Modeling Problems in Conservation Genetics Using Captive Drosophila Populations: Consequences of Equalization of Family Sizes

Equalization of family sizes is recommended for use in captive breeding programs, as it is predicted to double effective population sizes, reduce inbreeding, and slow the loss of genetic variation. The effects of maintaining small captive populations with equalization of family sizes versus random choice of parents on levels of inbreeding genetic variation, reproductive fitness, and effective population sizes ( N _e) were evaluated in 10 lines of each treatment maintained with four pairs of parents per generation. The mean inbreeding coefficient ( F ) increased at a significantly slower rate with equalization than with random choice (means of 0.35 and 0.44 at generation 10). Average heterozygosities at generation 10, based on six polymorphic enzyme loci, were significantly higher with equalization (0.149) than with random choice (0.085), compared to the generation 0 level of 0.188. The competitive index measure of reproductive fitness at generation 11 was more than twice as high with equalization as with random choice, both being much lower than in the outbred base population. There was considerable variation among replicate lines within treatments in all the above measures and considerable overlap between lines from the two treatments. Estimates of N _e for equalization were greater than those for random choice, whether estimated from changes in average heterozygosities or from changes in F. Equalization of family sizes can be unequivocally recommended for use in the genetic management of captive populations.     

Effective Population Size and Maintenance of Genetic Diversity in Captive-Bred Populations of a Lake Victoria Cichlid

Abstract: We used microsatellite DNA markers to investigate the maintenance of genetic diversity within and between samples of subpopulations (spanning five captive-bred generations) of the haplochromine cichlid Prognathochromis perrieri . The subpopulations are maintained as part of the Lake Victoria Cichlid species survival plan. Changes in the frequencies of 24 alleles, over four polymorphic loci, were used to estimate effective population size (   N _e   ). Point estimates of N _e ranged from 2.5 to 7.7 individuals and were significantly smaller than the actual census size (   N _obs  ) for all subpopulations (32–243 individuals per generation), with the corresponding conservative N _e   /  N _obs ratios ranging from 0.01 to 0.12. Approximately 19% of the initial alleles were lost within the first four generations of captive breeding. Between-generation comparisons of expected heterozygosity showed significant losses ranging from 6% to 12% per generation. Seven private alleles were observed in the last sampled generation of four subpopulations, and analysis of population structure by F _ST indicated that approximately 33% of the total genetic diversity is maintained between the subpopulations from different institutions. To reduce the loss of genetic variation, we recommend that offspring production be equalized by periodically removing dominant males, which will encourage reproduction by additional males. Consideration should also be given to encouraging more institutions to maintain populations, because a significant fraction of the genetic variation exists as among-population differences resulting from random differentiation among subpopulations.     

Quantitative and Molecular Genetic Variation in Captive Cotton-Top Tamarins (Saguinus oedipus)

Most genetic surveys of captive and endangered populations are carried out with single gene characters bearing no direct relationship to life history or other features for which genetic variation needs to be maintained. Quantitative genetic estimates of heritable variation for life-history traits may be a more direct and appropriate measure of genetic variation for some conservation purposes. Furthermore, recent theoretical and empirical results indicate that genetic variation measured on these two levels may not be concordant. We analyzed heterozygosity at 41 allozyme loci and heritability for body weight in captive cotton-top tamarins ( Saguinus oedipus ) from the Marmoset Research Center of the Oak Ridge Associated Universities in order to compare these two levels of genetic variation. Cotton-top tamarins are a highly endangered species native to Colombia. Many animals currently reside in research facilities and zoological parks. A total of 106 animals were used in the isozyme survey, while data on 364 animals contributed to the quantitative genetic study of body weight. We found a very low average heterozygosity ( H = 1%) for this colony. Body weight was moderately and significantly heritable ( h ² = 35%). This heritability is within the normal range for natural animal populations. The finding of biologically significant levels of heritability in a population with abnormally low allozyme heterozygosity illustrates the point that low levels of allozyme heterozygosity should not be taken as an indication of overall lack of genetic variation in important quantitative characters such as life-history traits. Genetic variation required for adaptation of species to future environmental challenges can exist despite low levels of enzyme heterozygosity.     

Relationship of Effective to Census Size in Fluctuating Populations

Abstract: The effective size of a population (    N _e   ) rather than the census size (    N ) determines its rate of genetic drift. Knowing the ratio of effective to census size, N _e  /   N , is useful for estimating the effective size of a population from census data and for examining how different ecological factors influence effective size. Two different multigenerational ratios have been used in the literature based on either the arithmetic mean or the harmonic mean in the denominator. We clarify the interpretation and meaning of these ratios. The arithmetic mean N _e  /   N ratio compares the total number of real individuals to the long-term effective size of the population. The harmonic mean N _e  /   N ratio summarizes variation in the N _e  /   N ratio for each generation. In addition, we show that the ratio of the harmonic mean population size to the arithmetic mean population size provides a useful measure of how much fluctuation in size reduced the effective size of a population. We discuss applications of these ratios and emphasize how to use the harmonic mean N _e  /   N ratio to estimate the effective size of a population over a period of time for which census counts have been collected.     

Estimating the Effective Population Size of Conserved Populations

Accurate estimation of effective population size is important in attempts to conserve small populations of animals or plants. We review the genetic and ecological methods that have been used to estimate effective population size in the past and suggest that, while genetic methods may often be appropriate for the estimation of N _e, and its monitoring, ecological methods have the advantage of providing data that can help predict the effect of a changed environment on N _e. Estimation of N _e, is particularly complex in populations with overlapping generations, and we summarize previous empirical estimates of N _e that used ecological methods in such populations. Since it is often difficult to assess what parameters and assumptions have been used in previous calculations, we suggest a method that provides a good estimate of N _e, makes clear what assumptions are involved, and yet requires a minimum of information. The method is used to analyze data from 14 studies. In 36% (5) of these studies, our estimate is in excellent agreement with the original, and yet we use significantly less information, in 21% (3) the original estimate is markedly lower, in 43% (6) it is markedly higher. Reasons for the discrepancies are suggested. Two of the underestimates involve a failure in the original to account for a long maturation time, and four of life overestimates involve problems in the original with the correction for overlapping generations.     

Unprecedented Low Levels of Genetic Variation and Inbreeding Depression in an Island Population of the Black-Footed Rock-Wallaby

Abstract: It has been argued that demographic and environmental factors will cause small, isolated populations to become extinct before genetic factors have a significant negative impact. Islands provide an ideal opportunity to test this hypothesis because they often support small, isolated populations that are highly vulnerable to extinction. To assess the potential negative impact of isolation and small population size, we compared levels of genetic variation and fitness in island and mainland populations of the black-footed rock-wallaby ( Petrogale lateralis [Marsupialia: Macropodidae]). Our results indicate that the Barrow Island population of P. lateralis has unprecedented low levels of genetic variation (  H _e = 0.053, from 10 microsatellite loci) and suffers from inbreeding depression (reduced female fecundity, skewed sex ratio, increased levels of fluctuating asymmetry). Despite a long period of isolation ( ∼ 1600 generations) and small effective population size (  N _e ∼ 15), demographic and environmental factors have not yet driven this population to extinction. Nevertheless, it has been affected significantly by genetic factors. It has lost most of its genetic variation and become highly inbred (  F _e = 0.91), and it exhibits reduced fitness. Because several other island populations of P. lateralis also exhibit exceptionally low levels of genetic variation, this phenomenon may be widespread. Inbreeding in these populations is at a level associated with high rates of extinction in populations of domestic and laboratory species. Genetic factors cannot then be excluded as contributing to the extinction proneness of small, isolated populations.     

Lineage Loss in Serengeti Cheetahs: Consequences of High Reproductive Variance and Heritability of Fitness on Effective Population Size

Abstract: In natural populations, many breeders do not leave surviving offspring, and as a result many potential genetic lineages are lost. I examined lineage extinction in Serengeti cheetahs ( Acinonyx jubatus ) and found that 76% of matrilines were lost over a 25-year period. Production of future breeders was nonrandom and generally confined to a few families. Five out of 63 matrilines accounted for 45% of the total cheetah population over the course of the study. Lineage persistence is perhaps best illustrated by the variance in lifetime reproductive success ( LRS) and heritability in this parameter. In female cheetahs, variance in LRS was high, and new data show that this LRS was heritable. Variance in LRS and heritability in LRS have dramatic consequences for effective population size, N _e. I calculated N _e for cheetahs, taking into account fluctuating population size, unequal sex ratio, non-Poisson distribution of reproductive success, and heritability of fitness. The N _e was most strongly affected by variance in reproductive success and especially heritability in reproductive success. The variance N _e was 44% of the actual population size, and the inclusion of heritability further reduced N _e to only 15% of the actual population, a ratio similar to that of a social carnivore with reproductive suppression. The current cheetah population in the Serengeti is below numbers suggested by N _e estimates as sufficient to maintain sufficient genetic diversity.     

Genetic Diversity, Population Size, and Fitness in Central and Peripheral Populations of a Rare Plant Lychnis viscaria

Abstract: Genetic diversity is expected to decrease in small and isolated populations as a consequence of bottlenecks, founder effects, inbreeding, and genetic drift. The genetics and ecology of the rare perennial plant Lychnis viscaria (Caryophyllaceae) were studied in both peripheral and central populations within its distribution area. We aimed to investigate the overall level of genetic diversity, its spatial distribution, and possible differences between peripheral and central populations by examining several populations with electrophoresis. Our results showed that the level of genetic diversity varied substantially among populations (  H _exp = 0.000–0.116) and that the total level of genetic diversity (mean H _exp = 0.056) was low compared to that of other species with similar life-history attributes. The peripheral populations of L. viscaria had less genetic variation (mean H _exp = 0.034) than the central ones (0.114). Analysis of genetic structure suggested limited gene flow (mean F _ST = 0.430) and high differentiation among populations, emphasizing the role of genetic drift (  N _e m = 0.33). Isolation was even higher than expected based on the physical distance among populations. We also focused on the association between population size and genetic diversity and possible effects on fitness of these factors. Population size was positively correlated with genetic diversity. Population size and genetic diversity, however, were not associated with fitness components such as germination rate, seedling mass, or seed yield. There were no differences in the measured fitness components between peripheral and central populations. Even though small and peripheral populations had lower levels of genetic variation, they were as viable as larger populations, which emphasizes their potential value for conservation.     

Influence of Fragmentation and Bottlenecks on Genetic Divergence of Wild Turkey Populations

Abstract: There are few empirical studies of the effects of human-induced fragmentation and bottlenecks on the genetic structure of field populations. Assessment of these effects is necessary to evaluate the relevance of predictions obtained from simulation and theoretical models to the management of wild populations. The genetic structure of populations of the Wild Turkey (Meleagris gallopavo), which have been heavily influenced by human activities, was examined using allozyme electrophoresis Allele frequencies at four polymorphic loci were estimated for 27 localities using 461 turkeys The estimated proportion of genetic variation resulting from differences among populations (F_st= 0.102) was one of the highest reported for any avian species and was also much higher than that observed between turkey populations that had not passed through known bottlenecks Almost all of the variation (89%) among populations was accounted for by differences between groups of populations with different histories of manipulation Fragmented distributions and population bottlenecks due to human activities appear to have increased genetic differentiation among populations of wild turkeys. This observation agrees with theoretical predictions concerning the effects of isolation and bottlenecks on the genetic structure of field populations.     

Population Demographics and Genetic Diversity in Remnant and Translocated Populations of Sea Otters

Abstract: The effects of small population size on genetic diversity and subsequent population recovery are theoretically predicted, but few empirical data are available to describe those relations. We use data from four remnant and three translocated sea otter ( Enhydra lutris ) populations to examine relations among magnitude and duration of minimum population size, population growth rates, and genetic variation. Mitochondrial (mt)DNA haplotype diversity was correlated with the number of years at minimum population size ( r _s = −0.741, p = 0.038) and minimum population size ( r _s = 0.709, p = 0.054). We found no relation between population growth and haplotype diversity, although growth was significantly greater in translocated than in remnant populations. Haplotype diversity in populations established from two sources was higher than in a population established from a single source and was higher than in the respective source populations. Haplotype frequencies in translocated populations of founding sizes of 4 and 28 differed from expected, indicating genetic drift and differential reproduction between source populations, whereas haplotype frequencies in a translocated population with a founding size of 150 did not. Relations between population demographics and genetic characteristics suggest that genetic sampling of source and translocated populations can provide valuable inferences about translocations.     

Genetic Effects of Multiple Generations of Supportive Breeding

Abstract: The practice of supporting weak wild populations by capturing a fraction of the wild individuals, bringing them into captivity for reproduction, and releasing their offspring into the natural habitat to mix with wild ones is called supportive breeding and has been widely applied in the fields of conservation biology and fish and wildlife management. This procedure is intended to increase population size without introducing exogenous genes into the managed population. Previous work examining the genetic effects of a single generation of supportive breeding has shown that although a successful program increases the census population size, it may reduce the genetically effective population size and thereby induce excessive inbreeding and loss of genetic variation. We expand and generalize previous analyses of supportive breeding and consider the effects of multiple generations of supportive breeding on rates of inbreeding and genetic drift. We derived recurrence equations for the inbreeding coefficient and coancestry, and thereby equations for inbreeding and variance effective sizes, under three models for selecting captive breeders: at random, preferentially among those born in captivity, and preferentially among those born in the wild. Numerical examples indicate that supportive breeding, when carried out successfully over multiple generations, may increase not only the census but also the effective size of the supported population as a whole. If supportive breeding does not result in a substantial and continuous increase of the census size of the breeding population, however, it might be genetically harmful because of elevated rates of inbreeding and genetic drift.     

Genetic Analyses Through DNA Fingerprinting of Captive Populations of Hawaiian Geese

DNA fingerprinting was used to assess levels of genetic variation in 106 Hawaiian Geese, or Nene ( Branta sandvicensis ), from two captive colonies in Hawaii and Slimbridge, England. Mantel tests were used to determine differences in mean similarity coefficients obtained from DNA fingerprints between unrelated and related Nene within and between captive colonies and to determine whether pedigree-based estimates of relatedness correlated with DNA fingerprint-based estimates. Between colonies, mean similarity coefficients for unrelated and related Slimbridge Nene were higher than those for Hawaiian Nene. Within each colony, related Nene bad higher mean similarity coefficients than did unrelated Nene. A positive relationship was found between coancestry coefficients and similarity coefficients. A greater number of founders for the Hawaiian colony contributed to the lower mean similarity coefficients. As genetic variation decreases, difficulty in distinguishing relatedness among individuals using DNA fingerprinting may increase. Lower genetic variation also may increase tine error in estimating the relationship between coancestry and similarity coefficients. DNA fingerprinting of Nene identified unique alleles and can determine optimal pairings between individuals. The calibrated similarity coefficient distributions can help determine the relatedness of individuals in wild populations of Nene.     

Genetic Variation for Morphological and Allozyme Variation in Relation to Population Size in Clarkia dudleyana, an Endemic Annual

Abstract: The maintenance of genetic variation within populations is expected to allow species to respond to evolutionary challenges such as selection and environmental stress. Larger populations are generally expected to maintain larger amounts of genetic variation. Although several studies have found a positive relationship between population size and levels of genetic variation for molecular markers such as allozymes, few comparisons have been made between molecular measures of variation and genetic variation that is likely to be ecologically important. Most ecologically important traits require quantitative genetic analyses. I examined the relationship between levels of genetic variation and population size for both allozymes and morphological traits in a California endemic annual plant, Clarkia dudleyana . Levels of genetic variation for allozymes did not show a significant positive relationship with population size. The level of genetic variance for all of the 18 morphological traits exhibited no significant relationship with population size. Further, allozyme heterozygosities were not related to levels of quantitative genetic variation. These results indicate that levels of allozyme variability do not predict levels of genetic variation for morphological traits in C. dudleyana , suggesting that molecular measures of variation, in general, differ from quantitative genetic measures. These results imply that conservation genetic studies should generally focus on aspects other than measuring levels of genetic variation found within populations.     

Comparisons of Genetic Diversity in the Endangered Adenophora lobophylla and Its Widespread Congener, A. potaninii

Abstract: Starch-gel electrophoresis was used to examine the levels and distribution of genetic diversity in two Adenophora species: the narrow endangered Adenophora lobophylla and its widespread congener, A. potaninii . Based on allozyme variation at 18 putative loci, we measured high levels of genetic variability both in the endangered and the widespread species, with 83.3% of the loci being polymorphic. The mean expected heterozygosity within populations (   H _ep  ) and within species (   H _es  ) were 0.234 and 0.244 for A. potaninii and were as high as 0.210 and 0.211 for A. lobophylla . There was higher differentiation among populations in A. potaninii (   F _ST = 0.155) than in A. lobophylla (   F _ST = 0.071). The high levels of genetic diversity in the present allozyme survey are consistent with the morphological variation observed in these species and may be attributed to high outcrossing rates in the Adenophora species. In addition, A. lobophylla was identified as a distinct species on the basis of Nei's genetic distances and thus should be given a high priority for protection. It is noteworthy that the endangered A. lobophylla maintains much higher genetic diversity than most endemic or narrowly distributed plant species in spite of its restricted distribution. We hypothesize that A. lobophylla has become endangered for ecological and stochastic reasons, including habitat destruction or environmental changes, mud slides, and human disturbance such as grazing and mowing. Consequently, habitat protection is of particular importance for conserving this endangered species.     

Is Mutation Accumulation a Threat to the Survival of Endangered Populations?

The accumulation of new deleterious mutations has been predicted to constitute a significant threat to the survival of finite sexually reproducing populations. Three measures of genetic load were made on populations of Drosophila melanogaster maintained at effective population sizes of 25, 50, 100, 250, and 500 for 45 or 50 generations and their outbred base population and a new sample from the same wild population. Genetic loads were measured as fitness differentials between inbred and non-inbred lines derived from each population under both benign ( productivity of single pairs) and competitive (competitive index) conditions. No trend of smaller populations exhibiting greater genetic loads than larger ones was observed under either benign or competitive conditions. Further, genetic loads were similar in captive and wild populations. Frequencies of deleterious and lethal alleles on chromosome II were measured by making the chromosome (approximately 40% of the genome) homozygous using a marked balancer stock. Neither deleterious nor lethal allele frequencies exhibited a relationship with population size. The accumulation of detrimental mutations does not appear to pose a significant threat to finite sexual populations with effective sizes of 25 or more over the 100–200 year time frames considered in most wildlife conservation programs.     

Effects of Population Size, Mating System, and Evolutionary Origin on Genetic Diversity in Spiranthes sinensis and S. hongkongensis

Hong Kong once supported more than 109 species of wild orchids, of which approximately 30% were endemic. Most of the local wild orchids have now become rare or endangered. I conducted a comparative study of genetic diversity in two closely related terrestrial orchids, an allotetraploid, Spiranthes hongkongensis , and its diploid progenitor, S. sinensis , to assess the effects of the population bottleneck associated with the origin of the polyploid and to investigate the relationships between number of breeding individuals, mating system, and level of isozyme variation in their populations. Nearly complete genetic uniformity was observed both within and among populations of S. hongkongensis . In contrast, S. sinensis had high levels of genetic variation for all of the genetic parameters examined. Regression analysis of population size and several components of genetic diversity in S. sinensis revealed that, among various measures of within-population variation, the proportion of polymorphic loci ( P ) and average number of alleles per locus ( A ) or per polymorphic locus ( A _p ) were the most sensitive to population size ( R ² = 0.942, p = 0.001; R ² = 0.932, p = 0.002; and R ² = 0.923, p = 0.002 respectively). The highly negative correlation ( r = −0.999, p < 0.01) between population size and the mean frequency of private alleles in pairwise population comparisons, p (1), indicated that population size may also be used to predict the extent of population differentiation caused by random genetic drift. Conservation of genetic diversity in S. sinensis could be maximized by protecting several of both large and small populations, whereas fewer populations may be needed to achieve this goal for S. hongkongensis.     

A Hierarchical View of Genetic Structure in the Rare Annual Plant Clarkia springvillensis

Genetic structure at several spatial scales was examined in the rare California annual, Clarkia springvillensis . Using seven isozyme-encoding loci as genetic markers, we assessed the amount and distribution of genetic variation among three populations and eight subpopulations. Total genetic variation was lower than in species with similar life history traits but equivalent to that of other endemic plants. Spatial autocorrelation showed some evidence for very limited differentiation within subpopulations at a scale of 1–2 m. The subpopulations, separated by tens of meters, were found to be more differentiated from each other ( F _sp = 0.084) on average than were populations ( F,pt = 0.017). This local genetic differentiation was not correlated with physical distance between subpopulations. The low F_pt estimates suggest that substantial gene flow is occurring among populations. However, the lack of correlation between genetic and geographic distances and the significant differentiation of subpopulations suggest that genetic drift is occurring within populations. Therefore, we believe the apparent homogeneity of populations is due to each population's gene frequencies' being an average of several divergent subpopulations. If drift is causing differentiation within populations, it may eventually cause differentiation between populations. The importance of using a hierarchical approach to evaluating genetic structure is clear. Patterns occurring at one spatial scale may not be evident at others. One should not necessarily conclude that gene flow is substantial and that the risk of genetic erosion via drift is negligible just because differentiation between populations is small; the system may not be at equilibrium. This lesson is particularly important when recent changes in climate or land use are apparent.     

Post-Perturbation Genetic Changes in Populations of Endangered Virgin River Chubs

A 34-kilometer reach of the Virgin River, Utah-Arizona-Nevada, was poisoned with rotenone in an attempt to eradicate non-native red shiners ( Cyprinella lutrensis ), a species implicated in the decline of native fish populations in the American West. An error in detoxification resulted in lethal concentrations of piscicide passing through an additional 50 kilometers of stream. We used allozyme electrophoresis to analyze genetic variation among pre- and post-poison samples of endangered Virgin River chubs ( Gila seminuda ). Pre-poison samples indicated a single panmictic population in the river. In contrast, fish subsequently produced through natural recruitment in poisoned reaches exhibited deviations from the original pattern of genetic variation. A genetic bottleneck caused by severe reduction in the number of spawning adults was indicated. The altered pattern persisted 2.5 years post-poisoning, indicating unexpectedly slow recolonization from the unpoisoned reach upstream. Genetic variation among hatchery-produced young was similarly unrepresentative of the original pattern because of the small number of brood fish used in propagation. Because of their small numbers and/or restricted distribution, endangered species are particularly vulnerable to natural or anthropogenic catastrophes. Assessment of the genetic impact of such events is essential but requires that baseline data are available.     

Effects of Habitat Fragmentation on Effective Population Size in the Endangered Rio Grande Silvery Minnow

Abstract:  We assessed spatial and temporal patterns of genetic diversity to evaluate effects of river fragmentation on remnant populations of the federally endangered Rio Grande silvery minnow ( Hybognathus amarus ). Analysis of microsatellite and mitochondrial DNA detected little spatial genetic structure over the current geographic range, consistent with high gene flow despite fragmentation by dams. Maximum-likelihood analysis of temporal genetic data indicated, however, that present-day effective population size ( N_eV ) of the largest extant population of this species was 78 and the ratio of effective size to adult numbers ( N_eV/N ) was ∼ 0.001 during the study period (1999 to 2001). Coalescent-based analytical methods provided an estimate of historical (river fragmentation was completed in 1975) effective size ( N_eI  ) that ranged between 10⁵ and 10⁶. We propose that disparity between contemporary and historical estimates of N_e and low contemporary N_e/N result from recent changes in demography related to river fragmentation. Rio Grande silvery minnows produce pelagic eggs and larvae subject to downstream transport through diversion dams. This life-history feature results in heavy losses of yearly reproductive effort to emigration and mortality, and extremely large variance in reproductive success among individuals and spawning localities. Interaction of pelagic early life history and river fragmentation has altered demographic and genetic dynamics of remnant populations and reduced N_e to critically low values over ecological time.     

Genetic Diversity in a Threatened Wetland Species, Helonias bullata (Liliaceae)

Genetic variation was examined in Helonias bullata , a threatened perennial plant species that occurs in isolated wetland habitats. Fifteen populations representing the species' geographic range were sampled. Genetic diversity was low for the species ( H _es = 0.053) as well as within populations ( H _ep = 0.029). Of the 33 allozyme loci examined, 11 (33%) were polymorphic, while on average only 12.8% (4) of the loci were polymorphic within populations. The number of alleles per polymorphic locus was 2.36 for the species and averaged 2.09 across populations. For every genetic parameter calculated, variation in H. bullata was lower than that typically found for narrowly distributed plant species. The lowest levels of genetic diversity were found in northern areas that were colonized following the last glacial epoch. The number of genotypes detected per population ranged from three to 21, with a mean of 13 for this clonally reproducing species. We found a relatively high proportion of total genetic diversity (30.6%) among populations and a significant correlation (p < 0.002) between genetic distance and geographic distance. Genetic drift phenomena appear to play a major role in the population genetics of this species. Anomalously, several populations that appeared most limited in size and vigor were genetically most variable, perhaps because they represent older, relictual populations. Life-history characteristics of H. bullata coupled with low levels of genetic diversity and the degradation and disappearance of wetlands threaten the existence of this species.