K (Ripley 1977) was chosen as the second-order
statistic. This test evaluates the expected number of points within
a distance t of an arbitrary point in the area being evaluated
(Haase 1995). If occurrences are randomly distributed, the expected
value of K(t)=πt2.
The unbiased estimate of K(t) for an observed spatial point pattern
is:
where n is the number of points in the landscape area A; uij is the
distance between events i and j; It(u), the counter variable, equals
1 if u<=t and 0 if u>t; wij is the proportion of the circumference
of a circle centered at the event i with radius uij lying within A;
and the summation is over all pairs of a given species occurrence
(Ripley 1977).
Random expectations for K(t) were based upon a random subsample
of n peatlands in the landscape, with n representing the
number of species occurrences, as is outlined in Kraft et al. (in
press). Randomly chosen sites were selected according to the
actual proportion of sites colonized by each species within each
peatland type. In this way, spatial pattern due to habitat distribution
and preference were factored out.
S-Plus was used for estimating Ripley’s K (Venables and
Ripley 1997). Null expectations were based on 5,000 Monte Carlo
simulations. Two-tailed 95% confidence intervals were generated
and plotted against observed values (Figs. 2, 3, 4, 5, 6, 7, 8, 9, 10,
11). From these, the total range of distances was determined over
which observed deviated from random.
Because our null expectations are based upon randomly selected
subsets of given peatland sites, the expectation that K(t) will
approximate πt2 is not valid. Thus, the boundary issues considered
by Haase (1995) do not apply to our analysis, as occurrences are
constrained to actual peatland locations. Any such boundary-based
biases should be accounted for by our use of Monte Carlo simulations
to estimate the 95% confidence interval for K(t).
Results were summarized across all species via histograms
documenting the total number of significant aggregations and segregations
at each analyzed distance for both methods (ca.
0–150 km for NNSA, 0–91 km for Ripley’s K).
56
Fig. 5 A Occurrence pattern of Clossiana titania within the study
region. Black circles represent occupied sites; open circles represent
sites vacant during flight time. B Black line shows distance to
nearest neighbor for observed distribution of occupied sites;
dashed lines show upper and lower 95% confidence intervals from
5,000 Monte Carlo NNSA simulations
Fig. 6 A Occurrence pattern of Coenonympha inornata within the
study region. Black circles represent occupied sites; open circles
represent sites vacant during flight time. B Black line shows distance
to nearest neighbor for observed distribution of occupied
sites; dashed lines show upper and lower 95% confidence intervals
from 5,000 Monte Carlo NNSA simulations. C Black line
shows K(t) for observed distribution of occupied sites; dashed
lines show upper and lower 95% confidence intervals from 5,000
Monte Carlo Ripley’s K simulations
Fig. 7 Occurrence pattern of Erebia discoidalis within the study
region. Black circles represent occupied sites; open circles represent
sites vacant during flight time
116
Results
Two taxa (Clossiana frigga and Erebia discoidalis) were
only twice encountered. Their populations were limited
to muskeg sites in the west and east. Although their distributions
have been mapped for sake of completeness
(Figs. 4, 7), their few occurrences precluded further statistical
analysis.
Significant deviations from random were noted in the
occurrences of seven of the remaining eight taxa. The
only species which did not significantly deviate from
random in either test was Coenonympha inornata, which
was located at nine scattered sites (4 muskeg, 1 kettlehole,
4 coastal; Fig. 7).
Clossiana eunomia dawsonii was located at 34 sites
(14 muskeg, 14 kettlehole, 4 coastal), with the majority
being located in the western half of the study region
(Fig. 2). NNSA identified significant aggregation in
these occurrences from the 6th–14th (min. P<0.0005;
21–44 km), 25rd–27th (min. P=0.008; 73–77 km), and
29th (P=0.021; 85 km) nearest neighbors. Ripley’s K
identified significant clustering (min. P<0.0005) in these
occurrences from 1–48 km and 67–87 km extents.
Clossiana freija was located at 12 muskeg sites, with
populations being restricted to the far west and east
(Fig. 3). NNSA identified significant aggregation over
the 1st–2nd (min. P=0.012; 5–11 km) and 8th (P=0.017;
47 km) nearest neighbors. Significant segregation was
noted at the 11th (P=0.001; 150 km) nearest neighbor.
Ripley’s K identified significant aggregation from
4–55 km (min. P<0.0005), and significant segregation
from 69–91 km (min. P<0.0005).
Clossiana titania was located at 6 muskeg sites in
the far west (Fig. 5). NNSA identified significant aggre-
57
Fig. 8 A Occurrence pattern of Incisalia augustinus within the
study region. Black circles represent occupied sites; open circles
represent sites vacant during flight time. B Black line shows distance
to nearest neighbor for observed distribution of occupied
sites; dashed lines show upper and lower 95% confidence intervals
from 5,000 Monte Carlo NNSA simulations. C Black line
shows K(t) for observed distribution of occupied sites; dashed
lines show upper and lower 95% confidence intervals from 5,000
Monte Carlo Ripley’s K simulations
Fig. 9 A Occurrence pattern of Lycaena dorcas within the study
region. Black circles represent occupied sites; open circles represent
sites vacant during flight time. B Black line shows distance to
nearest neighbor for observed distribution of occupied sites;
dashed lines show upper and lower 95% confidence intervals from
5,000 Monte Carlo NNSA simulations. C Black line shows K(t)
for observed distribution of occupied sites; dashed lines show upper
and lower 95% confidence intervals from 5,000 Monte Carlo
Ripley’s K simulations
117
gation (minimum p<0.0005) across all neighbors (5–
15 km). Too few occurrences were present to allow calculation
of Ripley’s K.
Incisalia augustinus was located at 39 sites (30 muskeg,
8 kettlehole, 1 coastal), which were most frequently
encountered in the west and east (Fig. 8). NNSA identified
significant segregation (min. P<0.0005) from the
22nd–38th nearest neighbors (81–133 km). Ripley’s K
identified significant aggregation in occurrence at 5 km
(P=0.023) and from 10–44 km (min. P<0.0005), and significant
segregation from 61–91 km (min. P<0.0005).
Lycaena dorcas was located at 13 muskeg sites, with
the majority of populations being found in the eastern
half (Fig. 9). NNSA identified significant (P=0.003) segregation
at the 3rd nearest neighbor (40 km). Ripley’s K
identified significant aggregation in its occurrence from
16–60 km (min. P<0.0005), and significant segregation
from 81–82 km (min. P=0.013) and 86–91 km (min.
P=0.006).
Lycaena epixanthe was located at 23 sites (7 muskeg,
7 kettlehole, 9 coastal), with populations sporadically occurring
throughout the entire region (Fig. 10). No significant
deviations from random were noted by NNSA.
However, Ripley’s K identified significant segregation
(min. P<0.0005) in its occurrences from 3 to 21 km.
Oeneis jutta was located at 18 sites (15 muskeg,
3 kettlehole), with populations being noted in the west
and south central (Fig. 11). NNSA identified significant
aggregation from the 3rd–5th (minimum p=0.009;
15–23 km) nearest neighbors. Ripley’s K identified significant
aggregation from 5–8 km (min. P=0.008), 10 km
(P=0.022), 13 km (P=0.022), 16 km (P=0.024), and
from 18–49 km (min. P=0.001).
Comparison of these results across species (Fig. 12)
show that significant deviation from random was most
apparent at three extents. At <20 km, NNSA identified
non-random aggregation for two taxa (C. freija, C. titania),
while Ripley’s K identified significant aggregation
58
Fig. 10 A Occurrence pattern of Lycaena epixanthe within the
study region. Black circles represent occupied sites; open circles
represent sites vacant during flight time. B Black line shows distance
to nearest neighbor for observed distribution of occupied
sites; dashed lines show upper and lower 95% confidence intervals
from 5,000 Monte Carlo NNSA simulations. C Black line
shows K(t) for observed distribution of occupied sites; dashed
lines show upper and lower 95% confidence intervals from 5,000
Monte Carlo Ripley’s K simulations
Fig. 11 A Occurrence pattern of Oeneis jutta within the study region.
Black circles represent occupied sites; open circles represent
sites vacant during flight time. B Black line shows distance to
nearest neighbor for observed distribution of occupied sites;
dashed lines show upper and lower 95% confidence intervals from
5,000 Monte Carlo NNSA simulations. C Black line shows K(t)
for observed distribution of occupied sites; dashed lines show upper
and lower 95% confidence intervals from 5,000 Monte Carlo
Ripley’s K simulations
118
within each peatland type, possessing similar (if not
identical) microhabitats, vascular plant, and bryophyte
species. Additionally, given the small total extent of
the sample region (ca. 150 × 75 km), it is unlikely that
strong climatic gradients exist. Thus, we believe that
the patterns of peatland butterfly occurrence are likely
related to other factors.
Assuming that peatlands within a given type have approximately
similar environments for these butterflies,
the existence of non-random occurrence patterns independent
of habitat location and preference would be unlikely
if dispersal allowed equal access to all suitable sites. The
frequent presence of significant non-random occurrence
pattern in these species suggests that within this landscape
spatial limitation of butterfly dispersal is common.
Inference of spatial process from spatial pattern
The local clustering of occurrences is not surprising as
many ecological processes show some degree of spatial
autocorrelation over relatively small extents (Shmida and
Ellner 1984; Burrough 1986; Okubo and Levin 1989). In
particular, limitations on butterfly movement contribute
to patchy occurrences in a number of European taxa, including
Clossiana (Boloria) aquilonaris (Mousson et al.
1999), C. (Proclossiana) eunomia (Nève et al. 1996a),
Hesperia comma (Hill et al. 1996), and Melitaea cinxia
(Hanski et al. 1994). When dispersal limitation is coupled
with other aspects of metapopulation dynamics,
local aggregations of butterfly occurrence, independent of
the physical environment, are often evident. For instance,
M. cinixa occurrences on the main Åland island in southwestern
Finland demonstrate what visually appears to be
strong aggregation over 1–10 km extents among appropriate
habitat patches (Moilanen and Hanski 1998). Occurrences
of H. comma in the North Downs, Surrey, UK,
also appear to be aggregated over 2–6 km extents within
appropriate sites (Hill et al. 1996). The metapopulations
of all five edaphically-restricted butterfly taxa of Sierra
Nevada (California) serpentine barrens also demonstrate
apparent spatial aggregation in their occurrences among
known sites (Gervais and Shapiro 1999).
Much of the previous work on occurrence patterns
in Lepidoptera has focused on the factors predicting habitat
occupancy or vacancy (Ehrlich and Murphy 1987;
Thomas et al. 1992; Hanski et al. 1994). While these
studies have elucidated the processes influencing individual
patches, none have investigated the spatial patterns
in occurrence generated by these processes across
an entire landscape. Analyzing such patterns may prove
a useful tool for identifying those species whose distribution
is influenced by spatial processes (Hanski 1999).
For instance, C. eunomia is known to exist in discrete
metapopulations within Belgium and France, with individuals
capable of traversing up to 8 km of unfavorable
habitat within a season (Nève et al. 1996b). Our analyses
demonstrate strong aggregation for this species at relatively
small (<50 km) scales. A previous investigation on
59
Fig. 12 Frequency of non-random aggregation and segregation
patterns from 0–100 km extents for both NNSA and Ripley’s K
analyses
in four taxa (C. eunomia dawsonii, C. freija, I. augustinus,
O. jutta). Additionally, Ripley’s K identified significant
segregation over this range for L. epixanthe. From
20–50 km, NNSA identified aggregation for three taxa
(C. eunomia dawsonii, C. freija, O. jutta), while Ripley’s
K identified aggregation for five taxa (C. eunomia dawsonii,
C. freija, I. augustinus, L. dorcas, O. jutta). Significant
segregation was identified at this scale by NNSA
for L. dorcas. Few non-random occurrence patterns were
identified by either method from 50–70 km. However, at
larger scales both methods identified significant aggregation
in C. eunomia dawsonii. Significant segregation at
this scale was identified by both methods for C. freija
and I. augustinus, and by Ripley’s K for L. dorcas.
Discussion
Perhaps the most obvious and important finding from
these analyses is the simple realization that in northwestern
Wisconsin, across a broad range of spatial scales
(0–90 km), most acid-peatland butterflies appear to possess
non-random occurrences independent of habitat location
and preference.
As we did not sample physical environmental variables
from these sites (e.g. soil and water chemistry,
local climate, amount and type of plant cover), it is possible
that some of these distribution patterns may be related
to such factors. If present, however, these factors
must be cryptic, as habitats appeared to be homogeneous
119
Coenonympha tullia (often lumped with C. inornata)
distribution in Northumberland, UK, documented that
occurrences were more strongly related to habitat quality
than geographic factors (Dennis and Eales 1999). In
northwestern Wisconsin, this species was the only one
that exhibited random occurrence patterns at all scales.
These examples, in conjunction with our analysis of spatial
distribution pattern, illustrate how analyses of spatial
pattern can reflect potential processes underlying organism
distribution. While Real and McElhaney (1996) note
that spatial pattern analysis, alone, cannot distinguish between
all processes capable of producing a given pattern,
we believe that the analysis of spatial pattern provides an
instructive first step that can help improve field experiments
or modeling efforts used to evaluate the processes
affecting distribution.
It is not clear for the majority of the fauna why aggregation
is essentially limited to <50 km extents. This is
especially interesting given the fact that many of these
species appear to share similar niches: in the region
Clossiana freija, C. eunomia, C. titania, L. dorcas, and
L. epixanthe all are obligate cranberry consumers, while
Coenonympha inornata, E. discoidalis, and O. jutta appear
limited to cottongrass and/or wiregrass (Nekola
1998). While these species might thus be expected to
demonstrate some form of competitive exclusion, they instead
appear to show largely coincident occurrence patterns
at regional scales. Such correspondence between occurrence
patterns is likely related to factors that have
equally influenced all taxa, such as habitat and host plant
distribution. If related to cryptic environmental gradients,
the location of colonies may be expected to remain static
over time. However, if shifting metapopulation dynamics
are responsible, the location of occurrence clusters may
change over ecological time scales within the landscape.
Another enigma is the presence of significant smallscale
segregation in occurrences of L. epixanthe. While
it is possible that colonies may have become segregated
due to competitive interactions, such processes are generally
thought to be limited to small extents within sites
(Shmida and Ellner 1984). It is difficult to envision how
such mechanisms could operate at distances up to 20 km.
A number of other spatial mechanisms may underlie
larger scale (70–150 km) segregated spatial distributions.
Migration history may provide one explanation. For instance,
migration of C. eunomia dawsonii out of Minnesota
(beyond the study area boundaries) could help
account for its higher occurrence frequencies in the west.
Likewise, the absence of species from the center (e.g.,
C. freija, L. dorcas, O. jutta) could be due to dual migration
pathways from peatland-rich landscapes to the west
and east. However, these patterns could also be related to
cryptic soil and water chemistry or climate gradients that
our statistical methods could not control for. In this
event, identification of such cryptic gradients would be a
useful result, given their difficulty in direct detection.
Teasing apart the historical and environmental influences
on distribution at large scales may prove difficult
due to covariation between these factors. Population genetics
analyses may provide one possible avenue to distinguish
between them. If historical factors are responsible
for the observed large-scale pattern, it seems likely that
significant genetic differences may occur across the study
region due to differences in initial population source
pools. However, if dispersal is not limiting and contemporaneous
environmental gradients are responsible for the
observed pattern, differences in allele frequency should be
diminished from more thorough genetic mixing.
Conservation implications
Relatively few populations (e.g., <25 stations) of these
taxa (except I. augustinus) have been previously reported
from Wisconsin (Ebner 1970; Masters 1971a, b, 1972;
Ferge and Kuehn 1976; Kuehn 1983; Swengel 1995). As
acid peatlands are frequent in the northern Wisconsin
landscape, it has been generally assumed that these species
were simply undercollected and would eventually be
found within most appropriate sites (Wisconsin Rare
Butterfly Working List 1996). As such, none of these
species have been granted endangered or threatened species
protection within the state (Wisconsin Endangered
Species List 1999).
The existence of complex spatial structure in the occurrence
of these taxa suggests that such expectations may be
false. The aggregation of most inventoried taxa indicates
that, at some scale, high-quality sites will be unoccupied.
Modifications to sites within limited areas of occurrence
would have profound implications to the long-term survival
of those species across the entire landscape. For example,
C. freija and C. titania are essentially restricted to a
small cluster of sites in the far west of the study region.
Unfortunately, peatlands in this same area were observed
to be undergoing the heaviest “all terrain vehicle” abuse
within the entire region. Thus, even though appropriate
habitat for these species exists throughout the study region,
the correspondence of high levels of disturbance in the region
where most colonies occur suggests that their continued
existence within this landscape may be in jeopardy.
In this region, the vulnerability of peatland butterfly
species appears to be generally unrelated to habitat frequency.
Such patterns may be common at range edges
where increased environmental stress and patch isolation
lead to lower habitat occupancy rates (Thomas et al.
1992). When species distribution does not simply mimic
habitat abundance, determination of population vulnerability
to extirpation can only be assessed by conducting
geographically systematic inventories and documenting
the actual distribution of taxa. Additionally, repeated
monitoring of distribution may be necessary if shifting
metapopulations exist.
Comparison of first- and second-order
spatial point statistics
Given the different assumptions and algorithms underlying
these methods, the general robustness of the results
60
120
suggests that the underlying patterns are quite strong.
Only one species (L. dorcas) exhibited a major discrepancy
in outcomes, with NNSA identifying segregation in
the same 30–40 km extent where Ripley’s K showed
strong aggregation. Based upon visual inspection of this
taxon’s occurrences, the Ripley’s K result appeared more
accurate. Perhaps NNSA was unable to identify aggregation
for this species as occurrences were concentrated on
a region of widely separated sites.
Comparing results across species, Ripley’s K identified
more non-random distributions at a given extent, and
identified larger regions for non-random distribution, as
compared to NNSA. It would thus appear that NNSA is
the more conservative of the two tests. However, visual
inspection of maps suggests that NNSA might be overly
conservative, as it failed to identify areas of apparent aggregation
or segregation at some scales (e.g., C. freija,
I. augustinus, L. epixanthe, O. jutta). A more rigorous
comparison of these methods would help to better determine
their relative strengths and weaknesses.
Acknowledgements The computer code used for conducting the
NNSA analyses was graciously provided by Jennifer Davis. Advice
on the implementation of Ripley’s K using S-Plus was provided by
Patrick Sullivan. Eric North helped in field sampling and Heather
Voster collected data from the Milwaukee Public Museum collections.
Patricia Schneider piloted the 20 May air reconnaissance of
the study region. Peter White and Michael Draney read and made
useful comments regarding earlier drafts. This research was funded
through a grant from the Lake Superior Coastal Wetland Inventory
which was administered by the Wisconsin Department of Natural
Resources and the U.S. Environmental Protection Agency.
References
Burrough PA (1986) Principles of geographical information systems
for land resources assessment. Oxford University Press,
New York
Davis JH, Howe RW, Davis GJ (2000) A multi-scale spatial analysis
method for point data. Landscape Ecol 15:99–114
Dennis RLH, Eales HT (1999) Probability of site occupancy in the
large heath butterfly Coenonympha tullia determined from
geographical and ecological data. Biol Conserv 87:295–301
Diggle PJ (1983) The analysis of spatial point patterns. Academic
Press, New York
Ebner JA (1970) Butterflies of Wisconsin. Milwaukee Public Museum
Popular Science Handbook 12, Milwaukee, Wisconsin
Ehrlich PR, Murphy DD (1987) Conservation lessons from longterm
studies of checkerspot butterflies. Conserv Biol 1:122–131
Ferge LA, Kuehn RM (1976) First records of Boloria frigga
(Nymphalidae) in Wisconsin. J Lepid Soc 30:233–234
Gervais BR, Shapiro AM (1999) Distribution of edaphic-endemic
butterflies in the Sierra Nevada of California. Glob Ecol
Biogeogr Lett 8:151–162
Haase P (1995) Spatial pattern analysis in ecology based on
Ripley’s K-function: introduction and methods of edge correction.
J Veg Sci 6:575–582
Hanski I (1999) Habitat connectivity, habitat continuity, and metapopulations
in dynamic landscapes. Oikos 87:209–219
Hanski I, Kuussaari M, Nieminen M (1994) Metapopulation structure
and migration in the butterfly Melitaea cinxia. Ecology
75:747–762
Hill JK, Thomas CD, Lewis OT (1996) Effects of habitat patch
size and isolation on dispersal by Hesperia comma butterflies:
implications for metapopulation structure. J Anim Ecol 65:
725–735
Kraft CE, Sullivan PJ, Karatayev AY, Burlakova LE, Nekola JC,
Johnson LE, Padilla DK (in press) Landscape patterns of zebra
mussel inland lake invasions: evaluating dispersal from distribution.
J Appl Ecol
Krebs CJ (1985) Ecology. Harper and Row, New York
Kuehn RM (1983) New Wisconsin butterfly records. J Lepid Soc
37:228–235
MacArthur RH (1972) Geographical ecology. Harper and Row,
New York
Manly BFJ (1991) Randomization and Monte Carlo methods in
biology. Chapman and Hall, New York
Masters JH (1971a) Ecological and distributional notes on Erebia
discoidalis (Satyridae) in the north central states. J Res Lepid
9:11–16
Masters JH (1971b) First records of Boloria eunomia (Nymphalidae)
in Wisconsin. J Lepid Soc 25:149
Masters JH (1972) Critical comments on James A. Ebner’s
“Butterflies of Wisconsin.” In: Keuhn RM, Masters JH (eds)
Wisconsin butterflies. Mid-Continent Lepidoptera Series 59,
pp 5-11
Miller LD, Brown FM (1981) A catalogue/checklist of the butterflies
of America north of Mexico. Lepidopterist’s Society
Memoir 2
Moilanen A, Hanski I (1998) Metapopulation dynamics: effects of
habitat quality and landscape structure. Ecology 79:2503–
2515
Mousson L, Nève G, Baguette M (1999) Metapopulation structure
and conservation of the Cranberry Fritillary Boloria aquilonaris
(Lepidoptera, Nymphalidae) in Belgium. Biol Conserv
87:285–293
Nekola JC (1998) Butterfly (Lepidoptera: Lycaeniidae, Nymphalidae,
and Satyridae) faunas of three peatland habitat types in
the Lake Superior drainage basin of Wisconsin. Great Lakes
Entomol 31:27–37
Nekola JC, White PS (1999) The distance decay of similarity in
bioigeography and ecology. J Biogeogr 26:867-878
Nève G, Mousson L, Baguette M (1996a) Adult dispersal and
genetic structure of butterfly populations in a fragmented landscape.
Acta Oecol 17:621–626
Nève G, Barascud B, Hughes R, Aubert J, Descimon H, Lebrun P,
Baguette M (1996b) Dispersal, colonization power and
metapopulation structure in the vulnerable butterfly Proclossiana
eunomia (Lepidoptera: Nymphalidae). J Appl Ecol
33:14–22
Okubo A, Levin SA (1989) A theoretical framework for data analysis
of wind dispersal of seeds and pollen. Ecology 70:329–
338
Real LA, McElhany P (1996) Spatial pattern and process in plantpathogen
interactions. Ecology 77:1011–1025
Ripley BD (1977) Modeling spatial patterns. J R Stat Soc Ser B
39:172–212
Roughgarden J, Gaines SD, Pacala SW (1987) Supply side ecology:
the role of physical transport processes. Br Ecol Soc Symp
27:491–518
Saur JD (1988) Plant migration. University of California Press,
Berkeley
Shmida A, Ellner S (1984) Coexistence of plant species with similar
niches. Vegetatio 58: 29–55
Swengel AB (1995) Butterflies of Northwestern Wisconsin. Madison
Audubon Society, Madison
Thomas CD, Thomas JA, Warren MS (1992) Distributions of occupied
and vacant butterfly habitats in fragmented landscapes.
Oecologia 92:563–567
Tilman D (1988) Plant strategies and the dynamics and structure
of plant communities. Princeton University Press, Princeton
Venables WN, Ripley BD (1997) Modern applied statistics with
S-Plus. Springer, Berlin Heidelberg New York
Wisconsin Endangered Species List (1999) Wisconsin Department
of Natural Resources, Madison
Wisconsin Rare Butterfly Working List (1996) Bureau of Endangered
Resources, Wisconsin Department of Natural Resources,
Madison
61
121
2459
CONCEPTS & SYNTHESIS
EMPHASIZING NEW IDEAS TO STIMULATE RESEARCH IN ECOLOGY
Ecology, 80(8), 1999, pp. 2459–2473
᭧ 1999 by the Ecological Society of America
PALEOREFUGIA AND NEOREFUGIA: THE INFLUENCE OF
COLONIZATION HISTORY ON COMMUNITY PATTERN AND PROCESS
JEFFREY C. NEKOLA1
Curriculum in Ecology, 229 Wilson Hall, CB 3275, University of North Carolina,
Chapel Hill, North Carolina 27514 USA
Abstract. Two types of biological refugia (habitats that support populations not able
to live elsewhere in a landscape) can be defined from relative refugium age as compared
to surrounding matrix age; paleorefugia are now-fragmented relicts of a formerly widespread
matrix community, whereas neorefugia have formed more recently than the matrix. This
difference should make extinction a relatively more important process in determining species
occurrence in paleorefugia, whereas immigration should be relatively more important
in neorefugia. Based on these differences, a series of eight a priori predictions relating to
the diversity and distribution patterns for the biota of such sites can be generated: (1) the
slope of the species–area relationship, and amount of variance explained by it, should be
greater in paleorefugia as compared to neorefugia; (2) the negative relationship between
habitat isolation and species richness should be stronger in neorefugia as compared to
paleorefugia; (3) species richness should be expected to decrease over time in paleorefugia,
but to increase over time in neorefugia; (4) the inverse correlation between site distance
and community similarity (distance decay) should be stronger in neorefugia as compared
to paleorefugia; (5) neorefugia should be enriched in highly vagile species relative to
paleorefugia, whereas paleorefugia should be rich in less vagile species relative to neorefugia;
(6) geographic factors should be more important predictors of species occurrence
for neorefugia than for paleorefugia; (7) paleorefuge sites should possess more and stronger
correlations between community composition and environmental covariables (such as soil
chemistry, climate, etc.) as compared to neorefuge sites; and (8) the number of competitive
co-equivalents held within a system of neorefugia should be greater then the number held
within a series of paleorefugia.
The most readily testable predictions (numbers 1, 2, 4, and 5) were evaluated by comparing
species-richness and community-composition patterns within two northeastern Iowa
refugia: algific talus slopes (paleorefugia) and fens (neorefugia). Results from these tests
were consistent with predictions. These results illustrate that colonization history may
influence contemporaneous species diversity and community-composition patterns. They
also suggest that (1) equilibrium has yet to be achieved in the example systems after 5000–
10 000 yr, (2) the ecological-biogeographic debate centered around the mutual exclusivity
of vicariance and dispersal is intrinsically flawed, and (3) optimum reserve-design strategies
for biodiversity protection within paleorefuge and neorefuge systems will differ.
Key words: biodiversity; biogeography, historical; colonization history, influence on community
pattern; conservation biology; dispersal; fens; Iowa (USA), northeastern; island biogeography; refugia,
paleorefugia vs. neorefugia; species richness; talus slopes, algific; vicariance.
INTRODUCTION
Past environmental changes have caused repeated
fluctuations in species and habitat distributions that can
Manuscript received 6 February 1998; revised 10 November
1998; accepted 27 November 1998; final version received
21 December 1998.
1 Present address: Department of Natural Applied Sciences,
University of Wisconsin, 2420 Nicolet Drive, Green Bay, Wisconsin
54311 USA. E-mail: nekolaj@uwgb.edu
lead to formation of island-like habitats and isolated
populations through two contrasting pathways. They
may represent remnants of once-more-widespread distributions
that have become fragmented, or they may
represent the de novo development of habitats or populations
in a landscape where they were previously
absent. Borrowing terminology used for endemic species
(Stebbins and Major 1965, Krukeberg and Rabinowitz
1985), I refer to habitats that have been colonized
through these processes, respectively, as ‘‘paleo-
122
2460 Ecology, Vol. 80, No. 8JEFFREY C. NEKOLA
Concepts&Synthesis
FIG. 1. Diagrammatic representation of the development of paleorefugia and neorefugia. At time period A, environmental
conditions are conducive to the existence of a particular habitat (shaded). After critical environmental change (time period
B), this habitat has retreated completely from the focal landscape (inner box) in the case of neorefugia but persists as isolated
patches in the case of paleorefugia. By time period C, the environment has changed again, allowing the habitat to exist again
within portions of its original range. At time period C, paleorefugia are found within essentially the same microenvironmentally
buffered patches present at time period B. Neorefugia, in contrast, have formed as appropriate microhabitats develop in
isolation from the main habitat area.
refugia’’ and ‘‘neorefugia.’’ More specifically, paleorefugia
represent habitats that are older than the
surrounding biological matrix, whereas neorefugia represent
habitats that are younger than the surrounding
biological matrix. Such habitats are considered refugia
as they support communities or populations unable to
survive elsewhere in the landscape (i.e., Pielou 1979).
As these terms reflect the relative age of a refugium to
its matrix, a paleorefugium in one landscape may be
of more recent origin than a neorefugium in another.
It is also possible that a particular habitat type that
developed as a paleorefugium in one landscape may
be a neorefugium elsewhere.
In this paper, I ask how such paleorefugia and neorefugia
would be expected to differ in their contemporaneous
species richness, species distribution, and
community-composition patterns. I develop eight predictions
about these differences and then investigate
the four most easily testable ones with data collected
from two types of insular habitats that co-occur in
northeastern Iowa.
THE INFLUENCE OF EXTINCTION AND IMMIGRATION
The relative importance of extinction and immigration
processes varies greatly between paleorefugia and
neorefugia. First, consider the development of paleorefugia
sites. Before environmental change, a given
community is continuous over a wide geographic extent,
harboring a diverse and well-mixed biota potentially
near equilibrium with the environment. Following
environmental change, the habitat becomes restricted
to a few discrete patches in the landscape (Fig.
1). As the community becomes fragmented, habitat size
decreases and isolation between fragments increases,
leading to increased extinction rates on individual fragments
(MacArthur and Wilson 1967, Diamond 1975,
Bierrgaard et al. 1992). The richness of individual sites
falls, with larger fragments retaining more taxa than
smaller ones (Diamond 1975). Although some recolonization
from nearby sites undoubtedly occurs (the rescue
effect of Brown and Kodric-Brown [1977]), this
should not greatly influence large-scale compositional
pattern as nearby sites should have been originally colonized
from similar species pools.
In contrast, neorefugia sites are created through the
development of novel environmental conditions in a
landscape (compare time periods B to C in Fig. 1).
Consequently, the composition of neorefugia must be
a product of immigration. Factors that effect immigration,
such as habitat isolation or age, should therefore
exert influence on community composition and species
richness (Carlquist 1974, Diamond 1975, Bush and
Whittaker 1993). Although extinction certainly occurs
on these sites as newly recruited populations disappear,
this process should be of limited importance as populations
can only become extinct if a successful immigration
has previously occurred, and as incomplete
dispersal between sites will limit rates of competitive
exclusion (Shmida and Ellner 1984).
The differential contributions of extinction and immigration
have been empirically documented for various
insular habitats. Fernald (1925) hypothesized (incorrectly)
about the role played by unglaciated nuntacks
(reputed paleorefugia) in the development of eastern
North American species ranges. Diamond (1973,
1975) noted that the species richness of land-bridge
islands (paleorefugia) is primarily determined by extinction
rates. Brown (1971) also discussed the role of
extinction in shaping rodent faunas of Great Basin
montane and alpine communities, which became frag-
123
December 1999 2461NEOREFUGIA AND PALEOREFUGIA
Concepts&Synthesis
mented following the close of Wisconsinan glaciation
(paleorefugia). The importance of immigration in the
development of the biota of young volcanic islands
(neorefugia) has been shown on Krakatau, Long, and
Ritter islands in the eastern Pacific (Diamond 1975,
Bush and Whittaker 1993). Similar patterns were demonstrated
by Simberloff and Wilson (1969) in their manipulative
experiments of mangrove islands. Research
on the importance of immigration in shaping community
composition and structure has also been referred
to as ‘‘supply-side ecology’’ (Roughgarden et
al. 1987). However, none of these previous works have
attempted to compare the contemporaneous ecological
patterns of isolated communities that possess different
colonization histories and that have, therefore, been
exposed to different levels of extinction and immigra-
tion.
PREDICTIONS ABOUT SPECIES DISTRIBUTION, SPECIES
RICHNESS, AND COMMUNITY COMPOSITION
My primary assumption is that across a large range
of taxonomic groups, extinction will most influence the
species composition of paleorefugia, whereas immigration
will most influence the species composition of
neorefugia. From this assumption, I have developed
eight a priori expectations regarding differential species
richness, distribution, and community-composition
patterns:
(1) All else being equal, the slope of the species–
area relationship, and the amount of variance explained
by it, should be greater in paleorefugia than in neorefugia.
Paleorefugia can be expected to have steeper
species–area curves than neorefugia, because for a given-size
island the paleorefuge will be approaching
equilibrium number through species loss, whereas the
neorefuge will be approaching this number through
species accretion. Paleorefugia were originally more
connected, and presumably once harbored similar numbers
of species per unit area. Consequently, differences
in species richness should be largely a function of postfragmentation
extinction rates. Such rates are believed
to be strongly correlated with habitat size (e.g., Diamond
1975, Bierrgaard et al. 1992). The species richness
of neorefugia, however, should also be a function
of immigration processes, which are related less to habitat
size than to other factors like habitat age and isolation.
Consequently, paleorefugia can be expected to
have less variance around the species–area curve than
neorefugia. This prediction may be violated for neorefugia
that have remained isolated over evolutionary
time scales, as strong species–area relationships may
be generated on island habitats through speciation
(Carlquist 1974).
(2) All else being equal, the negative relationship
between habitat isolation and species richness should
be stronger in neorefugia than in paleorefugia. I expect
this as propagule movement (and hence immigration
rate) is largely a function of barrier width (Okubo and
Levin 1989). Thus, less-isolated neorefugia should
have access to a larger propagule pool than more-isolated
ones. This result would not be anticipated from
paleorefugia as, over limited geographic extents, sites
were likely colonized from similar species pools. As
such, contemporaneous site isolation should not strongly
affect richness. Isolation could influence paleorefugia
richness when recolonization from nearby sites
(Brown and Kodric-Brown 1977) is important.
(3) Species richness should decrease over ecological
time scales in paleorefugia, while increasing over similar
time scales in neorefugia. Extinction, or ‘‘relaxation’’
(sensu Diamond 1975), should eliminate taxa
from paleorefugia, as is typical for land-bridge islands.
Such taxa may include those that are not well adapted
to local site conditions, that are competitively inferior
(e.g., Tilman 1988), or that exhibit area sensitivity
(McDonald and Brown 1992). However, as neorefugia
form de novo out of a hostile landscape, they will initially
be void of species. Over time their richness
should increase as taxa from the surrounding species
pools immigrate onto sites (Zobel 1997).
It is also possible that over time the species richness
of neorefugia and paleorefugia will converge. As time
increases, immigration between neorefugia should become
more and more complete, thereby increasing the
importance of extinction in the structuring of site biotas
and strengthening the species–area relationship. On paleorefugia,
the stochastic loss of species may eventually
lower site richness to levels anticipated for neorefugia.
It is not clear what factors will most strongly
influence this rate of convergence.
(4) The inverse correlation between site distance and
community similarity (distance decay) should be stronger
for neorefugia than for paleorefugia. A fundamental
principle of geography is the inverse correlation between
similarity and intersample distance (Tobler
1970). In the geographical literature this decrease in
similarity has been termed ‘‘distance decay’’ and been
applied to phenomena as diverse as human communication
networks, migration patterns (Bennett and
Gade 1979, Fotherinham 1981), and spatial interpolation
(Burrough 1986). The rate of distance decay
among natural communities has been shown to vary
with the dispersal strategy of organisms and with the
degree of landscape fragmentation (Nekola and White,
in press).
I expect that the rate of compositional distance decay
for paleorefugia should primarily reflect pre-fragmentation
rates as long as no strong environmental gradients
are present in the current landscape. Consequently,
in most cases the rates of compositional distance decay
for paleorefugia should be low within regions. However,
because immigration is required for development
of neorefugia biotas, sites across a region may be colonized
through different species pools when source areas
or migration corridors differ. As a result, similarity
124
2462 Ecology, Vol. 80, No. 8JEFFREY C. NEKOLA
Concepts&Synthesis
of neorefugia communities should be more strongly
correlated with intersample distance.
(5) Neorefugia should be enriched in highly vagile
species relative to paleorefugia, whereas paleorefugia
should be enriched in less vagile species relative to
neorefugia. I expect that species with good dispersal
abilities will be more frequently represented in neorefugia,
as colonization on such sites is only possible
through the crossing of dispersal barriers–effectively
eliminating poor dispersers while increasing the relative
importance of good dispersers in the neorefugia
species pool. Poor dispersers not only can be a component
of paleorefugia biotas (as sites were colonized
prior to fragmentation), but may be favored over time
on these sites if they lose a smaller percentage of propagules
beyond site boundaries (Carlquist 1974), and
have larger, more competitively fit progeny (Harper
1977). These mechanisms have been suggested by Carlquist
(1974) to explain the higher frequency of poorly
dispersing plant and animal species on remote oceanic
islands.
(6) Geographic factors (such as distance to nearest
population) should be more important predictors of
species occurrence on neorefugia over ecological time
scales. All paleorefuge species (theoretically) had similar
access to sites prior to fragmentation. For this reason,
contemporaneous geographic factors should be of
relatively limited importance in predicting current species
distribution. Those environmental factors that directly
affect the survival of populations should be of
pronounced importance in predicting species occurrence
on these sites due to the prolonged competitive
sorting that has occurred there. However, the distance
to contemporaneous propagule sources should be an
important predictor of species occurrence on neorefugia,
as the probability of colonization will increase
with decreasing dispersal distance. Additionally, direct
environmental factors may be less important in predicting
species occurrence in neorefugia due to increased
habitat breadths that result from limited competitive
sorting on these sites and incomplete immigration
of potential competitors (Cox and Ricklefs
1977).
(7) The higher levels of competitive sorting in paleorefugia
should lead to a stronger tie between environmental
and compositional gradients (Peet and Christensen
1988) than in neorefugia. Although community
composition within paleorefugia should initially contain
more and stronger correlations with environmental
covariables than neorefugia, convergence can be expected
over ecological time.
(8) The number of competitive co-equivalents held
within an archipelago of neorefugia should be greater
than that held within a similar set of paleorefugia. The
more intense competitive sorting within and the lower
levels of pre-fragmentation isolation between paleorefugia
should have resulted in similar initial compositions
and a strongly competitive environment. As a
consequence, there should have been little opportunity
for establishment of equivalents with varying fitnesses
(Shmida and Ellner 1984). However, the always-isolated
nature of neorefugia should allow for less complete
intersite dispersal, making it easier for less fit coequivalents
to persist in the landscape.
As is true for many ecological processes that operate
over large spatial and temporal extents, collection of
data to test these hypotheses may be difficult, or in
some cases impossible. And, even if appropriate data
exist, it may be equally difficult to control for confounding
factors (Brown 1995). This should not imply,
however, that hypotheses regarding differential processes
between paleorefugia and neorefugia are inherently
unfalsifiable. By comparing current ecological
patterns between sites within the same geographic region,
confounding effects of the local environment, climatic
history, and migration history can be at least
partially controlled.
The most easily tested hypotheses are those that concern
directly quantifiable variables such as current species
richness, community composition, habitat size,
habitat isolation, and habitat location. Consistency with
Hypothesis 1 can be easily tested as it is based on a
comparison of species richness to habitat area. Hypothesis
2, which is based on species richness and siteisolation
data, can also be readily assessed. Hypothesis
4 can also be easily tested, as it only requires community-composition
and geographic-location data.
Lastly, Hypothesis 5 can be easily assessed by investigation
of the dispersal strategies exhibited for paleorefuge
and neorefuge species.
Hypothesis 7 can also be relatively easily tested, as
it is based on analysis of how quickly and predictably
composition changes over a given amount of environmental-gradient
space. Testing of this question is
slightly more complex, however, as the documentation
of these patterns requires use of multivariate data-reduction
procedures such as ordination. It is possible
that use of different data-reduction techniques could
produce different outcomes. Previous knowledge of
which environmental variables are critical in the system
may also be necessary.
Hypotheses 6 will be more difficult to test. To assess
the effect of geography on species occurrence, it is first
necessary to control for variation in distribution related
to environmental gradients. Although in some cases it
may be relatively easy to quantify this through use of
parametric or nonparametric regression techniques
(Austin et al. 1990), such analyses will be problematic
when environmental variables demonstrate strong spatial
autocorrelation or when habitats are clustered in a
landscape. In such situations it is difficult to know what
the null expectation will be for the effect of geography
on species occurrence, independent of environmental
variation. While these problems are likely tractable
through use of Monte Carlo or other randomization
125
December 1999 2463NEOREFUGIA AND PALEOREFUGIA
Concepts&Synthesis
FIG. 2. Cross-section view of a Blanding
Formation algific talus slope showing major geomorphological
features (adapted from Frest
[1991: Fig. 2]). A complete system has small
sinks in the forested upland; vertical mechanical
karst fissures in the Hopkington Formation; ice
reservoirs in mostly horizontal mechanical karst
in the thin-bedded Blanding Formation; underlying
relatively impervious Tete des Mortes,
Mosalem, and/or Maquoketa formations.
techniques (Hilborn and Mangel 1997), the exact form
of these tests have yet to be formalized.
Hypotheses 3 and 8 will both be hard to test. Determination
of absolute habitat age will be difficult, if
not impossible, to measure in most cases. Even when
it is possible to age habitats through 14C dating of organically
rich basal sediments (peatlands, for example),
it would be difficult or impossible to document total
species richness over this time period. Hypothesis 8 is
difficult to address as it requires determination of which
species represent co-competitors. It is not clear how
one could adequately assess this without knowing, for
a given set of species, all interactions between all potential
limiting environmental variables.
Based on these factors, a priori Hypotheses 1, 2, 4,
and 5 appear to be the most easily testable. To provide
a preliminary assessment for these hypothesized differences,
I have used a paleorefugia system (algific
talus slopes; Frest 1982) and neorefugia system (fens;
Nekola 1994) that co-occur in northeastern Iowa to
examine four specific predictions: (1) the species–area
relationship is steeper and stronger in algific slopes as
compared to fens; (2) the species–isolation relationship
is stronger in fens as compared to algific talus slopes;
(3) distance-decay patterns are more important in fens
as compared to algific talus slopes; and (4) the occurrence
frequency of good and poor dispersers varies
between algific talus slopes and fens.
METHODS
Study sites
Algific talus slopes.—Algific talus slopes (‘‘algific’’
coined from the Greek root ‘‘algos’’ and translates to
‘‘cold producing’’) occur within a deciduous-forest matrix
on steep, usually north-facing, carbonate talus
slopes where cold air flows out of ice-filled caves (Frest
1981; Fig. 2). These caves developed through ice wedging
associated with periglacial activity (Hedges 1972),
and have maintained a below-freezing mean annual
temperature through thermal buffering provided by
overlying cliff-forming carbonates and subtending talus
slopes (Frest 1981, 1982). Algific talus slopes are
characterized by an unique buffered microclimate
where soil temperatures rarely exceed 15Њ C in the summer
(Frest 1982). Approximately 300 algific talus slope
sites have been located across a 10-county region of
northeastern Iowa (USA) in exposures of Prairie-duChien,
Galena, Blanding, or Hopkington Formation
carbonates (Fig. 3; Frest 1981, 1982, 1983, 1984,
1986a, b, 1987). A limited number of sites are also
known from adjacent portions of Minnesota, Wisconsin,
and Illinois, USA (Frest 1991).
Algific talus slope communities range from densely
forested to moderately open. Characteristic woody species
include Acer saccharum, A. spicatum, Abies balsamea,
Betula lutea, B. papyrifera, Fraxinus americana,
Juglans cinerea, Pinus strobus, Salix bebbiana,
Sambucus pubens, Taxus canadensis, and Tilia americana.
The ground layer is typically dominated by a
dense bryophyte cover that supports large populations
of pteridophytes such as Cystopteris fragilis, Gymnocarpium
robertianum, and Equisetum scirpoides.
Woody plant cover decreases and bryophyte/pteridophyte
cover increases in general with increasing intensity
of cold-air seepage. Algific talus slopes harbor populations
of over 60 vascular plant species that are disjunct
in Iowa from wet woods and bogs in northern or
western boreal forests, including Adoxa moschatellina,
Carex media, Cornus canadensis, Linnaea borealis,
Mertensia paniculata, Poa paludigena, Pyrola asarifolia,
Rhamnus alnifolia, Ribes hudsonianum, Streptopus
roseus, and Viola renifolia (Thorne 1964, Frest
1982). At least three near-endemic vascular plant taxa
(Aconitum noveboracense, Chrysosplenium iowense,
and Sullivantia renifolia) occur on these sites. Algific
talus slopes also harbor populations of at least eight
126
2464 Ecology, Vol. 80, No. 8JEFFREY C. NEKOLA
Concepts&Synthesis
FIG. 3. Distribution of algific talus slopes and major bedrock units in northeastern Iowa, USA.
land snail taxa once thought to have become extinct at
the end of the Wisconsinan (Frest 1991). Relict mites
in the family Rhagidiidae have also been discovered
on these sites (V. Ruzicka, personal communication).
These taxa were components of a full-glacial community
associated with quasi-maritime climates found
south of the ice margin (Frest and Dickson 1986). Populations
became restricted to algific talus slopes when
this climatic regime disappeared following the retreat
of glacial ice from the continental interior (Frest 1981).
Due to the microclimatic buffering provided by coldair
seepage, which mimics the full-glacial regional climate,
populations were able to persist on algific talus
slopes through the Hypsithermal to the cooler and wetter
conditions of modern times. Algific talus slopes thus
represent paleorefugia whose biota predates the surrounding
deciduous forest matrix.
Fens.—In northeastern Iowa, fens occur within a
tallgrass-prairie matrix at sites of local groundwater
discharge from pre-Illinoian tills, bedrock, aeolian
sands, fluvial sands, or oxbow-meander cutoffs (Nekola
1994, Fig. 4). Sites typically occur as hillside terraces,
or more rarely as flat expanses in low swales or
mounds. The constant issue of groundwater at fen sites
creates a buffered habitat somewhat similar to algific
talus slopes, with soil temperatures being cooler in the
summer, warmer in the winter, and with more constant
soil moisture than is otherwise found within the surrounding
landscape. Fens are scattered across a 30county
region where at least 2333 sites occurred as
recently as 50 yr ago and where 160 sites remain extant
(Fig. 5). The bulk of presettlement and extant fens are
found on the Iowan Erosional Surface, which was
formed during the Wisconsinan through intense periglacial
erosion that removed older tills, created a
stepped landscape surface, and prevented the accumulation
of deep loess deposits (Hallberg et al. 1978,
Prior 1991). Similar fen habitats are known from northwestern
Iowa (Anderson 1943), Illinois (Moran 1981),
Minnesota (Coffin and Pfannmuller 1988), South Dakota
(Ode 1985), and Wisconsin (Curtis 1959).
In northeastern Iowa, fen habitats are dominated by
a dense sedge turf, which includes Carex buxbaumii,
C. interior, C. lanuginosa, C. lasiocarpa, C. prairea,
C. stricta, C. suberecta, and C. tetanica. In discharge
zones Scirpus validus, Typha angustifolia, and T. la-
127
December 1999 2465NEOREFUGIA AND PALEOREFUGIA
Concepts&Synthesis
FIG. 4. Cross section of the five geologic features that
give rise to fen formation in northeastern Iowa: (A) pre-Illinoian
till sites, with groundwater seepage emanating from
sand and gravel lenses interbedded between till units; (B)
bedrock sites, with groundwater seepage emanating from
fractured carbonate, sandstone, or shale units; (C) aeolian
sand dune sites, with groundwater seepage emanating from
the contact between sand and an impervious till or paleosol;
(D) fluvial sand sites, with groundwater seepage emanating
from the base of fluvial or outwash terraces; and (E) oxbow
sites, which form through hydric succession of cutoff oxbow
meanders.
tifolia are often dominant. Scattered clumps of low
shrubs are also frequent, and include Cornus stolonifera,
Salix bebbiana, S. discolor, S. petiolaris, and S.
rigida. In areas of highest soil saturation, vascular-plant
growth often becomes stunted and sparse, with bryophytes
becoming dominant. Approximately 50% of the
vascular-plant species of these sites are regionally rare
(Nekola 1994). Over 80 of these species are disjunct
from open peatlands in boreal or northeastern North
American, including Aster junciformis, Betula pumila,
Carex sterilis, Epilobium strictum, Eriophorum angustifolium,
Galium labradoricum, Gentiana procera, Lobelia
kalmii, Menyanthes trifoliata, Mimulus glabratus
var. fremontii, Parnassia glauca, Rhynchospora capillacea,
Salix candida, S. pedicellaris, Solidago uliginosa,
Triglochin maritimum, and T. palustris. Northeastern
Iowa fens also harbor populations of 19 rare
Iowa butterfly and skipper species (Nekola 1994), and
seven rare Iowa land snail species (Frest 1990).
14
C dates of basal peat layers indicate that northeastern
Iowa fens initiated peat development following
the end of Hypsithermal warming, less than 6000 yr
ago (Thompson 1992). Although a few sites possess
basal peat dates extending to 10 000 yr ago (Thompson
and Bettis 1994), the presence of oxidized layers within
these beds (Hall 1971, Van Zant and Hallberg 1976)
indicate that deposition was not continuous, and that
during the Hypsithermal even these sites dried out.
Thus, although fens may have occurred at modern locations
in the late-glacial landscape, the geological record
indicates that they did not maintain environmental
buffering during the Hypsithermal. Because of this,
isolated populations of boreal taxa occurring within
these sites cannot represent late Pleistocene or early
Holocene relicts, but are instead relatively recent immigrants.
In northeastern Iowa, therefore, fens represent
neorefugia that are younger than their surrounding
tallgrass-prairie matrix.
Data sets
Species richness.—The ecological distance between
fen and algific talus slope habitats and the surrounding
matrix habitats is not so great as to preclude occurrence
of matrix taxa. Such species can colonize sites via
short-range dispersal or mass effect (Shmida and Ellner
1984), independent of the colonization history of a given
site. As such, inclusion of these ubiquitous or waif
species could seriously swamp any potential ecological
differences caused by differential colonization histories.
By limiting analysis to only those species restricted
to fens or algific talus slopes, this source of noise
can be filtered out. For this reason, observations of
species composition and richness for fen and algific
talus slope sites was limited to the universe of vascularplant
taxa restricted (Ͼ80% of known northeastern
Iowa occurrences) to either of these habitats. In total,
80 restricted vascular-plant species were identified
from fens and 51 from algific slopes, based on occurrence
records reported in Frest (1982), Howe et al.
(1984), and Nekola (1990, 1994), as well as field data
collected during this study. The presence or absence of
each of these taxa was subsequently noted for all 160
extant northeastern Iowa fens and all 76 extant algific
talus slopes in the Blanding Formation outcrop region.
Habitat size.—Sizes of fen sites were estimated
through digitization of soil pedons (named variously
128
2466 Ecology, Vol. 80, No. 8JEFFREY C. NEKOLA
Concepts&Synthesis
FIG. 5. Presettlement and 1991 distribution of northeastern Iowa fens within major landform boundaries, based on Prior
(1991).
129
December 1999 2467NEOREFUGIA AND PALEOREFUGIA
Concepts&Synthesis
as Palms Muck, Houghton Muck, Muck, or Peat soils)
associated with fen communities on USDA Soil Conservation
Service soil maps. Sizes of algific talus slope
sites were determined through digitization of site
boundaries recorded from field observations on 7.5minute
U.S. Geological Survey topographic maps.
Isolation.—To quantify the isolation of sites, a
weighted summation was calculated for the distance
between each site and all others, assuming that the
influence of surrounding sites exponentially decreases
as intersite distance increases:
n
Ϫcd
I ϭ e͸iϭ1
where I ϭ isolation index, n ϭ number of other sites
in a landscape, d ϭ distance between the reference site
and site i, and c ϭ a constant modifying the rate of
exponential decay
The smaller the value of this index for a site, the
less it is expected to interact with surrounding sites.
This index differs from a summed addition of site distances
as it down-weights the importance of distant
areas to the isolation of a given site. Justification of
this form of weighting is based upon theoretical (Okubo
and Levin 1989) and quantitative (Preston 1962, Nekola
and White, in press) studies that have shown that
dispersal ability and community similarity tend to exhibit
an exponential decay with increasing intersample
distance.
No a priori rules exist to suggest the appropriate
value of c, the constant that modifies the level of downweighting
of distant sites. To allow a robust test of the
effect of isolation on species richness, isolation-index
scores were calculated over five different values of c
(1.0, 0.5, 0.25, 0.1, and 0.05), corresponding to five
different neighborhood radii (roughly equivalent to 3,
6, 12, 30, and 60 km) that demarcate those sites that
most strongly contribute to isolation values. Sites that
fall outside these neighborhoods contribute very little
to the overall isolation-index values for a given loca-
tion.
Statistical analysis
Prediction 1: The species–area relationship will be
steeper and stronger in algific slopes as compared to
fens.—The slope of the vascular plant species–area relationship
was estimated by multiple least-squares linear
regression after both the species-richness and habitat-area
values had been log transformed. A double
log transform was used as these residuals were more
homoscedastic than those generated from untransformed
richness vs. log-transformed area. To allow inclusion
of sites with zero species, a one was added to
the species richness of each site prior to log transformation.
Tests for differences between the slopes and
intercepts of the two species–area relationships were
obtained by analysis of a binary variable representing
habitat type that was added into the model following
methods outlined in Kleinbaum et al. (1988). The form
of this regression model was
ln(species richness ϩ 1)
ϭ ␤0 ϩ ␤1(ln(habitat size)) ϩ ␤2(habitat type)
ϩ ␤3(habitat type ϫ ln(habitat size)) ϩ ␧ (1)
where ␤0 ϭ intercept for algific slopes, ␤1 ϭ slope for
algific slopes, ␤2 ϭ difference in intercept between algific
slopes and fens, and ␤3 ϭ difference in slope
between algific slopes and fens.
As fens can be significantly larger than algific talus
slopes, comparison of species–area slopes and intercepts
calculated from the entire data set may lead to
inaccurate conclusions (Conner and McCoy 1979). To
compensate, the regression analysis was repeated using
only those sites falling within the habitat-size overlap
between algific talus slopes and fens.
The relative strength of the species–area relationship
was determined through comparison of the correlation
coefficients generated through separate regressions of
log-transformed richness vs. log-transformed area for
each habitat. The significance of observed differences
between these r values was quantified using the correlation-coefficient
homogeneity test (Sokal and Rohlf
1981).
Prediction 2: Species–isolation relationships will be
stronger in fens as compared to algific talus slopes.—
The significance and amount of extra variance in vascular-plant
species richness explained by site isolation
was estimated by separately adding each of the five
isolation-index values into the previously defined species–area
model. The form of this regression model,
for each of the isolation-index values, was
ln(species richness ϩ 1) ϭ ␤0 ϩ ␤1(ln(habitat size))
ϩ ␤2(isolation index) ϩ ␧ (2)
where ␤0 ϭ intercept, ␤1 ϭ slope of habitat-size relationship,
and ␤2 ϭ slope of isolation relationship. The
amount of additional variance each of the isolationindex
values explained was determined through subtraction
of r2
values generated from these models from
the r2
values observed from the species-richness vs.
habitat-size regression.
Prediction 3: Distance-decay patterns will be more
pronounced in fens as compared to algific talus
slopes.—Distance-decay rates in the restricted vascular-plant
floras were estimated in both habitats using
the procedures outlined in Nekola and White (in press).
These analyses were conducted at five different sample
grain sizes: single sites, plus grid cells of 10, 20, 40,
and 80 km on a side laid over the study region. However,
80 ϫ 80 km grid cells were not analyzed for algific
talus slopes because of their more limited extent. The
flora of a given cell was determined by summing together
the floras of all sites whose centroids fell within
the boundaries of that cell. The centroid location of
each cell was set as the average location of all sites
130
2468 Ecology, Vol. 80, No. 8JEFFREY C. NEKOLA
Concepts&Synthesis
FIG. 6. Species–area curves (Eq. 1). Log-transformed restricted-species richness as a function of log-transformed habitat
size for northeastern Iowa algific talus slope and fen habitats. Each group of three curves shows the position of best-fit
regression line and 95% confidence interval around the mean of each.
found within that quadrat. Floristic similarity between
samples was calculated using Jaccard’s index, and distance
was calculated as the aerial separation between
cell centroids. The amount of variance in similarity
explained by distance was estimated through a linear
regression of natural-log-transformed similarity vs. untransformed
distance (Nekola and White, in press). The
significance of the similarity–distance relationship was
estimated by Mantel matrix randomization tests (RT
software package, Manley 1992). All Mantel P values
were generated using 10 000 replications, as suggested
by Jackson and Somer (1989) for biological-data sets.
Prediction 4: The proportional frequency of species
with good or poor dispersal abilities will vary between
the algific talus slope and fen flora.—Each restricted
vascular-plant species found in either of the two habitat
types was assigned to one of four dispersal classes
expected to differ in their dispersal abilities: (1) macroscopic
(Ͼ0.1 mm), non-plumose seeds in dry fruits
or capsules; (2) fleshy berries, fruits, or nuts; (3) microscopic
(Ͻ0.01 mm) seeds or spores; and (4) plumose
seeds. The frequency of these four categories in the
flora of each habitat was determined, and differences
tested using the Pearson chi-square statistic.
RESULTS
Prediction 1: The species–area relationship will be
steeper and stronger in algific slopes as compared
to fens
Both habitats demonstrate a significant positive relationship
between species richness and habitat size
(Fig. 6, Table 1), with both slope and intercept for
algific talus slopes being significantly (P Ͻ 0.0005)
larger than those reported for fens. When this analysis
was repeated for only those sites within the range of
habitat-size overlap, the slope and intercept were again
found to be significantly larger in algific slopes (P Ͻ
0.017 and P Ͻ 0.0005, respectively). In both analyses,
small algific talus slopes had fewer vascular-plant species
than fens of identical size, whereas large algific
talus slopes had greater richness than fens of identical
size.
The amount of variance explained in these regressions
also differed greatly (Table 2). Log-transformed
habitat size accounted for 32.6% of observed variance
in algific talus slope species richness but accounted for
only 6.3% of the variance in fen species richness. This
131
December 1999 2469NEOREFUGIA AND PALEOREFUGIA
Concepts&Synthesis
TABLE 1. Results of multiple-regression analysis of the effect of habitat area and habitat type
on restricted vascular-plant species richness in northeastern Iowa fen and algific talus slope
habitats (see Eq. 1). Restricted species are those present only in fens or algific talus slopes.
Habitat N Value 1 SE t P
Algific talus slope sites 76
Intercept, ␤0
Slope, ␤1
2.335
0.542
0.102
0.090
22.881
6.003
Ͻ0.0005
Ͻ0.0005
Difference between algific slopes
and fen sites 160
Intercept difference, ␤2
Slope difference, ␤3
Ϫ0.476
Ϫ0.428
0.115
0.097
Ϫ4.139
Ϫ4.421
Ͻ0.0005
Ͻ0.0005
Notes: N ϭ no. of observations. There were 160 extant fens and 76 extant algific talus slopes
in the Blanding Formation outcrop region.
TABLE 3. Summary statistics for restricted vascular-plant
species richness vs. area and isolation in northeastern Iowa
fen and algific talus slope habitats.
Exponen-
tial
decay
coefficient
Algific talus slopes
P
Additional
r2
Fens
P
Additional
r2
c ϭ 1.0
c ϭ 0.5
c ϭ 0.25
c ϭ 0.1
c ϭ 0.05
0.150
0.266
0.497
0.802
0.778
0.019
0.012
0.005
0.001
0.001
0.027
0.008
0.003
0.003
0.008
0.029
0.041
0.053
0.050
0.041
TABLE 2. Significance and r2 for the restricted vascular-plant
species richness vs. habitat-area relationship in northeastern
Iowa fen and algific talus slope habitats.
Habitat
No. of
observations P r2
Fens
Algific talus slopes
160
76
0.001
0.000
0.063
0.326
Note: Significance of the correlation-coefficient homogeneity
test: P ϭ 0.006.
difference was found to be highly significant (P ϭ
0.006).
Prediction 2: Species–isolation relationships will be
stronger in fens as compared to algific talus slopes
Isolation was found to be a significant additional
predictor of vascular-plant species richness in fens for
each of the calculated isolation-index values (P ranging
from 0.003–0.027; Table 3). The amount of additional
variance explained by isolation in fen habitats varied
between 0.026 and 0.053, or between 46% and 84% of
the variance explained by log-transformed habitat area.
However, for algific talus slopes, isolation was never
a significant additional predictor of species richness at
any of the isolation-index scores (P ranging between
0.150 and 0.802).
Prediction 3: Distance–decay patterns will be more
pronounced in fens as compared to
algific talus slopes
In fens, significant (P ϭ 0.0001) distance–decay relationships
were observed between the restricted species
floras at all grain sizes (Table 4). At the largest
grain (80 ϫ 80 km regions), this relationship explained
over one third of the variance in regional restricted
vascular-plant species composition (Table 5). At smaller
grains, distance accounted for Ͻ5% of observed variance
in floristic similarity. In algific slopes, only the
distance–decay relationship at the 20 ϫ 20 km sample
grain size was found to be significant at the 0.05 level.
At this scale, 23% of the observed variance in vascularplant
assemblage similarity was accounted for by intersample
distance.
Prediction 4: The proportional frequency of species
with good or poor dispersal abilities will vary
between the algific talus slope and fen flora
The frequencies of dispersal classes differed significantly
(P Ͻ 0.0005) between the two habitats (Table
6). Species with berries, fleshy fruits, or nuts were ϳ15
times more frequent in algific slopes than fens, whereas
species with plumose seeds were ϳ7 times more frequent
in fens than algific slopes.
DISCUSSION
Distinguishing among alternative explanations
In each case the results obtained agree with the a
priori hypotheses, which predicted differences between
the ecological patterns of paleorefugia and neorefugia
and suggest that extinction and immigration have
played differing roles in algific talus slope and fen
communities (Table 7). However, it is impossible to
control for all confounding factors in a natural experiment
such as this. By explicitly stating what some of
these factors may be, the preliminary nature of these
investigations, and potential avenues for future research,
may be made more clear.
Two potential confounders stem from the different
source pools for the flora of these two habitats, and
from the different landscape dynamics of the primary
disturbances that effect them. Although the flora of
each habitat is a subset of the boreal wetland flora, the
species of algific slopes have greater affinity with conifer
swamps while the species of fens have greater
132
2470 Ecology, Vol. 80, No. 8JEFFREY C. NEKOLA
Concepts&Synthesis
TABLE 4. P values for the distance–decay relationships of
the restricted vascular-plant floras of northeastern Iowa fen
and algific talus slope habitats at differing grain sizes.
Habitat
Grain size (km)
80 40 20 10
Individual
sites
Fen
Algific slope
0.0002
···
0.0001
0.6245
0.0001
0.0213
0.0001
0.2832
0.0001
0.2339
Note: P values were calculated using Mantel tests on distance
and similarity matrices for comparisons between each
region or site.
TABLE 5. The r2 values for the distance–decay relationships
of the restricted vascular-plant floras of northeastern Iowa
fen and algific talus slope habitats at differing grain sizes.
Habitat
Grain size (km)
80 40 20 10
Individual
sites
Fen
Algific slope
0.335
···
0.014
0.082
0.044
0.226
0.034
0.005
0.027
0.001
Note: The r2 values were calculated using linear regression
on log-transformed data.
TABLE 6. Frequency of four dispersal classes among restricted
vascular-plant species in northeastern Iowa fen and
algific slopes. The numbers in parentheses represent the
percentage of the total number of species recorded in each
habitat contained in each dispersal class.
Habitat
Dispersal class†
1 2 3 4
Fen
Algific
slope
42 (52.5%)
26 (51.0%)
1 (1.2%)
10 (19.6%)
17 (21.3%)
13 (25.4%)
20 (25.0%)
2 (3.9%)
Note: Chi-square statistic for differences in frequencies for
all dispersal classes ϭ 20.998; P Ͻ 0.0005.
† Dispersal classes: 1 ϭ macroscopic, non-plumose seeds
in dry capules; 2 ϭ berry/fleshy fruits; 3 ϭ spore or microscopic
seeds; and 4 ϭ plumose seeds.
similarity with open peatlands (e.g., Curtis 1959). Differences
in the composition, coexistence patterns, and
dispersal strategies of species in these pools could have
contributed to, or created, observed differences. Additionally,
algific slopes are occasionally impacted by
small-scale talus dislodgement from game trails or individual
tree-falls, whereas fens are more often affected
by larger-scale events such as fires and regional
droughts. Consequently, algific slopes seem more likely
than fens to be in a state of successional quasiequilibrium
(Shugart 1984, Turner et al. 1994). This
difference could potentially account for the stronger
species–area relationship in algific slopes, or for the
predominance of vagile species in fens.
Additional research will be necessary to tease apart
the effects of these and other factors from effects of
colonization history. For example, it would be instructive
to extend these analyses of algific slope and fen
vascular-plant floras to additional taxonomic groups
found on the same sites that possess varying dispersal
abilities, such as bryophytes, birds, lepidoptera, or terrestrial
gastropods. Such analyses would help document
whether the patterns observed for vascular plants
in this investigation are general in nature, and whether
species groups with different dispersal abilities will
respond to identical levels of geographic isolation and
patterns of habitat creation in the same way. It is possible
that poorly dispersing taxa (snails, for example)
may be more influenced by extinction and that morevagile
species (lepidoptera, for example) may be more
influenced by immigration processes on the same sites.
Important additional insights may also be garnered
if paleorefugia and neorefugia are studied over a more
diverse set of environmental situations and landscapes.
For instance, comparing identical habitats that have
formed at different times in different landscapes (creating
a neorefugia in one and a paleorefugia in the
other) may be a good way to control for the confounding
effects of species source pools or disturbance dynamics.
It may also be interesting to study the contrast
between these refugia types across a series of landscapes
in which sites vary from highly to less isolated.
Neorefugia and paleorefugia of very ancient origin may
also demonstrate different biogeographic patterns. For
instance, some ancient oceanic islands represent neorefugia
(like the Hawaiian archipelago, which formed
through volcanic eruptions) whereas others represent
paleorefugia (like New Zealand and New Caledonia,
which were created through fragmentation of continental
land masses). As time has been sufficient for
speciation to occur, a strong species–area relationship
should develop for both. However, the fact that dispersal
was originally more important for ancient oceanic
neorefugia should allow for stronger species–isolation
relationships, and different dispersal-strategy
frequencies in their contemporaneous biota as compared
to ancient oceanic paleorefugia. Hypotheses regarding
paleorefugia and neorefugia are also amenable
to some forms of manipulative experimentation for
those processes that operate over limited spatial and
temporal extents, or to computer simulations for largerscale
patterns. The future research priorities regarding
paleorefugia and neorefugia should include more taxa,
more habitats, more landscapes, and more times of origin.
Through such additional work a clearer understanding
of the role of colonization history on contemporaneous
ecological pattern can be achieved.
Implications for ecology, biogeography, and
conservation
The central question of this paper is whether the
historical contingency of habitat formation can have a
long-lasting impact on community patterns and processes.
The answer to this question has important implications
for ecological debates surrounding equilibrium
vs. non-equilibrium dynamics, vicariance vs.
long-range dispersal, and the delineation of optimum
reserve designs for maintenance of biological diversity.
133
December 1999 2471NEOREFUGIA AND PALEOREFUGIA
Concepts&Synthesis
TABLE 7. Summary outcomes of analysis: observed and predicted differences between algific talus slopes (paleorefugia)
and fens (neorefugia) in northeastern Iowa.
Factor
Response
Neorefugia Paleorefugia Result†
Relative steepness and strength of species–area relationship Less Greater Y
Strength of species–isolation relationship High Low Y
Change in richness over time Increasing Decreasing ···
Strength of distance-decay relations High Low Y
Dispersal ability of species in habitat Good Poor Y
Relative importance of geography in predicting species occurrence Strong Weak ···
Strength of community–environment correlation Weak Strong ···
Relative number of co-competitors in landscape More Less ···
† Y, relationship found statistically significant and in agreement with a priori predictions; ···, not addressed in this analysis.
Equilibrium vs. non-equilibrium dynamics.—Although
much classical ecological theory (e.g., MacArthur
and Wilson 1967, MacArthur 1972, Tilman
1988) assumes existence of equilibrium conditions, the
data presented above suggest that differences in colonization
history may still imprint biotas 5000–10 000
yr following habitat development. If ‘‘transient dynamics’’
continue to exist after such time periods, equilibrium
conditions cannot be assumed to dominate biogeographic
pattern. These results support paleoecological
work (Davis 1984) that suggests that disequilibrium
between climate and species range may persist
over relatively long temporal extents.
Vicariance vs. long-distance dispersal.—The relative
importance of vicariance vs. long-distance dispersal
in the development of biogeographic pattern has
been debated since the times of Wallace, Schimper,
Engler, and Hooker (Cain 1944). Many ecologists view
dispersal as so efficient that no barriers to movement
exist for any species within continental areas, making
biogeographic pattern simply the result of physiological
limitations and competitive interactions (e.g.,
Krebs 1985). Saur (1988) has summarized this belief
as Beijerinck’s Law: everything is everywhere but the
environment selects. However, some cladistic biogeographers
have stated that dispersal either does not occur
between isolated habitats, or is so unpredictable that it
cannot be studied. For these researchers, only the process
of vicariance can parsimoniously explain disjunct
distributions (e.g., Humphries and Parenti 1986, Brudin
1988).
Analysis of paleorefugia and neorefugia suggest that
such absolute positions may be fundamentally flawed
as both vicariance and dispersal can operate contemporaneously
in the same landscape. Instead of asking
which of these processes is responsible for observed
biogeographic patterns, it seems more productive to
determine how much each has contributed to that pattern.
If additional research confirms generality to the
patterns attributed above, it may be possible to use
similar analyses in additional landscapes to document
the occurrence of neorefugia and paleorefugia, and thus
the frequency of vicariance and long-distance dispersal.
Conservation of biological diversity.—These preliminary
analyses suggest that algific talus slopes are floristically
more self-contained than fens, with immigration
playing a relatively minor role in the determination
of species occurrence. Protection of the vascular flora
of these habitats should thus center on those sites that
harbor maximum diversity and the internal site dynamics
that are responsible for that diversity.
However, these preliminary analyses also suggest
that fens are not as self-contained, with immigration
from surrounding sites strongly influencing vascularplant
composition. Diversity in these communities varies
conversely with isolation, and community composition
is strongly influenced by the proximity of source
areas. To ensure adequate protection of the processes
that gave rise to site diversity in such systems, reserves
may have to be chosen to protect not only internal site
dynamics, but also the immigration potential between
sites. This will be possible only if many sites are protected
across the landscape, including those that may
not currently support diverse floras. Consequently, neorefugia
reserve design may require an integrated, regional
strategy.
Unfortunately, the reality of fen destruction in Iowa
may make such goals unrealistic. Since European settlement,
93% of northeastern Iowa fens have been destroyed
(Nekola 1994; Fig. 5). Even if all remaining
sites could be saved, immigration rates cannot be maintained
at presettlement levels. The ultimate effect of
this will be severely lowered species-immigration rates
compared to presettlement times, and low probabilities
that species will be able to reestablish themselves following
local extinction. If extinction rates are high,
strategies to guard fen biodiversity may require not
only protection of all remaining sites, but also the physical
transport of propagules between sites to reestablish
presettlement immigration levels. If this approach is
eventually deemed necessary, it would require creation
of stochastic immigration models to guide land managers
in their attempts to simulate ‘natural’ immigration
between sites.
ACKNOWLEDGMENTS
This research was supported by a National Science Foundation
Graduate Fellowship and by the McElroy Foundation
of Waterloo, Iowa. I wish to thank John and Gretchen Bray-
134
2472 Ecology, Vol. 80, No. 8JEFFREY C. NEKOLA
Concepts&Synthesis
ton, Bob Moats, Richard Baker, Willard Hawker, Louise
McEarchen, John Nehnevaj, Dean Roosa, Arlan Kirchman,
Dennis Schlicht, Bill Thomas, Robert Thomson, Dave Wendling,
and Herb Wilson for providing me with locality information
for extant fen sites. Terrence Frest was instrumental
in instructing me in the geology and ecology of algific talus
slopes, and initially located many of the Blanding outcrop
sites inventoried in this analysis. Peter White, Michael Palmer,
Susan Wiser, James Hazen Graves, Robert Howe, Clifford
Kraft, Ed Connor, Paul Keddy, three anonymous reviewers,
and especially Robert K. Peet read and made useful comments
regarding earlier drafts of this manuscript.
LITERATURE CITED
Anderson, W. A. 1943. A fen in northwestern Iowa. American
Midland Naturalist 29:787–791.
Austin, M. P., A. O. Nicholls, and C. R. Margules. 1990.
Measurement of the realized qualitative niche: environmental
niches of five Eucalyptus species. Ecological Monographs
60:161–177.
Bennett, D. G., and O. Gade. 1979. Geographic perspectives
in migration research. Department of Geography Studies
in Geography, Number 12. University of North Carolina,
Chapel Hill, North Carolina, USA.
Bierrgaard, R. O., Jr., T. E. Lovejoy, V. Kapos, A. Augusto
dos Santos, and R. W. Hutchings. 1992. The biological
dynamics of tropical rainforest fragments. BioScience 42:
859–866.
Brown, J. H. 1971. Mammals on mountaintops: non-equilibrium
insular biogeography. American Naturalist 105:
476–478.
. 1995. Macroecology. University of Chicago Press,
Chicago, Illinois, USA.
Brown, J. H., and A. Kodric-Brown. 1977. Turnover rates
in insular biogeography: effect of immigration on extinction.
Ecology 58:445–449.
Brudin, L. Z. 1988. Phylogenetic biogeography. Pages 343–
370 in A. A. Meyers and P. S. Giller, editors. Analytical
biogeography. Chapman & Hall, New York, New York,
USA.
Burrough, P. A. 1986. Principles of geographical information
systems for land resources assessment. Oxford University
Press, New York, New York, USA.
Bush, M. B., and R. J. Whittaker. 1993. Non-equilibrium in
island theory of Krakatau. Journal of Biogeography 20:
453–457.
Cain, S. A. 1944. Foundations of plant geography. Harper
& Row, New York, New York, USA.
Carlquist, S. 1974. Island biology. Columbia University
Press, New York, New York, USA.
Coffin, B., and L. Pfannmuller. 1988. Minnesota’s endangered
flora and fauna. University of Minnesota Press, Minneapolis,
Minnesota, USA.
Conner, E., and E. D. McCoy. 1979. The statistics and biology
of the species–area relationship. American Naturalist
113:791–833.
Cox, G. W., and R. E. Ricklefs. 1977. Species diversity,
ecological release, and community structuring in Caribbean
land bird faunas. Oikos 29:60–66.
Curtis, J. T. 1959. The vegetation of Wisconsin. University
of Wisconsin Press, Madison, Wisconsin, USA.
Davis, M. B. 1984. Quaternary history and the stability of
forest communities. Pages 132–151 in D. C. West, H. H.
Shugart, and D. B. Botkin, editors. Forest succession.
Springer-Verlag, New York, New York, USA.
Diamond, J. M. 1973. Distributional ecology of New Guinea
birds. Science 179:759–769.
. 1975. The island dilemma: lessons of modern biogeographic
studies for the design of nature reserves. Biological
Conservation 7:129–146.
Fernald, M. L. 1925. Persistence of plants in unglaciated
North America. Memoirs of the American Academy of Arts
and Sciences 15:241–342.
Fotherinham, A. S. 1981. Spatial structure and distance-decay
parameters. Annals of the Association of American
Geographers 71:425–436.
Frest, T. J. 1981. Final report, project SE-1-2 (Iowa Pleistocene
Snail). Iowa Conservation Commission, Des
Moines, Iowa, USA.
. 1982. Final report, project SE-1-4 (Iowa Pleistocene
Snail). Iowa Conservation Commission, Des Moines, Iowa,
USA.
. 1983. Final report (contract number 30181-1259)
northern Driftless Area survey. U.S. Fish and Wildlife Service,
Fort Snelling, Minnesota, USA.
. 1984. Final report, project SE-1-6 (Iowa Pleistocene
Snail). Iowa Conservation Commission, Des Moines, Iowa,
USA.
. 1986a. Final report, project SE-1-6 number 2 (Iowa
Pleistocene Snail). Iowa Department of Natural Resources,
Des Moines, Iowa, USA.
. 1986b. Final report, project SE-1-7 (Iowa Pleistocene
Snail). Iowa Department of Natural Resources, Des
Moines, Iowa, USA.
. 1987. Final report, project SE-1-8 (Iowa Pleistocene
Snail). Iowa Department of Natural Resources, Des
Moines, Iowa, USA.
. 1990. Final report (contract number 65-2454). Field
survey of Iowa spring fens. Iowa Department of Natural
Resources, Des Moines, Iowa, USA.
. 1991. Summary status reports on eight species of
candidate land snails from the Driftless Area (Paleozoic
Plateau), Upper Midwest. Final report (contract number
301-01366). U.S. Fish and Wildlife Service, Fort Snelling,
Minnesota, USA.
Frest, T. J., and J. R. Dickson. 1986. Land snails (Pleistocene-recent)
of the Loess Hills: a preliminary survey. Proceedings
of the Iowa Academy of Science 93:139–157.
Hall, S. A. 1971. Paleoecological interpretation of bison,
mollusks, and pollen from the Hughes peat bed, Linn County,
Iowa. Thesis. University of Iowa, Iowa City, Iowa, USA.
Hallberg, G. R., T. E. Fenton, G. A. Miller, and A. J. Lutenegger.
1978. The Iowan Erosion Surface: an old story,
an important lesson, and some new wrinkles. Pages 2.1 to
2.94 in R. Anderson, editor. Geology of East-central Iowa,
42nd annual tri-state geological field conference. Iowa
Geological Survey, Iowa City, Iowa, USA.
Harper, J. L. 1977. Population biology of plants. Academic
Press, New York, New York, USA.
Hedges, J. 1972. Expanded joints and other periglacial phenomena
along the Niagara Escarpment. Biuletyn Peryglacjalny
21:87–126.
Hilborn, R., and M. Mangel. 1997. The ecological detective.
Princeton Monographs in Population Biology number 28.
Princeton University Press, Princeton, New Jersey, USA.
Howe, R. W., M. J. Huston, W. P. Pusateri, R. H. Laushman,
and W. E. Schennum. 1984. An inventory of significant
natural areas in Iowa. Iowa Conservation Commission, Des
Moines, Iowa, USA.
Humphries, C. J., and L. R. Parenti. 1986. Cladistic biogeography.
Clarendon Press, Oxford, UK.
Jackson D. A., and K. M. Somer. 1989. Are probability
estimates from the permutation model of Mantel’s test stable?
Canadian Journal of Zoology 67:766–769.
Kleinbaum, D. G., L. L. Kupper, and K. E. Muller. 1988.
Applied regression analysis and other multivariate methods.
PWS-Kent Publishing, Boston, Massachusetts, USA.
Krebs, C. J. 1985. Ecology. Harper & Row, New York, New
York, USA.
Krukeberg, A. R., and D. Rabinowitz. 1985. Biological as-
135
December 1999 2473NEOREFUGIA AND PALEOREFUGIA
Concepts&Synthesis
pects of endemism in higher plants. Annual Review of
Ecology and Systematics 16:447–479.
Manley, B. F. J. 1992. RT—a program for randomization
testing. Version 1.02. The Centre for Applications of Statistics
and Mathematics, University of Otago, Dunedin,
New Zealand.
MacArthur, R. H. 1972. Geographical ecology. Harper &
Row, New York, New York, USA.
MacArthur, R. H., and E. O. Wilson. 1967. The theory of
island biogeography. Princeton University Press, Princeton,
New Jersey, USA.
McDonald, K. A., and J. H. Brown. 1992. Using montane
mammals to model extinctions due to global change. Conservation
Biology 6:409–415.
Moran, R. C. 1981. Prairie fens in northeastern Illinois: floristic
composition and disturbance. Pages 164–168 in R.
Stuckey and K. J. Reece, editors. Proceedings of the sixth
North American Prairie Conference. The Ohio State University,
Columbus, Ohio, USA.
Nekola, J. C. 1990. Rare Iowa plant notes from the R. V.
Drexler Herbarium. Proceedings of the Iowa Academy of
Science 97:55–73.
. 1994. The environment and flora of northeastern
Iowa fen communities. Rhodora 96:121–169.
Nekola, J. C., and P. S. White. In press.The distance decay
of similarity in ecology and evolution. Journal of Bioge-
ography.
Ode, D. J. 1985. Rediscovered plants of fen habitats of South
Dakota. Proceedings of the South Dakota Academy of Science
64:217.
Okubo, A., and S. A. Levin. 1989. A theoretical framework
for data analysis of wind dispersal of seeds and pollen.
Ecology 70:329–338.
Peet, R. K., and N. L. Christensen. 1988. Changes in species
diversity during secondary forest succession on the North
Carolina Piedmont. Pages 233–245 in H. J. During, M. J.
A. Werger, and J. H. Willems, editors. Diversity and pattern
in plant communities. SPB Academic Publishing, The
Hague, The Netherlands.
Pielou, E. C. 1979. Biogeography. John Wiley & Sons, New
York, New York, USA.
Preston, F. W. 1962. The canonical distribution of commonness
and rarity: part II. Ecology 43:410–432.
Prior, J. C. 1991. Landforms of Iowa. University of Iowa
Press, Iowa City, Iowa, USA.
Roughgarden, J., S. D. Gaines, and S. W. Pacala. 1987. Supply-side
ecology: the role of physical transport processes.
Symposium of the British Ecological Society 27:491–518.
Saur, J. D. 1988. Plant migration. University of California
Press, Berkeley, California, USA.
Shmida, A., and S. Ellner. 1984. Coexistence of plant species
with similar niches. Vegetatio 58:29–55.
Shugart, H. 1984. A theory of forest dynamics. SpringerVerlag,
New York, New York, USA.
Simberloff, D. S., and E. O. Wilson. 1969. Experimental
zoogeography of islands: a model for insular conservation.
Ecology 50:296–314.
Sokal, R. R., and F. J. Rohlf. 1981. Biometry. W. H. Freeman,
San Francisco, California, USA.
Stebbins, G. L., and J. Major. 1965. Endemism and speciation
in the California flora. Ecological Monographs 35:1–
35.
Thompson, C. A., and E. A. Bettis III. 1994. Age and development
of Iowa fens. Proceedings of the Iowa Academy
of Science 101:73–77.
Thorne, R. F. 1964. Relict nature of the flora of White Pine
Hollow Forest Reserve, Dubuque County, Iowa. State University
of Iowa Studies in Natural History 20:1–33.
Tilman, D. 1988. Plant strategies and the dynamics and structure
of plant communities. Princeton University Press,
Princeton, New Jersey, USA.
Tobler, W. R. 1970. A computer movie simulating urban
growth in the Detroit region. Economic Geography 46:234–
240.
Turner, M. G., W. H. Romme, and R. H. Gardner. 1994.
Landscape disturbance models and the long-term dynamics
of natural areas. Natural Areas Journal 14:3–11.
Van Zant, K. L., and G. R. Hallberg. 1976. A late-glacial
pollen sequence from northeastern Iowa: Sumner Bog revisited.
Iowa Geological Survey Technical Informational
Bulletin Series number 3.
Zobel, M. 1997. The relative role of species pools in determining
plant species richness: an alternative explanation
of species coexistence. Trends in Ecology & Evolution 12:
266–269.
136
Section IV: Biogeography and Macroecology
The ecological processes that most
influence global biodiversity operate at
spatial and temporal scales too large for
typical manipulative experiments. To
understand the impact of such drivers
requires observational studies based on
multiple sites that encompass large extents
and/or represent natural experiments.
Identification of general mechanism in
biogeography and macroecology requires
that pattern and process be investigated
over a wide range of organisms
representing many different body plans,
trophic states, dispersal strategies and
reproductive modes. In spite of this, the
great majority of work is limited to vertebrates, vascular plants, and a few anthropods (e.g.
lepidoptera). Land snails serve as an important potential alternative group because of their very
poor active but often excellent passive dispersal abilities and hermaphroditism with common
instances of unipartental reproduction. As a result, this group displays a number of unique
patterns. Companion articles published in 2014 document that continental (as well as most
regional) faunas display strongly bimodal body-size distributions, as opposed to the typical rightskewed
pattern. Additionally, they display strong scale-dependence, with site faunas tending to
be dominated by small (<5.7 mm³) species while regional/continental faunas are characterized by
large (>90.5 mm³) taxa. This pattern is more pronounced at lower latitudes and appears driven
by a strong inverse correlation between dispersal ability and body size, with small species
tending to have much larger ranges and lower levels of allopatric replacement across a given
geographic gradient Prior work has also shown that land snail communities demonstrate high
levels of microsympatry with up to 25% of a regional fauna co-existing within 0.04 m² areas.
My best known contribution in this field is my 1999 paper with Peter White which popularized
Distance Decay analyses to the field of Ecology. This work is now considered by many to be a
classic in the field with this pattern being one of the major statistics used to test Macroecological
hypotheses. I am also interested in documenting the nature and causes of spatial heterogeneity in
occurrence patterns, the spread of organisms across landscapes, and continental gradients of
body size variation within and between communities. And as intimated in Section III, I am using
macroecological pattern to inform about macroevolutionary process in Holarctic land snails.
Representative Publications
[number of citations as of October 27, 2017]
1. Edited Volumes
Cameron, R.A.D., J.C. Nekola, B.M. Pokryszko & F.E. Wells. 2005. Pattern and process in
land mollusc diversity. Records of the Western Australian Museum, Supplement #68.
137
2. Peer-reviewed Articles
Nekola, J.C. 2014. North American terrestrial gastropods through either end of a spyglass.
Journal of Molluscan Studies. 80:238-248. [7]
Nekola, J.C. 2013. Biodiversity refuges. pp. 141-148 in: MacLeod, N., D.J. Archibald & P.
Levin (eds), Grzimek's Animal Life Encyclopedia: Extinction. Gale Publishing,
Farmington Hills, Michigan. [6]
Nekola, J.C., G.M. Barker, R.A.D. Cameron & B.M. Pokryszko. 2013. Latitudinal variation of
body size in land snail populations and communities. pp. 62-82 in: F. Smith & K. Lyons
(eds), Global Patterns of Body Size. University of Chicago Press. [12]
Nekola, J.C. 2005. Geographic variation in richness, evenness, and shell size of eastern North
American land snail communities. Records of the Western Australian Museum.
Supplement 68:39-51. [39]
Nekola, J.C. 2003. Large-scale terrestrial gastropod community composition patterns in the
Great Lakes region of North America. Diversity and Distribution. 9:55-71. [52]
Nekola, J.C. & P.S. White. 1999. Distance decay of similarity in biogeography and ecology.
Journal of Biogeography. 26:867-878. [1069]
138
North American terrestrial gastropods through each end of a spyglass
Jeffrey C. Nekola
Biology Department, University of New Mexico, Albuquerque, NM 87131, USA
Correspondence: J. C. Nekola; e-mail: jnekola@unm.edu
(Received 30 December 2013; accepted 3 March 2014)
ABSTRACT
Some suggest that because of scale independence major biodiversity metrics can be estimated at large
scales from analysis of a well chosen suite of individual sites. Others have attempted to estimate individual
site patterns from analysis of the continental pool. But does such cross-scale extrapolation work?
This issue is addressed for the North American terrestrial gastropod fauna by comparison of family
representation, species richness and body-size patterns across site to continental scales. These data
demonstrate profound differences: while the continental fauna is dominated by large body-size families
such as the Polygyridae, Helminthoglyptidae, Oreohelicidae, Succineidae and Urocoptidae, average
site faunas are most frequently represented by small body-sized families like the Vertiginidae,
Gastrodontidae, Oxychilidae, Euconulidae, Punctidae, Valloniidae, Strobilopsidae and Ellobiidae.
Species richness within sites tends to be 2–7 times smaller than random draws of individuals of the same
number from regional or continental pools, indicating the potential for strong bias in the construction
of site faunas. And, while the body-size spectrum for average site faunas is strongly right-skewed, the
continental pool is strongly left-skewed. Thus, although taxa with biovolumes .16 mm3
dominate the
continental fauna (79.4% of total), they make up only a small average fraction (4.1%) of individual
site species lists. Within most regions, site faunas are overrepresented in species with biovolumes
,4 mm3
and underrepresented in species with biovolumes .128 mm3
as compared with the regional
pool. As a result, assumptions of self-similarity between observational scales in terrestrial gastropods are
inappropriate.
INTRODUCTION
The search for universal patterns and simplifying assumptions is
common practice among natural scientists, especially those working
in complex systems such as biology. One common form taken by this
search is the identification of patterns that are scale-independent, i.e.
patterns that remain largely similar across all observational scales.
Such relationships are often referred to as expressing ‘self-similarity’,
in which large scale datasets are made up of smaller scale components
possessing the same basic properties. In the realm of
fractal geometry, for instance, self-similarity across multiple scales
has been noted in crystals (Mandelbrot, 1983), clouds (Hentschel
& Procaccia, 1984), fluids (Nittman, Daccord & Stanley, 1985)
and water drainage networks (Milne, 1988). This assumption also
underlies most common algorithms generating fractal artwork,
lending aesthetic regularity to the generated images.
However, the assumption of self-similarity has been found
to be less useful in biological systems in which the major
drivers are themselves scale dependent (Wiens, 1989). For instance,
none of the vegetation patterns analysed by Palmer
(1988) demonstrated scale independence and self-similarity.
Yet, the siren-call of self-similarity has been especially strong
in biodiversity analyses as its presence would allow insights
regarding patterns and parameters present at one scale to be
accurately applied to adjacent scales without the need for empirical
observations. Direct measurement of various diversity
metrics at tractable (usually single site) scales could thus theoretically
be used to generate expectations at regional to global
scales where empirical observation may be logistically impossible.
Finlay et al. (2006), for instance, claimed that selfsimilarity
in various species composition, richness and bodysize
metrics allows for accurate extrapolation of insect diversity
patterns at Hilbre Island and Monks Woods Nature Reserves
in England to the entire globe.
Is it really the case that biodiversity patterns appear the
same when peering through each end of a spyglass? Can observation
of only a few sites allow for accurate estimation of
regional to global scale patterns? And, is it also possible that
knowledge of regional to global scale patterns informs accurately
about relationships within individual sites? These questions
will be empirically tested through an analysis of North
American terrestrial gastropod species composition, body-size
and abundance patterns across site, regional and continental
scales.
# The Author 2014. Published by Oxford University Press on behalf of The Malacological Society of London, all rights reserved
Journal of The Malacological Society of London
Molluscan Studies
Journal of Molluscan Studies (2014) 80: 238–248. doi:10.1093/mollus/eyu028
Advance Access publication date: 2 June 2014
atUniversityofNewMexicoonSeptember9,2014http://mollus.oxfordjournals.org/Downloadedfrom
139
MATERIAL AND METHODS
Faunistic enumeration
Site scale. Species lists and abundances were considered from
1,574 sites ranging from the Alaskan North Slope and southern
California coast east to central Quebec, the Atlantic seaboard
and the Florida Keys (Table 1; Fig. 1). These were sampled
according to methods presented by Nekola (2010), with individual
sample areas being no more than 1,000 m2
. An attempt was
made to sample sites from across the entire range of habitats
within a given biogeographic region and included 416 bedrock
outcrops, 455 upland forests, 335 lowland forests, 126 upland
grasslands/shrublands and 242 lowland grasslands/shrublands.
Regional scale. Using the maps of Pilsbry (1948), Burch (1962)
and observed community composition patterns as a guide, the
continent was subdivided into 21 biogeographic regions with
boundaries being set to enclose areas of similar faunistic composition
and to also allow each species to occur within at least one
region (Fig. 1). The total faunas for each region was then determined
through the county-scale distribution maps of Hubricht
(1985) for eastern North America (using updates provided by
Nekola & Coles, 2010 and J.C. Nekola, unpubl.), with Pilsbry
(1948) and various regional lists (e.g. Metcalf & Smartt, 1997;
Roth & Sadeghian, 2003; Forsyth, 2004) being used for western
North America. The number of survey sites per region ranged from
8 (southern Appalachians) to 252 (western Niagaran Escarpment).
Continental scale. The total North American fauna was based
on Nekola (in press). Only species having native or naturalized
populations in North America north of Mexico were considered;
species only occurring south of the USA boundary are considered
to belong to the Central American fauna. The initial
point of departure for this dataset is all terrestrial gastropods
listed by Turgeon et al. (1998). To this were added all subsequently
described species as determined via the Zoological Record.
Additionally, all species listed by Hubricht (1985)—but not by
Turgeon et al. (1998)—were included, as these represent dead
shells limited to drift along the south Texas Gulf Coast that
could easily have been sourced from local extant colonies. The
list was also expanded to include all subspecies-level entities of
Pilsbry (1948) that appear, based on their unique shell features,
ranges and/or ecological preferences, to represent valid species-level
entities. Finally, 14 undescribed new species encountered by the
author during field sampling across North America have also
been included, representing six Vertigo plus single representatives
of Columella, Daedalochila, Glyphyalinia, Hawaiia, Helicodiscus,
Paravitrea, Punctum and Succinea.
Taxonomy and body size
Family assignments for all species were based on those of Bouchet
& Rocroi (2005), with the placement of genera into these families
generally following Schileyko (2006 and previous works). Generic
level assignments generally followed Turgeon et al. (1998). Shell/
body-size dimensions for each taxon were determined from either
the published literature, or in the case of undescribed taxa from
lots held in the Nekola collection. Calculation of body volume (in
mm3
) was then determined for each species using the formulae
presented by McClain & Nekola (2008) and Nekola (in press).
Statistical analyses
All analyses were conducted in the R Statistical Environment
with scripts being available upon request.
Family-level composition. The proportional representation of
each family within each site was calculated by dividing the total
number of species present within each family into the total
species richness of that site. This process was repeated across all
sites, with the mean proportion for each family then being calculated
across all sites. This process was also repeated within each
biogeographic region. The proportion of each family within the
continental fauna was calculated by dividing the number of
species within each family into the continental species richness.
Table 1. Statistical summary for the continental and regional datasets of North American terrestrial gastropods.
Region No. of sites Regional richness Site richness No. of individuals
Reported Encountered Median Maximum Total Median/site
Whole continent 1,574 1,204 460 12 39 697,778 237
Churchill 23 17 15 6 12 9,303 261
Central Manitoba 24 53 41 10 21 11,014 288
Laurentian Plateau 48 80 40 8 18 9,729 131
Northwestern Minnesota 192 63 59 15 27 142,305 463
Western Superior Uplands 82 55 44 10 18 5,025 47
Eastern Niagaran Escarpment 73 149 86 16 31 26,804 305
Western Niagaran Escarpment 252 98 84 15 34 92,778 216
New England 194 133 82 13 34 84,661 217
Upper Mississippi Valley 131 103 90 20 39 105,075 587
Ozarks 35 149 88 20 34 20,601 255
Central Appalachians 22 246 86 16 35 7002 219
Southern Appalachians 8 278 42 12 33 2,181 210
Carolinas 38 131 62 8 31 9,298 132
Gulf Coast 37 156 34 3 11 2,260 43
Peninsular Florida 59 130 97 12 28 33,848 301
Southern Plains 14 154 58 12 23 3,379 128
Southern Rockies 161 308 113 7 21 72,116 230
California 30 271 51 5 10 8,219 99
Great Basin 39 76 45 7 13 12,635 206
Pacific Northwest 48 187 64 9 17 13,895 197
Alaskan Interior 64 33 32 7 16 25,644 216
NORTH AMERICA LAND SNAIL BIODIVERSITY
239
atUniversityofNewMexicoonSeptember9,2014http://mollus.oxfordjournals.org/Downloadedfrom
140
This process was also repeated at the regional level. All proportions
were multiplied by 100 to give percentage representation.
Statistical significance of the difference between site and regional
scale representation was calculated using a sign test for all
families occurring in at least 10 regions. The number of cases in
which mean site percentage was less than the regional percentage
was counted, as was the number of cases in which regional
percentage was less than mean site percentage. Under a null expectation
of no relationship, these numbers should be equal. The
P-value for deviations from this null was calculated using the binomial
distribution. Because the test was repeated on 18 different
families, the P ¼ 0.05 significance threshold was adjusted using a
Bonferroni correction to P ¼ 0.0028.
Site species richness. The ratio between observed and random
expected richness was calculated. Expected random richness was
determined by generating a vector representing the species identities
of all encountered individuals across either the entire continental
dataset or a respective region. For each site, n random
individuals were selected from this vector without replacement,
where n ¼ the number of all identified individuals from a given
site. The number of unique species in the random draw was then
reported, with this process being repeated 1,000 times. The
mean number of unique species across all random samples was
then recorded. This process was repeated across all sites.
Observed site richness was then divided into mean random richness.
A box plot was then generated showing the variation in
observed: random richness across all sites in the continent or
respective region. Statistical significance for differences in
observed vs randomized richness was estimated using the paired
Wilcoxon rank-sum test.
Body-size spectra. Observed body-size spectra were determined
across all sites by counting the number of species falling within
each of 41 log2 volume classes with bins starting at 22.5 and increasing
by 0.5 log2 units to 18.0. The first bin value ranged from
22.5 to 22.0 ( 0.18 to 0.25 mm3
) and the last bin from 17.5–
18.0 (185 363.8 to 262 144 mm3
). The proportion of the total site
fauna found within each bin was then determined by dividing the
observed bin richness by total site richness. Random expectations
were then generated by randomly drawing n species, without replacement,
from the continental or respective regional species
pool, where n ¼ number of recorded species from that site. The
proportion that each bin represented in the random draw was
then calculated. This procedure was repeated 1,000 times, with
the mean proportion across all randomizations being reported
across all 41 bins. This process was repeated for each site. From
this, body-size spectra graphs were generated showing across all
bins the mean observed site representation and the minimum and
maximum mean random values as drawn from the continental or
respective regional pool. Regional scale patterns were illustrated
through three latitudinal transects: western (Alaskan Interior,
Pacific Northwest, Great Basin and southern Rockies), central
(Churchill, Northwestern Minnesota, Upper Mississippi Valley
and Ozarks), and eastern (Laurentian Plateau, New England,
Carolinas, Peninsular Florida).
Figure 1. Location of the 1,574 sample sites and 21 biogeographic regions in North America. Regional codes are: A, Churchill; B, Central Manitoba;
C, Laurentian Plateau; D, Northwestern Minnesota; E, Lake Superior Uplands; F, Western Niagaran Escarpment; G, Eastern Niagaran Escarpment;
H, New England; I, Upper Mississippi Valley; J, Ozarks; K, Central Appalachians; L, Southern Appalachians; M, Carolinas; N, Gulf Coast; O,
Peninsular Florida; P, Southern Plains; Q, Southern Rockies; R, California; S, Great Basin; T, Pacific Northwest; U, Alaskan Interior.
J. C. NEKOLA
240
atUniversityofNewMexicoonSeptember9,2014http://mollus.oxfordjournals.org/Downloadedfrom
141
RESULTS
In all 1,204 species-level taxa are reported from North America,
partitioned among 170 genera and 51 families (Nekola, in press).
A total of 460 species (38% of the continental total) and 697,778
individuals were recorded from the 1,574 sites (Table 1).
Regional faunistic richness ranged from 17 (Churchill) to 308
(Southern Rockies), with encountered species ranging from 15
(Churchill) to 113 (Southern Rockies). Median site richness
ranged from 3 (Gulf Coast) to 20 (Upper Mississippi Valley and
Ozarks), while maximum richness ranged from 10 (California) to
39 (Upper Mississippi Valley). The total number of encountered
individuals ranged from 2,181 (Southern Appalachians) to
142,305 (Northwestern Minnesota), while the median number of
individuals encountered per site ranged from 43 (Gulf Coast) to
587 (Upper Mississippi Valley).
Family-level composition
For the entire continental fauna (Table 2), the ten most frequently
represented families were the Polygyridae (22.51%),
Helminthoglyptidae (14.78%), Vertiginidae (8.64%), Oreohelicidae
(6.98%), Pristilomatidae (5.40%), Oxychilidae (4.07%),
Succineidae (4.07%), Gastrodontidae (2.91%), Arionidae (2.82%)
and Urocoptidae (2.58%). This differs greatly from average
site faunas, which were most frequently represented by the
Vertiginidae (29.93%), Gastrodontidae (11.19%), Oxychilidae
(7.60%), Euconulidae (7.09%), Punctidae (5.44%), Discidae
(5.40%), Valloniidae (4.88%), Strobilopsidae (3.23%), Polygyridae
(3.20%) and Ellobiidae (2.97%).
Regional species pools demonstrated variability in familylevel
representation, with the most common family ranging from
the Vertiginidae (Churchill, Central Manitoba, Lake Superior
Uplands, Laurentian Plateau, New England, Northwestern
Minnesota, Eastern Niagaran Escarpment, Western Niagaran
Escarpment, Upper Mississippi Valley and Alaskan Interior) to the
Polygyridae (Southern Plains, Ozarks, Southern Appalachians,
Central Appalachians, Carolinas, Gulf Coast and Peninsular
Florida), Oreohelicidae (Pacific Northwest and Great Basin) and
Helminthoglyptidae (California and Southern Rockies).
The sign test demonstrates that five families show significant
(and another three marginally significant) differences in their
representation between site and regional scales (Table 3). The
Euconulidae, Punctidae and Vertiginidae all exhibited significantly
(Bonferroni adjusted P , 0.0028) more instances of
mean site fauna proportion exceeding the respective regional
species pool proportion than would be expected at random.
The Gastrodontidae, Oxychilidae and Strobilopsidae exhibited
this pattern to only a marginal (0.05 , P , 0.0028) degree.
However, the Polygyridae and Succineidae demonstrated
significantly more instances of mean site fauna proportion
being less than the respective regional pool proportion. Ten
families (Agriolimacidae, Cochlicopidae, Discidae, Ellobiidae,
Haplotrematidae, Helicodiscidae, Pristilomatidae, Pupillidae,
Valloniidae and Vitrinidae) exhibited similar proportional representations
between average site faunas and the regional species
pool. Because of limited occurrence (,10 regions occupied), statistical
power was too low to conduct this test on 25 families.
However, among these it should be noted that in all four regions
supporting members of the small body-size Thysanophoridae,
percent representation within site faunas always exceeded the
proportion in respective regional species pools, while in the large
body-size Helicidae, Helminthoglyptidae, Humboldtianidae,
Hygromiidae, Monadeniidae and Oreohelicidae site fauna representation
was always less than the respective regional pool. This
was also the case in 75% of the regions supporting large body-size
Orthalicidae.
Site richness
Observed site richness values were approximately seven times
smaller than random draws taken from the entire continental
dataset (Fig. 2). The paired Wilcoxon rank-sum test demonstrated
that this difference was highly significant (P , 0.0000001).
At the regional scale, observed richness generally ranged from 1.85
times smaller than random draws (Lake Superior Uplands) to 6.3
times smaller (Southern Rockies). Observed site faunas were 2.4
and 3.6 times less rich than average random draws from the regional
individual pool for Churchill and the Alaskan Interior, respectively.
This number ranged from 2.7 to 3.1 for Central
Manitoba, the Laurentian Plateau, New England, Northwestern
Minnesota, the Eastern and Western Niagaran Escarpments and
the Upper Mississippi Valley. In the Southern Plains, Ozarks and
Central Appalachians, individual site faunas were 2.6–2.2 times
smaller than random draws from their respective regional individual
pools. This number ranged from 3.4–3.9 in the Carolinas,
Gulf Coast and Peninsular Florida, and 3.1–4.7 in the Pacific
Northwest, Great Basin and California. The paired Wilcoxon
rank-sum test demonstrated that these differences were all highly
significant (P , 0.0000027 with a Bonferroni-adjusted P-value
threshold of 0.00238). The main exception was the Southern
Appalachians, which had observed richness ranging between
c. 1.3 times larger to 4.2 times smaller than the size of a random
draw from the regional individual pool (median ¼ 1.5 times
smaller). While sites were generally less rich than random draws
from the entire regional individual pool, this difference was only
marginally significant (P ¼ 0.007813).
Body-size spectra
Observed mean body-size spectra within sites differed strongly
from random draws made from the continental species pool
(Fig. 3). While observed site faunas were strongly right-skewed
with a mode at the 5th log2 body-size bin (0.71–1 mm3
),
random draws of the same richness from the continental species
pool were strongly left-skewed, with modal values ranging from
the 21st–24th log2 body-size bin (181–512 mm3
). Site faunas
were greatly enriched in species with biovolumes of 5.7 mm3
or
less, and greatly underrepresented in species with biovolumes of
90.5 mm3
or greater. These differences are highly significant,
with the site-scale mode for species with biovolumes ,5.7 mm3
being more than three times larger than the maximum observed
randomized mean at that same size class, and the minimum
score of the randomized means for species with biovolumes
.90.5 mm3
being at least five times larger than observed scores
for these same body-size classes. Species with biovolumes of 5.7–
90.5 mm3
were represented at approximately equal proportions
within observed and randomized site faunas.
The enrichment of small species and impoverishment of large
species at site scales was also present within most biogeographic
regions (Fig. 4). Among the 12 illustrated relationships in
Figure 4, the Pacific Northwest, Great Basin, Southern Rockies,
Upper Mississippi Valley, Ozarks, Laurentian Plateau, New
England and Carolinas all strongly demonstrate this pattern. It
is also more weakly exhibited in Northwestern Minnesota and
Peninsular Florida. Very similar observed vs randomized bodysize
spectra were noted for the Alaskan Interior and Churchill.
DISCUSSION
These analyses clearly show that family composition, body-size
spectra and species richness are not self-similar and scale independent
among North American terrestrial gastropods. While
the continental fauna is dominated by families with large-sized
individuals like the Polygyridae and Helminthoglyptidae,
NORTH AMERICA LAND SNAIL BIODIVERSITY
241
atUniversityofNewMexicoonSeptember9,2014http://mollus.oxfordjournals.org/Downloadedfrom
142
Table2.ComparisonofterrestrialgastropodfamilypercentrepresentationacrossobservationalscalesinNorthAmerica.
FamilyEntireContinentChurchillCentralManitobaLakeSuperior
Uplands
LaurentianPlateauNewEnglandNWMinnesotaENiagaraWNiagaraUpperMississippiSouthernPlainsOzarks
R%S%R%S%R%S%R%S%R%S%R%S%R%S%R%S%R%S%R%S%R%S%R%S%
Polygyridae22.513.205.460.3412.780.501.590.0312.083.988.162.7810.686.4832.4718.5429.5318.40
Helminthoglyptidae14.780.28
Vertiginidae8.6429.9364.7168.1135.8531.5627.2720.9626.2536.9315.7930.9331.7524.4615.4521.8620.4128.0621.3631.5812.9926.5515.4423.49
Oreohelicidae6.980.37
Pristilomatidae5.402.661.89–1.820.612.260.473.183.062.692.693.062.661.944.562.608.506.716.40
Oxychilidae4.077.603.7712.057.2711.785.0014.269.779.173.189.537.389.008.165.114.855.124.555.736.719.11
Succineidae4.072.7511.7711.1713.213.345.461.057.501.716.021.967.944.326.711.639.183.6210.682.247.149.006.710.52
Gastrodontidae2.9111.195.668.029.0921.716.257.537.5219.706.3511.927.3816.636.1212.615.835.663.255.815.379.75
Arionidae2.82–3.64–11.25–6.02–3.18–6.04–4.08–3.88–
Urocoptidae2.580.181.95–
Discidae2.245.405.88–3.7713.085.4612.703.758.662.263.734.766.704.037.904.087.184.855.141.300.993.362.36
Helicodiscidae2.082.763.77–3.643.741.250.983.013.543.181.762.013.933.064.483.885.955.205.772.694.68
Phylomicidae1.74–5.26–4.70–1.02–2.91–1.95–4.03–
Haplotrematidae1.330.440.750.090.670.291.020.070.971.130.671.70
Orthalicidae1.330.232.601.130.671.35
Monadeniidae1.330.03
Pupillidae1.161.885.882.711.89–1.250.281.500.413.180.081.340.922.040.051.941.770.653.730.672.50
Valloniidae1.084.889.434.477.276.578.756.714.514.669.526.634.703.477.143.735.835.083.901.562.691.19
Euconulidae0.917.095.8813.783.7710.255.465.332.5012.793.017.253.187.232.695.924.086.333.885.411.953.052.694.40
Ellobiidae0.912.973.773.733.642.122.50–1.503.683.186.332.014.452.045.081.943.981.300.480.672.21
Punctidae0.835.443.776.991.824.942.502.741.507.244.766.392.016.063.065.732.915.130.652.441.344.94
Humboldtianidae0.830.00
Helicinidae0.750.631.820.151.021.650.971.733.251.130.671.08
Binneyidae0.75–
Subulinidae0.660.280.65–
Helicidae0.660.053.750.143.760.034.030.591.02–2.60–1.34–
Hygromiidae0.660.023.75–1.500.142.690.06
Megomphicidae0.660.01
Veronicellidae0.66–
Agriolimacidae0.501.305.884.241.891.961.820.103.750.151.500.963.181.692.012.562.041.731.943.270.65–1.340.45
Strobilopsidae0.423.231.891.101.824.601.25–3.014.173.184.162.013.973.064.892.914.242.604.452.694.44
Thysanophoridae0.420.971.30–
Truncatellidae0.420.130.65–
Limacidae0.42–2.50–3.01–3.36–1.02–1.30–2.01–
Spiraxidae0.330.132.600.51
Pleurodontidae0.330.06
Cochlicopidae0.251.993.772.375.462.542.501.772.261.243.184.052.012.843.063.672.911.251.430.22
Vitrinidae0.171.431.891.101.820.761.255.360.750.151.591.650.671.231.020.420.970.11
Pomatiidae0.170.12
Pomatiopsidae0.170.070.75–0.670.051.020.141.940.180.670.83
Ferrussaciidae0.170.00
Charopidae0.170.09
Sagdidae0.080.07
Oleacinidae0.080.07
Bradybaenidae0.080.03
Streptaxidae0.080.02
Helicarionidae0.080.01
Cepolidae0.080.00
Cochlicellidae0.08–
Milacidae0.08–1.25–
Testacellidae0.08–1.25–0.67–
J. C. NEKOLA
242
atUniversityofNewMexicoonSeptember9,2014http://mollus.oxfordjournals.org/Downloadedfrom
143
FamilySouthern
Appalachians
Central
Appalachians
CarolinasGulfCoastPeninsularFloridaAlaskanInteriorPacificNorthwestGreatBasinCaliforniaSouthernRockies
R%S%R%S%R%S%R%S%R%S%R%S%R%S%R%S%R%S%R%S%
Polygyridae30.2212.8023.1711.0725.194.1630.138.3114.6212.709.634.348.862.9217.861.93
Helminthoglyptidae1.07–38.7512.1224.350.51
Vertiginidae6.1217.739.7623.4914.5036.8312.8241.3613.0817.5345.4659.0912.3025.6619.7419.837.7522.2312.6634.17
Oreohelicidae14.971.1028.957.470.37–12.341.46
Pristilomatidae12.9514.397.327.341.531.833.210.521.543.275.885.141.32–2.955.980.334.13
Oxychilidae12.5913.8210.9811.359.1618.208.9714.204.624.816.063.313.216.433.951.392.210.901.625.75
Succineidae1.80–4.891.227.63–8.330.396.931.309.099.946.422.0313.163.593.691.942.270.57
Gastrodontidae8.2717.9210.1610.876.119.997.6917.303.082.541.609.982.632.480.7411.310.657.21
Arionidae2.85–2.29–14.44–7.75–0.33–
Urocoptidae2.313.478.440.50
Discidae5.400.382.851.412.290.361.28–6.063.553.216.522.636.941.482.640.654.13
Helicodiscidae5.04–4.473.552.292.982.560.343.081.540.540.230.370.330.651.12
Phylomicidae4.68–4.89–3.05–3.21–1.54–0.33–
Haplotrematidae0.721.910.411.940.761.400.64–1.603.044.065.56
Orthalicidae0.36–1.28–5.393.831.620.49
Monadeniidae1.070.195.901.11
Pupillidae0.36–0.810.380.760.330.64–1.541.973.035.771.070.846.5814.950.370.673.907.31
Valloniidae1.08–2.030.642.290.641.281.681.544.316.061.342.144.176.5811.820.742.392.278.87
Euconulidae1.806.832.036.733.055.363.216.953.855.429.0914.271.078.632.636.680.744.200.656.95
Ellobiidae2.164.242.444.652.291.081.280.250.770.571.600.830.370.330.330.04
Punctidae1.448.461.637.773.056.141.282.430.770.246.060.881.6011.541.324.972.2120.870.972.87
Humboldtianidae3.250.04
Helicinidae0.72–0.410.520.64–2.314.77
Binneyidae4.28–0.37–
Subulinidae0.41–1.530.291.28–6.157.170.37–0.33–
Helicidae1.630.441.53–2.56–1.54–1.60–1.32–2.580.480.65–
Hygromiidae2.29–0.64–0.77–0.37–
Megomphicidae1.600.351.85–
Veronicellidae6.15–
Agriolimacidae0.72–0.811.311.530.910.640.340.770.133.031.342.670.312.631.091.111.220.970.31
Strobilopsidae1.08–1.633.192.299.092.565.612.312.86
Thysanophoridae1.544.311.076.191.324.680.654.78
Truncatellidae2.313.400.74–
Limacidae0.72–1.63–1.53–1.92–2.14–1.32–1.48–0.65–
Spiraxidae0.76–0.640.340.773.04
Pleurodontidae3.081.62
Cochlicopidae1.080.950.811.280.760.423.030.131.070.121.320.200.37–0.332.41
Vitrinidae0.41–3.030.390.541.151.3213.650.372.810.333.99
Pomatiidae1.543.34
Pomatiopsidae0.72–0.810.850.76–0.64–
Ferrussaciidae0.41–1.540.07
Charopidae0.541.201.320.260.330.45
Sagdidae0.771.77
Oleacinidae0.771.93
Bradybaenidae0.770.69
Streptaxidae0.64–0.770.62
Helicarionidae0.770.33
Cepolidae0.770.07
Cochlicellidae0.76–0.37–
Milacidae0.41–0.54–0.37–0.33–
Testacellidae0.54–0.37–
R%,thepercentageofspeciesinthecontinental/regionalfaunaoccurringwithinthatfamily.S%,thepercentagerepresentationofthatfamilywithinanaveragesitewithinthatregion.Dashesindicatefamiliesthatarepresentintherespectivefaunabuthavenotyetbeenencounteredwithinsites.
NORTH AMERICA LAND SNAIL BIODIVERSITY
243
atUniversityofNewMexicoonSeptember9,2014http://mollus.oxfordjournals.org/Downloadedfrom
144
typical site faunas are dominated by families with small-sized
individuals like the Vertiginidae and Gastrodontidae. Likewise,
while the continental fauna demonstrates a left-skewed body-size
spectrum dominated by species with biovolumes .90.5 mm3
,
average site faunas are right-skewed and dominated by species
with biovolumes ,5.7 mm3
. Random draws representing the
same number of individuals from the entire continental dataset
contained on average at least seven times more species than the
actual number observed at that site.
These discrepancies were generally also observed between
sites and their respective biogeographic regions. Typically, families
with small-sized individuals (Euconulidae, Gastrodontidae,
Oxychilidae, Punctidae, Strobilopsidae, Thysanophoridae and
Vertiginidae) were overrepresented in typical site faunas as
compared to their respective regional species pool, while families
with large-sized individuals (Helminthoglypidae, Monadeniidae,
Oreohelicidae, Polygyridae and Succineidae) were underrepresented.
This bias is demonstrated across most regions (e.g.
Pacific Northwest, Great Basin, Southern Rockies, New
England and Carolinas). However, both arctic regions tended to
show little discrepancy between typical site body-size spectra
and the regional pool. Site richness also generally ranged from
two to six times smaller than would be expected from random
draws from the respective regional individual pool. Even in the
case of the Southern Appalachians, where the eight sampled
sites were only collected across a 20-km extent and only 42 total
species were encountered (15% of the entire regional fauna) the
majority of sites were still found to have fewer species than
would be expected from a random draw of all encountered
individuals. As more samples are made across a larger extent
within this region and more of this regional fauna is encountered,
the distribution of observed to random ratio scores will be
lowered and made to fall in line with other regions.
These analyses indicate that site faunas are not a simple
random sample drawn from the continental or regional pool.
Rather, important biases are present. The most important of
these are not only the predictable filtering of species along environmental
gradients (principally soil architecture, moisture and
acidity; Nekola, 2003, 2010) but also purely geographic filtering
caused by dispersal limitation and other forms of spatial constraint
(Nekola & White, 1999; Hubbell, 2001).
As a result, it is not possible to extrapolate accurately these
macroecological metrics across scales for North American terrestrial
gastropods. Rather, the fauna looks considerably different
depending upon whether one is looking down from continental
scales, or up from site scales. From the continental-scale perspective,
the fauna is generally dominated by families with largesized
individuals. However, from the site scale the fauna is dominated
by a series of families with small body-sized individuals.
Documenting the faunas of a few individual sites will thus not
allow for an accurate portrayal of the continental fauna and
knowledge of the continental fauna informs little about what
one would expect within individual sites.
The lack of self-similarity within the North American terrestrial
gastropod fauna appears to be principally caused by a
strong inverse correlation between body and range size. In this
fauna, many small species (especially within the Euconulidae,
Gastrodotidae, Oxychilidae, Punctidae, Strobilopsidae and
Table 3. Comparison of family representation at regional vs site scales for terrestrial gastropods in North America.
A. 10 or more regions B. ,10 regions
Family R . S S . R P-value Family R . S S . R
Agriolimacidae 14 5 0.063568 Bradybaenidae 1 0
Cochlicopidae 11 5 0.210114 Cepolidae 1 0
Discidae 6 12 0.237885 Charopidae 1 2
Ellobiidae 11 8 1.0 Ferrussaciidae 1 0
Euconulidae 1 20 0.000021 Helicarionidae 1 0
Gastrodontidae 3 16 0.004425 Helicidae 5 0
Haplotrematidae 3 7 0.343750 Helicinidae 3 5
Helicodiscidae 7 9 0.803619 Helminthoglyptidae 2 0
Oxychilidae 5 15 0.041389 Humboldtianidae 1 0
Polygyridae 16 0 0.000031 Hygromiidae 2 0
Pristilomatidae 7 9 0.803619 Megomphicidae 1 0
Punctidae 2 18 0.000402 Monadeniidae 2 0
Pupillidae 10 7 0.629059 Oleacinidae 0 1
Strobilopsidae 1 12 0.003418 Oreohelicidae 3 0
Succineidae 17 2 0.000729 Orthalicidae 3 1
Valloniidae 12 7 0.359283 Pleurodontidae 1 0
Vertiginidae 3 18 0.001490 Pomatiidae 0 1
Vitrinidae 6 7 1.0 Pomatiopsidae 3 2
Sagdidae 0 1
Spiraxidae 2 1
Streptaxidae 1 0
Subulinidae 2 1
Thysanophoridae 0 4
Truncatellidae 0 1
Urocoptidae 1 1
R . S represents the number of regions in which family representation in the regional pool exceeded the representation on an average site. S . R represents the
number of regions in which family representation within an average site exceeded the representation in the regional pool. Assuming the null of R . S ¼ S . R,
P-values for families present in at least ten regions are based on the bionomial sign test. Families with P , 0.0028 have been highlighted in bold font; families with
0.05 , P , 0.0028 have been highlighted in italic font.
J. C. NEKOLA
244
atUniversityofNewMexicoonSeptember9,2014http://mollus.oxfordjournals.org/Downloadedfrom
145
Vertiginidae) tend to have component species harboring extensive
ranges. Examples include Euconulus fulvus (Mu¨ ller,
1774), E. alderi Gray, 1840, Punctum minutissumum (I. Lea, 1841)
and Vertigo arthuri von Martens, 1884, which range across the
entire continent north of 388N, while also extending south along
the crest of the Rockies to the Mexican border. Gastrocopta pellucida
(Pfeiffer, 1841), Glyphyalinia umbilicata (Singley, in Cockerell,
1893), Hawaiia miniscula (A. Binney, 1840) and Straitura meridionalis
(Pilsbry & Ferriss, 1906) likewise extend across much of
North America south of 358N. And, both Strobilops labyrinthica (Say,
1817) and Strobilops aenea Pilsbry, 1926 range across most of the
southeastern quarter of the continent. A number of genera with
large body-sized species, however, demonstrate significant levels of
allopatric replacement and local endemism. These are especially
pronounced within the Polygyridae (e.g. Ashmunella, Daedalochila,
Stenotrema, Inflectarius, Patera, Triodopsis, Trilobopsis and Vespericola),
Helminthoglyptidae (e.g. Helminthoglypta, Micrarionta and Sonorella),
Oreohelicidae (e.g. Oreohelix and Radiocentrum), Urocoptidae (e.g.
Holospira and Coelostemma), Monadeniidae (Monadena) and
Humboldtianidae (Humboldtiana), with it being uncommon for
species within these genera to be sympatric at the site scale.
It seems likely that the combined effects of body size on
passive dispersal and uniparental reproduction rates help
explain this counter-intuitive pattern. Because terrestrial gastropods
are among the poorest active dispersers known among terrestrial
animals, with individuals moving perhaps no more than
1–10 m over their lifetime (Schilthuizen & Lombaerts, 1994)
and populations being unable actively to cross barriers of only
100–1,000 m (Baur, 1988; Schilthuizen & Lombaerts, 1994),
Figure 2. Box plot showing distribution of observed site richness vs randomized expectations from the regional/continental individual pool of terrestrial
gastropods in North America. The site vs continental-scale comparison is identified by grey fill.
Figure 3. Site vs continental body-size spectra for terrestrial gastropods
of North America as represented by 41 log2 body-size classes. The solid
line represents the observed average for individual sites. The upper and
lower dashed lines represent the maximum and minimum mean scores
(respectively) observed for that size class across 1,000 random samples
drawn per site from the continental pool. The range between these
extremes has been filled with grey.
NORTH AMERICA LAND SNAIL BIODIVERSITY
245
atUniversityofNewMexicoonSeptember9,2014http://mollus.oxfordjournals.org/Downloadedfrom
146
passive dispersal takes precedence in determining population
movement and species range size. Small snails/slugs are much
more easily moved via passive vectors than larger ones, because
small-sized individuals are less likely to be pulled off by gravity
or fluid mechanics during movement. Additionally, a number of
North American genera with species of small body-size contain
some that are capable of uniparental reproduction (Pokryszko,
1987; Bulman, 1990). For these, movement of only a single
unmated individual is required to found a new population,
greatly increasing the effectiveness of long-range dispersal events
in causing range expansion. DNA sequence data illustrate, for
instance, that members of the genus Balea have been repeatedly
carried across 9,000 km of open sea in the eastern Atlantic
Ocean by migrating birds (Gittenberger et al., 2006). Species
within the Vertigo gouldii group in North America may possess
ranges exceeding 5,000 km in extent even though their modern
Figure 4. Site vs regional body-size spectra of terrestrial gastropods in North America, as represented by 41 log2 body-size classes across twelve selected
biogeographic regions. Panels are arranged in three columns (western, central and eastern North America) from most northern (top) to most southern
(bottom). In each panel, the solid line represents the observed average for individual sites in that region. The upper and lower dashed lines represent
the maximum and minimum mean scores (respectively) observed for that size class across 1,000 random samples drawn per site from the respective
regional pool. The range between these extremes has been filled with grey.
J. C. NEKOLA
246
atUniversityofNewMexicoonSeptember9,2014http://mollus.oxfordjournals.org/Downloadedfrom
147
ranges were covered by continental ice as recently as 12,000 yr
BP (Nekola, Coles & Bergthorsson, 2009). This is not a universal
pattern, however, with many Pristilomatidae and Helicodiscidae
exhibiting local endemism in the eastern USA (e.g. Helicodiscus,
Paravitrea, Pilsbryna and Polygyriscus). Comparison of the breeding
systems and ecology between large-range vs small-range small
body-sized species will undoubtedly document important differences.
For instance, many locally endemic Helicodiscus, Paravitrea
and Polygyriscus species tend to be restricted to highly insular rock
talus habitats, perhaps making it easier for their populations to
become isolated and speciate.
Because similar analyses have not yet been performed in other
regions, it cannot be conclusively stated whether strong scaledependence
is a general phenomenon or is unique to North
America terrestrial gastropods. Because a strong inverse correlation
between body and range size (Pokryszko & Cameron,
2005; Cameron, Pokryszko & Horsa´ k, 2010, 2012) in combination
with an overabundance of small individuals and large
species (Nekola et al., 2013) exists in the European fauna, it
seems likely that scale-dependence also occurs there.
Preliminary investigations of the New Zealand fauna (Nekola
et al., 2013), however, show apparent strong coupling in species
richness and individual abundance across the body-size spectrum.
Given that this fauna, as well as that of eastern Australia
(Stanisic et al., 2007, 2010), also harbours many narrow-range
endemics of small body size (especially in the Charopidae and
Punctidae), self-similarity within these faunas is possible. Yet,
the commonness in western Australia of large body-size camaenids
demonstrating strong alloptaric replacement with smaller
species demonstrating extensive ranges (Solem, 1988; Cameron,
1992) suggests that at least parts of this continental fauna may
behave in a fashion similar to that in North America.
Lastly, it is important to point out that at least some of the
‘self-similar’ patterns reported by Finlay et al. (2006) are almost
certainly not rooted in ecological mechanisms. For instance,
S-shaped species-abundance distributions along a rankabundance
axis and power-law species-area relationships have
both been shown to be common expectations across a wide
variety of complex systems spanning not only the physical, biological
and social sciences but also the arts (Nekola & Brown,
2007). As a result, the similarity of these patterns across scales
only means that these systems have remained complex across
these scales—and no more. Given that complex systems typically
demonstrate nesting, with a given system being made up of complex
parts, while at the same time being a component of even
larger-scale complex systems (Brown, 1994; West, 2006), these
findings are to be expected and are likely mathematically trivial.
Because of scale-dependence, no easy mathematical short-cut
exists for documenting the North American terrestrial gastropod
fauna. It is not only impossible to look at a few sites and be able
accurately to assemble continental or most regional scale biodiversity
patterns, but it is also equally impossible to deduce the
makeup of site faunas given knowledge of the continental or regional
pool. This fauna can only be accurately known by looking
at many sites spread across all biogeographic regions and then
summarizing these data to allow for expression of emergent properties.
The only way this can be accurately accomplished is thus
through extensive, time consuming and expensive fieldwork.
ACKNOWLEDGEMENTS
The generation of the continental and region-scale datasets was
made possible by the invaluable help of Gabe Martinez, Sarah
Martinez, and Khayman Mondragon of the Albuquerque
Institute for Math and Science. The generation of site-scale data
from across North America is due to the generous support of a
large number of organizations and people. While most have been
mentioned in previous papers, I would especially like to thank for
recent aid Ken Hotopp of Appalachian Conservation Biology for
fieldwork in the middle Appalachians, the Bailey-Matthew Shell
Museum for work in Peninsular Florida, the Valhalla Wilderness
Society for fieldwork in southeastern British Columbia and adjacent
areas in Idaho, Charles Drost of the USGS for fieldwork in
the Channel Islands and along the California coast. I would also
like especially to thank Michal Horsa´ k and Brian Coles for
helping keep the lights on in my lab though their generous personal
donations. This work would simply have been impossible
without their assistance.
REFERENCES
BAUR, B. 1988. Microgeographical variation in shell size of the land
snail Chondrina clienta. Biological Journal of the Linnean Society, 35:
247–259.
BOUCHET, P. & ROCROI, J.-P. 2005. Classification and nomenclator
of gastropod families. Malacologia, 47: 1–397.
BROWN, J.H. 1994. Complex ecological systems. In: Complexity:
metaphors, models, and reality (G.A. Cowan, D. Pines & D. Melzer,
eds), pp. 419–449. Santa Fe Institute Studies in the Science of
Complexity, Proceedings Volume XVIII. Addison-Wesley, Reading.
BULMAN, K. 1990. Life history of Carychium tridentatum (Risso, 1826)
(Gastropoda: Pulmonata: Ellobiidae) in the laboratory. Journal of
Conchology, 33: 321–333.
BURCH, J.B. 1962. The eastern land snails. W. C. Brown Co., Dubuque,
IA.
CAMERON, R.A.D. 1992. Land snail faunas of the Napier and Oscar
Ranges, Western Australia: diversity, distribution and speciation.
Biological Journal of the Linnean Society, 45: 271–286.
CAMERON, R.A.D., POKRYSZKO, B.M. & HORSA´ K, M. 2010.
Land snail faunas in Polish forests: patterns of richness and
composition in a post-glacial landscape. Malacologia, 53: 77–134.
CAMERON, R.A.D., POKRYSZKO, B.M. & HORSA´ K, M. 2012.
Forest snail faunas from Crimea (Ukraine), an isolated and
incomplete Pleistocene refugium. Biological Journal of the Linnean
Society, 109: 424–433.
FINLAY, B.L., THOMAS, J.A., MCGAVIN, G.C., FENCHEL, T. &
CLARKE, R.T. 2006. Self-similar patterns of nature: insect diversity at
local to global scales. Proceedings of the Royal Society B, 273: 1935–1941.
FORSYTH, R.G. 2004. Land snails of British Columbia. Royal British
Columbia Museum handbook. Royal British Columbia Museum,
Victoria.
GITTENBERGER, E., GROENENBERG, D.S.J., KOKSHOORN,
B. & PREECE, R.C. 2006. Molecular trails from hitch-hiking snails.
Nature, 439: 409.
HENTSCHEL, H.G.E. & PROCACCIA, I. 1984. Relative diffusion in
turbulent media: the fractal dimension of clouds. Physics Review A, 29:
1461–1470.
HUBBELL, S.P. 2001. The unified neutral theory of biodiversity and
biogeography. Monographs in Population Biology 32. Princeton
University Press, Princeton, NJ.
HUBRICHT, L. 1985. The distributions of the native land mollusks of
the eastern United States. Fieldiana, New Series, 24: 1–191.
MCCLAIN, C. & NEKOLA, J.C. 2008. The role of local-scale on
terrestrial and deep-sea gastropod body size distributions across
multiple scales. Evolutionary Ecology Research, 10: 129–146.
MANDELBROT, B. 1983. The fractal geometry of nature. W.H. Freeman,
San Francisco.
METCALF, A.L. & SMARTT, R.A. 1997. Land snails of New Mexico.
New Mexico Museum of Natural History and Science, Albuquerque,
New Mexico.
MILNE, B.T. 1988. Measuring the fractal geometry of landscapes.
Applied Mathematics and Computation, 27: 67–79.
NEKOLA, J.C. 2003. Large-scale terrestrial gastropod community
composition patterns in the Great Lakes region of North America.
Diversity and Distribution, 9: 55–71.
NEKOLA, J.C. 2010. Acidophilic terrestrial gastropod communities of
North America. Journal of Molluscan Studies, 76: 144–156.
NORTH AMERICA LAND SNAIL BIODIVERSITY
247
atUniversityofNewMexicoonSeptember9,2014http://mollus.oxfordjournals.org/Downloadedfrom
148
NEKOLA, J.C. in press. Overview of the North American terrestrial
gastropod fauna. American Malacological Bulletin.
NEKOLA, J.C., BARKER, G.M., CAMERON, R.A.D. &
POKRYSZKO, B.M. 2013. Latitudinal variation of body size in
land snail populations and communities. In: Global patterns of body size
(F. Smith & K. Lyons, eds), pp. 62–82. University of Chicago Press,
Chicago.
NEKOLA, J.C. & BROWN, J.H. 2007. The wealth of species:
ecological communities, complex systems, and the legacy of Frank
Preston. Ecology Letters, 10: 188–196.
NEKOLA, J.C. & COLES, B.F. 2010. Pupillid land snails of eastern
North America. American Malacological Bulletin, 28: 29–57.
NEKOLA, J.C., COLES, B.F. & BERGTHORSSON, B. 2009.
Evolutionary pattern and process in the Vertigo gouldii (Mollusca:
Pulmonata, Pupillidae) group of minute North American land
snails. Molecular Phylogenetics and Evolution, 53: 1010–1024.
NEKOLA, J.C. & WHITE, P.S. 1999. Distance decay of similarity in
biogeography and ecology. Journal of Biogeography, 26: 867–878.
NITTMAN, J., DACCORD, G. & STANLEY, H.E. 1985. Fractal
growth of viscous fingers: quantitative characterization of fluid
instability phenomenon. Nature, 314: 141–144.
PALMER, M.W. 1988. Fractal geometry: a tool for describing spatial
patterns in plant communities. Vegetatio, 75: 91–102.
PILSBRY, H.A. 1948. Land Mollusca of North America (north of Mexico).
Academy of Natural Sciences of Philadelphia Monograph, 1(1)–2(2).
POKRYSZKO, B.M. 1987. On aphally in the Vertiginidae (Gastrocopta:
Pulmonata: Orthurethra). Journal of Conchology, 32: 365–375.
POKRYSZKO, B.M. & CAMERON, R.A.D. 2005. Geographical
variation in the composition and richness of forest snail faunas in
northern Europe. Records of the Western Australian Museum, Supplement,
68: 115–132.
ROTH, B. & SADEGHIAN, P.S. 2003. Checklist of the land snails and
slugs of California. Santa Barbara Museum of Natural History
Contributions in Science, 3: 1–81.
SCHILEYKO, A.A. 2006. Treatise on recent terrestrial pulmonate molluscs,
Part 13. Helicidae, Pleurodontidae, Polygyridae, Ammonitellidae,
Oreohelicidae, Thysanophoridae. Ruthenica, Suppl. 2: 1765–1906.
SCHILTHUIZEN, M. & LOMBAERTS, M. 1994. Population
structure and levels of gene flow in the Mediterranean land snail
Albinaria corrugata (Pulmonata: Clausiliidae). Evolution, 48: 577–586.
SOLEM, A. 1988. Maximum in the minimum: biogeography of land
snail from the Ningbing Ranges and Jeremiah Hills, north-east
Kimberly, Western Australia. Journal of the Malacological Society of
Australia, 9: 59–113.
STANISIC, J., CAMERON, R.A.D., POKRYSZKO, B.M. &
NEKOLA, J.C. 2007. Forest snail faunas from S.E. Queensland and
N.E. New South Wales (Australia): patterns of local and regional
richness and differentiation. Malacologia, 49: 445–462.
STANISIC, J., SHEA, M., POTTER, D. & GRIFFITHS, O. 2010.
Australian land snails Volume 1: a field guide to eastern Australian species.
Queensland Museum, Brisbane.
TURGEON, D.D., QUINN, J.F., BOGAN, A.E., COAN, E.V.,
HOCHBERG, F.G., LYONS, W.G., MIKKELSEN, P., NEVES, R.J.,
ROPER, C.F.E., ROSENBERG, G., ROTH, B., SCHELTEMA, A.,
THOMPSON, F.G., VECCHIONE, M. & WILLIAMS, J.D. 1998.
Common and scientific names of aquatic invertebrates from the United States and
Canada: mollusks. Edn 2. American Fisheries Society Special Publication
26. Bethesda, MD.
WEST, B.J. 2006. Thoughts on modeling complexity. Complexity, 11:
33–43.
WIENS, J.A. 1989. Spatial scaling in ecology. Functional Ecology, 3:
385–397.
J. C. NEKOLA
248
atUniversityofNewMexicoonSeptember9,2014http://mollus.oxfordjournals.org/Downloadedfrom
149
Journal of Biogeography, 26, 867–878
The distance decay of similarity in biogeography
and ecology
Jeffrey C. Nekola∗ and Peter S. White Curriculum in Ecology, University of North Carolina
at Chapel Hill, Chapel Hill, North Carolina 27599, U.S.A.
Abstract
Aim Our aim was to understand how similarity changes with distance in biological
communities, to use the distance decay perspective as quantitative technique to describe
biogeographic pattern, and to explore whether growth form, dispersal type, rarity, or
support affected the rate of distance decay in similarity.
Location North American spruce-fir forests, Appalachian montane spruce-fir forests.
Methods We estimated rates of distance decay through regression of log-transformed
compositional similarity against distance for pairwise comparisons of thirty-four white
spruce plots and twenty-six black spruce plots distributed from eastern Canada to Alaska,
six regional floras along the crest of the Appalachians, and six regional floras along the
east–west extent of the boreal forest.
Results Similarity decreased significantly with distance, with the most linear models relating
the log of similarity to untransformed distance. The rate of similarity decay was 1.5–1.9
times higher for vascular plants than for bryophytes. The rate of distance decay was highest
for berry-fruited and nut-bearing species (1.7 times higher than plumose-seeded species and
1.9 times higher than microseeded/spore species) and 2.1 times higher for herbs than woody
plants. There was no distance decay for rare species, while species of intermediate frequency
had 2.0 times higher distance decay rates than common species. The rate of distance decay
was 2.7 times higher for floras from the fragmented Appalachians than for floras from the
contiguous boreal forest.
Main conclusions The distance decay of similarity can be caused by either a decrease in
environmental similarity with distance (e.g. climatic gradients) or by limits to dispersal and
niche width differences among taxa. Regardless of cause, the distance decay of similarity
provides a simple descriptor of how biological diversity is distributed and therefore has
consequences for conservation strategy.
Keywords
Similarity, spatial dependence, distance decay, biological diversity, boreal forest.
INTRODUCTION analyses of spatial autocorrelation and eventually to the field
of geostatistics, a field that has grown tremendously in the last
The similarity between two observations often decreases or
decade (Cressie, 1993).
decays as the distance between them increases, a pattern long
While ecologists and biogeographers have rarely referred to
recognized in the geographical literature and once called ‘the
distance decay per se, a negative relationship between distance
first law of geography’ (Tobler, 1970). That early interest in
and similarity is implicit in several ecological and evolutionary
the distance decay of similarity led subsequently to formal
phenomena. For example, species turnover along spatial
environmental gradients produces a decrease of similarity with
∗ Present address: Department of Natural and Applied Science, distance (e.g. Whittaker, 1975; Cody, 1985). Mass effect (Shmida
University of Wisconsin-Green Bay, Green Bay, Wisconsin 52311, U.S.A. & Ellner, 1984), source-sink dynamics in metapopulations
email: nekolaj@uwgb.edu
(Hanski & Gilpin, 1991), and supply side ecologyCorresponding author: Peter S. White, Department of Biology,
(Roughgarden, Gaines & Pacala, 1987) have stressed theCB#3280, University of North Carolina at Chapel Hill, Chapel Hill,
North Carolina 27599, U.S.A. email: pswhite@unc.edu importance of dispersal and therefore distance in explanations
© 1999
150
868 J. C. Nekola and P. S. White
of population dynamics and community composition (Palmer, DISTANCE DECAY OF BIOLOGICAL
SIMILARITY: CONCEPTUAL FRAMEWORK1988). Studies in island biogeography have found a decrease
in percent species saturation of oceanic or habitat islands as a
function of their distance from a source pool of immigrants Distance decay and geostatistics
(Vuilleumier, 1970; Kadmon & Pulliam, 1993), producing a
Our approach to the distance decay of similarity is parallel to,
decrease in the similarity of species lists with distance. Dispersal,
but methodologically different from, geostatistical approaches
migration, and gene flow produce geographic patterns in the
to the analysis of spatial autocorrelation (Cressie, 1993). It is
distribution of genetic variation and in the distribution of
now standard practice to use semi-variograms to characterize
species that result from vicariance and adaptive radiations
the relationship of variance of a parameter to intersample
(Meyers & Giller, 1988; Briggs, 1991), leading to spatially
distance or lag (Ripley, 1988), a critical step in the optimal
autocorrelated distributional patterns that were described in
interpolation technique called kriging (Burrough, 1986; Webster
the early biogeographic literature under the age and area and
& Oliver, 1990). To construct a semi-variogram, variance is
centre of origin hypotheses (Cain, 1944; Willis, 1922).
calculated by comparing values of the parameter at two
These examples suggest two very different causes of
locations of known distance apart. The statistic we model here,
biological distance decay. On the one hand, a decay of similarity
similarity, is a parallel concept to variance in the sense that it
results from a simple decrease in environmental similarity
results from comparison of two observations from different
with distance (e.g. movement along a topographic or climatic
locations. In our case the observations consist of species lists;
gradient). In this case, the underlying explanation is competitive
similarity is a complex function of the occurrence of many
sorting of species with different physiological abilities;
species (Faith, Minchin & Belbin, 1987). By contrast, variance
composition, at least after sufficient time has elapsed, is
is calculated for univariate factors (e.g. soil pH, rainfall, or
predictable from environment. We term this source of biological
abundance of a single species).
distance decay the niche difference model.
Variance and similarity both measure differences between
The second cause of distance decay is that the spatial
samples and can be modelled as a function of intersample
configuration (the size and isolation of habitats), spatial context
distance. In this paper, we use linear regression to model that
(the nature of the matrix surrounding the habitats), and time
relationship. The intercept of this model is similar to the nugget
influence species and gene movement across landscapes.
effect (residual intersample variance as distance decreases to
Variation in distance decay rates among landscapes occurs in
zero) in semi-variograms. The shape of the similarity-distance
this case because different landscapes have different resistances
relationship can be examined for evidence of a range and sill
to the movement of organisms. Variation in distance decay
which, in semi-variograms are, respectively, the distance at
rates among different species groups will also arise if the groups
which variance becomes independent of further increases in
have different dispersal abilities—in essence, different species
distance and the amount of variance at that distance. Regression
will perceive a given spatial configuration of habitats differently.
of similarity against distance has the advantage in our case of
Variation in landscape resistance to movement and in dispersal
producing a direct estimate of the rate of distance decay and
abilities suggest an interaction between the spatial template
thus the rate of composition change through space. One of our
and time: the greater the resistance to movement or the less
purposes is to ask whether species with different dispersal types
vagile the organism, the greater the amount of time required
or growth forms show different rates of compositional change.
for the organism to cross a given spatial template. As a result,
the nature of the spatial configuration of habitats will affect
how long the effects of historic events persists in terms of Scale dependence and distance decay
distributional patterns of species. When time or dispersal are
limiting, current environmental conditions will not fully explain Grain, extent and related geostatistical concepts
Observations of distance decay will show scale dependence.species distributions or local composition. We term this source
of distance decay the model of temporal and spatial constraint: Ecologists have divided scale into grain and extent (Allen &
Hoekstra, 1991; Weins, 1989). Grain, which is similar to thehistory and the nature of the spatial template explain species
distributions. At large scales of space and time, the imprint of concept of support or resolution in the geostatistical literature
(Odland, 1988; Webster & Oliver, 1990; Cressie, 1993), is thepast events has been termed historic biogeography (Schluter &
Ricklefs, 1993). contiguous area over which a single observation is made (e.g.
quadrat, plot, or pixel). In this sense it is a special case ofIn this paper we explore distance decay as a descriptor of
compositional variation. We first discuss scale dependence in support, which is defined as an n-dimensional volume within
which a linear averaged value of a regionalized value maythe detection of distance decay and develop a conceptual
framework for distance decay. We then examine distance decay be computed (Olea, 1991). Spatial extent, called area in the
geostatistical literature (Cressie, 1993), is the space over whichrates in boreal and montane spruce-fir forests, comparing the
rate of similarity decay in several species groups (vascular observations are made.
The grain size or support of samples will effect observedplants, mosses, species abundance classes, growth form classes,
and dispersal classes) across the same set of plots. Finally, similarities, a phenomenon known as the support effect in
the geostatistical literature. Consider two areas with identicalwe discuss the implications of distance decay perspective for
community composition, biogeographical pattern, and species lists. If a small quadrat is sampled from each area,
species-area relationships dictate that these quadrats will haveconservation.
© Blackwell Science Ltd 199926867–878
151
Distance decay of similarity 869
only a subset of the possible species and will contain identical
species lists only a portion of the time. If we increase grain
size, however, species lists will become more complete until,
at a grain size that is equivalent to the size of the study areas,
all possible species will be contained and the observed similarity
will equal the real similarity. If two regions have a similarity
greater than zero, small samples, by chance, can have either
higher or lower similarity than the real similarity. The
distribution of sample scores will vary with the distribution of
individuals among species, the spatial pattern of species and
individuals, the sampling design, and that component of the
species-area relationship that we term the species-grain
relationship (Palmer & White, 1994). In general, we expect
that the variation in observed similarities among samples will
increase as grain size decreases, assuming that the real similarity
is a positive value. This effect of change in support or grain size
on sample statistics has been well discussed in the geographical
literature as the modifiable areal unit problem (Openshaw,
1983). It has generally been found that variation increases as
support or grain size decreases.
Although distance decay proposes that similarity decreases
with extent, natural periodicities in the environment may cause
a lack of distance decay at some scales of observation (Fig. 1).
Geostatisticians have treated similar patterns of variance as
the problem of non-stationarity (Webster & Oliver, 1990). For
example, if extent is only large enough to include a mountain
crest and its adjacent valley, distance decay will likely occur
as a result of the strong environmental gradient that is present.
However, if extent is increased to include the next mountain
crest, similarity of community composition will likely increase
in relation to the first mountain summit. If extent is increased
further still to include a large number of mountain crests and
Figure 1 The influence of sample grain and extent on the
valleys, distance decay will likely re-emerge from either climatic
observation of distance decay. A, A hypothetical landscape with
gradients or the presence of dispersal barriers between places spatial periodicities in similarity controlled by strong environmental
with similar environments. gradients (elevation contours); B, the distance decay of similarity as
The detection of distance decay requires that the variation detected by samples of differing grain and extent.
due to small grain size or support be less than the variation
due to spatial extent. Increasing the grain size or support of a the spatial template) and two characteristics of the organisms
sample will decrease variation, and may allow detection of under study (niche-breadth and dispersal ability). The effects
weak distance decay trends. Alternatively, increasing extent of each of these factors are discussed below.
while holding sample grain or support constant may allow
identification of distance decay trends because the variation Environmental distance
due to distance will be larger than the noise present from small For a fixed spatial distance, the steeper the rate of environmental
grain size or support. The strongest distance decay relationships change, the more rapid the turnover in species composition
should be observed when the variation due to sampling error and the higher the expected rate of distance decay. Ecologists
is minimized (with large sample grains or support) and the often have assumed that composition is solely a function of
variation due to distance decay is maximized (with large sample environment (e.g. MacArthur, 1972; Tilman, 1988). For the
extents). rate of distance decay to be solely controlled by environmental
distance, all species must have access to all appropriate habitats
and time must be sufficient to allow competitive interactionsA conceptual model for variation in distance decay
to sort species as a function of environment. In essence, thisrates
summarizes Beijerinck’s Law: everything is everywhere but the
The rate of distance decay will vary with the strength of environment selects (Saur, 1988). Niche characteristics and
environmental gradients and will differ among taxa and competition determine composition.
between landscapes of different spatial configurations and
histories. More formally, we suggest that the rate of decay in The spatial template
Characteristics of the spatial template (habitat size, isolation,biological similarity will be correlated with two characteristics
of the environment (environmental distance and the nature of and the nature of the matrix) will effect the rate of distance
© Blackwell Science Ltd 199926867–878
152
870 J. C. Nekola and P. S. White
decay, all else being equal. The more isolated the habitat, the The second data set consisted of regional vascular plant lists
for the boreal forest and Appalachian montane spruce-fir. Whileless efficient dispersal will be. Some species will not locate all
available habitats (Huffaker, 1958; Shmida & Ellner, 1984), the boreal forest exists in a relatively continuous band from
Alaska to Newfoundland, montane spruce-fir is restricted toleading to lower levels of species overlap among sites and
higher rates of distance decay. Factors which would effect the highest peaks, with the areas dominated by spruce and fir
separated from one another as a series of island-like clustersdistance decay rates are: the unfavourability of the matrix
(which would affect rates of movement), the favourability of (Cogbill & White, 1991). Species lists for six boreal and six
Appalachian spruce-fir forest regions were compiled from thethe habitat patches (which would affect size, reproduction, and
persistence of the populations), and the presence of corridors literature and corrected to a common nomenclatural base
(Crandall, 1958; Curtis, 1959; Grigal, 1968; Lakela, 1965; LaRoi,or stepping stone connections. In a sense, these measure the
resistant to movement of the landscape. The percolation 1967; Maycock, 1961; Maycock & Curtis, 1960; McIntosh &
Hurley, 1964; Nicholson, Holway & Scott, 1969; Ohmann &coefficient is an example of a measure of the affect of spatial
characteristics on movement. This coefficient represents the Ream, 1971; Oosting & Billings, 1951; Ramseur, 1960; Siccama,
1974; White, 1982; White & Miller, 1988). Similarity waspermeability of a given landscape to unidirectional organism
movement (Gardner et al., 1989). calculated between all pairwise comparisons of these regions
within both groups. This data set has a larger support (each
list is a regional flora) than the one from LaRoi (1967).Niche breadth and overlap
The wider the average niche breadth and overlap for a group We addressed several questions using these data sets: Does
community similarity decrease with distance and, if so, whatof organisms, the lower the rate of distance decay which will
occur over a fixed amount of environmental distance. Realized statistical model best describes this relationship? Do isolated
habitats (Appalachian mountain summits and slopes) have aniche breadth and overlap may decrease over successional
time as species sort along gradients (Werner & Platt, 1976; higher rate of distance decay than more contiguous ones (boreal
forest)? Do data sets with larger grain size or support (theChristensen & Peet, 1984).
regional floristic lists), because they have more complete species
lists, have higher overall similarities and lower distance decayDispersal ability
Highly vagile species will be less affected by barriers and will rates than data sets with smaller grain size or support (boreal
forest plot data)? Do wind-dispersed species (bryophytes andbe able to locate most or all of the appropriate habitats in a
landscape (Bush & Whittaker, 1991; Dzwonko & Loster, 1992) wind-dispersed vascular plants) have higher rates of distance
decay than less vagile species? Do taller and dominant speciesand such dispersal may compensate for low persistence rates
(e.g. as suggested by the source-sink population model and the (which may have broader environmental tolerances) have lower
distance decay rates than smaller and understory species?rescue effect, Brown & Kodric-Brown, 1977). More vagile taxa
should therefore show lower rates of distance decay over a We addressed the questions about dispersal, dominance, and
frequency by using data from a single set of plots, thus holdinggiven landscape. Species with poorer vagilities will be unable
to cross barriers, leading to their absence from potential sites spatial configuration and environmental distance constant for
all groups compared. The distance decay rates for bryophytes(Weddell, 1991), contributing to higher distance decay rates.
and vascular plants were compared with data from both white
spruce and black spruce plots. Comparisons among vascular
DISTANCE DECAY IN SPRUCE-FIR FORESTS
plant classes were made with data from white spruce plots only.
The following paragraphs describe the questions addressed withThe boreal and montane spruce-fir forests of eastern and
northern North America have similar structures over large white spruce plots in more detail.
areas (Cogbill & White, 1991). Two data sets were constructed
from published studies of these forests. In the first, the vascular
plants and bryophytes of the North American boreal forests The contribution of species in different frequency classes to
distance decay ratewere analysed from data in LaRoi (1967) and LaRoi & Stringer
(1976). These papers present species lists for nine hectare plots Species were divided into three classes: species found on one
to three plots (139 species), four to fourteen plots (eighty-four(300×300 m) distributed from Newfoundland to Alaska in two
community types: upland white spruce forests (thirty-four plots species), and fifteen to thirty-four plots (twenty-nine species).
We also summed the last two classes to produce a fourth class,containing 252 vascular plant species and 118 bryophyte species)
and lowland black spruce forests (twenty-six plots containing those species found on four to thirty-four plots (113 species).
Common species are shared by many plots and therefore195 vascular plant species and seventy-nine bryophyte species).
Similarity and distances were calculated for all pairwise decrease the overall rate of distance decay (in the extreme, for
species found in all plots, intersample similarity is always 1.0combinations of plots within these two habitat types (a total
of 561 values for white spruce stands and 351 for black spruce and is independent of distance). Species found in only one plot
would lower the similarity between any two plots but wouldstands). These data sets were chosen because they allow us to
examine distance decay for relatively homogeneous forests not contribute to a spatial trend in similarity at this scale
(similarity based on these species would be zero at all distances).(each of the habitats is dominated by the same tree species
throughout) in relatively unbroken topography across a large We therefore hypothesized that species with intermediate
frequency would show the steepest rate of distance decay.geographic extent.
© Blackwell Science Ltd 199926867–878
153
Distance decay of similarity 871
Variation in distance decay rate among growth forms were assembled (we used non-zero values because we were
interested in the slope of the distance decay relationship overCasual observation suggested that trees in the boreal forest
have wider geographic distributions than plants with other those distances that have a measured similarity—presumably
similarity falls to zero at some distance and then isgrowth forms. This suggests wider environmental tolerances
and a lower distance decay rate. We therefore investigated independent of distance thereafter); (2) after testing the
difference in mean similarity for the two data subsets,whether there were systematic differences in distance decay
rates for plants with different growth forms. Species were similarity values within each subset were rescaled to a
common mean (we sought to test the difference in slopesdivided into five classes: trees (nineteen species), tall shrubs
(twenty-two species), low shrubs (thirty-five species), subshrubs of the distance decay relationships independent of differences
in the means); (3) for each pair of sites, the two similarity(sixteen species), and herbs (160 species). We also compared
the bryophyte data to the data from these vascular plant growth values (one for each of the two data subsets to be compared)
were randomly reassigned to the two data subsets; (4) afterform classes.
this randomization had been carried out for all pairs of
sites, linear regression was used to determine the slope ofVariation in distance decay rate among dispersal classes
the distance decay relationship for each of the randomizedMore vagile species should have lower distance decay rates
subsets; (5) step four was repeated to producethan less vagile species. Species were divided into three major
10,000 values for the slope of the distance decay relationshipand five total dispersal classes: wind dispersed species (a total
for randomized subsets; (6) finally, the absolute differenceof sixty-three species, comprised of two subcategories, twentybetween
the slopes in each of the 10,000 randomizationseight spore/microseed species and thirty-five plumose seed
was compared to the difference between the observed slopesspecies), berry-fruited/nut-bearing species (sixty-four species),
based on the original data (and thus the true assignmentsand species with other dispersal types (125 species). We also
to data subset categories) and we calculated how many timescompared the bryophyte data to the data from these vascular
(out of 10,000 trials) randomization produced a differenceplant dispersal classes.
in slope between the subsets that was at least as great as
the observed difference in slope based on the original data.
Similarity index If correct assignment to data categories is significant (if the
We used Jaccard’s Index (Mueller-Dombois & Ellenberg, 1974) categories are indeed different in slope), then the observed
due to its simplicity, its widespread use, its accessibility through difference should exceed the randomized differences a high
common statistical packages, and its use of presence/absence percentage of the time. We used this percentage as the P
data. We chose a presence-absence metric because it represents value for our tests.
a more conservative measure of community similarity than
ones based on species abundance data which are more sensitive
to disturbance and local environmental differences.
RESULTS
Data analysis
Distance decay in spruce-fir forests
For the largest empirical data sets (white spruce and black
spruce stands of the boreal forest), similarity and distance were In all data sets except for rare species in white spruce stands,
similarity decreased significantly with distance (Figs 2–6, Tablecalculated and then log-transformed. Linear regression was
used to calculate distance decay rates or all four combinations 1). The most linear relationship and the most heteroscedastic
regression residuals were found in models that used the logof these transformed and untransformed variables. Inspection
of the plots and regression residuals were used to determine transformation of similarity and untransformed distance. A
semi-log model was thus used in all of the regression modelswhich formulation most closely approximated a linear
relationship over most intersample distances. presented here. This implies an exponential rate of distance
decay. For the boreal forest plot data, the rate of similarityBecause the values in the matrices (561 for white spruce
stands, 351 for black spruce stands) are not independent, decay ranged from −0.17 (black spruce bryophytes; units are
in ln (similarity) per 1000 km of distance) to −0.55 (whiterandomization tests were used to establish the significance of
regression models and to make comparisons of distance decay spruce vascular plants of intermediate frequency) and r2
values
ranged from 0.24 (black spruce bryophytes) to 0.71 (commonrates among various subsets of the data. Mantel tests with
10,000 replications (Jackson & Somer, 1989) were used to vascular plants on white spruce plots). For the regional floras,
the rate of similarity decay ranged −0.25 (boreal forest, r2
=estimate significance of each model (Manly, 1991).
The comparison of distance decay rates among various 0.54) and −0.67 (Appalachian montane forest, r2
=0.80). All
of these relationships are fit well by a straight line in log-linearsubsets of the data was done with a randomization test
developed by Blake Weiler of the Ecology Curriculum at the space (Figs 2–6), with no obvious sill (i.e. no distance at which
the data imply a change of slope for the regression). Our closestUniversity of North Carolina at Chapel Hill. Our objective
was to compare the slopes of the distance decay relationship observations are not close enough for a detailed analysis of
the intercept value for the regressions, which is analogous tobetween two data subjects (e.g. trees v. herbs) across all
plots. The method followed this outline: (1) all non-zero the nugget effect in semi-variograms, although we make a few
comments on that parameter below.similarities based on the two data subsets to be compared
© Blackwell Science Ltd 199926867–878
154
872 J. C. Nekola and P. S. White
Figure 2 The log of similarity plotted
against distance for all pairwise comparisons
of plots sampled across the boreal forest
based on data in LaRoi (1967) and LaRoi &
Stringer (1976). A, Vascular plants in thrityfour
white spruce stands; B, bryophytes in
thirty-four white spruce stands; C, vascular
plants in twenty-six black spruce stands; D,
bryophytes in twenty-six black spruce stands.
Figure 3 The log of similarity plotted
against distance for species frequency classes
(white spruce stands only; data from LaRoi,
1967). A, Vascular plants found in four or
more plots; B, vascular plants found in fifteen
to thirty-four plots; C, vascular plants found
in four to fourteen plots; D, vascular plants
found in three or fewer plots.
The effect of frequency, growth form, dispersal type, distance decay rates for regional floras (Fig. 6) with those for
plot data (Fig. 2). Both include data taken over the geographicand habitat isolation on distance decay rates
extent of boreal forest, but the floristic lists include rare species
that are at their range limits at each location, whereas the plotsSpecies frequency
When rare species (those occurring in three or less of the 34 are less likely to sample species that are very rare in the
landscape as a whole. The fuller regional species lists (andwhite spruce plots) were eliminated from the data set, the rate
of distance decay was unaffected (Fig. 3A; Table 1), but larger support) of the floristic data set lowers the observed rate
of similarity loss per unit distance. In a sense, the larger supportsimilarities were generally higher, the intercept was closer to
unity (complete similarity at a distance of 0), and the r2
value reduces the observed rarity of species. Thus, the rate of distance
decay for boreal forest plot data was 1.5 times higher thanwas higher. Thus, rare species lower overall similarity but do
not affect the rate of distance decay. As might be expected that for boreal forest floristic lists across the same geographical
extent (Table 1).from this analysis, the rare species themselves showed no decay
of similarity as a function of distance (Fig. 3B).
The species occurring in four or more plots were divided Growth form
In the white spruce plots, herbs had 1.7–2.1 times higherinto two groups: common species (fifteen or more plots) and
intermediate species (four to fourteen plots). The common distance decay rate than woody plant classes (P < 0.005–0.0001;
Tables 1 and 2, Fig. 4). The higher distance decay rate for thespecies had a higher intercept value than intermediate species
and a rate of distance decay (−0.27) that was only one half smallest vascular plants was reversed for the contrast between
vascular plants and bryophytes: herbs had 1.6 times the ratesas great as the rate for intermediate species (−0.55) (Fig. 3C–D;
Table 1). for bryophytes in the same plots (Fig. 2, Table 1). Interestingly,
bryophytes had similar rates of distance decay to tall woodyThe influence of rare species can also be seen by comparing
© Blackwell Science Ltd 199926867–878
155
Distance decay of similarity 873
Figure 4 The log of similarity plotted
against distance for growth form classes
(white spruce stands only; data from LaRoi,
1967). A, Trees; B, tall shrubs; C, low shrubs;
D, subshrubs; E, herbs.
Figure 5 The log of similarity plotted
against distance for dispersal classes (white
spruce stands only; data from LaRoi, 1967).
A, Wind-dispersed species; B, species with
plumose seeds; C, species with spores or
microscopic seeds; D, species with nuts or
fleshy berries; E, species with other types of
seeds.
plants on the same plots; they also had similar rates to wind- Dispersal type
For vascular plants on white spruce plots, berry-fruited anddispersed vascular plants (see below). Vascular plants as a
group had 1.5 (white spruce plots) and 1.9 (black spruce plots) nut-bearing species had nearly twice the rate of distance decay
shown by spore/microseed and plumose seed species (−0.43 v.times higher distance decay rates than bryophytes of the same
plots (comparisons significant at the P < 0.001 level; Table 1). −0.23 and −0.25, P < 0.001; Tables 1 and 2, Fig. 5B, C, and
© Blackwell Science Ltd 199926867–878
156
874 J. C. Nekola and P. S. White
Figure 6 The log of similarity plotted
against distance for all pairwise comparisons
of vascular plant floras from six regions
across the northern boreal forest [data
modified from Curtis (1959), Grigal (1968),
Lakela (1965), LaRoi (1967), Maycock (1961),
Maycock & Curtis (1960) and Ohmann &
Ream (1971)] and six areas along the crest of
the Appalachian mountains [data modified
from Crandall (1958), LaRoi (1967),
McIntosh & Hurley (1964), Nicholson,
Holway & Scott (1969), Oosting & Billings
(1951), Ramseur (1960), White (1982)]. A,
Boreal spruce-fir forests; B, Appalachian
spruce-fir forests.
Table 1 Regression statistics for the decay of
similarity with distance in spruce-fir data sets.
The sample size (n) is the number of
similarity values and distances in each matrix.
The number of plots or areas are as follows:
white spruce plots, 34; black spruce plots, 26;
boreal spruce-fir areas, 6; Appalachian
montane spruce-fir areas, 6. For the three
data subsets of the white spruce plots, the
sample size is sometimes lower than 561
because of the removal of plot pairs with a
similarity of zero. Intercepts are in units of
ln(similarity). Slope values are in units of
ln(similarity) per 1000 km distance.
Data set Species n Intercept Slope r2
P
White spruce vascular plants 252 561 −0.75 −0.37 0.67 0.0001
White spruce bryophytes 118 561 −0.87 −0.25 0.53 0.0001
Black spruce vascular plants 195 351 −0.88 −0.33 0.57 0.0001
Black spruce bryophytes 70 351 −0.98 −0.17 0.24 0.0001
Data subsets (white spruce plots only)
Frequency class:
4–34 plots 113 561 −0.57 −0.37 0.71 0.0001
15–34 plots 29 561 −0.24 −0.27 0.59 0.0001
4–14 plots 84 467 −1.04 −0.55 0.66 0.0001
1–3 plots 139 63 −2.40 −0.006 0.00 n.s.
Growth form:
Trees 19 555 −0.51 −0.19 0.30 0.0001
Tall shrubs 22 271 −1.21 −0.20 0.17 0.0001
Low shrubs 35 472 −1.09 −0.21 0.20 0.0001
Subshrubs 16 528 −0.66 −0.28 0.45 0.0001
Herbs 160 554 −0.84 −0.40 0.52 0.0001
Dispersal class:
Wind 63 485 −1.14 −0.29 0.38 0.0001
Plumose 35 329 −1.09 −0.25 0.31 0.0001
Spore/microscopic 28 408 −1.02 −0.23 0.28 0.0001
Fleshy fruits, nuts 64 532 −0.62 −0.43 0.62 0.0001
Other 125 561 −0.84 −0.27 0.48 0.0001
Boreal forest regions 250 15 −0.76 −0.25 0.54 0.0039
Appalachian spruce-fir 226 15 −0.18 −0.67 0.80 0.0018
© Blackwell Science Ltd 199926867–878
157
Distance decay of similarity 875
Table 2 Significance levels for differences in
slopes for the growth form and dispersal
classes.
A. Growth form
Tall shrubs Low shrubs Subshrubs Herbs
Trees 0.05 0.05 0.0001 0.0001
Tall shrubs (0.18) 0.01 0.005
Low shrubs 0.005 0.0001
Subshrubs 0.0001
B. Dispersal class
Plumose Spore/micro. Fleshy Other
Wind (0.0649) 0.0001 0.0001 0.004
Plumose (0.645) 0.0001 (0.79)
Spore/microseeded species 0.0001 (0.306)
Fleshy fruited, nut-bearing 0.0001
species
E). Berry-fruited and nut-bearing species had the highest r2
change in similarity per unit distance. This is attractive because
it suggests that the processes underlying the distance decayvalues among the dispersal types (Table 1). These data suggest
that wind-dispersed species are both more widely dispersed relationship vary only with environmental distance, spatial
configuration, and organism attributes but not with distanceand more variable in their occurrences. Bryophytes had similar
distance decay rates to spore/microseeded and plumose seeded independent of these factors. Hubbell (1995) has produced a
simulation model of species colonization and extinction thatvascular plants on the same plots (Table 1).
results in just such a decrease in similarity with distance.
Habitat isolation
Appalachian forest communities had a rate of similarity decay
Variation in distance decay rates
2.7 times that observed for the continuous northern boreal
forest (Fig. 6, Table 1). The rate of distance decay varied among species groups in
expected ways: on the same plots, spore/microseeded and
plumose species had lower rates of distance decay than larger
DISCUSSION
fruited species. Trees, other woody plants, wind dispersed, and
more frequent species raised overall similarity and lowered the
Form of the distance decay model
rate of distance decay, while herbaceous plants, heavy fruited,
and less frequent species lowered the rate. Further, the moreIn the geographic literature, a distance effect on the interaction
between human populations has been modelled with a power fragmented montane forests dominated by the same genera of
trees as the boreal forest had much higher rates of distanceor gravity function (Bennett & Gade, 1979):
decay than the more contiguous boreal forest itself. We will
I=A∗d−c
consider alternative explanations for these patterns.
While it can be suggested that due to their small propagulewhere I is interaction, A is a constant, d is distance, and c
is the friction coefficient (correlated with technologies for size bryophytes should be more vagile and possess wider ranges
along geographic gradients than vascular plant species (Saur,communication and barriers to movement over a given
distance). This model implies an ever decreasing effect of a 1988), other explanations are reasonable. The smaller size of
bryophytes compared to vascular plants may allow persistenceunit distance as distance becomes larger and a linear relationship
only under double log transformation. in microsites not available to vascular plants, thus increasing
geographic range. Bryophytes may have broader physiologicalThe finding that a semi-log model showed the most linear
relationships suggests similarity could in fact be related to tolerances and higher persistence rates because they can cycle
rapidly between active and dormant states.distance through an exponential model:
In light of these possibilities, it is interesting to note that
S=S0e−cd
herbs, which are the smallest vascular plants (thus sharing,
with bryophytes, environments that have less fluctuating airwhere S is similarity at distance d; S0 is initial similarity or
similarity at distance 0; c is a constant for the rate of distance temperatures), showed the highest rates of similarity loss with
distance of all groups. By contrast, wind-dispersed vasculardecay; and d is distance. Exponential decay models have
previously been used in biogeography and ecology to model plants had very similar rates of distance decay to bryophytes
on these plots regardless of plant stature. Tall vascular plantsplant dispersal (Okubo & Levin, 1989) and to describe the
relationship between spatial distance and floristic similarity in had as low or lower distance decay rates than bryophytes.
In terms of the comparison of wind and non-wind dispersedthe Galapagos (Preston, 1962). The exponential model, in
contrast to the power model, implies a constant proportional species, dispersal type could be correlated with other traits that
© Blackwell Science Ltd 199926867–878
158
876 J. C. Nekola and P. S. White
influence the rate of geographic turnover. For example, wind- Regardless of whether we can evaluate causes, the decay of
similarity with distance is a relatively simple descriptor ofdispersed species could have wider niche breadth (as has been
argued for early successional species generally), thus showing variation in species composition and has consequences for
how we conserve biological diversity. Whether because ofa lower rate of distance decay. On the other hand, winddispersed
species often have small seeds and may therefore environmental distance decay or spatial and historical factors,
the distance decay paradigm suggests a natural world that is,have a lower probability of establishment from a given dispersal
event, particularly in undisturbed forest studies here (i.e. the at some scales, everywhere unique. Distance decay rates describe
the rate of turnover that underlies this uniqueness.probability of establishment in late successional stands could
be correlated with seed size). It is unclear how this would Simberloff & Abele (1970) argued that the theory of island
biogeography was silent on preserve design; that is, this theoryinfluence distance decay; a simple expectation is that it would
tend to make similarity more variable at any distance. could not be used to argue whether one large area would have
more species than several smaller areas that summed to theThe higher rate of distance decay in the Appalachian montane
spruce-fir forests could be due to less efficient dispersal among same total area. They suggested that the critical factor was the
degree of overlap (i.e. the similarity) among the smaller areas.fragmented habitats, but it could also be the result of history
(the southern and northern Appalachian spruce-fir forest may Shafer (1990) showed that whether one area or several small
areas of the same total area had more species or not washave had access to different refugial populations after climatic
warming started 15,000 years BP) and/or steeper rate of dependent on two factors: the slope of the log(species)-log(area)
relationship and the similarity among the smaller areas. Weenvironmental change with distance. In terms of environmental
distance, montane spruce-fir forests have very similar growing add here that if the smaller areas (which are by definition
non-contiguous and thus spatially dispersed compared to oneseason temperatures across the wide range of latitudes covered
in the data set (a mean July temperature of 17°C is correlated contiguous area) are more different from one another as a
function of the distance between them (see also Palmer &with their lower elevational boundaries), but are exposed to
increasingly low winter temperatures northward (Cogbill & White, 1994), then the decay of similarity with distance is a
central explanation of why some investigators (e.g. StohlgrenWhite, 1991).
Sorting out the relative importance of environment and & Quinn, 1991) find that smaller dispersed conservation areas
collectively have more species than large conservation areas.niche-based explanations from spatial configuration and
dispersal-based ones is a difficult challenge. Explanations of In essence, conservation strategy requires both grain size for
population persistence (Shoenwald-Cox, 1983; Newmark, 1987)species composition based on environmental factors and
competition are old ones in ecology. However, explanations and spatial extent (for capturing areas dissimilar to one
another). The SLOSS (single large or several small) debate canbased on history and dispersal have also been proposed from
the beginning. Gleason (1923) suggested that in the Midwest never be resolved because size (grain) and subdivision (extent)
maximize different aspects of how biological diversity ispassive dispersal of propagules was the major factor explaining
community composition. Palmgren (1926) found that many distributed. The distance decay of similarity, in this sense,
provides a metric and a conceptual base for such efforts asplant species of the A˚ land Archipelago, Finland, were absent
from seemingly suitable sites, and suggested that dispersal GAP analysis that seek to identify habitats and species left out
of existing conservation areas (Davis et al., 1990).limitation was responsible for this pattern.
When species do not disperse to all appropriate sites,
composition will be constrained by the spatial configuration
of sites and dispersal abilities of species. At one extreme (if all REFERENCES
species had the same dispersal abilities), community
Allen, T.F.H. & Hoekstra, T.W. (1991) Role of heterogeneity in scalingcomposition would be a stochastic subset of the species which
of ecological systems under analysis. Ecological heterogeneity (ed.can occupy that habitat. Shmida & Ellner (1984) and Shmida
by J. Kolasa and S.T.A. Pickett), pp. 47–68. Ecological Studies 86.
& Wilson (1985) have shown how such conditions may lead
Springer-Verlag, New York.
to the coexistence of co-competitive species in a landscape, Bennett, D.G. & Gade, O. (1979) Geographic perspectives in migration
which are allowed to co-exist only because populations are research. University of North Carolina at Chapel Hill Department
spatially separated from each other at any given moment. At of Geography Studies in Geography, Number 12.
the other extreme (if species have a graded series of dispersal Briggs, J.C. (1991) Historical biogeography: the pedagogical problem.
J. Biogeogr. 18, 3–6.abilities), community composition at each location would be
Brown, J. & Kodric-Brown, A. (1977) Turnover rates in insulara nested subset (varying with habitat isolation) of the full
biogeography: effect of immigration on extinction. Ecology, 58,species list. Hierarchical nesting of species lists has been much
445–449.discussed in the context of habitat area and area-sensitive
Burrough, P.A. (1986) Principles of geographical information systems
species (Patterson, 1987); similarly, there may be isolationfor
land resources assessment. Oxford University Press, New York.
sensitive species. Presences explained not by local environment
Bush, M.B. & Whittaker, R.J. (1991) Cracked: colonization patterns
but by continued dispersal has been termed mass effect (Shmida and hierarchies. J. Biogeogr. 18, 341–356.
& Wilson, 1985); here we add that absences can be the result, Cain, S.A. (1944) Foundations of plant geography. Hafner Publishing
not of local environment, but of dispersal limitations and of Co., New York.
stochastic patterns of original colonizing populations (e.g. the Christensen, N.L. & Peet, R.K. (1984) Secondary forest succession
on the North Carolina Piedmont. Forest succession: concepts andimprint of initial refuge competition and migration routes).
© Blackwell Science Ltd 199926867–878
159
Distance decay of similarity 877
application (ed. by D.C. West, H.H. Shugart and D.B. Botkin), biogeography. Analytical biogeography (ed. by A.A. Meyers and P.S.
Giller), pp. 3–12. Chapman and Hall, New York.pp. 230–245. Springer-Verlag, New York.
Cody, M.L. (1985) Towards a theory of continental species diversities. Mueller-Dombois, D. & Ellenberg, H. (1974) Aims and methods of
vegetation ecology. John Wiley and Sons, New York.Ecology and evolution of communities (ed. by M.L. Cody and J.M.
Diamond), pp. 214–257. Harvard University Press, Cambridge, Mass. Newmark, W.D. (1987) A land-bridge perspective on mammalian
extinctions in western North American parks. Nature, 325, 430–432.Cogbill, C.V. & White, P.S. (1991) The latitude-elevation relationship
for Appalachian spruce-fir. Vegetatio, 94, 153–175. Nicholson, S., Holway, J.G. & Scott, J.T. (1969) A floristic comparison
of undisturbed spruce-fir forests of the Adirondacks with four otherCrandall, D.L. (1958) Ground vegetation patterns of the spruce-fir area
of the Great Smoky Mountains National Park. Ecol. Monogr. 28, regions. Atm. Sci. Res. Center, State U. of N.Y. at Albany. Rept. 92.
Odland, J. (1988) Spatial autocorrelation. SAGE Publications, Newbury337–360.
Cressie, N.A. (1993) Statistics for spatial data. John Wiley & Sons, Park, California.
Ohmann, L.F. & Ream, R.R. (1971) Wilderness ecology: virgin plantNew York.
Curtis, J.T. (1959) The vegetation of Wisconsin. University of Wisconsin communities of the Boundary Waters Canoe Area. North Central
For. Exp. Sta. Rept. NC-63. St. Paul, Minnesota.Press, Madison.
Davis, F.W., Stoms, D.M., Estes, J.E., Scepan, J. & Scott, J.M. (1990) Okubo, A. & Levin, S.A. (1989) A theoretical framework for data
analysis of wind dispersal of seeds and pollen. Ecology, 70, 329–338.An information systems approach to the preservation of biological
diversity. Int. J. Geogr. Info. Syst. 4, 55–78. Olea, R.A. (1991) Geostatistical glossary and multilingual dictionary.
Oxford University Press, Oxford, England.Dzwonko, Z. & Loster, S. (1992) Species richness and seed dispersal
to secondary woods in southern Poland. J. Biogeogr. 19, 195–204. Oosting, H.J. & Billings, W.D. (1951) A comparison of virgin sprucefir
in the northern and southern Appalachian systems. Ecology, 32,Faith, D.P., Minchin, P.R. & Belbin, L. (1987) Compositional dissimilarity
as a robust measure of ecological distance. Vegetatio, 69, 84–103.
Openshaw, S. (1983) The modifiable areal unit problem. Concepts and57–68.
Gardner, R.H., O’Neill, ??, Turner, M.G. & Dale, V.H. (1989) Quan- Techniques in Modern Geography No. 38. Geobooks, Norwich,
England.tifying scale-dependent affects of animal movement with simple
percolation models. Landscape Ecol. 3, 217–227. Palmer, M.W. (1988) Fractal geometry: a tool for describing spatial
pattern of plant communities. Vegetatio, 75, 91–102.Gleason, H.A. (1923) The vegetation history of the middle west. Ann.
Assoc. Am. Geogr. 12, 39–85. Palmer, M.W. & White, P.S. (1994) Scale dependence and the speciesarea
relationship. Am. Nat. 144, 717–740.Grigal, D.F. (1968) The relationship between plants communities and
soils in northeastern Minnesota. PhD Dissertation, University of Palmgren, A. (1926) Chance as an element in plant geography. Proceedings
of the International Congress of Plant Sciences, Ithaca, NewMinnesota, Minneapolis.
Hanski, I. & Gilpin, M. (1991) Metapopulation dynamics: brief history York (ed. by B.M. Duggar), pp. 591–602. George Banta Publishing
Company, Menasha, Wisconsin.and conceptual domain. Biol. J. Linn. Soc. 42, 3–16.
Hubbell, S.P. (1995) Towards a theory of biodiversity and biogeography Patterson, B.D. (1987) The principle of nested subsets and its implications
for biological conservation. Conserv. Biol. 1, 323–334.on continuous landscapes. Preparing for global change: a midwestern
perspective (ed. by G.R. Carmichael, G.E. Folk and J.L. Schnoor), Preston, F.W. (1962) The canonical distribution of commonness and
rarity: parts I and II. Ecology, 43, 185–215; 410–432.pp. 173–201. Academic Publishing bv, Amsterdam.
Huffaker, C.B. (1958) Experimental studies on predation: dispersion Ramseur, G.S. (1960) The vascular flora of the high mountain communities
of the southern Appalachians. J. Elisha Mitchell Sci. Soc.factors and predator-prey oscillations. Hilgardia, 27, 343–383.
Jackson, D.A. & Somer, K.M. (1989) Are probability estimates from 76, 82–112.
Ripley, B.D. (1988) Statistical inference for spatial processes. Cambridgethe permutation model of Mantel’s test stable? Can. J. Zool. 67,
766–769. University Press, Cambridge.
Roughgarden, J., Gaines, S.D. & Pacala, S.W. (1987) Supply sideKadmon, R. & Pulliam, H.R. (1993) Island biogeography: effect of
geographical isolation on species composition. Ecology, 74, 977–981. ecology: the role of physical transport processes. Organization of
communities: past and present (ed. by J.H.R. Gee and P.S. Giller),Lakela, O. (1965) A flora of northeastern Minnesota. University of
Minnesota Press, Minneapolis. pp. 491–518. Blackwell Scientific, Oxford.
Saur, J.D. (1988) Plant migration. University of California Press,LaRoi, G.H. (1967) Ecological studies in the boreal spruce-fir forests
of the North American taiga. I. Analysis of the vascular flora. Ecol. Berkeley.
Schluter, D. & Ricklefs, R.E. (1993) Species diversity: an introductionMonogr. 37, 229–253.
LaRoi, G.H. & Stringer, M.H. (1976) Ecological studies in the boreal to the problem. Species diversity in ecological communities (ed. by
R.E. Ricklefs and D. Schluter), pp. 1–10. University of Chicago Press,spruce-fir forests of the North American taiga. II. Analysis of the
bryophyte flora. Can. J. Bot. 54, 619–643. Chicago.
Schoenwald-Cox, C. (1983) Guidelines to management: a beginningMacArthur, R.H. (1972) Geographical ecology. Harper and Row, New
York. attempt. Genetics and conservation (ed. by C. Schoenwald-Cox,
S.M. Chambers, B. MacBryde and W.L. Thomas), pp. 414–446.Manly, B.F.J. (1991) Randomization and Monte Carlo methods in
biology. Chapman and Hall, New York. Benjamin/Cummings Publishing, New York.
Shafer, C. (1990) Nature reserves: Island theory and conservationMcIntosh, R.P. & Hurley, R.T. (1964) The spruce-fir forests of the
Catskill Mountains. Ecology, 45, 314–326. practice. Smithsonian Institute Press, Washington, DC.
Shmida, A. & Ellner, S. (1984) Coexistence of plant species with similarMaycock, P.F. (1961) Botanical studies on Mont St. Hilaire, Rouville
County, Quebec. Can. J. Bot. 39, 1293–1325. niches. Vegetatio, 58, 29–55.
Shmida, A. & Wilson, M.V. (1985) Biological determinant of speciesMaycock, P.F. & Curtis, J.T. (1960) The phytosociology of boreal
conifer-hardwood forests of the Great Lakes region. Ecol. Monogr. diversity. J. Biogeogr. 12, 1–20.
Siccama, T.G. (1974) Vegetation, soil, and climate on the Green30, 1–35.
Meyers, A.A. & Giller, P.S. (1988) Process, pattern, and scale in Mountains of Vermont. Ecol. Monogr. 44, 325–349.
© Blackwell Science Ltd 199926867–878
160
878 J. C. Nekola and P. S. White
Simberlof, D.S. & Abele, L.G. (1976) Island biogeography and con- richness in the southern Appalachian high peaks. J. Ecol. 76, 192–199.
Whittaker, R.H. (1975) Communities and ecosystems. MacMillan Pub-servation practice. Science, 191, 285–286.
lishing, New York.Stohlgren, T.J. & Quinn, J.F. (1991) Evaluating the contribution of
Willis, J.C. (1922) Age and area: a study in geographical distributionnorthern California National Park areas to regional biodiversity.
and origin of species. Cambridge University Press, Cambridge.Proceedings of the Symposium on Biodiversity of Northwestern
California, 226–232.
Tilman, D. (1988) Plant strategies and the dynamics and structure of
plant communities. Princeton University Press, Princeton. BIOSKETCHES
Tobler, W.R. (1970) A computer movie simulating urban growth in
the Detroit region. Econ. Geogr. 46, 234–240.
Vuilleumier, F. (1970) Insular biogeography in continental regions. 1. Jeffrey C. Nekola received his Ph.D. in Ecology from the
The northern Andes of South America. Am. Nat. 104, 378–388. University of North Carolina at Chapel Hill in 1994. His
Webster, R. & Oliver, M.A. (1990) Statistical methods in soil and land previous work includes the documentation of the effect of
resource survey. Oxford University Press, Oxford. colonization history on community composition and the
Weddell, B.J. (1991) Distribution and movement of Columbian ground spatial distribution of rare vascular plant and butterfly
squirrels (Spermophilus columbianus (Ord)): are habitat patches like species within and between isolated habitats. He is
islands? J. Biogeogr. 18, 385–394. currently investigating species-rich land snail communities
Weins, J.A. (1989) Spatial scaling in ecology. Funct. Ecol. 3, 385–397. at scales ranging from 1 to 10,000 m.
Werner, P.A. & Platt, W.J. (1976) Ecological relationships of co- Peter S. White is Professor, Department of Biology,
occurring Goldenrods (Solidago: Compositae). Am. Nat. 110, 959– University of North Carolina at Chapel Hill and Director
971.
of the North Carolina Botanical Garden. His work ranges
White, P.S. (1982) The flora of the Great Smoky Mountains National
from studies of biogeography, rare species, exotic species
Park: an annotated checklist of the vascular plants and a review of
invasions, and species richness to vegetation dynamics and
previous floristic work. USDI, National Park Service, Southeast
disturbance.
Regional Office, Res./Resource Manage. Rept. SER-55.
White, P.S. & Miller, R.I. (1988) Topographic models of vascular plant
© Blackwell Science Ltd 199926867–878
161
Section V: Ecological Theory and Modeling
Without the context provided by theory, ecological research may become bogged down in trivial
detail. However, pure theory may easily become untethered from reality unless it is constantly
informed by real-world observations. For this reason, I have chosen actively pursue theoretical
issues while also maintaining an active research program spanning multiple frames of reference
across multiple taxa groups. There are a number of biodiversity patterns which appear to be
universal across multiple scientific disciplines. For instance, rarity-enriched Species Abundance
Distributions, power-law Species Area/Time Relationships, and non-linear Distance Decay
patterns appear ubiquitous across a wide array of complex systems, including not only ecology
but also meteorology, geoscience, materials science, economics, sociology, and even the arts.
Determining the common mechanisms underlying the distribution of trees in tropical forests, the
longevity of drink tumblers in restaurants
or of song performance frequencies in
Cowboy Junkies setlists will help
ecologists and others understand what
patterns are caused by statistical
mechanics rather than disciplinary
mechanisms such as birth, death,
immigration, or competition for limiting
resources. All disciplines must begin this
search for such fundamental mechanisms
which are both more universal and less
explicitly mechanistic than those typically
sought. In this way we will be able to
define the rigid rules for life’s sonnet,
within which there is an infinity of detail.
A particular foci for my theoretical interests is consideration of the role played by spatial or
temporal distance. In particular, I am interested in applying Tobler’s First Law of Geography:
“Everything is related to everything else, but near things are more related to each other” to
biogeographic analysis and modeling. For instance, this premise can be used, via gravity models,
to predict the spread of organisms. Such models offer an important counterpoint to diffusion
models as they assume landscape heterogeneity and do not require organisms to migrate via a
moving wavefront. Gravity models provide a much more accurate portrayal of Zebra Mussel
movement through small lakes of the upper Midwest USA than are possible via diffusion
models.
My latest contribution in this general topic area derives this relationship from individual species
occurrences and then illustrates that the functional form is related to sample scale, with powerlaw
decay dominating within and exponential-decay dominating between communities. It also
documents why distance decay is a universal expectation across a wide range of physical, social,
artistic and biological systems. There are perhaps a dozen or more other diversity metrics that
likewise appear to be universal statistical expectations, with rarity-enriched species abundance
distributions and power-law species area/ time relationships serving as classic examples. As a
result, the mere existence of these patterns likely say little about specific mechanism (e.g. neutral
162
vs. deterministic community assembly) other than illustrating that the system consists of many
quasi-independent parts. Thus, before we attempt to discern underlying mechanism we must: A.
determine which patterns are simply cross-disciplinary expectations derived from statistical
mechanics and thus say little per se about explicit ecological mechanism; B. determine what
parameters of these relationships, if any (e.g. functional form or rate of change), are most likely
to reflect differences in underlying ecological mechanism; C. investigate variation in these
parameters to determine their utility in differentiating competing ecological models.
Representative Publications
[number of citations as of October 27, 2017]
Nekola, J.C. & B. McGill. 2014. Scale dependency in the functional form of the distance decay
relationship. Ecography. 37:309-320. [18]
McGill, B. & J.C. Nekola. 2010. Mechanisms in macroecology: AWOL or purloined letter?
Towards a pragmatic view of mechanism. Oikos. 119:591-603. [57]
Nekola, J.C., A.L. Šizling, A.G. Boyer & D. Storch. 2008. Artifactions in the logtransformation
of species abundance distributions. Folia Geobotanica. 43:259-468. [21]
Moses, M.E., C. Hou, W.H. Woodruff, G.B. West, J.C. Nekola, W. Zuo & J.H. Brown. 2008.
Revisiting a model of ontogenetic growth: estimating model parameters from theory and
data. The American Naturalist. 171:632-645. [98]
Nekola, J.C. & J.H. Brown. 2007. The wealth of species: ecological communities, complex
systems, and the legacy of Frank Preston. Ecology Letters. 10:188-196. [101]
Bossenbroek, J.M., C.E. Kraft & J.C. Nekola. 2001. Prediction of long-distance dispersal using
gravity models: Zebra Mussel invasion of inland lakes. Ecological Applications.
11:1178-1788. [226]
Examples of linear, exponential, and power-law distance decay showing the best-fit regression to the
mean and the upper threshold (95% percentile).
163
309
Scale dependency in the functional form of the distance
decay relationship
Jeffrey C. Nekola and Brian J. McGill
J. C. Nekola (jnekola@unm.edu), Dept of Biology, Univ. of New Mexico, Albuquerque, NM 87131, USA. – B. J. McGill, School of Biology and
Ecology, Univ. of Maine, 5751 Murray Hall, Orono, ME 04469, USA.
We examine a novel mathematical approach which posits that the decay of similarity in community composition
with increasing distance (aka distance decay) can be modeled as the sum of individual species joint-probability vs
distance relationships. Our model, supported by analyses of these curves from three datasets (North American breeding
birds, North American taiga plants, and tropical forest trees), suggest that when sampling grain is large enough to
avoid absences due to stochastic sampling effects, and/or sampling extent is large enough to generate species turnover
through the deterministic crossing of environmental and/or geographical range limits, species joint-probability over
increasing distance will generally exhibit exponential decay. However, at small scales where occurrence is driven more
by stochastic sampling effects, species joint-probability curves exhibit a power-law decay form. Lacking a theoretical
prediction of how individual species joint-probability relationships combine to generate community distance decay,
we also performed a meta-analysis of 26 ecological and 4 human-system datasets, using non-linear regression to mean
and quantile non-linear regression at tau  0.95 for linear, exponential, and power-law decay forms. These analyses
demonstrate that the functional form of community distance decay – as shown by comparison of AIC ranks – is largely
determined by observational scale, with power law decay prevailing within domains where the species pool remains
constant, while exponential decay prevails at larger scales over which the species pool varies, paralleling the patterns
predicted in our mathematical approach.
The decay of compositional similarity with increasing
inter-observation distance (aka distance decay) is a ubiquitous
pattern found across a wide array of physical,
biological, social and other complex systems (Tobler 1970,
Nekola and Brown 2007). It has garnered considerable
interest within the fields of community ecology (Condit
et  al. 2002, Green et  al. 2004, Soininen et  al. 2007),
macroecology (McKnight et  al. 2007, Blanchette et  al.
2008, Qian et  al. 2009) and conservation biology (Bell
2003, Steinitz et al. 2005), and has been suggested to shed
important insights into underlying fundamental mechanisms
(Nekola and White 1999, Hubbell 2001, Tuomisto
et al. 2003, Gilbert and Lechowicz 2004).
In spite of this interest, controversy exists regarding its
expected functional form. While a number of empirical
analyses have firmly established exponential decay (Nekola
and White 1999, Qian et al. 2005, 2009, Jobe 2007), linear
(Blanchette et  al. 2008, Perez-del-Olmo et  al. 2009) and
power law (Harte et  al. 1999, Condit et  al. 2002, Green
et  al. 2004) forms have also been documented. In spite
of this, all theoretical derivations of the distance decay relationship
(Harte and Kinzig 1997, Hubbell 2001, Chave
and Leigh 2002, Houchmandzadeh 2008, Morlon et  al.
2008) generate power-law or power-law-like forms.
Here we begin to develop analytical underpinnings for a
theory of community distance decay based on jointprobabilities
for individual species. We show that: 1) expectation
of exponential or power-law forms for individual
species joint-probability curves is a function of sampling
scale, and 2) the community distance decay curve is a sum
of single species joint-probability curves. Because a theoretical
prediction for the shape of community distance
decay curves from individual species joint-probability curves
is not analytically tractable (except in a few specific
cases), we consider this issue through empirical metaanalysis
of 26 community datasets collected over both
spatial and temporal extents, covering a large range of ecosystems
(simulated data to abyssal plain to continental
mountains, arctic tundra to tropics), taxa groups (including
mammals, birds, lepidoptera, robber flies, land snails,
vascular plants, bryophytes), sample grains (100 m2 –
subcontinental and daily-yearly) and total extent (3–9000 km
and 1–25 yr). Four human system examples are also considered
(1 spatial, 3 temporal; Nekola and Brown 2007) to
Ecography 37: 309–320, 2014
doi: 10.1111/j.1600-0587.2013.00407.x
© 2014 The Authors. Ecography © 2014 Nordic Society Oikos
Subject Editor: Jens-Christian Svenning. Accepted 11 October 2013
164
310
identify if general patterns in community distance decay
functional forms exist.
Individualistic model of distance decay
Single species distance decay
We begin with consideration with the decay of jointprobability
(Palmer 2005) for a single species/agent/event
with increasing inter-observation distance. Specifically, if
comparing two plots L and M of the same size A at distance
d apart (Fig. 1) for species i, then:
pi(dLM)  Prob{species i present at M |	
     species i present at L}	 (1)
or more generally:
pi(d)  Prob{species present at any site distance	
    d from a site where species i is present}	
(2)
While this definition has a clear relationship to the
occupancy function used by numerous authors (Harte et al.
1999, He and Legendre 2002), in these previous formulations
p has been a function of the plot area, A, rather than
the distance, d, from reference plot L. Thus, our approach is
the first to explicitly consider spatial structure.
It is clear that under biologically realistic conditions
where species have finite ranges, pi(d ) needs to fit the
following criteria: 1) pi(0)  1; 2) pi()  0; 3) pi(dAB) 
 pi(dBA) (i.e. isotropy or direction does not matter); 4)
pi(d1)   pi(d2) if d1  d2 (i.e. pi monotonically decreasing);
5) pi(d1  d2)  g(p(d1), p(d2)) for some functional form g
if d1 and d2 are linear and contiguous as shown in Fig. 1
(i.e. decomposability with distance). Note that 1, 2 and
4 combined can be used to show that 0  pi(d )  1 if
0  d  ∞.
There are an infinite number of functional forms that
fit these criteria. Three that have been commonly used
to analyze community distance decay (Nekola and White
1999, Soininen et  al. 2007) include the linear [pi(d )  
1 2 cd], exponential [pi(d )  exp(2cd)], and power law
[pi(d )  d2c]. Additionally, Harte (2011) suggests a logarithmic
decay [pi(d )  1 2 c*log(d )] form. All contain a
single parameter, c, which denotes decay rate. Note that the
power law form violates criterion 1, the linear form violates
criterion 2, while the logarithmic form violates both. As a
result, these functional forms can in practice be used only
over limited intervals: [0, dmax] where dmax  d at which
pi(d )  0 for the linear, [∈, ∞] for the power law, and [∈,
dmax] for the logarithmic. While Palmer (2005) calculated
pi(d ) for tropical trees at the La Selva Biological Station in
Costa Rica, he did not propose that it take on any particular
mathematical form. However, his plot of pi(d) across all species
and sites appears to demonstrate an exponential trace.
What form should g in criterion 5 take? This function
describes how pi should behave when a long distance d is
decomposed into two smaller distances d1 and d2 (where
d1  d2  d and lie along a straight line; Fig. 1). Clearly it
should be commutative and associative. The most obvious,
simple forms are g(x,y)  x 1 y or g(x,y)  x*y. However, the
additive form violates criterion 4 [e.g. p(6)  p(4  2)  p(4) 
1 p(2)  p(4) but p(6)  p(4)] and can easily produce pi   1.
The multiplicative form fits well with our intuitive sense of
how similarity decays [i.e. pi(d1  d2)   pi(d1)*pi(d2)], and is
essentially the law of independent probabilistic events [P(A
and B)  P(A)*P(B)]. Note that pi(d1  d2)  pi(d1)pi(d2) can
only be satisified by pi(d )  exp(cd ) with criteria 1 and 2,
requiring that c  0. A negative exponential form for the
decay of joint occurrence probability with increasing distance
thus conforms well to criteria 1–5.
It is instructive to proceed with a more rigorous
examination of the form of g. Considering the straight line
connecting L to M to N in Fig. 1, via the law of total probability
(Karlin and Taylor 1975):
P(present at N|present at L)
    P(present at N|present at M)
    3 P(present at M|present at L) 	
   1 P(present at N|NOT present at M)	
   3 P(NOT present at M| present at L)	
(3)
or:
pi(d)  pi(d2)pi(d1)  X(d2)(1 2 pi(d1))	 (4)
where: X(d) is the probability of a species being present at
one point conditioned on it NOT being present at another
point [note this is not simply 1 2 pi(d)].
More simply:
pi(d)  pi(d1)pi(d2)  Y(d1,d2)	 (5)
where Y(d1,d2)  Prob{present at N, absent at M|present
at L} or essentially the probability of encountering an
‘occurrence hole’ (sensu Rapoport 1982, Hurlbert and White
2007) within the range of species i. Y  0 when species i is
never absent at M if present at L and N. In such cases,
we have the multiplicative form for g and pi(d)  exp(2cd).
If Y ≠ 0, then g is more complex, and pi(d) is not a simple
exponential.
Community distance decay
Assuming pi(d ) is exponential for one species, then what
distance decay form should be expected across a community
assemblage? As is common in theoretical ecology, we
will work with the Sørenson similarity measure due to its
tractable mathematics (Plotkin and Muller-Landau 2002,
Morlon et  al. 2008). Let SLM be the number of species
found in both plots L and M, while SL is the number of
species found in plot L and similarly for SM. We assume
approximately S  SL  SM. because systematic changes in
richness will lead to bias in parameter estimation (Jobe
dMNdLM
L M N
dLN
Figure 1. Heuristic diagram demonstrating the relationship
between sample plots and distances used in the theoretical model.
165
311
2007). Then Sørenson (dLM)  Sørenson (L,M)  SLM/average
(SL,SM). If we sum the probability that a species is present in
M over all species present in L, this gives:
S rensonø ( ) ( )L,M p d S
S
i LMi
SL


1
1
∑
eexp( ) exp( )  

c d E c di i i
i
S
1
∑
	
(6)
where Ei is the expectation (or averaging) operator taken
with respect to i or over all species. We call this the community
aggregated distance decay function. This makes the
strong assumption that the joint probability decay is independent
between all species.
A sum of exponentials as found in Eq. 6 can take on a
vast variety of shapes, dependent entirely on the distribution
of ci values. If we treat ci as a random variable, C, from which
the ci are drawn and using the tools for calculating the
expected value of a function of a random variable (Lindgren
1976 p. 113) we have:
S rø ( ) exp( ) exp( ) ( )d E c d xd f x dxi i   ∫ 	
(7)
Where f(x) is the pdf of the random variable describing
the distribution of C. A uniform distribution for C across [0,
c] yields 1 2 [exp(2c*d ))/(c*d)]. In the limiting case
of c having a Dirac delta distribution (i.e. ci is constant
with no variation), then Eq. (6) and (7) give community
Sørenson similarity as an exponentially decaying function of
distance. Assuming C is distributed according to an exponential
distribution yields Sør(d )  c/(c  d ) while assuming
C is distributed according to a power law yields
Sør(d )  G(c)c/(bc*dc). However, as will be shown in the
results, C is strongly right-truncated, making a Beta distribution
the most likely. While this does not yield an analytically
solvable integral, a partial understanding can realized by
taking a Taylor-series expansion of Ei exp(2cid). Let C now
be the mean value of ci (C  ci
–
) and Δi  ci 2 C, with the
expansion around C:
being the occupancy rate for species i (i.e. the fraction of
sites supporting species i) or equivalently the probability
that species i is found at site L. Then we have:
S rø ( ) exp( ) exp( )d
S
w c d E w c di i i i i
i
S
   

1
1
∑
	
(9)
We call this the weighted community aggregate distance
decay function. This curve should closely approximate the
traditional community-level decay of similarity curve
which is generated by calculating pairwise similarity between
samples (Nekola and White 1999).
Model assumptions
Two main assumptions underlie our approach. First, to
produce an exponential decay of joint-probability over
increasing distance for a single species we assumed that
Y(d1,d2)  0 or equivalently that a species must occur at
M if present at L and N. In effect Y informs about the
probability that a sample plot falls inside an occurrence
hole within a species range. The value of Y should show
strong scale dependence: if sample grain is small with significant
sampling error and extent does not exceed the
environmental and/or geographic range limits of species
within the pool, then it is likely that the main reason for
compositional turnover will be sampling effects (Nekola
and White 1999). In such instances Y ≠ 0, and because in
such situations pi(d) is not exponential, this form will
also not provide a good fit to the community data. The
expected shape in such cases should initially show a more
rapid decrease in joint-probability than the exponential
expectation because of occurrence holes, but will then ultimately
asymptote  0 because the probability of a given
species within the pool is also  0. Although not a rigorous
proof, at a minimum the above demonstrates that a more
power-law like than exponential form would result from
cases where Y ≠ 0. Multiple theoretical distance decay models
assuming such small grains and uniform species pools
E c d E C d Cd E
d d
i i i i i
i i
exp( ) exp( ( ) ) exp( )
( )
!
( )
!
        ∆
∆ ∆
1
3
2 3
2
(( )
!
( )
!
∆ ∆i id d4 5
4 5
 …





 (8)
The infinite series inside the brackets is a function of d that
alternates in sign, with kn/n! converging to zero for large
enough n (where k  Δid). Thus, while the series does
converge, it occurs in a way that depends on Δi and d. In
particular if Δid is  1 over much of the range of interest
for d, the possibility exists for very slow convergence and
strong deviations from exp(2Cd). Nothing further can be
said about this situation without a detailed knowledge of
the distribution and magnitude of Δi relative to d. After
examining empirical distributions in the Results, we will
present further analysis in the Discussion as to when convergence
on exp(2Cd) may be expected.
Equation 6 can be expanded for the case where the
evaluated species pool is not just the species present at L but
includes all taxa in the regional pool. To do this, we redefine
S to be the richness of the regional species pool, with wi
have all produced power law decay forms (Harte and Kinzig
1997, Chave and Leigh 2002, Houchmandzadeh 2008,
Morlon et al. 2008).
When sample grain is large relative to the organism/agent/
event (making for guaranteed detection), with extent being
large enough to pass beyond species environmental and/or
geographic range limits (allowing the community pool to
vary across the sample), then the main cause of species turnover
will be the exceeding of species distributional limits,
minimizing the chance that a species will be absent at M
but present at L and N. In these situations Y  0, and the
exponential form should represent a good approximation.
Note that these conditions may be met at various absolute
scales. For instance, if strong local spatial or temporal gradients
exist, even over very limited extents the community pool
will not be uniform and the exponential form should occur.
166
312
detection error and year-on-year variability, we took average
abundance over the years 2003–2007, and used only the
1430 routes that had a quality code reflecting appropriate
weather and observer ability for all five years. Thus, a bird
was counted as being present at a given route if it appeared
at least once in five years. The third dataset was the Barro
Colorado Island tree plot containing 225 species (Condit
et al. 2000). Every tree  10 cm DBH was recorded within
each of the 50 one hectare plots. Distance was recorded
between hectare midpoints.
Due to their continental extents, we a priori expect
the LaRoi and BBS datasets to have individual site compositions
drawn from varying species pools. Decay of individual
species joint-probabilities in these data should thus
generally take on an exponential form. However, because the
BCI dataset covers only a 1 km extent, all samples are likely
to be drawn from only two nested species pools (young
within old forest; Svenning et al. 2004). In this situation,
species absences are likely due to stochastic sampling
effects and/or disturbance history, with a power-law like
decay-form prevailing. In all three cases the community
distance decay curve should be the sum of the individual
species decay curves, although the fit should be worse for the
BCI data because of the greater impact of stochastic sampling
errors. While no explicit prediction can be made
for the functional form of the community distance decay
curve, the heuristic considerations given above suggest that
an exponential form may be more likely at large grains
and/or extents which sample across varying species pools.
We also predict the weighted community aggregate curve
should perform better for the La Roi and BBS data since
they have large enough extents/strong enough gradients to
exceed the limits of individual species, while conversely the
BCI data should be better fit by an unweighted curve.
To test the individual species portion of the model (prediction
1), for each dataset we fit an exponential decay
curve to the each species joint-probability by calculating
similarity (1 if the species was present in both plots; 0 of
not) and distance between each occupied site and all
others. Because calculations were only conducted from
occupied sites, we constrained similarity to equal 1 at the
origin and fit only a single parameter to the decay function.
Other functional forms (e.g. linear, power law, logarithmic)
were not fit because of their violations of criteria 1 and/or
2. We then calculated mean similarity across distance bins
of 200 km (LaRoi and BBS) or 200 m (BCI). Due to the
computational challenges in the 1403 site   600 species
BBS data (reflecting a total potential number of ~ 6  108
calculations), prior to model fit we randomly removed
absent sites in order to generate occupancy rates of approximately
10%. We then sampled 10 000 individuals from
all site by site pairs. We recorded the c and r2 value for
each species. To calculate the community aggregated curve
(Eq. 6), we predicted values every 200 km or 200 m, took
their average, and plotted a line through the data. We calculated
the weighted community aggregate curve (Eq. 7)
similarly except that each species was weighted by its
observed occupancy. Because singleton and doubleton species
do not possess enough degrees of freedom to accurately
fit an exponential model (singletons are points in space
while doubletons can only demonstrate linear patterns), we
The second major assumption is that the decay of each
species joint-probability is independent of that exhibited by
others. This allows summing of individual decay curves
to produce Eq. 6. Such an assumption of independence
between co-occurring species is one of the foundational
bases of modern community ecology (Gleason 1926, Curtis
1955, Whittaker 1975) that accurately reproduces a number
of community patterns (McGill 2010, 2011, He and
Legendre 2002, Green and Plotkin 2007). Although species
pairs can be identified which clearly show non-independent
distributions (e.g. red bellied woodpecker and red-headed
woodpecker – Root 1989; American beach and its obligate
parasite beachdrops – Voss 1985), these interactions represent
such a small fraction of all possible pairwise interactions
(Barker and Mayhill 1999) that the assumption of independence
is a valid approximation.
Summary of model predictions
1) Individual species will exhibit exponential decay in their
joint-probability with increasing distance if and only if
observational scale is large enough to allow a varying species
pool and to avoid quasi-stochastic occurrence holes.
2) The Sørenson decay of community similarity is equivalent
to the occupancy-weighted community average of the
individual pi(d) curves in the regional species pool.
3) The weighted community aggregate curve (Eq. 8) will
fit better than the unweighted curve in cases where species
are not equally likely across the species pool (or equivalently
that the species pool changes over the extent studied), but
will fit worse when the species pool is constant across sites.
4) The functional form for community distance decay
[Sør(d)] depends on the distribution of ci, the decay rates
of for individual joint-probability curves. Unfortunately
this distribution has never been studied and we have no
empirical knowledge of its shape, making impossible the
prediction of community distance decay functional form.
Even when all single species joint-probability curves are
exponential, community decay curves of innumerable forms
can be produced.
Methods
Test of model predictions
Three different datasets were used to test the general accuracy
of our approach. The first is the vascular plant data
of LaRoi (1967), which was sampled from 34 nine-ha
(300  300 m) mature, undisturbed upland white spruce
forest plots at relatively equal distances across boreal North
America. A total of 220 vascular plant species were encountered,
with individual plot richness ranging from 19–56.
The second is the North American Breeding Bird Survey
(BBS), which is run annually at over 2000 routes across the
continental US and southern Canada (Sauer et al. 2011). A
3-min point count is taken every half mile (approximately
0.8 km) for 50 stops along a transect of 25 miles (approximately
40 km). An individual route typically has 50–100
bird species detected and overall almost 600 species of
birds have been observed at least once. To avoid issues with
167
313
Preserve, Linn County, Iowa (F. Olsen pers comm.); and
seasonal robberfly emergences at the Sevilleta LTER
(H. Lease pers. comm.). Representative human-system
datasets include: concert setlists for Cowboy Junkies
( www.setlist.com ) and Phish ( www.ihoz.com/
PhishStats.html ) performances; all commercially available
garden vegetable varieties in the US and Canada as
reported by the Garden Seed Inventory of the Seed Savers
Exchange; all ingredients reported for the cuisines of
Ethiopia, Hungary, India, Iran, Ireland, Korea, Mexico,
Norway, Puerto Rico, and Thailand (Smith 1990). The
Cowboy Junkies setlist, Garden Seed Inventory, and global
cuisine distance decay analyses were previously reported in
Nekola and Brown (2007).
For each of these datasets, Jaccard similarity was
calculated between each pairwise combination of samples,
while distance was calculated as the linear difference in
space or time between observations. The Jaccard metric was
employed because of its common use in distance decay
analyses (Nekola and White 1999). The Jaccard and
Sørenson indices differ only slightly in their denominator
values, and perform in a highly similar (and for all intents,
identical) fashion. These data were then subjected to
non-linear regression to the mean using the ‘nls’ routine in
R, and quantile regression using the ‘nlrq’ routine in R
with tau  0.95. Quantile regression was used because the
central tendency may not be as important in the distance
decay relationship as the shape of the upper bound
(Rocchini and Cade 2008). For each of these regression
approaches three different functional forms were fit to the
data: Sd  Si 2 cd (linear decay); Sd  Si e2cd (exponential
decay) and Sd  Ad2c (power-law decay) where Sd  intersample
similarity at distance d; Si  initial similarity or
nugget (Nekola and White 1999); A  arbitrary intercept;
d  intersample distance; c  rate of similarity change.
We did not fit logarithmic decay [Sd  a 2 c(log(d))]
because we found it to neither outperform power-law fits
in the three intensively studied datasets (LaRoi, BBS, BCI)
nor in those data which appeared by eye to most likely to
express the logarithmic functional form (Tallgrass Prairie
Preserve, Iowa Butterfly County occurrences, Portal
rodents, and Cowboy Junkes setlists).
AIC values were calculated for each regression and ranked
from 1 (least) to 3 (largest) across the three functional forms
for each dataset. We then constructed a contingency table
for each of the regression methods, showing 1st, 2nd, and
3rd AIC ranks vs linear, exponential, and power-law fits.
A Fisher-exact test was conducted on these tables to determine
if cell frequency violated null expectations.
Results
Model test
All three model predictions were confirmed across the three
datasets. In general, the exponential decay form was found
to be a good approximation for taiga vascular plant and
continental bird species joint-probabilities with increasing
observational distance (Fig. 2). While the vast majority of
LaRoi and BBS species were fit by an exponential model
also calculated a community aggregate curve in which each
species was equally weighted (wi  1/S) after removal of
singletons and doubletons. We term this the filtered community
aggregate curve.
The traditional community distance decay curve was also
calculated for each dataset. Because of its use in model formulation,
Sørenson similarity was used. Since the total
BBS data would represent almost 2 000 000 pairwise comparisons,
we limited analysis to 10 000 randomly-drawn
pairs. To these data we then fit a linear decay curve using
ordinary least squares (OLS) regression and an exponential
and power-law decay curves using non-linear least squares
(NLS) regression. Note that under the assumption of normal
errors, OLS and NLS give maximum likelihood estimates
allowing the use of the Akiake information criteria (AIC),
which we used to compare the three models. We also plotted
a LOESS line on these plots to allow for easy visual comparison
of the central tendency in the empirical data vs the
community aggregated curves.
Meta-analysis of community similarity datasets
Lacking a theoretical prediction for the functional form of
community distance decay, we also conducted a metaanalysis
of 26 ecological and 4 human-system datasets. Two
datasets represent simulated community turnover data using
COMPAS (Minchin 1987), with composition lists being
generated for 50 randomly-spaced sites along a single gradient
containing 20 species. Two different turnover scenarios
were used: constant turnover where all species possess the
same modal abundance, standard deviation, and are equally
spaced along the gradient, and lognormal turnover, where
species modes, standard deviations, and gradient placement
follow a lognormal distribution. The remaining data are
empirical, represent examples of both spatial and temporal
decay, and attempt to maximize ecosystem range, taxa
groups, sample grains, and extents. Information regarding
the habitat type, taxa group, maximum extent, sample grain,
number of samples, and additional notes on these datasets is
found in Table 2. These data include: Galapagos Islands
vascular plant flora (Preston 1962); North America upland
taiga vascular plants (LaRoi 1967) and bryophytes
(LaRoi and Stringer 1976); fen-restricted vascular plants in
northeastern Iowa (Nekola 1994); Martha’s Head to
Bermuda abyssal plain bivalves (Allen and Sanders 1996);
Belize terra firma tropical forest (Bird 1998); entire vascular
plant floras for all algific talus slope sites along Buck Creek,
Clayton County, Iowa (Nekola 1999); Peruvian and
Panamanian terra firma forest (Condit et al. 2002); Barro
Colorado Island terra firma forest (Condit et  al. 2000);
New Zealand land snails (Barker 2005); eastern North
American and Asian vascular plants (Qian et al. 2005); Iowa
county butterfly faunas (Schlicht et al. 2007); small rodent
and winter annuals plants at Portal, Arizona (Ernest et al.
2009); eastern North American rock outcrop, Atlantic
coastal plain acidophile, and southwestern USA montane
land snail faunas (Nekola 2010); North American Breeding
Bird Survey (Sauer et al. 2011); southern Flint Hills tallgrass
prairie vascular plants (McGlinn and Palmer 2011);
seasonal butterfly and skipper emergences at the Rock Island
168
314
parameters in the weighted community aggregate curve
were adjusted to make them fit the true Sørenson decay
curve. Although no community aggregate curve provides a
good fit to the empirical BCI community decay data, the
unweighted curve is a better fit, confirming prediction 3.
Linear, exponential and power law forms were then fit to the
empirical Sørenson community distance decay (Fig. 4).
Community distance decay in the LaRoi and BBS data are
best fit by the exponential form, while the BCI data is best
fit by a power law (Table 1).
Meta-analyses
All 26 ecological and 4 human-system datasets demonstrated
a negative relationship between community similarity and
with r2  0.9, a noticeably higher frequency of poor fits
were noted in the BCI data (Fig. 3). These results confirm
prediction 1. It should be noted, however, that six BBS
species (1% of total), present at only a handful of sites,
demonstrated increasing similarity with distance. These
were removed from subsequent analysis. The distribution of
ci values was unimodal with a strong right truncation,
making exponential and power law distributions for C a
poor description of the empirical data (Fig. 3). Visual
comparison of the weighted community aggregate vs the
LOESS line for the true Sørenson decay (Fig. 4) shows a
remarkable convergence in the LaRoi and BBS data, confirming
prediction 2. Table 1 also demonstrates that the
community weighted aggregate produces exponential decay
parameters that are very close to the observed true community
decay values. This is quite remarkable given that no
0 5000
0
0.5
1
Sorenson
Viburdum edule
0 1000 2000
0
0.5
1
Sorenson
d (km)
Arctostaphylos rubra
0 1000 2000 3000
0
0.5
1
Sorenson
Dickcissel
0 1000 2000 3000
0
0.5
1
Sorenson
d (km)
Scissor-tail flycatcher
0 500 1000
0
0.5
1
Sorenson
Acalypha macrostachya
0 500 1000
0
0.5
1
Sorenson
d (m)
Adelia trioba
Figure 2. Individual species joint-probability decay with distance [pi(d )] curves for arbitrary species. Blue dots are observed similarities.
Red line is LOESS smoothed regression. Left column represents two species from the LaRoi dataset. Middle column represents two species
from the Breeding Bird Survey dataset. Right column represents two species from the Barro Colorado Island dataset. Note that for a few
species (like Viburnum edule) exponential decay is a poor fit. However, as seen in the bottom row of Fig. 3 there only a few such species.
–5 0 5
0
50
100
LaRoi
Ci Ci Ci
#spec
–5 0 5
x 10–3x 10–3
0
50
100
150
BBS
#spec
–0.01 0 0.01
0
50
100
150
BCI
#spec
0 0.5 1
0
50
100
150
r2
#spec
0 0.5 1
0
200
400
r2
#spec
0 0.5 1
0
50
100
r2
#spec
Figure 3. Results of exponential fits to individual species joint-probability occurrence decay. Top row shows distribution of estimated
coefficents, ci. Bottom row shows a histogram of r2 values for these fits. Left column is for LaRoi data, middle column is for the Breeding
Bird Survey, and the right column is for the Barro Colorado Island data.
169
315
distance (Table 2). Based on AIC ranks, each of the three
functional forms were found to provide a best-fit to at least
some of the ecological datasets under both non-linear
regression to mean and quantile regression to tau  0.95.
However, in all four human-system datasets the power-law
model was best fit to the data, followed by the exponential
model, with the linear model providing the poorest fit. For
ecological systems exponential decay was best fit 14 times to
the mean, and 13 times to tau  0.95 (Table 3). Powerlaw
forms were best fit in 8 and 6 cases, respectively, while
linear was the best fit in 4 and 7 cases. While exponential
equations never provided the poorest fits, in almost half of
the datasets linear fits performed the most poorly (13 and
10, respectively), while power-law fits performed worst
in at least 50% of datasets (13 and 16, respectively). Fisherexact
tests on these contingency tables indicate a highly
significant (p  0.00002 and p  0.000014) deviation of
cell frequencies from a uniform null model.
Discussion
These analyses clearly demonstrate that the distance
decay of community similarity reflects the aggregate of
individual species joint-probability decay functions across
the species pool. As shown in both large-scale detailed
datasets (BBS, LaRoi), when the grain size and/or extent
are large enough to generate exponential individual jointprobability
decay, the community decay curve is also best
fit by this form. This outcome is replicated in the metaanalysis,
where large scale samples crossing into multiple
species pools also exhibited exponential community distance
decay. Examples include: simulated communities
along a strong environmental gradient; vascular plant
flora of the Galapagos archipelago; vascular plant and bryophyte
assemblages across boreal North America; vascular
plants of North American and Asian temperate forests;
North American Atlantic coastal plain land snails; BBS
data; and the adult emergence of both butterflies/skippers
and robberflies along a seasonal gradient.
0 2000 4000 6000
0
0.5
1
Distance (km)
Sorenson
LaRoi(A)
(B)
(C)
0 2000 4000 6000 8000
0
0.5
1
Distance (km)
Sorenson
BBS
0 200 400 600 800 1000
0.4
0.6
0.8
1
Distance (km)
Sorenson
BCI
Raw Data
LOESS
Exponential
Power
CommAgg
CommAgg Wtd
CommAgg Filt
Figure 4. Empirical distance decay relationships, with each
blue dot representing a plot pair. Red solid line shows LOESS fit.
Red dash-dotted red line shows an exponential fit. Red dotted line
shows a power law fit. The green lines show the community aggregated
curves. The solid green line equally considers all species
(i.e. Eq. 6). The dotted green line weights species by their occupancy
(i.e. Eq. 7). The dash-dotted line includes all species equally
but eliminates species found in 2 or fewer sites. (A) LaRoi boreal
plant data. (B) North American Breeding Bird Survey (note only
10 000 random pairs were used out of the over 2 000 000 possible).
(C) Barro Colorado Island data.
Table 1. Summary of results for three datasets.
True Community aggregate
Weighted community
aggregate
C AIC C AIC C AIC
LaRoi
Linear 20.0001 22578.8 20.00012 2125.4 20.0001 2142.4
Exponential 20.00027 22590.4 20.00056 2183.5 20.00032 2185.1
Power 20.286 22355.1 20.482 2147.0 20.353 2149.4
BBS
Linear 20.00011 238294 20.00007 2156.0 20.00008 2167.5
Exponential 20.00039 239287 20.00047 2202.3 20.00034 2215.1
Power 20.370 237635 20.484 2211.5 20.409 2199.0
BCI
Linear 20.00006 27646 20.00039 231.0 20.00021 243.2
Exponential 20.00009 27647 20.00059 234.0 20.00024 245.4
Power 20.0367 27677 20.202 239.6 20.085 240.9
170
316
Table2.Source,system,samplescale,andregressioninformationfortheanalyzeddatasets.TheupperAICvaluesandranksrepresentstatisticsgeneratedfromnon-linearregressiontomean,whilethe
lowervaluesrepresentstatisticsgeneratedfromquantilenon-linearregressionfortau  0.95.
No.
samples
AICvalue/rankorder
SourceLocationHabitatTaxagroupMaxExtentGrainNotesLinearExponentialPower
COMPASSimulated50Constantturnover23665.7/2
4730.4/2
23956.1/1
3837.2/1
21478.3/3
5430.9/3
COMPASSimulated50Lognormalturnover21317.6/2
7018.5/2
21976.0/1
6937.4/1
2721.0/3
7372.3/3
Preston1962Galapagos
Islands
Volcanic
archipelago
Vascular
plants
286.2km1.1–4640km2182222.7/2
51.0/1
2224.6/1
51.6/2
2206.7/3
61.5/3
LaRoi1967North
America
UplandtaigaVascular
plants
5687.4km9ha3421212.6/3
1770.0/2
21244.1/1
1729.3/1
21027.2/2
2116.4/3
LaRoiandStringer1976North
America
UplandtaigaBryophytes5687.4km9ha3421205.1/2
1810.7/2
21221.7/1
1804.4/1
21122.2/3
1919.9/3
Nekola1994Northeastern
Iowa
FenSpecialist
vascular
plants
254.0km0.1–50.1ha10291.1/1
291.7/1
291.0/2
290.7/2
289.7/3
287.4/3
AllenandSanders1996Atlantic
abyssal
plain
TotalfaunaProtobranch
bivalves
777.5kmca0.1ha21 3000mdepthonly2286.8/1
498.3/1
2274.4/2
501.9/2
2208.7/3
522.6/3
Bird1998BelizeTropicalforestCanopy
trees
205.5km1ha302798.2/3
1290.1/3
2810.0/2
1282.6/2
2947.6/1
1068.2/1
Nekola1999BuckCreek
valley,Iowa
Algifictalus
slope
Vascular
plants
3.1km0.125–1ha7283.1/1
2109.2/1
282.9/2
2109.1/2
277.6/3
2107.5/3
Conditet al.2000Barro
Colorado
Island
TropicalforestTrees  
10cmdbh
1km1ha5024167.7/3
213446/3
24168.8/2
213448/2
24199.0/1
213496/1
Conditet al.2002PanamaTropicalforestCanopy
trees
110.6km1ha10029045.8/3
38969.4/3
211850/1
37667.5/1
211125/2
37820.6/2
Conditet al.2002PeruTropicalforestCanopy
trees
102.2km1ha152275.4/3
2105.6/3
2276.1/2
2107.8/2
2305.8/1
2136.2/1
Barker2005NewZealandTotalfaunaLandsnails1590.9km282460.6/2
1197.3/1
2463.0/1
1205.8/2
2416.5/3
1371.6/3
Qianet al.2005ENorth
America
TotalfloraVascular
plants
1895.3km67000–249000km22547chosen
comparisons
264.6/2
2107.2/1
267.0/1
2106.6/2
264.3/3
238.3/3
Qianet al.2005EAsiaTotalfloraVascular
plants
1870.6km67000–249000km21840chosen
comparisons
261.4/3
293.0/2
276.2/1
2116.9/1
270.9/2
210.2/3
Schlichtet al.2007IowaCounty
faunas
Butterflies
and
skippers
502.1km1042–2524km260Limitedtowell
sampled(40  taxa)
counties
23973.9/3
9840.0/3
23975.8/2
9839.1/2
23982.2/1
9838.4/1
Ernestet al.2009Portal,
Arizona
Desertscrub
and
grassland
Winter
annual
plants
16yrSeasonsummaryfor250m217Plotno.22from
1989–2005
10.4/3
264.7/3
10.3/2
265.3/1
9.7/1
265.4/2
Ernestet al.2009Portal,
Arizona
Desertscrub
and
grassland
Rodents25yrSeasonsummaryfor250m226Plotno.22from
1976–2005
2301.9/3
1227.5/3
2310.9/2
1219.0/2
2326.0/1
1197.8/1
(Continued)
171
317
Table2.(Continued).
No.
samples
AICvalue/rankorder
SourceLocationHabitatTaxagroupMaxExtentGrainNotesLinearExponentialPower
Nekola2010USAtlantic
Coast
AcidhabitatsLandsnails2013km10 000m2152147.1/2
70.9/2
2148.5/1
69.9/1
2146.5/3
72.0/3
Nekola2010ENorth
America
RockoutcropLandsnails3031.710 000m24821269.0/2
6446.9/1
21297.1/1
6462.6/2
21225.6/3
6553.0/3
Nekola2010SWNorth
America
Montane
forest
Landsnails841.1km10 000m2142104.9/1
26.4/2
2102.9/2
26.3/1
295.7/3
41.0/3
McGlinnandPalmer2011OklahomaTallgrass
prairie
Vascular
plants
15.6km100m2151Comparisonswithplot
no.5only
2358.0/3
7.8/3
2359.8/2
5.6/2
2367.5/1
211.0/1
McGlinnandPalmer2011OklahomaTallgrass
prairie
Vascular
plants
9yrSeasonsummaryfor100m210Plotno.5from
1997–2006
2149.5/3
2156.7/2
2151.0/2
2157.2/1
2155.6/1
2152.2/3
Saueret al.2011North
America
Totalfauna
2002–2007
Breeding
birds
6139.8km~ 16ha~ 37001000drawsfrom
allpairwise
comparisons
2968.9/2
28687.5/2
21084.8/1
28698.1/1
2878.4/3
28405.5/3
H.Leasepers.comm.Sevielleta
LTER
DesertscrubRobberflies193d1d262006season2173.9/3
1289.9/3
2231.7/1
1269.6/1
2227.5/2
1275.2/2
F.Olsenpers.comm.RockIsland
Preserve,
Iowa
Tallgrass
prairie
Butterflies
and
skippers
174d2–4h381990season2713.1/3
3301.1/3
2741.9/1
3244.2/1
2718.3/2
3277.2/2
NekolaandBrown2007Concert
setlists
Cowboy
Junkies
6856d1show331987to20062894.8/3
24559.7/3
2921.3/2
24657.3/2
21047.1/1
24910.7/1
www.ihoz.com/
PhishStats.html
Concert
setlists
Phish6070d1show181983–20002485.5/3
246.5/3
2511.5/2
251.0/2
2525.5/1
269.9/1
NekolaandBrown2007Commercially
available
varieties
GardenSeed
Inventory
23yr1GardenSeedInventory71981to2004240.4/3
2183.4/3
250.4/2
2199.4/2
254.4/1
2231.6/1
NekolaandBrown2007Ingredient
lists
Global
cusines
16500km1ethnicregion102126.1/3
2453.2/3
2128.5/2
2454.7/2
2136.6/1
2456.3/1
172
318
domesticates now feature prominently in new world cuisines –
e.g. wheat, cinnamon, anise, chickens and pigs in Mexico.
The existence of public and private seed banks has also
made it possible for ‘lost’ commercial varieties to be made
available after absences of a few decades. And, it is also highly
unlikely that either the Cowboy Junkies or Phish would be
unable to play any song in their repertoire given more than a
few days’ rehearsal. As a result, agents/events within the
analyzed human systems tend to be drawn from a uniform
pool, with absences being largely the result of occurrence
holes or stochastic sampling effects. Such situations should –
and do – lead to power-law community distance decay.
While we are unable to analytically prove that exponential
community distance decay should be expected in situations
where individual species joint-probability curves are
also largely exponential, consideration of ci values for
LaRoi and BBS in conjunction with Eq. 8 (the Taylor series
expansion), helps explain this correspondence. Recall that
convergence to exp(2Cd) is dependent upon how fast
(Δid)n/n! converges to zero as n increases. Our empirical
data (Fig. 3) indicate that the distribution of Δi (or equivalently,
ci) is strongly truncated on both tails, making the
values of ci strongly centered around the mean. As a result ci
(or more specifically Δi 5 ci 2 ci) is never larger than a small
multiple of ci (e.g. 2–3 times). Second we note that ci
is order-of-magnitude close to but slightly larger than 1/d
such that cid is slightly greater than 1 (i.e. approximately in
the interval [1,10] for most of the values of d in the decay
curve graphs). Since, by tail truncation, Δi is never larger
than 2ci or 3ci, then Δid also must fall in the range of
roughly 1–10. Additionally note that (Δid)n/n! converges
on zero quite quickly when Δid is in the range [1,10]. Also
note Δid is less than 1 when d is close to zero, which
also forces a fast convergence on zero. As a result, the Taylor
series expansion necessitates a community distance decay
function close to exp(2Cd) when species joint-probability
occurrence is exponential and the distribution of ci values is
strongly truncated and order of magnitude close to 1/d.
While providing a numerically based understanding for
the preponderance of observed exponential-like community
distance decay, this analysis can not be construed as a general
proof, as the end result is strongly dependent upon the
form of the joint-probability decay curve and ci distribution
properties.
These analyses also document two additional important
additional insights regarding the nature of ecological
communities. First, it is striking how similar the parameterized
and predicted coefficients are between the true and
weighted aggregate community distance decay function for
large-scale data (Fig. 4; Table 1). This correspondence provides
strong evidence of individualistic species sorting
(Gleason 1926). Second, the different expected forms for
distance decay within (power law) vs between (exponential)
communities allows for an empirical test of the observational
scales under which communities exist. For instance,
in Panamanian tropical forest, strong power-law decay
apparent at distances of 1 km or less indicate that samples
are being drawn from within the same community.
However, the exponential decay observed at distances of
5–100 km indicates that at this larger scale different communities
are being surveyed.
However when community data is sampled at small grains
or within a mostly uniform species pool at limited extents
(BCI), exponential decay is a poorer describer of jointprobability
decay with community distance decay exhibiting
a power-law like form. This result was also replicated in
the meta-analysis. Examples include: vascular plants within
100 m2 quadrats along a 15.6 km extent in the Tallgrass
Prairie Preserve in northeastern Oklahoma; canopy trees
within 1 ha tropical forest stands along a 1 km extent at BCI,
a 100 km extent in lowland Peru, and a 200 km extent in
Belize; seasonal summaries for rodents within a 250 m2
sample over a 25 yr extent in Portal, Arizona. While Iowa
county butterfly faunas would appear a violation, with power-law
decay being favored even though ~ 25% of the fauna
reaching its range limit within the state boundary, it is
also important to realize that range-limit species also tend to
be rare. Since rare species do not exhibit a distance decay
signal (Nekola and White 1999), they will not contribute
greatly to the overall shape of the decay function. As a result,
the observed power law shape is largely due to the ~ 75% of
the fauna which is cosmopolitan.
Power-law community distance decay has also previously
been shown to describe the variation of vascular plant composition
at  10 km extents within Colorado subalpine
meadows (Harte et  al. 1999) and for soil microbes over
~ 1000 km extents in Australia (Green et  al. 2004). It is
important to note in this latter example that because of
their exemplary long-distance dispersal abilities (Green and
Bohannan 2006) local microbe communities are likely
drawn from a much more extensive and uniform species
pool than would be the case for most macroscopic organism
groups.
The predominance of power-law community distance
decay in human-systems is also readily explained using this
framework. Because of the immense human capacity to
move ideas and goods across large spatial scales, and to store
and retrieve data across long time periods, absolute range
limits for commercial goods or ideas are likely not crossed at
either planetary spatial extents or decadal time spans. For
instance, global trade over the last 600 yr has allowed essentially
all domesticated crops to be available world wide. As
a result, many new world domesticates are now common
aspects in old world cuisines – e.g. tomatoes in Italian,
chiles in Indian/Ethiopian/Chinese/Thai, while old-world
Table 3. Contingency table of AIC rank vs functional form of the
distance decay relationship. The smallest AIC value among the three
functional forms was assigned a rank of ‘1’, the second smallest a
rank of ‘2’, with the largest being assigned a rank of ‘3’. The first
number in a cell represents the number of datasets achieving that
given rank based on AIC values calculated from non-linear regression
to mean, while the second represents the number achieving
that rank based on quantile non-linear regression for tau  0.95.
Mathematical distance decay form
AIC rank Linear Exponential Power law
1 4/7 14/13 8/6
2 9/9 12/13 5/4
3 13/10 0/0 13/16
Fisher exact test for identical cell frequencies: p  0.00001966
for non-linear regression to mean; p  0.00001374 for quantile
non-linear regression.
173
319
Green, J. L. and Plotkin, J. B. 2007. A statistical theory for sampling
species abundances. – Ecol. Lett. 10: 1037–1045.
Green, J. L. et  al. 2004. Spatial scaling of microbial eukaryote
diversity. – Nature 432: 747–750.
Harte, J. 2011. Maximum entropy and ecology. – Oxford
Univ. Press.
Harte, J. and Kinzig, A. P. 1997. On the implications of
species–area relationships for endemism, spatial turnover, and
food web patterns. – Oikos 80: 417–427.
Harte, J. et al. 1999. Estimating species–area relationships from
plot to landscape scale using species-turnover data. – Oikos
86: 45–54.
He, F. and Legendre, P. 2002. Species diversity patterns derived
from species–area models. – Ecology 83: 1185–1198.
Houchmandzadeh, B. 2008. Neutral clustering in a simple experimental
ecological community. – Phys. Rev. Lett. 101: 078103.
Hubbell, S. P. 2001. The unified neutral theory of biodiversity
and biogeography. Monographs in population biology no. 32.
– Princeton Univ. Press.
Hurlbert, A. H. and White, E. P. 2007. Ecological correlates
of geographical range occupancy in North American birds.
– Global Ecol. Biogeogr. 16: 764–773.
Jobe, R. T. 2007. Estimating landscape-scale species richness:
reconciling frequency- and turnover-based approaches.
– Ecology 89: 174–182.
Karlin, S. and Taylor, H. M. 1975. A first course in stochastic
processes, 2nd ed. – Academic Press.
LaRoi, G. H. 1967. Ecological studies in the boreal sprucefir
forests of the North American taiga. I. Analysis of the
vascular flora. – Ecol. Monogr. 37: 229–253.
LaRoi, G. H. and Stringer, M. H. 1976. Ecological studies in the
boreal spruce-fir forests of the North American taiga. II.
Analysis of the bryophyte flora. – Can. J. Bot. 54: 619–643.
Lindgren, B. W. 1976. Statistical theory. – Macmillan Publishing.
McGill, B. J. 2010. Towards a unification of unified theories of
biodiversity. – Ecol. Lett. 13: 627–642.
McGill, B. J. 2011. Linking biodiversity patterns by autocorrelated
random sampling. – Am. J. Bot. 98: 481–502.
McGlinn, D. J. and Palmer, M. W. 2011. Quantifying the influence
of environmental texture on the rate of species turnover:
evidence from two habitats. – Plant Ecol. 212: 495–506.
McKnight, M. W. et al. 2007. Putting beta-diversity on the map:
broad scale congruence and coincidence in the extremes.
– PLoS Biol. 5: 2424–2432.
Minchin, P. R. 1987. Simulation of multidimensional community
patterns: towards a comprehensive model. – Plant Ecol.
71: 145–156.
Morlon, H. et al. 2008. A general framework for distance-decay of
similarity in ecological communities. – Ecol. Lett. 11: 904–917.
Nekola, J. C. 1994. Environment and vascular flora of northeastern
Iowa fen communities. – Rhodora 96: 121–169.
Nekola, J. C. 1999. Paleorefugia and neorefugia: the influence
of colonization history on community pattern and process.
– Ecology 80: 2459–2473.
Nekola, J. C. 2010. Acidophilic terrestrial gastropod communities
of North America. – J. Molluscan Stud. 76: 144–156.
Nekola, J. C. and White, P. S. 1999. Distance decay of
similarity in biogeography and ecology. – J. Biogeogr. 26:
867–878.
Nekola, J. C. and Brown, J. H. 2007. The wealth of species:
ecological communities, complex systems, and the legacy
of Frank Preston. – Ecol. Lett. 10: 188–196.
Palmer, M. W. 2005. Distance decay in an old-growth neotropical
forest. – J. Veg. Sci. 16: 161–166.
Perez-del-Olmo, A. et al. 2009. Not everything is everywhere: the
distance decay of similarity in a marine host–parasite system.
– J. Biogeogr. 36: 200–209.
Since its popularization within ecological research over a
decade ago, there has been a mismatch between theoretical
models and empirical results of distance decay analysis. We
hope the approach described within this paper helps resolve
this conundrum. Specifically, sampling scales associated
with small grain, limited extents, and absent/limited
environmental gradients are expected to have power-law
community distance decay, while sampling scales associated
with large grain and strong environmental gradients are
expected to have exponential community distance decay.
Acknowledgements – The following graciously provided us with their
raw datasets for analysis: Gary Barker (New Zealand land snail
data); Matt Barthel (Garden Seed Inventory data); Neil Bird (Belize
tropical forest data); Jim Brown, Morgan Ernst and Tom Valone
(Portal rodent and vascular plant data); Richard Condit (Panama
and Peru tropical forest data); Hilary Lease (Sevilleta LTER
robberfly data); Frank Olsen (Rock Island Preserve lepidoptera
data); Mike Palmer and Dan McGlinn (Tallgrass Prairie Preserve
data); Peter White (eastern Asian and North American regional
flora data). Peter Minchin provided the COMPAS code for
simulating community turnover along environmental gradients.
Initial work was funded through a Santa Fe Inst. workshop grant.
References
Allen, J. A. and Sanders, H. L. 1996. The zoogeography, diversity
and origin of the deep-sea protobranch bivalves of the Atlantic:
the epilogue. – Prog. Oceanogr. 38: 95–153.
Barker, G. M. 2005. The character of the New Zealand land
snail fauna and communities: some evolutionary and ecological
perspectives. – Rec. West. Aust. Mus. Suppl. 63:
52–102.
Barker, G. M. and Mayhill, P. C. 1999. Patterns of diversity
and habitat relationships in terrestrial mollusc communities of
the Pukeamaru Ecological District, northeastern New Zealand.
– J. Biogeogr. 26: 215–238.
Bell, G. 2003. The interpretation of biological surveys. – Proc. R.
Soc. B 270: 2531–2542.
Bird, N. M. 1998. Sustaining the yield: improved timber
harvesting practices in Belize, 1992–1998. – Natural Resources
Inst., Univ. of Greenwich, NRI Catalog Services, CAB
International, Wallingford, UK.
Blanchette, C. A. et al. 2008. Biogeographical patterns of rocky
intertidal communities along the Pacific coast of North
America. – J. Biogeogr. 35: 1593–1607.
Chave, J. and Leigh, E. G. Jr 2002. A spatially explicit neutral
model of b-diversity in tropical forests. – Theor. Popul. Biol.
62: 153–168.
Condit, R. et  al. 2000. Spatial patterns in the distribution
of tropical tree species. – Science 288: 1414–1418.
Condit, R. et al. 2002. Beta-diversity in tropical forests. – Science
295: 666–669.
Curtis, J. T. 1955. A prairie continuum in Wisconsin. – Ecology
36: 558–566.
Ernest, S. K. M. et  al. 2009. Long-term monitoring and
experimental manipulation of a Chihuahuan Desert ecosystem
near Portal, Arizona, USA. – Ecology 90: 1708.
Gilbert, B. and Lechowicz, M. J. 2004. Neutrality, niches, and
dispersal in a temperate forest understory. – Proc. Natl Acad.
Sci. USA 101: 7651–7656.
Gleason, H. A. 1926. The individualistic concept of plant
association. – Bull. Torrey Bot. Club 53: 7–26.
Green, J. L. and Bohannan, B. J. M. 2006. Spatial scaling
of microbial biodiversity. – Trends Ecol. Evol. 21: 501–507.
174
320
Schlicht, D. W. et al. 2007. The butterflies of Iowa. – Univ. of
Iowa Press.
Soininen, J. et al. 2007. The distance decay of similarity in ecological
communities. – Ecography 30: 3–12.
Smith, J. 1990. The frugal gourmet on our immigrant ancestors.
– Harper-Collins.
Steinitz, O. et al. 2005. Predicting regional patterns of similarity
in species composition for conservation planning. – Conserv.
Biol. 19: 1978–1988.
Svenning, J.-C. et  al. 2004. Ecological determinism in plant
community structure across a tropical forest landscape.
– Ecology 85: 2526–2538.
Tobler, W. R. 1970. A computer movie simulating urban growth
in the Detroit region. – Econ. Geogr. 46: 234–240.
Tuomisto, H. et  al. 2003. Dispersal, environment, and floristic
varia­tion of western Amazonian forests. – Science 299: 241–244.
Voss, E. G. 1985. Michigan flora. – Cranbrook Inst. of Science,
Bloomfield Hills, MI, Bulletin 59.
Whittaker, R. H. 1975. Communities and ecosystems. – MacMillan
Publishing.
Plotkin, J. B. and Muller-Landau, H. C. 2002. Sampling
the species composition of a landscape. – Ecology 83:
3344–3356.
Preston, F. W. 1962. The canonical distribution of commonness
and rarity. – Ecology 43: 185–215, 410–432.
Qian, H. et al. 2005. Beta diversity of angiosperms in temperate
floras of eastern Asia and eastern North America. – Ecol.
Lett. 8: 15–22.
Qian, H. et al. 2009. The latitudinal gradient of beta diversity in
relation to climate and topography for mammals in North
America. – Global Ecol. Biogeogr. 18: 111–122.
Rapoport, E. H. 1982. Areography: geographical strategies of
species. – Pergamon Press.
Rocchini, D. and Cade, B. S. 2008. Quantile regression applied to
spectral distance decay. – Geosci. Remote Sens. Lett. 5: 640–643.
Root, T. 1989. Atlas of wintering North American birds. – Univ.
of Chicago Press.
Sauer, J. R. et al. 2011. The North American Breeding Bird Survey,
results and analysis 1966–2009. Version 3.23.2011. – USGS
Patuxent Wildlife Research Center, Laurel, MD.
175
IDEA AND
PERSPECTIVE The wealth of species: ecological communities,
complex systems and the legacy of Frank Preston
Jeffrey C. Nekola* and James H.
Brown
Department of Biology,
University of New Mexico,
Albuquerque, NM 87131, USA
*Correspondence: E-mail:
jnekola@unm.edu.
Abstract
General statistical patterns in community ecology have attracted considerable recent
debate. Difficulties in discriminating among mathematical models and the ecological
mechanisms underlying them are likely related to a phenomenon first described by Frank
Preston. He noted that the frequency distribution of abundances among species was
uncannily similar to the Boltzmann distribution of kinetic energies among gas molecules
and the Pareto distribution of incomes among wage earners. We provide additional
examples to show that four different Ôdistributions of wealthÕ (species abundance
distributions, species–area and species–time relations, and distance decay of compositional
similarity) are not unique to ecology, but have analogues in other physical,
geological, economic and cultural systems. Because these appear to be general statistical
patterns characteristic of many complex dynamical systems they are likely not generated
by uniquely ecological mechanistic processes.
Keywords
Community ecology, competitive sorting, complexity science, distance decay, ecological
theory, neutral models, species–abundance distribution, species–area relationship,
species–time relationship, statistical mechanics.
Ecology Letters (2007) 10: 188–196
Frank Preston (1950) published a little paper on ÔGas Laws
and Wealth LawsÕ in The Scientific Monthly. In it he
commented on the remarkable similarities between the
Boltzmann frequency distribution of molecular kinetic
energies in gasses, the Pareto frequency distribution of
personal incomes in countries, and the frequency distribution
of species abundances in ecological communities. After
noting how such distributions might arise, he went on to
remark – in comments which would now be politically
incorrect in reference to personal incomes – about whether
it is Ôwise to try to change the laws of natureÕ.
Interest in community Ôdistributions of wealthÕ (DOWs),
in particular the species–abundance distribution (SAD),
species–area relationship (SAR), species–time relationship
(STR), and distance decay of compositional similarity (DD),
have preoccupied community ecologists ever since Preston’s
seminal papers (Preston 1948, 1962a,b, 1980; see also
Williams 1964; MacArthur 1972; May 1975; Nee et al. 1991;
Rosenzweig 1995; Nekola & White 1999; Gaston &
Blackburn 2000; Hubbell 2001). As new, larger, and more
comprehensive data sets have become available, the
discipline of community ecology has been enlivened by
debate over which of several different conceptual frameworks
or formal mathematical models might be necessary or
sufficient to explain these patterns.
These debates have primarily had two foci. First, authors
have argued about the relative roles of ÔcompetitiveÕ
(deterministic ecological differences between species) vs.
ÔneutralÕ (stochastic processes of birth, death and dispersal)
mechanisms in community assembly. This issue has been
difficult to resolve as formal models invoking complete
neutrality generate very similar outcomes to models that
assume unique species-level abiotic requirements or strong
biotic interactions (e.g. Bell 2000; Chave et al. 2002;
Mouquet & Loreau 2003; Chave 2004; Tilman 2004; Gaston
& Chown 2005).
The second focus has been about which particular
mathematical distribution (and corresponding ecological
mechanism) best describes empirical DOWs. This debate
has been constrained by the fact that mathematical
distributions having different mechanistic interpretations
can be very difficult to empirically distinguish. For instance,
it has proved extremely difficult to determine whether a
particular SAD data set is best fit by a power law, truncated
Ecology Letters, (2007) 10: 188–196 doi: 10.1111/j.1461-0248.2006.01003.x
Ó 2007 Blackwell Publishing Ltd/CNRS
176
log-normal, or zero sum multinomial (ZSM) distribution
(e.g. McGill 2003a; Volkov et al. 2003; Chave 2004; Etienne
& Olff 2005), and similarly whether a power law or
exponential distribution best fits a given SAR/STR relationship
(Loehle 1990; White et al. 2006). May (1975) and
Connor & McCoy (1979) made these points decades ago for
the SAD and SAR.
We regard as healthy the renewed interest in community
ecology, and especially the focus on large-scale, multispecies
macroecological pattern and process which are often not
amenable to testing by manipulative experiments. But we
also wonder whether the current framing of questions and
tenor of debates are productive. Perhaps, as Preston
suggested, DOWs are not the unique provenance of
ecology, but rather are common properties of many
seemingly disparate systems. If so, it may be best to seek
explanations for these patterns that are as general as the
systems that exhibit them.
To address this issue, we have extended Preston’s
examples to a broader selection of data sets spanning a
wide range of physical and human economic, social and
artistic systems. Physical systems: (i) Yearly precipitation
averages for each of 1027 North American sites (see Nekola
2005). These data are based on 1-km resolution global
precipitation maps created by the WORLDCLIM Project
(http://www.worldclim.org); (ii) Mineral species richness
from county to global scales, as reported by the MINDAT
data base (http://www.mindat.org). Data represent the total
number of minerals for the entire terrestrial globe, for 10
countries (Australia, Canada, Chile, China, India, Mexico,
Namibia, Turkey, the UK and the USA), all the USA states
(including the District of Columbia), all Canadian provinces/
territories, and two selected counties per state. Areas for each
unit were based on data provided with ArcMap 9.1 (ESRI,
Redlands, California, USA). Economic systems: 2004 Stock
volumes for all publicly traded corporations in the USA,
based on data reported by Bloomberg (http://pages.stern.
nyu.edu/$adamodar/New_Home_Page/data.html). Social
systems: (i) Citation frequencies for all papers catalogued by
the Institute for Scientific Information from 1981 to 1997 (http://
physics.bu.edu/$redner/projects/citation/isi.html); (ii) the
list of all commercially sold garden vegetable varieties in
the USA and Canada from 1981 to 2004 as reported by the
Garden Seed Inventory of the Seed Savers Exchange; (iii) the
list of all ingredients reported in recipes for 10 global cuisines
(Ethiopia, Hungary, India, Iran, Ireland, Korea, Mexico,
Norway, Puerto Rico and Thailand) from Smith (1990).
Artistic systems: (i) The number of unique words and total
word length for the 1863 texts documented in Project
Gutenberg as of 2000 (http://www.mine-control.com/
zack/guttenberg/); (ii) concert setlists for 33 approximately
evenly spaced Cowboy Junkies performances from 1987 to
2006 (http://setlist.com).
Using standard ecological protocols (e.g. ÔoctavesÕ or log2
bin widths for SADs and Jaccard’s similarity in the
calculation of compositional similarity between all pairwise
combinations of observations for DD analyses), we treated
each as if they were ecological data sets. All DOW analogues
were calculated from the Cowboy Junkies example, as only
this data set provided abundance data across multiple
observations. For the remaining, analyses were limited to the
DOW equivalents that could be assessed from the given
data: SAD analogues for precipitation class frequencies,
stock volumes and scientific citations; SAR analogues for
mineral species richness and Project Gutenberg texts; STR
analogues for North American garden seed offerings; and
DD analogues for global cuisine and vegetable seed
offerings.
These analyses corroborate Preston’s observation that
typical DOW patterns are by no means unique to ecological
systems. North American precipitation classes, 2004 stock
volumes, scientific citation frequency and song frequencies
from Cowboy Junkies setlists (Fig. 1) are all similar to
truncated log-normal or ZSM distributions with rare events
being more frequent than abundant ones. Using code
developed by McGill (2003a) the ZSM can be shown to well
fit both the precipitation (h ¼ 55.173, m ¼ 0.179) and
Cowboy Junkies setlist (h ¼ 89.446, m ¼ 0.240) data. Power
law SAR and STR-like relationships are also widespread
(Fig. 2), being observed not only for the accumulation of
mineral species across space, but also for unique words as a
function of book size and for unique vegetable varieties and
Cowboy Junkies song performances as a function of time.
The fitted relationships explained 46–99% of observed
variance. Finally, nonlinear DD relationships are also
common (Fig. 3), being evident in cuisine ingredient lists,
the garden seed industry, and Cowboy Junkies setlists. For
these examples, power law decay models fit best, accounting
for 38–88% of observed variance.
The convergence between ecological and non-ecological
community patterns is not limited to these examples. In his
final Ecology paper, Preston (1981) documented rarityenriched
ZSM-like SAD analogues for the service life of
restaurant drink tumblers, the static fatigue of glass and other
materials, and first marriage age for Danish, UK and the USA
women. Power law STR analogues have also been noted in
comprehensive examination scores for degree candidates in
the University of Oslo, Department of Biology and for Norsk
Hydro stock prices on the Norwegian market (Ugland et al.
2005). And, given the convergence between the body-size
distribution of beetles in Borneo tree canopies and cars in
York and Heathrow Airport parking lots (Gaston et al. 1993),
it appears such similarities may not be limited to DOWs.
These examples caution that the general mechanisms
generating many familiar DOW patterns are not unique
to ecology. Certainly, the processes of birth, death,
Idea and Perspective The wealth of species 189
Ó 2007 Blackwell Publishing Ltd/CNRS
177
immigration and speciation invoked by Hubbell (2001), or
the competitive tradeoffs invoked by Tilman (2004) or
Chave et al. (2002) could not have generated the ZSM-like
SADs for precipitation classes, paper citations, stock
volumes, drink tumbler longevity, and marriage ages, or
power law SARs/STRs for unique mineral, word, and
garden seed occurrences, or comprehensive examination
scores. Conversely, it seems unlikely that Cowboy Junkies
performances mimic community ecology process, even
though they display rarity-enriched SAD, power law STD
and nonlinear DD patterns.
Preston (1950) concluded that the remarkable convergence
between the Boltzmann, Pareto and SAD might be
the result of Ôstatistical mechanicsÕ. Even though he could
not identify the common underlying mechanisms, he urged
scientists to Ôunderstand [this] law, and the causes that bring
it aboutÕ. We question whether the recent revival of interest
in ecological DOWs has advanced this goal. These
congruent patterns seem to beg for some kind of
explanation that is both more universal and also less
explicitly mechanistic than the kinds typically sought and
offered by community ecologists.
Over the last few decades investigations in the new
interdisciplinary field of complexity science have begun to
address such issues. Complexity science developed as
natural and social scientists sought to identify and explain
common features exhibited by complex dynamical systems
in such seemingly disparate fields as quantum physics,
computer science, economics, sociology, political science,
linguistics, astronomy, geology and meteorology. While a
single definition for a Ôcomplex systemÕ remains elusive, a
number of common features have been identified (Brown
1994a,b; West 2006a,b):
(1) They are composed of many components of many
different kinds.
(2) These components interact with each other and the
extrinsic environment in many different ways and on
multiple spatial and temporal scales.
(3) These interactions give rise to complex structures and
complicated nonlinear dynamics.
(4) These structures and dynamics are neither completely
stochastic nor entirely deterministic, but instead represent
a combination of randomness and order.
(5) They contain both positive and negative feedback
mechanisms, causing either amplification or damping
of temporal and spatial variation, depending on
conditions.
Figure 1 Species–abundance distribution analogues for non-ecological systems, including North American precipitation classes, stock
volumes for all publicly traded US corporations, song performances for Cowboy Junkies setlists, and citation frequencies for scientific papers.
All distributions have been binned into log2 ÔoctavesÕ. For precipitation classes and Cowboy Junkies performance frequencies, the best-fit zero
sum multinomial distribution has also been calculated using code provided by McGill (2003a). These best-fit curves are presented as the
graphed lines in both of these panels.
190 J. C. Nekola and J. H. Brown Idea and Perspective
Ó 2007 Blackwell Publishing Ltd/CNRS
178
(6) They are open systems which require exchanges of
energy, materials, and/or information from extrinsic
sources to maintain highly organized states far from
thermodynamic equilibrium.
(7) They are historically contingent, so that their present
configurations reflect the influence of initial conditions
and subsequent perturbations.
(8) They are often nested within other complex systems,
giving rise to hierarchical organizations that can be
approximated by fractal geometry and dynamic scaling
laws.
Ecological communities clearly demonstrate these features.
Even the simplest contain thousands to billions of
individuals of tens to thousands of different species,
ranging from unicellular procaryotes, protists, and fungi to
multicellular plants and animals. These individuals and
species interact with each other and their extrinsic abiotic
environment across multiple spatial and temporal scales.
These relationships are often inherently nonlinear, ranging
from Michaelis–Menten curves for nutrient uptake to
exponential or logistic population growth to the normal,
skewed or bimodal distribution of species along environmental
gradients. Feedbacks are prevalent, as demonstrated
by effects of keystone species on ecosystem function or
fuel buildup on fire return frequencies. Ecological communities
require the continual transformation of energy,
material and information to maintain their highly organized,
far-from-equilibrium thermodynamic states. Many
ecological processes are historically contingent, with
contemporaneous patterns reflecting legacies of past events
in both shallow (e.g. the sequence of habitat colonization)
and deep (e.g. plate tectonics and evolution) time. And
finally, ecological communities are composed of nested
hierarchies of complex components, ranging from organic
Figure 2 Species–area relationship and species–time relationship analogues for non-ecological systems, including the accumulation of unique
mineral varieties by sample area, unique words by total document words, and unique vegetable varieties and Cowboy Junkies songs by time.
All data are graphed on log-transformed axes. Best-fit linear regression lines have been plotted for each relationship.
Idea and Perspective The wealth of species 191
Ó 2007 Blackwell Publishing Ltd/CNRS
179
molecules and cells through populations, species, guilds
and trophic levels.
Because of these shared structural and dynamical
properties, ecological communities and other kinds of
complex systems tend to develop similar patterns of
whole-system structure and dynamics. An example of this
is the occurrence of many small and few large magnitude
events, which can be observed in the cumulative frequency
distributions of word use, scientific paper citations, Internet
web hits, copies of books sold, telephone calls received on a
single day, earthquake magnitudes, lunar crater diameters,
solar flare intensity, deaths in wars, wealth of rich people,
surname frequencies, and city populations (Newman 2005).
These patterns can often be nearly equally well fit by power
law or log-normal probability distribution functions (West &
Shlesinger 1989). The finding of DOW analogues (rarityenriched
SADs, roughly power law SAR and STR, and
nonlinear DD) across multiple non-ecological systems thus
suggests that these patterns represent shared statistical
properties of a large class of complex systems. To discover
the ÔlawÕ sought by Preston to explain these striking
convergences may therefore require an understanding of
the general principles that govern the structure and
dynamics of all these systems.
The uncanny similarity of ecological DOWs to those
exhibited by other complex systems has been largely ignored
by community ecologists, who for the most part have
sought explanation strictly from biological processes (but
see Limpert et al. 2001; McGill 2003b; Halloy & Whigham
2004; Ugland et al. 2005). A few authors have pointed out
phenomenological similarities between ecological and other
complex systems and suggested that these may hold clues to
common causes (e.g. Preston 1950, 1981; May 1975, 1981;
Bak et al. 1987; Brown 1994b; Brown et al. 2000, 2002; Allen
& Holling 2002; Sole´ & Bascompte 2006), but others have
suggested that such similarities are spurious (Root 1989).
Until such convergences are taken seriously and efforts are
made to identify their root causes, community ecology risks
becoming a myopic enterprise.
There are, in fact, a number of general processes that
might influence wealth accumulation across a wide variety of
systems. First, and perhaps most importantly, log-normal
and mathematically related power law probability distributions
are easily generated by the multiplicative rather than
additive interaction of variables. Such processes have been
shown to commonly apply to many physical (Meijer et al.
1981), biological (May 1975; McGill 2003b) and human
systems (Montroll & Shlesinger 1982). Even the log-normal
distribution of publication rates for researchers within
scientific institutions is governed by such multiplicative
processes (Shockley 1957). Similar distributions may also be
generated by the interaction of multiple agents governed by
nonlinear processes (Bak et al. 1987; West & Shlesinger
1989). Because of these processes, rarity-enriched SAD
analogues may be a universal expectation for many complex
systems (McGill 2003b). Second, many physical, biological
and human systems exist within environments that exhibit
power law or fractal-like spatial and temporal variation
Figure 3 Distance decay analogues for non-ecological systems,
including the change of compositional similarity of global cuisine
ingredient lists over space, and garden seed offerings and Cowboy
Junkies song performances over time. Best-fit exponential and
power law decay lines have been plotted for each relationship.
192 J. C. Nekola and J. H. Brown Idea and Perspective
Ó 2007 Blackwell Publishing Ltd/CNRS
180
(Mandelbrot 1982; Milne 1991; Ritchie & Olff 1999; Brown
et al. 2002). Not only can these environmental templates be
generated via simple hierarchical random models (Sizling &
Storch in press), but can also in turn give rise to typical
DOW patterns such as the power law SAR and nonlinear
DD (Sizling & Storch 2004; Harte et al. 2005). Third,
proximity effects, which occur when the influence of an
event on an agent is dependent upon the distance in space
and/or time between agent and event, occur in many
physical, biological and human systems. Such relationships
may be found across phenomena as diverse as electromagnetic
radiation and gravitation to natural disturbances,
dispersal and diffusion to migration, trade and wars.
Proximity effects not only can directly give rise to DD
relationships (Nekola & White 1999), but also can be an
important source of multiplicative relationships over space
or time. It is also possible to identify direct analogues in the
processes governing different systems. For instance, both
ecological and human systems exhibit hierarchical arrangements
of agents, including individuals within populations
and soldiers within armies and specialized classes of workers
within insect societies and manufacturing firms. The
structures and dynamics of both ecological and human
systems also reflect a complicated dynamic balance among
diverse cohesive and divisive forces, such as mutualism,
reciprocity, and cooperation on the one hand and selfinterest,
competition, and predatory or parasitic exploitation
on the other. The joint effects of such factors conspire to
ensure that ÔwealthÕ is distributed highly unequally among
agents and nonlinearly over space or time.
What lessons can be learned from viewing ecological
communities as complex systems and ecological DOWs as
shared statistical properties of such systems? First, the
mechanistic explanation for many of ecology’s most
venerable statistical patterns lie at a level of abstraction
extending far beyond the realm of ecological process. The
paradox that rarity-enriched SADs, power law SAR/STR,
and nonlinear DD are generated by models assuming either
complete neutrality or resource competition is therefore
explained by realizing both models generate dynamical
complexity. In competition models complexity arises as
multiple agents, each with individualistic resource utilization
and dispersal functions, interact in an elaborate network
with multiple other agents often over multiple temporal and
spatial scales. Such interactions can be made even more
complicated and realistic by inclusion of some degree of
stochasticity in resource requirements and dispersal processes
(Tilman 2004). In neutral models complexity arises as
multiple agents, each with a unique community and
metacommunity frequency, undergo a lottery for recruitment
of vacated spaces. These interactions are further
complicated as new species are supplied to the system across
multiple scales via dispersal and speciation. As a result it
may prove impossible to infer underlying mechanisms from
DOW mathematical forms, both in general and for specific
cases. However, this should not suggest that efforts to
adduce ecological community assembly mechanisms must
be abandoned. For instance, while a power law relationship
describes the relationship between mean and variance for
heartbeat frequency or interstride interval, the slope of this
function supplies information about the relative importance
of deterministic vs. stochastic drivers (West & Latka 2005;
West 2006b). Similar mechanistic inferences can be made in
ecological systems. In the case of the SAR, it has long been
recognized that differences in slope over local-to-global
scales or between islands and mainlands convey information
about the mechanisms that generate and maintain diversity
(Rosenzweig 1995). Similarly, DD rates provide information
about niche characteristics and dispersal capacities in
relation to the environmental template of spatial and
temporal variations. So, for example, isolated spruce-fir
forests of the Appalachians demonstrate an almost threefold
greater DD rate when compared with continuous northern
Taiga, and large-fruited, more dispersal-limited plant species
have almost twice the DD rate when compared with
smaller-seeded taxa (Nekola & White 1999).
Second, we question the utility of investing great effort to
determine whether competition or neutrality might be
responsible for observed DOWs when there is no logical
reason to expect these mechanisms to be mutually exclusive –
or, for that matter, to represent the only possible alternatives.
Such patterns are almost certainly generated by combinations
of stochastic and deterministic processes, local and regional
processes, current and historical events, biotic interactions
and abiotic factors, direct and indirect interactions, and
cohesive and divisive forces. Similarly, diverse combinations
of factors likely generate analogous properties in some
physical and many human social and economic systems.
Third, it is long overdue that community ecologists keep
abreast of developments in the field of complexity science.
We are not suggesting that complexity science has all the
solutions to the big important questions of community
ecology. Indeed, complexity science is struggling to address
similar problems and can currently offer few definitive
answers. However, we are suggesting that ecological,
physical and social scientists have much to learn from each
other. Ideas, models and data from community ecology have
the potential to make important contributions to complexity
science, just as theoretical and empirical advances in
complexity science have the potential to enlighten community
ecology. This focus on interdisciplinarity is not a
revolutionary concept. Community ecology has a long
tradition of borrowing ideas and techniques from many
other fields including economics (e.g. law of supply and
demand, MacArthur & Wilson 1967; game theory, Smith
1982; Gini coefficients, Weiner & Solbrig 1984; supply side
Idea and Perspective The wealth of species 193
Ó 2007 Blackwell Publishing Ltd/CNRS
181
ecology, Roughgarden et al. 1987), geography (DD, Nekola
& White 1999; gravity models, Bossenbroek et al. 2001); the
behavioural sciences (e.g. multidimensional scaling ordination,
Minchin 1987), and physics (e.g. diffusion models,
Skellam 1951; chaos, May 1976; percolation theory, O’Neill
et al. 1992).
Finally, complexity science suggests that certain aspects
of community ecology may never be highly predictable.
While qualitative forecasts may be possible, the precise
quantitative prediction of individual events may pose
insurmountable challenges. For example, while it might
be possible to predict qualitatively how community-wide
SADs and SARs will be altered by invasions of multiple
exotic species, it will be much more difficult to exactly
predict resultant species abundances and spatial distributions.
Indeed, such an effort may be as quixotic as
attempting to predict the exact location of an electron in a
shell or date and magnitude of the next earthquake in the
San Francisco Bay area or price of General Motors stock
in 2050. As in weather forecasting, some level of shortterm
predictability will often be possible from understanding
current drivers, recent trajectories, and spatial/temporal
autocorrelations. However, predictions based on these
dynamics will necessarily become increasingly imprecise as
the forecast period increases.
Rather than emphasizing prediction, community ecologists
perhaps should spend more effort on understanding
the mechanisms and events that have conspired to generate
current and past patterns. Even when it may be practically
impossible to predict the exact future trajectory of a system,
by looking backward it may be possible to deduce quite
accurately when and how specific mechanisms came into
play. In complex human systems mechanisms are deduced
by historians who analyse post hoc the particular combination
of initial conditions and drivers that generated pattern.
Similarly, community ecologists could enhance their understanding
of the spatial and temporal patterns of biodiversity
by becoming better natural historians, using post hoc analysis
to decipher how past events have left lasting influences. By
retrospectively studying systems, community ecologists may
thus be able to establish what factors lead to the dominance
of a particular taxon or to the occurrence of a particular
hantavirus outbreak or simply whether Schro¨dinger’s Cat
survived.
For community ecology to continue making exciting
advances, it must recognize that the distribution of
abundance among species has much in common with
distributions of ÔwealthÕ in many other non-ecological
systems. The fact that the distribution of species abundance
is very similar to the distribution of drinking glass
longevities or Cowboy Junkies song performances is thus
both empowering and humbling in equal parts. While such
similarities suggest a possible unification between ecological
communities and other complex systems, they also
suggest that many of ecology’s cherished fundamental
patterns may reflect general phenomena that have more in
common with statistical physics than species biology. The
apparent universality of the rarity-enriched SAD, power
law SAR/STR, and nonlinear DD suggests that community
ecologists should pay attention to these more fundamental
levels of inquiry and explanation. Complexity researchers
of all types, including community ecologists, would be well
advised to follow Frank Preston’s path and to focus on the
common factors that underlie these seemingly ubiquitous
distributions of wealth.
ACKN O WLEDG EMEN TS
We thank Matt Barthel of the Seed Savers Exchange for
granting us access to the digital version of their Garden Seed
Inventory. We also wish to thank Nicholas Gotelli, three
anonymous referees, Linda Fey and current and former
members of Brown Lab for helpful feedback regarding
earlier versions. This work was supported in part by a
National Science Foundation Biocomplexity grant (DEB
0083422) and a grant from the Los Alamos National
Laboratory (W-7405-ENG-36).
REFEREN CES
Allen, C.R. & Holling, C.S. (2002). Cross-scale structure and scale
breaks in ecosystems and other complex systems. Ecosystems, 5,
315–318.
Bak, P., Tang, C. & Wiesenfeld, K. (1987). Self-organized criticality:
an explanation of 1/f noise. Phys. Rev. Lett., 59, 381–384.
Bell, G. (2000). The distribution of abundance in neutral communities.
Am. Nat., 155, 606–617.
Bossenbroek, J.M., Kraft, C.E. & Nekola, J.C. (2001). Prediction of
long-distance dispersal using gravity models: Zebra Mussel
invasion of inland lakes. Ecol. Appl., 11, 1178–1788.
Brown, J.H. (1994a). Organisms and species as complex adaptive
systems: linking the biology of populations with the physics of
ecosystems. In: Linking Species and Ecosystems (eds Jones, C.G. &
Lawton, J.H.). Chapman and Hall, London, pp. 16–24.
Brown, J.H. (1994b). Complex ecological systems. In: Complexity:
Metaphors, Models, and Reality (eds Cowan, G.A., Pines, D. &
Melzer, D.). Santa Fe Institute Studies in the Science of Complexity,
Proceedings Volume XVIII. Addison-Wesley, Reading,
pp. 419–449.
Brown, J.H., West, G.B. & Enquist, B.J. (2000). Scaling in Biology:
Patterns and processes, causes and consequences. In: Scaling in
Biology (eds Brown, J.H. & West, G.B.). Oxford University Press,
Oxford, pp. 1–24.
Brown, J.H., Gupta, V.K., Li, B.L., Milne, B.T., Restrepo, C. &
West, G.B. (2002). The fractal nature of nature: power laws,
ecological complexity, and biodiversity. Proc. R. Soc. Lond. B, Biol.
Sci., 357, 619–626.
Chave, J. (2004). Neutral theory and community ecology. Ecol.
Lett., 7, 241–253.
194 J. C. Nekola and J. H. Brown Idea and Perspective
Ó 2007 Blackwell Publishing Ltd/CNRS
182
Chave, J., Muller-Landau, H.C. & Levin, S.A. (2002). Comparing
classical community models: theoretical consequences for patterns
of diversity. Am. Nat., 159, 1–23.
Connor, E.F. & McCoy, E.D. (1979). The statistics and biology of
the species–area relationship. Am. Nat., 113, 791–833.
Etienne, R.S. & Olff, H. (2005). Confronting different models of
community structure to species–abundance data: a Bayesian
model comparison. Ecol. Lett., 8, 493–504.
Gaston, K.J. & Blackburn, T.M. (2000). Pattern and Process in Macroecology.
Blackwell Science, Oxford.
Gaston, K.J. & Chown, S.L. (2005). Neutrality and the niche. Funct.
Ecol., 19, 1–6.
Gaston, K.J., Blackburn, T.M. & Lawton, J.H. (1993). Comparing
animals and automobiles: a vehicle for understanding body size
and abundance relationships in species assemblages? Oikos, 66,
172–178.
Halloy, S. R. P. & Whigham, P.A. (2004). The lognormal as universal
descriptor of unconstrained complex systems: a unifying
theory for complexity. In: Proceedings of the 7th Asia-Pacific Complex
Systems Conference. Cairns Convention Centre, Cairns, Qld, Australia,
pp. 309–320.
Harte, J., Conlisk, E., Ostling, A., Green, J.L., Smith, A.B. (2005).
A theory of spatial structure in ecological communities at multiple
spatial scales. Ecol. Monogr., 75, 179–197.
Hubbell, S.P. (2001). The unified neutral theory of biodiversity and
biogeography. Monographs in Population Biology #32. Princeton
University Press, Princeton, NJ.
Limpert, E., Stahel, W.A. & Abbt, M. (2001). Log-normal distributions
across the sciences: keys and clues. Bioscience, 51, 341–
352.
Loehle, C. (1990). Proper statistical treatment of species–area data.
Oikos, 57, 143–146.
MacArthur, R.H. (1972). Geographical Ecology. Harper and Row,
New York, NY.
MacArthur, R.H. & Wilson, E.O. (1967). Theory of island biogeography.
Monographs in Population Biology #1. Princeton University
Press, Princeton, NJ.
Mandelbrot, B.B. (1982). The Fractal Geometry of Nature. W.H.
Freeman & Co., New York, NY.
May, R.M. (1975). Patterns of species abundance and diversity.
In: Ecology and Evolution of Communities (eds Cody, M.L. & Diamond,
J.). Belknap Press of Harvard University Press, Cambridge,
MA, pp. 81–120.
May, R.M. (1976). Simple mathematical models with very complicated
dynamics. Nature, 261, 459–467.
May, R.M. (1981). Patterns in multi-species communities. In: Theoretical
Ecology (ed. May, R.M.). Blackwell Scientific, Oxford, pp.
197–227.
McGill, B.J. (2003a). A test of the unified neutral theory of biodiversity.
Nature, 422, 881–885.
McGill, B.J. (2003b). Strong and weak tests of macroecological
theory. Oikos, 102, 679–685.
Meijer, P.H.E., Mountain, R.D. & Soulen, R.J. Jr (1981). Sixth
International Conference on Noise in Physical Systems. National Bureau
of Standards Special Publication #614, Washington, DC.
Milne, B.T. (1991). Lessons from applying fractal models to
landscape patterns. In: Quantitative Methods in Landscape Ecology
(eds Turner, M.G. & Gardner, R.H.). Springer-Verlag, New
York, NY, pp. 199–235.
Minchin, P.R. (1987). An evaluation of the relative robustness of
techniques for ecological ordination. Vegetatio, 69, 89–108.
Montroll, E.W. & Shlesinger, M.F. (1982). On 1/f noise and other
distributions with long tails. Proc. Natl Acad. Sci. USA, 79, 3380–
3383.
Mouquet, N. & Loreau, M. (2003). Community patterns in sourcesink
metacommunities. Am. Nat., 162, 544–557.
Nee, S., Harvey, P.H. & May, R.M. (1991). Lifting the veil on
abundance patterns. Proc. R. Soc. Lond. B, Biol. Sci., 243, 161–163.
Nekola, J.C. (2005). Latitudinal richness, evenness, and shell size
gradients in eastern North American land snail communities.
Rec. West. Aust. Mus. Suppl., 68, 39–51.
Nekola, J.C. & White, P.S. (1999). Distance decay of similarity in
biogeography and ecology. J. Biogeogr., 26, 867–878.
Newman, M.E.J. (2005). Power laws, Pareto distributions, and
Zipf’s law. Contemp. Phys., 46, 323–351.
O’Neill, R.V., Gardner, R.H. & Turner, M.G. (1992). A hierarchical
neutral model for landscape analysis. Landscape Ecol., 7,
55–61.
Preston, F.W. (1948). The commonness, and rarity, of species.
Ecology, 29, 254–283.
Preston, F.W. (1950). Gas laws and wealth laws. Sci. Mon., 71, 309–
311.
Preston, F.W. (1962a). The canonical distribution of commonness
and rarity – I. Ecology, 43, 185–215.
Preston, F.W. (1962b). The canonical distribution of commonness
and rarity – II. Ecology, 43, 410–432.
Preston, F.W. (1980). Noncanonical distributions of commonness
and rarity. Ecology, 6, 88–97.
Preston, F.W. (1981). Pseudo-lognormal distributions. Ecology, 62,
355–364.
Ritchie, M.E. & Olff, H. (1999). Spatial scaling laws yield a synthetic
theory of biodiversity. Nature, 400, 557–560.
Root, R.B. (1989). Resolution of respect Frank W. Preston 1896–
1989. Bull. Ecol. Soc. Am. 70, 244–247.
Rosenzweig, M.L. (1995). Species Diversity in Space and Time. Cambridge
University Press, Cambridge, UK.
Roughgarden, J., Gaines, S.D. & Pacala, S.W. (1987). Supply-side
ecology: the role of physical transport processes. Symp. Br. Ecol.
Soc., 27, 491–518.
Shockley, W. (1957). On the statistics of individual variations of
productivity in research laboratories. Proc. Inst. Radio Eng., 45,
279–290.
Sizling, A.L. & Storch, D. (2004). Power-law species-area
relationships and self-similar species distributions within finite
areas. Ecol. Lett., 7, 60–68.
Sizling, A.L. & Storch, D. (in press). Geometry of species distributions:
random clustering and scale invariance. In: Scaling Biodiversity
(eds Storch, D., Marquet, P.A. & Brown, J.H.),
Cambridge University Press, Cambridge.
Skellam, J.G. (1951). Random dispersal in theoretical populations.
Biometrika, 38, 196–218.
Smith, J.M. (1982). Evolution and the Theory of Games. Cambridge
University Press, Cambridge.
Smith, J. (1990). The Frugal Gourmet on our Immigrant Ancestors.
Harper-Collins, New York, NY.
Sole´, R.V. & Bascompte, J. (2006). Self-organization in complex ecosystems.
Monographs in Population Biology #42. Princeton University
Press, Princeton, NJ.
Idea and Perspective The wealth of species 195
Ó 2007 Blackwell Publishing Ltd/CNRS
183
Tilman, D. (2004). Niche tradeoffs, neutrality, and community
structure: a stochastic theory of resource competition, invasion,
and community assembly. Proc. Natl Acad. Sci. USA, 101, 10854–
10861.
Ugland, K.I., Gray, J.S. & Lambshead, J.D. (2005). Species accumulation
curves analysed by a class of null models discovered by
Arrhenius. Oikos, 108, 263–274.
Volkov, I., Banavar, J.R., Hubbell, S.P. & Maritan, A. (2003).
Neutral theory and relative species abundance in ecology. Nature,
424, 1035–1037.
Weiner, J. & Solbrig, O.T. (1984). The meaning and measurement
of size hierarchies in plant populations. Oecologia, 61,
334–336.
West, B.J. (2006a). Thoughts on modeling complexity. Complexity,
11, 33–43.
West, B.J. (2006b). Complexity, scaling and fractals in biological
signals. In: Wiley Encyclopedia of Biomedical Engineering (ed. Akay
M.). Wiley-Interscience, New York, NY.
West, B.J. & Latka, M. (2005). Fractional Langevin model of gait
variability. J. Neuroeng. Rehabil., 2, 24.
West, B.J. & Shlesinger, M.F. (1989). On the ubiquity of 1/f noise.
Int. J. Mod. Phys. B, 3, 795–819.
White, E.P., Adler, P.B., Lauenroth, W.K., Gill, R.A., Greenberg,
D., Kaufman, D.M. et al. (2006). A comparison of the species–
time relationship across ecosystems and taxonomic groups.
Oikos, 112, 185–195.
Williams, C.B. (1964). Patterns in the Balance of Nature and Related
Problems in Quantitative Biology. Academic Press, London.
Editor, Nicholas Gotelli
Manuscript received 29 August 2006
First decision made 4 October 2006
Manuscript accepted 30 November 2006
196 J. C. Nekola and J. H. Brown Idea and Perspective
Ó 2007 Blackwell Publishing Ltd/CNRS
184
185
186
187
188
189
190
191
192
193
194
195
Section VI: Conservation Biology
Because of the precarious status of biodiversity across the globe, it is important for ecologists to
apply their pure research findings to the development of optimal conservation strategies. As
suggested by Susan Bratton, the current state of conservation biology is not unlike that of the
medical profession in 1700’s Vienna, when patients faced a lower chance of survival after
treatment than if they had let the disease run its course. Because of the complex nature of
ecological systems, we should not search for universally applicable and beneficial reserve design
or management strategies or seek exact quantitative predictions regarding particular
environmental perturbations. Rather, each reserve will face unique series of threats and will
require artist-like managers who develop unique management protocols for each site. Much of
my conservation biology research has been focused on documenting the unintended negative
impacts posed by fire management policies on biodiversity, especially in fragmented landscapes
with low potential for recolonization.
Total biodiversity maintenance in
these situations requires reduced use
of fire across both space and time,
with sensitive application of large
vertebrate grazing and use of more
labor intensive techniques such as
hand-removal of woody vegetation. I
have also published on the importance
of distance and colonization history in
reserve design, have documented the
negative impact of recreational rock
climbing and soil disturbance on land
snail diversity, and have investigated
the impact of powerline corridors on
biodiversity.
Representative Publications
[number of citations as of October 27, 2017]
Nekola, J.C. 2012. The impact of utility corridors on terrestrial gastropod biodiversity.
Biodiversity and Conservation. 21:781–795. [9]
McMillan, M., J.C. Nekola & D.W. Larson. 2003. Impact of recreational rock climbing on land
snail communities of the Niagara Escarpment, southern Ontario, Canada. Conservation
Biology. 17:616-621. [41]
Nekola, J.C. 2002. Effects of fire management on the richness and abundance of central North
American grassland land snail faunas. Animal Biodiversity and Conservation. 25:53-66.
[46]
Nekola, J.C. & P.S. White. 2002. Conservation, the two pillars of ecological explanation, and
the paradigm of distance. Natural Areas Journal. 22:305-310. [23]
White, P.S. & J.C. Nekola. 1992. Biological diversity in an ecological context. Pages 10-29
in: Barker, J.R. and D.T. Tingey (eds), Air pollution and biodiversity. Van Nostrand
Reinhold, New York. [6]
196
ORIGINAL PAPER
The impact of a utility corridor on terrestrial gastropod
biodiversity
Jeffrey C. Nekola
Received: 9 May 2011 / Accepted: 21 December 2011 / Published online: 11 January 2012
Ó Springer Science+Business Media B.V. 2012
Abstract Utility corridors are often thought to be disruptive to biodiversity because they
cause habitat fragmentation that may lead to increases in predation, parasitism, disease transmittance
and vagrant species while decreasing migration rates, gene flow and genetic diversity
for interior species. Species with poor dispersal abilities, sedentary lifestyles, and specialized
habitats have been thought to be potentially the most vulnerable to these effects. Terrestrial
gastropods thus serve as a valuable system in which to investigate these impacts because they
are among the poorest active dispersers in the animal kingdom. To document the impact of
corridor formation on land snail biodiversity, a 75-year old powerline right-of-way in the
eastern Upper Peninsula of Michigan was chosen for analysis. While terrestrial gastropod
richness and abundance was significantly reduced for corridor as compared to adjacent control
subsamples, with a 2/3 turnover in species composition, the corridor fauna is similar to nearby
nativegrasslandsitesintermsof speciescomposition,abundance distribution, andnumbersand
abundance of species of conservation concern. The fauna of control subsamples immediately
adjacent to the corridor remained similar to other undisturbed sites in the region, with multiple
speciesofconservationconcernpersistingatdistancesofonly30 mfromthecorridor.Thus,the
netimpactofcorridorgenerationhas beenarguably positive:while the surroundingforest fauna
has not been degraded, within the corridorthe reduction of forest species has beencompensated
for by establishment of even rarer grassland species.
Keywords Land snail Á Right-of-way Á Landscape Á Community ecology Á
Community structure Á Soil architecture
Introduction
Utility corridors have often been thought to be disruptive to biodiversity (Forman 1986;
Andrews 1990) because they fragment habitats, leading to increased predation, parasitism
and disease transmittance rates (Simberloff and Cox 1987; Yahner et al. 1989; Bennett
J. C. Nekola (&)
Department of Biology, University of New Mexico, Albuquerque, NM 87131, USA
e-mail: jnekola@unm.edu
123
Biodivers Conserv (2012) 21:781–795
DOI 10.1007/s10531-011-0216-8
Author's personal copy
197
1991), decreased migration abilities and lowering of gene flow and within-population
genetic diversity (Fahig and Mirriam 1985; Goosem and Marsh 1997; Gerlach and Mustov
2000), and expansion of vagrant species at the expense of interior species (Arnold et al.
1987). Species with poor dispersal abilities, sedentary habits, and specialized habitat
preferences have been thought to be the most vulnerable to these effects (Shaffer and
Samson 1985; Andrews 1990).
Terrestrial gastropods would therefore appear particularly sensitive to utility corridor
formation. The North American fauna, particularly at mid-to-high latitudes, is dominated
by species with shells \5 mm in maximum dimension that represent approximately
50–80% of regional species richness and 80–95% of individuals (Nekola 2005). Such
species have very poor active dispersal abilities: individuals move perhaps only 1–10 m
over a lifetime (Schilthuizen and Lombaerts 1994; Hausdorf and Hennig 2003). Even
active movement of larger species (maximum shell dimension [20 mm) has been effectively
inhibited by barriers of no more than 3–8 m (Baur 1988; Baur and Baur 1990;
Schilthuizen and Lombaerts 1994). Also, terrestrial gastropods exhibit a high degree of
habitat specialization, with order-of-magnitude changes in richness and abundance over as
little as 1 m (Nekola and Smith 1999).
Even though utility corridors may have serious negative impacts on terrestrial gastropod
biodiversity and community structure, no prior research has specifically addressed this
issue. This study therefore investigates changes in community structure and composition
between paired corridor and matrix habitats along a 75-year old utility corridor passing
through the Hiawatha National Forest in the eastern Upper Peninsula of Michigan.
Comparisons were also made with other undisturbed habitats in the surrounding landscape.
Materials and methods
Study system
ESE-6904/6905 is a 42 km powerline operated by the American Transmission Company in
Chippewa and Mackinac counties in the eastern Upper Peninsula of Michigan. The line
connects St. Ignace in the south to the Pine River Substation near Rudyard in the north. The
corridor was initially constructed in the 1930s and is 30–60 m in width. The line currently
runs at 69,000 volts with approximately 395 double-circuit wood pole structures. In
Mackinac County a considerable proportion of the corridor passes though the St. Ignace
Ranger District of the Hiawatha National Forest, where it has been maintained by mowing
and mechanical woody plant removal.
In total, 58 land snail species have been previously identified from the eastern Upper
Peninsula (Hubricht 1985; Nekola 2004). Of these, 13 are listed as of conservation concern
by the Michigan Natural Features Inventory (2011): Catinella exile—Threatened; Euconulus
alderi—Threatened; Planogyra asteriscus—Special Concern; Pupilla muscorum—
Special Concern; Vallonia gracilicosta—Endangered; Vertigo bollesiana—Threatened;
Vertigo cristata—Special Concern; Vertigo elatior—Special Concern; Vertigo hubrichti—
Endangered; Vertigo morsei—Endangered; Vertigo nylanderi—Endangered; Vertigo
paradoxa—Special Concern and Vertigo pygmaea—Special Concern. Six species (Cochlicopa
lubrica, Oxychilus draparnaudi, Pupilla muscorum, Vallonia costata, Vallonia
pulchella, Vertigo pygmaea) appear to be Eurasian exotics based on their preference for
anthropogenic habitats, known North American history and/or population genetics (Pilsbry
1948; Nekola and Coles 2010).
782 Biodivers Conserv (2012) 21:781–795
123
personal copy
198
Sampling design
Eighteen sample sites were identified along the ESE-6904/6905 route as it passes through
the Hiawatha National Forest in eastern Mackinac County, Michigan. Sample sites were
spread across the entire corridor route, were separated by at least 0.5 km, and represented
all natural habitats crossed by the corridor. These included marl ponds, fens, sedge
meadows, upland and lowland Northern White Cedar stands, sandy pine forest, mixed
deciduous forest, and rocky slopes. Sample sites were located in conjunction with pole
numbers 505, 520, 567, 600, 605, 623, 652, 680, 696, 705, 717, 726, 728, 755, 772, 790,
811 and 813 (Fig. 1). At most sites three subsamples were collected: one within the
existing corridor, and two controls located within intact vegetation 30 m to both the west
and east of the corridor. No private lands were surveyed. For this reason only the corridor
and west margin subsamples were collected at Pole 811 and only the corridor subsample
was collected at Pole 600. Also, because of abundance of Poison Ivy and lack of leaf litter
accumulation within the corridor, only the western margin at Pole 813 was sampled.
Comparisons were also made with 32 additional sites in the surrounding landscape,
including 11 wooded bedrock outcrops, 5 upland forests, 7 lowland forests, 8 lowland
grasslands and a single upland grassland. These capture the full compositional range of
regional land snail assemblages (Nekola 2003, 2004, 2005; Fig. 1).
Field methods
The land snail assemblage of each 100 m2
subsample was documented from September
25–30, 2009 through hand collection of larger shells and litter sampling for smaller taxa
following protocols of Cameron and Pokryszko (2005) and Oggier et al. (1998).
Fig. 1 Map of the eastern Upper Peninsula of Michigan showing location of the 18 sample sites analyzed
along the ESE-6904/6905 corridor along with locations of the 32 additional comparison sites
Biodivers Conserv (2012) 21:781–795 783
123
Author's personal copy
199
Approximately 30 min were spent collecting each subsample. Soil litter collections followed
protocols of Nekola and Coles (2010), and emphasized microsites known to support
high micro-snail densities (e.g., Emberton et al. 1996). Approximately 200 ml of litter
were collected per subsample. This generally captured ten times as many individuals as
species, which is advocated by Cameron and Pokryszko (2005) for accurate land snail
community documentation. The latitude-longitude position of each pole location was
determined using a Garmin 12XL hand-held GPS unit.
Sampling of comparison sites elsewhere in the eastern Upper Peninsula was conducted
from July 1997–October 2009 in a similar fashion (see Nekola 2003, 2004, 2005) although
at an order of magnitude larger spatial sampling grain. Any systematic bias would thus
result in underestimating the relative richness of ESE-6904/6905 subsamples through
random undersampling of the rarest species. However, this will not obscure community
composition gradients (Jobe 2007).
Laboratory procedures
As outlined by Nekola and Coles (2010), dried litter samples were passed through a
standard sieve series and handpicked against a neutral-brown background. All shells and
shell fragments were removed. All shells were assigned to species (or subspecies) using the
author’s reference collection, with the total numbers of shells per species per site being
recorded. The number of unassignable immature individuals and fragmentary shells was
recorded. Nomenclature is based principally on Hubricht (1985) with updates by Turgeon
et al. (1998), Nekola (2004) and Nekola and Coles (2010).
Data summary and statistical analysis
Species richness and abundance
For each species, the total number of subsamples in which it was recorded and the total
number of encountered individuals was calculated. Using the non-parametric Kruskal–
Wallis Rank Sum Test, these values were compared among the west of corridor, corridor,
and east of corridor subsamples, and were graphically illustrated using boxplots. To better
visualize the upper and lower bounds, boxplots were constructed using a natural-log scale.
However, raw values were used for statistical testing. Variation in community structure
among the three subsample positions was illustrated by dominance–diversity curves generated
from median proportional abundance values calculated within each of the three
subsample locations. Median values were used because they are more robust descriptors of
central tendency, being less influenced by outliers as compared to mean values. Variation
in richness and abundance of three soil architecture preference classes (duff specialist,
generalist, turf specialist) among the three subsample positions was also statistically tested
using the Kruskal–Wallis Rank Sum Test and illustrated with boxplots. As before, abundance
distributions were visualized using natural-log scaled box-plots, but tested using raw
data. The soil architecture nomenclature and preferences of Nekola (2003) were used, with
‘duff specialists’ being species that favor soils with loose organic horizons and ‘turf
specialists’ being those that favor organic horizons that are tightly bound with living plant
roots. The only exception was Striatura exigua, which is here considered a ‘generalist’
rather than a ‘turf’ species (Nekola 2003) based on additional eastern North American data
(Nekola 2010). Differences in richness and abundance for species of conservation concern
784 Biodivers Conserv (2012) 21:781–795
123
Author's personal copy
200
and alien species among the three subsample positions was also quantified using the
Kruskal–Wallis Rank Sum Test.
Community composition
The impact of the corridor on community composition was analyzed through global nonmetric
multidimensional scaling (NMDS) using DECODA (Minchin 1990). NMDS makes
no assumptions regarding the underlying nature of species distributions along compositional
gradients, and is thus the most robust form of ordination for detection of ecological
patterns (Minchin 1987). All 49 ESE-6904/6905 subsamples in addition to the 32 additional
regional comparison sites were included in this analysis. To ordinate sites a matrix of
dissimilarity coefficients was calculated based on species abundance data using the
Czekanowski index (Faith et al. 1987). Data were untransformed to avoid mathematical
artifacts (Nekola et al. 2008). NMDS in one through four dimensions was then performed
with 200 iterations, a stress ratio stopping value of 0.9999, and a small stress stopping
value of 0.01. Output was scaled in half-change units, so that points in the resulting
diagram separated by a distance of 1.0 will correspond, on average, to a 50% turnover in
species composition.
Because a given NMDS run may locate a local, rather than the global, stress minimum,
solution stability was assessed via 50 NMDS runs using different random initial starting
points (Minchin 1987). Solutions in each of the four dimensions were compared using a
Procrustes transformation to identify those that were statistically similar. The number of
unique solutions and number of individual runs that fell into each was then calculated. The
global optimum solution was identified as the smallest stress solution that was achieved in
a plurality of starts.
To assess the impact of corridor construction on community composition, a vector was
generated using a given control subsample as the origin and the adjacent corridor subsample
as the endpoint. Each of these vectors was plotted on the ordination diagram and its
length and angle from vertical calculated. This allowed a median vector of compositional
change to be determined across the entire dataset. The significance of these changes in
ordination space was estimated using the non-parametric paired Wilcoxon Rank Sum Test.
Fisher Exact contingency table tests were used to identify those species that were
statistically more abundant within the corridor or control subsamples as compared to a null
hypothesis of uniform distribution. Analysis was limited to paired subsamples (i.e. 600C
and 813W were eliminated). Species with fewer than ten total occurrences were not
analyzed due to the likelihood of Type-II errors. Because this test was repeated for each
species, a Bonferroni correction was used to modify the significance threshold. The results
of this analysis were then compared to the results of the duff vs. turf soil preference tests of
Nekola (2003) by use of a 3 9 3 contingency table. This table was compared to uniform
expectations using a Fishers exact test.
Results
In total, 44 species were observed from the 49 subsamples collected along the ESE-6904/
6905 corridor (Table 1). Of these, five (Gastrocopta pentodon, Oxychylus draparnaudi,
Vertigo bollesiana, Vertigo ovata and Vertigo pygmaea) are reported here for the first time
from Mackinac county (Hubricht 1985; Nekola 2004), raising the total county fauna to 54
species. Nine of the thirteen species of conservation concern known from the region
Biodivers Conserv (2012) 21:781–795 785
123
Author's personal copy
201
Table 1 Number of occupied subsamples and individuals for all encountered terrestrial gastropod species
Species name Soil architecture
preference
Number of subsamples/individuals
West of corridor Powerline corridor East of corridor
Anguispira alternata D 2/3 1/1 1/2
Carychium exiguum T 9/293 12/543 7/292
Carychium exile D 6/99 3/19 7/88
*Cochlicopa lubrica D 4/93 3/18 1/1
Columella simplex D 14/127 10/49 16/167
Deroceras spp. G 4/5 7/12 6/9
Discus catskillensis D 9/167 4/5 12/109
Euchemotrema fraternum D 0/0 1/2 1/1
Euconulus alderi T 6/26 8/57 5/19
Euconulus fulvus D 8/31 2/10 9/75
Euconulus polygyratus D 3/16 2/9 2/4
Gastrocopta pentodon D 1/1 1/33 0/0
Gastrocopta tappaniana T 2/25 7/101 3/24
Glyphyalinia indentata D 1/1 1/1 1/1
Helicodiscus parallelus G 0/0 1/2 0/0
Helicodiscus shimeki D 3/11 1/2 9/54
Nesovitrea binneyana D 8/102 0/0 9/144
Nesovitrea electrina T 4/28 10/158 4/43
*Oxychylus draparnaudi G 0/0 1/3 0/0
Oxyloma retusa T 1/1 4/6 0/0
Paravitrea multidentata D 1/8 0/0 0/0
Planogyra asteriscus G 8/814 2/7 8/603
Punctum minutissimum D 16/1,653 7/141 16/1,371
Punctum n.sp. T 6/84 8/410 6/70
*Pupilla muscorum G 1/9 1/8 0/0
Striatura exigua G 13/248 9/195 15/351
Striatura ferrea G 9/176 3/52 12/102
Striatura milium G 12/332 6/65 14/334
Strobilops labyrinthica D 15/1,279 14/275 16/993
Succinea ovalis D 0/0 3/4 0/0
*Vallonia costata G 2/55 1/5 0/0
*Vallonia pulchella T 0/0 2/4 0/0
Vertigo bollesiana D 6/15 0/0 6/14
Vertigo cristata D 7/33 2/20 5/19
Vertigo elatior T 5/42 11/118 7/17
Vertigo gouldii D 2/3 0/0 2/7
Vertigo morsei T 1/9 4/48 0/0
Vertigo nylanderi T 2/24 0/0 2/8
Vertigo ovata T 0/0 3/3 1/1
*Vertigo pygmaea G 1/4 2/2 0/0
Vitrina limpida G 2/8 2/7 0/0
Zonitoides arboreus D 6/14 3/5 5/9
786 Biodivers Conserv (2012) 21:781–795
123
Author's personal copy
202
(Euconulus alderi, Planogyra asteriscus, Pupilla muscorum, Vertigo bollesiana, Vertigo
cristata, Vertigo elatior, Vertigo morsei, Vertigo nylanderi, and Vertigo pygmaea) were
observed. In total, 14, 242 individuals were collected, of which 13,185 were identifiable to
species. Thirty-one of the subsamples included at least 200 individuals. The 32 eastern
Upper Peninsula comparison sites ranged in richness from 3 to 25 species and in abundance
from 19 to 1,480, with a total of 12,831 individuals being identified to species.
Impact on species richness and abundance
Based on paired observations, both species richness and abundance were lower within the
corridor than in adjacent control subsamples (Fig. 2): median richness fell by approximately
 (from 14 to 10; P = 0.015) and abundance by approximately ’ (from 400 to
100; P \ 0.00003). However, dominance-diversity curves demonstrated that median proportional
abundance was very similar for up to approximately the 7th most abundant
species (Fig. 3), after which median abundance fell off more rapidly for the corridor
subsamples.
Comparison of richness and abundance across species with differing soil architecture
preferences (Fig. 4) demonstrates that control subsamples were dominated by duff
Fig. 2 Boxplots showing variation in terrestrial gastropod richness and abundance among the three
subsample positions along the ESE-6904/6905 corridor. Abundance data are presented along a log-scaled
axis. The central line represents the median of the sample, the margins of each box represent the
interquartile distances, and the fences represent 1.5 times the interquartile distances. Outliers falling outside
of the fences are shown with circles
Table 1 continued
Species name Soil architecture
preference
Number of subsamples/individuals
West of corridor Powerline corridor East of corridor
Zonitoides nitidus T 1/4 0/0 0/0
Zoogenetes harpa D 2/4 1/4 1/2
Total terrestrial gastropods 16/5,847 17/2,404 16/4,934
Species underlined are species of conservation concern in Michigan (Michigan Natural Features Inventory
2011). Species preceded by an asterisk are Eurasian aliens. Soil architecture preference is based on Nekola
(2003) with minor revisions as per text; D Duff specialist, T turf specialist, G generalist
Biodivers Conserv (2012) 21:781–795 787
123
Author's personal copy
203
Fig. 3 Dominance-diversity curves demonstrating proportional abundance versus rank order of terrestrial
gastropod species for the three subample positions along the ESE-6904/6905 corridor
Fig. 4 Boxplots showing variation in duff specialist, generalist, and turf specialist terrestrial gastropod
species richness and abundance among the three subsample categories along the ESE-6904/6905 corridor.
Abundance data are presented along a log-scaled axis. The central line represents the median of the sample,
the margins of each box represent the interquartile distances, and the fences represent 1.5 times the
interquartile distances. Outliers falling outside of the fences are shown with circles
788 Biodivers Conserv (2012) 21:781–795
123
Author's personal copy
204
specialists (*57% of taxa and *80% of individuals), followed by generalists (*27% and
*18%) and turf specialists (*16% and *2%; Fig. 4). However, corridor subsamples
were dominated by turf specialists (*50% and *63%) followed by duff specialists
(*30% and *23%), and generalists (*20% and *14%). While these differences were
statistically significant for duff and turf specialists (P = 0.016 and P = 0.0020, respectively,
for richness, and P = 0.0095 and P = 0.0037 for individuals), generalist species
did not differ significantly (P = 0.1054 for richness and P = 0.0948 for individuals).
No significant differences were noted in richness (P = 0.7842) or abundance (P =
0.7350) of species of conservation concern or in exotic species richness (P = 0.2095) or
abundance (P = 0.2124) among corridor and adjacent control subsamples.
Impact on community composition
NMDS along two dimensions was chosen as the most robust solution, given that all five
minimum stress configurations fell into a single group. The 1-, 3-, and 4-dimensional
solutions were not considered as they either possessed larger stress values or had no single
modal solution. The chosen ordination documents a dramatic alteration of terrestrial
gastropod community composition within the corridor (Fig. 5). While control subsample
assemblages remain similar to other regional lowland conifer forest, upland forest, and
lowland grassland assemblages, corridor faunas tend to be displaced down and/or left in the
diagram, being similar to the faunas of native lowland grassland sites (e.g., sedge meadows,
fens). Displacement vectors from a given control subsample to its paired corridor
subsample ranged from 0.18 to 3.4, or 9% to 170% change in composition. Across all pairs,
the median distance was 1.27 (indicating roughly a 2/3 turnover in composition) at -110°
from the top of the diagram. This displacement was highly significant (P \ 0.00002) along
both axes.
Fig. 5 NMDS ordination of terrestrial gastropod faunas from the 49 ESE-6904/6905 subamples and 32
comparison sites. Axes are scaled in half-change units. a Optimum two-dimensional NMDS solution.
Powerline corridor subamples are represented by hatched diamonds; control subamples are indicated by a
circle surrounding the appropriate habitat icon; uncircled habitat icons represent comparison sites.
b Vectors demonstrating the change in composition between control (origin) and adjacent corridor
(endpoint) samples
Biodivers Conserv (2012) 21:781–795 789
123
Author's personal copy
205
Following removal of unpaired subsamples, fifteen species were eliminated from further
analysis because they were represented by ten or fewer individuals. Fisher exact tests of the
remaining 29 species (Table 2) demonstrated that eight either significantly favored
(Bonferroni corrected P-threshold = 0.0017) or tended to favor (0.05 [ P [0.0017) corridor
subsamples, 14 significantly favored the adjacent control subsamples, while the
remaining six were generalists (0.05 \ P). While three species of conservation concern
favored control subsamples (Planogyra asteriscus, Vertigo bollesiana, Vertigo nylanderi)
an additional three favored corridor subsamples (Euconulus alderi, Vertigo elatior, Vertigo
morsei) with two more being generalists (Pupilla muscorum, Vertigo cristata). Although
Vertigo pygmaea was not statistically analyzed because not enough individuals were
encountered, 2/3 of its occurrences were within the corridor.
Comparison of these responses to known soil architecture preferences (Nekola 2003)
indicates that the two factors are significantly linked (P = 0.0154; Table 3): species that
favored corridor subsamples tended to favor turf soils, while species favoring control
subsamples tended to favor duff soils.
Discussion
The sampling protocols used in this study were found to be more than adequate to accurately
characterize the land snail fauna and its response to powerline corridor formation.
First, roughly 80% of the known county and 75% of the eastern Upper Peninsula fauna was
encountered despite the fact that the corridor does not intersect the carbonate cliff sites that
serve as one of the most important regional reservoirs for land snail biodiversity (Nekola
1999). This sampling design also resulted in documenting well the regional species of
conservation concern. Only four such species were not encountered: Vallonia gracilicosta,
Vertigo hubrichti and Vertigo paradoxa were not observed because their stable carbonate
cliff habitat was not intersected by the ESE-6904/6905 corridor; and Catinella exile was
also not recorded even though it has been observed in low numbers on the marl fen at
Summerby Swamp to the northwest of the corridor. While seemingly appropriate habitat
for this species exists within the ESE-6904/6905 corridor and adjacent habitats, its absence
was not surprising as it is also absent from the immediately adjacent Martineau Creek fen
(Nekola 1998). Additionally, both the entire dataset as well as ’ of the individual subsamples
meet the criteria of Cameron and Pokryszko (2005) for accurate portrayal of land
snail community patterns.
These data show that creation of a treeless powerline corridor has led to profound
changes in the terrestrial gastropod community. Both the richness and abundance of land
snails within the corridor have fallen in comparison to adjacent habitats, with composition
turning over by roughly 2/3 as turf (i.e., grassland) specialists replace duff (i.e., forest)
specialists.
This change, however, has not led to a reduction in total land snail biodiversity. First,
faunas of control subsamples remain compositionally similar to those of other undisturbed
sites in the region, with multiple species of conservation concern persisting only 30 m from
the corridor edge. Second, species abundance distributions were generally unaffected, with
proportional abundance being similar for corridor and control subsamples up to the seventh-most
abundant species. Although abundance levels trailed off more rapidly for corridor
subsamples from the eighth-most-abundant species on, this general pattern is
characteristic for native grassland faunas throughout the region (Nekola 2002, 2010). As a
result, the more rapid fall off of abundance is probably not due to the corridor, per se, but
790 Biodivers Conserv (2012) 21:781–795
123
Author's personal copy
206
rather indirectly to the conversion to grassland habitats/faunas. Third, ordination demonstrates
that within 75 years the corridor fauna has evolved into assemblages very similar to
other native lowland grassland sites in the region, in particular sedge meadows and fens.
Table 2 Impact of powerline corridor on species abundance
Name Corridor samples Control samples P value
Obs. Exp. Obs. Exp.
A. Species with too few observations (10 or less) to characterize preference
Anguispira alternata; Euchemotrema fraternum; Glyphyalinia indentata; Helicodiscus parallelus; Oxychilus
draparnaudi; Oxyloma retusa; Paravitrea multidentata; Succinea ovalis; Vallonia costata; Vallonia
pulchella; Vertigo gouldii; Vertigo ovata; Vertigo pygmaea; Vitrina limpida; Zoogenetes harpa
B. Generalist species
Name Corridor Subamples Control Subamples P value
Obs. Exp. Obs. Exp.
Deroceras sp. 12 9 14 17 0.572488900
Euconulus polygyratus 9 10 20 19 1.000000000
Punctum n.sp. 85 81 154 158 0.773239300
Pupilla muscorum 8 6 9 11 0.728282100
Vertigo cristata 20 25 52 47 0.472278900
Zonitoides arboreus 5 10 23 18 0.226957500
C. Species that favor (P \ 0.0017) or tend to favor (0.05 [ P [ 0.0017) corridor subsamples
Carychium exiguum 465 357 585 693 0.000001676
Cochlicopa lubrica 18 8 6 16 0.008414732
Euconulus alderi 48 32 45 61 0.026030660
Gastrocopta pentodon 33 12 1 22 0.000000047
Gastrocopta tappaniana 95 49 49 95 0.000000090
Nesovitrea electrina 151 76 71 146 0.000000000
Vertigo elatior 118 60 59 117 0.000000001
Vertigo morsei 48 19 9 38 0.000000047
D. Species that favor (P \ 0.0017) or tend to favor (0.05 [ P [ 0.0017) control subsamples
Carychium exile 19 70 187 136 0.000000001
Columella simplex 49 116 293 226 0.000000002
Discus catskillensis 5 96 276 185 0.000000000
Euconulus fulvus 10 39 106 77 0.000004013
Helicodiscus shimeki 2 23 65 44 0.000002788
Nesovitrea binneyana 0 84 246 162 0.000000000
Planogyra asteriscus 7 485 1,417 939 0.000000000
Punctum minutissumum 141 1,028 2,879 1,992 0.000000000
Striatura exigua 195 270 599 524 0.000043710
Striatura ferrea 52 112 278 218 0.000000083
Striatura milium 65 249 666 482 0.000000000
Strobilops labyrinthica 275 867 2,271 1,679 0.000000000
Vertigo bollesiana 0 10 29 19 0.000767735
Vertigo nylanderi 0 11 32 21 0.000347028
Species considered of conservation concern by the Michigan Natural Features Inventory are underlined
Biodivers Conserv (2012) 21:781–795 791
123
Author's personal copy
207
The ultimate impact of corridor formation has been for the expansion of lowland
grassland land snail faunas into the previously largely forested matrix. Fourteen species
(*30% of total) were found to favor control subsamples, while eight (*20%) favored the
corridor itself. These changes are similar to those previously noted with forest clearing
(Nekola 2003). Because species of conservation concern in this region are roughly equally
partitioned between forests and open habitats, the decrease in population occurrence/size of
forest-favoring rare species (Planogyra asteriscus, Vertigo bollesiana, Vertigo nylanderi)
in the corridor has been compensated by the increase of grassland-favoring rare species
(Euconulus alderi, Vertigo elatior, Vertigo morsei), with Vertigo morsei representing one
of the rarest land snail species in the Great Lakes region (Nekola 2004). As a result, no
significant change in either the total richness or abundance of species of conservation
concern was noted between corridor and adjacent forest habitats. However, the increase in
coverage and population sizes of grassland species that would otherwise be rare or absent
from the corridor has enriched total biodiversity. This result holds even when the two alien
species listed by the Michigan Natural Features Inventory (Pupilla muscorum, Vertigo
pygmaea) are removed from analysis.
Even though terrestrial gastropods appear to be ideal candidates for documenting detrimental
biodiversity impacts from corridor formation, these data demonstrate that systemwide
diversity has not been lowered by corridor formation. Rather, the net impact appears
positive, as the corridor has allowed for an expansion of grassland/turf specialists,
including one of the rarest species in the landscape, while at the same time not seriously
impacting the faunas of adjacent habitats.
How have small grassland snails with poor active dispersal abilities been able to occupy
the new grassland habitats created by corridor formation within the relatively short ecological
time frame that the ESE-6904/6905 corridor has been in existence? The explanation
is probably that small land snails, such as those that dominate the Upper Peninsula fauna,
possess great passive dispersal capabilities. For example, DNA sequence data demonstrate
that members of the genus Balea have been repeatedly carried across 9,000 km of open
eastern Atlantic Ocean by migrating birds (Gittenberger et al. 2006). Similarly, in North
America many members of the Vertigo gouldii group have ranges exceeding 5,000 km in
extent even though they inhabit areas covered by continental ice as recently as 10 ka
(Nekola et al. 2009). Passive dispersal of some small land snail species is also facilitated by
their ability to reproduce uniparentally (Pokryszko 1987), allowing only single individuals
to found populations. As a result, small land snails tend to possess much larger ranges and
to more completely saturate their available habitats than larger taxa (McClain and Nekola
2008; Nekola 2009).
Table 3 Species response to subsamples placement versus duff/turf soil architecture preference, with
numbers in each cell representing number of species
Soil architecture preference
Duff species Generalist species Turf species
Corridor preference
Control subsamples 9 4 1
Generalist 3 1 1
Corridor subsamples 2 0 6
Fisher’s exact test: P = 0.0154
792 Biodivers Conserv (2012) 21:781–795
123
Author's personal copy
208
The most likely vectors for passive land snail dispersal in the ESE-6904/6905 corridor
are birds and small mammals. For both of these groups, utility corridors represent travel
paths for grassland species through forested landscapes (Kroodsma 1982; Gates 1991).
This will allow for movement of grassland/turf specialist land snails through corridors as
well. Additionally, narrow corridors such as that for ESE-6904/6905 have been found to
not restrict movements of small mammals in eastern North American forests (Schreiber
and Graves 1977; Gates 1991), and gliding mammals (Asari et al. 2010), small mammals,
reptiles, and amphibians (Carthew et al. 2009) in Australia. Consequently, the ESE-6904/
6905 corridor probably does not serve as a barrier to passive movements of minute forest
snails. It should thus not be surprising that the richness, abundance, and community
structure of land snail assemblages on either side of the corridor were found to be very
similar.
These results mirror numerous studies that have shown that utility corridors serve as
refuges for grassland species within forested landscapes. This effect has been repeatedly
shown for birds (Kroodsma 1982; King et al. 2009; Meehan and Haas 1997) and mammals
(Gates 1991) in eastern North America and for a variety of small vertebrates in Australia
(Goosem and Marsh 1997; Clarke et al. 2006; Clarke and White 2008a, b). As was found in
the current study, these grassland species may be quite uncommon elsewhere within the
landscape (Clarke and White 2008b). Similar patterns exist for vascular plants in the
southeastern USA, where powerline corridors serve as refuges in an otherwise forested
matrix for a number of critically endangered grassland/savanna species (Sorrie and
Weakley 2001).
Thus, as long as a given utility corridor (1) does not substantially impact the overall
population of a given forest-specialist species, (2) exists within a forest-dominated landscape
that also includes native grassland habitats and (3) is maintained through manual
vegetation cutting, it may actually benefit terrestrial gastropod biodiversity by maintaining
refuges for both common and rare grassland species while not leading to the extirpation of
sensitive forest species. For the current alignment of ESE-6904/6905 this equates to
roughly 3–6 ha of grassland habitat per kilometer of corridor. However, it should not be
assumed that these positive benefits will also accrue in landscapes lacking native grassland
habitats, or for corridors maintained through herbicide application (Nowak et al. 1993).
Although quantitative sampling for land snails was not conducted in areas subjected to this
type of management, it appeared that terrestrial gastropod richness and abundance was
much lower in herbicide-managed sections of the corridor just outside of the National
Forest boundary.
Acknowledgments Thanks are extended to Todd Miller and the American Transmission Company for
facilitating access to the study site and for underwriting field work and analysis expenses. Joe Pagliara of
Natural Resources Consulting, Inc./Stantec helped interface with ATC and also assisted with field work.
Melinda Fey, Robert Cowie, and two anonymous reviewers provided invaluable comments on earlier
manuscript drafts.
References
Andrews A (1990) Fragmentation of habitat by roads and utility corridors: a review. Aust Zool 26:130–141
Arnold GW, Algar D, Hobbs RJ, Atkins L (1987) A survey of vegetation and its relationship to vertebrate
fauna present in winter on road verges in the Kellerberrin District, Western Australia. Technical Report
18, Department of Conservation and Land Management, Western Australia
Asari Y, Johnson CN, Parsons M, Larson J (2010) Gap-crossing in fragmented habitats by mahogany gliders
(Petaurus gracilis). Do they cross roads and powerline corridors? Aust Mammol 32:10–15
Biodivers Conserv (2012) 21:781–795 793
123
Author's personal copy
209
Baur B (1988) Microgeographical variation in shell size of the land snail Chondrina clienta. Biol J Linn Soc
35:247–259
Baur A, Baur B (1990) Are roads barriers to dispersal in the land snail Arianta arbustorum? Can J Zool
68:613–617
Bennett AF (1991) Roads, roadsides, and wildlife conservation: a review. In: Saunders DA, Hobbs RJ (eds)
Conservation 2: the role of corridors. Surrey Beatty & Sons, Chipping Norton
Cameron RAD, Pokryszko BM (2005) Estimating the species richness and composition of land mollusc
communities. J Conchol 38:529–547
Carthew SM, Horner B, Jones KMW (2009) Do utility corridors affect movements of small terrestrial fauna?
Wildl Res 36:488–495
Clarke DJ, White JG (2008a) Towards ecological management of Australian powerline corridor vegetation.
Landsc Urban Plan 86:257–266
Clarke DJ, White JG (2008b) Recolonization of powerline corridor vegetation by small mammals: timing
and influence of vegetation management. Landsc Urban Plann 87:108–116
Clarke DJ, Pearce KA, White JG (2006) Powerline corridors: degraded ecosystems or wildlife havens?
Wildl Res 33:615–626
Emberton KC, Pearce TA, Randalana R (1996) Quantitatively sampling land-snail species richness in
Madagascan rainforests. Malacologia 38:203–212
Fahig L, Mirriam G (1985) Habitat patch connectivity and population survival. Ecology 66:1762–1768
Faith DP, Minchin PR, Belbin L (1987) Compositional dissimilarity as a robust measure of ecological
distance. Vegetatio 69:57–68
Forman RTT (1986) Landscape ecology. Wiley, New York
Gates JE (1991) Powerline corridors, edge effects, and wildlife in forested landscapes of the central
Appalachians. In: Rodiek JE, Bolen EG (eds) Wildlife and habitats in managed landscapes: an
overview. Island Press, Washington
Gerlach G, Mustov K (2000) Fragmentation of landscape as a cause for genetic subdivision in bank voles.
Conserv Biol 14:1066–1074
Gittenberger E, Groenenberg DSJ, Kokshoorn B, Preece RC (2006) Molecular trails from hitch-hiking
snails. Nature 439:409
Goosem M, Marsh H (1997) Fragmentation of a small mammal community by a powerline corridor through
tropical rainforest. Wildl Res 24:613–629
Hausdorf B, Hennig C (2003) Nestedness of north-west European land snail ranges as a consequence of
differential immigration from Pleistocene glacial refuges. Oecologia 35:102–109
Hubricht L (1985) The distributions of the native land mollusks of the eastern United States. Fieldiana New
Ser 24:1–191
Jobe RT (2007) Estimating landscape-scale species richness: reconciling frequency- and turnover-based
approaches. Ecology 89:174–182
King DI, Chandler RB, Collins JM, Petersen WR, Lautzenheiser TE (2009) Effects of width, edge and
habitat on the abundance and nesting success of scrub-shrub birds in powerline corridors. Biol Conserv
142:2672–2680
Kroodsma RL (1982) Edge effect on breeding forest birds along a power-line corridor. J Appl Ecol
19:361–370
McClain C, Nekola JC (2008) The role of local-scale on terrestrial and deep-sea Gastropod body size
distributions across multiple scales. Evol Ecol Res 10:129–146
Meehan AL, Haas CA (1997) Use of a powerline corridor by breeding and wintering birds in Giles County,
Virginia. Va J Sci 48:259–264
Michigan Natural Features Inventory (2011) Michigan’s Special Animals. http://web4.msue.msu.edu/mnfi/
data/specialanimals.cfm. Accessed 7 Mar 2011
Minchin PR (1987) An evaluation of the relative robustness of techniques for ecological ordination. Vegetatio
69:89–108
Minchin PR (1990) DECODA software. Australian National University, Canberra
Nekola JC (1998) Terrestrial gastropod inventory of the Niagaran escarpment and Keweenaw volcanic belt
in Michigan’s Upper Peninsula. Final Report, Michigan Department of Natural Resources Small
Grants Program, Lansin
Nekola JC (1999) Terrestrial gastropod richness of carbonate cliff and associated habitats in the Great Lakes
region of North America. Malacologia 41:231–252
Nekola JC (2002) Effects of fire management on the richness and abundance of central North American
grassland land snail faunas. Animal Biodivers Conserv 25:53–66
Nekola JC (2003) Large-scale terrestrial gastropod community composition patterns in the Great Lakes
region of North America. Divers Distrib 9:55–71
794 Biodivers Conserv (2012) 21:781–795
123
Author's personal copy
210
Nekola JC (2004) Terrestrial gastropod fauna of northeastern Wisconsin and the southern Upper Peninsula
of Michigan. Am Malacol Bull 18:21–44
Nekola JC (2005) Latitudinal richness, evenness, and shell size gradients in eastern North American land
snail communities. Rec West Aust Mus Suppl 68:39–51
Nekola JC (2009) Big ranges from small packages: North American vertiginids more widespread than
thought. Tentacle 17:26–27
Nekola JC (2010) Acidophilic terrestrial gastropod communities of North America. J Mollus Stud
76:144–156
Nekola JC, Coles BF (2010) Pupillid land snails of eastern North America. Am Malacol Bull 28:29–57
Nekola JC, Smith TA (1999) Terrestrial gastropod richness patterns in Wisconsin carbonate cliff communities.
Malacologia 41:253–269
Nekola JC, Sˇizling AL, Boyer AG, Storch D (2008) Artifactions in the log-transformation of species
abundance distributions. Folia Geobot 43:259–468
Nekola JC, Coles BF, Bergthorsson U (2009) Evolutionary pattern and process in the Vertigo gouldii
(Mollusca: Pulmonata, Pupillidae) group of minute North American land snails. Mol Phylogenet Evol
53:1010–1024
Nowak CA, Abrahamson LP, Raynal DJ (1993) Powerline corridor vegetation management trends in New
York State: has a post-herbicide era begun? J Arboric 19:20–26
Oggier P, Zschokke S, Baur B (1998) A comparison of three methods for assessing the gastropod community
in dry grasslands. Pedobiologia 42:348–357
Pilsbry HA (1948) Land Mollusca of North America (North of Mexico). Academy of Natural Sciences of
Philadelphia, Monograph #3
Pokryszko BM (1987) On aphally in the Vertiginidae (Gastrocopta: Pulmonata: Orthurethra). J Conchol
32:365–375
Schilthuizen M, Lombaerts M (1994) Population structure and levels of gene flow in the Mediterranean land
snail Albinaria corrugata (Pulmonata: Clausiliidae). Evolution 48:577–586
Schreiber RK, Graves JH (1977) Powerline corridors as possible barriers to movements of small mammals.
Am Midl Nat 97:504–508
Shaffer ML, Samson FB (1985) Population size and extinction: a note on determining critical sizes. Am Nat
125:144–152
Simberloff D, Cox J (1987) Consequences and costs of conservation corridors. Conserv Biol 1:63–71
Sorrie BA, Weakley AS (2001) Coastal plain vascular plant endemics and phytogeographic patterns.
Castanea 66:50–82
Turgeon DD, Quinn JF Jr, Bogan AE, Coan EV, Hochberg FG, Lyons WG, Mikkelsen P, Neves RJ, Roper
CFE, Rosenberg G, Roth B, Scheltema A, Thompson FG, Vecchione M, Williams JD (1998) Common
and scientific names of aquatic invertebrates from the United States and Canada: Mollusks, 2nd edn.
Bethesda: American Fisheries Society Special Publication 26
Yahner RH, Morrell TE, Rachael JS (1989) Effects and edge contrast on depredation of artificial avian nests.
J Wildl Manag 53:1135–1138
Biodivers Conserv (2012) 21:781–795 795
123
Author's personal copy
211
Research Note
616
Conservation Biology, Pages 616–621
Volume 17, No. 2, April 2003
Effects of Rock Climbing on the Land Snail
Community of the Niagara Escarpment in Southern
Ontario, Canada
MICHELE A. MCMILLAN,* JEFFREY C. NEKOLA,† AND DOUGLAS W. LARSON*‡
*Department of Botany, University of Guelph, Guelph, Ontario N1G 2W1, Canada
†Department of Natural and Applied Sciences, University of Wisconsin, Green Bay, WI 54311, U.S.A.
Abstract: The cliffs of the Niagara Escarpment provide habitat for extremely diverse communities of land
snails that may be at risk as a result of recreational rock climbing. We examined the effects of rock climbing
on the density, richness, diversity, and community composition of snails on the Niagara Escarpment in southern
Ontario, Canada. We sampled from randomly selected climbed and unclimbed sections of cliffs on the plateau
(cliff edge), cliff face, and talus (cliff base). Snail density, richness, and diversity were lower along climbing
routes than in unclimbed areas, and community composition differed between climbed and unclimbed
samples. These results suggest that rock climbing has significant negative effects on all aspects of the snail community
on cliffs; therefore, we recommend the inclusion of gastropods in conservation plans for protected areas
containing cliffs.
Efectos del Alpinismo en la Comunidad de Caracoles Terrestres del Acantilado del Niagara en Ontario Meridional,
Canadá
Resumen: Las barrancas del acantilado del Niagara proveen hábitat para comunidades extremadamente
diversas de caracoles terrestres que pueden estar en riesgo debido al alpinismo recreativo. Examinamos los
efectos del alpinismo en la densidad, riqueza, diversidad y composición de comunidades de caracoles del acantilado
del Niagara en Ontario Meridional, Canadá. Tomamos muestras de las mesetas (bordes de acantilados),
la cara del acantilado y el talud (base del acantilado) de secciones de barrancas usadas y no usadas para el alpinismo
y seleccionadas al azar. La densidad, la riqueza y la diversidad de caracoles fueron más bajas en las
rutas escaladas que en aquellas áreas no escaladas y la composición de la comunidad difirió entre muestras
escaladas y no escaladas. Estos resultados sugieren que el alpinismo tiene impactos negativos significativos en
todos los aspectos de la comunidad de caracoles en acantilados; por lo tanto, recomendamos la inclusión de
gasterópodos en los planes de conservación para áreas protegidas que contengan acantilados.
‡Address correspondence to D.W. Larson, email dwlarson@uoguelph.ca
Paper submitted July 25, 2001; revised manuscript accepted May 31, 2002.
Introduction
Since its introduction to eastern North America 50 years
ago, recreational rock climbing has continually increased
in popularity, with the most dramatic increases occurring
over the past 20 years (Valis 1991). Cliff ecosystems contribute
greatly to the regional biodiversity of plants and
animals (Larson et al. 2000). Research conducted thus
far has demonstrated that rock climbing can lead to
decreased abundance and richness of vascular and nonvascular
plants and lichens (Nuzzo 1995; Herter 1996;
Nuzzo 1996; Kelly & Larson 1997; Camp & Knight 1998;
Farris 1998; McMillan & Larson 2002), suggesting that
the entire biotic community might be affected.
212
Conservation Biology
Volume 17, No. 2, April 2003
McMillan et al. Effects of Rock Climbing on Land Snails 617
The potential exists for rock climbing to have a particularly
strong influence on land snails because land snail
diversity is generally higher on limestone outcrops than
in other habitats (Burch 1962; Baur et al. 1995). In fact,
maximum global richness levels for terrestrial gastropods
are often associated with wooded carbonate cliffs and
talus slopes. This has been demonstrated in New South
Wales (Stanisic 1997), Scotland (Cameron & Greenwood
1991), Germany (Schmid 1966), Sweden (Waldén 1981),
the Appalachian mountains (Getz & Uetz 1994), and the
Great Lakes region of North America (Nekola 1999).
The Niagara Escarpment is a series of dolomitic limestone
outcrops that extends from the Bruce Peninsula to
the Niagara Region in Ontario and through Michigan,
Wisconsin, Illinois, and Iowa. These cliffs support the most
ancient forest ecosystem east of the Rocky Mountains (Larson
& Kelly 1991), with eastern white cedar trees (Thuja
occidentalis) living in excess of 900 years (Larson et al.
2000). These cliffs also support extremely high levels of
snail diversity. Up to 34 species have been observed in
soil collected from a surface area of 1000 m2
, up to 23
species in 1 m2
, and up to 21 species in 0.04 m2
(Nekola
1999; Nekola & Smith 1999). Unfortunately, these cliffs
are also exposed to the highest levels of recreational
rock climbing in Ontario (Bracken et al. 1991).
The importance of the Niagara Escarpment as a reservoir
for biotic biodiversity necessitates the creation of
formal conservation policies that protect this community.
Here, we document the effects of rock climbing
on the density, richness, diversity, and community composition
of land snails on dolomitic limestone cliffs in
Ontario.
Methods
Study Site
We conducted field work from late August to early October
1998 along a 35-km section of the Niagara Escarpment
near Milton, Ontario (Universal Transverse Mercator
Zone 17, 4806800–4818000 mN, 587500–591500 mE).
The public and privately owned cliffs we sampled are composed
of Silurian dolomitic limestone (Niagara Escarpment
Commission 1979) surrounded by a green belt of secondgrowth
deciduous forest that runs through larger areas of
urban development and agricultural land.
Climbed cliffs were sampled at Buffalo Crag (90 graded
routes), Rattlesnake Point (142 graded routes), Mt. Nemo
(236 graded routes), and Kelso (54 graded routes) conservation
areas. Unclimbed cliffs were sampled at Mt. Nemo
and Crawford Lake conservation areas and from an adjacent
cliff on the property of Milton Limestone Incorporated. Sections
of climbed and unclimbed cliffs were within 15 km of
one another.
Sampling Design
We determined the impact of climbing through the comparison
of “climbed” and “unclimbed” areas. A region was
classified as climbed when climbing manuals described
an established climbing route for that area (Bracken et al.
1991; Oates & Bracken 1997). A section of the cliff was
classified as unclimbed when no routes were described
for the area and when no other obvious signs of climbing
were present (e.g., implanted climbing hardware,
chalk marks on hand holds).
We sampled from 50 vertical belt transects, each 1 m
in width, that extended the height of the cliff. Twentyfive
climbed transects were chosen randomly from a pool
of 101 climbing routes, and 25 unclimbed transects were
chosen randomly from a pool of 106 transects. The pool
of climbed transects included all documented climbing
routes for the area with a difficulty rating between 5.7
and 5.9 and a star rating in the climbing manual (Bracken
et al. 1991). We chose these routes because they represented
the best routes for climbers with an intermediate
skill level. As such, they likely attract higher and more
regular amounts of climbing activity than other routes.
We restricted unclimbed transects to sections of the cliff
that were Ն7 m wide, that appeared physically suitable
for climbing, and that two experienced climbers and the
first author agreed would be rated between 5.7 and 5.9
if graded. Climbed and unclimbed transects were rejected
if they were Ͻ10 m in height or contained a roof or overhang
of more than 1 m, continuous water seeps, or loose
rocks.
Land snails were sampled from three positions within
each transect for a total of 150 observations: cliff face
(vertical surface), plateau (or cliff edge), and talus (cliff
base). These adjacent habitats were sampled in addition
to the cliff face because they are also subjected to climbing-related
disturbances such as being used as a place from
which to belay (hold the ropes for the climber), to anchor
climbing ropes, to store packs and other equipment, and
for climbers to rest between ascents.
Transects extended from 2 m beyond the cliff face on
the plateau to 2 m beyond the cliff base in the subtending
talus. We collected 250 mL of soil from each of the
three positions within each transect. In plateau and talus
habitats, we collected soil from the upper 5 cm of the
organic soil horizon within a 1 ϫ 2 m area. Because of
the paucity of soil on the cliff face, we collected soil samples
from ledges and cracks extending the entire height
of the cliff within the 1-m transect (ranging from 11.5 to
30 m). Although the sample extent for the cliff-face samples
was greater, the total surface area over which the
samples were taken did not exceed 2 m2
. To minimize the
effects of sampling on the cliff community, no ledge was
deprived of more than half its soil.
Each 1 ϫ 2 m quadrat sampled from the plateau and
talus was divided into 50 20 ϫ 20 cm subquadrats, and
213
618 Effects of Rock Climbing on Land Snails McMillan et al.
Conservation Biology
Volume 17, No. 2, April 2003
Table1.Frequency(%ofquadrats)inwhicheachspeciesoflandsnailwaspresentforunclimbed(U)andclimbed(C)quadratssampledin1998fromtheNiagaraEscarpmentinsouthern
Ontario,Canada.
Freq.(%)Freq.(%)
SpeciesGranka
Srankb
UC␹2c,d
SpeciesGranka
Srankb
UC␹2c,d
AnguispiraalternataG5S549.324.09.30*NesovitreabinneyanaG?S4-S528.02.716.64**
CarychiumexileG4?S3-S428.09.37.42ParavitreamultidentataG4-G5S2-S36.72.70.60
CarychiumnannodesG4-G5S1-S26.71.31.56Punctumminutissimum//85.32.750.03****
CochlicopalubricaG4-G5S514.70.09.81*Pupillamuscorum//1.332.023.23****
CochlicopalubricellaG4-G5S42.71.30.00Stenotremafraternumfraternum//5.30.02.31
Columellasimplex//28.00.022.15***StriaturaexiguaG4S417.01.39.53*
DiscuscatskillensisG3-G5S572.032.022.46****StriaturaferreaG4-G5S51.30.00.00
EuconulusfulvusG4-G5S4-S528.00.022.15***StriaturamiliumG4S4-S540.02.728.96****
EuconuluspolygyratusG?S46.70.03.31StrobilopslabyrinthicaG5S518.710.71.33
GastrocoptacontractaG5S514.718.70.19SuccineaovalisG5S3-S420.00.014.52**
GastrocoptacorticariaG4-G5S232.038.70.47Triodopsisalbolabris//2.70.00.51
GastrocoptaholzingeriG4-G5S2-S353.349.30.11Triodopsisdenotata//1.30.00.00
GastrocoptapentodonG5S561.317.328.61****TriodopsistridentataG5S3-S46.74.00.13
Glyphyaliniaindentata//4.01.30.26ValloniacostataG4-G5S50.01.30.00
GlyphyalinarhoadsiG5S3-S416.04.04.74ValloniagracilicostaG?S1-S265.372.00.50
GuppyasterkiiG4-G5SH5.31.30.83ValloniapulchellaG4-G5S50.01.30.00
HawaiiaminisculaG5S30.08.04.34VertigobollesianaG3S326.75.311.16*
HelicodiscusparallelusG5S512.06.70.71VertigogouldiG4-G5S378.733.329.46****
HelicodiscusshimekiG4S4-S52.76.70.60VertigoparadoxaG2-G4S2-S376.042.715.91**
MesomphixcupreusG5S31.30.00.00ZonitoidesarboreusG5S550.126.78.12
a
Globalconservationrank:G1,extremelyrare;G2,veryrare;G3,raretouncommon;G4,common;G5,verycommon;G?,unrankedornotlisted.
b
Provincialconservationranking:S1,extremelyrare;S2,veryrare;S3,raretouncommon;S4,common;S5,verycommon;SH,historicallypresentinOntariobutnotobservedrecently(typ-
icallynotrecordedfor20years)ornotlisted.
c
AdjustedusingYates’correctionforcontinuityassuggestedforcasesinwhichthereisasingledegreeoffreedom.TheBonferronicorrectionformultiplecomparisonswasusedtocalculate
significancelevels(MinistryofNaturalResources2002).
d
Probability:*pϽ0.1,**pϽ0.01,***pϽ0.001,****pϽ0.0001.
214
Conservation Biology
Volume 17, No. 2, April 2003
McMillan et al. Effects of Rock Climbing on Land Snails 619
the number of subquadrats containing exposed rock
was recorded. A 1 ϫ 2 m quadrat was similarly sampled
from the vertical center of each cliff-face transect. We
recorded this data prior to soil collection.
We collected soil samples in muslin bags. We placed
fresh samples in a drying oven at 80Њ C for 4–7 days and
then stored them at room temperature for several months.
We soaked each bag in water from 12–24 hours, after
which we removed the soil and washed it through a
graduated sieve series (4, 2, 1, and 0.42 mm). We discarded
particles that passed through the smallest sieve.
We returned the soil to muslin bags and redried them.
We then dry-sieved the soil through the graduated sieve
series and examined the factions of soil under an 8ϫ dissecting
microscope against a neutral brown background.
We removed intact shells from the soil with a moist paintbrush
and identified them to species (or subspecies) following
the nomenclature of Hubricht (1985). We checked our
identifications against the Hubricht Collection at the Field
Museum of Natural History and the second author’s reference
collection at the University of Wisconsin, Green
Bay. We counted and recorded the number of shells for
each species. We also calculated the species richness
and heterogeneity (Shannon-Wiener index) of each sample.
The reference collection from this study is housed
in the Department of Botany at the University of Guelph.
Statistical Analysis
We used a split-plot experimental design (Steel & Torrie
1980; Kuehl 1994) to compare the density, richness, and
heterogeneity of the sampled snails. We used climbing
intensity (climbed vs. unclimbed) as the whole-plot factor
and habitat (plateau, cliff face, and talus) as the splitplot
factor.
To minimize the impact of factors other than climbing
and habitat on the outcome of the statistical analysis, we
included cliff height, percent canopy cover, and compass
direction of the cliff face in the model as covariates.
The compass direction of the cliff was broken down
into a north component and an east component, with
the former being equal to the cosine (compass reading *
␲) and the latter being equal to sine (compass reading *
␲). We used these conversions because the use of an untransformed
compass reading would determine 1Њ and
359Њ to be very different from one another, when they
are both almost due north.
We analyzed the effects of rock climbing and habitat
type on snail density, richness, and heterogeneity
through analysis of variance (ANOVA) tables generated
by SAS computer software (Anonymous 1995). We used
the MIXED procedure due to the combination of covariates
within the split-plot design. We analyzed the results
for both main and simple effects, the main effects with
F statistics and the simple effects with Tukey’s procedure
(Steel & Torrie 1980).
We employed chi-square tests to determine the impact
of climbing and habitat type on the abundance of individual
species. We used the Bonferroni correction for multiple
comparisons (Kuehl 1994) and Yates’s correction for
continuity when there was only a single degree of freedom
(Steel & Torrie 1980).
Results
We retrieved 14,023 intact shells from the 37.5 L of soil
collected. These individuals belonged to 40 different species,
which represents almost half the diversity of land
snails in Ontario (88 taxa). The species varied in global
conservation ranking from G3 (Vertigo bollesiana) to G5
and in provincial ranking from S1/S2 (Carychium nannodes,
Vallonia gracilicosta) to S5 (Table 1).
Most of the species were Ͻ5 mm at their widest point
and are thus referred to as “minute” snails (Emberton
1995). We found a large amount of broken shell material
but did not include this in our analysis.
Shell density was over five times greater in unclimbed
samples than in climbed samples (78.26 vs. 14.86 individuals
per 250 mL of soil; p Ͻ 0.00001). Shell density
was significantly lower along climbing routes in each of
the three habitat positions (Fig. 1). In addition, shell density
varied significantly between the three habitat positions
( p Ͻ 0.005). The average density of shells found in
Figure 1. Mean density of land snails per 250 mL soil
sample collected from the Niagara Escarpment, Ontario,
Canada. Each bar represents the mean from 25
samples. Bars that share a letter code are not significantly
different from one another at ␣ ϭ 0.05.
215
620 Effects of Rock Climbing on Land Snails McMillan et al.
Conservation Biology
Volume 17, No. 2, April 2003
cliff face and talus samples (46.63 and 45.41 individuals
per 250 mL soil) was more than twice the average density
of shells found in plateau samples (19.15 individuals
per 250 mL soil; p Ͻ 0.005).
The species richness of land snails was significantly
lower along climbing routes (p Ͻ 0.00001). Unclimbed
areas supported almost two times more species than undisturbed
areas (9.84 vs. 5.20 species per 250 mL soil).
This significant reduction in richness along climbing routes
occurred in all habitat positions (Fig. 2). Similarly, climbed
areas showed significantly lower levels of Shannon-Wiener
diversity than did unclimbed areas (1.11 vs. 1.53; p Ͻ
0.001). Species richness was lowest on the plateau (4.94
species per 250 mL soil) and highest on the talus (10.56
species per 250 mL soil; p Ͻ0.00001).
Fourteen of the 40 species had significantly greater
frequency in unclimbed samples, whereas only one species
had a significantly greater frequency in climbed areas
(Table 1). The remaining species showed no statistical
preference for either climbed or unclimbed areas.
The proportion of subquadrats containing exposed
rock was significantly higher in climbed than unclimbed
quadrats (79.6% vs. 48.0%; p Ͻ 0.00001). This trend occurred
in each of the three habitat types, but was significant
for the plateau (10.6% vs. 66.8%; p Ͻ 0.0001) and
talus (40.2% vs. 73.9%; p Ͻ 0.0001) habitats only. Unclimbed
and climbed quadrats contained similar amounts
of exposed rock (93.1% vs. 98.0%; p ϭ 0.98).
Discussion
Our results suggest that rock climbing has strong negative
effects on the extremely diverse and abundant community
of land snails that normally occurs on undisturbed
cliffs of the Niagara Escarpment. Snail density, richness,
and diversity were all significantly lower in climbed areas
than in undisturbed areas, in spite of the fact that the
amount of soil collected was constant. Because the surface
area of soil was also lower in climbed areas, it is likely
that rock climbing has an even larger impact on land
snails than is suggested by our results.
The enormous local abundance of land snails and the
unusually high species richness in undisturbed sites suggest
that snails may be processing energy and matter at
high rates and therefore may be important natural components
of cliff-ecosystem food webs. Whatever their
role, their recolonization and the subsequent restoration
of function are likely to occur very slowly on denuded
cliffs in view of the slow rates of dispersal of such snails
( Baur & Baur 1994).
Rock climbing also appears to affect the composition of
land snail communities. There are several possible mechanisms
that might account for this change. First, the ledges
and cracks in the vertical rock surface that are the primary
habitat for cliff land snails are also the primary means by
which climbers ascend the cliff face. The use of these microsites
for hand and foot holds causes removal of unconsolidated
soil and organic matter, thereby decreasing the
amount and quality of available habitat for snails. In some
instances, climbers purposefully remove soil to increase
the security of the hold and to reduce the chance of slipping.
The soil that is left is then subject to direct pressure
from the hands and feet of climbers, which probably
causes soil compaction and a reduction in organic material,
as is found in other disturbed sites. Soil compaction and removal
from staging areas in the talus and at tie-off sites on
the plateau also decrease the amount of habitat and resources
available for these species. Future research is required
to determine the effects of climbing on physical soil
properties and chemistry.
Due to the detrimental effect of rock climbing on the
land snail community we observed in this study, we recommend
that management plans for the Niagara Escarpment
be modified to include specific policies regarding
recreational rock climbing. The most effective management
practices for the protection of gastropod species
on the Niagara Escarpment would limit the aerial extent
of climbing activity. We believe that land snail communities
on other cliffs may react similarly to climbing pressure,
because other cliffs around the world also support
extremely diverse communities of terrestrial gastropods
(Schmid 1966; Waldén 1981; Cameron & Greenwood
1991; Getz & Uetz 1994; Stanisic 1997; Nekola 1999).
We therefore recommend that gastropods also be included
in management policies for other areas containFigure
2. Mean number of land snail species present in
250 mL soil samples collected from the Niagara Escarpment,
Ontario, Canada. Each bar represents the mean
from 25 samples. Bars that share a letter code are not
significantly different from one another at ␣ ϭ 0.05.
216
Conservation Biology
Volume 17, No. 2, April 2003
McMillan et al. Effects of Rock Climbing on Land Snails 621
ing cliffs. Special attention should be given to areas in
which rare species are known to exist.
Acknowledgments
Funding for this project was provided by the Natural Sciences
and Engineering Research Council of Canada, the
Mountain Equipment Co-op, the Lattornell Graduate Student
Travel Fund, and the Access Fund. We thank X. Benjamin,
I. Brodo, J. Buck, P. Clarke, J. Crowe, A. Cullen, M.
deGruchy, C. Fietsch, S. Girard, K. Gormley, P. Kelly, C. A.
Lacroix, G. Lecanu, J. Lundholm, G. Mackie, U. Matthes, A.
Millward, A. Nassuth, S. Newman, R. Pal, R. Reader, J.
Stewart, S. Thevenet, and P. Wong for their help in the
field and laboratory. We thank Halton Region Conservation
Authority and Milton Limestone Ltd. for allowing us to
sample on their properties.
Literature Cited
Anonymous. 1995. SAS system for Windows. Version 6.11. SAS Institute,
Cary, North Carolina.
Baur, B., and A. Baur. 1994. Dispersal of the land snail Chondrina avenacea
on vertical rock walls. Malacological Review 27:53–59.
Baur, B., L. Fröberg, and A. Baur. 1995. Species diversity and grazing
damage in a calcicolous lichen community on top of stone walls in
Öland, Sweden. Annales Botanici Fennici 32:239–250.
Bracken, M., J. Barnes, and C. Oates. 1991. The escarpment: a climbers
guide. Borealis Press, Toronto.
Burch, J. B. 1962. How to know the eastern land snails. WM.C. Brown,
Dubuque, Iowa.
Cameron, R. A. D., and J. J. D. Greenwood. 1991. Some montane and
forest molluscan fauna from eastern Scotland: effects of altitude,
disturbance, and isolation. Pages 437–442 in Proceedings of the
tenth international malacological congress. Unitas Malacologica,
Tübingen, Germany.
Camp, R. J., and R. L. Knight. 1998. Effects of rock climbing on cliff
plant communities at Joshua Tree National Park, California. Conservation
Biology 12:1302–1306.
Emberton, K. C. 1995. Land-snail community morphologies of the
highest-diversity sites of Madagascar, North America, and New
Zealand, with recommended alternatives to height-diameter plots.
Malacologia 36:43–66.
Farris, M. A. 1998. The effects of rock climbing on the vegetation of
three Minnesota cliff systems. Canadian Journal of Botany 76:
1981–1990.
Getz, L. L., and G. W. Uetz. 1994. Species diversity of terrestrial snails
in the southern Appalachian mountains, U.S.A. Malacological Review
27:61–74.
Herter, W. 1996. Die Xerothermvegetation des Oberen Donautals—
Gefaerhrdung der Vegetation durchMensch und Wild sowie
Schutz—und Erhaltungsvorschlaege. Veroeffentlichungen des projekts
AAngewandte Oekologie@ PAoe 10. Veroef, Karlsruhe, Ger-
many.
Hubricht, L. 1985. The distributions of the native land mollusks of the
eastern United States. Fieldiana 24:1–191.
Kelly, P. E., and D. W. Larson. 1997. Effects of rock climbing on populations
of presettlement eastern white cedar (Thuja occidentalis)
on cliffs of the Niagara Escarpment, Canada. Conservation Biology
11:1125–1132.
Kuehl, R. O. 1994. Statistical principles of research design and analysis.
Wadsworth Publishing, Belmont, California.
Larson, D. W., and P. E. Kelly. 1991. The extent of old-growth Thuja
occidentalis on cliffs of the Niagara Escarpment. Canadian Journal
of Botany 69:1628–1636.
Larson, D. W., U. Matthes, and P. E. Kelly. 2000. Cliff ecology: pattern
and process in cliff ecosystems. Cambridge University Press, Cambridge,
United Kingdom.
McMillan, M. A., and D. W. Larson. 2002. The effects of rock climbing
on the vegetation of the Niagara Escarpment in Southern Ontario,
Canada. Conservation Biology 16:389–398.
Ministry of Natural Resources. 2002. Natural heritage information centre
list of Ontario mollusks. Ministry of Natural Resources, Peterborough,
Ontario, Canada. Available from http://www.mnr.gov.on.
ca/MNR/nhic/nhic.html (accessed March 2002).
Nekola, J. C. 1999. Terrestrial gastropod richness of carbonate cliff and
associated habitats in the great lakes region of North America. Malacologia
41:231–252.
Nekola, J. C., and T. M. Smith. 1999. Terrestrial gastropod richness patterns
in Wisconsin and carbonate cliff communities. Malacologia
41:253–269.
Niagara Escarpment Commission. 1979. Geology: the Niagara Escarpment.
Niagara Escarpment Commission, Georgetown, Ontario, Canada.
Nuzzo, V. A. 1995. Effects of rock climbing on cliff goldenrod (Solidago
sciaphila Steele) in Northwest Illinois. American Midland
Naturalist 133:229–241.
Nuzzo, V. A. 1996. Structure of cliff vegetation on exposed cliffs and
the effect of rock climbing. Canadian Journal of Botany 74:607–617.
Oates, C., and M. Bracken. 1997. A sport climbers guide to Ontario
limestone. Borealis Press, Toronto.
Schmid, G. 1966. Die mollusken des Spitzbergs. Der Spitzberg bei
Tübingen, Natur und Landschaftsschutzgebiete. Baden-Würtemburg
3:596–701.
Stanisic, J. 1997. An area of exceptional land snail diversity: the Macleay
Valley, north-eastern New South Wales. Memoirs of the Museum
of Victoria 56:441–448.
Steel, R. G. D., and J. H. Torrie. 1980. Principles and procedures of statistics:
a biometrical approach. 2nd edition. McGraw-Hill, New York.
Valis, T. 1991. Escarpment climbing history. Pages 3–6 in M. Bracken,
J. Barnes, and C. Oates, editors. The escarpment: a climbers guide.
Borealis Press, Toronto.
Waldén, H. W. 1981. Communities and diversity of land molluscs in Scandinavian
woodlands. I. High diversity communities in taluses and
boulder slopes in SW Sweden. Journal of Conchology 30:351–372.
217
53Animal Biodiversity and Conservation 25.2 (2002)
© 2002 Museu de Ciències NaturalsISSN: 1578–665X
Nekola, J. C., 2002. Effects of fire management on the richness and abundance of central North American
grassland land snail faunas. Animal Biodiversity and Conservation, 25.2: 53–66.
AbstractAbstractAbstractAbstractAbstract
Effects of fire management on the richness and abundance of central North American grassland land snail faunas.— The land
snail faunas from 72 upland and lowland grassland sites from central North America were analyzed. Sixteen of these had been
exposed to fire management within the last 15 years, while the remainder had not. A total of 91,074 individuals in 72 different
species were observed. Richness was reduced by approximately 30% on burned sites, while abundance was reduced by 50–90%.
One–way ANOVA of all sites (using management type as the independent variable), a full 2–way ANOVA (using management and
grassland type) of all sites, and a 2–way ANOVA limited to 26 sites paired according to their habitat type and geographic location,
demonstrated in all cases a highly significant (up to p < 0.0005) reduction in richness and abundance on fire managed sites.
Contingency table analysis of individual species demonstrated that 44% experienced a significant reduction in abundance on firemanaged
sites. Only six species positively responded to fire. Comparisons of fire response to the general ecological preferences
of these species demonstrated that fully 72% of turf–specialists were negatively impacted by fire, while 67% of duff–specialists
demonstrated no significant response. These differences were highly significant (p = 0.0006). Thus, frequent use of fire
management represents a significant threat to the health and diversity of North American grassland land snail communities.
Protecting this fauna will require the preservation of site organic litter layers, which will require the increase of fire return intervals
to 15+ years in conjunction with use of more diversified methods to remove woody and invasive plants.
Key words: Land snail, Biodiversity, Conservation, Fire management, Grassland, North America.
ResumenResumenResumenResumenResumen
Efectos de la gestión con fuego sobre la riqueza y abundancia de la fauna de caracoles terrestres de las praderas de
América del Norte.— Se analiza la fauna de caracoles terrestres de 72 praderas en mesetas y llanuras de la región
central de América del Norte. En 16 de ellas se habían efectuado intervenciones de incendio controlado durante los
últimos 15 años, mientras en el resto no. Se observaron un total de 91.074 individuos de 72 especies diferentes. La
riqueza en especies estaba reducida en un 30% en las áreas quemadas, mientras que la abundancia de individuos
estaba reducida en un 50–90%. Un ANOVA unidireccional de todas las áreas (usando como variable independiente el
tipo de intervención), un ANOVA bidireccional completo (usando el tipo de intervención y el tipo de pradera) en todas
las áreas y un ANOVA bidireccional limitado a 26 áreas agrupadas según su tipo de hábitat y localización geográfica,
demostró en todos los casos una reducción altamente significativa de la riqueza y de la abundancia (hasta p < 0,0005)
en áreas sometidas a incendio. Un análisis individual de las especies mediante tablas de contingencia demostró que
el 44% experimentaron una reducción significativa de su abundancia en las áreas quemadas. Sólo seis especies
respondieron positivamente al fuego. Comparando la respuesta al fuego con las preferencias ecológicas generales de
estas especies se demostró que al menos el 72% de las especialistas que viven en sustrato herbáceo fueron afectadas
negativamente por el fuego mientras que el 67% de las que viven en sustrato húmico no demostraron ninguna
respuesta significativa. Estas diferencias fueron altamente significativas (p = 0,0006). Así pues, el uso frecuente del
fuego representa una amenaza significativa para la salud y diversidad de las comunidades de caracoles terrestres de
las praderas de América del Norte. La protección de esta fauna requerirá la preservación de las capas de materia
orgánica y la ampliación de los intervalos entre las actuaciones de quema a periodos superiores a 15 años, así como
el uso de métodos más diversos para eliminar las plantas leñosas e invasivas.
Palabras clave: Caracol terrestre, Biodiversidad, Conservación, Gestión con fuego, Praderas, América del Norte.
(Received: 9 IV 02; Final acceptance: 18 VI 02)
Jeffrey C. Nekola, Dept. of Natural and Applied Sciences, Univ. of Wisconsin–Green Bay, Green Bay, Wisconsin
54311 USA.
E–mail: nekolaj@uwgb.edu
Effects of fire management on the
richness and abundance of
central North American
grassland land snail faunas
J. C. Nekola
218
54 Nekola
Introduction
Fire has long been implicated in the maintenance
of central North American grassland communities
(WEAVER, 1954; CURTIS, 1959). Numerous native
plant species respond to fire by increasing their
growth and reproductive rates (EHRENREICH &
AIKMAN, 1963; KUCERA & KOELLING, 1964; TOWNE &
OWENSBY, 1984). One of the most direct effects
of prairie fire is the removal of the soil mulch
layer, which has been implicated in the
‘stagnation’ of prairie plant communities through
the delay of initial spring growth, thinning of
grass stem density, and prevention of herbaceous
understory development (WEAVER & ROWLAND,
1952; KUCERA & KOELLING, 1964). Fire is also
thought to limit invasion of woody and exotic
plants into native prairie habitats (e.g., PAULY,
1985; ROOSA, 1984). For these reasons, prescribed
fire has become the management tool of choice
by prairie conservation groups throughout the
midwestern USA (COLLINS & WALLACE, 1990).
However, an increasing body of research suggests
that fire is not universally beneficial all prairie
biota. Fire depresses growth and reproductive rates
of native C3
prairie plants (DIX, 1960; HADLEY, 1970;
HILL & PLATT, 1975), which make up at least 50% of
the native flora north of 44o
N (STOWE & TEERI, 1978;
SIMS, 1988). Fire has also been implicated in the loss
and/or reduction of numerous native prairie
invertebrate species including Lepidoptera,
Homoptera, Hymenoptera, and Araneae (SWENGEL,
1996, 1998; HARPER et al., 2000). The effects of such
practices on prairie soil biodiversity are largely
undocumented. Combustion of mulch through
repeated fire episodes will remove the detritusphere,
one of the most important reservoirs for soil
biodiversity (COLEMAN & CROSSLEY, 1996). HARPER et
al. (2000) documented significant reductions in
Collembola following Illinois prairie fires. As the
soil fauna represents one of the largest species
pools in terrestrial ecosystems (BEHAN–PELLETIER &
NEWTON, 1999), the potential impacts of such
processes on total site biodiversity may be large.
Although not as hyper–diverse as bacteria,
fungi, nematodes, and arthropods, molluscs still
represent one of the more important components
of soil biodiversity (RUSSELL–HUNTER, 1983). Almost
600 species are known from eastern North
America (HUBRICHT, 1985), with up to 21 taxa cooccurring
within 400 cm2
microhabitats (NEKOLA
& SMITH, 1999). Most of these taxa represent
generalist detritivores that live in and on dead
organic material (BURCH & PEARCE, 1990)
As almost 90% of snails occur within 5 cm of
the soil surface (HAWKINS et al., 1998), protection
of this fauna will likely be tied to the fate of
mulch layers. Disturbances such as logging,
recreational or urban development, or bedrock
and soil removal cause dramatic changes in
woodland snail communities with duff soil
surfaces (NEKOLA, in press a). The impact of fire,
and associated detritusphere removal, on snail
communities is unclear. Fire has been suggested
to negatively influence the faunas of Aegean
islands (WELTER–SCHULTES & WILLIAMS, 1999),
Queensland fens (STANISIC 1996), and Tasmanian
woodlands (REGAN et al., 2001). However, FREST &
JOHANNES (1995) state that molluscs are able to
survive natural fires in northwestern North
America, and THELER (1997) argued that xeric
prairie faunas in Wisconsin owe their existence
to frequent fires that keep grassland areas
treeless. Unfortunately, no data was presented
by these various authors to validate such
conflicting statements.
To evaluate this issue, the richness and
abundance of land snails was quantitatively
compared between unburned and recently
(< 15 year) burned sites in the midwestern USA,
including 13 pairs of sites which possess similar
habitats and are spatiall proximate. From these,
the following questions will be considered: 1. Is
there a significant difference in land snail
community richness between burned and
unburned grasslands? 2. Is there a significant
difference in land snail abundance between
burned and unburned grasslands? 3. What species
show positive, negative, or no response to fire?
What ecological factors (if any) may help explain
these responses?
Materials and methods
Study Sites
Seventy two grassland sites were surveyed
between May 1996–November 2001 for terrestrial
molluscs across a 850 km extent of central North
America (fig. 1, table 1). Sites are generally
centered on northwestern Minnesota and
northeastern Iowa. Forty–two occur in Minnesota,
25 in Iowa, and 5 in Wisconsin. Thirty–two sites
represent upland habitats (including tallgrass
prairie, sand prairie, and bedrock glades), while
the remaining 40 are lowland sites (including
wet prairie, sedge meadow, and fens). Previous
use of fire management on sites was assessed by
either observing carbonized woody plant stems
or other debris on the ground surface, or through
interviews with site managers or other
knowledgeable individuals. No use of fire
management was noted from 56 sites (88% of
total), while 16 (22%) had been subjected to
some amount of prescribed burning. Eleven of
these burned sites occur in Minnesota, while the
remaining five occur in Iowa. The latitude–
longitude location of each site was determined
using either USGS 7.5 minute topographic maps
or a hand–held GPS.
Field Methods
Documentation of terrestrial gastropod faunas
from each site was accomplished by hand
219
Animal Biodiversity and Conservation 25.2 (2002) 55
collection of larger shells and litter sampling for
smaller taxa within 100–1,000 m2
areas that
contained examples of all major microhabitats
and were thus representative of the larger site.
The actual grain size employed was determined
by the minimum size necessary to emcompass all
microhabitats. Soil litter sampling was primary
used as it provides the most complete assessment
of grassland faunas (OGGIER et al., 1998). A single
site sample consisted of a composite of individual
soil litter subsamples of approximately 200 ml
collected from appropriate microhabitats. As
suggested by EMBERTON et al. (1996), litter
collections were made at places of high micromollusc
density, with a constant volume
(approximately 4 liters) being gathered from
each site. Sampling was generally comprised of:
1. Small blocks (ca. 125 cm3
) of turf; 2. Loose soil
and leaf litter accumulations under or adjacent
to shrubs, cobbles, boulders, and/or hummocks;
and 3. Other microsites supporting relatively
thick mulch layers.
LLLLLaboratory procedures
Samples were slowly and completely dried in
either a low–temperature soil oven (ca. 80–95o
C)
or in full sun in a greenhouse. Dried samples
were then soaked in water for 3–24 hours, and
subjected to careful but vigorous water
disaggregation through a standard sieve series
(ASTME 3/8" (9.5 mm), #10 (2.0 mm), #20 (0.85),
and #40 (0.425 mm) mesh screens). Sieved sample
fractions were then dried and passed again
through the same sieve series. These dry, resorted
fractions were hand picked against a neutralbrown
background. All shells and shell fragments
were removed.
All identifiable shells from each site were
assigned to species (or subspecies) using the
author’s reference collection and the Hubricht
Collection at the Field Museum of Natural History
(FMNH), with the total number of shells per
species per site being recorded. The total number
of unassignable, immature individuals was also
counted from each site. All specimens have been
catalogued and are housed in the author’s
collection at the University of Wisconsin–Green
Bay. Nomenclature generally follows that of
HUBRICHT (1985), with updates and corrections by
FREST (1990, 1991) and NEKOLA (in press b). The
general ecological preferences (turf specialist,
duff–specialist or generalist) of each species is
based upon analyses presented in NEKOLA (in
press a).
Statistical procedures
Differences in species richness and total shell
abundance between burned and unburned
grassland sites were analyzed via ANOVA. Initially,
1–way ANOVAs were preformed on the entire
dataset. However, the effect of fire may be
obscured in this analysis due to confounding
effects of habitat type and geographic location.
To help control for this, two additional sets of
ANOVAs were conducted. First, full 2–way ANOVAs
were calculated for all sites using grassland type
(upland vs. lowland) and management history
(burned vs. unburned) as the independent
variables. Second, 13 pairs of sites representing
closely similar habitats within the same geographic
region, but differing in their fire management
history, were selected. These site pairs are (first
site is burned, second is unburned): Malmberg
Prairie vs. Sandpiper Prairie; Pankratz Mesic Prairie
vs. Radium NE; Pankratz Low Prairie vs. Bjornson
WMA; Pankratz Fen vs. Faith South; Marcoux
WMA vs. Cyr Creek; East Park WMA vs. Goose
Lake; Felton Fen 1 vs. Ogema West; Waubun SE
vs. Eastlund Lake; Chicog vs. Tansen; Beemis Creek
vs. Hampton East; Fayette vs. Decorah Glade; Baty
Glade vs. Canton Glade; Brayton–Horsley vs.
Stapleton Church. A 2–way ANOVA without
interaction was then calculated for these sites,
with site pair identity and management type
representing independent variables.
Fig. 1. Map of study region, showing location
of surveyed grassland sites: F Unburned
upland; M Burned upland; H Unburned
lowland; F Burned lowland.
Fig. 1. Mapa del área de estudio que muestra
la localización de las praderas estudiadas:
F Meseta no quemada; M Meseta quemada;
H Llanura no quemada; F Llanura
quemada.
220
56 Nekola
Table 1. Location, grassland type, management, species richness and total number of collected
individuals from sample sites: GT. Grassland type; M. Management; R. Richness; I. Individuals.
Tabla 1. Localización, tipo de pradera, gestión, riqueza de especies y número total de individuos
recogidos en cada área de estudio: GT. Tipo de pradera; M. Gestión; R. Riqueza; I. Individuos.
State / County / Site Name Location GT M R I
Iowa
Allamakee County
Fish Farm Mounds 91°17'12" W – 43°27'13" N Upland Unburned 21 632
Williams Creek 3 91°29'1" W – 43°8'1" N Upland Unburned 23 2,708
Bremer County
Brayton–Horsley Fen 92°6'29" W – 42°48'36" N Lowland Burned 16 627
Buchanan County
Rowley Fen 91°51'7" W – 42°22'27" N Lowland Unburned 16 3,217
Rowley North Fen 91°51'3" W – 42°22'35" N Lowland Unburned 17 3,231
Rowley West Fen 91°54'40" W – 42°22'15" N Lowland Unburned 22 2,250
Cerro Gordo County
Buffalo Slough 93°11'11" W – 43°10'36" N Lowland Unburned 19 4,770
Chickasaw County
Stapelton Church Fen 92°6'14" W – 43°1'35" N Lowland Unburned 18 1,065
Clayton County
Postville Fen 91°33'59" W – 43°2'3" N Lowland Unburned 12 252
Turkey River Mounds 91°2'11" W – 42°42'46" N Upland Unburned 22 870
Clinton County
Maquoketa South 90°39'5" W – 42°1'12" N Upland Unburned 12 310
Dubuque County
Roosevelt Road 90°44'30" W – 42°32'55" N Upland Unburned 18 375
Fayette County
Fayette 91°47'28" W – 42°50'11" N Upland Burned 13 254
Turner Creek 1 Fen 91°52'11" W – 42°58'15" N Lowland Unburned 16 1,071
Floyd County
Beemis Creek 93°1'18" W – 42°59'39" N Upland Burned 8 192
Juniper Hill 92°59'2" W – 43°3'10" N Upland Unburned 12 206
Franklin County
Hampton East 93°8'13" W – 42°43'42" N Upland Unburned 15 381
Howard County
Hayden Prairie 92°23'4" W – 43°26'30" N Upland Burned 12 132
Staff Creek Fen 92°30'34" W – 43°26'41" N Lowland Unburned 15 1,599
Jackson County
Hamilton Glade 90°34'9" W – 42°4'23" N Upland Unburned 15 340
Jones County
Canton Glade 90°59'52" W – 42°10'46" N Upland Unburned 19 446
Linn County
Baty Glade 91°39'14" W – 42°11'44" N Upland Burned 16 345
Paris Fen 91°35'42" W – 42°13'40" N Lowland Unburned 12 1,254
221
Animal Biodiversity and Conservation 25.2 (2002) 57
State / County / Site Name Location GT M R I
Mitchell County
Stone School Fen 92°38'11" W – 43°22'49" N Lowland Unburned 18 2,926
Winneshiek County
Decorah Glade 91°46'11" W – 43°18'55" N Upland Unburned 18 605
Minnesota
Becker County
Audubon South Fen 95°58'47" W – 46°49'58" N Lowland Unburned 15 1,816
Callaway North 95°55'22" W – 47°3'57" N Upland Unburned 19 362
Greenwater Lake Fen 95°29'59" W – 46°59'20" N Lowland Unburned 20 2,132
Ogema West Fen 95°55'59" W – 47°6'32" N Lowland Unburned 16 5,001
Straight Lake 95°18'40" W – 46°58'40" N Upland Unburned 13 281
Beltrami County
Fourtown Fen 95°18'21" W – 48°15'56" N Lowland Unburned 14 1,403
Clay County
Barnesville WMA 96°17'34" W –46°43'5" N Upland Unburned 11 469
Barnesville WMA Fen 96°17'38" W – 46°43'9" N Lowland Unburned 13 436
Bjornson WMA 96°21'24" W – 46°45'44" N Lowland Unburned 14 436
Bluestem Prairie 96°28'45" W – 46°51'18" N Upland Burned 15 371
Felton Prairie 1 Fen 96°26'21" W – 47°3'51" N Lowland Burned 15 2,370
Felton Prairie 2 Fen 96°26'20" W – 47°4'0" N Lowland Unburned 14 3,131
Felton Prairie 96°26'1" W – 47°3'34" N Upland Unburned 5 63
Tansen 96°11'17" W – 46°42'14" N Upland Unburned 10 146
Clearwater County
Bagley Lake Fen 95°14'35" W – 47°45'41" N Lowland Unburned 9 126
Filmore County
Vesta Creek 91°45'0" W – 43°40'5" N Upland Unburned 21 1,151
Houston County
Twin Pines Farm 91°22'45" W – 43°44'48" N Upland Unburned 24 591
Yucatan Twp. 91°38'28" W – 43°43'23" N Upland Unburned 20 765
Mahnomen County
Eastlund Lake 95°47'5" W – 47°26'41" N Upland Unburned 13 490
Mahnomen North 95°58'8" W – 47°21'27" N Upland Unburned 18 806
Waubun SE 95°54'55" W – 47°9'57" N Lowland Unburned 18 2,915
Waubun SE 95°55'4" W – 47°10'5" N Upland Burned 8 220
Marshall County
East Park WMA 96°16'44" W – 48°31'57" N Lowland Burned 14 735
Florian WMA 96°33'21" W – 48°26'33" N Lowland Unburned 17 3,923
Radium NE 96°32'38" W – 48°16'49" N Upland Unburned 12 493
Norman County
Faith South 96°5'12" W – 47°15'42" N Lowland Unburned 16 3,047
Prairie Smoke Dunes 96°18'22" W – 47°27'44" N Upland Unburned 3 19
Sandpiper Prairie 96°24'22" W – 47°14'43" N Lowland Unburned 12 1,261
Table 1. (Cont.)
222
58 Nekola
The central tendencies in these various
relationships were graphically represented via box
plots. In box plots, the central line represents the
median of the sample, the margins of the box
represent the interquartile distances, and the fences
represent 1.5 times the interquartile distances. For
data having a Gaussian distribution, approximately
99.3% of the data will fall inside of the fences
(VELLEMAN & HOAGLIN, 1981). Outliers falling outside
of the fences are shown with asterisks.
The average number of individuals per species
per site was determined for burned uplands,
unburned uplands, burned lowlands, and
unburned lowlands. The average proportion of
each species in the total community for each site
was calculated for each management/habitat type.
These proportions were placed in rank order, and
plotted vs. log–transformed frequency to create
dominance–diversity curves (WHITTAKER, 1975).
The response of individual species to fire was
analyzed through log–linear modelling, as
predicted values in the associated contingency
table were sparse (< 5) in more than one–fifth of
cells (ZAR, 1984). The total number of individuals
within all burned or unburned sites was compared
to a null expectation of equal occurrence
Pennington County
Goose Lake 96°27'44" W – 48°5'37" N Lowland Unburned 17 996
Higenbotham WMA 96°17'41" W – 48° 0'22" N Lowland Unburned 22 1,114
Sanders Fen 96°21'9" W – 48°3'52" N Lowland Unburned 15 2,218
Polk County
Chicog Prairie 96°23'14" W – 47°35'53" N Upland Burned 2 153
Erskine North 96°0'3" W – 47°44'17" N Lowland Unburned 19 741
Gulley Fen 95°37'22" W – 47°48'13" N Lowland Unburned 19 2,032
Malmberg Prairie 96°49'25" W – 47°43'52" N Lowland Burned 7 563
Pankratz Prairie 96°26'37" W – 47°43'23" N Lowland Burned 12 314
Pankratz Prairie 96°26'31" W – 47°43'23" N Upland Burned 11 159
Pankratz Prairie 96°26'48" W – 47°43'9" N Lowland Burned 7 190
Red Lake County
Crane WMA 95°42'49" W – 47°53'27" N Lowland Unburned 15 425
Cyr Creek 96°16'12" W – 47°48'10" N Lowland Unburned 22 1,845
Marcoux WMA 96°13'27" W – 47°47'55" N Lowland Burned 12 688
Winona County
Great River Bluffs 91°23'28" W – 43°56'53" N Upland Burned 19 788
Wisconsin
Green Lake County
Berlin Fen 88°54'20" W – 43°57'47" N Lowland Unburned 20 3,454
Manitowoc County
Point Beach St. Forest 87°30'40" W – 44°11'52" N Upland Unburned 4 6
Walworth County
Bluff Creek Fen 88°40'54" W – 42°48'2" N Lowland Unburned 20 1,106
Washington County
Allenton Fen 88°18'25" W – 43°22'42" N Lowland Unburned 20 2,858
Waushara County
Bass Lake Fen 89°16'59" W – 44°0'16" N Lowland Unburned 19 1,466
Table 1. (Cont.)
State / County / Site Name Location GT M R I
223
Animal Biodiversity and Conservation 25.2 (2002) 59
Table 2. List of encountered species, with their average abundances from burned and unburned
sites. P–values are based on log-likelihood ratio tests, with the two–tailed significance threshold
being lowered to p = 0.000347 to account for the 72 tested species. General ecological preferences
are based on NEKOLA (in press a). Turf–specialists represent those species demonstrating at least
a p < 0.05 preference to sites with a friable upper A soil horizon supporting few living plant roots.
Turf specialists represent those species demonstrating at least a p < 0.05 preference to sites with
an upper A soil horizon that is bound together with living plant roots. Species without preferences
were too infrequently encountered by NEKOLA (in press a) to be statistically assigned: AvU. Average
unburned; Abb. Abundance burned; Ecp. Ecological preference (T. Turf; D. Duff; G. Generalist.)
Tabla 2. Lista de especies detectadas, con sus abundancias medias en áreas quemadas y no
quemadas. Los valores de P se basan en tests de cociente de probabilidad logarítmica, con el umbral
de significación de doble cola reducido hasta p = 0,000247 para las 72 especies estudiadas. Las
preferencias ecológicas generales se basan en NEKOLA (in press a). Las especies que viven en sustrato
herbáceo presentan una preferencia de al menos p < 0,05 por las zonas con horizonte de tierra
friable superior A provisto de escasas raíces de plantas vivas. Las especies que viven en sustrato
herbáceo presentan una preferencia de al menos p < 0,05 por las zonas cuyo horizonte se mantiene
unido por raíces de plantas vivas. Las especies sin preferencias resultaron excesivamente infrecuentes
según NEKOLA (in press a) para consignarlas estadísticamente: AvU. Medias en áreas no quemadas;
Abb. Abundancia en áreas quemadas; Ecp. Preferencia ecológica (T. Sustrato herbáceo; D. Sustrato
húmico; G. Generalista.)
Species AvU Abb P–value Ecp
Negative responses
Carychium exiguum (Say, 1822) 273.607 90.250 0.0000000 T
Carychium exile H. C. Lea, 1842 5.196 0.000 0.0000000 D
Catinella exile (Leonard, 1972) 58.446 1.625 0.0000000 T
Catinella "vermeta" 1.482 0.000 0.0000001 T
Deroceras laeve (Müller, 1774) 4.036 1.188 0.0000050 G
Discus cronkhitei (Newcomb, 1865) 16.143 5.000 0.0000000 G
Euconulus alderi (Gray, 1840) 43.054 8.375 0.0000000 T
Gastrocopta contracta (Say, 1822) 21.232 5.875 0.0000000 G
Gastrocopta holzingeri (Sterki, 1889) 36.196 17.938 0.0000000 D
Gastrocopta pentodon (Say, 1821) 9.661 4.875 0.0000072 D
Gastrocopta procera Gould, 1840 2.304 0.000 0.0000000 T
Gastrocopta rogersensis Nekola & Coles, 2001 5.518 0.062 0.0000000 T
Gastrocopta similis (Sterki, 1909) 21.518 4.125 0.0000000 T
Gastrocopta tappaniana (C. B. Adams, 1842) 112.929 33.375 0.0000000 T
Hawaiia minuscula (A. Binney, 1840) 36.286 23.062 0.0000000 G
Helicodiscus n. sp. 1.071 0.062 0.0001509 T
Nesovitrea binneyana (Morse, 1864) 1.500 0.000 0.0000001 D
Nesovitrea electrina (Gould, 1841) 80.179 22.688 0.0000000 T
Oxyloma retusa (I. Lea, 1834) 21.268 8.750 0.0000000 T
Pomatiopsis lapidaria (Say, 1817) 1.196 0.000 0.0000021 –
Punctum minutissimum (I. Lea, 1841) 26.286 10.500 0.0000000 D
Punctum n. sp. 41.982 12.625 0.0000000 T
Punctum vitreum H. B. Baker, 1930 16.536 2.812 0.0000000 D
Stenotrema leai leai (A. Binney) 2.696 0.375 0.0000004 T
Striatura milium (Morse, 1859) 0.714 0.000 0.0002470 G
Strobilops affinis Pilsbry, 1893 74.911 4.312 0.0000000 T
Triodopsis multilineata (Say, 1821) 1.571 0.062 0.0000016 G
Vallonia pulchella (Müller, 1774) 23.321 11.688 0.0000000 T
224
60 Nekola
Species AvU Abb P–value Ecp
Vertigo elatior Sterki, 1894 41.875 5.875 0.0000000 T
Vertigo milium (Gould, 1840) 59.357 36.375 0.0000000 T
Vertigo morsei Sterki, 1894 3.589 0.375 0.0000000 T
Vitrina limpida Gould, 1850 1.143 0.000 0.0000035 G
No response
Anguispira alternata (Say, 1817) 0.018 0.000 0.5622176 D
Catinella avara (Say, 1824) 7.286 8.000 0.5185276 T
Cochlicopa lubrica (Müller, 1774) 0.464 0.000 0.0031253 D
Cochlicopa lubricella (Porro, 1838) 1.714 1.938 0.6798546 G
Columella simplex (Gould, 1841) 0.071 0.188 0.4068651 D
Discus catskillensis (Pilsbry, 1898) 0.357 0.000 0.0095467 D
Euconulus fulvus (Müller, 1774) 3.429 1.625 0.0040433 D
Euconulus polygyratus (Pilsbry, 1899) 0.018 0.000 0.5622176 D
Gastrocopta abbreviata (Sterki, 1909) 0.000 0.062 0.2039785 –
Gastrocopta armifera (Say, 1821) 1.232 1.188 0.9197258 G
Glyphyalinia indentata (Say, 1823) 2.732 3.250 0.4543545 D
Haplotrema concavum (Say, 1821) 0.411 0.000 0.0054455 G
Hawaiia n. sp. 2.571 1.750 0.1616094 T
Helicodiscus inermis H. B. Baker, 1929 0.679 1.812 0.0089186 –
Helicodiscus parallelus (Say, 1817) 5.393 6.438 0.2831711 G
Helicodiscus shimeki Hubricht, 1962 0.286 0.000 0.0204383 D
Helicodiscus singleyanus (Pilsbry, 1890) 0.375 1.125 0.0247064 G
Hendersonia occulta (Say, 1831) 0.036 0.062 0.7606162 D
Mesodon clausus clausus (Say, 1821) 0.054 0.000 0.3154693 D
Oxyloma peoriensis (Wolf in Walker, 1892) 0.125 0.000 0.1251903 –
Pupoides albilabris (C. B. Adams, 1821) 8.393 7.062 0.2324971 T
Stenotrema barbatum (Clapp, 1904) 0.107 0.000 0.1557229 D
Stenotrema fraternum fraternum (Say, 1824) 0.054 0.062 0.9262276 D
Succinea indiana Pilsbry, 1905 0.000 0.188 0.0277912 –
Succinea ovalis Say, 1817 0.143 0.188 0.7834904 D
Triodopsis alleni (Wetherby in Sampson, 1883) 0.071 0.000 0.2464148 D
Vallonia gracilicosta Reinhardt, 1883 11.500 11.062 0.7459095 D
Vertigo arthuri (von Martens, 1884) 0.643 0.000 0.0005065 –
Vertigo gouldi (A. Binney, 1843) 0.018 0.000 0.5622176 D
Vertigo nylanderi Sterki, 1909 0.036 0.000 0.4124393 T
Vertigo ovata Say, 1822 5.750 4.000 0.0472312 T
Vertigo tridentata Wolf, 1870 0.375 0.062 0.0738709 D
Zonitoides arboreus (Say, 1816) 4.143 4.625 0.5650976 D
Zonitoides nitidus (Müller, 1774) 0.464 0.000 0.0031253 T
Positive response
Gastrocopta corticaria (Say, 1816) 0.732 3.500 0.0000003 D
Strobilops labyrinthica (Say, 1817) 9.661 22.375 0.0000000 D
Vallonia costata (Müller, 1774) 1.804 6.438 0.0000000 G
Vallonia parvula Sterki, 1892 8.250 13.250 0.0001267 T
Vallonia perspectiva Sterki, 1892 2.143 5.625 0.0000055 D
Vertigo pygmaea (Draparnaud, 1801) 0.000 0.562 0.0001385 G
Tabla 2. (Cont.)
225
Animal Biodiversity and Conservation 25.2 (2002) 61
Fig. 2. Box–plot diagram of the response of species richness and abundance to management type
on all sampled sites.
Fig. 2. Diagrama de la respuesta en riqueza y abundancia de especies al tipo de actuación
desarrollado en todas las áreas estudiadas.
3030303030
2020202020
1010101010
00000
SpeciesrichnessSpeciesrichnessSpeciesrichnessSpeciesrichnessSpeciesrichness
Unburned BurnedUnburned BurnedUnburned BurnedUnburned BurnedUnburned Burned Unburned BurnedUnburned BurnedUnburned BurnedUnburned BurnedUnburned Burned
Management typeManagement typeManagement typeManagement typeManagement type Management typeManagement typeManagement typeManagement typeManagement type
6,0006,0006,0006,0006,000
5,0005,0005,0005,0005,000
4,0004,0004,0004,0004,000
3,0003,0003,0003,0003,000
2,0002,0002,0002,0002,000
1,0001,0001,0001,0001,000
00000
TTTTTotal#ofindividualsotal#ofindividualsotal#ofindividualsotal#ofindividualsotal#ofindividuals
p = 0.001p = 0.001p = 0.001p = 0.001p = 0.001 p = 0.008p = 0.008p = 0.008p = 0.008p = 0.008
frequency. This null expectation was calculated
by assigning 88% of all encountered individuals
to unburned sites, with the remaining 22% to
burned sites. This procedure was necessary as
the number of unburned vs. burned sites was
not balanced (88% vs. 22%). A two–tailed
significance threshold was employed so that
species with positive and negative responses to
fire could both be identified. As these analyses
were repeated for each species, a Bonferroni
correction was used to adjust the significance
threshold. Differences between fire responses
across the three general ecological preference
types were documented via a contingency table,
with significance being estimated using both
log–linear modelling and Fisher’s Exact test.
Results
These grassland habitats were generally found
to support a diverse and abundant land snail
fauna. A total of 91,074 individuals in 72 different
species were recovered from the 72 surveyed
sites (tables 1, 2). The number of species per
each 4 l litter sample ranged from two (Chicog
gravel prairie) to 24 (Twin Pines Farm sandstone
glade). Average richness ranged from roughly 15
in upland sites, to 17 in lowland. Snail abundance
per site ranged from 6 (Point Beach State Forest
dunes) to 5,001 (Ogema West fen). Average
abundance ranged from roughly 500 in upland
sites to 2,000 in lowlands.
One–way ANOVA, using all sites, demonstrated
that both species richness (p = 0.001) and
abundance (p = 0.008) were significantly lower
on sites that had experienced fire management
(fig. 2). Median species richness was approximately
18 on unburned vs. 12 on burned sites. Likewise,
median shell abundance was 1,000 on unburned
vs. 300 on burned sites.
Full 2–way ANOVA, using all sites and
considering both management type and habitat
type (upland vs. lowland) as independent
variables, demonstrated a highly significant
(p = 0.002) reduction (approximately 30%) in
species richness in both upland and lowland sites
(fig. 3). Habitat type and the interaction between
habitat and fire history were not significant
predictors (p = 0.209 and p = 0.628, respectively).
Likewise, a significant (p = 0.010) reduction in
shell abundance (50–70%) was noted on burned
sites (fig. 3). In this case, however, habitat type
was a more significant (p < 0.0005) predictor,
with lowlands having 4–10 times the number of
shells as uplands. Additionally, a marginally
significant (p = 0.088) interaction between
management and habitat was observed, with
the reduction appearing to be roughly 50%
greater in lowlands.
Two–way ANOVA restricted to the 26 paired
sites (fig. 4) demonstrated that even after
blocking of variation due to site pair identity, a
significant reduction in richness (p < 0.0005) and
abundance (p = 0.015) still occurred on fire–
managed sites.
226
62 Nekola
Fig. 3. Box–plot diagram of the response of species richness and abundance to management and
habitat type on all sampled sites.
Fig. 3. Diagrama de la respuesta en riqueza y abundancia de especies al tipo de actuación
desarrollado y el hábitat en todas las áreas estudiadas.
Comparison of dominance–diversity diagrams
for these sites (fig. 5) demonstrates that both
burned upland and lowland sites have truncated
curves, with the rarest 40–50% of species being
much less common as compared to unburned
sites. However, the more common species appear
to have largely similar dominance–diversity
diagrams.
Contingency table analysis of individual species
responses to fire (table 2) indicate that 32 (44%)
experience a significant reduction in abundance
on fire–managed sites, even following use of a
Bonferroni–corrected two–tailed significance
threshold (p = 0.000347). Only six species (8%)
demonstrated positive responses to fire, while
the remaining 34 (47%) demonstrated no
ANOVANOVANOVANOVANOVA results: fire p = 0.002A results: fire p = 0.002A results: fire p = 0.002A results: fire p = 0.002A results: fire p = 0.002
habitat p = 0.209habitat p = 0.209habitat p = 0.209habitat p = 0.209habitat p = 0.209
interaction p = 0.628interaction p = 0.628interaction p = 0.628interaction p = 0.628interaction p = 0.628
ANOVANOVANOVANOVANOVA results:A results:A results:A results:A results: fire p = 0.010fire p = 0.010fire p = 0.010fire p = 0.010fire p = 0.010
habitat p < 0.0006habitat p < 0.0006habitat p < 0.0006habitat p < 0.0006habitat p < 0.0006
interaction p = 0.088interaction p = 0.088interaction p = 0.088interaction p = 0.088interaction p = 0.088
Lowland sitesLowland sitesLowland sitesLowland sitesLowland sites Upland sitesUpland sitesUpland sitesUpland sitesUpland sites
6,0006,0006,0006,0006,000
5,0005,0005,0005,0005,000
4,0004,0004,0004,0004,000
3,0003,0003,0003,0003,000
2,0002,0002,0002,0002,000
1,0001,0001,0001,0001,000
00000
3333300000
2020202020
1010101010
00000
SpeciesrichnessSpeciesrichnessSpeciesrichnessSpeciesrichnessSpeciesrichness
TTTTTotal#ofindividualsotal#ofindividualsotal#ofindividualsotal#ofindividualsotal#ofindividuals
Unburned BurnedUnburned BurnedUnburned BurnedUnburned BurnedUnburned Burned Unburned BurnedUnburned BurnedUnburned BurnedUnburned BurnedUnburned Burned
Management typeManagement typeManagement typeManagement typeManagement type Management typeManagement typeManagement typeManagement typeManagement type
227
Animal Biodiversity and Conservation 25.2 (2002) 63
Fig. 4. Box–plot diagram of the response of species richness and abundance to management on
26 sites paired by habitat type and geographic location.
Fig. 4. Diagrama de la respuesta en riqueza y abundancia de especies a la actuación desarrollada
en 26 áreas emparejadas según el tipo de hábitat y localización geográfica.
Fig. 5. Dominance–diversity curve for upland/lowland sites which have been burned/unburned.
Fig. 5. Curva de la dominancia–diversidad para las mesetas/llanuras que hayan sido quemadas/
no quemadas.
3030303030
2020202020
1010101010
00000
SpeciesrichnessSpeciesrichnessSpeciesrichnessSpeciesrichnessSpeciesrichness
6,0006,0006,0006,0006,000
5,0005,0005,0005,0005,000
4,0004,0004,0004,0004,000
2,0002,0002,0002,0002,000
3,0003,0003,0003,0003,000
1,0001,0001,0001,0001,000
00000
TTTTTotal#ofindividualsotal#ofindividualsotal#ofindividualsotal#ofindividualsotal#ofindividuals
Unburned BurnedUnburned BurnedUnburned BurnedUnburned BurnedUnburned Burned UnburnedUnburnedUnburnedUnburnedUnburned BurnedBurnedBurnedBurnedBurned
Management typeManagement typeManagement typeManagement typeManagement type Management typeManagement typeManagement typeManagement typeManagement type
ANOVANOVANOVANOVANOVA results:A results:A results:A results:A results: ANOVANOVANOVANOVANOVA results:A results:A results:A results:A results:
Pairs p = 0.001Pairs p = 0.001Pairs p = 0.001Pairs p = 0.001Pairs p = 0.001 Pairs p = 0.009Pairs p = 0.009Pairs p = 0.009Pairs p = 0.009Pairs p = 0.009
Fire p < 0.0005Fire p < 0.0005Fire p < 0.0005Fire p < 0.0005Fire p < 0.0005 Fire p = 0.015Fire p = 0.015Fire p = 0.015Fire p = 0.015Fire p = 0.015
Lowland sitesLowland sitesLowland sitesLowland sitesLowland sites Upland sitesUpland sitesUpland sitesUpland sitesUpland sites
–1–1–1–1–1
–4–4–4–4–4
–7–7–7–7–7
–10–10–10–10–10
00000 1010101010 2020202020 3030303030 4040404040 5050505050 6060606060 00000 10 20 3010 20 3010 20 3010 20 3010 20 30 4040404040 5050505050 6060606060
Rank orderRank orderRank orderRank orderRank order Rank orderRank orderRank orderRank orderRank order
BurnedBurnedBurnedBurnedBurned
UnburnedUnburnedUnburnedUnburnedUnburned
ln(%frecuency)ln(%frecuency)ln(%frecuency)ln(%frecuency)ln(%frecuency)
228
64 Nekola
significant changes in population size. Contingency
table analysis of ecological preference vs. fire
response indicated that fully 72% of turf–
specialists were negatively impacted by fire
(table 3). However, only 22% of duff–specialists
exhibited negative responses. While 67% of duff–
specialists demonstrated no significant response
to fire only 24% of turf–specialists were
unaffected. Generalist species demonstrated little
discernable trend to fire, with seven decreasing,
two increasing, and five with no response. Log–
likelihood ratio and Fisher’s Exact tests both
indicated these differences as being highly
significant (p = 0.0006 and p = 0.004, respectively).
Discussion
These data clearly indicate that fire management
causes significant reductions in land snail
community richness and abundance in both
upland and lowland grasslands throughout a
significant section of the tallgrass prairie biome
in central North America. At a species–level, fire
most strongly impacts the rarest species, and
causes significant population reductions in 44%
of the 72 encountered taxa. These negative
impacts were most strongly felt in turf–specialists,
where almost 75% experienced significant
reductions. Thus, statements regarding the
benign nature of fire on snail populations (FREST
& JOHANNES, 1995), and the beneficial impact of
fire on North American grassland faunas (THELER,
1997) can be proven false. Rather, frequent use
of fire management appears to represent a
significant threat to the health and diversity of
North American grassland land snails.
It is not possible through these analyses to
definitively identify the factors that directly lead
to these impacts. However, at least part of the
answer must lay in grassland detritusphere
removal. This will lead to direct mortality, as the
great majority of land snails are limited to this
layer (HAWKINS et al., 1998). As land snail
abundance (BERRY, 1973), diversity (CAIN, 1983;
LOCASCIULLI & BOAG, 1987), and composition
(CAMERON & MORGAN–HUWS, 1975; BAUER et al.,
1996; BARKER & MAYHILL, 1999) is often positively
correlated with litter depth, detritusphere removal
would be expected to have a strong negative
impact on land snail community structure.
Redevelopment of an equilibrium thickness of
organic detritus takes at least five years in
southern Plains grasslands (KUCERA & KOELLING,
1964), with even longer intervals being required
in more northern locations (HILL & PLATT, 1975).
The optimal interval between fires for land snails
might be even longer, depending upon the time
required for more refractory plant debris (such as
lignified grass stems) to break down, allowing a
complete suite of decompositional microhabitats
to develop. Litter architecture is known to effect
snail community composition in forests of Virginia
(BURCH, 1956), British Columbia (CAMERON, 1986),
and Puerto Rico (ÁLVAREZ & WILLIG, 1993) and
grasslands of England (YOUNG & EVANS, 1991). It
should thus not be surprising that in the current
data set, sites burned up to 15 years ago have
maintained lowered land snail richness and
abundance as compared to unburned sites.
As grassland land snails presumably evolved in
conjunction with natural fire regimes, it is also
intriguing to note that turf–specialists experienced
the most severe negative impacts to fire. If fire
was a common process structuring central North
American grasslands, evolution should have selected
for individuals that were more tolerant of, or
favored by, this disturbance. Like other native
grassland invertebrate groups (SWENGEL, 1996;
HARPER et al., 2000), land snails in the presettlement
landscape may have been able to tolerate fires by
being able to easily recolonize from source pools
in adjacent unburned areas. Even when such
adjacent source pools are present, recolonization
may take over a dozen years (MÄND et al., 2001). In
modern landscapes, where grasslands are highly
fragmented and surrounded by agricultural, urban,
or forest habitats, such recolonization has become
much more difficult. Thus, turf–specialist taxa may
continue to decrease in burned grasslands due to a
lack of recolonization sources, while generalist
and duff-specialist woodland taxa, which are more
common in the surrounding landscape, may be
able to maintain their populations through mass
effect (SHMIDA & ELLNER, 1984).
The depression of land snail richness and
abundance following fire episodes, the length
of time required to redevelop a mature
detritusphere, and the greater sensitivity of turf–
specialist taxa to fire casts doubt on the wide–
Table 3. Contingency table analysis of fire
response vs. general ecological preferences.
Log–likelihood ratio p = 0.000634; Fisher’s
Exact Test p = 0.004 (Ecological preferences:
T. Turf; D. Duff; G. Generalist.)
Tabla 3. Análisis de la tabla de contingencia
de la respuesta al fuego frente a las
preferencias ecológicas generales. Logaritmo
de la razón de verosimilitudes p = 0,000634;
Test exacto de Fisher p = 0,004 (Preferencias
ecológicas: T. Sustrato herbáceo; D. Sustrato
húmico; G. Generalista.)
Ecological preferences
Fire response T D G
Negative 18 6 7
None 6 18 5
Positive 1 3 2
229
Animal Biodiversity and Conservation 25.2 (2002) 65
held belief (e.g., PAULY, 1985) that North American
grasslands should be burned at 2–6 year intervals.
Rather, these data support the contention that
presettlement return intervals ranged between
20–30 years (SIMS, 1988). These data also strongly
suggest that other factors, such as large herbivore
grazing (COLLINS et al., 1998) and periodic drought
(BORCHERT, 1950), may have also played essential
roles in keeping prairies treeless, as these
processes do not lead to the wholesale
detritusphere removal.
Protecting the health of North American
grassland land snail populations will require the
preservation of mulch layers on sites. Such efforts
will also help protect a large percentage of the
entire grassland soil biota. The detritusphere can
only be protected if more realistic fire return
intervals (20–30 years) are adopted by conservation
agencies, and used in conjunction with more
diversified approaches towards woody and
invasive plant removal. Activities like grazing,
haying, and hand cutting/pulling will not cause
widespread removal of the detritusphere, and
should thus be more compatible with land snail
(and soil biodiversity) conservation.
Acknowledgements
Alyssa Barnes, Tracy Kuklinski, J. J. Schiefelbein
and Angela Sette helped processed many soil
litter samples, and assisted in shell counting.
Additional assistance in litter processing was also
provided by students of the Land Snail Ecology
Practicum at the University of Wisconsin – Green
Bay. Funding was provided by the Minnesota
Nongame Wildlife Tax Checkoff and Minnesota
State Park Nature Store Sales through the
Minnesota Department of Natural Resources
Natural Heritage and Nongame Research Program.
References
ÁLVAREZ, J. & WILLIG, M. R., 1993. Effects of treefall
gaps on the density of land snails in the
Luquillo Experimental Forest of Puerto Rico.
Biotropica, 25: 100–110.
BARKER, G. M. & MAYHILL, P. C., 1999. Patterns of
diversity and habitat relationships in terrestrial
mollusc communities of the Pukeamaru
Ecological District, northeastern New Zealand.
Journal of Biogeography, 26: 215–238.
BAUER, B., JOSHI, J., SCHMID, B., HÄNGGI, A., BORCARD,
D., STARY, J., PEDROLI–CHRISTEN, A., THOMMEN, G.
H., LUKA, H., RUSTERHOLZ, H. P., OGGIER, P.,
LEDERGERBER, S. & ERHARDT, A., 1996. Variation
in species richness of plants and diverse groups
of invertebrates in three calcareous grasslands
of the Swiss Jura mountains. Revue Suisse de
Zooligie, 103: 801–833.
BEHAN–PELLETIER, V. & NEWTON, G., 1999. Linking
soil biodiversity and ecosystem function–the
taxonomic dilemma. Bioscience, 49: 149–153.
BERRY, F. G., 1973. Patterns of snail distribution at
Ham Street Woods National Nature Reserve,
East Kent. Journal of Conchology, 28: 23–35.
BOUCHERT, J. R., 1950. The climate of the central
North American grassland. Annals of the
Association of American Geographers, 40: 1–39.
BURCH, J. B., 1956. Distribution of land snails in
plant associations in eastern Virginia. The
Nautilus, 70: 60–64.
BURCH, J. B. & PEARCE, T. A., 1990. Terrestrial
Gastropoda. In: Soil biology guide: 201–309 (D.
L. Dindal, Ed.). John Wiley & Sons, New York.
CAIN, A. J., 1983. Ecology and ecogenetics of
terrestrial molluscan populations. In: The
Mollusca, Volume 6, Ecology: 597-647 (W. D.
Russell–Hunter, Ed.). Academic Press, New York.
CAMERON, R. A. D., 1986. Environment and
diversities of forest snail faunas from coastal
British Columbia. Malacologia, 27: 341–355.
CAMERON, R. A. D. & MORGAN–HUWS, D. I., 1975.
Snail faunas in the early stages of a chalk
grassland succession. Biological Journal of the
Linnean Society, 7: 215–229.
COLEMAN, D. C. & CROSSLEY, D. A. JR., 1996.
Fundamentals of soil ecology. Academic Press,
New York.
COLLINS, S. L., KNAPP, A. K., BLAIR, J. M. & STEINAUER,
E. M., 1998. Modulation of diversity by grazing
and mowing on native tallgrass prairie.
Science, 280: 745–747.
COLLINS, S. L. & WALLACE, L. L., 1990. Fire in North
American tallgrass prairie. University of
Oklahoma Press, Norman, Oklahoma.
CURTIS, J. T., 1959. The vegetation of Wisconsin.
U. Wisconsin Press, Madison.
DIX, R. L., 1960. The effects of burning on the
mulch structure and species composition of
grasslands in western North Dakota. Ecology,
41: 49–56.
EHRENREICH, J. H. & AIKMAN, J. M., 1963. An
ecological study of the effect of certain
management practices on native prairie in
Iowa. Ecological Monographs, 33: 113–134.
EMBERTON, K. C., PEARCE, T. A. & RANDALANA, R.,
1996. Quantitatively sampling land–snail
species richness in Madagascan rainforests.
Malacologia, 38: 203–212.
FREST, T. J., 1990. Final report, field survey of
Iowa spring fens, contract #65–2454. Iowa
Department of Natural Resources, Des Moines.
– 1991. Summary status reports on eight species
of candidate land snails from the Driftless
Area (Paleozoic Plateau), Upper Midwest. Final
Report, Contract #301–01366, USFWS Region
3, Ft. Snelling, Minnesota.
FREST, T. J. & JOHANNES, E. J., 1995. Interior Columbia
Basin mollusc species of special concern. Final
Report, Contract #43–0E00–4–9112, Interior
Columbia Basin Ecosystem Management Project,
Walla Walla, Washington.
HADLEY, E. B., 1970. Net productivity and burning
responses of native eastern North Dakota
230
66 Nekola
prairie communities. American Midland
Naturalist, 84: 121–135.
HARPER, M. G., DIETRICH, C. H., LARIMORE, R. L. &
TESSENE, P. A., 2000. Effects of prescribed fire
on prairie Arthropods: an enclosure study.
Natural Areas Journal, 20: 325–335.
HAWKINS, J. W., LANKESTER, M. W. & NELSON, R. R.
A., 1998. Sampling terrestrial gastropods using
cardboard sheets. Malacologia, 39: 1–9.
HILL, G. R. & PLATT, W. J., 1975. Some effects of
fire upon a tallgrass prairie plant community
in northwestern Iowa. In: Prairie: a multiple
view: 103–113 (M. K. Wali, Ed.). University of
North Dakota Press, Grand Forks.
HUBRICHT, L., 1985. The distributions of the native
land molluscs of the eastern United States.
Fieldiana, n.s., 24: 1–191.
KUCERA, C. L. & KOELLING, M., 1964. The influence of
fire on composition of central Missouri prairie.
American Midland Naturalist, 72: 142–147.
LOCASCIULLI, O. & BOAG, D. A., 1987. Microdistribution
of terrestrial snails (Stylommatophora) in forest
litter. Canadian Field Naturalist, 101: 76–81.
MÄND, R., EHLVEST, A. & KIRISTAJA, P., 2001. Land
snail in an afforested oil–shale mining area.
Proceedings of the Estonian Academy of
Sciences, Biology and Ecology, 50: 37–41.
NEKOLA, J. C. (in press a). Large–scale Terrestrial
Gastropod Community Composition Gradients
in the Great Lakes region of North America.
Journal of Biogeography.
– (in press b). Terrestrial gastropod fauna of
northeastern Wisconsin and the southern
Upper Peninsula of Michigan. American
Malacological Bulletin.
NEKOLA, J. C. & SMITH, T. A., 1999. Terrestrial
gastropod richness patterns in Wisconsin
carbonate cliff communities. Malacologia, 41:
253–269.
OGGIER, P., ZSCHOKKE, S. & BAUR, B., 1998. A
comparison of three methods for assessing
the gastropod community in dry grasslands.
Pedobiologia, 42: 348–357.
PAULY, W. R., 1985. How to manage small prairie
fires. Dane County Environmental Council,
Madison, Wisconsin.
REGAN, T. J., REGAN, H. M., BONHAM, K., TAYLOR, R.
J. & BURGMAN, M. A., 2001. Modelling the impact
of timber harvesting on a rare carnivorous land
snail (Tasmaphena lamproides) in northwest
Tasmania, Australia. Ecological Modelling, 139:
253–264.
ROOSA, D. M., 1984. Fury on the prairie. Iowa
Conservationist, 43: 13.
RUSSELL–HUNTER, W. D., 1983. Overview: planetary
distribution of and the ecological constraints
upon the mollusca. In: The Mollusca, Volume
6, Ecology: 1–27 (W. D. Russell–Hunter, Ed.).
Academic Press, New York.
SHMIDA, A. & ELLNER, S., 1984. Coexistence of
plant species with similar niches. Vegetatio,
58: 29–55.
SIMS, P. L., 1988. Grasslands. In: North American
terrestrial vegetation: 266–286 (M. G. Barbour
& W. D. Billings, Eds.). Cambridge University
Press, New York.
STANISIC, J., 1996. New land snails from boggomoss
environments in the Dawson Valley, southeastern
Queensland (Eupulmonata: Charopidae
and Camaenidae). Memoirs of the Queensland
Museum, 39: 343–354.
STOWE, L. G. & TEERI, J. A., 1978. The geographic
distribution of C4
species of the Dicotyledonae
in relation to climate. American Naturalist,
112: 609–621.
SWENGEL, A. B., 1996. Effects of fire and hay
management on the abundance of prairie
butterflies. Biological Conservation, 76: 73–85.
– 1998. Comparisons of butterfly richness and
abundance measures in prairie and barrens.
Biodiversity and Conservation, 7: 1,639–1,659.
THELER, J. L., 1997. The modern terrestrial
gastropod (land snail) fauna of western
Wisconsin’s hill prairies. The Nautilus, 110:
111–121.
TOWNE, G. & OWENSBY, C., 1984. Long–term effects
of annual burning at different dates in
ungrazed Kansas tallgrass prairie. Journal of
Range Management, 37: 392–397.
VELLEMAN, P. F. & HOAGLIN, D. C., 1981. Applications,
basics, and computing of exploratory data
analysis. Addison–Wesley Press, Reading,
Massachusetts.
WEAVER, J. E., 1954. North American prairie.
Johnson Publishing Company, Lincoln, Nebraska.
WEAVER, J. E. & ROWLAND, N. W., 1952. Effects of
excessive natural mulch on development, yield,
and structure of native grassland. Botanical
Gazette, 114: 1–19.
WELTER–SCHULTES, F. W. & WILLIAMS, M. R., 1999.
History, island area and habitat availability
determine land snail species richness of Aegean
islands. Journal of Biogeography, 26: 239–249.
WHITTAKER, R. H., 1975. Communities and
ecosystems. MacMillan Publishing, New York.
YOUNG, M. S. & EVANS, J. G., 1991. Modern land
mollusc communities from Flat Holm, South
Glamorgan. Journal of Conchology, 34: 63–70.
ZAR, J. H., 1984. Biostatistical analysis. Prentice–
Hall, Inc., Englewood Cliffs, New Jersey.
231
Section VII: Human macroecology and sustainability
The goal of creating a human society which exists in a sustainable fashion with the finite Earth
requires not only that our actions be assessed for individual farms, villages, and cities, but also
collectively across humanity. However, such a macroecological perspective is largely missing
from sustainability studies. Because human
civilization is a complex adaptive system which is
maintained far from thermodynamic equilibrium, it
requires energy inputs via vast quantities of
increasingly exhausted fossil fuel stocks. It also
requires other essential and non-substitutable
commodities such as metal ores, radio-nucleotides,
rare earths, phosphate fertilizer, arable land, and
fresh water that are becoming evermore scarce. The
dynamics of such systems are highly unpredictable.
Small perturbations can cause wholesale changes,
including total collapse. Ensuring that global
civilization does not suffer this fate will require that
valuable insights from the natural – and not just
social – sciences be considered. Our most recent
paper is a ecological consideration of the history
food storage and transportation, published in
Bioscience in the summer of 2015 (see right). The
next project considers the ecological dynamics of
human innovation, in particular whether high
population density cities are needed to generate the
technological advancements needed to support the
ever growing human population.
Representative Publications
[number of citations as of October 27, 2017]
Hammond, S.T., J.H. Brown, J.R. Burger, T.P. Flanagan, T.S. Fristoe, N. Mercado-Silva, J.C.
Nekola & J.G. Okie. 2015. Food spoilage, storage and transport: implications for a
sustainable future. Bioscience. 65:758–768. [12]
Brown, J.H., J.R. Burger, W.R. Burnside, M. Chang, A.D. Davidson, T.S. Fristoe, M.J.
Hamilton, S.T. Hammond, A. Kodric-Brown, N. Mercado-Silva, J.C. Nekola & J.G.
Okie. 2014. Macroecology meets macroeconomics: resource scarcity and global
sustainability. Ecological Engineering. 65:24-32. [32]
Nekola, J.C., C.D. Allen, J.H. Brown, J.R. Burger, A.D. Davidson, T.S. Fristoe, M.J. Hamilton,
S.T. Hammond, A. Kodric-Brown, N. Mercado-Silva & J.G. Okie. 2013. The
Malthusian-Darwinian dynamic and the trajectory of civilization. Trends in Ecology and
Evolution. 28:127-130. [41]
Nekola, J.C., J.H. Brown, A. Kodric-Brown & J.G. Okie. 2013. Global sustainability vs. the
MDD: a response to Rull. Trends in Ecology and Evolution. 28: 444. [3]
Relationship between energetic costs and speed
for goods transportation fueled by animal
metabolism (black dots) and extrametabolic
processes in marine (blue), terrestrial (blue), and
aerial (red) domains. The trajectory of the
relationship over time is shown within each of the
latter three groups by best-fit vectors.
232
Burger, J.R., J.H. Brown, C.D. Allen,, W.R. Burnside, A.D. Davidson, T.S. Fristoe, M.J.
Hamilton, N. Mercado-Silva, J.C. Nekola, J.G. Okie & W. Zuo. 2012. The
macroecology of sustainability. PLoS Biology. 10:e1001345. [84]
Brown, J.H., W.R. Burnside, A.D. Davidson, J.P. DeLong, W.C. Dunn, M.J. Hamilton, J.C.
Nekola, J.G. Okie, N. Mercado-Silva, W.H. Woodruff & W. Zuo. 2011. Energetic
Limits to Economic Growth. Bioscience. 61:19-26. [179]
Smith, V.H., G.D. Tilman & J.C. Nekola. 1999. Eutrophication impacts of excess nutrient
inputs on fresh water, marine, and terrestrial ecosystems. Environmental Pollution. 100:
179-196. [2017]
233
The Malthusian–Darwinian dynamic and the trajectory
of civilization
Jeffrey C. Nekola1
, Craig D. Allen2
, James H. Brown1,3
, Joseph R. Burger1
,
Ana D. Davidson1,4
, Trevor S. Fristoe1
, Marcus J. Hamilton1,3,5
, Sean T. Hammond1
,
Astrid Kodric-Brown1
, Norman Mercado-Silva1,6
, and Jordan G. Okie1,7
1
Department of Biology, University of New Mexico, Albuquerque, NM, USA
2
US Geological Survey, Fort Collins Science Center, Jemez Mountains Field Station, Los Alamos, NM, USA
3
Santa Fe Institute, Santa Fe, NM, USA
4
Department of Ecology and Evolution, Stony Brook University, Stony Brook, NY, USA
5
Department of Anthropology, University of New Mexico, Albuquerque, NM, USA
6
School of Natural Resources and the Environment, University of Arizona, Tucson, AZ, USA
7
School of Earth and Space Exploration, Arizona State University, Tempe, AZ, USA
Two interacting forces influence all populations: the
Malthusian dynamic of exponential growth until resource
limits are reached, and the Darwinian dynamic
of innovation and adaptation to circumvent these limits
through biological and/or cultural evolution. The specific
manifestations of these forces in modern human
society provide an important context for determining
how humans can establish a sustainable relationship
with the finite Earth.
Malthus, Darwin, and population dynamics
In 1798 Thomas Malthus laid out the concept of exponential
population growth that became the foundation of demography
and population biology. He noted that the ‘increase of
population is necessarily limited by the means of subsistence.’
Population growth can thus continue only as long as
environmental conditions remain favorable. As numbers
increase, sooner or later environmental limits cause birth
rates to decrease and/or death rates to increase, ultimately
leading to an end to population growth. These concepts
profoundly influenced Charles Darwin half a century later:
because more offspring are born than can survive, only the
fittest individuals reproduce and pass their superior traits
on to their offspring. The result is adaption or innovation in
the form of either genetic or cultural evolution.
The Malthusian dynamic pushes a population to increase
until it reaches its environmental limits. The Darwinian
dynamic pushes against these limits by incorporating new
traits and technologies that enhance survival and reproduction.
There are restrictions to this Malthusian–Darwinian
Dynamic (MDD), however: it is logically, physically, and
biologically impossible for exponential growth to continue
indefinitely within a finite world.
Ecological and historical perspective on the rise
of human civilization
Humans are an exceptional species. Our population has
increased almost continuously from less than a million
individuals in sub-Saharan Africa 50 000 years ago to a
current population of 7 billion spanning the entire globe.
The human population is projected to reach between 9 and
10 billion by 2050 [1]. In the process, humans have created
complex social, technological, and economic systems. We
have transformed the atmosphere, water, land, and biodiversity
of the planet. We have become the most dominant
single species the earth has ever seen. Our success to
date is a consequence of our ability to continually develop
new innovations that push back environmental limits
and to transfer this information across generations
(Figure 1). Operating in this way, the MDD has led
to rapid cultural evolution and made possible the transition
from hunter–gatherer to agricultural to industrial–
technological–informational economies.
A central feature of human ecology has been the positive
feedback between growth and innovation. As populations
grew and aggregated into larger and more complex social
groups, more information was acquired and processed. This
led to new technologies that further pushed back ecological
limits, allowing for continued population growth. The result
has been an ascending spiral of exponential processes feeding
back on each other: population growth and aggregation
begot technological innovation, which in turn allowed for
more resource extraction and a greater ability to overcome
ecological constraints, begetting still more population
growth and socioeconomic development [2].
Our ability to evade local resource shortages and population
crashes through trade, migration, and innovation
has allowed for continued growth not just of the global
population but also of its resource use and economic
activity [3]. For more than 200 years, so-called Malthusians
[4] have argued, however, that this cannot indefinitely
continue because essential resources supplied by
the finite Earth will ultimately become limiting. This
Malthusian perspective has historically been countered
by so-called Cornucopians, who have argued that there is
no hard limit to human population size and economic
activity because human ingenuity and technological innovation
provide an effectively infinite capacity to increase
resource supply [5].
Forum: Science & Society
Corresponding author: Nekola, J.C. (jnekola@unm.edu).
Keywords: resource limitation; behavioral constraints; sustainability science; macroecology;
evolutionary biology.
127
Author's Personal Copypy234
Malthus warned of population and economic collapse
in the British Isles because the exponentially increasing
population would eventually outstrip linearly increasing
food supplies. Although millions did die from widespread
famines in Ireland and the Scottish Highlands in the
mid-19th century, more widespread starvation was
averted because of newly acquired agricultural technologies
and emigration of people to overseas colonies [2].
As a result, the British economy did not collapse and by
1900 had become the dominant global empire. In the
1960s and 1970s Malthusians again warned that the
global human population was nearing its environmental
limits. Outbreaks of famine and disease did occur,
but advances in agriculture, medicine, technology, and
commerce – most notably the green revolution and economic
globalization – pushed back the limits again and
our population has more than doubled from approximately
three billion in 1960 to more than seven billion
today [1].
Our ability to evade previous environmental limits
through the MDD should not imply, however, that such
outcomes are inevitable. First, although the MDD has
served as the engine for our extraordinary success, it also
contains within it the potential seeds for our ultimate
downfall. Motives driven by the MDD, such as selfishness
and cheating, benefit individuals at the expense of society
as a whole (Box 1). Second, because natural selection can
only operate on current conditions, there is little tendency
to recognize approaching environmental limits and
curb growth before resources are overexploited [6]. Third,
the assertion that adaptation and innovation will always
prevent collapse – because they have in the past – is
logically untenable and akin to the statistical fallacy of
extrapolating beyond a data set. Whether we wish to
acknowledge it or not, there are hard limits to innovation:
according to the first law of thermodynamics, we cannot
create energy or matter; according to the second law of
thermodynamics, large energy inputs are required to
maintain highly organized systems.
The ruins of Mohenjo Daro, Mesopotamia, Egypt,
Greece, Rome, the Maya, Hohokam, Angkor Wat, and
Easter Island are enduring evidence that many earlier
societies were unable to innovate their way out of local
limits and therefore collapsed despite attaining dense
populations and advanced cultures. Although the proximate
causes of these declines are debated and undoubtedly
differ, the ultimate causes lie in an inability to sustain
resource supplies and protect against parasites, diseases,
and other human groups [7].
Until now, both Malthusians and Cornucopians have
been correct: some populations have crashed and cultures
have vanished, but our species has endured because
these events have been localized. However, behavioral
changes and technological innovations over the last century
now intricately interconnect us in a single global
society. As a result, local perturbations currently have
the ability to reverberate across all of humanity. For
instance, the 2008 meltdown of the USA real estate
and mortgage markets and the 2011 To¯hoku earthquake
and tsunami in Japan both led to global interruptions
in economic activity, impacting most individuals in
some way.
Within the context of our now highly globalized society,
the essential question is how much potential exists for the
Bronze:
IrrigaƟon:
CiƟes:
Wheel:
Bow & arrow:
Animal domesƟcaƟon:
5000
5000
5700
6000
10000
10000
0
10000 8000 6000 4000
Years before present
GlobalpopulaƟon(Billions)
2000 0
1
2
3
4
5
6
7
Steam engine:
Windmill:
Gunpowder:
Water mills:
Roads:
Metal plows:
300
1200
1200
2300
2500
2500
Green revoluƟon:
Computer:
AnƟbioƟcs:
PlasƟcs:
Airplane:
Internal combusƟon engine:
Vaccine:
40
60
70
90
100
150
200
years ago
TRENDS in Ecology & Evolution
Bronze:
IrrigaƟon:
CiƟes:
Wheel:
Bow & arrow:
Animal domesƟcaƟon:
5000
5000
5700
6000
10000
10000
Steam engine:
Windmill:
Gunpowder:
Water mills:
Roads:
Metal plows:
300
1200
1200
2300
2500
2500
Green revoluƟon:
Computer:
AnƟbioƟcs:
PlasƟcs:
Airplane:
Internal combusƟon engine:
Vaccine:
40
60
70
90
100
150
200
years ago
Figure 1. Global human population over the last 10 000 years. Data are based on
estimates from [11], the US census bureau, and http://en.wikipedia.org/wiki/
World_population_estimates. Average estimates of population size across these
sources are plotted. Examples of major innovations that have helped to expand
human carrying capacity (based in part on [12]) are listed for three time periods in
years before present (ybp): 10 000–3000 ybp (green), 3000–300 ybp (orange), and
300 ybp–present (red).
Box 1. The MDD and challenges to global sustainability
Although the MDD has spurred innovations and adaptations that
have allowed us to colonize the entire planet, it has also shaped our
biological makeup, selecting for traits and behaviors that favor
selfishness and may ultimately make us ill-equipped to act in
concert to address global issues.
Natural selection generates the tendency for individuals to
preferentially promote themselves and their families. Social groups
often establish acceptable behaviors to minimize the inherent
conflict between such individual self-interest and group welfare.
Although disobeying group rules can result in increased individual
fitness, this is counteracted by the detection and punishment of
cheaters by social groups. Whether through luck or drive, some will
still manage to possess a disproportionate share of resources.
Although the degree of inequality can be moderated, highly
asymmetric wealth distributions have proven robust to the best
efforts of social philosophies and political movements to impose
more egalitarian patterns [13].
All of these factors contribute to the tragedy of the commons,
which occurs when selfish individuals reap the benefits from shared
resources while spreading the costs across the group [14]. The
increase in global atmospheric CO2 concentrations and the reduction
in global fisheries stocks (Figure 2) both represent pressing
global environmental issues related to this mechanism. In the case
of CO2 emissions, individuals benefit from using carbon-based
energy while global society is burdened by the costs related to
increased atmospheric CO2 concentrations. Similarly, a single
fishing vessel has no incentive to curb harvests because the costs
of reduced stocks are paid for by humanity as a whole. Efforts to
implement and enforce international regulations to moderate these
two activities have failed because of the selfishness of countries and
individuals in combination with the difficulty of detecting cheating
at the global scale [15].
Forum: Science & Society Trends in Ecology & Evolution March 2013, Vol. 28, No. 3
128
Author's Personal Copypy235
Darwinian side of the MDD to allow for continued adaptation
and innovation to push back against global scale
constraints. Many are now asking questions such as ‘What
are the limits to growth?’ and ‘What will happen when
these limits are met?’ [5,8].
The road forward
We cannot provide definitive answers to these questions.
Contemporary human civilization is a complex adaptive
system, maintained far from thermodynamic equilibrium
largely via the throughput of vast quantities of increasingly
exhausted fossil fuel stocks [9]. This system also requires
other essential and non-substitutable commodities such as
metal ores, radionucleotides, rare earth elements, phosphate
fertilizer, arable land, and fresh water that are
becoming ever more scarce [10]. The dynamics of such
systems are highly unpredictable. Small perturbations
can cause wholesale changes, including total collapse.
The bad news is that the MDD has left humans ill
prepared to make the necessary ecological and behavioral
changes required to avoid civilization collapse. The universal
underlying biological imperatives of the MDD lie at the
root of many of the most challenging impediments to longterm
sustainability: exponential population growth, exploitation
of all available resources, and the expression
of behaviors that promote the competitive abilities of
individuals, their families, and social groups over the
species as a whole.
However,thegoodnewsisthattheMDDmayalsoprovide
valuable insights into potential solutions from both natural
(in particular evolutionary biology and ecology) and social
(in particular economics and sociology) science perspectives.
We must recognize that a sustainable future will ultimately
require: (i) negative population growth for a number of
generations, followed by zero growth; (ii) a steady-state
economy based on sustainable use of renewable energy
and material resources; and (iii) new social norms that favor
the welfare of the entire global population over that of
specific individuals and groups. It is also essential that we
recognize that humanity has not yet evolved the genetic or
cultural adaptations needed to accomplish these tasks.
Our exceptional brains give us the ability to appreciate
the present situation and envision alternate futures before
catastrophe occurs. The challenge will be to facilitate a
rapid cultural evolution that, for the good of the entire
species, rewards individual sacrifices in fitness and quality
of life. Genuine collaboration between natural and social
scientists will be essential to inform society as a whole –
and policy makers specifically – of this difficult but necessary
adaptation required to accommodate our species in a
finite and now full world.
References
1 Bongaarts, J. (2009) Human population growth and the demographic
transition. Philos. Trans. R. Soc. B: Biol. Sci. 364, 2985–2990
2 Boserup, E. (1981) Population and Technological Change, University of
Chicago Press
3 Hall, C.A.S. and Klitgaard, K.A. (2011) Energy and the Wealth of
Nations: Understanding the Biophysical Economy, Springer
4 Ehrlich, P.R. (1968) The Population Bomb, Sierra Club/Ballantine
Books
5 Simon, J.L. (1981) The Ultimate Resource, Princeton University Press
6 Levin, S.A. (2000) Fragile Dominion: Complexity and the Commons,
Perseus Publishing
7 Tainter, J.A. (1988) The Collapse of Complex Societies, Cambridge
University Press
8 Heinberg, R. (2011) The End of Growth: Adapting to Our New Economic
Reality, New Society Publishers
1960 1970 1980
Year
Wildfisheriesharvest(tonnes)
AtmosphericconcentraƟonofCO2(ppm)
GlobalhumanpopulaƟon
1990 2000 2010
2e+6
3e+6
4e+6
5e+6
6e+6
7e+6
8e+6
1e+7
2e+7
3e+7
4e+7
5e+7
6e+7
7e+7
8e+7
300
400
380
360
340
320
Human populaƟon
Key:
Fisheries
CO2
TRENDS in Ecology & Evolution
Figure 2. The trajectories of atmospheric CO2 and wild fisheries harvest in relation to global population since 1955. Note that the increase in CO2 concentration has
accelerated, whereas fisheries harvest reached a peak in the 1990s and has since declined. The world population size is from the World Resources Institute (http://
earthtrends.wri.org). The wild fisheries harvest data are from the FAO Fishery Statistical Collection Global Capture Production Database (http://www.fao.org/fishery/
statistics/global-capture-production/en) and are limited to diadromous and marine species. Yearly mean CO2 concentrations as measured at the Mauna Loa Observatory
were obtained from ftp://ftp.cmdl.noaa.gov/ccg/co2/trends/co2_annmean_mlo.txt.
Forum: Science & Society Trends in Ecology & Evolution March 2013, Vol. 28, No. 3
129
Author's Personal Copypy236
9 Smil, V. (2008) Energy in Nature and Society: General Energetics of
Complex Systems, MIT Press
10 Burger, J.R. et al. (2012) The macroecology of sustainability. PLoS Biol.
10, e1001345
11 Durand, J.D. (1977) Historical estimates of world population: an
evaluation. Popul. Dev. Rev. 3, 253–296
12 Vermeij, G.J. and Leigh, E.G., Jr (2011) Natural and human economies
compared. Ecosphere 2, art39
13 Chatterjee, A. et al., eds (2005) Econophysics of Wealth Distributions,
Springer Verlag
14 Hardin, G. (1968) The tragedy of the commons. Science 162, 1243–1248
15 Ostrom, E. et al. (1999) Revisiting the commons: local lessons, global
challenges. Science 284, 278–282
0169-5347/$ – see front matter ß 2012 Elsevier Ltd. All rights reserved.
http://dx.doi.org/10.1016/j.tree.2012.12.001 Trends in Ecology & Evolution, March 2013,
Vol. 28, No. 3
Forum: Science & Society Trends in Ecology & Evolution March 2013, Vol. 28, No. 3
130
Author's Personal Copypy237
Essay
The Macroecology of Sustainability
Joseph R. Burger1
*, Craig D. Allen2
, James H. Brown1,3
, William R. Burnside1
, Ana D. Davidson1,4
,
Trevor S. Fristoe1
, Marcus J. Hamilton1,3,5
, Norman Mercado-Silva1,6
, Jeffrey C. Nekola1
, Jordan G. Okie1¤
,
Wenyun Zuo1
1 Department of Biology, University of New Mexico, Albuquerque, New Mexico, United States of America, 2 Geological Survey, Fort Collins Science Center, Jemez
Mountains Field Station, Los Alamos, New Mexico, United States of America, 3 Santa Fe Institute, Santa Fe, New Mexico, United States of America, 4 Instituto de Ecologı´a,
Universidad Nacional Auto´noma de Me´xico, Me´xico Federal, Me´xico, 5 Department of Anthropology, University of New Mexico, Albuquerque, New Mexico, United States
of America, 6 School of Natural Resources and the Environment, University of Arizona, Tucson, Arizona, United States of America
abstract: The discipline of sustainability
science has emerged in
response to concerns of natural
and social scientists, policymakers,
and lay people about whether the
Earth can continue to support
human population growth and economic
prosperity. Yet, sustainability
science has developed largely independently
from and with little
reference to key ecological principles
that govern life on Earth. A
macroecological perspective highlights
three principles that should
be integral to sustainability science:
1) physical conservation laws govern
the flows of energy and materials
between human systems and
the environment, 2) smaller systems
are connected by these flows to
larger systems in which they are
embedded, and 3) global constraints
ultimately limit flows at
smaller scales. Over the past few
decades, decreasing per capita rates
of consumption of petroleum, phosphate,
agricultural land, fresh water,
fish, and wood indicate that the
growing human population has
surpassed the capacity of the Earth
to supply enough of these essential
resources to sustain even the current
population and level of socioeconomic
development.
‘‘Sustainability’’ has become a key
concern of scientists, politicians, and lay
people—and for good reason. There is
increasing evidence that we have approached,
or perhaps even surpassed, the
capacity of the planet to support continued
human population growth and socioeconomic
development [1–3]. Currently,
humans are appropriating 20%–40% of
the Earth’s terrestrial primary production
[4–6], depleting finite supplies of fossil fuels
and minerals, and overharvesting ‘‘renewable’’
natural resources such as fresh water
and marine fisheries [7–10]. In the process,
we are producing greenhouse gases and
other wastes faster than the environment
can assimilate them, altering global climate
and landscapes, and drastically reducing
biodiversity [2]. Concern about whether
current trajectories of human demography
and socioeconomic activity can continue in
the face of such environmental impacts has
led to calls for ‘‘sustainability.’’ A seminal
event was the Brundtland commission
report [11], which defined ‘‘sustainable
development (as) development that meets
the needs of the present without compromising
the ability of future generations to
meet their own needs.’’
One result has been the emergence of
the discipline of sustainability science. ‘‘Sustainability
science (is) an emerging field of
research dealing with the interactions
between natural and social systems, and
with how those interactions affect the
challenge of sustainability: meeting the
needs of present and future generations
while substantially reducing poverty and
conserving the planet’s life support systems’’
(Proceedings of the National Academy
of Sciences of the USA [PNAS], http://www.
pnas.org/site/misc/sustainability.shtml).
It is the subject of numerous books, at least
three journals (Sustainability Science [Springer];
Sustainability: Science, Practice, & Policy
[ProQuest-CSA]; International Journal of
Sustainability Science and Studies [Polo Publishing]),
and a special section of the PNAS.
In ‘‘A Survey of University-Based Sustainability
Science Programs’’, conducted
in 2007, (http://sustainabilityscience.org/
content.html?contentid =1484), the American
Association for the Advancement of
Science listed 103 academic programs,
including 64 in the United States and
Canada, and many more have been established
subsequently.
Interestingly, despite the above definition,
the majority of sustainability science
appears to emphasize social science while
largely neglecting natural science. A survey
of the published literature from 1980
through November 2010 using the Web of
Science reveals striking results. Of the
23,535 published papers that include ‘‘sustainability’’
in the title, abstract, or key
Essays articulate a specific perspective on a topic of
broad interest to scientists.
Citation: Burger JR, Allen CD, Brown JH, Burnside WR, Davidson AD, et al. (2012) The Macroecology of
Sustainability. PLoS Biol 10(6): e1001345. doi:10.1371/journal.pbio.1001345
Academic Editor: Georgina M. Mace, Imperial College London, United Kingdom
Published June 19, 2012
This is an open-access article, free of all copyright, and may be freely reproduced, distributed, transmitted,
modified, built upon, or otherwise used by anyone for any lawful purpose. The work is made available under
the Creative Commons CC0 public domain dedication.
Funding: Financial support was provided by the Howard Hughes Medical Institute (Interfaces grant to JHB,
JGO, and WZ), National Institutes of Health (grant T32EB009414 to JHB and JRB), National Science Foundation
(grant OISE-0653296 to ADD and grant DEB-0541625 to MJH), and Rockefeller Foundation (grant to MJH). The
content is solely the responsibility of the authors and does not necessarily reflect the views of the organizations
listed above. The funders had no role in study design, data collection and analysis, decision to publish, or
preparation of the manuscript.
Competing Interests: The authors have declared that no competing interests exist.
Abbreviations: GDP, gross domestic product
* E-mail: jrburger@unm.edu
¤ Current address: School of Earth and Space Exploration, Arizona State University, Tempe, Arizona, United
States of America
PLoS Biology | www.plosbiology.org 1 June 2012 | Volume 10 | Issue 6 | e1001345
238
words, 48% include ‘‘development’’ or
‘‘economics’’. In contrast, only 17% include
any mention of ‘‘ecology’’ or ‘‘ecological’’,
12% ‘‘energy’’, 2% ‘‘limits’’, and fewer than
1% ‘‘thermodynamic’’ or ‘‘steady state’’.
Any assessment of sustainability is necessarily
incomplete without incorporating these
concepts from the natural sciences.
Human Macroecology
A macroecological approach to sustainability
aims to understand how humans are
integrated into and constrained by the
Earth’s systems [12]. In just the last
50,000 years, Homo sapiens has expanded
out of Africa to become the most dominant
species the Earth has ever experienced.
Near-exponential population growth, global
colonization, and socioeconomic development
have been fueled by extracting
resources from the environment and transforming
them into people, goods, and
services. Hunter-gatherers had subsistence
economies based on harvesting local biological
resources for food and fiber and on
burning wood and dung to supplement
energy from human metabolism. With the
transition to agricultural societies after the
last ice age [13] and then to industrial
societies within the last two centuries, per
capita energy use has increased from
approximately 120 watts of human biological
metabolism to over 10,000 watts,
mostly from fossil fuels [3,14]. Modern
economies rely on global networks of
extraction, trade, and communication to
rapidly distribute vast quantities of energy,
materials, and information.
The capacity of the environment to
support the requirements of contemporary
human societies is not just a matter of
political and economic concern. It is also a
central aspect of ecology— the study of the
interactions between organisms, including
humans, and their environments. These
relationships always involve exchanges of
energy, matter, or information. The scientific
principles that govern the flows and
transformations of these commodities are
fundamental to ecology and directly relevant
to sustainability and to the maintenance of
ecosystem services, especially in times of
energy scarcity [15]. A macroecological
perspective highlights three principles that
should be combined with perspectives from
the social sciences to achieve an integrated
science of sustainability.
Principle 1: Thermodynamics
and the Zero-Sum Game
The laws of thermodynamics and conservation
of energy, mass, and chemical
stoichiometry are universal and without
exception. These principles are fundamental
to biology and ecology [16–18]. They
also apply equally to humans and their
activities at all spatial and temporal scales.
The laws of thermodynamics mean that
continual flows and transformations of
energy are required to maintain highly
organized, far-from-equilibrium states of
complex systems, including human societies.
For example, increased rates of
energy use are required to fuel economic
growth and development, raising formidable
challenges in a time of growing energy
scarcity and insecurity [3,15,19]. Conservation
of mass and stoichiometry means
that the planetary quantities of chemical
elements are effectively finite [15,18].
Human use of material resources, such
as nitrogen and phosphorus, alters flows
and affects the distribution and local
concentrations in the environment [18].
This is illustrated by the Bristol Bay
salmon fishery, which is frequently cited
as a success story in sustainable fisheries
management [20,21]. In three years for
which good data are available (2007–
2009), about 70% of the annual wild
salmon run was harvested commercially,
with one species, sockeye, accounting for
about 95% of the catch [22]. From a
management perspective, the Bristol Bay
sockeye fishery has been sustainable,
because annual runs have not declined.
Additional implications for sustainability,
however, come from considering the effect
of human harvest on the flows of energy
and materials in the upstream ecosystem
(Figure 1). When humans take about 70%
of Bristol Bay sockeye runs as commercial
catch, this means a 70% reduction in the
number of mature salmon returning to
their native waters to spawn and complete
their life cycles. It also means a concomitant
reduction in the supply of salmon to
support populations of predators, such as
grizzly bears, bald eagles, and indigenous
people, all of which historically relied on
salmon for a large proportion of their diet
[23,24]. Additionally, a 70% harvest means
annual removal of more than 83,000 metric
tonnes of salmon biomass, consisting of
approximately 12,000, 2,500, and 330
tonnes of carbon, nitrogen, and phosphorus,
respectively (see Text S1 for sources
and calculations). These marine-derived
materials are no longer deposited inland
in the Bristol Bay watershed, where they
once provided important nutrient subsidies
to stream, lake, riparian, and terrestrial
ecosystems [24–27]. So, for example, one
apparent consequence is that net primary
production in one oligotrophic lake in the
Bristol Bay watershed has decreased ‘‘to
about 1/3 of its level before commercial
fishing’’ [28]. Seventy percent of Bristol
Bay salmon biomass and nutrients are now
exported to eastern Asia, western Europe,
and the continental US, which are the
primary markets for commercially harvested
wild Alaskan salmon. Our macroecological
assessment of the Bristol Bay fishery
suggests that ‘‘sustainable harvest’’ of the
focal salmon species does not consider the
indirect impacts of human take on critical
resource flows in the ecosystem (Figure 1).
So the Bristol Bay salmon fishery is
probably not entirely sustainable even at
the ‘‘local’’ scale.
Principle 2: Scale and
Embeddedness
Most published examples of sustainability
focus on maintaining or improving
environmental conditions or quality of life
in a localized human system, such as a
farm, village, city, industry, or country
([29,30] and articles following [31]). These
socioeconomic systems are not closed or
isolated, but instead are open, interconnected,
and embedded in larger environmental
systems. Human economies extract
energy and material resources from the
environment and transform them into
goods and services. In the process, they
create waste products that are released
back into the environment. The laws of
conservation and thermodynamics mean
that the embedded human systems are
absolutely dependent on these flows:
population growth and economic development
require increased rates of consumption
of energy and materials and increased
production of wastes. The degree of
dependence is a function of the size of
the economy and its level of socioeconomic
development [3]. Most organic farms
import fuel, tools, machinery, social services,
and even fertilizer, and export their
products to markets. A small village in a
developing country harvests food, water,
and fuel from the surrounding landscape.
Large, complex human systems, such as
corporations, cities, and countries, are
even more dependent on exchanges with
the broader environment and consequently
pose formidable challenges for sustainability.
Modern cities and nation states are
embedded in the global economy, and
supported by trade and communication
networks that transport people, other
organisms, energy, materials, and information.
High densities of people and
concentrations of socioeconomic activities
require massive inputs of energy and
materials and produce proportionately
large amounts of wastes. Claims that such
PLoS Biology | www.plosbiology.org 2 June 2012 | Volume 10 | Issue 6 | e1001345
239
systems are ‘‘sustainable’’ usually only
mean that they are comparatively
‘‘green’’—that they aim to minimize
environmental impacts while offering their
inhabitants happy, healthy lifestyles.
A macroecological perspective on the
sustainability of local systems emphasizes
their interrelations with the larger systems
in which they are embedded, rather than
viewing these systems in isolation. Portland,
Oregon offers an illuminating example.
The city of Portland and surrounding
Multnomah County, with a population of
715,000 and a median per capita income
of US$51,000, bills itself and is often
hailed by the media as ‘‘the most sustainable
city in America’’ (e.g., SustainLane.com,
2008). On the one hand, there can
be little question that Portland is relatively
green and offers its citizens a pleasant,
healthy lifestyle, with exemplary bike paths,
parks, gardens, farmers’ markets, and recycling
programs. About 8% of its electricity
comes from renewable non-hydroelectric
sources (http://apps3.eere.energy.gov/
greenpower/resources/tables/topten.shtml).
On the other hand, there also can be no
question that Portland is embedded in and
completely dependent on environments and
economies at regional, national, and global
scales (Figure 2). A compilation and quantitative
analysis of the flows into and out of
the city are informative (see Text S1 for
sources and calculations). Each year the
Portland metropolitan area consumes at
least 1.25 billion liters of gasoline, 28.8
billion megajoules of natural gas, 31.1 billion
megajoules of electricity, 136 billion liters of
water, and 0.5 million tonnes of food, and
the city releases 8.5 million tonnes of carbon
as CO2, 99 billion liters of liquid sewage,
and 1 million tonnes of solid waste into the
environment. Total domestic and international
trade amounts to 24 million tonnes of
materials annually. With respect to these
flows, Portland is not conspicuously ‘‘green’’;
the above figures are about average for a US
city of comparable size (e.g., [32]).
A good way to see the embedding
problem is to imagine the consequences
of cutting off all flows in and out, as
military sieges of European castles and
cities attempted to do in the Middle Ages.
From this point of view and in the short
term of days to months, some farms and
ranches would be reasonably sustainable,
but the residents of a large city or an
apartment building would rapidly succumb
to thirst, starvation, or disease.
Viewed from this perspective, even though
Portland may be the greenest and by some
definitions ‘‘the most sustainable city in
America’’, it is definitely not self-sustaining.
Massive flows of energy and materials
across the city’s boundaries are required
just to keep its residents alive, let alone
provide them with the lifestyles to which
they have become accustomed. Any complete
ecological assessment of the sustainability
of a local system should consider its
connectedness with and dependence on
the larger systems in which it is embedded.
Figure 1. Pictorial illustration of important flows of salmon and contained biomass, energy, and nutrients within and out of the
Bristol Bay ecosystem. Brown arrows depict the flows within the ecosystem, green arrows depict inputs due to growth in fresh water or the sea,
and red arrows represent human harvest. Seventy percent of salmon are extracted by humans and are no longer available to the Bristol Bay
ecosystem.
doi:10.1371/journal.pbio.1001345.g001
PLoS Biology | www.plosbiology.org 3 June 2012 | Volume 10 | Issue 6 | e1001345
240
Principle 3: Global Constraints
For thousands of years, humans have
harvested fish, other animals, and plants
with varying degrees of ‘‘sustainability’’
and lived in settlements that depend on
imports and exports of energy and materials.
Throughout history, humans have
relied on the environment for goods and
services and used trade to compensate for
imbalances between extraction, production,
and consumption at local to regional
scales. What is different now are the
enormous magnitudes and global scales
of the fluxes of energy and materials into
and out of human systems. Every year
fisheries export thousands of tonnes of
salmon biomass and the contained energy
and nutrients from the Bristol Bay ecosystem
to consumers in Asia, Europe, and the
US. Every year Portland imports ever
larger quantities of energy and materials to
support its lifestyle and economy. Collectively,
such activities, replicated thousands
of times across the globe, are transforming
the biosphere.
Can the Earth support even current
levels of human resource use and waste
production, let alone provide for projected
population growth and economic development?
From our perspective, this should be
the critical issue for sustainability science.
The emphasis on local and regional
scales—as seen in the majority of the
sustainability literature and the above two
examples—is largely irrelevant if the human
demand for essential energy and materials
exceeds the capacity of the Earth to supply
these resources and if the release of wastes
exceeds the capacity of the biosphere to
absorb or detoxify these substances.
Human-caused climate change is an
obvious and timely case in point. Carbon
dioxide has always been a waste product of
human metabolism—not only the biological
metabolism that consumes oxygen and
produces carbon dioxide as it converts
food into usable energy for biological
activities, but also the extra-biological
metabolism that also produces CO2 as it
burns biofuels and fossil fuels to power the
maintenance and development of huntergatherer,
agricultural, and industrial-technological
societies. Only in the last century
or so, however, has the increasing production
of CO2 by humans overwhelmed
the Earth’s capacity to absorb it, increasing
atmospheric concentrations and
warming the planet more each decade.
So, for example, efforts to achieve a
‘‘sustainable’’ local economy for a coastal
fishing village in a developing country will
be overwhelmed if, in only a few decades,
a rising sea level caused by global climate
change inundates the community. This
shows the importance of analyzing sustainability
on a global as well as a local
and regional scale.
A macroecological approach to sustainability
science emphasizes how human
socioeconomic systems at any scale depend
on the flows of essential energy and
material resources at the scale of the
biosphere as a whole. The finite Earth
system imposes absolute limits on the
ecological processes and human activities
embedded within it. The impossibility of
continued exponential growth of population
and resource use in a finite world has
long been recognized [33–35]. But repeated
failures to reach the limits in the
predicted time frames have caused much
of the economic establishment and general
public to discredit or at least discount
Malthusian dynamics. Now, however,
there is increasing evidence that humans
are pushing if not exceeding global limits
[2,3,36,37]. For example, the Global
Footprint Network estimates that the
ecological footprint, the amount of land
required to maintain the human population
at a steady state [9], had exceeded the
available land area by more than 50% by
2007, and the imbalance is increasing (http://
www.footprintnetwork.org/en/index.php/
GFN).
Here we present additional evidence that
humans have approached or surpassed the
capacity of the biosphere to provide
essential and often non-substitutable natural
resources. Figure 3 plots trends in the
total and per capita use of agricultural land,
fresh water, fisheries, wood, phosphate,
petroleum, copper, and coal, as well as
gross domestic product (GDP), from 1961
to 2008. Note that only oil, copper, coal,
and perhaps fresh water show consistent
increases in total consumption. Consumption
of the other resources peaked in the
1980s or 1990s and has since declined.
Dividing the total use of each resource by
the human population gives the per capita
rate of resource use, which has decreased
conspicuously for all commodities except
copper and coal. This means that production
of these commodities has not kept pace
with population growth. Consumption by
the present generation is already ‘‘compromising
the ability of future generations to
meet their own needs.’’ And this does not
account for continued population growth,
which is projected to increase the global
population to 9–10 billion by 2050 and
would result in substantial further decreases
in per capita consumption.
Figure 3 shows results consistent with
other analyses reporting ‘‘peak’’ oil, fresh
water, and phosphate, meaning that global
stocks of these important resources have
been depleted to the point that global
consumption will soon decrease if it has
not already done so [10,37]. Decreased
per capita consumption of essential resources
might be taken as an encouraging
sign of increased efficiency. But the
increase in efficiency is also a response to
higher prices as a result of decreasing
supply and increasing demand. We have
included plots for copper and coal to show
that overall production of some more
abundant commodities has kept pace with
population growth, even though the richest
stocks have already been exploited.
This is typical in ecology: not all essential
resources are equally limiting at any given
time. Diminishing supplies of some critical
Figure 2. Pictorial illustration of important flows of resources into and wastes out of
Portland, Oregon. This ‘‘most sustainable city in America’’ depends on exchanges with the
local, regional, and global environments and economies in which it is embedded.
doi:10.1371/journal.pbio.1001345.g002
PLoS Biology | www.plosbiology.org 4 June 2012 | Volume 10 | Issue 6 | e1001345
241
Figure 3. Global trends in total and per capita consumption of resources and GDP from 1961 to 2008. Total global use/production is
represented by the grey line using the axis scale on the left side of each diagram. Per capita use/production is represented by the black line using the axis
scale on the right side of each diagram. Per capita values represent the total values divided by global population size as reported by the World Resources
Institute (http://earthtrends.wri.org/). The y-axes are untransformed and scaled to allow for maximum dispersion of variance. Individual sources for global
use/production values are as follows: Agricultural land in square-km is from the World Development Indicators Database of the World Bank (http://data.
worldbank.org/data-catalog/world-development-indicators) and represents the sum of arable, permanent crop, and permanent pasture lands (see also
[46]). Freshwater withdrawal in cubic-km from 1960, 1970, 1980, and 1990 is from UNESCO (http://webworld.unesco.org/water/ihp/db/shiklomanov/
part%273/HTML/Tb_14.html) and for 2000 from The Pacific Institute (http://www.worldwater.org/data.html). Wild fisheries harvest in tonnes is from the
FAO Fishery Statistical Collection Global Capture Production Database (http://www.fao.org/fishery/statistics/global-capture-production/en) and is limited
to diadromous and marine species. Wood building material production in tonnes is based on the FAO ForeSTAT database (http://faostat.fao.org/site/626/
default.aspx), and represents the sum of compressed fiberboard, pulpwood+particles (conifer and non-conifer [C & NC]), chips and particles, hardboard,
insulating board, medium density fiberboard, other industrial roundwood (C & NC), particle board, plywood, sawlogs+veneer logs (C & NC), sawnwood (C &
NC), veneer sheets, and wood residues. Phosphate, copper, and combustible coal production in tonnes is based on World Production values reported in the
USGS Historical Statistics for Mineral and Material Commodities (http://minerals.usgs.gov/ds/2005/140/). Global coal production data is limited to 1966–
2008. Petroleum production in barrels from 1965 to 2008 is based on The Statistical Review of World Energy (http://www.bp.com/sectiongenericarticle800.
do?categoryId=9037130&contentId=7068669) and represents all crude oil, shale oil, and oil sands plus the liquid content of natural gas where this is
separately recovered. These data are reported in 1,000 barrels/day units, and were transformed to total barrels produced per year. GDP in 1990 US dollars
are from the World Resources Institute (http://earthtrends.wri.org/). All data were accessed May 15–June 15, 2011.
doi:10.1371/journal.pbio.1001345.g003
PLoS Biology | www.plosbiology.org 5 June 2012 | Volume 10 | Issue 6 | e1001345
242
resources, such as oil, phosphorus, arable
land, and fresh water, jeopardize the
capacity to maintain even the current
human population and standard of living.
What are the consequences of these
trends? Many economists and sustainability
scientists suggest that there is little
cause for concern, at least in the short
term of years to decades. They give several
reasons: i) the finite stocks have not been
totally exhausted, just depleted; there are
still fish in the sea, and oil, water,
phosphate, copper, and coal in the
ground; they are just getting harder to
find and extract; ii) conservation and
substitution can compensate for depletion,
allowing economies to grow and provide
for increases in population and standard of
living; iii) production depends more on the
relationship between supply and demand
as reflected in price than on absolute
availability; and iv) the socioeconomic
status of contemporary humans depends
not so much on raw materials and
conventional goods as on electronic information,
service industries, and the traditional
economic variables of money, capital,
labor, wages, prices, and debt.
There are several reasons to question
this optimistic scenario. First, the fact that
GDP has so far kept pace with population
does not imply that resource production
will do likewise. Indeed, we have shown
that production of some critical resources
is not keeping pace. Second, there is
limited or zero scope to substitute for
some resources. For most of them, all
known substitutes are inferior, scarcer, and
more costly. For example, there is no
substitute for phosphate, which is an
essential requirement of all living things
and a major constituent of fertilizer. No
other element has the special properties of
copper, which is used extensively in
electronics. Despite extensive recycling of
copper, iron, aluminum, and other metals,
there is increasing concern about maintaining
supplies as the rich natural ores
have been depleted (e.g., [38], but see
[39]). Third, several of the critical resources
have interacting limiting effects. For
example, the roughly constant area of land
in cultivation since 1990 indicates that
modern agriculture has fed the increasing
human population by achieving higher
yields per unit area. But such increased
yields have required increased inputs of oil
for powering machinery, fresh water for
irrigation, and phosphate for fertilizer.
Similarly, increased use of finite fossil fuels
has been required to synthesize nitrogen
fertilizers and to maintain supplies of
mineral resources, such as copper, nickel,
and iron, as the richest ores have been
depleted and increased energy is required
to extract the remaining stocks. An
optimistic scenario would suggest that
increased use of coal and renewable
energy sources such as solar and wind
can substitute for depleted reserves of
petroleum, but Figure 3 shows a similar
pattern of per capita consumption for coal
as for other limiting resources, and the
capacity of renewables to substitute for
fossil fuels is limited by thermodynamic
constraints due to low energy density and
economic constraints of low energy and
monetary return on investment [40–43].
Fourth, these and similar results (e.g., [3])
are starting to illuminate the necessary
interdependencies between the energetic
and material currencies of ecology and the
monetary currencies of economics. The
relationship between decreasing supply and
increasing demand is causing prices of
natural resources to increase as they are
depleted, and also causing prices of food to
increase as fisheries are overharvested and
agriculture requires increasing energy and
material subsidies [2,8,43]. The bottom line
is that the growing human population and
economy are being fed by unsustainable use
of finite resources of fossil fuel energy,
fertilizers, and arable land and by unsustainable
harvests of ‘‘renewable resources’’
such as fish, wood, and fresh water. Furthermore,
attaining sustainability is additionally
complicated by inevitable yet
unpredictable changes in both human
socioeconomic conditions and the extrinsic
global environment [44]. Sustainability will
always be a moving target and there cannot
be a single long-term stable solution.
Most sustainability science focuses on
efforts to improve standards of living and
reduce environmental impacts at local to
regional scales. These efforts will ultimately
and inevitably fail unless the global
system is sustainable. There is increasing
evidence that modern humans have already
exceeded global limits on population
and socioeconomic development, because
essential resources are being consumed at
unsustainable rates. Attaining sustainability
at the global scale will require some
combination of two things: a decrease in
population and/or a decrease in per capita
resource consumption (see also [45]).
Neither will be easy to achieve. Whether
population and resource use can be
reduced sufficiently and in time to avoid
socioeconomic collapse and attendant
human suffering is an open question.
Critics will point out that our examination
of sustainability from a macroecological
and natural science perspective
conveys a message of ‘‘doom and gloom’’
and does not offer ‘‘a way forward’’. It is
true that humanity is faced with difficult
choices, and there are no easy solutions.
But the role of science is to understand how
the world works, not to tell us what we want
to hear. The advances of modern medicine
have cured some diseases and improved
health, but they have not given us immortality,
because fundamental limits on human
biology constrain us to a finite lifespan.
Similarly, fundamental limits on the flows
of energy and materials must ultimately
limit the human population and level of
socioeconomic development. If civilization
in anything like its present form is to persist,
it must take account of the finite nature of
the biosphere.
Conclusion
If sustainability science is to achieve its
stated goals of ‘‘dealing with the interactions
between natural and social systems’’
so as to ‘‘[meet] the needs of present and
future generations while substantially reducing
poverty and conserving the planet’s
life support systems’’, it must take account
of the ecological limits on human systems
and the inherently ecological nature of the
human enterprise. The human economy
depends on flows of energy and materials
extracted from the environment and
transformed by technology to create goods
and services. These flows are governed by
physical conservation laws. These flows
rarely balance at local or regional scales.
More importantly, however, because these
systems are all embedded in the global
system, the flows of critical resources that
currently sustain socioeconomic systems at
these scales are jeopardized by unsustainable
consumption at the scale of the
biosphere. These ecological relationships
will determine whether ‘‘sustainability’’
means anything more than ‘‘green’’, and
whether ‘‘future generations [will be able]
to meet their own needs’’.
Supporting Information
Text S1 Supplementary data. Calculations
for salmon nutrient inputs to
terrestrial and riparian ecosystems.
(DOCX)
Acknowledgments
This paper is a product of the Human
Macroecology Group, an informal collaboration
of scientists associated with the University
of New Mexico and the Santa Fe Institute. It
benefits from discussions with many colleagues
on the important but controversial topic of
sustainability. We thank Woody Woodruff,
Matthew Moerschbaecher, and John Day for
providing helpful comments on the manuscript.
PLoS Biology | www.plosbiology.org 6 June 2012 | Volume 10 | Issue 6 | e1001345
243
References
1. Goodland R (1995) The concept of environmental
sustainability. Ann Rev Ecol System 26: 1–24.
2. Rockstro¨m J, Steffen W, Noone K, Persson A,
Chapin FS 3rd, et al. (2009) A safe operating
space for humanity. Nature 461: 472–475.
3. Brown JH, Burnside WR, Davidson AD, DeLong
JP, Dunn WC, et al. (2011) Energetic limits to
economic growth. BioScience 61: 19–26.
4. Vitousek PM, Mooney HA, Lubchenco J, Melillo
JM (1997) Human domination of Earth’s ecosystems.
Science 277: 494–499.
5. Imhoff ML, Bounoua L, Ricketts T, Loucks C,
Harriss R, et al. (2004) Global patterns in human
consumption of net primary production. Nature
429: 870–873.
6. Haberl H, Erb KH, Krausmann F, Gaube V,
Bondeau A, et al. (2007) Quantifying and
mapping the human appropriation of net primary
production in earth’s terrestrial ecosystems. Proc
Natl Acad Sci U S A 104: 12942–12947.
7. Pauly D, Watson R, Alder J (2005) Global trends
in world fisheries: impacts on marine ecosystems
and food security. Phil Trans Roy Soc B 360: 5–
12.
8. Worm B, Barbier EB, Beaumont N, Duffy JE,
Folke C, et al. (2006) Impacts of biodiversity loss
on ocean ecosystem services. Science 314: 787–
790.
9. Wackernagel M, Rees WE (1996) Our ecological
footprint: reducing human impact on the earth.
New Society Publications.
10. Gleick PH, Palaniappan M (2010) Peak water
limits to freshwater withdrawal and use. Proc Natl
Acad Sci U S A 107: 11155–11162.
11. Brundtland GH (1987) Our common future.
World Commission on Environment and Devel-
opment.
12. Burnside WR, Brown JH, Burger O, Hamilton
MJ, Moses M, et al. (2012) Human macroecology:
linking pattern and process in big picture
human ecology. Biol Rev 87: 194–208.
13. Day JW Jr, Boesch DF, Clairain EJ, Kemp GP,
Laska SB, et al. (2007) Restoration of the
Mississippi Delta: lessons from hurricanes Katrina
and Rita. Science 315: 1679–1684.
14. Moses ME, Brown JH (2003) Allometry of human
fertility and energy use. Ecol Let 6: 295–300.
15. Day JW Jr, et al. (2009) Ecology in times of
scarcity. BioScience 59: 321–331.
16. Odum HT (1971) Environment, power and
society. New York: Wiley-Interscience.
17. Odum HT (2007) Environment, power, and
society for the twenty-first century: the hierarchy
of energy. Columbia University Press.
18. Sterner RW, Elser JJ (2002) Ecological stoichiometry:
the biology of elements from molecules to
the biosphere. Princeton: Princeton University
Press.
19. Czu´cz B, Gathman JP, Mcpherson GUY (2010)
The impending peak and decline of petroleum
production: an underestimated challenge for
conservation of ecological integrity. Con Biol
24: 948–956.
20. Hilborn R, Quinn TP, Schindler DE, Rogers DE
(2003) Biocomplexity and fisheries sustainability.
Proc Natl Acad Sci U S A 100: 6564–6568.
21. Hilborn R (2006) Salmon-farming impacts on
wild salmon. Proc Natl Acad Sci U S A 103:
15277.
22. ADG&F (2010) Alaska historical commercial salmon
catches, 1878–2010. Alaska Department of
Game and Fish, Division of Commercial Fisheries.
23. Coupland G, Stewart K, Patton K (2010) Do you
never get tired of salmon? Evidence for extreme
salmon specialization at Prince Rupert harbour,
British Columbia. J Anthro Arch 29: 189–207.
24. Cederholm CJ, Kunze MD, Murota T, Sibatani
A (1999) A Pacific salmon carcasses: essential
contributions of nutrients and energy for aquatic
and terrestrial ecosystems. Fisheries 24: 6–15.
25. Gende SM, Edwards RT, Willson MF, Wipfli MS
(2002) Pacific salmon in aquatic and terrestrial
ecosystems. BioScience 52: 917–928.
26. Naiman RJ, Bilby RE, Schindler DE, Helfield JM
(2002) Pacific salmon, nutrients, and the dynamics
of freshwater and riparian ecosystems. Ecosystems
5: 399–417.
27. Schindler DE, Scheuerell MD, Moore JW, Gende
SM, Francis TB, et al. (2003) Pacific salmon and
the ecology of coastal ecosystems. Front Ecol
Enviro 1: 31–37.
28. Schindler DE, Leavitt PR, Brock CS, Johnson SP,
Quay PD (2005) Marine-derived nutrients, commercial
fisheries, and production of salmon and
lake algae in Alaska. Ecology 86: 3225–3231.
29. International Council for Science (2002) Science
and technology for sustainable development.
Paris: International Council for Science.
30. Millenium Ecosystem Assessment (2005) Ecosystems
and human well-being: synthesis. Island
Press.
31. Turner BL, Lambin EF, Reenberg A (2007) The
emergence of land change science for global
environmental change and sustainability. Proc
Natl Acad Sci U S A 104: 20666–20671.
32. Hillman T, Ramaswami A (2010) Greenhouse gas
emission footprints and energy use benchmarks
for eight US cities. Environ Science & Tech 44:
1902–1910.
33. Malthus TR (1798) An essay on the principle of
population. Prometheus.
34. Ehrlich PR (1968) The population bomb. New
York.
35. Meadows DH (1972) The limits of growth.
A report for The Club of Rome.
36. Holdren JP (2008) Science and technology for
sustainable well-being. Science 319: 424–434.
37. Nel WP, Van Zyl G (2010) Defining limits: energy
constrained economic growth. Applied Energy
87: 168–177.
38. Gordon RB, Bertram M, Graedel TE (2006)
Metal stocks and sustainability. Proc Natl Acad
Sci U S A 103: 1209–1214.
39. Kesler SE, Wilkinson BH (2008) Earth’s copper
resources estimated from tectonic diffusion of
porphyry copper deposits. Geology 36: 255.
40. Fargione J, Hill J, Tilman D, Polasky S,
Hawthorne P (2008) Land clearing and the
biofuel carbon debt. Science 319: 1235–1238.
41. Hall CAS, Day JW (2009) Revisiting the limits to
growth after peak oil. American Scientist 97: 230–
237.
42. Smil V (2008) Energy in nature and society:
general energetics of complex systems. MIT Press.
43. The Royal Society (2009) Reaping the benefits:
science and the sustainable intensification of
global agriculture. London: The Royal Society.
44. Milly PC, Betancourt J, Falkenmark M, Hirsch
RM, Kundzewicz ZW, et al. (2008) Stationarity is
dead: whither water management? Science 319:
573–574.
45. Cohen JE (1996) How many people can the earth
support? WW Norton & Company.
46. Foley JA, Ramankutty N, Brauman KA, Cassidy
ES, Gerber JS, et al. (2011) Solutions for a
cultivated planet. Nature 478: 337–342.
PLoS Biology | www.plosbiology.org 7 June 2012 | Volume 10 | Issue 6 | e1001345
244
Articles
www.biosciencemag.org 	 January 2011 / Vol. 61 No. 1 • BioScience 19
Energetic Limits to Economic Growth
James H. Brown, William R. Burnside, Ana D. Davidson, John P. DeLong, William C. Dunn,
Marcus J. Hamilton, Jeffrey C. Nekola, Jordan G. Okie, Norman Mercado-Silva,
William H. Woodruff, and Wenyun Zuo
The human population and economy have grown exponentially and now have impacts on climate, ecosystem processes, and biodiversity far
exceeding those of any other species. Like all organisms, humans are subject to natural laws and are limited by energy and other resources. In
this article, we use a macroecological approach to integrate perspectives of physics, ecology, and economics with an analysis of extensive global
data to show how energy imposes fundamental constraints on economic growth and development. We demonstrate a positive scaling relationship
between per capita energy use and per capita gross domestic product (GDP) both across nations and within nations over time. Other indicators of
socioeconomic status and ecological impact are correlated with energy use and GDP. We estimate global energy consumption for alternative future
scenarios of population growth and standards of living. Large amounts of energy will be required to fuel economic growth, increase standards of
living, and lift developing nations out of poverty.
Keywords: energy, economic growth, economy, human macroecology, scaling
metabolic ecology (McMahon and Bonner 1983, Schneider
and Kay 1995, Brown et al. 2004) and empirical approaches
from macroecology (Brown 1995) to document energetic
constraints on human ecology that have important implications
for modern humans.
The central role of energy
Economic growth and development require that energy
and other resources be extracted from the environment to
manufacture goods, provide services, and create capital. The
central role of energy is substantiated by both theory and
data.
Key theoretical underpinnings come from the laws of
thermodynamics: first, that energy can be neither created
nor destroyed, and second, that some capacity to perform
useful work is lost as heat when energy is converted from
one form to another. Complex, highly organized systems,
including human economies, are maintained in states far
from thermodynamic equilibrium by the continual intake
and transformation of energy (Soddy 1926, Odum 1971,
Georgescu-Roegen 1977, Ruth 1993, Schneider and Kay
1995, Hall et al. 2001, Chen 2005, Smil 2008).
Empirically, the central role of energy in modern human
economies is demonstrated by the positive relationship
between energy use and economic growth (Shafiee and Topal
2008, Smil 2008, Payne 2010). Here, we take a macroecological
perspective and quantify statistical relationships between
energy use and economic activity for 220 nations over 24
years, using data from the International Energy Agency (IEA;
www.iea.org/stats/index.asp) and World Resources Institute
(WRI; http://earthtrends.wri.org/index.php).Per capita energy
consumption for each country is calculated as the sum of
human biological metabolism plus the energy obtained from
The human species has an interesting duality. On the one 
hand, Homo sapiens is just another species, subject to the
same scientific laws as the millions of other animals, plants,
and microbes. On the other hand, humans are unique. No
other species in the history of Earth has achieved such ecological
dominance and created such complex socioeconomic
systems. Because of this duality, humans have been studied
by both natural and social scientists, but often from very different
perspectives (Arrow et al. 2004).
In just a few thousand years the human population has
colonized the entire world and grown to almost 7 billion.
Humans now appropriate 20% to 40% of terrestrial annual
net primary production, and have transformed the atmosphere,
water, land, and biodiversity of the planet (Vitousek
et al. 1997, Haberl et al. 2007). For centuries some have
questioned how long a finite planet can continue to support
near-exponential population and economic growth
(e.g., Malthus 1798, Ehrlich 1968, Meadows et al. 1972).
Recent issues such as climate change, the global decline in
population growth rate, the depletion of petroleum reserves
and resulting increase in oil prices, and the recent economic
downturn have prompted renewed concerns about
whether longstanding trajectories of population and economic
growth can continue (e.g., Arrow et al. 2004). These
serious issues fall within the purview of both the natural and
social sciences—especially ecology and economics.
This article integrates perspectives from physics, ecology,
and economics with an analysis of extensive global data to
show how scientific laws governing the flows of energy in
the biosphere affect socioeconomic activity. Our purpose
is neither to pit ecology against economics nor to predict
future population and economic trends; rather, we use theoretical
perspectives from thermodynamics, allometry, and
BioScience 61: 19–26. ISSN 0006-3568, electronic ISSN 1525-3244. © 2011 by American Institute of Biological Sciences. All rights reserved. Request permission
to photocopy or reproduce article content at the University of California Press’s Rights and Permissions Web site at www.ucpressjournals.com/reprintinfo.asp.
doi:10.1525/bio.2011.61.1.7
245
20 BioScience • January 2011 / Vol. 61 No. 1	 www.biosciencemag.org
Articles
all other sources. Individual biological metabolism was estimated
from data on daily caloric intake for each country and
converted to watts per individual. Energy use from all other
sources, including fossil fuels and renewable energy supplies,
was obtained from the IEA. For our measure of economic
activity, we used the WRI’s data for gross domestic product
(GDP), the market value of all goods and services produced
within a country per year (Mankiw 2006). Both energy use
and GDP are expressed on a per capita basis. So, per capita
GDP can be thought of as an index of an average individual’s
share of his or her country’s
economy, and per capita energy
use as the power required to
sustain that level of economic
activity.
 Figure 1 shows the relationship
between energy use and
GDP plotted on logarithmic axes,
with each colored line indicating
the trajectory for a single country
over the period 1980–2003.
A regression through the mean
GDP, G for mean energy consumption
and E for each country
over the 24-year period, accounts
for 76% of the variation. The
fitted regression describes the
scaling of per capita energy use
with per capita GDP as a power
law: E = 4.13G 0.76
(figure 1; see
also figure S1c in supplemental
online materials at http://caliber.
ucpress.net/doi/suppl/10.1525/
bio.2011.61.1.7). The sublinear
slope, 0.76, indicates that the
rate of per capita energy consumption
associated with greater
economic activity increases less
rapidly than GDP itself. Countries
with larger economies
take advantage of economies of
scale and new technologies to
use energy more efficiently on
a per capita basis (Hoffert et al.
1998). For example, there is both
a positive relationship and an
economy of scale between economic
growth and the amount
of infrastructure, such as roads,
pipelines, and power lines, that
distributes energy resources
(Easterly and Rebelo 1993). The
relationship between energy use
and GDP holds across countries
spanning the entire range of economic
development from poorest
to richest, encompassing two orders of magnitude in
both energy use (100 to 10,000 watts) and wealth ($500 to
$50,000).
 A similar trend occurs within countries over time. The
vast majority of nations we analyzed (74%) increased both
energy use and GDP from 1980 to 2003 and exhibited positive
correlations across the 24 years (mean slope = 0.59; 95%
confidence interval = 0.45–0.72; figure 2; see also figure S1 at
http://caliber.ucpress.net/doi/suppl/10.1525/bio.2011.61.1.7).
Countries notable for sustained recent development, such as
Figure 1. The relationship between per capita energy use and per capita gross
domestic product (GDP; in US dollars) of countries, plotted on logarithmic axes,
from 1980 to 2003. Note that the slope or exponent, 0.76 (95% confidence interval:
0.69–0.82), is close to three-quarters, which is the canonical value of the exponent for
the scaling of metabolic rate with body mass in animals. If per capita GDP is taken
as the size of an average individual’s economy and per capita energy use as the rate of
energy consumption required to support that economy, this relationship may not be
coincidental. Total per capita energy consumption is calculated as the caloric intake
of humans (about 130 watts) plus the energy derived from all other sources, including
fossil fuels and renewables. The thin colored lines show trends for individual
countries from 1980 to 2003. The thick black line is a regression model fit to the mean
values for each nation during this period. GDP data are from the World Resources
Institute (http://earthtrends.wri.org/index.php). Total energy consumption data are
calculated from the sum of energy consumption from eating (data from the World
Resources Institute) plus all other sources of energy consumed for other purposes
such as utilities, manufacturing, and transportation. Source: Data are from the
International Energy Agency at www.iea.org/stats/index.asp.
246
www.biosciencemag.org 	 January 2011 / Vol. 61 No. 1 • BioScience 21
Articles
China and India, show trajectories of continually increasing
energy use. The patterns in figures 1 and 2 are consistent
with the increasing energy use that fueled socioeconomic
development throughout history (Tainter 1988, Smil 2008).
For example, from 1850 to 2000, while the global human
population grew fivefold, world energy use increased 20-fold
and fossil fuel use rose more than 150-fold (Holdren 2008).
 The relationship between energy use and GDP depicted
in figure 1 raises several important issues. One is the considerable
variation around the regression line: Countries with
similar per capita GDPs differ by more than an order of magnitude
in per capita energy consumption.We did not analyze
this residual variation quantitatively, although it would be
illuminating to do so. For example, several oil-exporting
nations (e.g., the United Arab Emirates, Bahrain, Tajikistan,
and Azerbaijan) are consistent outliers, with high values of
energy use relative to GDP. We hypothesize that much of
this energy is used to extract oil that is then exported to and
consumed by industrialized nations. Although the United
States is among the countries with highest per capita energy
use, several other developed countries have comparably
high values. We hypothesize that in some of them, such as
Iceland, Sweden, and Norway, large quantities of energy are
used to heat houses and workplaces during cold winters at
high latitudes.
 Additionally, some of the variation in figure 1 may reflect
error or a lack of standardization in the ways different countries
have calculated energy use and GDP. The data sources
that we used (IEA and WRI), although probably the best
available, have their limitations. The original data are selfreported
by the countries, and some of the abrupt, seemingly
inexplicable changes in the trajectories in figure 1 may simply
reflect errors or changing methods of estimating energy
use, GDP, or both. One strength of our macroecological
approach (Brown 1995), however, is that the hundreds of
data points distributed over orders of magnitude variation
mean that modest errors and other sources of uncontrolled
noise in the data do not obscure the strong signals that are
manifest in the robust patterns within and across countries
over time.
 Another question is, what are the independent and
dependent variables? Does energy use support economic
development or does economic development drive energy
consumption? Financial and energy economists have used
econometric techniques to analyze time series of energy
consumption and economic growth within countries in an
effort to assess causal relationships, but they have reached no
clear consensus about whether energy use causes economic
growth, or vice versa (see Mahadevan and Asafu-Adjaye
2007, Payne 2010). By analogy to biological allometry, we
plotted per capita energy use as the dependent variable
and per capita GDP as the independent variable; this is
analogous to plotting the rate of energy use of an animal
as a function of its body size. The exponent for the scaling
of energy use as a function of GDP, 0.76, is reminiscent
of the three-quarter-power scaling of metabolic rate with
body mass in animals (Kleiber 1961, McMahon and Bonner
1983). This may not be coincidental. In a very real sense
both animals and economies have “metabolisms.” Both consume,
transform, and allocate energy to maintain complex
adaptive systems far from thermodynamic equilibrium.
The energy and other resources that sustain these systems
are supplied by hierarchically branching networks, such as
the blood vessels and lungs of mammals and the oil pipelines,
power grids, and transportation networks of nations.
Models of these networks suggest that three-quarter-power
scaling optimizes distribution of resources (West et al. 1997,
Banavar et al. 2010).
 Some may be concerned that the relationships in figures
1 and 2 are “just correlations” that do not necessarily
imply any underlying mechanism or causality. We disagree.
All science is ultimately based on correlations—between
dependent and independent variables, model predictions
and empirical measurements, or experimental treatments
and controls. Any mechanism or causation comes from
logical inference. We infer that energy limits economic
activity through direct causal mechanisms. The evidence
for this inference is presented above and comes from three
sources: (1) theory, the application of the second law of
thermodynamics to complex adaptive systems; (2) data,
the robust relationship between per capita energy use and
per capita GDP across both space (the 220 nations of the
world) and time (24 years); and (3) analogy, the similarity
Figure 2. The frequency distribution of exponents (slopes
from log-log plots) of economic trajectories of per capita
energy use as a function of per capita gross domestic product
(GDP) within countries over the time series from 1980 to
2003. The average slope is approximately 0.6 and nearly all
countries showed positive growth over the 24-year period.
Source: Data are from the World Resources Institute (http://
earthtrends.wri.org/index.php) and the International
Energy Agency (www.iea.org/stats/index.asp).
247
22 BioScience • January 2011 / Vol. 61 No. 1	 www.biosciencemag.org
Articles
between biological and socioeconomic metabolism. We
find the last to be especially compelling. Just as a body has
a metabolism that burns food energy to survive and grow,
a city or national economy has a metabolism that must
burn fuel in order to sustain itself and grow. Just as higher
metabolic rates are required to sustain and grow larger,
more complex bodies (Kleiber 1961, McMahon and Bonner
1983), so higher rates of energy consumption are required
to sustain and grow larger, more developed economies that
provide greater levels of technological development and
higher standards of living.
Quantitative relationships among energy use, GDP,
and other socioeconomic indicators
Some may be concerned that the relationships in figures 1
and 2 do not reflect what is “really important,” which might
be some aspect of quality of life rather than GDP. However,
nearly all measures of economic activity and standard of
living are closely correlated with both GDP and energy
use (figure 3; for additional variables, see figure S2 in the
supplemental online materials at http://caliber.ucpress.net/
doi/suppl/10.1525/bio.2011.61.1.7). These include measures
of nutrition, education, health care, resource use, technology,
and innovation. These relationships are not surprising
and reflect mechanistic underpinnings. It takes money and
energy to train engineers, MDs, and PhDs; to produce vaccines,
drugs, and medical equipment; and to construct and
maintain road, rail, airplane, cell phone, and Internet networks,
hospitals and research centers, parks and conservation
areas, and modern buildings and cities. The ecological
footprint, an aggregate measure of per capita resource consumption
and waste production, also increases with energy
use and GDP (figure 3; Dietz et al. 2007). Figure 3 shows that
it has not been possible to increase socially desirable goods
and services substantially without concomitantly increasing
the consumption of energy and other natural resources and
without increasing environmental impacts that now include
climate change, pollution, altered biogeochemical cycles, and
reduced biodiversity.
Energetic implications for future economic growth
These empirical patterns, together with their theoretical
underpinnings, raise the question of whether economic
growth and associated increases in human population,
resource use, technological development, and standard of
living can continue their present trajectories (Grossman
and Krueger 1995, Ausubel 1996). In figure 4 we develop
some quantitative scenarios. We caution that these are
not intended to be predictions of the future; rather, they
are extrapolations of the power-law relationship shown
in figure 1 to estimate the quantity of energy that would
be required to support different global populations and
levels of economic development. So, for example, raising
the current global population to the standard of living in
the United States would require a nearly fivefold increase
in the rate of energy consumption, from 17 to 77 terawatts
(1 terawatt = 1012
watts). Population growth must also be
considered in any future scenario. To support a projected
global population of 9.5 billion in 2050 with an average standard
of living equivalent to the current US lifestyle would
require about 268 terawatts, 16 times the current global
energy use. Even maintaining this increased population at
the more modest Chinese standard of living would require
2.5 times more energy than is used today (figure 4).
 There are good reasons, however, such simple scenarios
based on extrapolations of current population and economic
trends may be imprecise. Our calculations incorporate
the economy of scale implicit in the nonlinear scaling of
energy use with GDP, but do not take into account many
potentially important factors, such as greater efficiency
that may be triggered by energy shortages; technological
innovations that may increase energy supplies; and socioeconomic,
demographic, and behavioral changes. Indeed,
the global human socioeconomic system is complex, poised
far from thermodynamic equilibrium by high rates of energy
input and transformation. Such systems have unpredictable
nonlinear dynamics, making it nearly impossible to predict
very far into the future (Schneider and Kay 1995).
 One thing is clear: If the relationships depicted in
figures 1–3 characterize fundamental causal relationships
among the rate of energy use, level of economic activity,
and standard of living, then additional economic growth
and development will require some combination of (a)
increased energy supply, (b) decreased per capita energy
use, and (c) decreased human population. We consider
each in turn.
Increased energy supply.  The sources of energy that may
be used to support future economic growth include finite
stocks of fossil fuels as well as nuclear, renewable, and other
proposed but unproven technologies. Fossil fuels currently
provide 85% of humankind’s energy needs (figure 5), but
they are effectively fixed stores that are being depleted
rapidly (Heinberg 2003, IEA 2008, Hall and Day 2009).
Conventional nuclear energy currently supplies only about
6% of global energy; fuel supplies are also finite, and future
developments are plagued by concerns about safety, waste
storage,anddisposal(NelandCooper2009).Abreakthrough
in nuclear fusion, which has remained elusive for the last
50 years, could potentially generate enormous quantities of
energy, but would likely produce large and unpredictable
socioeconomic and environmental consequences. Solar,
hydro, wind, and tidal renewable energy sources are abundant,
but environmental impacts and the time, resources,
and expenses required to capture their energy limit their
potential (Hall and Day 2009). Biofuels may be renewable,
but ecological constraints and environmental impacts
constrain their contribution (Fargione et al. 2008). More
generally, most efforts to develop new sources of energy face
economic problems of diminishing returns on energy and
monetary investment (Hall et al. 1986, Tainter 1988, Allen
et al. 2001, Tainter et al. 2003).
248
www.biosciencemag.org 	 January 2011 / Vol. 61 No. 1 • BioScience 23
Articles
Decreased per capita energy use.  The Malthusian-Darwinian
dynamic that has shaped the evolution of human behavior
and demography has created powerful tendencies for individuals
and societies to exploit all available resources and
use all available technologies to enhance personal status,
biological fitness, and societal wealth (Lotka 1922). Poor
people migrate to cities and to other countries to improve
their prospects. Citizens of developing countries such as
China and India are not usually satisfied with the status
quo, and understandably want to live like those in the deFigure
3. Variables reflecting socioeconomic status and standard of living are strongly correlated with per capita energy
use (upper panels) and per capita gross domestic product (GDP; lower panels). The variables include measures of health
and wellness (population growth rate, doctors per 100,000 people, life expectancy, infant mortality, caloric intake, national
poverty), energy use (electricity [kilowatt hours], residential energy [kilograms of oil equivalent]), resource consumption
(meat, televisions, aluminum [kilotons], waste [kilograms]), intellectual and technological contributions (Nobel Prizes,
patents), and ecological impacts (ecological footprint [in hectares]). All correlations (r) are significant (P < 0.05). Data
sources are provided in supplemental online materials.
249
24 BioScience • January 2011 / Vol. 61 No. 1	 www.biosciencemag.org
Articles
veloped world. People in the richest nations are reluctant to
sacrifice economic growth—much less give up their automobiles,
electronics, and organ transplants—so that people
in poorer countries can have bicycles, personal computers,
and flu shots.
Decreased human population.  With growing standards of
living and rates of energy use, parents tend to invest more
resources in fewer children (Moses and Brown 2003). This
trade-off between the number and quality of offspring
contributes to demographic transitions, where family size
and the rate of population growth decrease with increasing
economic development (Thompson 1929). The global
population growth rate has declined in the last decade,
but only a few developed countries currently have zero or
negative population growth (WRI, http://earthtrends.wri.
org/index.php). The relationship between family size and per
capita energy use suggests that five times the current rates
of energy supply will be required to achieve a global level
of socioeconomic development capable of stabilizing the
human population without infringing on the freedom of
individuals to have as many children as they choose (Moses
and Brown 2003, DeLong et al. 2010).
 The bottom line is that an enormous increase in energy
supply will be required to meet the demands of projected
population growth and lift the developing world out of
poverty without jeopardizing current standards of living
in the most developed countries. And the possibilities for
substantially increasing energy supplies are highly uncertain.
Moreover, the nonlinear, complex nature of the global
economy raises the possibility that energy shortages might
trigger massive socioeconomic disruption. Again, consider
the analogy to biological metabolism: Gradually reducing
an individual’s food supply leads initially to physiological
adjustments, but then to death from starvation, well before
all food supplies have been exhausted.
 Mainstream economists historically have dismissed
warnings that resource shortages might permanently limit
economic growth. Many believe that the capacity for
technological innovation to meet the demand for resources
is as much a law of human nature as the MalthusianDarwinian
dynamic that creates the demand (Barro and
Sala-i-Martin 2003, Durlauf et al. 2005, Mankiw 2006).
However, there is no scientific support for this proposition;
it is either an article of faith or based on statistically flawed
extrapolations of historical trends. The ruins of Mohenjo
Daro, Mesopotamia, Egypt, Rome, the Maya, Angkor,
Easter Island, and many other complex civilizations provide
incontrovertible evidence that innovation does not always
prevent socioeconomic collapse (Tainter 1988, Diamond
2004).
Conclusions
We are by no means the first to write about the limits to
economic growth and the fundamental energetic constraints
that stem directly from the laws of thermodynamics and
the principles of ecology. Beginning with Malthus (1798),
both ecologists and economists have called attention to the
essential dependence of economies on natural resources
and have pointed out that near-exponential growth of
the human population and economy cannot be sustained
indefinitely in a world of finite resources (e.g., Soddy 1922,
Odum 1971, Daly 1977, Georgescu-Roegen 1977, Cleveland
et al. 1984, Costanza and Daly 1992, Hall et al. 2001, Arrow
et al. 2004, Stern 2004, Nel and van Zyl 2010). Some
ecological economists and systems ecologists have made
Figure 4. Current and projected global energy
consumption based on alternative scenarios of population
growth (2006, 2025, and 2050) and standard of living
(equivalent to contemporary Uganda, China, and United
States). Dashed line is total global terrestrial net primary
productivity (NPP), 75 terawatts (Haberl et al. 2007).
Data sources and calculation methods can be found in
supplemental online materials.
Figure 5. Sources of energy currently consumed by the
global human economy. Total annual consumption is
approximately 15.9 terawatts (TW, 1 terawatt = 1012
watts),
of which about 85% comes from fossil fuels, 6% from nuclear
energy, and the remaining 9% from solar, hydro, wind, and
other renewable sources (BP 2009, REN21 2009).
250
www.biosciencemag.org 	 January 2011 / Vol. 61 No. 1 • BioScience 25
Articles
similar theoretical arguments for energetic constraints
on economic systems (e.g., Odum 1971, Hall et al. 1986).
However, these perspectives have not been incorporated
into mainstream economic theory, practice, or pedagogy
(e.g., Barro and Sala-i-Martin 2003, Mankiw 2006), and
they have been downplayed in consensus statements by
influential ecologists (e.g., Lubchenco et al. 1991, Palmer et
al. 2004, ESA 2009) and sustainability scientists (e.g., NRC
1999, Kates et al. 2001, ICS 2002, Kates and Parris 2003,
Parris and Kates 2003, Clark 2007).
 Our explicitly macroecological and metabolic approach
uses new data and analyses to provide quantitative, mechanistic,
and practically relevant insights into energetic limits
on economic growth. We hope the evidence and interpretations
presented here will call the attention of scientists,
policymakers, world leaders, and the public to the central
but largely underappreciated role of energetic limits to
economic growth.
Acknowledgments
For support we thank the Howard Hughes Medical
Institute and National Institute of Biomedical Imaging
and Bioengineering Interfaces grant to JHB, JGO, and WZ;
National Science Foundation (NSF) Grant DEB-0541625
and the Rockefeller Foundation to MJH; NSF Grant
OISE-0653296 to ADD; and National Institutes of Health
Grant DK36263 to WHW. We thank the many colleagues
who have discussed these ideas with us and encouraged us to
write this article. Charles A. S. Hall, Charles Fowler, Joseph
A. Tainter, and several anonymous reviewers provided
helpful comments on earlier drafts of the manuscript.
References cited
Allen TFH, Tainter JA, Pires JC, Hoekstra TW. 2001. Dragnet ecology—
“Just the facts, ma’am”: The privilege of science in a postmodern world.
BioScience 51: 475–485.
Arrow K, Dasgupta P, Goulder L, Daily G, Ehrlich P, Heal G, Levin S,
Schneider S, Starrett D, Walker B. 2004. Are we consuming too much?
Journal of Economic Perspectives 18: 147–172.
Ausubel JH. 1996. Can technology spare the Earth? American Scientist 84:
166–178.
Banavar JR, Moses ME, Brown JH, Damuth J, Rinaldo A, Sibly RM,
Maritan A. 2010. A general basis for quarter power scaling in
animals. Proceedings of the National Academy of Sciences 107:
15816–15820.
Barro RJ, Sala-i-Martin X. 2003. Economic Growth. 2nd ed. MIT Press.
[BP] British Petroleum.2009.BP Statistical Review of World Energy June 2006.
BP. (11October2010;www.bp.com/liveassets/bp_internet/globalbp/globalbp_
uk_english/reports_and_publications/statistical_energy_review_2006/
STAGING/local_assets/downloads/spreadsheets/statistical_review_full_
report_workbook_2006.xls)
Brown JH. 1995. Macroecology. University of Chicago Press.
Brown JH, Gillooly JF, Allen AP, Savage VM, West GB. 2004. Toward a metabolic
theory of ecology. Ecology 85: 1771–1789.
Chen J. 2005. The Physical Foundation of Economics: An Analytical Thermodynamic
Theory. World Scientific.
Clark WC. 2007. Sustainability science: A room of its own. Proceedings of
the National Academy of Sciences 104: 1737–1738.
Cleveland CJ, Costanza R, Hall CAS, Kaufmann RK. 1984. Energy and the
U.S. economy: A biophysical perspective. Science 225: 890–897.
Costanza R, Daly HE. 1992. Natural capital and sustainable development.
Conservation Biology 6: 37–46.
Daly HE. 1977. Steady-state Economics. Freeman.
DeLong JP, Burger O, Hamilton MJ. 2010. Current demographics suggest
that future energy supplies will be inadequate to slow human population
growth. PLoS ONE 5: e13206. doi:10.1371/journal.pone.0013206
Diamond J. 2004. Collapse: How Societies Choose to Fail or Succeed.Viking
Adult.
Dietz T, Rosa EA, York R. 2007. Driving the human ecological footprint.
Frontiers in Ecology and the Environment 5: 13–18.
Durlauf SN, Johnson PA, Temple JRW. 2005. Growth econometrics. Pages
555–677 in Aghion P, Durlauf SN, eds. Handbook of Economic Growth.
Elsevier.
Easterly W, Rebelo S. 1993. Fiscal policy and economic growth: An empirical
investigation. Journal of Monetary Economics 32: 417–458.
Ehrlich P. 1968. The Population Bomb. Ballantine.
[ESA] Ecological Society of America. 2009. Ecological Impacts of Economic
Activities. ESA. (14 October 2010; www.esa.org/pao/economic_activities.
php)
Fargione J, Hill J, Tilman D, Polasky S, Hawthorne P. 2008. Land clearing
and the biofuel carbon debt. Science 319: 1235–1238.
Georgescu-Roegen N. 1977. The steady state and ecological salvation: A
thermodynamic analysis. BioScience 27: 266–270.
Grossman GM, Krueger AB. 1995. Economic growth and the environment.
Quarterly Journal of Economics 110: 353–377.
Haberl H, Erb K-H, Krausmann F, Gaube V, Bondeau A, Plutzar C, Gingrich
S, Lucht W, Fischer-Kowalski M. 2007. Quantifying and mapping
the human appropriation of net primary production in Earth’s terrestrial
ecosystems. Proceedings of the National Academy of Sciences 104:
12942–12947.
Hall CAS, Day JW Jr. 2009. Revisiting the limits to growth after peak oil.
American Scientist 97: 230–237.
Hall CAS, Cleveland CJ, Kaufmann RK. 1986. Energy and Resource Quality:
The Ecology of the Economic Process. Wiley.
Hall C, Lindenberger D, Kümmel R, Kroeger T, Eichhorn W. 2001. The
need to reintegrate the natural sciences with economics. BioScience 51:
663–673.
Heinberg R. 2003. The end of the oil age. Earth Island Journal 18: 3.
Hoffert MI, et al. 1998. Energy implications of future stabilization of atmospheric
CO2
content. Nature 395: 881–884.
Holdren JP. 2008. Science and technology for sustainable well-being.
Science 319: 424–434.
[ICS] International Council for Science. 2002. Report of the Scientific and
Technological Community to the World Summit on Sustainable Development
(WSSD). ICS. Report no. 1.
[IEA] International Energy Agency. 2008. World Energy Outlook 2008.
Organisation for Economic Co-operation and Development and IEA.
Kates RW, Parris TM. 2003. Long-term trends and a sustainability transition.
Proceedings of the National Academy of Sciences 100: 8062–8067.
Kates RW, et al. 2001. Environment and development: Sustainability
science. Science 292: 641–642.
Kleiber M. 1961. The Fire of Life: An Introduction to Animal Energetics.
Wiley.
Lotka AJ. 1922. Contribution to the energetics of evolution. Proceedings of
the National Academy of Sciences 8: 147–151.
Lubchenco J, et al. 1991. The sustainable biosphere initiative: An ecological
research agenda: A report from the Ecological Society of America.
Ecology 72: 371–412.
Mahadevan R, Asafu-Adjaye J. 2007. Energy consumption, economic
growth and prices: A reassessment using panel VECM for developed
and developing countries. Energy Policy 35: 2481–2490.
Malthus TR. 1798. An Essay on the Principle of Population. Prometheus.
Mankiw NG. 2006. Macroeconomics. 6th ed. Worth.
McMahon TA, Bonner JT. 1983. On Size and Life. Scientific American Library.
Meadows DH, Meadows DL, Randers J, Behrens WW. 1972. The Limits to
Growth: A Report for the Club of Rome’s Project on the Predicament
of Mankind. Earth Island.
251
26 BioScience • January 2011 / Vol. 61 No. 1	 www.biosciencemag.org
Articles
———. 1926. Wealth, Virtual Wealth, and Debt: The Solution of the Economic
Paradox. Omni.
Stern DI. 2004. Economic growth and energy. Pages 35–78 in Encyclopedia
of Energy, vol 2. Academic Press.
Tainter JA. 1988. The Collapse of Complex Societies. Cambridge University
Press.
Tainter JA, Allen THF, Little A, Hoekstra TW. 2003. Resource transitions
and energy gain: Contexts of organization. Conservation Ecology 7: 4.
Thompson WS. 1929. Population. American Journal of Sociology 34:
959–975.
Vitousek PM, Mooney HA, Lubchenco J, Melillo JM. 1997. Human
domination of Earth’s ecosystems. Science 277: 494–499.
West GB, Brown JH, Enquist BJ. 1997. A general model for the origin of
allometric scaling laws in biology. Science 276: 122–126.
James H. Brown (jhbrown@unm.edu) is a distinguished professor at the
University of New Mexico and external faculty of the Santa Fe Institute.
William R. Burnside, William C. Dunn, Jordan G. Okie, and Wenyun Zuo
are PhD candidates in the Department of Biology at the University of New
Mexico. Ana D. Davidson is a postdoctoral researcher at the National University
of Mexico and adjunct professor of biology at the University of New
Mexico. John P. DeLong is a postdoctoral associate at Yale University in the
Department of Ecology and Evolutionary Biology. Marcus J. Hamilton is an
archaeological anthropologist at the University of New Mexico and the Santa
Fe Institute. Jeffrey C. Nekola is an ecologist at the University of New Mexico.
Norman Mercado-Silva is a research specialist with the School of Natural
Resources and the Environment, Arizona Cooperative Fish and Wildlife
Research Unit, at the University of Arizona, in Tucson. William H. Woodruff
is a scientist at Los Alamos National Laboratory and external faculty at the
Santa Fe Institute.
Moses ME, Brown JH. 2003. Allometry of human fertility and energy use.
Ecology Letters 6: 295–300.
[NRC] National Research Council. 1999. Our Common Journey. The
National Academies Press.
Nel WP, Cooper CJ. 2009. Implications of fossil fuel constraints on
economic growth and global warming. Energy Policy 37: 166–180.
Nel WP, van Zyl G. 2010. Defining limits: Energy constrained economic
growth. Applied Energy 87: 168–177.
Odum HT. 1971. Environment, Power and Society. Wiley.
Palmer M, et al. 2004. Ecology: Ecology for a crowded planet. Science 304:
1251–1252.
Parris TM, Kates RW. 2003. Characterizing a sustainability transition: Goals,
targets, trends, and driving forces. Proceedings of the National Academy
of Sciences 100: 8068–8073.
Payne JE. 2010. Survey of the international evidence on the causal
relationship between energy consumption and growth. Journal of
Economic Studies 37: 53–95.
[REN21] Renewable Energy Policy Network for the 21st Century.2006.Renewables,
Global Status Report: 2006 Update. REN21 Steering Committee.
Ruth M. 1993. Integrating Economics, Ecology, and Thermodynamics.
Springer.
Schneider ED, Kay JJ. 1995. Order from Disorder: The Thermodynamics of
Complexity in Biology. Pages 161–172 in Murphy MP, O’Neill LAJ, eds.
What Is Life: The Next Fifty Years. Reflections on the Future of Biology.
Cambridge University Press.
Shafiee S, Topal E. 2008. An econometrics view of worldwide fossil fuel
consumption and the role of US. Energy Policy 36: 775–786.
Smil V. 2008. Energy in Nature and Society: General Energetics of Complex
Systems. MIT Press.
Soddy F. 1922. Cartesian Economics: The Bearing of Physical Science upon
State Stewardship. Hendersons.
252
CONCLUDING REMARKS
What does ‘mechanism’ mean within an ecological context? It is interesting to note the strongly
negative reactions of some ecologists to even the idea that stochastic process may underlie a
number of classic ecological patterns. There appear to be two main reasons ecologists have
tended to reject stochastic mechanism when attempting to explain ecological pattern: they are
‘too’ random and they ‘have nothing to do with ecology’.
With regards to the first critique, why should pattern-producing random process not be
considered a type of mechanism? This bias against stochastic models may be partly driven by
the field of ecology having historically adopted a definition of ‘mechanism’ used by biological,
molecular and cellular biologists, in which mechanism is quite explicit, concrete and
deterministic – this protein bumps into that protein causing a conformation change and exposing
an enzymatic site. But physics has seen a progression from deterministic laws such as those of
Newton to the stochastic techniques and laws of Heisenberg and Schrödinger. Now as
molecular and cellular biologists have begun scaling up into systems biology, stochastic models
are of increasing importance. Why should ecology be any different? Perhaps the increasing
incorporation of stochasticity is a sign of disciplinary maturity. Stochastic models should be
seen as just another class of possible explanatory models that should be judged on their ability to
elucidate our understanding of and make novel predictions about ecological systems.
Although almost never verbalized, it also seems likely that considerable resistance to stochastic
thinking is also based on the “it has nothing to do with ecology” argument. Ecology is relatively
unique as a field because many scientists choose their discipline not because they are enamored
with the process of science but also because they are enthralled with the organisms they study
and the landscapes those organisms live in. Few physicists felt a loss at reducing atoms to
randomly colliding abstractions, but many ecologists feel a loss at similarly reducing organisms
to grist for an abstract, general principle like the central limit theorem or maximum entropy. But
therein lies an irreconcilable problem: many of the patterns ecologists like to hold up as their
own, such as the species abundance distribution, species area relationship or distance decay of
similarity are in fact frequently repeated across entirely non-ecological systems such as geology,
meteorology, economics, computer science, sociology, and the arts. Are we willing to throw
these essential ecological descriptors away as not being truly ‘ecological’? Or, should we be
willing to accept mechanisms that are general enough to be common between all these fields?
Through my eclectic research program, I have rounded up my ‘blind men’ (the various fields that
I conduct active research in) and asked them to report back on the nature of this thing called
‘ecology’. When viewed across multiple fields and types of observation, some general
perspectives come into view – perspectives which are generally not articulated by practitioners in
the field. Yet they appear to help define ‘ecology’:
(1) The more general the pattern the more likely its ultimate cause lies in statistical
mechanics. In ecological subdisciplines like community ecology and macroecology –
which by their vary nature consider hundreds to billions of interacting individuals and
dozens to thousands of species – it is highly probable that almost all emergent pattern is
253
generated from principles of statistical mechanics. This realization underlies the fact that
the distribution of species abundance is very similar to the distribution of drinking glass
longevities and Cowboy Junkies song performances.
(2) Conversely, the smaller the system under investigation, and the more specific /
idiosyncratic its patterns, the more likely it is governed by system-restricted process.
Process restricted to purely ecological process (i.e. the interactions between individuals
and their environment) is most apt to operate at the small scale including few interacting
individuals and species and may not be generalizable. While typically such research has
been dismissed (especially by theorists) as ‘natural history’ and ‘just-so stories’ yet it
remains the most likely path to identifying ecological mechanism.
(3) The presence of deterministic mechanistic process at the small scale does not mean
that the sum across all such processes at larger scales will not be stochastic. While the
evolutionary trajectory for a given species or clade and the explicit makeup of a particular
individual ecological community may well be explained via deterministic, mechanistic
ecological process, the emergent patterns created across the larger system including these
components may well take on patterns indicative of stochastic process. This will happen
when individual mechanistic processes – and their individual/species-specific parameters
– are themselves randomly distributed.
(4) The existence of universal patterns cannot be used to test for underlying mechanism
beyond that of stochastic logic. The paradox that the same statistical distributions can be
generated in model communities from complete resource competition or neutral models
is explained by realizing both generate dynamical complexities, and it is these which
generate expected patterns. In competition models complexity arises as multiple agents,
each with individualistic resource utilization and dispersal functions, interact in an
elaborate network with multiple other agents often over multiple temporal and spatial
scales. Such interactions can be made even more complicated, but also more realistic, by
inclusion of some degree of stochasticity in resource requirements and dispersal
processes. In neutral models complexity arises as multiple agents, each with a unique
community and metacommunity frequency, undergo a lottery process of replacement in
space. These interactions are further complicated as new species are supplied to the
system across multiple scales via dispersal and speciation.
(5) Important mechanistic insights may be gained by looking for differences in the
parameterized values and functional form of universal distributions, and deviations of
pattern from universal expectations. For instance, in the case of the species area
relationship, it has long been recognized that differences in slope on continents vs. across
archipelagos conveys information about the mechanisms that generate and maintain
diversity. Similarly, distance decay rates provide vital information about niche
characteristics and dispersal capacities in relation to the environmental template of spatial
and temporal variation. So, for example, isolated spruce-fir forests of the Appalachians
demonstrate an almost 3-fold greater DD rate as compared to continuous northern taiga,
and large-fruited, more dispersal-limited plant species have almost twice the DD rate as
compared to smaller-seeded taxa. And the shape of the distance decay relationship
254
informs about whether samples were collected within the same (power-law) or across
different (exponential) communities. Important insights may also be gained by
identifying domains where universal expectations are not generated, such as the lower
than expected frequency of rare and abundant organisms in ecological communities vs.
other complex systems.
(6) Efforts to accurately predict ecological outcomes, even at small deterministic scales,
must be held with great suspect. The mathematics for >2 interacting individuals is
highly complicated and falls within the
framework of the Three Body Problem
in astrophysics, which shows there is no
single general outcome emerging from
the gravitational interaction of 3 or more
bodies (see right panel). Accurate
prediction in such systems is entirely
dependent upon exact knowledge of
starting parameters, and as shown by the
Heisenberg Uncertainty Principle and
Observer Effects, such knowledge is
impossible to obtain. While qualitative
forecasts may be possible, precise,
quantitative predictions of individual events pose great challenges. For example, it might
be possible to predict qualitatively how community-wide species abundance distributions
will be altered by invasions of multiple exotic species. It will be much more difficult,
however, to predict the relative abundance and extent of spatial distribution of each
exotic and native species. Indeed, such an effort may be as Quixotic as attempting to
predict the exact date and magnitude of the next earthquake in the San Francisco Bay
area or the price of a given stock in 2050. As in weather forecasting, while some level of
short-term predictability will often be possible due to understanding current drivers,
recent trajectories, and spatial/temporal autocorrelations, predictions will necessarily
become increasingly imprecise as the forecast period increases. Rather than emphasizing
prediction, ecologists perhaps should spend more effort on understanding the
mechanisms and events that have conspired to generate current and past patterns. Even
when it may be practically impossible to predict the future trajectory of a system, by
looking backward it may be possible to deduce quite accurately when and how specific
mechanisms came into play. In complex, human systems mechanisms are deduced by
historians who analyze post hoc the particular combination of initial conditions and
drivers that generated pattern. Similarly, community ecologists could enhance their
understanding of the spatial and tempral patterns of biodiversity by becoming better
natural historians, and using post hoc analysis to decipher how past events that have left a
lasting influence on current conditions. While it may not be possible to predict future
events with precision, by studying systems over time community ecologists may be able
to establish what factors caused a particular outbreak of hantavirus or bird flu or simply
whether Schrödinger's Cat survived.
255
(7) Statistical mechanics is not physics even though physicists got there first. Statistical
mechanics simply represents the fundamental rules of logic for systems made of a large
number of quasi-independent and variable parts. While this exists for physical systems
(for instance the speed, trajectory, and atomic weights of gas molecules) it is even more
applicable for ecological systems where agents possess many additional degrees of
freedom. In some ways statistical mechanics may become even more fully expressed in
ecological systems as compared to the simpler systems expressed by electrons, gasses,
and fluids.
(8) When searching for mechanism for a given ecological pattern, we must first consider
processes larger than ecology. These include: mathematical logic; statistical mechanics
(mathematical logic of large stochastic systems); physical laws; and chemical laws.
What is left unexplained after this may be due to explicit ecological process (the
interaction of organisms with themselves and the physical environment).
256