Ann. Rev. Eco/. Svst. 1985. 16:447-79
Copyright © 1985 by Annual Reviews Inc. All rights reserved
BIOLOGICAL ASPECTS OF
ENDEMISM IN HIGHER PLANTS
Arthur R. Kruckeberg
Department of Botany, University of Washington, Seattle, Washington 98195
Deborah Rabinowitz
Section of Ecology and Systematics, Cornell University, Ithaca, New York 14853
INTRODUCTION
Rarity of all sorts piques the curiosity of humans. Rare objects, whether
discoveries of natural origin or artifacts of cultures, arc avidly sought by
collectors and are treasured, housed, and exhibited for the benefit of others.
This fascination with that which is rare may partially explain the naturalists' and
systematists' time-honored preoccupation with rare and endemic taxa. To be
sure, there are sound justifications for acquiring and studying rare organisms.
But we suspect that the rarity-seeking syndrome in humans has fostered much
of the biologist's preoccupation with rare and endemic plants and animals.
The late twentieth century has witnessed exceptional concern for rare organisms
and at the same time has seen many rare plants and animals brought close
to extinction (22). The extirpation of raretaxa as well as efforts to preserve rare
organisms is the product of the vast human influence on the world's biota.
Threatened or endangered taxa have received prime press coverage and, in
some instances, legal protection. The Rare and Endangered Species Act of
1973 (US 92nd Congress) and subsequent congressional actions constitute one
such response to the threatened loss of the rarer examples of biological diversity.
We have no problem, then, in finding a wealth of literature on efforts to
conserve rare species. Sources of information on what is rare and on the
magnitude of the threat to survival range from the Federal Register's list of
447
0066-4162/85/1120-0447$02.00
Annu.Rev.Ecol.Syst.1985.16:447-479.Downloadedfromwww.annualreviews.org
AccessprovidedbyMasarykUniversityon12/04/19.Forpersonaluseonly.
Quick links to online content
FurtherANNUAL
REVIEWS
448 KRUCKEBERG & RABINOWITZ
candidate organisms to the popular press, conservation society journals, symposia
(e.g. 68, 92, 100, 105, 1 14), and books (22, 26, 55, 65).
Biologists welcome the public attention given to the urgent need for action
against the threatened extinction of rare biota. But aside from defining their
rarity, what can biologists say about the attributes--ecological and evo1utionary--ofrare
taxa? Though we recognize that endemism and rarity have several
meanings (20, 21, 37, 72), we direct our reply to this question primarily to that
form of rarity called narrow endemism. Narrow endemic taxa are those that
occur in one or a few small populations (21) and hence are confined to a single
domain or to a few localities. We consider only vascular plants.
If, from some magical Landsat image, we knew the ranges to the nearest
square kilometer of all of the world's higher plant species (putting aside the
difficulties in delimiting ranges), we couldlook at theirfrequency distributions.
The variation in ranges would be huge, from plants that are nearly "cosmopolites,"
such as Chenopodium album, to plants that occupy small areas, like
Betula uber, which resides in less than a hectare. This distribution of ranges
would have two tails; although we are concerned mainly with the endemics of
the left-hand tail of the curve, comparisons of the two extremes would be of
interest too.
Are plants with small ranges a heterogeneous group without other common
characteristics or does the term endemic reflect an assemblage with special
attributes? This review focuses on the biological implications ofendemism: (a)
How and where does endemism occur? (b) What properties do endemics have?
(c) What processes occur uniquely in their populations?
BRIEF REVIEW OF IDEAS ON ENDEMISM
Naturalists and botanists have recognized the existence of rare or endemic
plants for centuries. Cain (12) ascribes the origin of the word endemic as it is
applied to the distribution of organisms to A. de Candolle (16). The great
voyages of discovery from the seventeenth through the nineteenth centuries
brought to light countless rare and endemic taxa. Though Linnaeus's Species
Plantarum of 1753 (53) lists no rarities for North America, his South African
listing of Proteaceae contains local endemics in Protea and Leucadendron. In
North America, Torrey & Gray (104) record several rarities; notable among
them are Franklinia aLtamaha (their Gordonia pubescens) and Silene polypetaLa
(their S. baldwinii), both from Georgia and one of which is now extinct, the
other still rare. However, instances where botanists have addressed the questions
of how rare taxa arise, exist, and go extinct are seldom found in the vast
number of tabulations of rare and endemic taxa.
Adolph Engler (23) appears to be the originator of the dichotomy of old vs
new endemics, which has been used extensively by other plant geographers
Annu.Rev.Ecol.Syst.1985.16:447-479.Downloadedfromwww.annualreviews.org
AccessprovidedbyMasarykUniversityon12/04/19.Forpersonaluseonly.
ENDEMISM IN HIGHER PLANTS 449
ever since (e.g. 94, 97, 101, 1 12, 1 15). Willis (112) quantified the idea of the
youthful endemic with his I-shaped, or "hollow" curves; they became the
backbone of his controversial and largely discredited theory of age and area.
The taxon cycle (78, 79, 113) is a modemversion ofthe idea that insular species
have an evolutionary progression ofstages analogous to the development of an
organism (1 1). The new formulation is as unfalsifiable as earlier concepts.
Stebbins (94) provided a genetical explanation for the epibiotic or relictual
endemic. Such taxa, he argued, havedepleted their store ofgenetic variation, a
process called biotype depletion, and thus they are unable to expand their
ranges. In 1980 Stebbins relied on better genetic evidence to modify the
hypothesis of biotype depletion (96).
Stanley Cain's now classic work on plant geography ( 12) set forth the then
prevailing notions on the kinds and origins of endemic plant species. His three
dicta on endemics merit repeating: (a) Endemism includes two types of plants
that are confined to single regions--endemics, sensustricto, which are relatively
youthful species, and epibiotics, which are relatively old relict species. (b)
"Youthful endemics may or may not have attained their complete areas by
having migrated to their natural barriers. Epibiotics may, but frequently do not,
contain the biotype richness that will allow or has allowed them an expansion of
area, following their historical contraction of area." (c) "A high degree of
endemism is usually correlated with age and isolation of an area, and with the
diversification of its habitats, as these factors influence both evolution (the
formation of new endemics) and survival (the production of relic endemics)"
(12, p. 212).
Stebbins & Major (97) recast Cain's two categories as paleoendemism and
neoendemism. These authors point to persistent defects in the new vs old
endemic dichotomy. First, often little basis exists on which tojudge whether an
endemic taxon is new or old. On this point, Cain (12, p. 227) has the last word:
". , , the relicnature ofan endemic should never be accepted without some form
of positive evidence." Second, the mode of origin of either type of endemic
taxon remains obscure.
Stebbins & Major based their classification upon the way in which narrow
endemics have achieved their restricted distributions, since this varies among
species; this system was also proposed by Favarger & Contandriopouplos (24).
Stebbins & Major's system incorporates the age of the endemic, its systematic
position, and cytological data (chromosome number and ploidal level), In
groups of related species, diploids are older than derivative low polyploids,
while both high polyploids and diploids are paleoendemics. The several kinds
of endemics and their attributes are arrayed in Table 1. Endemics with more
than one disjunct population are most likely relictual (paleoendemics), while
endemics confined to a single popUlation can be either paleoendemics or
neoendemics,
Annu.Rev.Ecol.Syst.1985.16:447-479.Downloadedfromwww.annualreviews.org
AccessprovidedbyMasarykUniversityon12/04/19.Forpersonaluseonly.
�
Table 1 A typology of narrow endemics based on three characteristics: systematic position, ploidallevel, and age (adapted from 97)
Isolated Closely related
Age Diploid Low polyploid High polyploid Diploid Low polyploid High polyploid
Ancient Schizoondemic Paleoendemic Schizoendemic
Patroendernic
Recent Schizoendemic Apoendemic Apoendemic
Annu.Rev.Ecol.Syst.1985.16:447-479.Downloadedfromwww.annualreviews.org
AccessprovidedbyMasarykUniversityon12/04/19.Forpersonaluseonly.
ENDEMISM IN HIGHER PLANTS 45 1
Stebbins & Major (97) use the ploidal level and its modes of origin both to
categorize endemics and to explain their origins. Paleoendemics are ancient
vestiges of taxa that were once more widespread. Their present relictual status
is presumably the result of the increasing constriction of their specialized
habitats over time. The neoendemics, on the other hand, are recent in origin,
have just split off from a parental entity, and may be poised for a further
expansion of their ranges and gene pools. Plantago cordata (6 1) and Stephanomeria
malheurensis (31) demonstrate that both paleoendemics and neoendemics
indeed occur, but we currently have no way to evaluate their relative
proportion in floras. Between the two extremes in age ofendemics, there are, of
course, endemics of intermediate age; they remain narrow endemics, confined
to a restricted, local habitat.
Recent reviews agree that there are multiple causes of rarity and endemism
(20,21,73,96). Neither genetics, ecology, nor history alone will suffice to
explain the origin of endemic taxa. Moreover, the interplay among various
causal factor� will vary in intensity, depending on the particular taxon under
scrutiny. Stebbins (96) proposes the gene pool/niche interaction theory to
explain origin or rarity and endemism. His notion is grounded on the assumption
of mUltiple causation:
"According to this theory . the primary cause of localized or endemic distribution patterns is
adaptation to a combination of ecological factors that are themselves localized. Factors of
soil texture or chemical composition arc the most common but by no means the only ones. .. .
Next to climatic and edaphic factors, those inherent in the gene pool of the population are of
critical importance.They include the total amount of variability , the amount of variability
that can be released at any one time, and the amount of variation that can be generated with
respect to those particular characteristics that affect most strongly the establishment of new
populations" (96, pp. 82-83).
Stebbins (96) uses an abstraction to envision the interaction of these factors:
". . . the niche (is seen) as a depression that is partly filled with a liquid, the gene
pool. Variations in the size and steepness of the sides of the niche and the depth
to which the liquid gene pool occupies the niche provide the variables that can
lead to rarity" (96, p. 83). A deep pool in a niche with a precipitous rim would
represent a species that is highly specialized, i.e. endemic.
THE GEOGRAPHY OF ENDEMISM
The narrow or local endemic is the one that best fits the colloquial notion of
rarity. However, the term endemism, in its classical biogeographic usage, does
not necessarily imply rarity or even small range. Thus, continental or regional
endemics need not be, and in fact seldom are, rare. Douglas fir (Pseudotsuga
Annu.Rev.Ecol.Syst.1985.16:447-479.Downloadedfromwww.annualreviews.org
AccessprovidedbyMasarykUniversityon12/04/19.Forpersonaluseonly.
452 KRUCKEBERG & RABINOWITZ
menziesii), canyon live oak (Quercus chrysolepis), and Scots pine (Pinus
sylvestris) are endemic to their respective continents or extensive regions, but
they are hardly rare. Sequoia sempervirens is acommon conifer endemic to the
coastal forests of California; Saponaria ocymoides is widespread though
endemic to the European Alps. But our intuitive perception ofrarity is at a much
smaller scale of magnitude. A rare species occupies only a small fraction of a
regional floristic or geographic province. Streptanthus niger. an endemic
crucifer, occurs on only one ofthe many serpentine outcrops in California. It is
a recognized rarity, one that Mason (59) would call a narrow endemic.
Cosmopolitan species constitute avery small proportion ofthe world's flora.
Wulff (115) goes so far as to say that there are no true cosmopolites among
flowering plants. A few more than 200 species have ranges that occupy each a
quarter or more ofthe earth'ssurface (0.00 1% ofthe flowering plants) ( 1 15) . A
vast majority of the world's seed-plant flora are endemic on some scale, from
regional to local. Good (28) states thatendemism, in the sense ofrestriction to a
floristic province, accounts for more than 90% of the world's plant species.
Further, endemism manifests itself at various taxonomic levels from variety to
higher category. Many of the smaller angiosperm families are endemic [sensu
Good (28)] and are found in the tropics and Australasia.
Three primary factors-geographic area, ecological breadth, and isolation
(13)--describe the distribution of endemics. Endemics are found on all landmasses
ofthe world, both continents and islands, and in all major biomes. More
curious is the well-known fact, first identified by Charles Darwin, that the
quantity and quality of endemism differ among the major geographic, topographic,
and vegetation types. For instance, while species number is smaller for
islands than for areas ofcomparable size on continents, the islands have higher
proportions of endemics ( 13, 28). Most oceanic (i.e. isolated) islands are far
richer in endemics than islands near continents. And by virtue of the small size
of most oceanic islands, their endemics would perforce be narrowly distributed-i.e.
true rarities.
Endemic plants also are distributed unevenly across the land areas of the
world. Some places, like mountains and islands, are rich in endemics, while
boreal and arctic regions are relatively poor in them. Many parts of the world
are well-known centers of endemism: California, the European Alps, the
Mediterranean region, alpine regions of central Africa, New Caledonia,
Hawaii, the Cape region of South Africa, and the Sino-Himalayan region.
Nonetheless, examples ofnarrow endemics abound in nearly all floras. A small
sampling includes Shortia galacifolia and Pedicularisfurbishiae in the eastern
United States; Fremontodendron decumbens and Carpenteria californica in the
western United States; and Silene diclinis and Daphne rodriguezii in the
Mediterranean region of Europe.
Annu.Rev.Ecol.Syst.1985.16:447-479.Downloadedfromwww.annualreviews.org
AccessprovidedbyMasarykUniversityon12/04/19.Forpersonaluseonly.
ENDEMISM IN HIGHER PLANTS 453
TAXONOMIC BIAS AND ENDEMISM
Rare plants hold a special fascination for taxonomists. The quest for local
endemics, marginal populations, and kindred rarities is a major concern of
taxonomists studying regional floristics, once they have described the common
taxa. Indeed, the taxonomist has made the greatest initial contribution to
cataloging rare, endangered, or threatened plants, thus providing the basis for
their conservation.
A binomial thatappears in print is not necessarily an absolute, eternal verity.
To paraphrase an old adage, "one person's species is another person's variety,"
or one may simply demote the binomial to conspecific status. So in the
taxonomy of rareplants, thedilemma of evaluating taxonomicjudgment is ever
present. We like to think that the concept behind any legally named taxon is a
hypothesis testable to some degree. The name Berberis nervosa implies the
existence of a system of populations in nature having particular attributes. Yet
most binomials (or trinomials) that appear in the taxonomic literature represent
little more than opinions that are subject to counteropinions and to taxonomic
inflation or deflation by yet another expert.
The very naming of an cxceptional organism or group of organisms confers
upon it a sense of uniqueness and discreteness. But a subsequent name change
or its relegation to synonomy may submerge a local rarity to obscurity. Rarity
surely is not a mere artifact of taxonomicjudgment. For well-worked temperate
areas of the world, rare taxa at the generic, specific, and infraspecific levelse.g.
Darlingtonia, Pedicularis furbishiae, and Pinus contorta subsp. bolanderi-are
often unequivocal concepts.
The potential subjectivity of taxonomic status of rarities may express itself in
ambiguities at various levels of the taxonomic hierarchy. The compilation of
global endemism at the family or subfamily level is a case in point. What may
be a small or evenmonotypic family to one taxonomist is simply a member of a
larger family to another expert. Two monotypic families endemic to North
America, the Koberliniaceae and Canotiaceae (as used in 28), are just as
justifiably placed in a larger family, Celastraceae, which is more cosmopolitan
(48). Examples at the genus level are known to all taxonomists--e.g. Benthamidia
into Cornus, Roxira into Madia, Vulpia into Festuca. At the species
level, differences in taxonomicjudgment can alter the apparentstatus from rare
to abundant, or the reverse.
Fluctuations in taxonomic status are recorded in the many catalogs of rare,
threatened, or endangered plants that have been compiled in recent years. For
examples, Holodiscus discolor var. delnortensis, a rosaceous shrub, was synonymized
as ordinary H. discolor; andIris tenax var. gormanii was demoted to
aforma of I. tenax (87).
Annu.Rev.Ecol.Syst.1985.16:447-479.Downloadedfromwww.annualreviews.org
AccessprovidedbyMasarykUniversityon12/04/19.Forpersonaluseonly.
454 KRUCKEBERG & RABINOWITZ
Since taxonomists have been the chief recorders of rarity, their familiarity
with taxonomic groups containing endemics should shed light on the origin and
diversification of endemic angiosperms. Narrowly construed, the work of the
taxonomist is simply to name and classify groups of organisms separated by
recognizable discontinuities. But, ofcourse, taxonomists usually do more with
their materials. Seeking answers to "Why?" and "How come?" questions, they
embark on excursions in ecology, biogeography, and evolution. Systematics is
the broad field that attempts to synthesize data from diverse disciplines in order
to interpret relationships. Broadly speaking, then, systematic studies ofendemics
should be able to address questions oftheir origins and radiations, that is, of
their evolutionary history and phylogeny (77). Monographs that have utilized
endemics and their attributes in accounting for relationships include those on
Epilobium (74), Clarkia (52), Lasthenia (66), Nigella (99), and numerous
others.
Insufficient field records and changes in taxonomic status owing to
monographic revision are other sources of taxonomic error in studies of endemism
(77). On the basis ofonly a single or a few collections, local endemics are
identified in lightly explored areas, especially in the tropics; further exploration
often discloses that the distribution of the putative local rarity has been underestimated.
Of course, as Richardson (77) points out, further field work in a
floristically rich but little known region can also increase the number ofknown
narrow endemics. Richardson also gives examples ofreductions in the numbers
of endemics following monographic revision ofa group--for instance, Canarium
(Burseraceae), earlier thought to consist of45 species that were all endemic-is
now considered to have only 9 species, 4 of which are endemic.
Some kinds of endemism permanently escape taxonomic recognition. Disjunct
or outlier/marginal populations, aberrant individuals, physiological
races, ecotypes, and even sibling species may be rare in nature and possess
unique genetic attributes. Harper (37) stresses this overlooked aspect of rarity
and cites the example of Agrostis tenuis, with its remarkable heavy metaltolerantraces;
thesevariants do notreceivetaxonomicrecognition, and thus, he
claims, may not be included on lists of rare species.
Native plants, rare to a localor regional flora, receive the primary attention of
the botanist. Less frequently, the botanists' perception of endemism embraces
taxa that are introduced from other floras. Lists of endangered plants, for
example, simply do not include localized introduced plants. This sentiment is
supportable iftherare adventiveis notrare elsewhere, either inits native land or
as an adventive in some other place. When stripped of our bias against weeds,
however, some introduced plants can be considered genuine rarities, either in
their native state or in their adopted environments. Klamath weed, Hypericum
perforatum, once so common a weed in parts of the western United States, has
Annu.Rev.Ecol.Syst.1985.16:447-479.Downloadedfromwww.annualreviews.org
AccessprovidedbyMasarykUniversityon12/04/19.Forpersonaluseonly.
ENDEMISM IN HIGHER PLANTS 455
become less common locally as a result ofbiological control, although its range
probably has not altered greatly. Should it be on an endangered list too?
THE EVOLUTION AND GENETICS OF ENDEMICS
Rarity enters the evolutionary domain in two distinct ways. The first is the
scientists' preoccupation with how narrow endemics come into being. The
second concerns the fate, i.e. the evolutionary potential and future, of the local
rarity.
Origins of Endemism
Opportunism, an ever-present prospect for living systems, plays a major role in
the genesis, perpetuation, and extinction of rare taxa. Given time, space, and
environmental heterogeneities, discrete biological capacities will develop opportunistically.
Mangroves have arisen independently in 23 families (14), and
many have small ranges despite water dispersal, e.g. Pelliciera rhizophorae.
Genetic potential, tempered by developmental and phenotypic constraints, is
the only arbiter on the part of the organism. We see no special, novel mechanism
that is needcd to generatc endemic taxa. Narrow endemics are the
product of speciation; in their origins and continued rarity they differ only in
degree from commoner species.
Two series of studies provide us with models of the evolution of narrow
endemism. A major achievement of field biologists is the observation in nature,
several times, of the origins of a geographically restricted taxon. The first series
deals with the evolution of heavy-metal tolerance by plants on polluted soils,
paralleling the pathway to endemism associated with environmental, especially
edaphic, peculiarities (9). The second series focuses on the origin of Stephanomeria
malheurensis from S. exigua ssp. coronaria and illustrates the likely
developmental path for species that occupy a fraction of a larger potential range
(31).
Very localized adaptation to soil toxicity is induced by a variety of substances---e.g.
copper, zinc, lead, and nickel. Grasses such as Anthoxanthum
odoratum andAgrostis stolonifera can evolve a tolerance to heavy metals in less
than a decade. The work of Bradshaw, McNeilly, Antonovics, and their
colleagues, summarized in a recent review (9), has demonstrated that tolerance
is a direct selective consequence of the presence of these metals and has high
heritability. Populations in nontoxic environments may contain tolerant
genotypes preadaptively, so natural variation in the capacity to withstand the
metals is present in natural popUlations. Thc local differentiation of tolerant
genotypes can be maintained despite considerable gene flow from neighboring,
Annu.Rev.Ecol.Syst.1985.16:447-479.Downloadedfromwww.annualreviews.org
AccessprovidedbyMasarykUniversityon12/04/19.Forpersonaluseonly.
456 KRUCKEBERG & RABINOWITZ
nontolerant populations; limited reproductive isolation by differentiation in
flowering times has evolved.
While the selective regime on toxic mine tailings might seem to be an
extreme case, we have no reason to suspect that the evolution of serpentine
endemics, for instance, is any different from the pathway to tolerance of toxic
soil shown by widespread colonists, such as Anthoxanthum odoratum. The
studies on heavy-metal tolerance have illuminated evolutionary mechanisms
and provided a base of information that workers on edaphic endemism can use
for comparison.
Less direct observation is relevant for studying the evolution of endemics that
occupy a small portion of a potential range. Some fraction of these endemics
must arise rapidly; these derived species would reside within or adjacent to the
range of their progenitors. Two annual composites, Stephanomeria exigua ssp.
coronaria and its presumed derivative S. malheurensis, are examples of this
(31). S. e. ssp. coronaria has an enormous range and occupies numerous
habitats in California and neighboring states. S. malhr;urensisoccurs on a single
hilltop of 150 acres in Oregon, at the northern edge of S. exigua ssp. coronaria's
range. In any year, there are probably fewer than 750 individuals of S.
malheurensis in existence.
A variety of genetic and morphological indicators verify the progenitorderivative
relationship. Although the two species resemble one another very
closely, key reproductive traits differ. The achene weight is doubled in the
derivative, and its seed number is lower. Self-pollination is the rule in the
derivative, but the progenitor is an obligate outcrosser. Several chromosomal
differences, including a reciprocal translocation, occur in the derivative. S.
malheurensis probably experienced a genetic bottleneck and evolved "after a
rapid and abrupt series of events initiated by the occurrence of a mutation at the
self-incompatibility locus" (31). Presumably, a self-pollinating individual of S.
exigua ssp. coronaria arose, and this led to the development ofan inbreeding
lineage and to chromosomal rearrangements that produced reproductive isolation.
Although the information on other documented cases is not as complete,
this evolutionary route to neoendemism is undoubtedly representative of other
cases.
Genetic Variation in Endemics
Are endemics genetically depauperate? Stebbins (94) and others have previously
argued that the answer is yes. We need to distinguish a variety of
possible causes of lowered genetic variation in endemics. Reduction in genetic
variability, or in polymorphism, may be the result of reduced heterozygosity,
decrease in mean number of alleles per locus, or reduction in the proportion of
polymorphic loci. One explanation for a reduction in polymorphism is that
lowered heterozygosity is a responsc to selection, for both neoendemic and
Annu.Rev.Ecol.Syst.1985.16:447-479.Downloadedfromwww.annualreviews.org
AccessprovidedbyMasarykUniversityon12/04/19.Forpersonaluseonly.
ENDEMISM IN HIGHER PLANTS 457
paleoendemic species. Precise adaptation to narrow ecological conditions-the
niche-variation hypothesis (106)-is the most commonly cited route, but there
are several possibilities.
Factors other than selection, however, affect heterozygosity. Under selectively
neutral circumstances, heterozygosity is influenced by effective population
size. Endemics may have small effective population sizes for three reasons.
The simplest is that endemics may have smaller total populations than widespread
species. In this case, the lowered heterozygosity may be equilibrial and
is likely to be permanent. Another explanation is our previous scenario (the
Stephanomeria case) that implies a genetic bottleneck as the origin of
neoendemism. Such a bottleneck associated with rapid speciation might produce
a current low level of genetic variation, but this condition could be
transitory. With expansion after a bottleneck, heterozygosity would rise and
equilibrate as a result of mutation. A third possible avenue to neocndemism is
inbreeding. Selfingreduceseffective populationsizeandhenceheterozygosity.
Endemics might be seifers because their progenitors as founders benefited
from self-compatibility (43). At a later stagein divergence, seifers more readily
become reproductively isolated from their parent popUlation. The degree of
selfing also may vary within a restricted population. Wyatt (116) found that
peripheral populations were more self-compatible than central populations in
granite outcrop species ofArenaria. Some endemics may be seifers because of
their mode of origin, not because of their range restriction per se.
We can turn this last point around and ask a related but separate question: Can
lowered genetic variation be the cause of restricted range? A lack of ecological
flexibility may preclude range expansion in a neoendemic or it may cause range
contractionin a paleoendemic. Thus, the answer is aconditional "yes" for some
cases, but we have no reliable information by which to judge.
Several examples will illustrate levels of genic variation in endemics.
Stephanomeria malheurensis shows reduced electrophoretic variation in comparison
with its progenitor, S. exigua ssp. coronaria (3 1). The alleles in S.
malheurensis are a subset of those in S. exigua ssp. coronaria. In the widespread
progenitor, 60% of the 25 loci were polymorphic with 45 alleles, but
only 12% of the loci with 7 alleles were polymorphic in the endemic derivative.
Here, the endemic's lower variability may be a product of a genetic bottleneck
associated with the speciation event and/or a result of selfing in the derivative in
contrast to the progenitor's outcrossing (30).
Lupinus subcarnosus and L. texensis provide similar, but not as extreme,
results. The former is an edaphic endemic on sandy soils in east-central Texas;
the latter occurs on a variety of soils over a larger range (2). For 6 loci, the
average number of alleles per locus was 1.84 for the endemic and 3.12 for the
more widespread species. Again, the endemic lupine is probably more inbred
than the widespreadlupine. We cannot associatelowered geneticvariationwith
Annu.Rev.Ecol.Syst.1985.16:447-479.Downloadedfromwww.annualreviews.org
AccessprovidedbyMasarykUniversityon12/04/19.Forpersonaluseonly.
458 KRUCKEBERG & RABINOWITZ
endemism per se, however, because we cannot discern whether these endemics
are any more genetically depauperate than selfing, widespread nonendemics.
Conifers ofrestricted range may also have reduced genetic diversity. Ledig
& Conkle (49) found no heterozygosity or polymorphisms in the two populations
of the highly local :rorrey pine (Pinus torreyana). The same workers
report (personal communication) that another narrow endemic conifer, Montereycypress
(Cupressus macrocarpa) is quite variable, as are the less restricted
California conifers, P. sabiniana and P. coulteri.
Oenothera organensis also is allozymieally depauperate, being polymorphic
at only 1 of the 15 loci examined; that single locus displayed only2 alleles (50).
In contrast to the two previous cases, this narrow endemic of the Organ
Mountains in New Mexico has an extensive self-incompatibility system and
functions as a panmictic unit over the estimated 5000 individuals in existence.
Still other endemics-e.g. Oenothera argillicola (51) and Clarkia Ungulata
(29)-exhibit moderate genic variation (see 50 for a recent review). All seven
species of the California endemic grass genus, Orcuttia, had as much genetic
variability as more widely distributed members ofthe Gramineae (33). Because
such electrophoretic evidence is equivocal, Stebbins (96, p. 80) concluded that
"there appears to be no recognizable correlation, either positive or negative,
between the amount ofgenetic variation within popUlations ofplant species and
the rarity or commonness ofthe species as a whole." See recent reviews on the
genetic structure of plant populations (35, 54).
Genetic systems that promote inbreeding oroutbreeding often are associated
with particular reproductive strategies and life histories (95). Do rare taxa show
a particular bias towards one kind of reproductive biology and life history?
Though a number of the case histories discussed in this review examine this
question for the particular taxon, we know of only one attempt at synthesis of
relevant data (38). It appears from this compilation that "a syndrome of
life-form and reproductive characteristics separates a rare and common species."
Using lists of rare and endemic taxa of Utah, California, and Colorado,
K. T.,Harper found woody plants to be underrepresented and herbs to be in
excess ofexpectation (38). Flowers ofrare speciesaremoreoften bilateral than
radially symmetrical. From this, he infers that most rare taxa that are bilateral
also are outcrossing and depend on healthy insect popUlations for pollination.
Any factors that diminish effective pollination would adversely affect the
rarity.
THE DYNAMICS OF ENDEMICS
Shifts from Common to Rare and the Reverse
We endorse the notion that rarity need not be permanent, as Harper (37)
'" reminds us. Taxa once common become rare, and local endemics can become
widespread in time. The fossil record abounds with instances of formerly wide
Annu.Rev.Ecol.Syst.1985.16:447-479.Downloadedfromwww.annualreviews.org
AccessprovidedbyMasarykUniversityon12/04/19.Forpersonaluseonly.
ENDEMISM IN HIGHER PLANTS 459
distributions that have become markedly constricted over time. The three
sequoias are good examples. Sequoia sempervirens, now restricted to coastal
California, was once found throughout western North America, from Alaska to
southern California and the Rocky Mountains. The highly restricted big tree,
Sequoiadendron giganteum, now in isolated groves in California's Sierra
Nevada, was once more extensive in the California Floristic Province and
eastward to Nevada (73, 115). Metasequoia glyptostroboides, now found only
in a very local relictual area in central China, once grew in both the Old and the
New Worlds (25).
The packrat midden record also illustrates this shift from a common to a
restricted distribution. Spaulding et al (93) found that the bristlecone pine
(Pinus longaeva) and the Texas pinyon pine (P. remota) were widespread in the
Great Basin andnorthern Chihauhuandeserts during the lateWisconsin glacial
period; their present ranges have been drastically reduced. The authors further
suggest that the reverse trend, from rare to common, has been the fate of the
creosote bush (Larrea divaricata) and the ponderosa pine (P. ponderosa).
A. S. Watt (109) provides the best documented example on a very local scale
of the increase and decrease in the abundance of several species. When he
began to observe a single 6 x 6 m plot of East Anglian breckland in 1 936,
Hieracium pilosella was absent and then quite rare after an initial invasion. In
1 947, it began to spread; from 1950 to 1958, it was the dominant species in the
plot. The incidence of H. pilosella then declinedvery rapidly after 1 963, and it
remained rare for the next decade, after which observations ceased.
Rarity followed by commonness is a pattern that is implicit in our notions of
speciation, yet difficult to detect in the fossil record. If we subscribe to the
orthodoxy of monophyletic origins, every new taxon must start out as a local
endemic. Yet because of their very scarcity, the beginnings of new species are
seldom recovered, let alone recognized, from fossil beds.
This sequence-from rare to common-is not documented often, either
because it happens infrequently or is difficult to observe. Harper (37) cites a few
examples, some of which must be considered in the special category of
introductions to otherfloras, i.e. of initiallyrareadventives whosenumbersand
ranges subsequently increased dramatically. No doubt many weedy taxa
throughoutthe world began as tenuous local introductions, which then spread in
varying degrees. Newly arisen alloploid taxa--e.g. Spartina townsendii and
Tragopogon mirus-similarly start as rarities and become more widespread
over time. But what about rare native taxa? Are there records of their numbers
increasing to the point that they are no longer considered rare? Drury (21 ) gives
the example of Epigaea repens. trailing arbutus, once considered rare in New
England; it became "an abundant weed after a fire in the oak-pine woods of
southern New England" (21, p. 2 1).
There is good reason to predict tnat some rarities will remain rare indefinitely.
Insular endemics like the Haleakala silversword (Argyroxyphium
Annu.Rev.Ecol.Syst.1985.16:447-479.Downloadedfromwww.annualreviews.org
AccessprovidedbyMasarykUniversityon12/04/19.Forpersonaluseonly.
460 KRUCKEBERG & RABINOWITZ
macrocephala) of Hawaii or Lobelia elgonensis ofthe Afro-alpine are unlikely
. to increase or decrease their ranges so long as their local habitats are maintained.
But one must acknowledge that evenstatus quo rarities may fortuitously
break out of their limited ranges. Accidental dispersal to another compatible
habitat, increase in the area of a specialized habitat, or a genetic revolution
accompanying change in selection within the endemic may trigger the expansion
of a local range. Another possible fate for an endemic is a variant of the
stasis condition. We can envision a rare taxon moving gradually, all the while
closely tracking an environment that itselfis moving spatially. At no time then,
would the endemic appreciably change in popUlation size or range, only in
location.
Extinction: Are Endemics Precarious?
Are small populations in greater hazard ofgoing locally extinct than large ones?
Several ideas and some evidence suggest that the answer is "yes." A limited
range means that a single disturbance can carry away an endemic. A team of
woodcutters could easily remove the entire habitat of Shortia galacifolia, and
dammed lakes can be larger than the range of the organisms they might
drown--e.g. Pedicularisfurbishiae. Further, we think that taxa on their way to
extinction for whatever reason must pass through a period of contraction.
Therefore, somewhere between high levels ofabundance and extinction, small
populations rnust appear. The notion of demographic stochasticity is relevant
here (60, 62). Small populations, if they are as responsive to environmental
variation as large ones, are more likely to hit the zero point and go locally
extinct. This small population process is analogous to genetic drift.
Observing contemporary natural extinctions is indeed difficult (26), and it is
unclear whether extinctions caused by hurnans-for instance, Hawaiian
Malvaceae-function like natural ones. Simberloff (89) provides the only
relevant experimental data. According to his study, when tiny mangrove
islands were cut in half, the smaller insectpopulations that resulted were more
likely to go locally extinct. Extrapolating from these studies to endemic plants
is clearly risky.
Other Forms of Rarity Useful for Understanding Endemism
Having concluded that there is no one kind of rarity, we think it is useful to
organize the attributes that generate different kinds of rarity into the form of a
matrix (72; Table 2). The cells yield an array of rarity in seven forms, as the
outcome of local population size (large vs small), geographic range (large vs
srnall), and habitat specificity (wide vs narrow). Of the kinds of rarity that
emerge from the matrix, the categories of sparse species are particularly
interesting. A taxon that is subordinate in the community, common nowhere,
and yet widespread and with rather wide habitat tolerances, could be viewed as
Annu.Rev.Ecol.Syst.1985.16:447-479.Downloadedfromwww.annualreviews.org
AccessprovidedbyMasarykUniversityon12/04/19.Forpersonaluseonly.
Table 2 A typology of rare species based on three characteristics: geographic range, habitat selection, and local population size (from 72)
Large, dominant in some
places
Small, nondominant
Large and wide
Locally abundant over a
large range in several
habitats
Constantly sparse over a
l a r g e r a n g e a n d in
several habitats
Geographic range and habitat specificity
Large and narrow
Locally abundant over a
large range in specific
habitat
Constantly sparse in a
specific habitat but over
a large range
Small and wide
Locally abundant in several
habitats but restricted
geographically
Constantly sparse and
geographically restricted
in several habitats
Small and narrow
Locally abundant in a �
specific habitat but reiilstricted
geographically ::
Constantly sparse and til
geographically restrict-
::
ed in a specific habitat Z
::c
Ci
::c
tTl
;:0
tE
:>
z
;A
oj:::.
0\
.....
Annu.Rev.Ecol.Syst.1985.16:447-479.Downloadedfromwww.annualreviews.org
AccessprovidedbyMasarykUniversityon12/04/19.Forpersonaluseonly.
462 KRUCKEBERG & RABINOWITZ
rare by local naturalists. Yet, in fact, sparsi,ty
mism!
Because the narrow endemic species is only one manifestation of rarity, we
would like to know whether information on other small populations is instructive
about plants with small ranges. Harper (37) contends that rarity
depends on the scale of observations: What is rare in a hectare of a county may
not be rare over a larger region. Furthermore, this scale factor may have a
political or chauvinistic flavor. Pointing to the rare occurrence of Fritillaria
camschatcensis in Washington State, for instance, ignores its increasing commonness
to the north, as well as in eastern Asia. Harper, writing as a population
biologist, also notes that in examining rarity a taxonomic bias leads the scientist
to overlook manifest or hidden genic diversity within species, some forms of
which can be exceedingly rare.
Two other sorts of small populations are especially relevant for an understanding
of endemism: populations on the margin of a large range (91) and
isolated disjuncts distant from a larger central range (34). The genetic structure
of marginal populations has revealed how divergence and isolation arise.
Ecological processes in smalldisjunct populations should be similar to those for
narrow endemics, but little is known about either group. A notable exception is
Donald Pigott's work on Tilia platyphyllos (67). Pigott pieces together climatic,
ethnographic, edaphic, biogeographic, systematic, and physiological information
on widespread European populations to infer the native status and
historic pathway to extreme isolation of "perhaps the rarest native species in
northernEurope" (67, p. 305). Thus, isolated disjuncts provide amodel, not yet
used, for understanding endemics,because in some cases theyprovide replicate
populations for experimentation; the consequences of manipulating them are
not as hazardous as those stemming from tampering with a single remaining
population.
THE ENVIRONMENTAL DETERMINANTS OF
NARROW ENDEMISM
While numerous factors--e.g. history, area, and isolation-limit distribution,
the ultimate arbiter of the success in occupying space of any organism, whether
endemic or not, is its inherited tolerance to environmental factors. Organisms
do not occur where they cannot, but often they do not occur where they might.
Thus a major question about endemics is: Are they restricted to the places in
which they reside because they cannot exist beyond these bounds, or could they
occur over large areas if they were brought there? Are endemics ecologically or
physiologically narrow or fastidious? Are the restricted ranges of endemics a
reflection' of their 'small nic,hes?
through climate, geology (both topography and soils), or the presence of other
Annu.Rev.Ecol.Syst.1985.16:447-479.Downloadedfromwww.annualreviews.org
AccessprovidedbyMasarykUniversityon12/04/19.Forpersonaluseonly.
ENDEMISM IN HIGHER PLANTS 463
organisms; when environments are narrow in breadth, compatible organisms
may be similarly restricted.
Climate
When microclimatic diversity is superimposed on a regional climate, the
diversity of organisms can escalate. Differences in microclimates arrayed from
forest floors to upper canopies in the tropics promote microhabitat variation that
is exploited by distinct biotas. Rock outcrops in deserts create local climates, as
do waterfalls, streamsides, and lakes in forest habitats. When local climates are
highly discontinuous or uniquely set apart from prevailing climates, some
plants will be endemic to the site. In the tropics, epiphytic orchids, aroids,
gesneriads, and bromeliads can be microclimate specialists; vernal pool species
(in Downingia, Navarretia, and Pogogyne) and terrestrial orchids found in
bogs are good microclimate specialists in temperate floras.
Geological Determinants ofRarity
Geological phenomena relevant to rarity include events associated with the
plate tectonics that create mountains and raft continents, with the genesis of
discrete lithologies, with processes of land formation, and with the chemical
and physical constitution of rocks. Soil formation, especially the genesis of
unique soil types, is but one geological setting for the evolution of discrete
forms of plant life.
When the products of geological processes are displayed discontinuously, as
discrete land forms or as chemically and physically distinct substrates, they
may provide the necessary isolation for the genesis of unique biotas. Island
endemics are spectacular outcomes of geological events that are arrayed discontinuously
over time and space. On continents, topographic, lithological,
and pedological discontinuities achieve the ecological isolation essential to
species diversity. Isolated mountain ranges on continents often harbor local
endemics, owing to the geological events of discrete orogenies--e.g. isolated
batholiths, emergence of volcanoes, and separated cordilleras. The Cascade
Range of western North America provides some good examples, especially on
its isolated volcanic peaks. Some taxa are endemic to a single volcano (e.g.
Pedicularis rainierensis on Mount Rainier), while other endemics have less
narrow ranges and occur on two or more of the isolated volcanoes (e.g. Arnica
viscosa on Mount Shasta, the Three Sisters, Mount Thielson, and Mount
Mazama). Discontinuous distributions similar to those on isolated volcanic
peaks are known for the remarkable tree lobeliads of the alpine zone of central
African peaks--e.g. Lobelia elgonensis on Mount Elgon vs Lobelia gibberoa
on several Afro-alpine peaks (28).
Edaphic factors-i.e. chemical, physical, and biological properties of
soils-are the phenomena most commonly used to draw the link between
Annu.Rev.Ecol.Syst.1985.16:447-479.Downloadedfromwww.annualreviews.org
AccessprovidedbyMasarykUniversityon12/04/19.Forpersonaluseonly.
464 KRUCKEBERG & RABINOWITZ
environments and endemic taxa. Endemics that are associated with unusual
substrates like gypsum, serpentine, limestone, alkaline, and heavy metal soils
are well known to field botanists in many parts of the world. Mason (58, 59)
was one of the first to articulate the close tie between unusual soils and narrow
endemics. He described instances of narrow edaphic endemism in order to
illustrate the interplay between genetically determined tolerances and edaphic
influences.
Plants endemic to soils derived from serpentinite and other ferromagnesian
rocks are found in many parts of the world. Two areas in the tropics are noted
for their high incidences of endemics to serpentine: The Great Dyke of Zimbabwe
has at least 20 species restricted to serpentine (III), while in New
Caledonia2 monotypic families, more than 30 genera and 900 species (60% of
the island's flora), are restricted to serpentine outcrops (42). Examples from
these two classic sites are Dicoma niccolifera, Pearsonia metallifera, and
Lotononis serpentinicola from Zimbabwe; and Geissois pruinosa, Sebertia
acuminata, and Xylosma serpentinum from New Caledonia. Several of these
taxa are known to be accumulators of heavy metals in unusually high amounts
(hyperaccumulators with > 1000 ppm).
Serpentine endemism in temperate areas is exemplified by the monotypic
borage, Halacsya sendteri of Yugoslavia, and by the several species of the
cruciferous genus Streptanthus (section Euclisia) of California. When the
serpentine outcrop is highly local (i.e. "insular"), narrow endemism can be
expected. Examples of the Californian local serpentine endemic include Streptathus
niger of Tiburon Peninsula, S. batrachopus of Mount Tamalpais, and
Layia discoidea at New Idria (45).
Denton's ( 1 7) study of Sedum (section Gormania) illustrateswellthe edaphic
influence on narrow endemism. All taxa grow on rock outcrops in California
and Oregon; some are restricted to ultramafic rocks-i.e. serpentinite and
peridotite. Denton judged the most highly restricted entities to be relictual,
based on their low (diploid) ploidal level, reproduction (self-compatibility and
reduced cloning ability), and restriction to the most severe substrates. She
challenges the notion that the highest concentrations of endemic taxa "are found
in regions that are floristically rich and diverse and where extreme environmental
gradients exist" (p. 1 51). While this relationship may hold for the broad
region of the Klamath-Siskiyou mountain area, she found that it does not for
local rock outcrop habitats: "The narrowest endemics of section Gormania are
not found on richer or more diverse outcrops than the more widely distributed
taxa" (p. 1 5 1 ) .
Two spectacular cases of edaphic restriction in California are in Streptanthus
(Cruciferae) andLinum (Linaceae). Of the 16 taxa in sectionEuclisia (Streptanthus),
14 are serpentine species, and 10 are narrow endemics (45, 46). Sharsmith's
(86) monograph of Hesperolinon (sometimes treated as a section of
Annu.Rev.Ecol.Syst.1985.16:447-479.Downloadedfromwww.annualreviews.org
AccessprovidedbyMasarykUniversityon12/04/19.Forpersonaluseonly.
ENDEMISM IN HIGHER PLANTS 465
Linum) reveals that 8 of its 12 taxa are obligate or near-obligate serpentine
endemics, some of which are quite local.
In a very real sense, the multitude of discontinuities created by geological
processes is perhaps the ultimate cause of local rarity and narrow endemism.
Given a regional climate, the fractionation ofthe landscape within that climatic
zone by physical and chemical irregularities provides a host ofdiscrete habitats.
No doubt some landscape-making geological processes are the root cause of
local to microclimate variations. Hence when microclimates foster a unique
flora, the ultimate cause must be considered geological. In the absence of
geomorphological processes that operate discontinuously in time and space,
there would be no irregularity in land forms to produce local climates, and the
world's landscapes would be pretty boring too. Humans also create geochemical
discontinuities through toxic mine tailings that have provided the settings
where local adaptation has been most clearly illuminated (9).
The edaphic specialist leads us to the distinction between two sorts ofnarrow
endemics. Some are closely associated with some distinctive environmental
feature that is usually edaphic--e.g. serpentine species and mangroves. In the
absence ofthe feature--e.g. saline soils-the endemic is absent. Other endemics
have ranges that simply occupy a tiny fraction of an extensive habitat. In
many cases, the unoccupied portion ofthe habitat seems perfectly appropriate,
and the absence of the species is particularly puzzling. Some species of
Astragalus show this (4).
Plantago cordata and Zizania texana, both stream-dwelling endemics, provide
good examples for comparison. Within an enormous range (from the
Hudson River to northern Florida to Missouri to Ontario), Plantago cordata
occurs only in scattered, isolated popUlations and is clearly a paleoendemic
plant. Three dozen sites have been reported historically, and in some areas
(Michigan and Virginia) the species now appears to be extinct. Meagher et al
(61) showed that the heart-leaved plantain has considerable physiological
plasticity and is capable of surviving under a variety of conditions. The sites
where P. cordata occurs do not have any features that conspicuously distinguish
them from nearby sites where the plant might grow but does not. These
authors point to demographic features such as a low reproductive output and
poor dispersal as the causes for the rarity.
In contrast, Zizania texana has only a single population on the San Marcos
River in Texas, 640 km from the nearest congener (Z. aquatica) in Louisiana,
and it is probably a neoendemic (103). The site is distinctive for its water
chemistry and its equitable environmental conditions; a spring from a limestone
source produces neutral to alkaline water with high flow rates and creates a
thermal regime varying only 5° C annually. Thus the habitat of this extreme
endemic is itselfrare, and we do not need to invoke additional factors to explain
why Z. texana lives nowhere else.
Annu.Rev.Ecol.Syst.1985.16:447-479.Downloadedfromwww.annualreviews.org
AccessprovidedbyMasarykUniversityon12/04/19.Forpersonaluseonly.
466 KRUCKEBERG & RABINOWITZ
Rarity Resulting from Organism Interactions
RARITY AND COEVOLUTION Synergisms and antagonisms of all sorts between
species may result in the genesis of a rare member of any given
interactive syndrome. In the following interactions, one of the coevolved pair
of organisms might be rare: host-pathogen, two symbionts, two commensals,
herbivore-herb, pollinator-plant, host-epiphyte, host-parasite (both flowering
plants), and others. At least one member of a synergistic pairis most likely to be
rare when the relationship is new. A case in point is the putativehybrid species,
Penstemon spectabilis. whose origin Straw (98) postulated as follows: The FI
hybrid between P. centranthifolius and P. grinnellii was transformed into a
species only when an insect vector different from either pollinator of the parent
species mediated homogamous matingsof the hybrid entity. At an early stage of
this speciational event, the new taxon must have been rare.
Another situation where rarity is promoted and even maintained by biotic
interaction could well be the host plant-pathogen syndrome. Several years ago,
Gillett (27) proposed that pest pressure by bacteria and fungi as well as by
invertebrates like insects and roundworms can account for "the apparently
pointless multiplicity of species in areas where it has had time to operate" (p.
40). Gillett's paper, written well before the current boom in coevolutionary
studies, provided a theoretical, and largely untested, basis for explaining the
high incidence of both speciation and rarityin some parts ofthe world-e.g. the
tropics, South Africa. Whenever a pest or pathogen finds a suitable host, the
host may escape by evolving an ability to resist or avoid it, this in time may
warrant its recognition as a distinct taxon. In such situations, not only is the
initial divergent population rare, but it may never survive pest pressure to
become common.
The same model may be invoked for other adverse interactions, such as
herbivory and competitive, or even allelopathic, inhibition. Though current
evidence is still meager, we see this as a potential explanation of initial or
ongoing endemism that arises out of different kinds of coevolutionary rela­
tionships.
SYNECOLOGICAL FACTORS PROMOTING RARITY Endemic taxa are rarely if
ever found as isolated individuals or as monocultures in nature. The rare, local
endemic coexists with other organisms. Sometimes the community can be as
unique as the rare taxon itself-e.g. Lobelia elgonensis is restricted to the
distinctive alpine flora of central Africa; Arctostaphylos myrtifolia occurs with
other acid heath flora only on azonal soils derived from Eocene laterites and
acid sericitic schist at lone, California; Becium homblei is an endemic of
serpentine communities on the ultramafic Great Dyke in Zimbabwe. In other
Annu.Rev.Ecol.Syst.1985.16:447-479.Downloadedfromwww.annualreviews.org
AccessprovidedbyMasarykUniversityon12/04/19.Forpersonaluseonly.
ENDEMISM IN HIGHER PLANTS 467
instances, an endemic is simply a unique member of a common or widespread
community type. Poapachypholis occurs only at llwaco, Washington, but it is
found in thecommongrass-forb-seacliff community ofthenorthPacific Coast.
Tauschia stricklandii grows only at one place in Mount Rainier National Park
and at one other site in Washington, but it is a member of the widespread
mountain meadow community of the Pacific Northwest.
Considering plants as members of biotic communities poses special questions.
Do rare taxa tend to occur more often in unusual or unique communities,
rather than in common ones? Does a system of interacting organisms influence
the population size and dynamics of rare taxa? How is plant succession related
to rarity?
Trulynarrow endemics (i.e. extremely local rarities) are mostoftenmembers
of distinctive communities, or at least of singular habitats. This assertion is
impossible to verify on a global basis, but some statistics from regional floras
do provide corroboration. We have used inventory catalogs of rare and endangeredplants
in some states of the western United States to generate the data
in Table 3. Since the areal extent of unique habitats is at least an order of
magnitude less than the area of common habitats, these species' counts reflect
an underlying asymmetry in the distribution of endemics. Of course, while
unique habitats delimit endemics, the small size of an endemic's range per se
may also limit the number of available habitats.
The process of compiling ecological life histories has favored common
species. For years, the Journal ofEcology has published detailed accounts of
the attributes of selected taxa in the British flora. Often included in the species
portraits are synccological data: community status, competitive ability, interactions
with other organisms, etc. Rarely has the series included narrow
endemics. Occasionally, an intensive study of a rare taxon takes on some of the
features of an ecological life history; examples include Lobelia Rattingeri (6),
Table 3 Occurrence of endemicsa in unique or common habitats, based on data from inventories of rare
and endangered taxa.
Location or Taxon in
Inventory (Reference) unique habitat
Number %
California (90) 560 49
Oregon (87) 139 49
Washington (108) 65 47
Yukon Territory ( 1 8) 5 38
aDisjuncts ur peripherab umitted.
Taxon in
common habitat
Number %
388 34
I I I 39
74 53
8 62
Uncertain
Number
1 85
34
0
0
%
1 6
12
0
0
Total
(Number)
1 133
284
1 39
13
Annu.Rev.Ecol.Syst.1985.16:447-479.Downloadedfromwww.annualreviews.org
AccessprovidedbyMasarykUniversityon12/04/19.Forpersonaluseonly.
468 KRUCKEBERG & RABINOWITZ
Pedicularis !urbishiae (56), Silene diclinis (69), and Ranunculus ophioglossoides
(19).
Pedicularisfurbishiae, consisting of around 600 known individual plants,
occupies a precarious habitat along the St. John River in northern Maine. It
prefers a calcareous substrate situated in the transient river terraces between the
water's edge and the adjacent spruce forests. Its constant association with Alnus
crispa suggests a host-parasite relationship, which has not yet been confirmed.
None of the associated plant species is narrowly restricted to the Pedicularis
furbishiae localities. Moreover, the transient river terrace habitat is more
widespread than the lousewort populations. Neither unique pollinators nor
highly local habitat conditions explain its narrow range, and the cause of the
restriction is not yet known. Silene diciinis. equally rare in southeastern Spain
(ca. 600 individuals), occupies a similarly transient habitat--old cultivated
orchards of olive and carob on limestone (69). Yet there is nothing to suggest
that the habitat is unique either in terms ofits associated flora and fauna or in its
physical features. It has been suggested, without direct evidence, that both
rarities are glacial relicts.
These two examples suggest that there is a common denominator in the
synecology ofcertain rarities. Their occupation oftransient or unstable habitats
marks them as successional taxa, whose existence could be obliterated when
vegetation "matures" (34). As long as the successional stage persists, the
rarity'S habitat is preserved. Of course, this explanation does not address the
issue ofhow the narrow endemic originated; it simply accounts for its perpetua­
tion.
Quite another view ofthe successional position ofrare plants is espoused by
Mahler (57). His brief note merits quoting in toto: "Species occurring in the
lower seral stages of plant succession are not apt to become extinct. However,
endemic species ofrare occurrence in the climax stage ofthe site . . . are the taxa
most susceptible to reduction in population numbers and with catastrophic
events or man's activity are apt to become extinct." Mahler (personal communication)
believes rare plants of specialized climax habitats are more likely
to go extinct if the habitats are narrowly limited in extent.
Rarities also may occupy extensive, stable, climax-forest situations. Their
presentrange restriction may be dueeitherto the pasthistory ofthe areaor to the
destruction of their specialized habitat within the climax forest. The welldocumented
case of the semiaquatic Plantago cordata shows how change in a
habitat, in this case produced by human activity, can reduce the population size
of the entity (61 ) . This highly specialized plantain has not been able to adapt to
the shifting man-made habitats now available to it.
The rich species diversity so often found in tropical forests is enmeshed with
notions ofrarity. Atlowlandsites, thenumberoftree species can be remarkably
large,--e.g. 502 woody species in a2000 m2 area in a central Amazonian forest
Annu.Rev.Ecol.Syst.1985.16:447-479.Downloadedfromwww.annualreviews.org
AccessprovidedbyMasarykUniversityon12/04/19.Forpersonaluseonly.
ENDEMISM IN HIGHER PLANTS 469
(Klinge cited in 47). Furthermore, in any such sample, a small number of
individuals often is taken as a sign that the taxon is rare. Representation of a
species may be less than one individual per hectare (47). Kubitzki (47) takes
issue with the assumption that low population densities in the tropics signify
rarity: "Since the tropical lowland forests often encompass huge areas, a low
population density does not necessarily imply rarity in the sense of a low global
abundancy." Hubbell (41) has shown that the shapes of dominance-diversity
curves are similar in tropical andtemperatezones. This result indicates, not that
the proportion ofrare species in tropical forests is higher, but simply that there
are more species.
The Habitat Attributes ofRare Taxa
Conservationists who want to protect an endangered endemic species usually
emphasize the need to preserve its habitat. This idea is based on the presumption
that the vitality ofan endemic is dependent on its continued occupancy of a
convivial habitat. This notion has not been tested explicitly, though perhaps the
reintroduction of a rare taxon into its natural habitat may be construed as an
unintentional test. We must ask therefore: What attributes of an endemic's
habitat are crucial for its survival? To what degree can the habitat be modified
and still hORt the rarity?
One approach to the problem of how to characterize a habitat for an endemic
is to ask what it takes to provide the lebensraumfor a narrow endemic. The first
requirement surely must be knowledge of its tolerance ranges. What does it
require in the way of climate (i.e. temperature, light, and water), edaphic
conditions, and interactions with other organisms? Since all organisms need
some particular mix of these ingredients, the rare plant would seem to require
some unique qualitative and quantitative combination of these same elements.
In addition, the areal extent of the habitat where these unique requirements are
met is likely to be abruptly discontinuous with neighboring areas where the
same factors may be present in different proportions or with distinct absolute
values. In other words, the habitat of the rarity is expected to be sharply
discontinuous with neighboring habitats of contrasting attributes.
For endemics associated with particular habitats, it is surprising that little
evidence supports the view that they have narrowed tolerances or that,
equivalently, rare species have small niches. Arboreta demonstrate that trees,
both endemic and widespread, are capable of growing in habitats and conditions
well beyond those in which they naturally occur. Mangroves and other
halophytes do not require salt water (3), nor do serpentine plants require
magnesium ( 107). These plants clearly possess the capacity to tolerate conditions
that are toxic to, and thus exclude, other species. But the converse necd
not be true: Species with these unusual capacities are not intolerant of more
common environments. Mangroves can grow in the absence of a tidal influx,
Annu.Rev.Ecol.Syst.1985.16:447-479.Downloadedfromwww.annualreviews.org
AccessprovidedbyMasarykUniversityon12/04/19.Forpersonaluseonly.
470 KRUCKEBERG & RABINOWITZ
but in nature they simply do not. Why these endemics do not grow in more
ordinary habitats is a puzzle; their absence is usually attributed to competition
from other vegetation or to the high cost of maintaining their special abilities.
Understanding the exclusion of endemics from more common habitats is an
area that needs further investigation.
The size of the endemic's area, its ecogeographic isolation, its temporal
duration (steady state or seral), and its intrinsic attributes are essential features
of its generalized habitat and they determine the endemic's fitness within it. A
complete roster of information on a plant's habitat and inherent fitness is never
available for a rare taxon, and it is uncertain whether we would really understand
the endemic even if we had this information. But perhaps there is a
way around the impasse of our gaps in information. The bioassay method, used
so successfully inotherareas ofappliedbiology, could beutilizedas a surrogate
test of a species' capacities; the rare taxon itself becomes the bioassay unit. The
success or failure of introducing (or reintroducing) the rare plant both into its
preferred habitat and into a less convivial habitat, as a control, then measures
the fit between the organism and its habitat ( 10).
THE POPULATION BIOLOGY OF RARE TAXA
Plant demography over the last decade has developed into a sophisticated
discipline with a substantial body of data (36, 88). Although we are gaining
information on the life histories of endemics from numerous studies (e.g. see
5-7 forLobelia gattergeri, Leavenworthia exigua var. laciniata, and L. stylosa;
70, 71 for Plantago spp.), none of the demographic studies on endemics have
been as thorough or complete as Sarukhan' s classic works (80-82) onRanuncuIus.
Our current knowledge of the demography of endemics does not permit us
to answer questions such as whether survivorship, dispersal, pollination, seedling
establishment, or other properties differ systematically from plants with a
small range compared to those with a larger one.
Two studies on rarity, neither of whichconcerns an endemic, provide models
for future work on endemism, both because they yield interesting results and
because they illustrate divergent approaches. Wells ( 1 10) observed the development,
phenology, mortality, and reproduction of the widespread orchids
Spiranthes spiralis. Aceras anthropophorum. and Herminium monorchis in
Bedfordshire chalk grassland in England. He showed how changes in management,
such as brush clearing and grazing that occurred more than a decade prior
to the emergence of the orchids from their belowground mycorrhizome phase,
affected their population recruitment and current dynamics. Sophisticated
observational studies and field natural histories such as Wells' work are critically
important to achieving the conservation management of endemics.
How local population size is regulated for endemics and why we see many
Annu.Rev.Ecol.Syst.1985.16:447-479.Downloadedfromwww.annualreviews.org
AccessprovidedbyMasarykUniversityon12/04/19.Forpersonaluseonly.
ENDEMISM IN HIGHER PLANTS 471
plants in one m2 and none in the adjoining m2 are among the most puzzling
ecological phenomenaforall sorts ofplants. Carlina vulgarisis widespread and
often common. On Anglesey dunes in Wales, Greig-Smith & Sagar (32)
investigated patches where C. vulgaris had small populations in order to
explore the factors that restricted its local increase. Through a series of experiments
in removal, nutrient enrichment, seed sowing, and caging, they
inferred that competition from neighbors and low fertility were responsible for
local rarity. Sowing additional fruits didincrease population size, and predation
by small mammals on seed heads limited propagule production. Thus work on
plants that are not generally rare can tell us a great deal about dynamic processes
in small populations. Applied to endemics an experimental approach such as
Greig-Smith & Sagar's would provide much useful information.
SOME CASE HISTORIES
Recognition that some rare taxa are in danger of extinction has provoked
intensive studies of the biology of particular rarities. These efforts often are
stimulated by governmental orders that certain rare taxa be preserved in their
natural habitats. Ecological life-history profiles, with data on demographics,
habitat status and potential vitality ofthe plants, have been compiled for several
such taxa. Most such studies are not usually published in accessible journal
form but rather, appear as mimeographed status reports to the contracting
agency. We have had access to a number of these reports through the kindness
of various regional offices of the Rare and Endangered Species Program, US
Fish and Wildlife Service. We now examine some of those rarities for which
fairly complete status reports have been compiled.
1 . Astragalus phoenix Bameby-Ash Meadows milk vetch (Leguminosae).
This edaphic endemic closely matches the Mentzelia case (see below) in
its narrow restriction to flats, washes, and knolls of calcareous alkaline soils at
Ash Meadows, Nevada (75). Its present population is probably less than 600
individuals, mostly mature (three or more years old). Reproduction is sparse;
parent plants produce few pods, and these are entrapped by the parent plant.
Seed production, though low, is assured by effective self-pollination. The
major impediment to survival-let alone spread-ofA . phoenix is disturbance
ofthe substrate. Plants do not establish on sites where the distinctive soil crust is
broken.
2. Betula uber (Ashe) Fernald-Virginia Round-Leaf Birch (Betulaceae).
By 1980 a single population of 20 individuals was all that remained of this
highly publicized rarity of Smyth County, Virginia (84). Despite its near
extinction, implementation ofa recoveryplan ( 1982) may fostcrits survival. So
few individuals survive that research on the biology ofB. uber must be limited
to nondestructive sampling and to propagules in cultivation. From 1975 to
Annu.Rev.Ecol.Syst.1985.16:447-479.Downloadedfromwww.annualreviews.org
AccessprovidedbyMasarykUniversityon12/04/19.Forpersonaluseonly.
472 KRUCKEBERG & RABINOWITZ
1980, half of the remaining individuals (trees, saplings, and seedlings) had
been eliminated by natural and human causes. No new recruits have been
observed, and the remaining 20 plants are of reduced vigor. Suppression by the
overstory and repeated sampling by those studying the restricted endemic
appearto account for the reduced vigor. Reproductive output is presumed to be
extremely low. Deliberatesowings of seed collected in the wild yielded ca. 1 %
germination, and of 300 seedlings, only 3 were judged B. uber; the rest were
similar to the sympatric and widespread B. lenta.
There appears to be nothing exceptional in the habitat of B. uber to account
for its restriction. Its association on disturbed alluvial flats with the successional
B. lenta suggests that the habitat is transient. Moreover, its co-occurrence with
B. lenta and B. aUeghaniensis suggests a hybrid origin for the rarity.
Given its sympatry and resemblance to B. lenta (series Costatae, dark-barked
tree birches), is B . uber a good spccies? A morphometric analysis (85) supported
its specific status as well as its affinity to B . lenta. But even if it were
reduced to infraspecific status (or considered conspecific with the variable B.
lenta), the uniqueness of the population would persist. Harper (37) reminds us
that rare gene pools may escape taxonomic recognition and still be worthy of
preservation.
3. Hudsonia montana Nutt.-Mountain Golden Heather (Cistaceae). The
status report (63) on this rare North Carolina plant is a model of its kind, for
conservation purposes. It contains the best biological information available for
the taxon and provides us with a well-documented case history. The plant is
restricted to quartzite ledges of the Table Rock area, Burke County, North
Carolina. The habitat, unique and local within the southern Appalachian
Mountains, is described as "an open heath-bald vegetation type, with scattered
shrubs (mostly Erieaeeae), few herbs or forbs and much bare ground" (63 , p.
291). H. montana is restricted to the more open areas of the ledges by
competition (overtopping and shading) from other shrubs in the adjacent areas.
The four known populations in the Table Rock area consist of 100-200
individuals, from mature clones (4--8 dm in diameter) to juvenile plants and
seedlings. Mature clones are much more common than juveniles or seedlings;
nothing is known of the plants "stored" in the seed bank, though four-year-old
seed could be germinated. Reproduction is both sexual (selfing and outcrossing)
and vegetative (rooting at the edge of a growing clump). The clonal mode
predominates. Seed is dispersed locally (precinctively).
H. montana appears to be relictual, surviving in restricted sites where its
locally xeric habitat may persist. Barring changes to the surrounding terrain,
and with Federal protection, this remarkable rarity should persist indefinitely as
a narrow edaphic endemic.
4. Lupinus padre-crowleyi C. P. Smith-DeDecker's lupin syn. L. dedeckerae
(Fabaceae). This rare and distinct Mono Basin (California) lupin
Annu.Rev.Ecol.Syst.1985.16:447-479.Downloadedfromwww.annualreviews.org
AccessprovidedbyMasarykUniversityon12/04/19.Forpersonaluseonly.
ENDEMISM IN HIGHER PLANTS 473
appears to be a newly evolved entity (102). Its small but vigorous populations
are not confined to exceptional edaphic habitats and are associated with both
common as well as locally rare species of the region. Changes in late Pleistocene
climate following hybridization may have left the new entity stranded
following retreat of one of its parents. Height-class distribution for 8 subpopulations
reveals that the normal-to-Ieft-skewed cohorts contain young recruits.
Thus the total population size of the species (100,000± individuals)
appears to be large enough, with vigorous replacement, to avoid extinction.
The plastic breeding system (insect pollination plus self-compatibility) of this
wholly sexual species supplies sufficient seed for continuity, if seed-toestablished-plant
conversion is successful. Thus, barring natural or human
disturbance, L. padre-crowleyi should expand its range into nearby compatible
habitats.
5 . Mentzelia leucophylla Brandegee-Ash Meadows stick-leaf
(Loasaceae). This is one of several plant and animal taxa narrowly confined to
the local spring-fed "oasis" in the desert, Ash Meadows, Nye County, Nevada
(76). "Ash Meadows is an extensive lowland plains area of the Amargosa
Valley (around 75 sq. mi.), which is the more or less dissected remnants of a
large Pleistocene playa that extended well into California. Hence the soils are
extremely fine textured (silts and clays) and have slow internal drainage, and
the water table is near the surface in much of the area . . . The soils are light
colored and have a high salt content; many are heavily salt encrusted at the
surface" (8). The present population ofthe plant is estimated to be less than 100
individuals, occupying an area less than one mile square. M. leucophylla is
presumed to be a biennial, though individuals may flower the first year;
nonbolting rosettes also are present. Reproduction is presumed to be by seed
from outcrossed individuals. The taxon may be genuinely evanescentdoomed
to extinction aft�r its short life span. It is thought to have arisen afterthe
pluvial lake of 10,000 years BP disappeared. M. leucophylla is a polyploid (n =
18), exceptional in its group for having a base number of 9, rather than 1 1, the
usual base number among its congeners. It is very sensitive to disturbance and
also is subject to decimation in drought years. Should the climate change,
survival of the species would be further impaired. M. leucophylla is thus a
relictual, highly local, edaphic endemic, whose survival is made precarious by
continuing climatic deterioration, aggravated by human disturbance.
6. Shortia galacifolia T. & G.-Oconee Bells (Diapensiaceae). This
clump-forming herbaceous perennial, a plant ofgreat ornamental horticultural
value, is endemic to small portions of the southern Appalachian Mountains of
the eastern United States. It is restricted to rhododendron thickets or to mixed
(cove) hardwood stands, often in close proximity to watercourses. The status
report (1) is based on 1979 field studies of20 historic sites. The report lists three
major habitat types (Angiosperm forest, mixed forest and shrub), which can be
Annu.Rev.Ecol.Syst.1985.16:447-479.Downloadedfromwww.annualreviews.org
AccessprovidedbyMasarykUniversityon12/04/19.Forpersonaluseonly.
474 KRUCKEBERG & RABINOWITZ
found in a wide range of topographies (mountains, hills, and ridges; scarps,
bluffs, cliffs, and escarpments; valleys, gorges, and channels; plains and flats;
ravines) and on a wide variety of metamorphic substrates. The species successional
status "ranges from mid/transient to late/climax stages." Eleven
localities were censused, using the clump or colony (from a few cm2 to many m2
in size) as the population unit. At 7 localities, the sample consisted of all the
population units, 2-1 25 in number; some localities showed a range of cover
classes but 6 localities had none (i.e. they were all of one class?). Population
areas for the 1 1 localities ranged in size from 1 m2 to 1 730 m2, with 5 localities
of less than 10 m2 in area. While the large size of some clumps are the result of
clonal increase via underground stems, some sexual reproduction does occur.
In some sites very young seedlings can be found on late-stage, decaying logs.
Shortia galacijolia has lost territory largely due to human interventions-sites
inundated by dams, disturbed by logging and .collecting for garden use, and
other habitat intrusions. The only natural disturbance making inroads on population
size is recurrent stream bank slippage. With sufficient habitat protection,
Oconee Bells should be able to maintain itself in the discrete sites in
perpetuity.
7 . Silene polypetala (Walt.) Fernald & Schubert-Fringed campion
(Caryophyllaceae). Ever since its discovery in the late eighteenth century ( 15),
this distinctive silene appears to have continued as a rare, local endemic.
Presently it is confined to two disjunct regions of the southcastcrn United
States: west centralGeorgiaand the southwestern Georgia-Florida border area.
The habitat for the more northern localities (Talbot and Crow counties, Georgia)
is "rich deciduous woods on river banks and on hardwood bottom lands
near the Flint River" ( 15) . The more southerly populations are in deep
calcareous ravines along Lake Seminole and the Apalachicola River. Both
localities support mixed communities of hardwood tree species, with a shrub
understory. Population sizes range from 25 to 250 individuals, and one population
consists of several hundred individuals. All populations were deemed
vigorous, when seen in flower. No seedlings were observed; individuals in
some populations are propagated by the layering of decumbent stems. S.
polypetala is rare presumably because of its close trackingof a unique habitatforested
ravines and hardwood bottoms bordering certain rivers in the southern
United States. Though other eastern North American silenes are known from
woodland habitats (S. virginica, S. ovata, and S. rotundifolia), the very distinct
S. polypetala appears to be the most mesic taxon of them all. Its origin and
relationships are obscure; with two other eastern North American taxa, it is
linked with species in western United States-So laciniata and S. parishii (40).
Crosses of S. polypetala with several eastern North American silenes yielded
vigorous but wholly sterile F' hybrids (44).
Annu.Rev.Ecol.Syst.1985.16:447-479.Downloadedfromwww.annualreviews.org
AccessprovidedbyMasarykUniversityon12/04/19.Forpersonaluseonly.
ENDEMISM IN HIGHER PLANTS 475
RARITY AND CONSERVATION
Tl).e conservation of narrow endemics that are threatened or endangered has
become a major concern shared by governments, conservation organizations,
and individuals. An axiom for preservation is "Know thy organism. " An
oft-repeatedjustification for preserving rare populations is the argument for the
conservation of gene pools-i.e. the unique genetic resources ofrare organisms
may be useful in breeding programs. Two recent works (26, 83) address the
relationship of genetics and evolution to conservation. At symposia on rare
plant conservation, detailed protocols on the information-gathering process are
provided that could yield valuable biological information (64, 68 , 100). Papers
by Henefin et al (39), Morse (63), and others (64) provide procedures for
studying the biological attributes of rare taxa. Morse's (63) account ofHudsonia
montana is a good example of a biological life history of a rare taxon.
Because we are convinced that biology as a whole will be enriched by the study
of rare organisms, we encourage the biological community to undertake more
extensive research on the genetic, demographic, reproductive, and habitat
characteristics of rare plant populations. More comparative studies to contrast
the biologies ofrare taxa with those of related common ones would be particularly
valuable.
ACKNOWLEDGMENTS
We thank Montgomery Slatkin, Jody Rapp, Martin Lechowicz, Bill DiMichele,
and Jeff Karron for discussions and comments on the manuscript.
This work was supported by grants from the National Science Foundation
(DEB-821 1500, BSR-8406l 26) .
Literature Cited
I . Anon. n.d. [Status report on Shortia
galacifolia Torrey & Gray]
2. Babbel, G. R . , Selander, R. K. 1 974.
Genetic variability in cdaphically restricted
and widespread plant species.
Evolution 28:61 9-30
3. Barbour, M. G. 1 970. Is any angiosperm
an obligate halophyte? Am. Midi. Nat.
84: 105-20
4. Bameby, R. C. 1 964. Atlas of North
American Astragalus. Mem. NY Bot.
Card. 13: 1-1 1 88
5. Baskin, J. M . , Baskin, C. C. 1 978. The
seed bank in a population of an endemic
plant species and its ecological significance.
BioI. Conserv. 1 4: 1 25-30
6. Baskin, J. M . , Baskin, C. C. 1 979. The
ecological life cycle of the cedar glade
endemic Lobelia gattingeri. Bull. Torrey
Bot. Club 1 06: 1 76-81
7. Baskin, J. M . , Baskin, C. C. 1 98 1 . Geographic
distribution and notes on the
ecology of the rare endemic, Leavenworthia
exigua var. laciniata. Castanea
46:243-47
8. Beatley, J. C. 1 976. Vascular plants of
the Nevada Test Site and centralsouthern
Nevada: Ecologic and geographic
distributions. Div. Biomed. and
Environ. Res. US Energy Res. Devel.
Admin . , TID-2688 1 . Tech. Info. Ctr.
Office of Tcch. Info . , Springfield, Va.
9. Bradshaw, A. D., McNeilly , T. 1 98 1 .
Evolution and Pollution. London: Ar­
nold
10. Brookes, B. S. 198 1 . The discovery, extermination,
translocation and eventual
survival of Schoenusferrugineus in Britain.
See Ref. 100, pp. 421-28
I I . Cain, S. A. 1 940. Some observations on
Annu.Rev.Ecol.Syst.1985.16:447-479.Downloadedfromwww.annualreviews.org
AccessprovidedbyMasarykUniversityon12/04/19.Forpersonaluseonly.
476 KRUCKEBERG & RABINOWITZ
the concept of species senescence. Ecology
21 :213-15
1 2. Cain, S. A. 1 944. Foundations of Plant
Geography. New York: Harper
13. Cariquist, S. 1974. Island Biology. New
York: Columbia Univ. Press
14. Chapman, V. J. 1 976. Mangrove Vegetation.
Vaduz, Liechtenstein: Cramer
1 5 . Columbus College. n.d. Endangered and
. Threatened Plant Status Surveys: Region
IV. Silenepolypetala (Walt. ) Fernald and
Schubert. Columbus College, Ga. , Contract
No. 14-16-0004-79-102
1 6. de Candolle, A. 1 855. Geographie Botanique
Raisonnee. Vols. 1 , 2. Paris: Mas­
son
17. Denton, M. F. 1979. Factors contributing
to evolutionary divergence and
endemism in Sedum section Gormania
(Crassulaceae). Taxon 28: 149-55
18. Douglas, G. W . , Argus, G. W. , Dickson,
H. L., Brunton, D. G. 198 1 . The
rare vascular plants of the Yukon. Syllogeus
28: [-61
19. Dring, M. J., Frost, L. C. [97 1 . Studies
of Ranunculus ophioglossifolius in relation
to its conservation at the Badgeworth
Nature Reserve, G[oucestershire, England.
Bioi. Conserv. 4:48-56
20. Drury, W. H. 1 974. Rare species. Bioi.
Conserv. 6:1 62-69
2 1 . Drury, W. H. 1 980. Rare species of
plants. Rhodora 82(829):3-48
22. Ehrlich, P., Ehrlich, A. 198 1 . Extinction.
The Causes and Consequences of
the Disappearance of Species. New
York: Random House
23. Engler, A. 1 882. Versuch einer Entwicklungsgeschichte
der Pflanzenwelt. Leipzig:
Engelmann
24. Favarger, C . , Contandriopoup[os, J.
1961 . Essai sur ['endemisme. Bull. Soc.
Bot. Suisse 7 1 :384-408
25. Florin, R. 1952. On Metasequoia. living
and fossil. Bot. Not. I: 1-30
26. Frankel, O. H . , Soule, M. E. 1 98 1 . Conservation
and Evolution. New York:
Cambridge Univ. Press
27. G illett, 1. B. 1962. Pest pressure, an underestimated
factor in evolution. In Taxonomy
and Geography. Syst. Assoc.
Publ. 4:37-46, London
-
28. Good, R . 1974. The Geography of
Flowering Plants. London: Longman.
4th ed.
29. Gottlieb, L. D. 1974. Genetic confirmation
of the origin of Clarkia Iingulata.
Evolution 28:244-50
30. Gott[ieb, L. D. 1 977. Electrophoretic
evidence and plant systematics. Ann.
Mo. Bot. Gard. 64: [61-80
3 1 . Gott[ieb, L. D. 1 979. The origin of phenotype
in a recently evolved species. In
Topics in Plant Population Biology. ed.
O. T. Solbrig, S. Jain, G. B. Johnson, P.
H. Raven, pp. 264-86. New York:
Columbia Univ. Press
32. Greig-Smith. J . , Sagar. G. R. 198 1 .
Biologica[ causes of local rarity in Carlina
vulgaris. See Ref. 100, pp. 389-400
33. Griggs, F. T., Jain, S. K. 1983. Conservation
of vernal pool plants in Ca[ifornia.
II. Population biology of a rare and
unique grass genus Orcuttia. Bioi. Conservo
27: 1 7 1-193
34. Griggs. R. F. 1940. The ecology of rare
plants. Bull. Torrey Bot. Club 67:575-94
35. Hamrick, J. L. 1 983. The distribution of
genetic variation within and among natural
plant populations. In Genetics and
Conservation. ed. C. M. SchonewaldCox,
S. M. Chambers, B. MacBryde, L.
Thomas, pp. 335-48. Menlo Park, Calif:
Benjamin/Cummings
36. Harper, J. L. 1 977. Population Biology
of Plants. London: Academic
37. Harper, J. L. [98 1 . The meanings of rarity.
See Ref. [ 00. pp. 1 89-203
38. Harper, K. T. 1979. Some reproductive
and life history characteristics of rare
plants and implications of management.
In The Endangered Species: A Symposium.
Great Basin Naturalist Memoirs
ed. Provo, Utah: B. Young Univ. 3: 1 29-
37.
39. Henefin, M . S . , Morse, L. E . , Reveal, J.
L., MacBryde, B . , Lawyer, J. I. 1 98 1 .
Guidelines for the preparation of status
reports on rare or endangered plant species.
See Ref. 64, pp. 261-82
40. Hitchcock, C. L., Maguire, B. 1974. A
revision of the North American species of
Silene. Univ. Wash. Publ. Bioi. 13: 1-
73
4 1 . Hubbell, S. P. [979. Tree dispersion,
abundance, and diversity in a tropical dry
forest. Science 203: 1 299-[ 309
42. Jaffre, T. 1 98 1 . Etude ecologique du
peuplement vegetal des sols derives des
roches ultrabasiques en Nouvelle Caledonie.
Paris/Noumea:Office Rech. Sci.
Techno!. Outre Mere
43. Jain, S. K. 1 976. The evolution of inbreeding
in plants. Ann. Rev. Eco!. Syst.
7:469-95
44. Kruckeberg, A . R. 1964. Artificial
crosses involving eastern North American
Silenes. Brittonia 1 6:95-105
45. Kruckeberg. A. R. 1984. California serpentines:
Flora, vegetation, geology,
soils, and management problems. Univ.
Calif. Publ, Bot. 78: 1-180
46. Kruckeberg, A. R . , Morrison, J. L.
1 983. New Streptanthus taxa (Cru-
Annu.Rev.Ecol.Syst.1985.16:447-479.Downloadedfromwww.annualreviews.org
AccessprovidedbyMasarykUniversityon12/04/19.Forpersonaluseonly.
ENDEMISM IN HIGHER PLANTS 477
ciferae) from California. Madroiio 30:
230-44
47. Kubitzki, K. 1 977. The problem of rare
and of frequent species: the monographer's
view. See Ref. 68, pp. 331-36
48. Lawrence, G. H. M. 1 95 1 . Taxonomy of
Vascular Plants. New York: Macmillan
49. Ledig, F. T . , Conkle, M. T. 1 983. Gene
diversity and genetic structure in a narrow
endemic, Torrey pine (Pinus torreyana
Parry ex. Carr). Evolution 37:79-
85
50. Levin, D. A . , Ritter, K . , Ellstrand, N. C.
1 979. Protein polymorphism in the narrow
endemic Oenothera organensis.
Evolution 33:534-42
5 1 . Levy, M . , Levin, D. A. 1 975. Genic
heterozygosity and variation in permanent
translocation heterozygotes of
the Oenothera biennis complex. Genetics
79:493-5 1 2
52. Lewis, H . , Lewis, M . 1 955. The genus
Clarkia. Univ. Calif. Pub!. Bot. 20:241-
392
53. Linnaeus, C. 1 753. Species Piantarum,
Vols. 1 40, 1 42. Stockholm: Salvii. (Facsimile
ed. , Ray Society)
54. Loveless, M. D . , Hamrick, J. L. 1 984.
Ecological determinants of genetic structure
in plant populations. Ann. Rev. Ecol.
Syst. 15:65-96
55. Lucas, G . , Synge, H . 1 978. The IUCN
Plant Red Data Book. Morge�, Switzerland:
Int. Union Conserv. Nat.
56. Macior, L. W. 1 978. The pollination
ecology and endemic adaptation of Pedicularis
furbishiae S. Wats. Bull. Torrey
Bot. Club 1 05:268-77
57. Mahler, W. F. 1 983. The role of plant
succession in the extinction of plant species.
Sida 1 0 : 1 9 1
5 8 . Mason, H . L . 1 946. The edaphic factor in
narrow endemism. I. The nature of environmental
influences. MadrOllo 8:209-
26
59. Mason, H. L. 1 946. The edaphic factor in
narrow endemism. n. The geugraphic
occurrence of plants of highly restricted
patterns of distribution. Madroiio 8:241-
57
60. May, R. M. 1973. Stability and Complexity
in Model Ecosystems. Princeton,
NJ: Princeton Univ. Press. 2nd ed.
6 1 . Meagher, T. R . , Antonovics, J., Primack,
R. 1 978. Experimental ecological
genetics in Plantago. III. Genetic variation
and demography in relation to survival
ofPlantago cordata, a rare species.
Bioi. Conserv. 14:243-57
62. Mertz, D. B . , Dawthon, D. A . , Park, T.
1976. An experimental analysis of cumpetitive
indeterrninancy in Tribolium.
Proc. Nat!. Acad. Sci. USA 73: 1 368-
72
63. Morse, L. E. 1 98 1 . Report on the COnservation
status of Hudsonia montana, a
candidate endangered species. See Ref.
64, pp. 283-308
64. Morse, L. E. , Henefin, M. S . , eds. 1 98 1 .
Rare Plant Conservation: Geographical
Data Organization. Bronx: NY Bot.
Gard.
65 . Myers, N. 1 979. The Sinking Ark. New
York: Pergamon
66. Ornduff, R. 1966. A bio�ystematic survey
of the gold field genus Lasthenia.
Univ. Calif. Publ. Bot. 40: 1-92
67 . Pigott, C. D. 198 1 . The status, ecology
and conservation of Tilia platyphyllos in
Britain. See Ref. 1 00, pp. 305-1 7
68 . Prance, G. T. , Elias, T. S . , eds. 1 977.
Extinction is Forever, Threatened and
Endangered Species of Plants in the
Americas and Their Significance in
Ecosvstems Todav and in the Future.
Bron'x: NY Bot. Gard.
69. Prentice, H. C. 1 976. A study in endemi�m:
Silene diclinis. Bioi. Conserv.
10: 15-30
70. Primack, R. B. 1 979. Reproductive
effort in annual and perennial species of
Plantago (Plantaginaceae). Am. Nat.
1 14:5 1-62
7 1 . Primack, R. B . 1 980. Phenotypic variation
of rare and widespread species of
Plantago. Rhodora 82(829):87-95
72. Rabinowitz, D. 198 1 . Seven forms of
rarity. See Ref. 100, pp. 205-\ 7
7 3 . Raven, P . H . , Axelrod, D. I . 1 978. Origin
and relationships of the California
flora. VI/iv. Calif. Publ. Bot. 72: 1-1 34
74. Raven, P. H . , Raven, T. E. 1976. The
genus Epilobium in Australasia. NZ Uep.
Sci, Ind, Res. Bull, 216: 1-321
75. Reveal, J. L. n.d. Status report on Astragalus
phoenix Barneby (Ash Meadow
milk vetch). US Dep. Interior, Washington,
DC. (Mimeogr.)
76. Reveal, J. L. n .d. Status report on Melltzelia
leucophylla Brandegee (Ash
Meadows stick-leaf). US Dep. Inter. ,
Washington, DC. (Mimeogr.)
77. Richardson, I. B. K, 1978. Endemic taxa
and the taxonomist. In Essavs in Plant
Taxonomy, ed. H. E. Street, pp. 245-62.
London: Academic
78. Ricklefs, R. E" Cox, G. W. 1 972. Taxon
cycles in the West Indian avifauna. Am.
Nat. 1 06: 1 95-2 1 9
79. Ricklefs, R. E., Cox, O. W . 1 978.
Stages of taxon cycle, habitat distribution,
and population density in the avifauna
of the West Indies. Am. Nat.
1 1 2:875-95
Annu.Rev.Ecol.Syst.1985.16:447-479.Downloadedfromwww.annualreviews.org
AccessprovidedbyMasarykUniversityon12/04/19.Forpersonaluseonly.
478 KRUCKEBERG & RABINOWITZ
80. Sarukhan, J. 1974. Studies on plant demography:
Ranunculus repens L. , R.
bulbosus L., and R . acris L. II. Reproductive
strategies and seed population
dynamics. 1. Eml. 62: 1 5 1-77
8 1 . Sarukhan, J. , Gadgil, M. 1 974. Studies
on plantdemography: Ranunculus repens
L. , R. bulbosus L. , and R. acris L. III. A
mathematical model incorporating multiple
modes of reproduction. J. Ecof.
62:921-36
82. Sarukhan. J . , Harper, J. L. 1 973. Studies
on plant demography: Ranunculus repens
L, R. bulbosus L., and R. acris L. 1.
Population flux and survivorship. J.
Ecol. 61 :675-716
83 . Schoenewald-Cox, c . , Chambers, S .
M . , MacBryde, B . , Thomas, W. L . , eds.
1 983. Genetics and Conservation. Reading,
Mass: Addison-Wesley
84 . Sharik, T. L. 1982. Virginia round-leaf
birch recovery plan. Region V, US Fish
Wildlife Serv. , Washington, DC.
(Mimeogr.)
85. Sharik, T. L. , Ford, R. H. 1 984. Variation
and taxonomy of Betu/a uber, R.
lenta, and B. alleghaniensis. Brittonia
36:307-16
86. Sharsmith, H. K. 1 96 1 . The genus Hesperolinon
(Linaceae). Univ. Calif. Pub/.
Bot. 32:235-3 1 4
87 . Siddall , 1. L., Chambers, K. L., Wagner,
D. H. 1979. Rare, Threatened and
Endangered Vascular Plants in Oregon .
Oreg. Nat. Area Preser. Advis. Com.
Salem: Oreg. State Land Bd.
88. Silvertown, J. W. 1982. Introduction to
Plant Population Ecology. London:
Longman
89. Simberloff, D. S. 1974. Experimental
zoogeography of islands and effects of
island size. Ecology 57:629-48
90. Smith, 1. P. Jr. , Cole, R. 1 . , Sawyer, J.
O. Jr. (In collaboration with W. R.
Powell). 1980. Inventory a/Rare and Endangered
Vascular Plants of California.
Berkeley: Calif. Native Plant Soc. 2nd
ed.
9 1 . Soule, M. E. 1 973. The epistasis cycle: a
theory of marginal populations. Ann.
Rev. Ecol. Syst. 4:1 65-87
92. Soule, M. E., Wilcox, B. A . , eds. 1 980.
Conservation Biology: An EvolutionaryEcological
Perspective. Sunderland,
Mass: Sinauer
93. Spaulding, W. G . , Leopold. E. B . , VanDevander,
T. R. 1983. Late Wisconsin
paleoecology of the American Southwest.
In Late Pleistocene Environments
ofthe United States, ed. S . C. Porter, pp.
259-93. Minneapolis: Vniv. Minn. Press
94. Stebbins, G. L. 1942. The genetic approach
to problems of rare and endemic
species. Madrono 6:24 1-58
95 . Stebbins, G. L. 1 950. Variation and
Evolution in Plants. New York: Columbia
Univ. Press
96.\ Stebbins, G. L. 1 980. Rarity of plant
species: a synthetic viewpoint. Rhodora
82(829):77-86
97 . Stebbins, G. L . , Major, J. 1 965 . Endemism
and speciation in the California
flora. Eco!. Monogr. 35: 1-35
98. Straw, R. M. 1 955. Hybridization ,
homogamy and sympatric speciation.
Evolution 9:441-44
99. Strid, A. 1 970. Studies in the Aegean
flora XVI: Biosystematics of the Nigella
arvensis complex with special reference
to the problem of non-adaptive radiation.
Opera Bot. 28: 1-169
100. Synge, H . , ed. 1 98 1 . The Biological
Aspects of Rare Plant Conservation.
New York: Wiley
101 . Szymkiewicz, D. 1 933. Contributions a
la geographie des plantes. I. Phytogeographie
floristique et ecologique. Kosmos
[Stockholm] 58 :405-24
1 02 . Taylor, D. . ·W. 198 1 . Endangerment
status of Lupinus dedeckerae on the
Inyo National Forest, California. Mono
Basin Res. Group Contrib. No. 4, Inyo
Nat. For., US For. Serv. , Washington,
DC
103 . Terrell, E. E., Emery, W. H. P . , Beaty,
H. E. 1978. Observations on Zizania texana
(Texas wild rice), an endangered
species. Bull. Torrey Bot. Club 105:50-
57
1 04. Torrey, J . , Gray, A . 1838. A Flora of
North America. New York: Wiley &
Putnam
1 05 . Tryon, R. M . , ed. 1980. Rare and endangered
plant species in New England.
A symposium. Rhodora 82(829): 1-
237
106. Van Valen, L. 1965. Morphological variation
and width of ecological niche. Am.
Nat. 99:377-90
107 . Walker, R. B . o Walker, H. M . , Ashworth,
P. R. 1 955. Calcium-magnesium
nutrition with special reference to serpentine
soils. Plant Phvsiol. 30:214-21
108. Washington Natura'l Heritage Program.
1 98 1 . An ll/usrrated Guide to the Endangered,
Threatened and Sensitive
Vascular Plants of Washington. Olympia:
Wash. Dep. Nat. Resources
109. Watt, A. S. 198 1 . Further observations
on the effects of excluding rabbits from
grassland A in East Anglian breckland:
the pattern of change and factors affecting
it (1 936-1973). J. Ecof. 69:509-
36
Annu.Rev.Ecol.Syst.1985.16:447-479.Downloadedfromwww.annualreviews.org
AccessprovidedbyMasarykUniversityon12/04/19.Forpersonaluseonly.
ENDEMISM IN HIGHER PLANTS 479
l l O. Wells, T. C. E. 1 98 1 . Population ecology
of terrestrial orchids. See Ref. 100,
pp. 28 1-95
l l l . Wild, H . , Bradshaw, A. D. 1977. The
evolutionary effects of metalliferous and
other anomalous soils in south central
Africa. Evolution 3 1 :282-93
l l2. Willis, J. C. 1 922. Age andArea:A Study
in Geographical Distribution and Origin
of Species. London: Cambridge Univ.
Press
1 1 3 . Wilson, E. O. 196 1 . The nature of the
taxon cycle in the Melanesian ant fauna.
Am. Nat. 95: 1 69-93
1 1 4. Wood, S . L., ed. 1979. The endangered
species: a symposium. Great Basin Nat.
Mem. 3 : 1- 1 7 1
1 1 5. Wulff, E. V. 1 943. An Introduction to
Historical Plant Geography. Waltham,
Mass: Chronica Botanica
1 1 6. Wyatt, R. 1 984. The evolution of selfpollination
in granite outcrop species of
Arenaria (Caryophyllaceae) . I. Morphological
correlates. Evolution 38:804-16
Annu.Rev.Ecol.Syst.1985.16:447-479.Downloadedfromwww.annualreviews.org
AccessprovidedbyMasarykUniversityon12/04/19.Forpersonaluseonly.
Annu.Rev.Ecol.Syst.1985.16:447-479.Downloadedfromwww.annualreviews.org
AccessprovidedbyMasarykUniversityon12/04/19.Forpersonaluseonly.
Annu.Rev.Ecol.Syst.1985.16:447-479.Downloadedfromwww.annualreviews.org
AccessprovidedbyMasarykUniversityon12/04/19.Forpersonaluseonly.