Electrophoretic examination of the population structure of brown trout, Salmo trutta L., from the Lough Neagh catchment, Northern Ireland

Samples of brown trout, Salmo trutia L., from 34 locations throughout the Lough Neagh system in north-east Ireland were electrophoretically examined for genetic variation at 28 enzyme loci. Patterns of allelic variation at 12 polymorphic loci indicated the existence of genetic differentiation within as well as among several river systems, suggesting the existence of multiple brown trout populations. Significant gene frequency differences were detected over distances as little as 3 km, demonstrating the propensity of this species for microgeographic genetic differentiation. This was confirmed by a hierarchical analysis of genetic variance, some 35% of among-sample variance being distributed within tributaries. Within Lough Neagh itself significant genetic differentiation was detected between two morphotypes (dollaghan and salmon-trout) and for one of these (dollaghan) among samples from different years and from different areas of the lough. This suggests the existence of genetically differentiated subpopulations of originating from separate river catchments.     

Population structure of loggerhead shrikes in the California Channel Islands

The loggerhead shrike (Lanius ludovicianus), a songbird that hunts like a small raptor, maintains breeding populations on seven of the eight California Channel Islands. One of the two subspecies, L. l. anthonyi, was described as having breeding populations on six of the islands while a second subspecies, L. l. mearnsi, was described as being endemic to San Clemente Island. Previous genetic studies have demonstrated that the San Clemente Island loggerhead shrike is well differentiated genetically from both L. l. anthonyi and mainland populations, despite the fact that birds from outside the population are regular visitors to the island. Those studies, however, did not include a comparison between San Clemente Island shrikes and the breeding population on Santa Catalina Island, the closest island to San Clemente. Here we use mitochondrial control region sequences and nuclear microsatellites to investigate the population structure of loggerhead shrikes in the Channel Islands. We confirm the genetic distinctiveness of the San Clemente Island loggerhead shrike and, using Bayesian clustering analysis, demonstrate the presence and infer the source of the nonbreeding visitors. Our results indicate that Channel Island loggerhead shrikes comprise three distinct genetic clusters that inhabit: (i) San Clemente Island, (ii) Santa Catalina Island and (iii) the Northern Channel Islands and nearby mainland; they do not support a recent suggestion that all Channel Island loggerhead shrikes should be managed as a single entity.     

Adaptive divergence despite strong genetic drift: genomic analysis of the evolutionary mechanisms causing genetic differentiation in the island fox (Urocyon littoralis)

The evolutionary mechanisms generating the tremendous biodiversity of islands have long fascinated evolutionary biologists. Genetic drift and divergent selection are predicted to be strong on islands and both could drive population divergence and speciation. Alternatively, strong genetic drift may preclude adaptation. We conducted a genomic analysis to test the roles of genetic drift and divergent selection in causing genetic differentiation among populations of the island fox (Urocyon littoralis). This species consists of six subspecies, each of which occupies a different California Channel Island. Analysis of 5293 SNP loci generated using Restriction‐site Associated DNA (RAD) sequencing found support for genetic drift as the dominant evolutionary mechanism driving population divergence among island fox populations. In particular, populations had exceptionally low genetic variation, small N_e (range = 2.1–89.7; median = 19.4), and significant genetic signatures of bottlenecks. Moreover, islands with the lowest genetic variation (and, by inference, the strongest historical genetic drift) were most genetically differentiated from mainland grey foxes, and vice versa, indicating genetic drift drives genome‐wide divergence. Nonetheless, outlier tests identified 3.6–6.6% of loci as high F_ST outliers, suggesting that despite strong genetic drift, divergent selection contributes to population divergence. Patterns of similarity among populations based on high F_ST outliers mirrored patterns based on morphology, providing additional evidence that outliers reflect adaptive divergence. Extremely low genetic variation and small N_e in some island fox populations, particularly on San Nicolas Island, suggest that they may be vulnerable to fixation of deleterious alleles, decreased fitness and reduced adaptive potential.     

MICROGEOGRAPHIC PATTERNS OF GENETIC AND MORPHOLOGICAL VARIATION IN SAVANNAH SPARROWS (PASSERCULUS SANDWICHENSIS)

Surveys of genetic population structure are often limited to large geographic scales because geographically close populations are indistinguishable. Genetic uniformity across adjacent demes can be interpreted as evidence for cohesion (panmixia) or recent divergence. However, poor genetic resolution at microgeographic scales can also arise from the use of overly conservative (slowly evolving) markers. This study examines the ability of hypervariable, minisatellite loci to discriminate among geographically close populations of Savannah sparrows (Passerculus sandwichensis) and to track morphological differentiation at a microgeographic scale (interregional distance < 55 km). Savannah sparrows breeding at five island and two mainland sites in the Bay of Fundy, New Brunswick, Canada, show concordant patterns of variation in external morphology (seven characters) and multilocus DNA fingerprinting profiles (S_xy): island sparrows are phenotypically larger and genetically more similar to each other than they are to mainland sparrows. This pattern of variation is consistent with both adaptive (natural selection) and nonadaptive (genetic drift) mechanisms of population divergence. Based on minisatellite diversity, the effective size of both island and mainland populations is 37, an estimate substantially lower than census population sizes. These data are discordant with observations of sparrow vagility and abundance and suggest a closer examination of microgeographic patterns in avian systems.     

Population Genetic Structure and Conservation of the Galápagos Petrel (Pterodroma phaeopygia)

The Galápagos petrel (Pterodroma phaeopygia) is endemic to the Galápagos archipelago, where it is known to breed only on five islands. The species has been listed as critically endangered due to habitat deterioration and predation by introduced mammals. Significant morphological and behavioural differences among petrels nesting on different islands suggest that island populations may differ genetically. Furthermore, nesting phenology suggests that genetically differentiated seasonal populations may exist within at least one island. We analysed variation in six microsatellite loci and part of the mitochondrial ATPase 6/8 gene in 206 Galápagos petrels sampled from all five islands. No evidence of genetic structuring within islands was found, although statistical power was low. In contrast, significant differences occurred among island populations. For the microsatellite loci, private alleles occurred at all islands, sometimes at high frequency; global and pairwise estimates of genetic differentiation were all statistically significant; Bayesian analysis of genotypes frequencies provided strong support for three genetic populations; and most estimates of migration between populations did not differ significantly from zero. Only two ATPase haplotypes were found, but the geographic distribution of haplotypes indicated significant differentiation among populations. For conservation purposes, populations from Floreana, Santa Cruz, San Cristóbal and Santiago should be regarded as separate genetic management units. Birds from Isabela appear to be derived recently from the Santiago population, and the population on San Cristóbal appears to be a mixture of birds from other populations. However, considering ecological and behavioural differences among birds from different islands, we recommend that all five populations be protected.     

Population structure and genetic diversity of trout (<Emphasis Type="Italic">Oncorhynchus mykiss</Emphasis>) above and below natural and man-made barriers in the Russian River,California

The effects of landscape features on gene flow in threatened and endangered species play an important role in influencing the genetic structure of populations. We examined genetic variation of trout in the species Oncorhynchus mykiss at 22 microsatellite loci from 20 sites in the Russian River basin in central California. We assessed relative patterns of genetic structure and variation in fish from above and below both natural (waterfalls) and man-made (dams) barriers. Additionally, we compared sites sampled in the Russian River with sites from 16 other coastal watersheds in California. Genetic variation among the 20 sites sampled within the Russian River was significantly partitioned into six groups above natural barriers and one group consisting of all below barrier and above dam sites. Although the below-barrier sites showed moderate gene flow, we found some support for sub-population differentiation of individual tributaries in the watershed. Genetic variation at all below-barrier sites was high compared to above-barrier sites. Fish above dams were similar to those from below-barrier sites and had similar levels of genetic diversity, indicating they have not been isolated very long from below-barrier populations. Population samples from above natural barriers were highly divergent, with large F _st values, and had significantly lower genetic diversity, indicating relatively small population sizes. The origins of populations above natural barriers could not be ascertained by comparing microsatellite diversity to other California rivers. Finally, below-barrier sites farther inland were more genetically differentiated from other watersheds than below-barrier sites nearer the river’s mouth.     

Genetic and historical evidence disagree on likely sources of the Atlantic amethyst gem clam Gemma gemma (Totten, 1834) in California

Aim Historical information about source populations of invasive species is often limited; therefore, genetic analyses are used. We compared inference about source populations from historical and genetic data for the oyster‐associated clam, Gemma gemma that invaded California from the USA Atlantic coast. Location Mid‐Atlantic (North Carolina, Maryland), Northeastern (New Jersey, New York, Massachusetts) and the California coasts (Elkhorn Slough, San Francisco Bay, Bolinas Lagoon, Tomales Bay, Bodega Harbor). Methods The documented history of transplantation of Eastern oysters to California was reviewed. Cytochrome c oxidase subunit I (COI) sequences from recent and archived clams were examined in a haplotype network. We used AMOVA to detect geographic genetic structure and a permutation test for significant reductions in diversity. Results Chesapeake Bay oysters were transplanted to New York prior to shipment to San Francisco Bay and from there to peripheral bays. Gemma in the Northeastern and Mid‐Atlantic regions were genetically differentiated. In California, populations in Bodega Harbor and Tomales Bay were genetically similar to those in the Mid‐Atlantic area while clams in San Francisco Bay, Elkhorn Slough and Bolinas Lagoon resembled populations in the Northeastern region. In California, genetic variation was not highest in San Francisco Bay despite greater magnitude of oyster plantings. Haplotypes varied over time in native and introduced populations. Main Conclusions Historical records and inferences from genetics agree that both Northeastern and Mid‐Atlantic regions were sources for Gemma in California. Only complex genetic hypotheses reconcile the strong segregation of haplotypes in California to the historical evidence of mixing in their proximate source (New York). These hypotheses include sorting of mixtures of haplotypes or selection in non‐native areas. Haplotype turnover in San Francisco and Massachusetts samples over time suggests that the sorting hypothesis is plausible. We suggest, however, that Gemma was introduced independently and recently to Tomales Bay and Bodega Harbor.     

Reconstruction of the Sudden Oak Death epidemic in California through microsatellite analysis of the pathogen Phytophthora ramorum

The genetic structure of the clonally reproducing Sudden Oak Death (SOD) pathogen in California was investigated using seven variable microsatellites. A total of 35 multilocus genotypes were identified among 292 samples representative of populations from 14 forest sites and of the nursery trade. amova indicated significant genetic variability both within (44.34%) and among populations (55.66%). Spatial autocorrelation analyses indicated that Moran's index of similarity reached a minimum of 0.1 at 350 m, increased to 0.4 at 1500 m and then decreased to zero at 10 km. These results suggest a bimodal pattern of spread, with medium range dispersal (1500–10 000 m) putatively attributed to the presence of strong winds. Lack of genetic structure was identified for three groups of populations. One group notably included the nurseries' population and two forest populations, both linked to early reports of the pathogen. A neighbour-joining analysis based on pairwise Φ_ST values indicated that the clade inclusive of the nurseries' populations is basal to all California populations. A network analysis identified three common genotypes as the likely founders of the California infestation and proposes a stepwise model for local evolution of novel genotypes. This was supported by the identification in the same locations of novel genotypes and of their 1- or 2-step parents. We hypothesize that the few undifferentiated population groups indicate historical human spread of the pathogen, while the general presence of genetically structured populations indicates that new infestations are currently generated by rare medium or long-range natural movement of the pathogen, followed by local generation of new genotypes.     

Population genetic structure of annual and perennial populations of Zostera marina L. along the Pacific coast of Baja California and the Gulf of California

The Baja California peninsula represents a biogeographical boundary contributing to regional differentiation among populations of marine animals. We investigated the genetic characteristics of perennial and annual populations of the marine angiosperm, Zostera marina, along the Pacific coast of Baja California and in the Gulf of California, respectively. Populations of Z. marina from five coastal lagoons along the Pacific coast and four sites in the Gulf of California were studied using nine microsatellite loci. Analyses of variance revealed significant interregional differentiation, but no subregional differentiation. Significant spatial differentiation, assessed using θ_ST values, was observed among all populations within the two regions. Z. marina populations along the Pacific coast are separated by more than 220 km and had the greatest θ_ST (0.13–0.28) values, suggesting restricted gene flow. In contrast, lower but still significant genetic differentiation was observed among populations within the Gulf of California (θ_ST = 0.04–0.18), even though populations are separated by more than 250 km. This suggests higher levels of gene flow among Gulf of California populations relative to Pacific coast populations. Direction of gene flow was predominantly southward among Pacific coast populations, whereas no dominant polarity in the Gulf of California populations was observed. The test for isolation by distance (IBD) showed a significant correlation between genetic and geographical distances in Gulf of California populations, but not in Pacific coast populations, perhaps because of shifts in currents during El Niño Southern Oscillation (ENSO) events along the Pacific coast.     

Genetic structure in urban and rural populations of Apodemus agrarius in Poland

The aims of this study were (i) to determine the genetic variability of the striped field mouse (Apodemus agrarius Pallas, 1771) population in the area of the Warsaw conurbation and the genetic differentiation among local populations of this species within the city, and (ii) to examine the influence of the degree of isolation on the appearance of such genetic structure. Comparison of the genetic structure of the urban population with that of populations inhabiting rural areas was intended to shed light on the depth of genetic changes in this species caused by life in an urban environment. Eight microsatellite loci were used for genotyping and these molecular analyses were successfully performed for 269 individuals. Genetic differentiation (F_ST) was much higher in the case of the urban population. STRUCTURE analysis indicated that this population was subdivided into several local populations, whereas for the rural locations, the highest likelihood was for a single genetic group (one genetically unstructured population). Moreover, in contrast to the rural population, we found no isolation by distance in the urban population. Despite clear subdivision of the urban population, the level of genetic variability was very similar in both analyzed groups. Our findings indicate the occurrence of an advanced process of differentiation of the urban population by microevolution, whereas populations from rural locations displayed very small and mainly insignificant genetic differentiation. Urban local populations, situated near the city borders or close to the natural ecological corridor (banks of the Vistula River), showed higher genetic variability and were less differentiated from populations inhabiting rural sites than from local populations in the city centre. These results provide support for the importance of ecological corridors in preserving the genetic variability of the urban striped field mouse population.     

Lineage diversification on an evolving landscape: phylogeography of the California newt, Taricha torosa (Caudata: Salamandridae)

We used mitochondrial cytochrome  b sequences (up to 778 bp) and starch gel electrophoresis (45 loci) to examine the phylogeographical history of 39 populations of the California newt, Taricha torosa . Phylogenetic and population genetic methods were integrated to infer history at deep and shallow time depths. Using a molecular clock, the subspecies T. t. torosa and T. t. sierrae were estimated to have diverged 7–13 Mya. Within T. t. torosa, genetically differentiated groups were identified along coastal California, in southern California, and in the southern Sierra Nevada. The coastal group exhibited isolation by distance, but a lack of genetic variation north of present-day Monterey was indicative of a recent range expansion. In southern California, a disjunct population in central San Diego County was genetically diverged from coastal populations to the north (Nei's genetic distance of 0.113). However, mtDNA and protein data were geographically discordant regarding the boundary between the coastal and southern Californian groups, and a biogeographical scenario was developed to account for this discordance. The southern Sierran clade of T. t. torosa was weakly diverged from coastal populations for mtDNA sequence variation, yet was strongly differentiated for allozyme variation (Nei's genetic distance of 0.17–0.20). Populations of T. t. sierrae exhibited substantial population structure, and showed a steeper pattern of isolation by distance than did coastal populations of T. t. torosa . These results are interpreted in consideration of the known geomorphological history of California.  © 2006 The Linnean Society of London, Biological Journal of the Linnean Society , 2006, 89 , 213–239.     

Spatial genetic structuring in a vagile species, the European wood mouse

We examined the genetic structure of natural populations of the European wood mouse Apodemus sylvaticus at the microgeographic (<3 km) and macrogeographic (>30 km) scales. Ecological and behavioural studies indicate that this species exhibits considerable dispersal relative to its home-range size. Thus, there is potential for high gene flow over larger geographic areas. As levels of population genetic structure are related to gene flow, we hypothesized that population genetic structuring at the microgeographic level should be negligible, increasing only with geographic distance. To test this, four sites were sampled within a microgeographic scale with two additional samples at the macrogeographic level. Individuals ( n =415) were screened and analysed for seven polymorphic microsatellite loci. Contrary to our hypothesis, significant levels of population structuring were detected at both scales. Comparing genetic differentiation with geographic distance suggests increasing genetic isolation with distance. However, this distance effect was non-significant being confounded by surprisingly high levels of differentiation among microgeographic samples. We attribute this pattern of genetic differentiation to the effect of habitat fragmentation, splitting large populations into components with small effective population sizes resulting in enhanced genetic drift. Our results indicate that it is incorrect to assume genetic homogeneity among populations even where there is no evidence of physical barriers and dispersal can occur freely. In the case of A. sylvaticus , it is not clear whether dispersal does not occur across habitat barriers or behavioural dispersal occurs without consequent gene flow.     

Hookworm enteritis with bacteremia in California sea lion pups on San Miguel Island

Large breeding populations of California sea lions (Zalophus californianus) are located on San Miguel and San Nicolas Islands in the Southern California Bight. In 2001, there was a substantial increase in pup mortality in late summer and fall. From June 2002 to January 2003, 208 freshly dead pups were examined on San Miguel Island, the most western of the Channel Islands off the coast of southern California. Tissues from 186 of these pups were examined histologically. The primary lesions in 133 (72%) of the pups were an enteritis associated with hookworms and infections in major organs. Emaciation/starvation in 43 pups (26%) was the second most important cause of death.     

Biogeography of the Indonesian archipelago: mitochondrial DNA variation in the fruit bat, Eonycteris spelaea

The fruit bat, Eonycteris spelaea , occurs from India through the Philippines to the southeast limit of its distribution in the Lesser Sunda islands of Indonesia. Mitochondrial DNA (mtDNA) variation was examined in Indonesian E. spelaea island populations by amplification of the D-loop and digestion with restriction endonucleases. In addition, microgeographic variation was assessed by investigation of three cave populations within one island. A total of 24 genotypes, comprising two broad clades, was detected. The pattern of mtDNA variation reflects the colonization history of E. spelaea with estimates of haplotype and sequence diversity highest in the older western populations and lowest at the eastern periphery of the species' distribution. These findings may also be associated with an environmental cline from west to east. There is also evidence that genetic distance between populations reflects geographic relationships, especially historical connectedness, as measured by Pleistocene sea-crossing distances. At the microgeographic level, cave populations were heterogeneous and composed of diverse lineages suggesting restricted local interchange.     

Genetic diversity through life history of Dioon edule Lindley (Zamiaceae, Cycadales)

The distribution of genetic diversity and structure for three populations of Dioon edule Lindley (Zamiaceae) at Monte Oscuro (MO), El Farallón (EF) and Rancho del Niño (RN) in Veracruz, Mexico was studied using 20 allozyme loci, considering four life history classes: seeds, seedlings, juveniles and adults. The MO population is genetically less diverse than the EF and RN populations. Total and local inbreeding differ significantly between life history classes. An increment of inbreeding among all classes was observed, and genetic differentiation among populations was higher in seeds and seedlings than in juveniles and adults. In terms of percentage of polymorphic loci, the MO seeds showed least (80%), followed by RN (95%) and EF had the highest values (100%), probably because of a reduction in effective population size and habitat fragmentation processes. In this context, the mean effective population size was 23.2 ± 11.3 for all populations. We conclude that seed cohorts in EF and RN represent a reservoir of genetic diversity within these two populations. Also, preservation of adult plants is an essential aspect to consider in management and conservation efforts for populations of Dioon edule in natural conditions.     

Molecular markers reveal differentiation among isolates of Coccidioides immitis from California, Arizona and Texas

Coccidioides immitis causes coccidioidomycosis, a fungal disease of both immunocompromised and otherwise healthy people; it is capable of causing large epidemics and the disease is often refractory to chemotherapy. To quantify the magnitude of population differentiation and estimate levels of gene flow in C. immitis , multilocus genotypes were scored for 20–25 clinical isolates from each of Bakersfield (California), Tucson (Arizona), and San Antonio (Texas). The molecular markers used were PCR products with polymorphic restriction endonuclease sites, found and characterized in a previous study of the Tucson population. The data show very highly significant differences in allele frequencies between all three populations, and suggest very low levels of migration between populations. One isolate in the San Antonio sample was an outlier, showing the California-specific allele at all four of the loci distinguishing the two populations, and subsequent inquiries indicated that the infection had indeed been acquired in California. Thus, genetic information can be used to infer the geographical origin of a fungal infection.     

Landscape genetics of alpine-snowbed plants: comparisons along geographic and snowmelt gradients

The genetic structure of three snowbed-herb species (Peucedanum multivittatum, Veronica stelleri, and Gentiana nipponica) was analyzed using allozymes across nine populations arranged as a matrix of three snowmelt gradients x three geographic locations within 3 km in the Taisetsu Mountains, northern Japan. Phenologically asynchronous populations are packed within a local area in alpine snowbeds, because flowering season of alpine plants depends strongly on the timing of snowmelt. Moderate genetic differentiation was detected among local populations in every species (FST=0.03-0.07). There was a significant correlation between the geographic distance and genetic distance in the P. multivittatum populations, but not in the V. stelleri and G. nipponica populations. On the other hand, a significant correlation between the phenological distance caused by snowmelt timing and genetic distance was detected in the V. stelleri and G. nipponica populations, but not in the P. multivittatum populations. The snowmelt gradient or geographic separation influenced hierarchical genetic structure of these species moderately (FRT <0.04). Restriction of gene flow due to phenological separation and possible differential selection along the snowmelt gradient may produce genetic clines at microgeographic scale in these species.     

Spatial and temporal variation in the seroprevalence of canine heartworm antigen in the island fox

Island foxes (Urocyon littoralis) are endemic to six of the eight California Channel Islands (USA). The island fox is classified as a threatened species by the State of California, and recently three of the six subspecies have experienced abrupt population declines. As part of a continuing effort to determine the cause of the declines, we tested island fox serum samples collected in 1988 (n = 176) and 1997-98 (n = 156) over the entire geographic range of the species for seroprevalence of canine heartworm (Dirofilaria immitis) antigen. Using a commercially available enzyme-linked immunosorbent assay (PetChek, Idexx Laboratories, Westbrook, Maine, USA) we detected heartworm antigen in four of the six populations of island foxes. On San Miguel and Santa Rosa Islands, seroprevalence in adult foxes was >85% (n = 62) in 1988 and increased to 100% (n = 24) in 1997-98. On Santa Cruz Island, seroprevalence in adult foxes decreased from 83% (n = 30) to 58% (n = 26), whereas on San Nicolas Island, seroprevalence increased from 25% (n = 32) to 77% (n = 30) during the same period. All of the pups assayed (n = 33) were seronegative. The seroprevalences of heartworm reported herein for the four populations of island foxes are the highest yet reported for a fox species. However, additional demographic data reported elsewhere suggests that heartworm has not been a major factor in the recent declines of island fox populations.     

Dispersal ability and genetic structure in aquatic invertebrates: a comparative study in southern California streams and reservoirs

1. The natural seasonal drying and flooding of southern California streams have been altered over the past century by activities related to agriculture, flood control, and reservoir construction. The genetic structure and diversity of aquatic invertebrates inhabiting these environments is largely unexplored, and may be important for conservation. 2. We sampled two species of aquatic invertebrates with different dispersal abilities to assess genetic structure and diversity, and make inferences about the evolutionary processes that underlie these genetic patterns. The mayfly Fallceon quilleri, which has a winged terrestrial stage, was sampled from perennial and intermittent streams from three catchments across San Diego County. The amphipod Hyalella azteca was sampled from streams (perennial and intermittent) and reservoirs in a single catchment (San Dieguito). Because it is obligately aquatic throughout its life‐cycle, H. azteca was assumed to disperse less than F. quilleri. 3. Intrapopulation and overall genetic diversity was higher in F. quilleri than in H. azteca. In F. quilleri there was very little genetic divergence among populations, and most of the genetic differentiation that was observed could be attributed to a single population. In H. azteca, populations were markedly differentiated between the upper and lower segments of the San Dieguito basin, which are separated by a c. 10 km section of stream that rarely has surface flow. Within both segments, genetic divergence between sites connected by reservoirs and perennial streams was not significantly different. 4. Our results suggest that F. quilleri disperses widely and thus avoids genetic bottlenecks and marked levels of population differentiation that may be expected from frequent extinctions and recolonizations. In contrast, restricted dispersal in H. azteca is associated with relatively low genetic diversity and high genetic divergence across a portion of the catchment in which surface flow is rare.     

Genetic diversity and microgeographic differentiation of Yushan cane (Yushania niitakayamensis; Poaceae) in Taiwan

Yushan cane ( Yushania niitakayamensis ) is distributed in southeast Asia. In Taiwan, the species occurs in mountains 1000–3600 m above sea level. The species appears to spread mainly by rhizomes and flowers only rarely. Nine locations across its distribution range in Taiwan were sampled. Locations at higher altitudes generally consist of grassland and forest undergrowth habitats while those of lower altitudes generally consist of forest undergrowth only. Thus two sampling sites (montane grassland and forest undergrowth) were selected from each location at higher altitudes while only one sampling site was selected from each location at lower altitudes, resulting in a total of 13 sampling sites. Within each sampling site, 20 individual plants were sampled. The results of the cluster analysis and the principal coordinate analysis based on random amplified polymorphic DNA (RAPD) indicated that the populations are generally differentiated according to geographical separation and altitudinal differences that interrupt gene flow. The populations at higher altitudes, where the species is distributed somewhat contiguously, were found to be more similar genetically. Analysis of molecular variance ( AMOVA ) revealed that the among-location, between sampling sites within location, and among individuals within sampling site components accounted for 15.27%, 4.80% and 79.93% of the total variance, respectively. For locations with two sampling sites, two-level AMOVA revealed that the diversities between sampling sites (sun and shade habitats) within locations ranged from 2.91% to 7.99% of the total diversity. Random permutation tests revealed that these diversities were significant, implying that there is microgeographic differentiation due to habitat differences.