Genetic diversity and differentiation among Lagopus lagopus populations in Scandinavia and Scotland: evolutionary significant units confirmed by SNP markers

Single Nucleotide Polymorphism in four Scandinavian populations of willow grouse (Lagopus lagopus) and two Scottish populations of red grouse (Lagopus lagopus scoticus) were assessed at 13 protein‐coding loci. We found high levels of diversity, with one substitution every 55 bp as an average and a total of 76 unlinked parsimony informative SNPs. Different estimators of genetic diversity such as: number of synonymous and non‐synonymous sites, average number of alleles, number and percentage of polymorphic loci, mean nucleotide diversity (π_s, π_a) and gene diversity at synonymous and non‐synonymous sites showed higher diversity in the northern populations compared to southern ones. Strong levels of purifying selection found in all the populations together with neutrality tests conforming to neutral expectations agree with large effective population sizes. Assignment tests reported a clear distinction between Scandinavian and Scottish grouse suggesting the existence of two different evolutionary significant units. The divergence time between willow and red grouse ranging between 12 500 and 125 000 years, in conjunction with the presence of ‘specific’ markers for each subspecies prompt a reassessment of the taxonomical status of the Scottish red grouse.     

Looking into the past – the reaction of three grouse species to climate change over the last million years using whole genome sequences

Tracking past population fluctuations can give insight into current levels of genetic variation present within species. Analysing population dynamics over larger timescales can be aligned to known climatic changes to determine the response of species to varying environments. Here, we applied the Pairwise Sequentially Markovian Coalescent (psmc ) model to infer past population dynamics of three widespread grouse species; black grouse, willow grouse and rock ptarmigan. This allowed the tracking of the effective population size (N_e) of all three species beyond 1 Mya, revealing that (i) early Pleistocene cooling (~2.5 Mya) caused an increase in the willow grouse and rock ptarmigan populations, (ii) the mid‐Brunhes event (~430 kya) and following climatic oscillations decreased the N_e of willow grouse and rock ptarmigan, but increased the N_e of black grouse and (iii) all three species reacted differently to the last glacial maximum (LGM) – black grouse increased prior to it, rock ptarmigan experienced a severe bottleneck and willow grouse was maintained at large population size. We postulate that the varying psmc signal throughout the LGM depicts only the local history of the species. Nevertheless, the large population fluctuations in willow grouse and rock ptarmigan indicate that both species are opportunistic breeders while black grouse tracks the climatic changes more slowly and is maintained at lower N_e. Our results highlight the usefulness of the psmc approach in investigating species’ reaction to climate change in the deep past, but also that caution should be taken in drawing general conclusions about the recent past.     

Temporal changes in allele frequencies in two reciprocally selected maize populations

The effects of breeding on allele frequency changes at 82 restriction fragment length polymorphism (RFLP) loci were examined in two maize (Zea mays L.) populations undergoing reciprocal recurrent selection, Iowa Stiff Stalk Synthetic and Iowa Corn Borer Synthetic #1. After 12 cycles of selection, approximately 30% of the alleles were extinct and 10% near fixation in each population. A test of selective neutrality identified several loci in each population whose allele frequency changes cannot be explained by genetic drift; interpopulation mean expected heterozygosity increased for that subset of 28 loci but not for the remaining 54 loci. Mean expected heterozygosity within the two subpopulations decreased 39%, while the between-population component of genetic variation increased from 0.5% to 33.4% of the total. Effective population size is a key parameter for discerning allele frequency changes due to genetic drift versus those resulting from selection and genetic hitchhiking. Empirical estimates of effective population size for each population were within the range predicted by the breeding method. Received: 10 June 1998 / Accepted: 29 April 1999     

GENETIC VARIATION AND TERRITORIALITY IN WILLOW PTARMIGAN (LAGOPUS LAGOPUS LAGOPUS)

We examined eight polymorphic esterase loci in 526 juvenile and adult willow ptarmigan (Lagopus lagopus lagopus) collected during autumn and spring over five years. The genetic structure of territorial birds during spring differed from birds on the study area in autumn. This can not be explained by selective winter mortality since juvenile birds in the autumn had the same genetic structure as the adults who had lived through at least one winter. In the spring, birds with intermediate heterozygosity had the largest territories and were more frequently mated than expected from random mating among autumn birds. The results suggest selective access to territories by genotype and stabilizing selection, especially among males, since natural selection is assumed to favor large territories and reproduction. Our data also suggest that the spring population consisted of both territorial and non-territorial birds. The genetic relationship between birds of established pairs in spring was lower than that between randomly drawn birds on the study area in the autumn. This together with the finding that parents with an intermediate level of genetic relationship produced the largest broods, suggest that optimal rather than maximal outbreeding is the most successful breeding strategy in this species.     

Temporal variation of allele frequencies in populations of Akodon dolores (Rodentia,Cricetidae)

Summary Six population samples of the South American cricetid rodent Akodon dolores were collected at the same site at six-month intervals over a three year period. Changes in density were detected. Seven out of 18 loci analyzed by means of starch gel electrophoresis were polymorphic. Only two of these loci (Est-4 and G6pdh) showed statistically significant variation in allele frequencies following a seasonal pattern. There was no correlation between allele frequencies and population density. When animals were grouped into two classes according to body weight, a clear difference in allele distribution at the Est-4 and G6pdh loci was observed between individuals 39 g or less and those heavier than 39 g. As the first group comprises predominantly younger animals, the data indicate that changes in the age-structure of population, rather than density variations, are responsible for the cyclic pattern of allele frequencies fluctuations.     

Local genetic structure in red grouse (Lagopus lagopus scoticus): evidence from microsatellite DNA markers

Allelic variation at seven hypervariable tri- and tetranucleotide microsatellite loci was used to determine levels of population differentiation between 14 populations of red grouse ( Lagopus lagopus scoticus ) in northeast Scotland, UK. Despite the potential for long-distance dispersal in grouse, and a semicontinuous habitat, significant population divergence was observed (mean R _ST = 0.153; P < 0.01) and an isolation-by-distance effect detected (Mantel test: P < 0.001). Examination of the spatial trend in principal component scores derived from allele frequencies among populations highlighted a barrier to gene flow that was confounding a simple isolation-by-distance effect. This barrier corresponded to an area of unsuitable habitat for grouse associated with a river system that bisected the study area. Mean genetic relatedness was higher for males than for females in all but one of the study populations, suggesting that the territorial behaviour and natal philopatry displayed by cocks have a manifold effect in generating the observed spatial genetic structure. Lower female relatedness values suggest a higher level of female-mediated gene flow, which is sufficient to prevent the loss of genetic variation from within populations and the onset of inbreeding effects. The potential consequences of local subdivision for red grouse populations are discussed.     

Microevolution in time and space: SNP analysis of historical DNA reveals dynamic signatures of selection in Atlantic cod

Little is known about how quickly natural populations adapt to changes in their environment and how temporal and spatial variation in selection pressures interact to shape patterns of genetic diversity. We here address these issues with a series of genome scans in four overfished populations of Atlantic cod (Gadus morhua) studied over an 80‐year period. Screening of >1000 gene‐associated single‐nucleotide polymorphisms (SNPs) identified 77 loci that showed highly elevated levels of differentiation, likely as an effect of directional selection, in either time, space or both. Exploratory analysis suggested that temporal allele frequency shifts at certain loci may correlate with local temperature variation and with life history changes suggested to be fisheries induced. Interestingly, however, largely nonoverlapping sets of loci were temporal outliers in the different populations and outliers from the 1928 to 1960 period showed almost complete stability during later decades. The contrasting microevolutionary trajectories among populations resulted in sequential shifts in spatial outliers, with no locus maintaining elevated spatial differentiation throughout the study period. Simulations of migration coupled with observations of temporally stable spatial structure at neutral loci suggest that population replacement or gene flow alone could not explain all the observed allele frequency variation. Thus, the genetic changes are likely to at least partly be driven by highly dynamic temporally and spatially varying selection. These findings have important implications for our understanding of local adaptation and evolutionary potential in high gene flow organisms and underscore the need to carefully consider all dimensions of biocomplexity for evolutionarily sustainable management.     

Successful adult willow grouse are exposed to increased harvest risk

Age and sex ratios in bag records are frequently used as indices of population composition for harvested populations. However, vulnerability to harvest may differ by age and sex thereby producing bias in population estimates. We assessed whether age and sex affected vulnerability to harvest for willow grouse (Lagopus lagopus) where adult density and brood size were known in the harvested populations. We collected bag records during 2 days of controlled hunting in 4 areas in 2 years (2007 and 2008) in Jämtland county, Sweden. We found that vulnerability to harvest was different for chicks and adults, but not between male and female adults. Hunters encountered broods at a higher rate than single birds compared to personnel conducting pre-harvest counts along line transects. Furthermore, the probability of shooting a grouse was higher in encounters of broods than individual grouse. Proportionally, we calculated about a 50% probability of a hunter shooting either a chick or an adult independent of encountering a single bird or broods of 2–10 grouse. Increasing adult density also increased the vulnerability to harvest for adults relative to chicks, independent of the chick to adult ratio in the pre-harvest population. The different vulnerability of adults and chicks to harvest observed in this study will dampen variation in age classes in bag records compared to the population, and we caution against extrapolation of age ratios in bag records to harvested populations. © 2012 The Wildlife Society.     

Willow grouse bag size is more sensitive to variation in hunter effort than to variation in willow grouse density

Annual variation in harvested animals (hereafter bag size) is often used as an index of population abundance when investigating population dynamics. Few studies have evaluated how well bag size tracks population change despite its widespread use. Two recent studies on grouse harvest statistics have reached contrasting conclusions. Th ere is limited information about the functional response of hunters in relation to varying game densities, and effort is seldom recorded. We investigated how much of the variation in bag size (total number of harvested grouse km^?2) is explained by variation in willow grouse Lagopus lagopus density (adult and young grouse km^?2) and hunting eff ort (total number of hunting days km^?2). We also evaluated catch per unit effort (CPUE) as an index of grouse abundance, and estimated the response in harvest rate (total bag size in relation to total grouse density) to varying hunting effort. We used data from the 88 management areas on state land in Jämtland county, Sweden (1996–2007), where hunting days and bag size are recorded in detail. Willow grouse density was estimated in four of these management areas in August using line transects and distance sampling. The hunting effort and total grouse density explained most of the variation in bag size (R²= 0.89). Bag size was twice as sensitive to changes in hunting effort compared to changes in grouse density. More than a ten times change in the grouse population density was required to one unit change in bag size. The use of CPUE did not provide a better alternative index of grouse density, and variation in density only explained 23% of the variation in CPUE. Harvest rate showed a strong relationship with hunting effort, and we suggest that an upper limit in hunting effort can be used to reduce the risk of high harvest rates. Hunters became more efficient at low densities and controlling hunting effort is most important when there are indications of population lows and/or poor breeding. CPUE may be less sensitive to changes in game abundance than previously assumed, and bag size as a proxy for population density would then depend on the ability of hunters to adjust their effort according to population change. We speculate that this ability will depend on whether or not hunters have long‐term experience of a hunting area where they can return to hunt throughout the hunting season. We propose that recording hunting eff ort should be encouraged and possible correlations with game abundance and other factors such as weather should be investigated for game species.     

Fine‐scale genetic structure in an eastern Alpine black grouse Tetrao tetrix metapopulation

Understanding genetic consequences of habitat fragmentation is crucial for the management and conservation of wildlife populations, especially in case of species sensitive to environmental changes and landscape alteration. In central Europe, the Alps are the core area of black grouse Tetrao tetrix distribution. There, black grouse dispersal is limited by high altitude mountain ridges and recent black grouse habitats are known to show some degree of natural fragmentation. Additionally, substantial anthropogenic fragmentation has occurred within the past ninety years. Facing losses of peripheral subpopulations and ongoing range contractions, we explored genetic variability and the fine‐scale genetic structure of the Alpine black grouse metapopulation at the easternmost fringe of the species’ Alpine range. Two hundred and fifty tissue samples and non‐invasive faecal and feather samples of eleven a priori defined subpopulations were used for genetic analysis based on nine microsatellite loci. Overall, eastern Alpine black grouse show similar amounts of genetic variation (H_O = 0.65, H_E = 0.66) to those found in more continuous populations like in Scandinavia. Despite of naturally and anthropogenically fragmented landscapes, genetic structuring was weak (global F_ST < 0.05), suggesting that the actual intensity of habitat fragmentation does not completely hamper dispersal, but probably restricts it to some extent. The most peripheral subpopulations at the edge of the species range show signs of genetic differentiation. The present study gives new insights into the population genetic structure of black grouse in the eastern Alps and provides a more fine‐scale view of genetic structure than previously available. Our findings will contribute to monitor the current and future status of the population under human pressures and to support supra‐regional land use planning as well as decision making processes in responsibilities of public administration.     

Genetic variability in white-tailed deer (Odocoileus virginianus) and its relationship to environmental parameters and herd origin (Cervidae)

Allozyme variation was examined in 1571 white-tailed deer (Odocoileus virginianus) from 29 localities in Tennessee by starch gel electrophoresis. For 11 polymorphic loci, sex-related, age-related and temporal differences were minimal. However, significant spatial hererogeneity was evident in genotypes (contingency table results), allele frequencies (F _ST=0.057) and heterozygosity. Heterozygosity ranged from 16.9% to 26.8% with a mean of 22.9%. The spatial pattern of allele frequencies determined from Rogers' coefficients of genetic similarity indicated associations based on geographic proximity and stocking history. In hierarchial analyses, physiographic regions accounted for more of the total gene diversity than herd origin groups (populations of similar origin) but less than individual populations. For five loci, physiographic regions accounted for more of the gene diversity than populations, suggesting a selection role in the observed genetic variability. Bivariate and canonical correlation analyses revealed significant associations between environmental and genetic variables. Temperature variables and allele frequencies for three loci (alcohol dehydrogenase, alpha-glycerophosphate dehydrogenase, sorbitol dehydrogenase) had the prominent roles in the multivariate association between environmental and genetic variables. Herd origin, gene flow and selection appear to be involved in the gene diversity in deer from Tennesee.     

Microsatellites and allozymes as the genetic memory of habitat fragmentation and defragmentation in populations of the ground beetle Carabus auronitens (Col., Carabidae)

Aim This paper has three aims: (1) to reconstruct the population history of a flightless silvicolous (forest) ground beetle in a region where strong anthropogenic activity has altered the connectivity of the landscape; (2) to estimate the effects of both contemporary and historical landscape structure on the genetics of populations; and (3) to investigate the reasons for clinal variation in one gene locus found in an earlier study carried out in the same geographical location. Location Münster (Westphalia), north‐west Germany. Methods We investigated 26 populations of the ground beetle Carabus auronitens Fabricius, 1792 by analysing seven polymorphic microsatellite loci and an allozyme locus. Samples of at least 16 individuals per site were studied. These were obtained from dry pitfall traps placed at 23 sites and from three additional (refuge) populations. We used regression and correlation analyses to investigate the effects of both historical and contemporary landscape structure on the allele frequency distributions of the investigated loci. Spatial autocorrelation analysis was used to study possible clinal variations, and admixture rates were calculated in order to assess the genetic influence of populations from possible refuges. Possible reasons for the development of the cline were examined using simulation models. Results The allele frequency distributions at the investigated loci could not be explained by selection or isolation by distance. We found clinal variation in 50% of the investigated loci and our simulations indicated that this was unlikely to have developed by chance. Admixture rates show significant influences of the investigated refuge populations on the populations under study. Main conclusions The findings strongly suggest that the clinal variation is secondary. It results from recolonization of the area by C. auronitens from multiple refuges after anthropogenic landscape changes caused forest fragmentation and led to species isolation.     

The genetic underpinnings of population cyclicity: establishing expectations for the genetic anatomy of cycling populations

Despite extensive research into the mechanisms underlying population cyclicity, we have little understanding of the impacts of numerical fluctuations on the genetic variation of cycling populations. Thus, the potential implications of natural and anthropogenically‐driven variation in population cycle dynamics on the diversity and evolutionary potential of cyclic populations is unclear. Here, we use Canada lynx Lynx canadensis matrix population models, set up in a linear stepping‐stone, to generate demographic replicates of biologically realistic cycling populations. Overall, increasing cycle amplitude predictably reduced genetic diversity and increased genetic differentiation, with cyclic effects increased by population synchrony. Modest dispersal rates (1–3% of the population) between high and low amplitude cyclic populations did not diminish these effects suggesting that spatial variation in cyclic amplitude should be reflected in patterns of genetic diversity and differentiation at these rates. At high dispersal rates (6%) groups containing only high amplitude cyclic populations had higher diversity and lower differentiation than those mixed with low amplitude cyclic populations. Negative density‐dependent dispersal did not impact genetic diversity, but did homogenize populations by reducing differentiation and patterns of isolation by distance. Surprisingly, temporal changes in diversity and differentiation throughout a cycle were not always consistent with population size. In particular, negative density‐dependent dispersal simultaneously decreased differences in genetic diversity while increasing differences in genetic differentiation between numerical peaks and nadirs. Combined, our findings suggest demographic changes at fine temporal scales can impact genetic variation of interacting populations and provide testable predictions relating population cyclicty to genetic variation. Further, our results suggest that including realistic demographic and dispersal parameters in population genetic models and using information from temporal changes in genetic variation could help to discern complex demographic scenarios and illuminate population dynamics at fine temporal scales.     

TEMPORAL VARIATION IN POPULATIONS OF THE MARINE ISOPOD EXCIROLANA: HOW STABLE ARE GENE FREQUENCIES AND MORPHOLOGY?

Excirolana braziliensis is a dioecious marine isopod that lives in the high intertidal zone on both sides of tropical America. It lacks a dispersal phase and displays a remarkable degree of genetic divergence even between localities less than 1 km apart. Nine populations of this nominal species from both sides of the Isthmus of Panama and one population of the closely allied species, Excirolana chamensis, from the eastern Pacific were studied for 2 yr for allozymic temporal variation in 13 loci and for 3 to 4 yr for morphological variation in nine characters. The genetic and morphological constitution of 9 out of 10 populations remained stable. Allele frequencies at two loci and overall morphology in a tenth beach occupied by E. braziliensis changed drastically and significantly between 1986 and 1988. The change in gene frequency is too great to explain by genetic drift occurring during a maximum of 14 generations regardless of assumed effective population size; drift is also unlikely to have caused observed changes in morphology. Selective survival of a previously rare genotype is more plausible but still not probable. The most credible explanation is that the resident population at this locality became extinct and that the beach was recolonized by immigrants from another locality. Such infrequent episodes of extinction and recolonization from a single source may account for the large amount of genetic divergence between local populations of E. braziliensis. However, the low probability of large temporal genetic change even in a species such as this, in which gene flow between local demes is limited and generation time is short, suggests that a single sample through time is usually adequate for reconstructing the genetic history of populations.     

A Population Genomic Assessment of Three Decades of Evolution in a Natural Drosophila Population

Population genetics seeks to illuminate the forces shaping genetic variation, often based on a single snapshot of genomic variation. However, utilizing multiple sampling times to study changes in allele frequencies can help clarify the relative roles of neutral and non-neutral forces on short time scales. This study compares whole-genome sequence variation of recently collected natural population samples of Drosophila melanogaster against a collection made approximately 35 years prior from the same locality—encompassing roughly 500 generations of evolution. The allele frequency changes between these time points would suggest a relatively small local effective population size on the order of 10,000, significantly smaller than the global effective population size of the species. Some loci display stronger allele frequency changes than would be expected anywhere in the genome under neutrality—most notably the tandem paralogs Cyp6a17 and Cyp6a23, which are impacted by structural variation associated with resistance to pyrethroid insecticides. We find a genome-wide excess of outliers for high genetic differentiation between old and new samples, but a larger number of adaptation targets may have affected SNP-level differentiation versus window differentiation. We also find evidence for strengthening latitudinal allele frequency clines: northern-associated alleles have increased in frequency by an average of nearly 2.5% at SNPs previously identified as clinal outliers, but no such pattern is observed at random SNPs. This project underscores the scientific potential of using multiple sampling time points to investigate how evolution operates in natural populations, by quantifying how genetic variation has changed over ecologically relevant timescales.     

Effective population size and temporal genetic change in stream resident brown trout (Salmo trutta, L.)

Temporal genetic data may be used forestimating effective population size (N _e) and for addressing the `temporal stability' of population structure, two issues of central importance for conservation and management. In this paper we assess the amount of spatio-temporal genetic variation at 17 di-allelic allozyme loci and estimate current N _e in two populations of stream resident brown trout (Salmo trutta) using data collected over 20 years. The amount ofpopulation divergence was found to bereasonably stable over the studied time period.There was significant temporal heterogeneitywithin both populations, however, and N _e was estimated as 19 and 48 for the twopopulations. Empirical estimates of theprobability of detecting statisticallysignificant allele frequency differencesbetween samples from the same populationseparated by different numbers of years wereobtained. This probability was found to befairly small when comparing samples collectedonly a few years apart, even for theseparticular populations that exhibit quiterestricted effective sizes. We discuss someimplications of the present results for browntrout population genetics and conservation, andfor the analysis of temporal genetic change inpopulations with overlapping generations ingeneral.     

Molecular evidence for historical and recent population size reductions of tiger salamanders (Ambystoma tigrinum) in Yellowstone National Park

Population declines caused by natural and anthropogenic factors can quickly erode genetic diversity in natural populations. In this study, we examined genetic variation within 10 tiger salamander populations across northern Yellowstone National Park in Wyoming and Montana, USA using eight microsatellite loci. We tested for the genetic signature of population decline using heterozygosity excess, shifts in allele frequencies, and low ratios of allelic number to allelic size range (M-ratios). We found different results among the three tests. All 10 populations had low M-ratios, five had shifts in allele frequencies and only two had significant heterozygosity excesses. These results support theoretical expectations of different temporal signatures among bottleneck tests and suggest that both historical fish stocking, recent, sustained drought, and possibly an emerging amphibian disease have contributed to declines in effective population size.     

GENETIC DIVERSITY AND POPULATION STRUCTURE IN CAMELLIA JAPONICA L. (THEACEAE)

Camellia japonica is a widespread and morphologically diverse tree native to parts of Japan and adjacent islands. Starch gel electrophoresis was used to score allelic variation at 20 loci in seeds collected from 60 populations distributed throughout the species range. In comparison with other plant species, the level of genetic diversity within C. japonica populations is very high: 66.2% of loci were polymorphic on average per population, with a mean number of 2.16 alleles per locus; the mean observed and panmictic heterozygosities were 0.230 and 0.265, respectively. Genotypic proportions at most loci in most populations fit Hardy-Weinberg expectations. However, small heterozygote deficiencies were commonly observed (mean population fixation index = 0.129). It is suggested that the most likely cause of the observed deficiencies is population subdivision into genetically divergent subpopulations. The overall level of population differentiation is greater than is typically observed in out-breeders: The mean genetic distance and identity (Nei's D and I) between pairs of populations were 0.073 and 0.930, respectively, and Wright's Fst was 0.144. Differences among populations appeared to be manifested as variation in gene frequencies at many loci rather than variation in allelic composition per se. However, the patterns of variation were not random. Reciprocal clinal variation of gene frequencies was observed for allele pairs at six loci. In addition, principal components analysis revealed that populations tended to genetically cluster into four regions representing the geographic areas Kyushu, Shikoku, western Honshu, and eastern Honshu. There was a significant relationship between genetic and geographic distance (r = 0.61; P < 0.01). Analysis of variance on allozyme frequencies showed that there was approximately four times as much differentiation among populations within regions, as among regions. It is likely that the observed patterns of population relationships result from the balance between genetic drift in small subpopulations and gene flow between them.     

A population genetic evaluation of ecological restoration with the case study on <Emphasis Type="Italic">Cyclobalanopsis myrsinaefolia</Emphasis> (Fagaceae)

The ultimate goal of ecological restoration is to create a self-sustaining ecosystem that is resilient to perturbation without further assistance. Genetic variation is a prerequisite for evolutionary response to environmental changes. However, few studies have evaluated the genetic structure of restored populations of dominant plants. In this study, we compared genetic variation of the restored populations with the natural ones in Cyclobalanopsis myrsinaefolia, a dominant species of evergreen broadleaved forest. Using eight polymorphic microsatellite loci, we analyzed samples collected from restored populations and the donor population as well as two other natural populations. We compared the genetic diversity of restored and natural populations. Differences in genetic composition were evaluated using measurements of genetic differentiation and assignment tests. The mean number of alleles per locus was 4.65. Three parameters (A, A _R, and expected heterozygosity) of genetic variation were found to be lower, but not significantly, in the restored populations than they were in the natural populations, indicating a founder effect during the restoration. Significant but low F _ST (0.061) was observed over all loci, indicating high gene flow among populations, as expected from its wind-pollination. Differentiation between the two restored populations was smallest. However, differences between the donor population and the restored populations were higher than those between other natural populations and the restored populations. Only 13.5% and 25.7% individuals in the two restored populations were assigned to the donor population, but 54.1 and 40% were assigned to another natural population. The genetic variation of the donor population was lowest, and geographic distances from the restoration sites to the donor site were much higher than the other natural populations, indicating that the present donor likely was not the best donor for these ecological restoration efforts. However, no deleterious consequences might be observed in restored populations due to high observed heterozygosity and high gene flow. This study demonstrates that during the restoration process, genetic structures of the restored populations may be biased from the donor population. The results also highlight population genetic knowledge, especially of gene flow-limited species, in ecological restoration.