Bot. Bull. Acad. Sin. (2005) 46: 61-69

QIU et al. Genetic variation in Dysosma versipellis

A preliminary study of genetic variation in the endangered, Chinese endemic species Dysosma versipellis (Berberidaceae)

Ying-Xiong QIU1, Xin-Wen ZHOU1, Cheng-Xin FU1,*, and Yuk-Sing Gilbert CHAN2

1Laboratory of Systematic and Evolutionary Botany, Department of Biology, College of Life Sciences, Zhejiang University, Hangzhou 310029, P. R. China

2Department of Applied Biology and Chemical Technology, The Hong Kong Polytechnic University, Hung Hom, Hong Kong

(Received December 23, 2003; Accepted September 30, 2004)

Abstract. This study represents a preliminary analysis of allozyme variation in Dysosma versipellis (Berberidaceae), an endangered plant species endemic to China. Five populations of D. versipellis and one population of D. pleiantha were sampled and analyzed using starch gel electrophoresis of nine enzymes that corresponded to nine interpretable loci. Levels of genetic polymorphism within populations (means: P = 15.54%, A = 1.16, He = 0.045) were much smaller than values for seed plants in general (P = 34.2%, A = 1.53, He = 0.113), as well as values for other endemic species (P = 26.3%, A = 1.39, He = 0.063). Mean values for the FST across all D. versipellis populations tended to be high (FST = 0.468). An indirect estimate of the number of migrants per generation (Nm = 0.284) indicated that gene flow is low among populations of D. versipellis. Additionally, analysis of genetic variation revealed a substantial heterozygosity deficiency in all analyzed populations except HB. Genotype frequencies within D. versipellis populations indicate that they may be severely inbred, making inbreeding depression a possible explanation for the low seed set observed in this species. Likewise, the low level of genetic diversity observed within D. versipellis populations suggests that clonal reproduction might be more important than sexual reproduction for D. versipellis. In comparison, genetic variation observed in one population of the closely related species D. pleiantha was much higher than the variation within D. versipellis populations. On the basis of these observations, we suggest that in situ conservation will be an important and practical measure for maintaining this species. If ex situ conservation is pursued, sampling should cover all populations across the species' distribution so as to retain as much genetic diversity as possible.

Keywords: Allozyme; Asexual reproduction; Dsma versipellis; Endangered species; Genetic variation.


Plants in the genus Dysosma (Berberidaceae), with seven species, occur solely in China. Populations of the endangered species D. versipellis, an herbaceous perennial species that grows in the understory of mixed evergreen and deciduous forests (Ying et al., 1993), have a further restricted distribution in the East and South China. Dysosma pleiantha is another rare species in the genus, but little is known about its biology. However, it is clear that these two species have non-overlapping distributions and differ substantially in both leaf and flower morphology (Ying et al., 1993). These observations suggest that the breeding systems of some Dysosma species may have undergone divergent evolution resulting from adaptation to different ecological conditions. Dysosma versipellis reproduces sexually through cross-pollination and asexually by spreading rhizomes. Field observations indicates that although the plants flower almost every year, seed production is limited, and seedling establishment is poor.

Propagation, therefore, appears to occur mostly by vegetative means (Ma, 2000; Qiu and Qiu, 2002). In recent years, natural populations of this species have declined considerably due to anthropogenic activities like habitat destruction and overcollecting for medicinal applications. The rhizome of Dysosma has been found to be a source of podophyllotoxin, the active ingredient used as a starting compound for the chemical synthesis of etoposide (VP-16-213) and teniposide (VM-26), effective agents in the treatment of lung cancer, a variety of leukemias, and other solid tumor diseases (Jackson and Dewick, 1984, 1985). Nearly all remaining populations of D. versipellis are now located within protected nature reserves, and the species is classified as endangered in the Chinese Plant Red Book (Fu, 1992). Because the species is endangered, and harvesting of rhizomes continues to exceed the rate of natural regeneration, immediate attention should be given to conserving the species through in situ and/or ex situ approaches.

Resource managers responsible for the conservation and recovery of threatened plant species frequently need to make choices regarding which areas are in need of greatest protection, which populations should have priority for

*Corresponding author. Tel: +86-571-86971576; Fax: 86-571-86432273; E-mail:

Botanical Bulletin of Academia Sinica, Vol. 46, 2005

preservation, what the minimum number of individuals required to avoid inbreeding and to sustain genetic variation is, as well as how germplasm should be collected from a population to capture representative genetic diversity for ex situ conservation (Torres et al., 2003). Thus, understanding how genetic variation is distributed among individuals and populations must be considered together with evolutionary history, population dynamics, breeding system, and species/population structure in evaluating the critical attributes of rare species, and such knowledge is of paramount importance for developing recovery plans that can meaningfully sample and preserve genetic diversity (Falk and Holsinger, 1991). To our knowledge, this study represents the first attempt to characterize the genetic diversity within and among populations of D. versipellis.

Among the many molecular techniques available for evaluating plant genetic variation, allozyme electrophoresis is more than adequate to address many conservation genetics issues. This technique is amenable for conservation genetic surveys because data can be obtained quickly for many individuals. The approach involves analysis of single-gene codominant molecular markers (loci) that are biparentally inherited and easily assessed by visualizing band patterns and intensities on starch electrophoretic gels. This approach has made it possible to better understand the spatial distribution of clones and genetic diversity maintained within and among plant populations (e.g., Parker and Hamrick, 1992; Berg and Hamrick, 1994).

In the present study, we used enzyme electrophoresis to assess whether patterns of genetic diversity are consistent with clonal structure and low levels of sexual reproduction observed across D. versipellis populations in the field. Specifically, we addressed the following questions: (1) Are D. versipellis populations typically composed of one or multiple genets? (2) What are the levels and distribution of genetic variability within and among populations of D. versipellis? (3) What are the possible factors that might explain the patterns and levels of genetic variation observed? Additionally, we present measures of allozyme variation within and among populations of D. versipellis, and compare these to data published for other plant taxa with similar characteristics (Ellstrand and Roose, 1987; Hamrick and Godt, 1990). It is our hope that such information will be used by conservation biologists to formulate effective management strategies based on genetic data for the conservation of this endangered species.

Materials and Methods

Plant Material and Population Sampling

Dysosma versipellis is a perennial rhizomatous herb that undergoes both sexual and asexual reproduction. Individual plants grow from rhizome and typically reach 40 to 150 cm in height, with an unforked stem bearing one or two alternately arranged leaves, which are rounded, centrally peltate, and 4-9-lobed with finely dentate margin. Plants remain in a juvenile phase for 4-5 years. When

mature, plant sends out a shoot in early March, and a terminal cyme emerges on the shoot in April with 5 to 8 drooping red-purple flowers. Fruits mature in late June (Figure 1) (Ying et al., 1993; Qiu and Qiu, 2002).

Dysosma versipellis occurs in the undergrowth of subtropical forests between 200 and 3,500 m elevation, and is distributed across Sichuan, Anhui, Hunan, Guanxi, Jianxi and Hubei Provinces. From 1998 to 2002 several extensive collecting trips for D. versipellis were conducted across China. The sampling included five fragmented populations located across the species entire range. Samples from one population of D. pleiantha were also collected (Figure 2). Due to the different sizes of the populations, the number of samples collected ranged from 12 to 22. The distance between adjacent samples was at least 10 m to increase the likelihood of sampling non-clonal, inter-individual variation within each population. Samples were taken from the edges, as well as the interior of populations. A total of 100 young leaf samples were taken from randomly chosen putative individuals. The leaf material was transported in a hand cooler to the laboratory, where it was stored at -80C.

Figure 1. Dysosma versipellis. a, rhizome and roots; b, leaf; c, cyme; d, sepals; e, petal; f, fruit.

QIU et al. Genetic variation in Dysosma versipellis

polymorphic loci (P) (95% criterion) per subpopulation, the observed heterozygosity (Ho), and the mean expected heterozygosity (He). Differences in both population and sample size confound direct comparison of the genetic diversity in this sample site with that of other sites. To more directly compare levels of genetic diversity, we controlled for differences in sample size by randomly choosing twelve individuals from each site and calculating all genetic diversity summary statistics on this reduced data set. This comparison allowed us to determine whether the observed levels of diversity in the small population were comparable to those found in equivalent samples from larger populations.

Wright's F [F = (1- Ho / He)], the inbreeding coefficient, measures the deviation of population genotypic composition from Hardy-Weinberg (H-W) expectations. The inbreeding coefficient was calculated at each polymorphic locus and tested for significant deviation using chi-square tests (Li and Horvitz, 1953). The average fixation indices were also calculated for each population and tested for significant difference from zero.

The partition of total genetic diversity into within- and among-population components was examined using Nei's (1973, 1978) genetic diversity statistics. For each polymorphic locus, total genetic diversity (HT) was partitioned into diversity within populations (HS) and diversity among populations (and DST) as HT = HS + DST. A measure of genetic differentiation among populations relative to the total genetic diversity (GST) was calculated at each polymorphic locus (GST = DST / HT). The genetic structure within and among populations was also evaluated using Wright's (1965) F-statistics: FIT, FIS and FST. The FIT and FIS coefficients measure excesses of homozygotes or heterozygotes relative to the panmictic expectations within the entire samples and within populations, respectively. The FST coefficient estimates relative population differentiation. Deviation of FIT, FIS and FST from zero were tested using chi-square tests (Li and Horvitz, 1953; Workman and Niswander, 1970). A rough estimation of the quantity Nm (N = population size, m = migration rate) was also estimated using Wright's (1951) formula FST = 1 / (1+4Nm).

To examine the genetic relationship among populations, a dendrogram was constructed using a UPGMA analysis as implemented by NTSYS-pc, Version 2.02c (Rohlf, 1997). In order to test for a correlation between genetic distances (GD) and geographical distances (measured in km) among populations, a Mantel test was performed using the program TFPGA (Miller, 1997). The null hypothesis refers to the absence of association between the elements of the pairs of matrices. A normalized Z test was performed in which the observed value after 1000 permutations should be significantly larger than that expected by chance in order for an association to be accepted as valid.

Clonal diversity (Fager, 1972; Ellstrand and Roose, 1987) was evaluated by the following indices: (1) number of genotypes, G; (2) the mean clone size, Nc = N / G, where N represents the sample size; (3) proportion of distinguishable genotypes, PD = G / N ; (4) a modified version of the

Figure 2. Location of the five natural populations of Dysosma versipellis and one population of Dysosma pleiantha (ZJ) included in this study. SC = Sichun population; HB = Hubei population; HN = Hunan population; AH = Aihui population; JX = Jiangxi population; and ZJ = Zhejiang population.

Enzyme Extraction and Electrophoresis

Leaves were ground in cold extraction buffer consisting of 0.2 M Tris-HCl (pH = 7.5), 2 mM EDTA, 0.12 M Na2S2O5, 1 M MgCl2, 80 mg/mL, PVP, and 40 L/mL mercaptoethanol. Extracts were absorbed onto Whatman No. 1 filter papers subjected to electrophoresis on horizontal 12.5% starch gels. Enzymes were resolved and scored by starch-gel electrophoresis. The electrophoresis methods followed those of Soltis et al. (1983). After initially screening 15 enzymes, nine were chosen for further analysis. Alcoholdehydrogenase (Adh), isocitrate dehydrogenase (Idh) and malic enzyme (Me) were resolved on buffer system 6 (S11) (Soltis et al., 1983) while phosphoglucoisomerase (Pgi), nadh-diaphorase (Dia), shikimate dehydrogenase (Skd), 6-phosphogluconate dehydrogenase (Pgd), triose-phosphate isomerase (Tpi) and phosphoglucomutase (Pgm) were resolved on buffer system 12 (S12) according to Soltis et al. (1983). After electrophoresis, gels were sliced into layers for staining following the protocols reported by Weeden and Wendel (1989) with slight modifications. Putative loci were designated sequentially. (The most anodally migrating isozyme was designated "1", the next "2", and so on.) Similarly, allelic variation at a locus was coded alphabetically, with the most anodally migrating allozyme designated as allele "a".

Data Analysis

Allozyme diversity was estimated for each subpopulation using BIOSYS-1, Version 1.7 (Swofford and Selander, 1989). Data were entered as genotype numbers from which allele frequencies were calculated. Unbiased estimates of Nei's genetic identity (I) and genetic distance (GD) (Nei, 1972) were calculated. The levels of genetic variability within subpopulations were estimated using four variables: the mean number of alleles (A) per locus, proportion of

Botanical Bulletin of Academia Sinica, Vol. 46, 2005

Simpson diversity index, D = 1- [Ni(Ni - 1) / N(N - 1)] , where Ni is the number of samples of the ith genotype; and (5) a Fager index, E = (D - Dmin) / (Dmax - Dmin), where Dmin = (G - 1)(2N - G) / N(N - 1) and Dmax = (G - 1) N / G (N - 1). The Simpson diversity index (D) was originally developed as a measure of species diversity and has been employed to measure clonal diversity within populations (Parker, 1979; Ellstrand and Roose, 1987). The D value ranges from 0 to 1, where 0 reflects all individuals containing the same multilocus genotype, and 1 is for each individual having a unique multilocus genotype. In addition, Fager's E value describes the evenness of distribution of genotypes within a population and varies between 0 (when all individuals in a population possess the same genotype) and 1 (when a population has completely uniform genotype frequencies).


Within-Population Variation and Population-level Homozygosity

Of the nine loci resolved, seven were polymorphic across the range of D. versipellis. Percentage of polymorphic loci at the population level for D. versipellis ranged from 0.0% at JX, to 33.3% at AH, with a mean of 15.54% across populations. The mean number of alleles (A)

per locus ranged from 1.0 at JX, to 1.3 at AH for D. versipellis. The mean within population genetic diversity (He) was 0.045. Population AH had the highest expected diversity (0.061) while Population JX had the lowest (0.000) (Table 1). Genetic variation within the one sampled population of D. pleiantha (means: P = 55.60%, A = 1.80, He = 0.208) was much higher than that of examined populations of D. versipellis. Population level values for D. versipellis (means: P = 15.54%, A = 1.16, He = 0.045) were much smaller than average values measured for other seed plants (P = 34.2%, A = 1.53, He = 0.113), as well as for other endemic species (P = 26.3%, A = 1.39, He =0.063) (see Hamrick and Godt, 1990). Across all populations, the mean number of alleles per locus for D. versipellis was 1.80 (N = 84), and the percentage of polymorphic loci (95% criterion) was 77.78%. Since the genetic diversity results may correlate with sample size of a population, an equal number of individual (twelve) randomly selected from each population was analyzed. When N = 12 in all populations, the average value of P was 15.54%, A was 1.14, He was 0.047, and Ho was 0.022. Results showed that the values of A and Ho obtained from the twelve random individuals of each population were a little lower and He a little higher than the original ones, but the trend was the same (Table 1). Thus, the apparently high levels of diversity at AH population (Table 1) appear not to be the result of a large sample size.

QIU et al. Genetic variation in Dysosma versipellis

In general, observed genotype frequencies were significantly different from H-W expectations. Indeed, of eight inbreeding coefficients calculated, only one was not significantly different from zero. Such results were obtained from locus Skd in population HB, which showed a negative but insignificant F index (F = -0.429; P = 0.119). Accordingly, the average fixation index (F = 0.467) is significantly higher than zero for the analyzed populations, except for HB (Table 1).

Clonal Diversity

The clonal diversity and evenness of genotype distributions of the five populations of D. versipellis are summarized in Table 2. Ten genotypes were identified in 84 individuals of D. versipellis. Four populations were comprised of more than one genet while the JX population appeared to consist of only one. The largest number of genets, five, was found in the AH population. Among the five populations of D. versipellis, average clone sizes ranged from 3.00 (AH) to 15.00 (JX), and the mean values of D and E were 0.458 and 0.528, respectively (Table 2).

Distribution of Genetic Variation and Population Divergence

The estimates of population genetic structure using Nei's genetic diversity statistics are shown in Table 3. The average of total heterozygosity (HT) and intrapopulation genetic diversity (HS) were 0.083 and 0.046, respectively. The coefficient of genetic differentiation among population (GST) varied from 0.004 (Idh) to 0.668 (Dia), with a mean of 0.445. The results indicated that 44.5% of total genetic diversity is among populations, 55.5% representing intrapopulation genetic diversity. In the five populations of D. versipellis studied here, the mean overall FIS of 0.455 was statistically different from zero, suggesting that most of the populations deviate from Hardy-Weinberg expectation within populations (Table 3). The mean overall FIT of 0.710 was also statistically different from zero, indicating nonequilibrium conditions across populations of the species and a deficiency of heterozygotes. The mean FST over all D. versipellis populations was 0.468. The D. pleiantha population (from ZJ) differs considerably from all D. versipellis populations, as depicted in the UPGMA clus

Botanical Bulletin of Academia Sinica, Vol. 46, 2005

ter analysis (Table 4; Figure 3). The mean genetic identity between the two species was 0.885 (Nei's genetic distance = 0.115). No significant correlation was found between genetic distance and geographic distance (r=0.155; P = 0.652) based on the Mantel test. The one-tail probability, P coefficient [random Z observed Z], indicated that the null hypothesis can not be rejected, suggesting no clear geographical pattern of isolation-by-distance in the distribution of the species' genetic variability.


As it has been well documented that breeding system and geographic range are good predictors of genetic parameters at the population level (Hamrick and Godt, 1990), it is useful to compare the level of genetic variation in D. versipellis to the levels documented for other seed plants with a similar life history, geographic range, and breeding system. The percentage of polymorphic loci (15.54%) found in D. versipellis is lower than what was reported for other endemic species (26%) and predominantly clonal species (29%). Similarly, the genetic diversity value (He = 0.045) is much lower than the mean value calculated by Hamrick and Godt (1990) for 338 species of dicotyledon

ous plants (He = 0.09) and 56 species of sexual and asexual taxa (He = 0.10). The mean "proportion of distinguishable genotypes" per population for D. versipellis was 0.119, which is also lower than the value of 0.17 for the 21 species summarized by Ellstrand and Roose (1987), and well below the mean of 0.27 for 45 species reported by Widen et al. (1994). However, it is quite similar to the mean value of 0.13 for asexual species in which sexual reproduction is uncommon (Widen et al., 1994).

In addition, the fixation indices for the sampled D. versipellis populations indicate that most deviate from Hardy-Weinberg equilibrium, and that there is a substantial deficiency of heterozygotes. The high FIS levels in the species probably indicate some selfing and/or intra-clone outcrossing events caused by clonal reproduction and limited seedling recruitment. The species is self-incompatible, and insect pollinators tend to visit adjacent flowers (Qiu and Qiu, 2002). Fruits and developing seeds resulting from these pollinations may abort at various stages of development as a result of self-incompatibility in this taxon. Limited seedling recruitment undoubtedly also contributes to the maintenance of low within-population heterozygosity in D. versipellis. Further loss of genetic variability may also result from succession because of reduced seedling establishment combined with the elimination of genotypes via competition, poor adaptation, or stochastic events (McNeilly and Roose, 1984).

Clonal diversity and genetic structure in populations of clonal species varies greatly. Some endangered plant species, especially those with little or no sexual reproduction, such as Taraxacum obliquum (Van Oostrum et al., 1985; Ellstrand and Roose, 1987) and Haloragodendron lucasii (Sydes and Peakall, 1998), have only one or a few genets across all populations. For other species, the number of genets varies greatly among populations, with some populations consisting of only one genet and others supporting many (Aspinwall and Christian, 1992; Eckert and Barrett, 1993; Ayres and Ryan, 1997). In asexual populations, genotypic diversity is expected to reach a climax over time until a single genotype reaches fixation (Parker, 1979). Dysosma versipellis is described as self-incompatible, and Ma (2000) believed that colonies in the wild may derive from a single seedling, which corresponds to a single genotype growing in clonal patches. However, contrasting this view, only the JX popu

Figure 3. UPGMA dendrogram showing relationships among five populations of Dysosma versipellis (HB-HN) and one of Dysosma pleiantha (ZJ) based on Nei's unbiased genetic distances (GD).

QIU et al. Genetic variation in Dysosma versipellis

the loss of genetic diversity and alteration of population genetic structure (Ellstand and Elam, 1993). Overall, the low allozyme diversity documented for this species is likely a result of habitat fragmentation followed by genetic drift and limited gene flow among small populations.

Continued exploitation of wild individuals for the traditional Chinese medicine trade remains an important factor affecting the rarity and endangered status of the species. Since low clone diversity and large clone sizes indicate that vegetative reproduction is more important than sexual reproduction for D. versipellis, the species may be able to survive in the wild despite being collected by humans, as long as harvesting is managed in a sustainable manner and habitats are preserved. In this regard, in situ conservation and regular monitoring of natural populations will be an important strategy for conserving this species. If ex situ strategies are developed and pursued, we recommend that sampling include all known localities across the species' distribution, since the species now occurs in fragmented and genetically distinct populations.

Acknowledgements. The authors thank Dr. Zhou Shiliang (Institute of Botany, CAS) for assistance in conducting the electrophoresis; Dr. Jin Xiaofeng for preparation of drawing. We are also grateful to Dr. Kenneth M. Cameron, Dr. Joongku Lee, and two anonymous reviewers for valuable comments on the manuscript. This work was supported by the State Key Basic Research and Development Plan of China (G2000046806), and Zhejiang Provincial Natural Science Foundation of China (No. M303092).

Literature Cited

Aspinwall, N. and T. Christian. 1992. Clonal structure, genotypi diversity, and seed production in populations of Filipendula rubra (Rosaceae) from the northcentral United States. Am. J. Bot. 79: 294-299.

Ayres, D.R. and F.J. Ryan. 1997. The clonal and population strture of a rare endemic plant, Wyethia reticulata (Asteraceae): allozyme and RAPD analysis. Mol. Ecol. 6: 761-772.

Berg, E.E. and J.L. Hamrick. 1994. Spatial and genetic structure of two sandhills oaks: Quercus laevis and Q. margaretta (Fagaceae). Am. J. Bot. 81: 7-14.

Eckert, C.G. and S.C.H. Barrett. 1993. Clonal reproduction and patterns of genotypic diversity in Decodon verticillatus (Lythraceae). Am. J. Bot. 80: 1175-1182.

Ellstrand, N.C. and D.R. Elam. 1993. Population genetic consequences of small population size: Implications for plant conservation. Annu. Rev. Ecol. Syst. 24: 217-225.

Ellstrand, N.C. and M.L. Roose. 1987. Patterns of genotypic diversity in clonal plant species. Am. J. Bot. 74: 123-131.

Eriksson, O. and B. Bremer. 1993. Genet dynamics of the clonal plant Rubus saxatilis. J. Ecol. 81: 533-542.

Fager, E.W. 1972. Diversity a sampling study. Am. Nat. 106: 293-310.

Falk, D. and K. Holsinger. 1991. Genetics and Conservation of Rare Plants. Oxford University Press, New York.

Fischer, M. and D. Matthies. 1998. RAPD variation in relation

lation of D. versipellis comprised one genet while others appeared to consist of several genets in this study. Plausible explanations for such results could be (a) mutation; (b) occasional occurrence of sexual reproduction resulting in unique genotypes; or (c) differential selection in a spatially heterogenous environment (Ellstrand and Roose, 1987). Mutation is unlikely to play an important role in maintaining genetic variation in D. versipellis owing to low frequency of mutations. Occasional occurrence of sexual reproduction leading to seed set may contribute towards maintaining genetic diversity in some populations of D. versipellis. Overall, the low level of genetic diversity in D. versipellis provides evidence of a bias towards asexual reproduction although some sexual reproduction apparently does take place.

For species that reproduce sexually, populations usually contain many genets (Ellstrand and Roose, 1987; Eriksson and Bremer, 1993). The ZJ population of D. pleiantha that we examined appears to consist of multiple genets. The relatively high clone diversity and small clone sizes of D. pleiantha indicate that sexual reproduction is likely more important than clonal reproduction in this particular species; indeed, the high fruit production observed in natural populations of D. pleiantha confirms this hypothesis.

Further comparison of the observed genetic variation among D. versipellis populations with values reported for other plant taxa (reviewed by Hamrick and Godt, 1990) indicates that D. versipellis populations are highly differentiated. As stated earlier, the population genetic differentiation in the present study (GST = 0.455) is much higher than the average for perennial-herbaceous plants (GST = 0.233) and endemic plants (GST = 0.25). A high level of population differentiation may be explained by several factors, including the species' breeding system, genetic drift, or geographic isolation of populations (Hogbin and Peakall, 1999). Low levels or absence of gene flow among populations is also characteristic of many rare species (Slatkin, 1985), and a number of studies have documented high levels of genetic differentiation among populations of rare species (Fischer and Matthies, 1998; Sun and Wong, 2001). Nm, the number of migrants per generation (Nm = 0.284), for D. versipellis was less than one successful migrant per generation, indicating limited gene flow among populations. This value corresponds well with the frequent inbreeding, considerable clonal reproduction, and isolated nature of D. versipellis populations. However, it is noteworthy that genetic differentiation among populations of D. versipellis does not appear to be correlated with geographic distance among populations.

The relatively low genetic diversity and high genetic differentiation documented for D. versipellis are most likely a result of habitat fragmentation (Qiu and Qiu, 2002). The subtropical forests necessary for the survival of this species have been destroyed and disturbed by anthropogenic activity at an alarming rate during the past century. As a result, the numbers and sizes of extant D. versipellis populations have decreased greatly, which in turn has led to

Botanical Bulletin of Academia Sinica, Vol. 46, 2005

to population size and plant fitness in the rare Gentianella germanica (Gentianaceae). Am. J. Bot. 85: 811-819.

Fu, L.G. 1992. Chinese Plant Red Book. Science Press, Beijing, China.

Hamrick, J.L. and M.J.W. Godt. 1990. Allozyme diversity in plant species. In A.H.D. Brown, M.T. Clegg, A.L. Kahler, and B.S. Weir (eds.), Plant Population Genetics, Breeding, and Genetic Resources. Sinauer, Sunderland, Massachusetts USA, pp. 43-63.

Hogbin, P.M. and R. Peakall. 1999. Evaluation of the conservation of genetic research to the management of endangered plant Zieria prostrata. Conserv. Biol. 13: 514-522.

Jackson, D.E. and P.M. Dewick. 1984. Aryltetralin lignans from Podophyllum hexandrum and Podophyllum peltatum (isolated from the roots). Phytochemistry 23: 1147-1152.

Jackson, D.E. and P.M. Dewick. 1985. Tumor-inhibitory aryltetralin lignans from Podophyllum pleianthum. Phytochemistry 24: 2407-2409.

Li, C.C. and D.G. Horvitz. 1953. Some methods of estimating the inbreeding coefficient. Am. J. Hum. Genet. 5: 107-117.

Miller, M.P. 1997. Tools for Population Genetic Analysis (TEPGA), Version 3. Department of Biological Sciences, Northern Arizona University, Arizona, USA.

Ma, S.B. 2000. A contribution to the reproductive ecology of Dysosma veitchii. Acta Phytoecol. Sin. 24: 748-753.

Nei, M. 1972. Genetic distance between populations. Am. Nat. 106: 283-292.

Nei, M. 1973. Analysis of gene diversity in subdivided populations. Proc. Natl. Acad. Sci. 70: 3321-3323.

Nei, M. 1978. Estimation of average heterozygosity and genetic distance from a small number of individuals. Genetics 89: 583-590.

McNeilly, T. and M.L.Roose. 1984. The distribution of perennial ryegrass genotypes in swards. New Phytol. 98: 503-513.

Parker, E.D. 1979. Ecological implications of clonal diversity in parthenogenetic morphospecies. Am. Zool. 19: 753-762.

Parker, K.C. and J.L. Hamrick. 1992. Genetic diversity and clonal structure in a columnar cactus, Lophocereus schottii. Amer. J. Bot. 79: 86-96.

Qiu, H.H., and Y.X. Qiu. 2002. Advances in research on the endemic and endangered plant of Dysosma in China and its exploitation prospects. J. Anqing Teachers College (Natural Science) 8: 91-94.

Rohlf, F.J. 1997. NTSYS-pc: Numerical Taxonomy and Multivariate Analysis System, Ver. 2.02. Exeter Ltd, Setauket, NY, USA.

Slatkin, M. 1985. Gene flow in natural populations. Annu. Rev. Ecol. Syst. 16: 393-430.

Soltis, D.E., C.H. Haufler, D.C. Darrow, and G.J. Gastony. 1983. Starch gel electrophoresis of ferns; A compilation of grinding buffers, and staining schedules.

Sun, M. and K.C. Wong. 2001. Genetic structure of three orchid spcies with contrasting breeding systems using RAPD and allozyme markers. Am. J. Bot. 88: 2180-2189.

Swofford, D.L. and R.B. Selander. 1989. BOSYS-1, A Computer Program for the Analysis of Allelic Variation in Population Genetics and Biochemical Systematics. Release 1.7. University of Illinois, Urbana, IL.

Sydes, M.A. and R. Peakall. 1998. Etensive clonlity in the endangered shrub Haloragodendron lucasii (Haloragaceae) revealed by allozymes and RAPDs. Mol. Ecol. 7: 87-93.

Torres, E., J.M. Iriondo, and C. Prez. 2003. Genetic structure of an endangered plant, Antirrhinum microphyllum (Scrophulariaceae): allozyme and RAPD analysis. Am. J. Bot. 90: 85-92.

Van Oostrum, H., A.A. Sterk, and H.J.W. Wusman. 1985. Genetic variation in agamospermous microspecies of Taraxacum sect. Erythrosperma and sect. Obliqua. Heredity 55: 223-228.

Weeden, N.F. and J.F. Wendel. 1989. Genetics of plant isozymes. In D.E. Soltis and P.S. Soltis (eds.), Isozymes in Plant Biology, Dioscorides. Portland, Oregon, pp. 46-72.

Widen, B., N. Cronberg, and M. Widen. 1994. Genotypic diversity, molecular markers and spatial distribution of genets in clonal plants, a literature survey. Folia Geobot. Phytoaxon. 29: 245-263.

Wright, S. 1951. The genetic structure of populations. Ann. Eugen. 15: 313-354.

Workman, P.L. and J.D. Niswander. 1970. Population studies on southwestern Indian tribes. II. Local genetic differentiation in the Papago. Amer. J. Human Genet. 22: 24-49.

Wright, S. 1965. The interpretation of genetic population structure by F-statistics with special regard to systems of mating. Evolution 19: 395-420.

Ying, T.S., Y.L. Zhang, and D.E. Boufford. 1993. The Endemic Genera of Seed Plants of China. Science Press, Beijing.

QIU et al. Genetic variation in Dysosma versipellis