Contributed Papers
Conservation Lessons from Long-Term Studies
of Checkerspot Butterffies
PAUL R. EHRLICH*
Center for Conservation Biology
Department of Biological Sciences
Stanford University
Stanford, California 94305
DENNIS D. MURPHY
Center for Conservation Biology
Department of Biological Sciences
Stanford University
Stanford, California 94305
Abstract: Long-term studies of checkerspot butterflies (Euphydryas
spp.) are used to draw conservation lessons with
implications for the design and management of reserves for
invertebrates. We discuss;
1. The importance of determining the structure of and
patterns of mating within demographic units.
2. The lack of congruence between migration and gene
flow.
3. The crucial role of habitat diversity in the protection
of species which otherwise require only small areas of
habitat.
4. The importance of identifying and preserving "reservoir"
populations within metapopulations.
5. The necessity of identifying subtle habitat requirements.
6 The key role of environmental stochasticity in the extinction
of small populations.
7. The difficulty of protecting metapopulations and of introducing
populations into "empty" habitat patches.
We conclude, among other things, that the size of reserves
required to preserve herbivorous insects is often underestimated;
that a "minimum viable metapopulation" approach
to their conservation may be needed; that long-term studies
of the biology of populations of sample invertebrate groups
would greatly aid conservation biologists; and that more
effort is needed to inform the public and decision makers
about the need to conserve populations as well as species.
*Correspondence and requests for reprints should be addressed to
this author.
Paper submitted 1/5/87; revised manuscript accepted 4/7/87.
Resumen: Se discuten las implicaciones que los estudios
sobre las mariposas del genero Euphydryas tienen para el
diseno y manejo de reservas para invertebrados. Se desarrollan
los siguientes temas:
1. La importancia que tiene el determinar la estructura y
los patrones de apareamiento dentro de las unidades
demograficas.
2. La falta de congruencia entre la migraci6n y el intercambio
de genes.
3. La importancia de la diversidad de habitat en la protecci6n
de aquellas especies que requieren areas pe-
quenias.
4. La importancia de identificar y proteger "reservas" de
poblaciones dentro de metapoblaciones.
5. La necesidad de identificar aquellas caracteristicas sutiles
del babitat que requiere la especie.
6 La importancia que tienen los eventos estocasticos del
medio ambiente en la extinci6n de las poblaciones pe-
quenias.
7. La dificultadparaproteger metapoblacionesypara introducirpoblaciones
enfragmentos de habitat "vacios"
("empty" habitat patches).
Conclutmos, entre otras cosas, que es comun el subestimar
el tamanio de las reservas necesario para la sobrevivencia de
las poblaciones de insectos herbivoros; que se necesita enfocar
la conservacion de estos insectos desde un punto de
vista de "metapoblacion minima viable'; que estudios a largo
plazo de la biologia de poblaciones de grupos de invertebrados
selectos seran de mucho valor para los bi6logos; y
que es necesario informar al publico y a las personas que
toman decisiones en el campo de la conservaci6n sobre la
necesidad de conservar poblaciones y especies.
122
Conservation Biology
Volume 1, No. 2, August 1987
This content downloaded from 147.251.87.8 on Tue, 03 Dec 2019 13:41:14 UTC
All use subject to https://about.jstor.org/terms
Ehrlich & Murphy Long-Term Studies of Checkerspot Butterflies 123
Introduction
Field research in population biology tends to be short
in duration and narrow in geographic scope. This situation
presents a major impediment to the application of
basic research to problems confronting conservation biologists,
since conservation biology is by definition concerned
with the maintenance of populations and species
over the long term. "Snapshots" of the histories of single
populations over brief periods repeatedly have proven
to give at best a partial, and at worst an erroneous,
impression of the factors controlling the distribution and
abundance of plants and animals.
One of the few long-term studies of a suite of invertebrate
species over a substantial portion of their ranges
has been carried out on checkerspot butterflies (genus
Euphydryas). The ecology and population genetics of
checkerspot butterflies in the western United States have
been under intensive study for the past 27 years (e.g.,
Ehrlich 1961, 1965; Singer 1971; Ehrlich et al. 1975;
Singer & Ehrlich 1979; Brown & Ehrlich 1980; Singer
1982; Murphy et al. 1983, 1984, 1986; Murphy 1984;
Brussard et al. 1985). Here we summarize some of that
work, highlighting findings of particular importance to
conservation biology, including the design and management
of nature reserves. Our findings should give particular
insight into the conservation biology of herbivorous
insects.
Population Biology
Physical Environment and Population Structure
1. An important lesson for conservation biology from
the earliest studies of Euphydryas is that the movement
of individuals between habitatpatches or reserves does
not necessarily mean that gene flow is occurring. It
also does not mean that, in case of extinction, recolonization
of one patch by individuals from the other
will occur readily.
The most extensively studied colony of Euphydryas
is that of E editha on Stanford University's Jasper Ridge
Biological Preserve. There the butterflies occupy an island
of serpentine-based grassland approximately 9 hectares
in extent, surrounded by chaparral and sandstonebased
grassland. Despite essentially continuous habitat,
uninterrupted by barriers to the butterfly's flight (Ehrlich
1961), the colony was found to consist of three
separate demographic units-populations sufficiently
isolated to have independent dynamics, though not necessarily
enough to be independent genetic units. In each
season, fewer than 5 percent of the individuals marked
in one demographic unit were recaptured in another,
so migration between extant units did not dramatically
influence their dynamics. Nonetheless, this consistent
finding suggests that E. editha exists as "metapopulations";
that is, populations subdivided into groups that
occupy clusters of habitat patches and interact exten-
sively.
Furthermore, because of the combination of late-season
timing of migration and the reproductive strategy
of the butterflies, individual movements are clearly much
more frequent than the movement of genes. Males "plug"
females after they inseminate them, effectively preventing
another insemination for a week or more (Labine
1964). Most females are mated and plugged within hours
of eclosion, so that after the first week of the female
flight period an immigrant male has a comparatively
small chance of being able to inseminate females. Later,
when females become receptive to second inseminations,
the probability of offspring survival becomes very
small because of time constraints on larval development
(discussed below). These same constraints make it unlikely
that any but the earliest migrant females will make
a genetic contribution to the recipient population, or
serve as effective founders.
2. Understanding the scale and structure of populations
is crucial to designing and managing reserves.
The dynamic histories of the three demographic units
on Jasper Ridge, in areas C, G, and H (Ehrlich 1965),
have been documented since 1960 (Fig. 1). Note that
Jasper Ridge area G went extinct in 1964, was re-established
in 1966, and then went extinct again in 1974. If
the mark-recapture studies had addressed the entire
grassland as a single area, and the butterflies had been
considered as a single demographic unit (solid line in
Fig. 1), the extinction events would have been overlooked,
and the safety of the Jasper Ridge population
from stochastic extinctions of its component demographic
units would have been overestimated.
3. Understanding the patterns of mating within demographic
units is also required if conservation efforts
are to succeed, because those patterns may change with
density.
We have found population structure in some locations
other than Jasper Ridge to be much looser (demographic
units much less well defined), and this structure is reflected
in the genetics of the populations. Spatial autocorrelation
of allele frequencies in Great Basin
populations, for example, suggests that discrete demographic
units may occupy habitat areas of dozens of
square kilometers and individuals may move much
greater distances (Wright et al. 1987). Yet, even in more
or less continuous habitat, special physical features may
be crucial to population structure. In Colorado, where
diffuse demographic units of Euphydryas editba occupy
many square kilometers of montane sagebrush communities,
hilltops appear to constitute important rendezvous
points at which males aggregate when population
densities are low and to which females travel to seek
mates (Ehrlich & Wheye 1986).
Conservation Biology
Volume 1, No. 2, August 1987
This content downloaded from 147.251.87.8 on Tue, 03 Dec 2019 13:41:14 UTC
All use subject to https://about.jstor.org/terms
124 Long-Term Studies of Checkerspot Butterflies Ehrlich & Murphy
10,000 I If I I I I I I I I I I I I I I I 1
5000 6646 6877 747 788 82 45166266 6877277AREA C 'it
3000*
1000. I
CIO 500
300 .
Fiur 1.JseIig'ouainsiecagshdtesprtedmgahcuisntbe idnife(lt)
100 r
50 o AREA H
30
Ii I
1960 62 64 66 68 70 72 74 76 78 80 82 84 85 1960 62 64 66 68 70 72 74 76 78 80 82 84 85
Figure 1. Jasper Ridge population size changes had the separate demographic
Population size changes of the demographic units in Jasper Ridge areas C, G, and H (right). Note extinction
and recolonization events in area G.
At low population densities, such mating aggregations
may play a crucial role in keeping a population viable.
At high densities, however, males that move to hilltops
may have relatively small chances of mating (Ehrlich &
Wheye 1986). Thus, while such an environmental feature
may seem unimportant if the study of a demographic
unit was done in a year of abundance, its inclusion
may prove critical if a reserve is sliced out of continuous
habitat.
4. Topographic heterogeneity or other aspects of
habitat diversity may be key determinants of the habitat
suitability of an area for the maintenance of populations
of many insect herbivores that otherwise
require only small areas of habitat.
Topography is a crucial consideration in the conservation
biology of threatened populations of Euphydryas
editha bayensis in the San Francisco Bay area of California,
the ecotype represented atJasper Ridge. For more
than a century, the ecotype has been confined to patches
of serpentine-based grassland. Earlier it may have been
much more widespread, but lack of knowledge of the
California grassland flora before the introduction of now
virtually ubiquitous Eurasian weeds makes reconstruction
of prior distributions impossible. For the last century,
areas of suitable habitat have been declining in the
face of development, and populations in those serpentine
habitats that remain are often subject to extinction
as a result of climatic events such as the California drought
of the mid-1970s (Ehrlich et al. 1980) and the "El Nino"
weather of 1982-1983 (Dobkin et al. 1987).
The sizes of populations of the Bay Area ecotype are
controlled in a density-independent fashion. Eggs are
laid in the spring, and upon hatching, larvae are immediately
in a race to reach the size at which they can
enter an obligatory diapause in which they pass the summer
dry season of California's Mediterranean climate.
The larvae lose this race if they cannot grow large enough
before their annual food plants senesce. There is a complex
phase relationship in the timing of the insect's developmental
cycle and that of the plant (Singer 1972,
Singer & Ehrlich 1979). This relationship is mediated
by the macroclimate interacting with the microclimate,
and the latter varies significantly with slope and exposure
(Dobkin et al. 1987).
The remaining patches of serpentine habitat most suitable
for the long-term maintenance of populations are
the largest and most topographically diverse. A large area
of habitat alone is clearly not sufficient to guarantee the
persistence of checkerspots. Extinctions have been observed
in large habitats that lack topographic heteroConservation
Biology
Volume 1, No. 2, August 1987
This content downloaded from 147.251.87.8 on Tue, 03 Dec 2019 13:41:14 UTC
All use subject to https://about.jstor.org/terms
Ebrlich & Murphy Long-Term Studies of Checkerspot Butterflies 125
geneity. The Coyote Reservoir population, the most
densely populated of known Bay Area colonies in 1971,
was extinct by 1976. The Coyote Reservoir site consists
almost entirely of an east-facing slope. Topographic heterogeneity
allows some areas within a serpentine patch
to have a favorable microclimate for larval development
in any weather regime; north-facing slopes will be favored
in most years, while south-facing slopes contribute
to population increases in wet years. High survival
rates on south-facing slopes in wet years appear to push
population numbers high enough in topographically diverse
locations to allow demographic units to survive
substantial declines during inevitable drought years.
5. If conservation options are limite4 top priority
should be given to identifying and preserving reservoir
populations, if any.
The overall pattern in the San Francisco Bay area appears
to be one in which just one population south of
the bay (Morgan Hill, Santa Clara County) is large enough
(Fig. 2) and stable enough to act as a "reservoir population"
of E editha bayensis (Fig. 3). A reservoir population
may be defined as a demographic unit which can
be expected to persist through essentially all natural
environmental perturbations. A reservoir population is
one that can act as a relatively permanent source of
colonists to habitats that cannot permanently maintain
what might be called "satellite" populations. We suggest
EWN
JR /
~MH
i a r-~~~ -\\ "^ E~~CR
Figure 2. Distribution of San Francisco Bay Area serpentine
soils, including grassland habitats of Euphydryas
editha. EW = Edgewood Park (part of the San
Mateo-Woodside grassland system), JR = Jasper
Ridge, CR = Coyote Reservoir, and MH = Morgan
Hilt
that since all demographic units will eventually become
extinct, that the terms "reservoir population" and "satellite
population" only be used where it is estimated that
the former will persist at least 10 times as long as the
latter.
The extensive San Mateo-Woodside system of serpentine-based
grasslands has been severely fragmented, and
only at Edgewood Park is the remnant habitat greater
than 10 ha in extent. The ability of the population there
to function as a reservoir is doubtful, as it has undergone
dramatic population fluctuations itself (> 105 individuals
in 1981 to <102 in 1985). Al other sites, including
Jasper Ridge Reserve, are too small for truly long-term
maintenance of the butterfly's populations through extended
drought. It seems likely that if the California
drought had continued for another year, all three demographic
units in the reserve might have become extinct.
Indeed, weather conditions seem not to have been optimal
since 1981, and neither the Jasper Ridge area C
nor area H population has been able to rebuild to the
high density levels of the early 1970s.
Resources and Population Dynamics
6. Natural history investigations should be undertaken
before a reserve is designed in an attempt to
identify subtle habitat requirements that could potentially
play key roles in the long-term survival of populations
to be protected
Independent fluctuations in size of the demographic
units on Jasper Ridge result, at least in part, from rather
subtle differences in the means by which larvae survive
to diapause each spring. In both populations, survival is
greatest for eggs laid on north-facing slopes or other
comparatively mesic areas by early emerging females
from south-facing slopes. In the area H demographic
unit, many survivors are larvae that wander from drying
Plantago and locate the flowers of later-senescing Orthocarpus
densiflorus on which the larvae feed to diapause.
At nearby Edgewood Park, females often oviposit
directly on Orthocarpus (Singer 1971). In contrast, many
survivors to diapause in Jasper Ridge area C come from
eggs laid on Plantago on soil tilled by gophers (Thomomys),
where plants are large and exhibit delayed senescence-although
Orthocarpus plays a role as a
secondary food source in this area also.
While the primary resource determinants of habitat
suitability clearly are oviposition and larval host plants,
other resources may play crucial roles in population
persistence and habitat suitability. The availability of
nectar, for instance, is very important in maintaining egg
production by females in the latter stages of reproduction.
A paucity of middle- and late-season nectar has been
implicated in the extinction of the Jasper Ridge area G
population (Murphy et al. 1983). The presence of standing
water has been shown to affect population structure
Conservation Biology
Volume 1, No. 2, August 1987
This content downloaded from 147.251.87.8 on Tue, 03 Dec 2019 13:41:14 UTC
All use subject to https://about.jstor.org/terms
126 Long-Term Studies of Checkerspot Butterflies Ebrlicb & Murphy
Figure 3. Newly emerged Euphydryas editha bayensis female in the area H serpentine grassland of tbe Jasper
Ridge Biological Preserve.
inEuphydryas (Ehrlich & White 1980) and in Chlosyne,
a genus closely related to Euphydryas (Schrier et al.
1976). Males travel relatively long distances to sip from
moist earth around puddles, presumably to obtain sodium
(Arms et al. 1974, Adler & Pearson 1982). The
presence of both nectar and standing water, additionally,
have been shown to be determinants of overall butterfly
species numbers in Great Basin riparian communities
(Murphy & Wilcox 1986).
7. Under severe environmental conditiomn the degree
of oligophagy mtay be a crucial determinant of
extinction pronenes&
We used mark-recapture studies and detailed statistical
analysis in an attempt to identify adult population
dynamic parameters that might be unique to particularly
extinction-prone species. But, population studies of
highly endangered Euphydryas editha and ubiquitous,
unthreatened Euphydryas chalcedona on Jasper Ridge
revealed few differences (Murphy et al. 1986). Indeed,
both species appear to move freely within suitable habitat
and exchange individuals at low rates with other
colonies at distances of several kilometers or more.
Ease of recolonization well may be determined, however,
primarily by patterns of resource use, rather than
interspecific differences in vagility. The Bay Area ecotypes
of Euphydryas chalcedona feed on perennial host
plants and appear to accept a wide variety of hosts. E.
editha by contrast, feeds on annuals and exhibits narrower
host tolerance. Furthermore, within the California
range of E editha populations feeding on perennials
(e.g., onPedicularis densiflora in the inner coast ranges
and on Castilleja nana in the alpine Sierra Nevada) have
suffered fewer documented "natural" extinctions than
those feeding on annuals (e.g., on Plantago and Orthocarpus
in outer coast ranges and on Collinsia tinctoria
in the Sierra foothills). These observations suggest that
wide host tolerance and the incorporation of less ephemeral
hosts into the larval diet may confer resistance to
extinction.
Dispersal, Patch Dynamics, and Gene Flow
8. Recolonization of empty patches may not occur in
most years; many populations may be re-established
in rare years of explosive dispersaL
Conservation Biology
Volume 1, No. 2, August 1987
This content downloaded from 147.251.87.8 on Tue, 03 Dec 2019 13:41:14 UTC
All use subject to https://about.jstor.org/terms
Ehrlich & Murphy Long-Term Studies of Checkerspot Butterflies 127
The need to monitor populations for many generations
is underscored by our observation of populations
of Euphydryas editha in southern California. Individuals
in those populations exhibit rather sedentary behavior
similar to Jasper Ridge butterflies. Dispersal in southern
California populations increases, however, in particularly
dry years when oviposition plants and adult nectar
sources are sparse. Yet, in 1977, a year of substantial
rainfall in southern California, the populations exploded
in size, and adults dispersed en masse from population
centers (Murphy & White 1984). These circumstances
(not repeated since) allowed insight into the abiotic and
biotic forces that can determine patterns of dispersal in
the species.
Dry winters seem to result in increased vagility because
larval host plants and adult nectar sources are
sparse from the lack of precipitation and because postdiapause
larvae defoliate a higher proportion of available
host plants before adult females emerge to oviposit. When
wet years precede dry years, defoliation is intensified in
the dry years because larval populations are larger. Furthermore,
consecutive years of adequate rainfall apparently
result in butterfly population build-ups such that
entire standing crops of host plants may be defoliated;
massive postdiapause mortality of larvae then occurs and
virtually all emerging adults disperse. Population size,
dispersal, and the likelihood of local extinction, therefore,
are determined by density-independent factors in
most years, but by density-dependent factors in a very
few years. These rare years, however, appear to be one
of the keys to the natural recolonization of habitats unable
to support populations over the long term. Thus
not only may gene flow and recolonization potential be
difficult to estimate from data on movement of individuals,
but the amount of dispersal itself may vary dramatically
from year to year.
Minimum Viable Population Size
9. Environmental stochasticity, especially variation in
host quantity, quality, and phenology, is the major
cause of extinction, particularly in smallpopulations.
A general constancy has been found in allozyme frequencies
and in average heterozygosity in California Euphydryas
populations, despite wide fluctuations in
population sizes and after local "bottlenecks" where Ne
may have fallen well below 50. This suggests that deleterious
genetic effects, such as inbreeding depression
or loss of variation, may be relatively unimportant in the
persistence of Euphydryas populations.
Although rates of dispersal from Jasper Ridge demographic
units appear to be inversely correlated with population
size (Gilbert & Singer 1973), recent observations
in area H indicate that E editha populations can persist
for numerous generations at surprisingly small sizes. It
appears that levels of dispersal in Euphydryas are complexly
related both to population density itself, and to
the abundance (or scarcity) of resources. The population
in Jasper Ridge area H has been estimated to consist
of less than 100 individuals for four consecutive years
(in two of those years, Ne was certainly less than 10!).
Thus, while the role of demographic stochasticity has
not been determined and could be significant in certain
circumstances, our present evidence indicates that small
population size in and of itself will not ordinarily lead
to population extinction. Small populations, of course,
are particularly susceptible to extinction due to environmental
events such as drought.
Even large populations are at risk, however, when
drought reaches dramatic proportions (on the order of
half "normal" rainfall), or if several less extreme drought
years occur in succession. Such conditions are effectively
"natural" catastrophes. Likewise, the major cause
of anthropogenic extinction of Euphydryas editha populations
has not been a gradual deterioration of local
environments leading eventually to population extinctions,
but rather a catastrophic loss of habitats. For example,
in the San Francisco Bay area the paving over of
patches of serpentine grassland has led to the endangerment
of E editha bayensis.
Management and Politics
10. Habitatpatches and metapopulations are difficult
to protect
In the Outer Coast Ranges of central California, native
grassland persists as remnant patches on serpentine and
other highly weathered soils. Unique chemical characteristics,
including high levels of magnesium, nickel, and
chromium, and low levels of calcium, nitrogen, and
phosphorus, appear to account for the resistance of serpentine
soils to invasion by non-native grasses and forbs,
and the resultant exclusion of many native plant species.
The high degree of ecotypic differentiation and endemism
in serpentine-based grassland communities make
them an obvious target for conservation efforts, but they
exist as dispersed small units, ranging from less than a
hectare to a few thousand hectares in extent, with most
less than 10 ha. If a continuous national park or other
reserve of reasonable configuration was established that
contained all the serpentine patches in the San Francisco
Bay area, it would necessarily include the homes and
businesses of millions of people.
Most of the native grassland remnants are privately
owned, hence conservation efforts must follow one of
two paths: acquisition of property, obviously an expensive
proposition; or invocation of the Endangered Species
Act. Unfortunately, the 1982 reauthorization of the
Act largely removed protection from "take" for endangered
and threatened plants. Euphydryas editha bayensis
is, at present, the only animal restricted to the
native grassland community that has been proposed for
federal endangered species protection.
Conservation Biology
Volume 1, No. 2, August 1987
This content downloaded from 147.251.87.8 on Tue, 03 Dec 2019 13:41:14 UTC
All use subject to https://about.jstor.org/terms
128 Long-Term Studies of Checkerspot Butterflies Ehrlich & Murphy
Two major problems exist in using this butterfly as a
tool to protect serpentine-based grassland habitat. First,
because some number of demographic units within the
metapopulation are always extinct and their sites awaiting
recolonization, the butterfly may not be present in
many of the remnants with the highest plant diversity
(Murphy & Ehrlich 1987). Since the Endangered Species
Act focuses on protecting "species," temporarily vacant
habitat patches have no legal protection. Second, invertebrates
can be protected only if an entire described
subspecies is threatened or endangered. An opponent
to the listing of Euphydryas editha bayensis has challenged
the validity of the traditional taxonomic treatment
of the subspecies. As biologists well know, the
subspecies is an arbitrary taxonomic category. This challenge,
which promises to become a protracted battle,
ultimately may compromise future attempts to protect
infraspecific genetic diversity of invertebrates.
Finally, because small reserves in largely urban areas
have neighbors, maintaining good relations with those
neighbors may be critical to the integrity of the reserve.
Jasper Ridge has an extremely active docent program,
and docents lead carefully guided tours of the facility
on which, among other things, the importance of Euphydryas
work is explained. Tours were routed away
from area H after it was discovered that, even with the
most careful supervision, it was impossible to prevent
trails from widening in response to increased foot traffic.
The area around the preserve is suburban, and some
homes share common boundaries with it. Educational
programs, fencing, and patrolling at Jasper Ridge have
generally been able to keep intrusion to a minimum in
the past decade, but constant vigilance has been required,
and in one case Stanford University has had to
resort to extensive and expensive litigation to maintain
the integrity of the preserve.
Simply finding the funds to establish small reserves to
protect the elements of a metapopulation is only the
first, and in many cases the easiest, step. Finding the
additional funds and engaging the time of interested
people to manage them are clearly major challenges.
Furthermore, it has yet to be determined whether, even
if the sites that can support the demographic units of a
metapopulation (such as that of E e. bayensis) can be
preserved, the metapopulation itself can be protected.
It may be, for example, that complete urbanization of
intervening areas would so interfere with dispersal and
recolonization that the entire system would collapse to
extinction.
Thus, political problems often involve more than finding
ways to protect reserves. Conservation biologists
working with butterflies and other vagile organisms must
always be concerned with land use patterns adjacent to
reserves. They must work to make those areas more
hospitable to organisms that disperse out of the protected
area. Conservation biologists also must actively
campaign for abatement threats such as acid precipitation
and rapid anthropogenic climate change. No amount
of careful management of reserves will prevent extinctions
from such adverse conditions.
11. The ease of maintaining a metapopulation or
re-establishing extinct metapopulations by introducing
insects into "empty" habitat is difficult to predict
It may be a long time before the success of a transplant
experiment can be evaluated, even when a univoltine
insect is involved. Re-establishment of demographic units
known to have gone extinct and their establishment in
"empty" habitats that seem suitable both appear to be
possible management options for the conservation of
Euphydryas Two transplantation experiments have been
carried out in connection with our studies of the allozyme
genetics of E editha The failure of one transplant
of 500 larvae to Agua Fria in the foothills of the California
Sierra Nevada was due, at least in part, to human disturbance
of a portion of the habitat. Another, to Arroyo
Bayou in the Inner Coast Range of central California,
again with 500 larval colonists, appears to have resulted
in establishment of a demographic unit, although there
are signs that natural immigration also played a part.
Our most interesting result with transplant experiments
thus far have been achieved with another checkerspot
species, Euphydryas gillettii. One of two attempts
to transplant a colony for northwestern Wyoming to
central Colorado was successful (Holdren & Ehrlich
1981). A population founded in 1977 with 83 egg masses
(about 104 eggs) has persisted through the 1986 field
season. Most important to conservation is that in nine
generations the population has fluctuated in size, but
has never exceeded 200 individuals, and has not spread
from the immediate vicinity of the release, despite extensive
adjacent and apparently suitable habitat (Ehrlich
& Holdren, unpublished data). The population thus has
never expanded to occupy a variety of sites and slopes
and, therefore, may yet go extinct. Indeed, this underscores
the difficulty of assessing habitat quality or suitability
in general.
Unfortunately, population density constitutes the best
bioassay of habitat quality. Furthermore, even if habitat
is adequate to sustain a population for many ordinary
years, the key to long-term habitat suitability is whether
a habitat can support a population during extreme years
(for E editha, years of drought or deluge).
We are now embarking on a series of experiments to
re-establish E editha bayensis populations in isolated
patches of habitat that we know previously supported
demographic units, and in others that appear suitable
but whose previous record of occupancy (or lack of it)
is unknown. This work, which may involve as many as
70 patches, should greatly expand our understanding of
the problems that managers will face in maintaining these
and related organisms with similar "metapopulation dynamics."
It is, however, already clear from work with
Conservation Biology
Volume 1, No. 2, August 1987
This content downloaded from 147.251.87.8 on Tue, 03 Dec 2019 13:41:14 UTC
All use subject to https://about.jstor.org/terms
Ehrlicb & Murphy Long-Term Studies of Checkerspot Butterflies 129
Euphydryas and with many other organisms (e.g., Ehrlich
1986, 1987) that simply introducing a group of
colonists into entirely suitable habitat is no guarantee
of establishment.
12. Introduced mammalian herbivores may have
complex impacts on endangered herbivorous insects.
In the Bay Area, grassland habitat supporting high
densities of Euphydryas editha host plants are early and
intermediate successional stages leading to a stable climax
community dominated by perennial bunchgrasses
(Murphy & Ehrlich 1987). Native ruminant grazers (including
the now extirpated tule elk and pronghorn
antelope) undoubtedly played a role in creating successional
habitat. In the absence of these species, domesticated
livestock now affect the landscape physically and
biologically and can alter profoundly the plant species
composition, hence suitability of habitat for Euphydryas
editha
The physical presence of cattle directly affects the
butterfly in its nonvagile stages. Mortality from being
stepped on or buried under cow pats obviously can be
significant, and is proportional to the density of cattle
in a given area (White 1986). A more subtle effect of
cattle is the physical alteration of the landscape from
the action of hooves, including the terracing of steep
slopes. Larvae are often found exclusively on cattle terraces
on north-facing slopes in winter when the sun is
low in the sky, sunlight is at a premium, and the terraces
receive more isolation than the slopes.
The effect of cattle on the species composition of
grasslands is more complicated. The physical action of
hooves provides continual ground disturbance, which
may or may not favor the butterfly's host plants. Selective
grazing of introduced annual grasses reduces their ability
to invade and outcompete the native flora, although native
forbs and bunchgrasses are also consumed by cattle.
Cattle droppings add concentrated nutrients that favor
the establishment of introduced grasses in localized areas.
The areas around salt licks and watering troughs can be
almost completely dominated by introduced grasses. On
both grazed and ungrazed sites, the butterfly host plants
are interspersed with patches of introduced grasses. In
ungrazed sites, invasion of introduced grasses appears
to be facilitated by large gopher mounds. The dynamic
balance of these processes is poorly understood.
13. Endangered populations of insects may be
"studied to death."
In the course of the intensive study in 1981 in Jasper
Ridge area H (above), it became apparent that
roughlyl50-200 investigator hours spent collecting in
an area of approximately 2.5 ha had a significant impact
on the vegetation of the area. Although investigators
took care always to wear tennis shoes (rather than cleated
boots), some crushing of areas of herbage was obvious
at the end of the experiment, although no long-term
damage to the vegetation appears to have been done.
We were concerned, however, that some pupae, egg
masses, and larvae of the butterflies (which are close to
the soil surface) may have been destroyed by trampling,
which, in turn, could have serious consequences for the
population. Sadly, research areas, especially those close
to major centers, tend to be small and thus vulnerable
to serious disturbance from research activities themselves
and to interference between research projects. At
the Jasper Ridge Preserve and the Rocky Mountain Biological
Laboratory, the field stations we know best, both
problems are causing growing concern among researchers
and managers. Since it is questionable whether intensive
mark-recapture experiments in such limited
habitat areas can add to one's knowledge of the population
(Murphy et al. 1986), it seems to us that managers
should advocate instead the use of noncapture methods
to estimate population sizes over the long-term.
Conclusions
What are the most important lessons for conservation
biology to come out of the Euphydryas work? We think
the first is that the size requirements of reserves designed
to protect herbivorous insects may often be
underestimated. Populations of such insects seem often
to be subject primarily to density-independent population
regulation, and thus their waiting time to extinction
may be relatively short. A reserve that simply encompasses
the location of a single demographic unit, even
if it provides adequate marginal buffering and security
from intrusion, may soon lose the organisms it was designed
to protect. If such insects are to be conserved,
in many if not most cases, a series of demographic units,
selected for the diversity of their habitats, should be set
aside. Where necessary, either corridors among the sites
should be established or active management planned so
that stochastic extinctions of demographic units can be
followed by assured recolonizations.
In short, the "minimum viable population" sort of
approach that many conservation biologists favor for
vertebrate species, while risky enough for them, may
not work at all for many invertebrates. Instead, concern
there should be shifted toward "minimum viable metapopulations."
This means that, in some cases, vertebrates
or plants will prove inadequate as "umbrella species"
for invertebrates. While the grizzly bear, because of its
gigantic home range requirements, may protect many
invertebrates, including checkerspot butterflies, in the
Yellowstone ecosystem, preserving one patch of serpentine
soil-based grassland for the last known population
of the San Mateo Thornmint (Acanthomintha
obovata duttoni) will do little to aid in the long-term
maintenance of Euphydryas editha bayensis.
It is also clear that the level of effort that has gone
into uncovering the details of Euphydryas biology cannot
be expended on more than a minute fraction of
Conservation Biology
Volume 1, No. 2, August 1987
This content downloaded from 147.251.87.8 on Tue, 03 Dec 2019 13:41:14 UTC
All use subject to https://about.jstor.org/terms
130 Long-Term Studies of Checkerspot Butterflies Ehrlich & Murphy
potentially endangered invertebrate species most of
which, even in the temperate zones, have yet to be
recognized as endangered. This leads to two conclusions.
First, strategies for invertebrate conservation must
be established on the basis of long-term studies of a few
species or taxonomic groups chosen as representative.
Second, in the absence of better information, conservative
assumptions should be made about the long-term
requirements of invertebrates. We probably should make
a best guess about the habitat area required to preserve
a species, and then at least double that area to reach a
recommended minimum reserve size.
We cannot emphasize too strongly the importance of
making studies of sample groups over the long term.
We have, for example, drastically altered our original
conclusions repeatedly after observing populations under
different climatic regimes. Doing a long-term study
also often allows time for observing the behavior of
different ecotypes of the same species, which in the
Euphydryas system led to drastic revisions of many early
ideas about the biology of these organisms. Fortunately,
the National Science Foundation is already aware of the
need to conduct long-term studies (e.g., Callahan 1984),
and conservation biologists in the United States should
do all they can to encourage increased funding by Congress
of the Foundation's program to support such re-
search.
Finally, there is a great need for conservation biologists
to emphasize the conservation of populations along
with the conservation of species. We have seen here
how the attrition of populations of Euphydryas edit*a
bayensis has endangered the entire ecotype. A similar
attrition, we believe, is gradually endangering many, if
not most, other organisms. Focusing on the loss of species
is focusing on the closing scene of what is ordinarily
a long process. It leads to underestimates of the scale
and severity of the current extinction crisis. When public
statements are made on that crisis, the critical importance
of the loss of populations should always be
mentioned.
Literature Cited
Adler, P.E.; Pearson, D.L. Why do male butterflies visit mud
puddles? Can. j Zool. 60:322-325;1982.
Arms, K; Feeney, P.; Lederhouse, R. Sodium: Stimulus for puddling
behavior by tiger swallowtail butterflies (Papilio glaucus).
Science 185:373-374;1974.
Brown, I.L.; Ehrlich, P.R. Population biology of the checkerspot
butterfly Euphydryas chalcedona Structure of the Jasper
Ridge colony. Oecologia 47:239-251;1980.
Brussard, P.F; Ehrlich, P.R.; Murphy, D.D.; Wilcox, B.A.; Wright,
J. Genetic distances and the taxonomy of checkerspot butterflies
(Nymphalidae: Nymphalinae).j Kans. Ent Soc. 58:403-
412;1985.
Callahan, J.T. Long term ecological research. BioScience 34:363-
367;1984.
Dobkin, D.S.; Olivieri, I.; Ehrlich, P.R. Rainfall and the interaction
of microclimate with larval resources in the population
dynamics of checkerspot butterflies (Euphydryas editha) inhabiting
serpentine grasslands. Oecologia 71:161-166;1987.
Ehrlich, P.R. The attributed of invaders and invasion processes:
vertebrates. In: H.A. Mooney, J.A. Drake, eds. Biological Invasions:
A Global Perspective.
Ehrlich, P.R. Which animal will invade? Il: H.A. Mooney, J.A.
Drake, eds. Ecology of Biological Invasions of North America
and Hawaii. Ecological Studies, Vol. 58 New York: SpringerVerlag;
1986:79-95.
Ehrlich, P.R. The population biology of the butterfly, Euphydryas
editha II. The structure of the Jasper Ridge colony.
Evolution 19:327-336;1965.
Ehrlich, P.R.; Murphy, D.D. The population biology of checkerspot
butterflies (Euphydryas) - a review. Biol. Zentral.
100:613-629;1981.
Ehrlich, P.R; Murphy, D.D.; Singer, M.C.; Sherwood, C.B.; White,
R.R.; Brown, I.L. Extinction, reduction, stability and increase:
The responses of checkerspot butterfly populations to the California
drought. Oecologia 46:101-105;1980.
Ehrlich, P.R.; White, R.R. Colorado checkerspot butterflies: Isolation,
neutrality, and the biospecies. Amer. Nat. 115:328-
341;1980.
Ehrlich, P.R.; White, R.R.; Singer, M.C.; McKechnie, S.W.; Gilbert,
L.E. Checkerspot butterflies: A historical perspective.
Science 188:22 1-228;1975.
Ehrlich, P.R.; Launer, A.E.; Murphy, D.D. Can sex ratio be determined?
The case of a population of checkerspot butterflies.
Amer. Nat. 124:527-539;1984.
Ehrlich, P.R.; Wheye, D. Non-adaptive "hilltopping" behavior
in male checkerspot butterflies (Euphydryas editba). Amer.
Nat. 127:447-483;1986.
Gilbert, L.E.; Singer, M.C. Patterns of dispersal and gene flow
in a butterfly species. Amer. Nat 107:58-72;1973.
Holdren, C.E.; Ehrlich, P.R. Long-range dispersal in checkerspot
butterflies: Transplant experiments with Euphydryas gillettii.
Oecologia 50:125-129;1981.
Holdren, C.E.; Ehrlich, P.R. Ecological determinants of food
plant choice in the checkerspot butterfly Euphydryas editha
in Colorado. Oecologia 52:417-423;1982.
Labine, P.A. Population biology of the butterfly, Euphydryas
editba I. Barriers to multiple inseminations. Evolution 18:335-
336;1964.
McKechnie, S.W.; Ehrlich, P.R.; White, R.R. Population genetics
of Euphydryas butterflies. I. Genetic variation and the neutrality
hypothesis. Genetics 81:571-594;1975.
Conservation Biology
Volume 1, No. 2, August 1987
This content downloaded from 147.251.87.8 on Tue, 03 Dec 2019 13:41:14 UTC
All use subject to https://about.jstor.org/terms
Ehrlich & Murphy Long-Term Studies of Checkerspot Butterflies 131
Mueller, L.E.; Wilcox, B.A.; Ehrlich, P.R.; Murphy, D.D. A direct
assessment of the role of genetic drift in determining allele
frequency variation in populations of Eupbydryas editba Genetics
110:495-51 1;1985.
Murphy, D.D. Nectar sources as constraints on the distribution
of egg masses by the checkerspot butterfly, Eup*ydryas cbalcedona
(Lepidoptera: Nymphalidae). Environ. Entomol.
12:463-466;1983.
Murphy, D.D. Butterflies and their nectar plants: The role of
the checkerspot butterfly Eupbydryas editba as a pollen vector.
Oikos 43:113-117;1984.
Murphy, D.D.; Launer, A.E.; Ehrlich, P.R. The role of nectar
feeding in egg production and population dynamics of the
checkerspot butterfly, Eupbydryas editba Oecologia 56:257-
263;1983.
Murphy, D.D.; Menninger, M.S.; Ehrlich, P.R. Nectar source
distribution as a determinant of oviposition host species in
Euphydryas cbalcedona Oecologia 62:269-271 ;1984.
Murphy, D.D.; Menninger, M.S.; Ehrlich, P.R.; Wilcox, B.A. Local
population dynamics of adult butterflies and the conservation
status of two closely related species, San Francisco Bay, USA.
Biol. Conserv. 37:201-233;1986.
Murphy, D.D.; White, R.R. Rainfall, resources, and dispersal in
southern populations of Eupbydryas editba (Lepidoptera:
Nymphalidae). Pan-Pac. Ent. 60:350-354;1984.
Murphy, D.D.; Wilcox, B.A. Butterfly diversity in natural forest
fragments: A test of the validity of vertebrate-based management.
In:J. Verner et al., eds. Wildlife 2000: Modeling Wildlife
Habitat Relationsbips. Madison, WS: Univ. Wisconsin
Press;1986:287-292.
Murphy, D.D.; Ehrlich, P.R. The conservation biology of California's
remnant native grasslands. In: Mooney, H.A., L.F. Huenneke,
eds. California Grasslands: Structure and Productivity;
(in press)1987.
Schrier, R.D.; Cullenward, Mj.; Ehrlich, P.R.; White, R.R. The
structure and genetics of a montane population of the checkerspot
butterfly, Cblosynepalla Oecologia25:279-289;1976.
Singer, M.C. Evolution of foodplant preferences in the butterfly
Eupbydryas editba Evolution 35:383-389;1971.
Singer, M.C. Complex components of habitat suitability within
a butterfly colony. Science 173:75-77;1972.
Singer, M.C. Quantification of host specificity by manipulation
of oviposition behavior in the butterfly Eupbydryas editba
Oecologia 52:224-229;1982.
Singer, M.C.; Ehrlich, P.R. Population dynamics of the checkerspot
butterfly Eupbydryas editba Fortrscbr. Zool. 25:53-
60;1979.
Wheye, D.; Ehrlich, P.R. The use of fluorescent pigments to
study insect behavior: Investigating mating patterns in a butterfly
population. Ecol. Entomol. 10:231-234;1985.
White, R.R. Pupal mortality in the bay checkerspot butterfly.
j Res. Lepid. 25:52-62;1986.
White, R.R.; Singer, M.C. Geographical distribution of host plant
choice in Eupbydryas editba (Nymphalidae). j Lepid. Soc.
28:103-107;1974.
Wilcox, B.A.; Murphy, D.D. Conservation strategy: The effects
of fragmentation on extinction.Amer. Nat 125:879-887;1985.
Wright, J.; Brussard, P.F.; Murphy D.D.; Ehrlich, P.R.; Wilcox,
B.A. Genetic population structure in Eupbydryas editba: The
search for an appropriate spatial scale. Evolution (submitted
for publication)
Conservation Biology
Volume 1, No. 2, August 1987
This content downloaded from 147.251.87.8 on Tue, 03 Dec 2019 13:41:14 UTC
All use subject to https://about.jstor.org/terms