| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
2 Department of Environmental Science and Policy, University of California Davis, Davis, California 95616 USA; 3 Department of Natural and Applied Sciences, University of Wisconsin-Green Bay, Green Bay, Wisconsin 54311 USA; and 4 Department of Botany and Genetics, University of Georgia, Athens, Georgia 30602 USA
Received for publication May 11, 1999. Accepted for publication September 24, 1999.
ABSTRACT
We used enzyme electrophoresis to evaluate genetic diversity in 32 populations of Calystegia collina, a clonal plant species endemic to serpentine outcrops in northern California (USA). Of 34 loci examined 56% were polymorphic, but on average only 17% were polymorphic within local populations. Neither the total number of alleles nor the number of multilocus genotypes differed significantly between populations in small vs. large serpentine outcrops. Genetic and geographic distances between populations were positively correlated, but this relationship was not significantly affected by the isolation of serpentine outcrops. Populations were highly differentiated (Fst = 0.417) and little genetic variation was explained by geographic region or serpentine outcrop.
Observed heterozygosity within populations almost always exceeded Hardy-Weinberg expectations. In many populations, all 30 sample ramets were uniformly heterozygous at one or more loci yet were genetically variable at other loci. These results imply that many C. collina populations originate from one or a few genetic founders, with little recruitment from seeds. Genetic variation within uniformly heterozygous populations must be the product of multiple, closely related founders or somatic mutations within the population. We conclude that vegetative reproduction, perhaps coupled with somatic mutation, helps maintain genetic diversity in these isolated but long-lived populations.
Key Words: Calystegia collina • clonal plant • Convolvulaceae • habitat fragmentation • genetic diversity • population structure • serpentine endemic • somatic mutation
Small populations of narrowly distributed species are expected to exhibit low levels of genetic variation within populations but high levels of genetic variation among populations because of genetic drift and restricted gene flow (Hamrick and Godt, 1989
; Karron, 1991
; Young, Boyle, and Brown, 1996
). Genetic impoverishment, together with demographic stochasticity, is expected to increase the risk of local extinction in these small populations (Hanski and Gilpin, 1991
). Yet many narrowly endemic plant species persist for long periods in small, naturally isolated populations (Ellstrand and Elam, 1993
; Holsinger, 1993
). Clonal reproduction is one mechanism that might counteract the extinction of small populations by preserving genetic variation and mitigating the effects of demographic stochasticity (Schaal and Leverich, 1996
; D'Amato, 1997
). Few studies, however, have examined genetic variation in small clonal plant populations (Harper, 1977
; Barrett and Kohn, 1991
; McLellan et al., 1997
), and even fewer have compared genetic variation among clonal populations at both local and landscape scales.
Serpentine soils in the North Coast Ranges of California provide an ideal environment for studying plant population genetics at multiple spatial scales. Serpentine outcrops occur in areas of tectonic activity, where ultramafic minerals high in iron and magnesium but low in calcium have intruded or displaced overlying sedimentary rocks (Kruckeberg, 1984
; Brooks, 1987
). Outcrops of serpentine in California's North Coast Ranges number in the thousands, ranging in area from <1 ha to >10 000 ha. Serpentine substrates support a large number of endemic plant species, many of which are globally rare or endangered (Skinner and Pavlik, 1994
). The naturally patchy distribution of serpentine outcrops is compounded for many endemic plant species by the patchiness of populations within outcrops. This multiscale pattern provides the setting for our study of local and landscape-level population genetics.
We investigate genetic diversity, population structure, and gene flow among populations of a diploid plant species, Calystegia collina (E. Greene) Brummitt (Convolvulaceae), endemic to serpentine soils in California's North Coast Range. Calystegia collina occurs on both small and large serpentine outcrops, where it exists in discrete patches, which we call local populations. These local populations consist of one or more genetic individuals with multiple ramets, connected to varying degrees by underground rhizomes. Populations in large and small outcrops do not differ significantly in either the number of ramets or the number of genotypes (Wolf, Harrison, and Hamrick, 2000
). Previous experimental studies (Wolf, 1998
) have demonstrated that C. collina is an obligate outcrosser. Highest rates of seed production occur in flowers that are fertilized with pollen from a different population. Nearby populations are more numerous on large serpentine outcrops, where natural rates of flower and seed production are significantly higher than on small outcrops.
We used protein electrophoresis to help address several questions about the genetic structure of C. collina populations: (1) What proportion of the species' genetic variation is represented within local populations? (2) Do populations on small serpentine outcrops contain as much genetic diversity as populations on large outcrops? (3) How are genetic relationships among populations affected by the geographic distances between populations? (4) Is the nonserpentine matrix separating small outcrops a more effective barrier to gene flow than is the serpentine habitat between populations in large outcrops? Answers to these questions can help explain how the geographic configuration of serpentine habitat affects the evolutionary history and population dynamics of C. collina.
Results from our investigation have implications for conservation strategies of serpentine habitats and may be applicable to other clonal plants that occur in small, local populations. We are particularly interested in the conservation value of small serpentine outcrops. If small outcrops support populations with little genetic variation and no novel alleles, then conservation efforts should be more effectively directed to large continuous serpentine habitats. On the other hand, if populations on small outcrops harbor a significant fraction of the species' genetic variation or contain novel alleles, then conservation plans should include both large and small serpentine outcrops.
Clonal reproduction enables local populations and individual genets to persist for a very long time. Under these circumstances, traditional models (e.g., Wright, 1931
) may not adequately describe the population genetics of C. collina. Another objective of our study was to identify the evolutionary forces that best explain patterns of genetic variation within and between local C. collina populations. An understanding of the genetic composition of local populations and a comparison of populations in large and small serpentine outcrops helps us understand how gene flow, genetic drift, and other factors have shaped the patterns of genetic variation that we see today.
MATERIALS AND METHODS
Study sites
Calystegia collina has a narrow distribution in the North Coast Ranges of Lake, Napa, and Sonoma Counties, California (Fig. 1). The species is endemic to serpentine outcrops or soils derived from these formations (Maxwell clay). The sparse chaparral on serpentine contrasts sharply with the oak woodland on adjacent nonserpentine soils. Raven and Axelrod (1978)
argue that serpentine in California has been available for plant colonization since the Pleistocene, and at some localities since the Miocene. In northern California serpentine is typically occupied by chaparral, dominated by the endemic shrubs Quercus durata Jepson, Arctostaphylos viscida C. Parry, and Ceanothus jepsonii Greene as well as nonendemic chaparral species such as ghost pine (Pinus sabiniàna Dougl.) and chamise (Adenostoma fasciculatum Hook. and Arn.). Within serpentine outcrops Calystegia collina populations are discrete and patchily distributed in open rocky gaps in the chaparral. Serpentine outcrops on which C. collina populations are found range from <1 ha to >2500 km2.
|
). Successful pollination results in a capsule containing one to four seeds. When mature, the capsule splits and seeds drop to the ground. The seeds of this species exhibit no obvious long-distance dispersal mechanisms. Calystegia collina populations ranging in size from <50 ramets (<2 m2) to >3000 ramets (up to 400 m2) occur on both large and small outcrops. During May 1997 we collected young leaf tissue from 30 randomly marked ramets in 32 C. collina populations, half from small outcrops (<5 ha) and the other half from large outcrops (>1000 ha). We sampled four populations in the northeast part of C. collina's range, six each in the northwest and central west and eight each in the central east and south central (Fig. 1). Maps of each population were constructed by recording the distance from the center of the population to the farthest ramet at 30° intervals. We located sample ramets by overlaying a grid on each population map, selecting random coordinates from a list of random numbers, and finding the nearest plant to the random coordinates.
Isozyme analysis
Leaf tissue was kept on ice and flown to the University of Georgia for electrophoretic analysis. In order to extract enzymes, leaf tissue was crushed with mortar and pestle using the extraction buffer of Wendel and Parks (1982)
. Enzyme extracts were absorbed onto chromatography wicks and stored in an ultracold freezer at -70°C. Starch-gel electrophoresis (10% starch gels) was used to estimate genetic diversity. Gels were stained for ten enzyme systems to resolve 19 polymorphic allozyme loci: amino acid transferase (AAT, three loci), uridine glucose pyrophosphatase (UGPP, two loci), 6-phosphogluconase dehydrogenase (6PGD, one locus), phosphoglucose isomerase (PGI, three loci), aconitase (ACO, two loci), isocitrate dehydrogenase (IDH, one locus), menadione reductase (MNR, two loci), florescent esterase (FE, two loci), colormetric esterase (CE, one locus), triose-phosphate isomerase (TPI, two loci). Stain recipes were taken from Soltis et al. (1983)
except for MNR, which was from Cheliak and Pitel (1984)
.
Data analysis
We calculated standard measures of genetic diversity for each population, including the mean number of alleles per locus (A) and per polymorphic locus (AP), the percentage of polymorphic loci (P), and mean observed heterozygosity (Ho) and expected heterozygosity (He) at the species and within-population levels. Wright's fixation index (Wright, 1922
) was used to calculate deviations from Hardy-Weinberg equilibrium for each polymorphic locus within populations [(F = He - Ho)/He]. Chi-square tests were used to test for significant deviations in the fixation indices from the expected value F = 0 (Li and Horvitz, 1953
). Chi-square tests were also used to test for heterogeneity in allele frequencies among populations (Workman and Niswander, 1970
). Total genetic heterozygosity (HT), heterozygosity within populations (HS), genetic diversity among populations (DST), and the proportion of genetic diversity found among populations (GST) were calculated following the equations of Nei (1973, 1977)
.
Deviations of genotypic frequencies from Hardy-Weinberg expectations were tested using exact tests with the alternative hypothesis of heterozygote deficiency using GENEPOP v1.2 (Raymond and Rousset, 1995
). This is a more powerful test than the two-tailed probability test (Raymond and Rousset, 1995
).
Genetic statistics for large and small outcrops were compared using nonparametric Mann-Whitney tests because genetic measures included proportions and indices that are not expected to be normally distributed.
A dendrogram based on genetic distance was created using the UPGMA (unweighted pair group method, arithmetic mean) method. Genetic divergence among populations was estimated by calculating Nei's genetic distance (D) and identity (I) for all pairs of populations (Nei, 1972
). Nei's genetic distance measure ranges from 0 for populations with identical allele frequencies to infinity for populations that do not share any alleles. Identity measures range from 0 for populations with no alleles in common to unity for populations with identical allele frequencies. We also calculated the overall divergence (Gst) for each polymorphic locus by calculating the proportion of total genetic variation found among populations and averaging across loci (Nei, 1973, 1977
).
Relationships between genetic and geographic distances were analyzed using Mantel's test, a permutation procedure that overcomes the lack of independence in comparisons of distance and dissimilarity matrices (Manly, 1991
).
F statistics, which measure departures from expected levels of heterozygosity, were calculated within and among populations using the program GDA (Lewis and Zaykin, 1996
). Genetic structure (amount and distribution of genetic varation) at several spatial scales was examined among five geographic regions and 32 populations. F statistics that do not use the observed heterozygosity within populations (i.e., which are not directly affected by clonal reproduction and obligate outcrossing) provide more meaningful indices of population differentiation. The fixation index, FST, compares the expected heterozygosity within populations or subpopulations (HS) with HT. Since both HS and HT are based on the assumption of random mating within their respective scales of reference (i.e., within populations vs. across the entire study area), their magnitudes are directly related to patterns of differentiation in allele frequencies (Hartl, 1988
). Hierarchical F statistics were computed using the methods of Weir (1990)
for populations and subpopulations. We use hierarchical F statistics to partition variation from expected heterozygosity among all populations (i.e, subpopulations) into the effects of regions (FRT), outcrops within regions (FOR), populations within outcrops (FSO), and populations within the entire study area (FST). Calculations follow the procedures of Weir and Cockerham (1984)
and parallel the analysis of McCue, Buckler, and Holtsford (1996)
. Variances of the F statistics were estimated by jackknifing across loci (Weir and Cockerham, 1984
). Standard errors of the jackknife statistics were estimated following Sokal and Rohlf (1995)
.
Indirect estimates of gene flow were calculated using Wright's equation (1951), as modified by Crow and Aoki (1984)
: Nm
= 1/4(1/Fst - 1), where = [n/(n - 1)]2, n = the number of subpopulations (here called populations), and Nm is the average number of migrants exchanged per generation. A second estimate of Nm, using the distribution of "private alleles" (alleles found in only one population), was calculated by the computer program GENEPOP (Raymond and Rousset, 1995
) following the procedure of Barton and Slatkin (1986)
.
RESULTS
Local patterns of genetic diversity
Of the 34 loci resolved, 19 (56%) were polymorphic across the range of C. collina and 16.9% were polymorphic, on average, at the population level (Table 1). The mean number of alleles per polymorphic locus was 2.63 at the species level and 2.11 at the population level. Including the 15 monomorphic loci, the average number of alleles per locus was 1.91 at the species level and 1.19 at the population level. We found 65 distinct alleles overall (among 34 loci), and the average number of alleles per sample of 30 ramets was ~40. Only seven of the 65 alleles were unique to a single population (= "private" alleles), while 40% of the alleles (52% of the alleles at polymorphic loci) were found in ten or fewer populations (Fig. 2). Including the 15 alleles at monomorphic loci, 29 alleles (44.6%) were found in all 32 populations.
|
|
). The six populations with unique or "private" alleles (C2A, HW2, HW1, HS3, SS2, and MTA; one population had two private alleles) contain 84% of the total allelic richness. All of the alleles encountered in this study can be found in eight populations, including the six with "private" alleles and two others (MT3 and PL2). These populations represent all but the Central East (CE) geographic region (Fig. 1).
Expected heterozygosity for the species was somewhat low (0.11); in other words, only 11% of the genetic individuals are expected to be heterozygous at a given locus under random mating conditions. At the population level, average expected heterozygosity was 0.067, similar to values found in local populations of other endemic plant species (Hamrick and Godt, 1989
). Observed levels of heterozygosity in populations (Table 1), however, were almost always higher than Hardy-Wienberg expectations, averaging 0.108 and ranging from 0.029 (B1A) to 0.181 (IC2).
Comparisons of small vs. large serpentine outcrops
No measure of genetic diversity (allelic richness, % polymorphic loci, observed heterozygosity) differed significantly between populations in large vs. small serpentine outcrops (P > 0.20 in all cases, Mann-Whitney test). The combination of all 16 populations on small outcrops contained 60 of the 65 recorded alleles, representing 84% of the total allelic diversity; the 16 populations on large outcrops also contained 60 of the 65 alleles. Levels of genetic variation did differ, however, among geographic regions within the study area (P < 0.05, Kruskal-Wallis test). In particular, populations in the Central West (CW) and North West (NW) regions had more alleles, more multilocus genotypes, more polymorphic loci, and higher levels of observed heterozygosity than were present in other regions (Table 1). The most isolated region (SC) had the lowest values for all genetic diversity measures.
Genetic relationships among populations
A UPGMA dendrogram illustrating genetic relationships among populations (Fig. 3) shows a complex pattern that only partly reflects the spatial configuration of populations. Three of the four populations in the North East Region (NE), for example, are grouped in the same cluster of sites, while the fourth patch is classified in the adjacent cluster (Fig. 3). Likewise, populations from the South Central Region (SC) tend to be grouped together in the central portion of the dendrogram, but they are included in several lower order clusters. Populations from the other regions are scattered across the dendrogram. In some cases, nearby sites within these regions were very similar genetically (e.g., CR1 and CR2), but in other cases they were very different (e.g., MH1 and MH2).
|
) are significantly associated with geographic distances (r = 0.17, P < 0.002, Mantel test), but the relationship is weak and considerable scatter exists at all geographic scales (Fig. 4). On a smaller scale, only a single Mantel test involving populations within the same large outcrop (the northwestern outcrop in Region CE) yielded a statistically significant correlation between geographic and genetic distance (r = 0.10, P = 0.04), and none of the Mantel tests comparing genetic and geographic distances among small outcrops in the same region yielded statistically significant correlations. If we compare genetic distances among all population pairs inhabiting the same large outcrop (N = 16) and all pairs inhabiting small outcrops in the same region (N = 25), we find no significant relationship between genetic and geographic distances (P > 0.67), and no significant difference between the two categories (P > 0.93, Generalized Linear Model). Although the distance measures are not strictly independent, this analysis is conservative, because the violation of statistical independence would tend to favor Type I statistical error (Sokal and Rohlf, 1995
).
|
) were highly significant for all polymorphic loci. In most cases, at least one and sometimes five or more of the 32 populations included unusual alleles or allele combinations. Often these rare or uncommon alleles were quite common and widespread at sites where they occurred.
Mean genetic identity (I) among pairs of populations, on the other hand, was usually high (
= 0.947; SD = 0.023), suggesting that relatively few alleles contributed significantly to the among-population genetic variation. Lowest genetic identity (I = 0.866) was detected between population MT3 (one of the genetically richest populations) and population C2A (one of the most genetically impoverished populations). The highest identity (0.998) was recorded between populations IC4 and PL1 and populations B1A and C1A, pairs that were separated by 14.8 and 27.6 km, respectively. Population B1A, which is homozygous for the most common allele at all but one locus, shows high identity with most of the other populations ( = 0.964, SD = 0.02).
F statistics
More than 40% of the genetic variation in our sample can be attributed to variation among populations (Gst =0.417). Identification of other sources of variation by hierarchical F statistics (Weir, 1990
) is complicated by C. collina's clonal reproduction and obligate outcrossing. The inbreeding coefficients (FIS) for nearly all of the 19 polymorphic loci were < 0 (Table 2), reflecting that observed levels of heterozygosity within populations (HI) were greater than would have been expected if ramets were the products of random sexual reproduction, which of course is unlikely given the known ability of C. collina to reproduce vegetatively. This also affected values of FIT, which compare observed heterozygosity within populations (HI) with expected heterozygosity based on random mating within the entire study area (HT). High FST values for virtually all loci (Table 2) indicate significant genetic differentiation at the population level. If this variation is partitioned into geographic regions and outcrops (Table 2), we find only minor additional sources of differentiation. For example, analysis of HS relative to expected heterozygosity within outcrops yields FSO, which tends to closely resemble the corresponding value of FST. In other words, populations are differentiated from one another within outcrops roughly to the same extent as they are differentiated within the entire study area. A similar result is obtained when we consider the expected heterozygosity in outcrops relative to regions, FOR, or the expected heterozygosity in regions relative to the entire study area, FRT (Table 2).
|
|
|
model gave an estimate of Nm = 0.35. The low value of Nm again reflects the high value of FST. An independent estimate of gene flow (Barton and Slatkin, 1986
) based on private alleles gave an even lower estimate of gene flow (Nm = 0.054). The disparity between these results indicates a higher number of private alleles than expected according to simulations and theoretical relationships between FST and the number of private alleles (Slatkin and Barton, 1989
). Both estimates of Nm < 1 suggest that gene flow between populations is inadequate to counter the effects of genetic drift in local populations (Real, 1994
).
Separate calculations of gene flow for populations of large vs. small outcrops suggest that migration barriers between populations in the same large outcrops are little different than those between populations in small outcrops. Using FST to estimate gene flow (Crow and Aoki, 1984
), Nm = 0.32 (on average) for large outcrops and Nm = 0.36 for small outcrops. The private allele method gave similarly small differences between large and small outcrops; of the seven private alleles, four were found in large outcrops, while three were found in small outcrops.
DISCUSSION
Measures of genetic variation in Calystegia collina are similar to values for other narrowly endemic plant species reported by Hamrick and Godt (1989)
. The distribution of genetic diversity within and among populations of C. collina, however, follows an unexpected pattern. Most of the species' genetic variation occurs among populations, independently of the serpentine outcrop or region in which they occur. Gene flow between populations is low, not only between small outcrops but also between populations occupying the same large outcrop.
Populations in small serpentine outcrops do not appear to be substantially impoverished compared with populations within extensive serpentine outcrops even though small outcrops are separated by extensive areas of nonserpentine vegetation, often disturbed by grazing and human activities. This result suggests that losses of genetic diversity through the combined effects of genetic drift, inbreeding, and local selection are no more severe on small serpentine outcrops than on larger serpentine outcrops, where populations are more numerous. Likewise, higher rates of pollination from nearby populations (Wolf, 1998
) do not generate significantly higher levels of genetic diversity in large outcrops and do not increase the genetic similarity between geographically proximate populations. Seeds produced by interpopulation pollination apparently rarely germinate and grow within populations where they are produced; if they did, uniformly heterozygous genotypes would be highly unlikely, and genetic diversity would be higher on large outcrops, where production of seeds (which incorporate new alleles into the population) is demonstrably greater (Wolf, 1998
).
How, then, does genetic diversity arise in C. collina populations? The genetic composition of local populations provides important clues. Two of the 32 local populations consisted of a single genotype. All of our sample ramets from these populations must have been the products of vegetative (clonal) reproduction from a single founder; genetic diversity in the "population" merely reflects the genotype of the founder individual. Other populations also had a high degree of genetic uniformity. In large outcrops, the median number of genotypes in our samples of 30 ramets was only 4.5; in small outcrops the median was 5.0. Clearly, vegetative reproduction has played a role in the production of ramets in most, if not all, of these C. collina populations.
Perhaps more revealing was the occurrence of identical heterozygous genotypes in all 30 sample ramets (i.e., uniform heterozygotes). All but seven of the 32 populations were uniformly heterozygous at one or more loci, and 16 populations were uniformly heterozygous at two or more loci. The probability that repeated sexual reproduction could produce 30 ramets with identical heterozygous genotypes and no homozygotes is exceedingly small. This result again implies that vegetative reproduction must play a major role in the origin of C. collina ramets.
If our sample ramets originated through vegetative reproduction, however, these ramets should be uniform at all loci, which was rarely the case. Populations that are uniformly heterozygous at one locus (e.g, Table 3) might be variable at other loci if: (1) homozygotes at the one locus are nonviable or are eliminated by strong selection, or (2) all 30 sample ramets are descended from parents which are homozygous for two different alleles at one locus, or (3) new alleles are created occasionally by somatic mutations and are subsequently propagated by vegetative reproduction (Klekowski, 1997
).
Explanation 1 is unlikely because homozygotes at the loci in question often occur in other populations. Explanation 2 would require unique circumstances. According to this scenario, all ramets in the population are descendants of two founding parents, each homozygous for a different allele. Indeed, experimental studies (Wolf, 1998
) have shown that C. collina is an obligate outcrosser, which supports the idea that seeds are formed mainly by pollination between individuals from two different populations. In order to account for the lack of uniformity in all alleles, multiple founders (with slightly different genetic ancestry) must have become established at the site. Recruitment of seeds that are produced by cross-pollination within the site must be very rare, because no homozygotes are found at the uniformly heterozygous loci. Explanation 3 also implies very little recruitment of seeds produced within the site. In this case, however, the population may have originated from a single founder. According to this explanation genetic variation in the population is attributable to somatic mutations, which are maintained in the population by vegetative reproduction. Note that explanations 2 and 3 are not mutually exclusive. Yet occasional somatic mutations, coupled with clonal reproduction, are consistent with high Fst values, high mean genetic identity among populations, as well as lack of significant differences in allelic diversity between small vs. large serpentine outcrops.
Somatic mutations are widely studied by crop geneticists (e.g., Maluszynski, Ahloowalia, and Sigurbjornsson, 1995
; Richards, 1996
; Salomonson, 1996
; Kinoshita and Mori, 1998
) and can be an important source of genetic variation in long-lived plants (D'Amato, 1997
; Klekowski, 1997
). Much less is known about the relevance of somatic mutations for the evolution of natural plant populations (Schaal and Leverich, 1996
; de Kroon and van Groenendael, 1997
; McLellan et al., 1997
). Clonal plants like C. collina are excellent candidates for the expression of somatic mutations because individual genets occur almost exclusively in small, long-lived populations. None of our 32 C. collina populations has gone extinct during 4 yr of observation, and we have encountered no new populations in the vicinity of these populations. New populations probably originate as a result of rare founder events involving only one or a small number of seeds, similar to the situation reported for island populations of Silene dioica in Sweden (Giles and Goudet, 1997
). Vegetative reproduction subsequently leads to genetically uniform clones, with self-incompatible flowers (Wolf, Harrison, and Hamrick, 2000
). Our results imply that most local populations of C. collina consist of only a few of these clones.
The pattern of genetic relationships among populations shows some effects of geography, but the relationship between genetic and geographic distance explains only a small fraction of the genetic variation in C. collina. Hierarchical F statistics show that most of the genetic heterogeneity occurs among populations; in fact, the Fst values reported here are among the highest values reported for any outcrossing plant species (Hamrick and Godt, 1989
). Similar results were reported by Kim and Chung (1995)
for Calystegia species in Korea.
Gene flow appears to be very low between C. collina populations based on indirect population genetics statistics (Nm). One mechanism of gene flow, seed migration, is very unlikely given the extensive spaces separating neighboring populations, even within large serpentine outcrops. Pollen dispersal by bees clearly is the most likely mechanism of gene flow between populations. Broyles, Schnabel, and Wyatt (1994)
reported that bees are able to disperse pollinia as far as 1 km. This distance would link virtually all of the populations in this study to at least one other C. collina population. Genetic composition (e.g., presence of uniform heterozygotes), on the other hand, argues that seeds produced by successful pollinations are rarely recruited into the population. Perhaps the most important role of seeds is to establish new populations under rare circumstances of recruitment or dispersal. Local populations of C. collina tend to show unusual genotype distributions, perhaps because somatic mutations tend to increase the uniqueness of local populations or because populations originate from a limited numbers of founders from the same two parent populations.
Bossart and Prowell (1998)
emphasize the limitations of population genetic metrics (e.g., Fst, Nm,) that use Hardy-Weinberg equilibria as baseline conditions. Long-lived, clonal plant species like C. collina provide additional reasons why these approaches should be interpreted cautiously. Processes that generate and maintain genetic diversity in C. collina populations include self-incompatability, vegetative reproduction, and perhaps somatic mutations in long-lived clones. These processes complicate the application of traditional population genetics. Indirect estimates of gene flow (Nm) are likely to underestimate the amount of gene flow because within-population events contribute significantly to differences between populations. Generation time itself is difficult to establish in clonal species because vegetatively reproducing genets might persist for hundreds or thousands of years (Klekowski, 1997
). Our results suggest that the importance of clonal reproduction and somatic mutation for plant population genetics deserves considerable future study and may provide important insights into the longer term evolutionary diversification of C. collina and other long-lived plant species.
Regardless of the evolutionary interpretation, information about the genetic composition of local populations provides an important tool for the conservation of plant species like C. collina (Holsinger and Gottlieb, 1991
). Individual populations contained as much as 42% and as little as 3% of the allelic richness recorded in our genetic sample. All of the observed allelic richness at 34 loci (19 of which are polymorphic) can be preserved in eight of the 32 sampled populations. Perhaps more importantly, we have shown that populations in small serpentine outcrops are often not genetically impoverished, and hence they should be included in any comprehensive conservation plan.
FOOTNOTES
1 The authors thank S. Harrison, J. Quinn, and S. Strauss for comments on the manuscript; M. J. Godt and M. Burke for assistance with gel electrophoresis; G. Starr for assistance with computer programs; and J. Callizo and N. Jurjavcic for help with field work and logistics. This research was supported by funds from the Graduate Group in Ecology at the University of California Davis and National Science Foundation Grant DEB 94-24137. 
LITERATURE CITED
Barrett, S. C. H., and J. R. Kohn. 1991 Genetic and evolutionary consequences of small population size in plants: implications for conservation. In D. A. Falk and K. E. Holsinger [eds.], Genetics and conservation of rare plants, 3–30. Oxford University Press, New York, New York, USA.
Barton, N. H., and M. Slatkin. 1986 A quasi-equilibrium theory of the distribution of rare alleles in a sub-divided population. Heredity 56: 409–415.
Bossart, J. L., and D. P. Prowell. 1998 Genetic estimates of population structure and gene flow: Limitations, lessons and new directions. Trends in Ecology and Evolution 13: 202–206.
Brooks, R. R. 1987 Serpentine and its vegetation: a multidisciplinary approach. Dioscorides Press, Portland, Oregon, USA.
Broyles, S. B., A. Schnabel, and R. Wyatt. 1994 Evidence for long-distance pollen dispersal in milkweeds (Asclepias exaltata). Evolution 48: 1032–1040.[CrossRef][ISI]
Cheliak, W. M., and J. A. Pitel. 1984 Techniques for starch gel electrophoresis of enzymes from forest tree species. Petawawa National Forestry Institute, Information Report P1-X-42. Canadian Forestry Service, Agriculture, Canada.
Crow, J. F., and K. Aoki. 1984 Group selection for a polygenic behavioral trait: estimating the degree of population subdivision. Proceedings of the National Academy of Sciences, USA 81: 6073–6077.
D'Amato, F. 1997 Role of somatic mutations in the evolution of higher plants. Caryologia 50: 1–15.
de Kroon, H., and J. van Groenendael. 1997 The ecology and evolution of clonal plants. Backhuys Publishers, Leiden, The Netherlands.
Ellstrand, N. C., and D. R. Elam. 1993 Population genetic consequences of small population size: implications for plant conservation. Annual Review of Ecology and Systematics 24: 217–242.[CrossRef][ISI]
Giles, B. E., and J. Goudet. 1997 Genetic differentiation in Silene dioica metapopulations: estimation of spatiotemporal effects in a successional plant species. American Naturalist 149: 507–526.[CrossRef][ISI]
Hamrick, J. L., and M. J. Godt. 1989 Allozyme diversity in plant species. In A. D. H. Brown, M. T. Clegg, A. L. Kahler, and B. S. Weir [eds.], Plant population genetics, breeding and genetic resources, 43–63. Sinauer, Sunderland, Massachusetts, USA.
Hanski, I. A., and M. E. Gilpin. 1991 Metapopulation biology: ecology, genetics, and evolution. Academic Press, San Diego, California, USA.
Harper, J. L. 1977 Population biology of plants. Academic Press, London, UK.
Hartl, D. L. 1988 A primer of population genetics. Sinauer, Sunderland, Massachusetts, USA.
Hedrick, P. W. 1985 Genetics of populations. Jones and Bartlett, Boston, Massachusetts, USA.
Holsinger, K. E. 1993 The evolutionary dynamics of fragmented plant populations. In P. M. Kareiva, J. G. Kingsolver, and R. B. Huey [eds.], Biotic interactions and global change, 198–216. Sinauer, Sunderland, Massachusetts, USA.
———, and L. D. Gottlieb. 1991 Conservation of rare and endangered plants: principles and prospects. In D. A. Falk and K. E. Holsinger [eds.], Genetics and conservation of rare plants, 195–208. Oxford University Press, New York, New York, USA.
Karron, J. D. 1991 Patterns of genetic variation and breeding systems in rare plant species. In D. A. Falk and K. E. Holsinger [eds.], Genetics and conservation of rare plants, 87–98. Oxford University Press, New York, New York, USA.
Kim, S. T., and M. G. Chung. 1995 Genetic and clonal structure in Korean populations of Calystegia japonica (Convolvulaceae). Botanical Bulletin Academia Sinica 36: 135–141.
Kinoshita, T., and K. Mori. 1998 Somaclonal selection of physiological mutants and several problems related to rice cell breeding. Journal of Applied Genetics 39: 9–35.
Klekowski, E. J., Jr. 1997 Somatic mutation theory of clonality. In H. de Kroon and J. van Groenendael [eds.], The ecology and evolution of clonal plants, 227–241. Backhuys Publishers, Leiden, The Netherlands.
Kruckeberg, A. R. 1984 California serpentines: flora, vegetation, geology, soils and management problems. University of California Press, Berkeley, California, USA.
Lewis, P. O., and D. Zaykin. 1996 Genetic Data Analysis: Software for the analysis of discrete genetic data. Computer program distributed by the authors. University of Connecticut, Storrs, Connecticut, USA.
Li, C. C., and D. G. Horvitz. 1953 Some methods of estimating the inbreeding coefficient. American Journal of Human Genetics 5: 107–117.[ISI][Medline]
Maluszynski, M., B. S. Ahloowalia, and B. Sigurbjornsson. 1995 Application of in vivo and in vitro mutation techniques for crop improvement. Euphytica 85: 303–315.[CrossRef][ISI]
Manly, B. F. J. 1991 Randomization and Monte Carlo methods in biology. Chapman and Hall, London, UK.
McCue, K. A., E. S. Buckler, and T. P. Holtsford. 1996 A hierarchical view of genetic structure in the rare annual plant Clarkia springvillensis. Conservation Biology 10: 1425–1434.
McLellan, A. J., D. Prati, O. Kaltz, and B. Schmid. 1997 Structure and analysis of phoenotypic and genetic variation in clonal plants. In H. de Kroon, and J. van Groenendael [eds.], The ecology and evolution of clonal plants, 185–210. Backhuys Publishers, Leiden, The Netherlands.
Nei, M. 1972 Genetic distance between populations. American Naturalist 106: 283–291.[CrossRef][ISI]
———. 1973 Analysis of gene diversity in subdivided populations. Proceedings of the National Academy of Sciences, USA 70: 3321–3323.
———. 1977 F-statistics and analysis of gene diversity in subdivided populations. Annals of Human Genetics 41: 225–233.[ISI][Medline]
Petit, R. J., A. El Mousadik, and O. Pons. 1998 Identifying populations for conservation on the basis of genetic markers. Conservation Biology 12: 844–855.[CrossRef][ISI]
Raven, P. H., and D. I. Axelrod. 1978 Origin and relationships of the California flora. University of California Publications in Botany 72: 1–134.
Raymond, M., and F. Rousset. 1995 GENEPOP (Version 1.2): population genetics software for exact tests and ecumenicism. Journal of Heredity 86: 248–249.
Real, L. A. 1994 Ecological genetics. Princeton University Press, Princeton, New Jersey, USA.
Richards, A. J. 1996 Genetic variability in obligate apomicts of the genus Taraxacum. Folia Geobotanica et Phytotaxonomica 31: 405–414.
Salomonson, A. 1996 Interactions between somatic mutations and plant development. Vegetatio 127: 71–75.[CrossRef][ISI]
Schaal, B. A., and W. J. Leverich. 1996 Molecular variation in isolated plant populations. Plant Species Biology 11: 33–40.
Skinner, M. W., and B. M. Pavlik. 1994 California Native Plant Society's inventory of rare and endangered plants of California. California Native Plant Society Special Publication Number 1, Sacramento, California, USA.
Slatkin, M., and N. H. Barton. 1989 A comparison of three indirect methods for estimating average levels of gene flow. Evolution 43: 1349–1368.[CrossRef][ISI]
Sokal, R. R., and F. J. Rohlf. 1995 Biometry: the principles and practice of statistics in biological research. W. H. Freeman, New York, New York, USA.
Soltis, D. E., C. H. Haufler, D. C. Darrow, and G. J. Gastony. 1983 Starch gel electrophoresis of ferns: a compilation of grinding buffers, gel and electrode buffers, and staining schedules. American Fern Journal 73: 9–27.[CrossRef][ISI]
Wagner, D. L., and E. J. Bortugno. 1982 Geologic map of the Santa Rosa Quadrangle, California, 1:250,000. State of California, Resources Agency, Sacramento, California, USA.
Weir, B. S. 1990 Genetic data analysis. Sinauer Associates, Sunderland, Massachusetts, USA.
———, and C. C. Cockerham. 1984 Estimating F-statistics for the analysis of population structure. Evolution 38: 1358–1370.[CrossRef][ISI]
Wendel, J. F., and C. R. Parks. 1982 Genetic control of isozyme variation in Camillia japonica L. Heredity 73: 197–204.
Wolf, A. T. 1998 Population structure and reproductive ecology of serpentine endemic plant species in California's North Coast Range. Ph.D. Dissertation, University of California, Davis, California, USA.
———, S. P. Harrison, and J. L. Hamrick. 2000 Influence of habitat patchiness on genetic diversity and spatial structure of a serpentine endemic plant. Conservation Biology 14: 452–463.
Workman, P. L., and J. D. Niswander. 1970 Population studies of southwestern Indian tribes: 2. Local genetic differentiation in the Papago. American Journal of Human Genetics 22: 4–49.
Wright, S. 1922 Coefficients of inbreeding and relationship. American Naturalist 56: 330–338.[CrossRef][ISI]
———. 1931 Evolution in Mendelian populations. Genetics 16: 97–159.
———. 1951 The genetical structure of populations. Annals of Eugenics 15: 324–354.
Young, A., T. Boyle, and T. Brown. 1996 The population genetic consequences of habitat fragmentation for plants. Trends in Ecology and Evolution 11: 413–418.[CrossRef]
This article has been cited by other articles:
|
|
 |
|
  K. Kurata, T. Jaffre, and H. Setoguchi Genetic diversity and geographical structure of the pitcher plant Nepenthes vieillardii in New Caledonia: A chloroplast DNA haplotype analysis Am. J. Botany, December 1, 2008; 95(12): 1632 - 1644. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. Medrano and C. M. Herrera Geographical Structuring of Genetic Diversity Across the Whole Distribution Range of Narcissus longispathus, a Habitat-specialist, Mediterranean Narrow Endemic Ann. Bot., August 1, 2008; 102(2): 183 - 194. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
C. Torres-Diaz, E. Ruiz, F. Gonzalez, G. Fuentes, and L. A. Cavieres Genetic Diversity in Nothofagus alessandrii (Fagaceae), an Endangered Endemic Tree Species of the Coastal Maulino Forest of Central Chile Ann. Bot., July 1, 2007; 100(1): 75 - 82. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M. V. Cattell and S. A. Karl Genetics and morphology in a Borrichia frutescens and B. arborescens (Asteraceae) hybrid zone Am. J. Botany, November 1, 2004; 91(11): 1757 - 1766. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
); populations on large outcrops are represented by solid circles (
). Locations of serpentine outcrops are associated with areas of serpentine bedrock (shaded) derived from USGS Bedrock Geology Map (Wagner and Bortugno, 1982
) and pairs of populations occupying the same large outcrop (