Spatially explicit models of divergence and genome hitchhiking

Strong barriers to genetic exchange can exist at divergently selected loci, whereas alleles at neutral loci flow more readily between populations, thus impeding divergence and speciation in the face of gene flow. However, ‘divergence hitchhiking’ theory posits that divergent selection can generate large regions of differentiation around selected loci. ‘Genome hitchhiking’ theory suggests that selection can also cause reductions in average genome‐wide rates of gene flow, resulting in widespread genomic divergence (rather than divergence only around specific selected loci). Spatial heterogeneity is ubiquitous in nature, yet previous models of genetic barriers to gene flow have explored limited combinations of spatial and selective scenarios. Using simulations of secondary contact of populations, we explore barriers to gene flow in various selective and spatial contexts in continuous, two‐dimensional, spatially explicit environments. In general, the effects of hitchhiking are strongest in environments with regular spatial patterning of starkly divergent habitat types. When divergent selection is very strong, the absence of intermediate habitat types increases the effects of hitchhiking. However, when selection is moderate or weak, regular (vs. random) spatial arrangement of habitat types becomes more important than the presence of intermediate habitats per se. We also document counterintuitive processes arising from the stochastic interplay between selection, gene flow and drift. Our results indicate that generalization of results from two‐deme models requires caution and increase understanding of the genomic and geographic basis of population divergence.     

Landscape genetics of alpine Sierra Nevada salamanders reveal extreme population subdivision in space and time

Quantifying the influence of the landscape on the genetic structure of natural populations remains an important empirical challenge, particularly for poorly studied, ecologically cryptic species. We conducted an extensive microsatellite analysis to examine the population genetics of the southern long‐toed salamander (Ambystoma macrodactylum sigillatum) in a naturally complex landscape. Using spatially explicit modelling, we investigated the influence of the Sierra Nevada topography on potential dispersal corridors between sampled populations. Our results indicate very high‐genetic divergence among populations, high within‐deme relatedness, and little evidence of recent migration or population admixture. We also discovered unexpectedly high between‐year genetic differentiation (F_ST) for breeding sites, suggesting that breeding groups vary over localized space and time. While environmental factors associated with high‐elevation montane habitats apparently play an important role in shaping population differentiation, additional, species‐specific biological processes must also be operating to account for observed deviations from temporal, among‐year panmixia. Our study emphasizes the population‐level insights that can be gained from high‐density sampling in space and time, and the highly substructured population biology that may characterize amphibians in extreme montane habitats.     

Application of network methods for understanding evolutionary dynamics in discrete habitats

In populations occupying discrete habitat patches, gene flow between habitat patches may form an intricate population structure. In such structures, the evolutionary dynamics resulting from interaction of gene‐flow patterns with other evolutionary forces may be exceedingly complex. Several models describing gene flow between discrete habitat patches have been presented in the population‐genetics literature; however, these models have usually addressed relatively simple settings of habitable patches and have stopped short of providing general methodologies for addressing nontrivial gene‐flow patterns. In the last decades, network theory – a branch of discrete mathematics concerned with complex interactions between discrete elements – has been applied to address several problems in population genetics by modelling gene flow between habitat patches using networks. Here, we present the idea and concepts of modelling complex gene flows in discrete habitats using networks. Our goal is to raise awareness to existing network theory applications in molecular ecology studies, as well as to outline the current and potential contribution of network methods to the understanding of evolutionary dynamics in discrete habitats. We review the main branches of network theory that have been, or that we believe potentially could be, applied to population genetics and molecular ecology research. We address applications to theoretical modelling and to empirical population‐genetic studies, and we highlight future directions for extending the integration of network science with molecular ecology.     

Genetic load,inbreeding depression and heterosis in an age‐structured metapopulation

In a metapopulation, the process of recurrent local extinction and recolonization gives rise to an age structure among demes. Recently established demes will tend to differ from older demes in terms of the levels of genetic diversity found within them and the way this diversity is distributed among demes in the same and different ages. The effects of population turnover on average levels of genetic diversity among demes in a metapopulation have been the focus of much attention, both for neutral and nonneutral loci, but much less is known about the distribution of nonneutral genetic diversity among demes of different ages. In this paper, we used computer simulations to study the distribution of genetic load, inbreeding depression and heterosis in an age‐structured metapopulation. We found that, for mildly deleterious mutations, within‐deme inbreeding depression increased, whereas heterosis and genetic load decreased with deme age following severe colonization bottlenecks. In contrast, recessive lethal alleles tended to be purged during colonization, with older populations showing higher genetic load and higher within‐deme inbreeding depression. Heterosis caused by recessive lethal alleles and resulting from gene flow among different demes tended to be greatest for young demes, because the mutations responsible tended to be purged in the first few generations after colonization, but its effects increased again as populations grow older as a result of immigration. Our results point to a need for estimates of genetic diversity, genetic load, within‐deme inbreeding depression and heterosis in demes of different age classes separately.     

MetaPopGen: an r package to simulate population genetics in large size metapopulations

Population genetics simulation models are useful tools to study the effects of demography and environmental factors on genetic variation and genetic differentiation. They allow for studying species and populations with complex life histories, spatial distribution and many other complicating factors that make analytical treatment impracticable. Most simulation models are individual‐based: this poses a limitation to simulation of very large populations because of the limits in computer memory and long computation times. To overcome these limitations, we propose an intermediate approach that allows modelling of very complex demographic scenarios, which would be intractable with analytical models, and removes the limitations imposed by large population size, which affect individual‐based simulation models. We implement this approach in a software package for the r environment, MetaPopGen. The innovative concept of this approach with respect to the other population genetic simulators is that it focuses on genotype numbers rather than on individuals. Genotype numbers are iterated through time by using random number generators for appropriate probabilistic distributions to reproduce the stochasticity inherent to Mendelian segregation, survival, dispersal and reproduction. Features included in the model are age structure, monoecious and dioecious (or separate sexes) life cycles, mutation, dispersal and selection. The model simulates only one locus at a time. All demographic parameters can be genotype‐, sex‐, age‐, deme‐ and time‐dependent. MetaPopGen is therefore indicated to study large populations and very complex demographic scenarios. We illustrate the capabilities of MetaPopGen by applying it to the case of a marine fish metapopulation in the Mediterranean Sea.     

Extranuclear differentiation and gene flow in the finite island model

Use of sequence information from extranuclear genomes to examine deme structure in natural populations has been hampered by lack of clear linkage between sequence relatedness and rates of mutation and migration among demes. Here, we approach this problem in two complementary ways. First, we develop a model of extranuclear genomes in a population divided into a finite number of demes. Sex-dependent migration, neutral mutation, unequal genetic contribution of separate sexes and random genetic drift in each deme are incorporated for generality. From this model, we derive the relationship between gene identity probabilities (between and within demes) and migration rate, mutation rate and effective deme size. Second, we show how within- and between-deme identity probabilities may be calculated from restriction maps of mitochondrial (mt) DNA. These results, when coupled with our results on gene flow and genetic differentiation, allow estimation of relative interdeme gene flow when deme sizes are constant and genetic variants are selectively neutral. We illustrate use of our results by reanalyzing published data on mtDNA in mouse populations from around the world and show that their geographic differentiation is consistent with an island model of deme structure.     

THE EFFECT OF COLLECTIVE DISPERSAL ON THE GENETIC STRUCTURE OF A SUBDIVIDED POPULATION

Correlated dispersal paths between two or more individuals are widespread across many taxa. The population genetic implications of this collective dispersal have received relatively little attention. Here we develop two‐sample coalescent theory that incorporates collective dispersal in a finite island model to predict expected coalescence times, genetic diversities, and F‐statistics. We show that collective dispersal reduces mixing in the system, which decreases expected coalescence times and increases F_ST. The effects are strongest in systems with high migration rates. Collective dispersal breaks the invariance of within‐deme coalescence times to migration rate, whatever the deme size. It can also cause F_ST to increase with migration rate because the ratio of within‐ to between‐deme coalescence times can decrease as migration rate approaches unity. This effect is most biologically relevant when deme size is small. We find qualitatively similar results for diploid and gametic dispersal. We also demonstrate with simulations and analytical theory the strong similarity between the effects of collective dispersal and anisotropic dispersal. These findings have implications for our understanding of the balance between drift–migration–mutation in models of neutral evolution. This has applied consequences for the interpretation of genetic structure (e.g., chaotic genetic patchiness) and estimation of migration rates from genetic data.     

Linkage disequilibrium clustering‐based approach for association mapping with tightly linked genomewide data

Genomewide association studies (GWAS) aim to identify genetic markers strongly associated with quantitative traits by utilizing linkage disequilibrium (LD) between candidate genes and markers. However, because of LD between nearby genetic markers, the standard GWAS approaches typically detect a number of correlated SNPs covering long genomic regions, making corrections for multiple testing overly conservative. Additionally, the high dimensionality of modern GWAS data poses considerable challenges for GWAS procedures such as permutation tests, which are computationally intensive. We propose a cluster‐based GWAS approach that first divides the genome into many large nonoverlapping windows and uses linkage disequilibrium network analysis in combination with principal component (PC) analysis as dimensional reduction tools to summarize the SNP data to independent PCs within clusters of loci connected by high LD. We then introduce single‐ and multilocus models that can efficiently conduct the association tests on such high‐dimensional data. The methods can be adapted to different model structures and used to analyse samples collected from the wild or from biparental F₂ populations, which are commonly used in ecological genetics mapping studies. We demonstrate the performance of our approaches with two publicly available data sets from a plant (Arabidopsis thaliana) and a fish (Pungitius pungitius), as well as with simulated data.     

Divergence and evolution of assortative mating in a polygenic trait model of speciation with gene flow

Assortative mating is an important driver of speciation in populations with gene flow and is predicted to evolve under certain conditions in few‐locus models. However, the evolution of assortment is less understood for mating based on quantitative traits, which are often characterized by high genetic variability and extensive linkage disequilibrium between trait loci. We explore this scenario for a two‐deme model with migration, by considering a single polygenic trait subject to divergent viability selection across demes, as well as assortative mating and sexual selection within demes, and investigate how trait divergence is shaped by various evolutionary forces. Our analysis reveals the existence of sharp thresholds of assortment strength, at which divergence increases dramatically. We also study the evolution of assortment via invasion of modifiers of mate discrimination and show that the ES assortment strength has an intermediate value under a range of migration‐selection parameters, even in diverged populations, due to subtle effects which depend sensitively on the extent of phenotypic variation within these populations. The evolutionary dynamics of the polygenic trait is studied using the hypergeometric and infinitesimal models. We further investigate the sensitivity of our results to the assumptions of the hypergeometric model, using individual‐based simulations.     

Genetic structure of a native cyprinid in a reservoir‐altered stream network

1. Reservoirs modify riverine ecosystems worldwide, and often with deleterious impacts on native biota. The immediate effects of reservoirs on native fish species below dams and in impounded reaches have received considerable attention, but it is unclear how reservoirs may affect fish species at larger spatial and temporal scales. Documented declines of stream fish populations in direct tributaries of reservoirs suggest reservoir pools may reduce gene flow among historically connected populations. 2. Because of increased predator densities in reservoirs and the extent of habitat alteration in impounded reaches, I predicted reservoir habitats would reduce gene flow among small‐bodied fish populations separated by reservoir habitat. I used microsatellite markers to assess the spatial genetic structure of populations of the red shiner (Cyprinella lutrensis), in a reservoir‐fragmented stream network (Lake Texoma, U.S.A.). I also tested the prediction that populations in two direct tributaries that have experienced population declines would have low genetic diversity. Individuals were collected from six sites upstream of the reservoir, three sites in the reservoir and three sites in direct tributaries of the reservoir during 2008 and 2009. 3. Results indicate that most populations were isolated by distance with little divergence among populations. In one direct tributary population, however, there was substantial genetic divergence, and genetic diversity was significantly lower than in other populations. Gene flow also seemed to be lower in reservoir habitats than in intact stream habitats, suggesting reservoir habitats may be reducing gene flow among the reservoir‐separated populations. These results indicate that reservoirs may reduce gene flow among reservoir‐fragmented stream fish populations, altering the evolutionary trajectories of fragmented populations.     

Fine-scale spatial genetic structure and dispersal among spotted salamander (Ambystoma maculatum) breeding populations

We examined fine-scale genetic variation among breeding aggregations of the spotted salamander (Ambystoma maculatum) to quantify dispersal, interpopulation connectivity and population genetic structure. Spotted salamanders rely on temporary ponds or wetlands for aggregate breeding. Adequate breeding sites are relatively isolated from one another and field studies suggest considerable adult site fidelity; therefore, we expected to find population structure and differentiation at small spatial scales. We used microsatellites to estimate population structure and dispersal among 29 breeding aggregations in Tompkins County, New York, USA, an area encompassing 1272 km(2). Bayesian and frequency-based analyses revealed fine-scale genetic structure with two genetically defined demes: the North deme included seven breeding ponds, and the South deme included 13 ponds. Nine ponds showed evidence of admixture between these two genetic pools. Bayesian assignment tests for detection of interpopulation dispersal indicate that immigration among ponds is common within demes, and that certain populations serve as sources of immigrants to neighbouring ponds. Likewise, spatial genetic correlation analyses showed that populations < or = 4.8 km distant from each other show significant genetic correlation that is not evident at higher scales. Within-population levels of relatedness are consistently larger than expected if mating were completely random across ponds, and in the case of a few ponds, within-population processes such as inbreeding or reproductive skew contribute significantly to differentiation from neighbouring ponds. Our data underscore the importance of these within-population processes as a source of genetic diversity across the landscape, despite considerable population connectivity. Our data further suggest that spotted salamander breeding groups behave as metapopulations, with population clusters as functional units, but sufficient migration among demes to allow for potential rescue and recolonization. Amphibian habitats are becoming increasingly fragmented and a clear understanding of dispersal and patterns of population connectivity for taxa with different ecologies and life histories is crucial for their conservation.     

Functional connectivity in replicated urban landscapes in the land snail (Cornu aspersum)

Urban areas are highly fragmented and thereby exert strong constraints on individual dispersal. Despite this, some species manage to persist in urban areas, such as the garden snail, Cornu aspersum, which is common in cityscapes despite its low mobility. Using landscape genetic approaches, we combined study area replication and multiscale analysis to determine how landscape composition, configuration and connectivity influence snail dispersal across urban areas. At the overall landscape scale, areas with a high percentage of roads decreased genetic differentiation between populations. At the population scale, genetic differentiation was positively linked with building surface, the proportion of borders where wooded patches and roads appeared side by side and the proportion of borders combining wooded patches and other impervious areas. Analyses based on pairwise genetic distances validated the isolation‐by‐distance and isolation‐by‐resistance models for this land snail, with an equal fit to least‐cost paths and circuit‐theory‐based models. Each of the 12 landscapes analysed separately yielded specific relations to environmental features, whereas analyses integrating all replicates highlighted general common effects. Our results suggest that urban transport infrastructures facilitate passive snail dispersal. At a local scale, corresponding to active dispersal, unfavourable habitats (wooded and impervious areas) isolate populations. This work upholds the use of replicated landscapes to increase the generalizability of landscape genetics results and shows how multiscale analyses provide insight into scale‐dependent processes.     

Genetic variation in an ephemeral mudflat species: The role of the soil seed bank and dispersal in river and secondary anthropogenic habitats

Many ephemeral mudflat species, which rely on a soil seed bank to build up the next generation, are endangered in their natural habitat due to the widespread regulation of rivers. The aim of the present study was to elucidate the role of the soil seed bank and dispersal for the maintenance of genetic diversity in populations of near‐natural river habitats and anthropogenic habitats created by traditional fish farming practices using Cyperus fuscus as a model. Using microsatellite markers, we found no difference in genetic diversity levels between soil seed bank and above‐ground population and only moderate differentiation between the two fractions. One possible interpretation is the difference in short‐term selection during germination under specific conditions (glasshouse versus field) resulting in an ecological filtering of genotypes out of the reservoir in the soil. River populations harbored significantly more genetic diversity than populations from the anthropogenic pond types. We suggest that altered levels and patterns of dispersal together with stronger selection pressures and historical bottlenecks in anthropogenic habitats are responsible for the observed reduction in genetic diversity. Dispersal is also supposed to largely prohibit genetic structure across Europe, although there is a gradient in private allelic richness from southern Europe (high values) to northern, especially north‐western, Europe (low values), which probably relates to postglacial expansion out of southern and/or eastern refugia.     

Habitat connectivity for conserving cervids in a multifunctional landscape

Habitat fragmentation is a considerable threat to biodiversity worldwide. To minimize the effects of fragmentation, it is important to identify and conserve the existing habitat connections that facilitate dispersal and gene flow among populations. Connected populations are more resilient to the changing environment that affects local populations due to greater demographic stability and higher genetic diversity. Our study is the first attempt to identify the crucial habitats facilitating the dispersal of two key sympatric cervids - spotted deer Axis axis and sambar Rusa unicolor in central India. We use species distribution models followed by landscape pattern analyses and connectivity analyses to delineate the essential habitats. Thereafter, we estimated the relative contribution of habitats outside protected areas in maintaining the ecological network, using graph-based metrics. We then locate and predict the areas that have a high risk of human-influenced cervid mortality using a Bayesian regression model that accounts for spatial structure in the data. The results show that about 55% of the core habitats, integrated across both species, lie outside the protected areas and are important in maintaining the ecological network for these cervids. Some peripheral habitats have an increased risk of anthropogenic cervid mortality, which poses high demographic risk. There is an urgent need to regulate the nature and intensity of human activities in areas of human-wildlife coexistence to maintain habitat connectivity and ensure the survival of wildlife populations. Our results on cervids complement analyses on connectivity for large carnivores and thus enables one to account for important trophic interactions among wildlife species in land use planning.     

Spatial scale affects landscape genetic analysis of a wetland grasshopper

Most landscape genetic studies assess the impact of landscape elements on species' dispersal and gene flow. Many of these studies perform their analysis on all possible population pairs in a study area and do not explicitly consider the effects of spatial scale and population network topology on their results. Here, we examined the effects of spatial scale and population network topology on the outcome of a landscape genetic analysis. Additionally, we tested whether the relevant spatial scale of landscape genetic analysis could be defined by population network topology or by isolation‐by‐distance (IBD) patterns. A data set of the wetland grasshopper Stethophyma grossum, collected in a fragmented agricultural landscape, was used to analyse population network topology, IBD patterns and dispersal habitats, using least‐cost transect analysis. Landscape genetic analyses neglecting spatial scale and population network topology resulted in models with low fits, with which a most likely dispersal habitat could not be identified. In contrast, analyses considering spatial scale and population network topology resulted in high model fits by restricting landscape genetic analysis to smaller scales (0–3 km) and neighbouring populations, as represented by a Gabriel graph. These models also successfully identified a likely dispersal habitat of S. grossum. The above results suggest that spatial scale and potentially population network topology should be more explicitly considered in future landscape genetic analyses.     

Contrasting patterns of nuclear microsatellite genetic structure of Fraxinus mandshurica var. japonica between northern and southern populations in Japan

Aim The aim of this study is to detect extant patterns of population genetic structure of Fraxinus mandshurica var. japonica in Japan, and to provide insights into the post‐glacial history of this species during the Holocene. Location Hokkaido and Honshu islands, Japan (including the Oshima and Shimokita peninsulas). Methods We examined nine polymorphic nuclear microsatellite loci to assess genetic variation within and among 15 populations across almost the entire range of the species in Japan. Extant patterns of geographical structure were analysed using Bayesian clustering, Monmonier’s algorithm, analysis of molecular variance, Mantel tests and principal coordinates analysis. Recent bottlenecks within populations and regional genetic variation were also assessed. Results Northern populations (Hokkaido Island and the Shimokita Peninsula) formed a single homogeneous deme, maintaining the highest level of allelic diversity on the Oshima Peninsula. By contrast, southern populations (Honshu Island) demonstrated strong substructure on both coasts. Specifically, populations on the Pacific side of Honshu exhibited significant bottlenecks and erosion of allelic diversity but preserved distinct subclusters diverging from widespread subclusters on the Japan Sea side of this island. Main conclusions Genetic evidence and life history traits suggest that F. mandshurica occupied cryptic northern refugia on the Oshima Peninsula during the Last Glacial Maximum, which is reflected in the species’ extant northern distribution. Strong geographical structure in southern populations, in agreement with fossil pollen records, suggests geographical isolation by mountain ranges running north–south along Honshu. Given that this tree species is cold‐adapted and found in riparian habitats, populations on the Pacific side of Honshu probably contracted into higher‐elevation swamps during warm post‐glacial periods, leading to a reduction of effective population sizes and rare allelic richness.     

Quaternary climate instability is correlated with patterns of population genetic variability in Bombus huntii

Climate oscillations have left a significant impact on the patterns of genetic diversity observed in numerous taxa. In this study, we examine the effect of Quaternary climate instability on population genetic variability of a bumble bee pollinator species, Bombus huntii in western North America. Pleistocene and contemporary B. huntii habitat suitability (HS) was estimated with an environmental niche model (ENM) by associating 1,035 locality records with 10 bioclimatic variables. To estimate genetic variability, we genotyped 380 individuals from 33 localities at 13 microsatellite loci. Bayesian inference was used to examine population structure with and without a priori specification of geographic locality. We compared isolation by distance (IBD) and isolation by resistance (IBR) models to examine population differentiation within and among the Bayesian inferred genetic clusters. Furthermore, we tested for the effect of environmental niche stability (ENS) on population genetic diversity with linear regression. As predicted, high‐latitude B. huntii habitats exhibit low ENS when compared to low‐latitude habitats. Two major genetic clusters of B. huntii inhabit western North America: (a) a north genetic cluster predominantly distributed north of 28°N and (b) a south genetic cluster distributed south of 28°N. In the south genetic cluser, both IBD and IBR models are significant. However, in the north genetic cluster, IBD is significant but not IBR. Furthermore, the IBR models suggest that low‐latitude montane populations are surrounded by habitat with low HS, possibly limiting dispersal, and ultimately gene flow between populations. Finally, we detected high genetic diversity across populations in regions that have been climatically unstable since the last glacial maximum (LGM), and low genetic diversity across populations in regions that have been climatically stable since the LGM. Understanding how species have responded to climate change has the potential to inform management and conservation decisions of both ecological and economic concerns.     

Analyses of genetic ancestry enable key insights for molecular ecology

Gene flow and recombination in admixed populations produce genomes that are mosaic combinations of chromosome segments inherited from different source populations, that is, chromosome segments with different genetic ancestries. The statistical problem of estimating genetic ancestry from DNA sequence data has been widely studied, and analyses of genetic ancestry have facilitated research in molecular ecology and ecological genetics. In this review, we describe and compare different model‐based statistical methods used to infer genetic ancestry. We describe the conceptual and mathematical structure of these models and highlight some of their key differences and shared features. We then discuss recent empirical studies that use estimates of genetic ancestry to analyse population histories, the nature and genetic basis of species boundaries, and the genetic architecture of traits. These diverse studies demonstrate the breadth of applications that rely on genetic ancestry estimates and typify the genomics‐enabled research that is becoming increasingly common in molecular ecology. We conclude by identifying key research areas where future studies might further advance this field.     

Seasonal effects and fine‐scale population dynamics of Aedes taeniorhynchus,a major disease vector in the Galapagos Islands

Characterization of the fine‐scale population dynamics of the mosquito Aedes taeniorhynchus is needed to improve our understanding of its role as a disease vector in the Galapagos Islands. We used microsatellite data to assess the genetic structure of coastal and highland mosquito populations and patterns of gene flow between the two habitats through time on Santa Cruz Island. In addition, we assessed possible associations of mosquito abundance and genetic diversity with environmental variables. The coastal and highland mosquito populations were highly differentiated from each other all year round, with some gene flow detected only during periods of increased precipitation. The results support the hypothesis that selection arising from ecological differences between habitats is driving adaptation and divergence in A. taeniorhynchus, and maintaining long‐term genetic differentiation of the populations against gene flow. The highland and lowland populations may constitute an example of incipient speciation in progress. Highland populations were characterized by lower observed heterozygosity and allelic richness, suggesting a founder effect and/or lower breeding site availability in the highlands. A lack of reduction in genetic diversity over time in highland populations suggests that they survive dry periods as dormant eggs. Association between mosquito abundance and precipitation was strong in the highlands, whereas tide height was the main factor affecting mosquito abundance on the coast. Our findings suggests differences in the infection dynamics of mosquito‐borne parasites in the highlands compared to the coast, and a higher risk of mosquito‐driven disease spread across these habitats during periods of increased precipitation.     

Population Dynamic and Genetic Consequences of Spatial Density-Dependent Dispersal in Patchy Populations

Predictions about sex-specific, spatial density-dependent dispersal and their demographic and genetic consequences were tested in experimental populations of root voles (Microtus oeconomus). Each population consisted of two demes inhabiting equal-sized habitat patches imbedded in a barren matrix area. We used a neutral two-allele allozyme marker to monitor gene flow. Initially, the two demes were genetically distinct and had different densities so that the size of a high-density deme (genotype bb) was four times larger than that of a low-density deme (genotype aa). The sex-specific dispersal pattern was in accordance with our prediction. Male dispersal was unconditional on deme-specific densities, and the majority of the first-generation males became dispersed from both demes, whereas female dispersal was strongly density dependent, so that dispersal took place exclusively from the high-density to the low-density deme. The demographic implication of this dispersal pattern was that the initial density difference between the demes was quickly canceled out. We built a mathematical model that predicted that the initially rare allele (a) would increase in frequency given the dispersal pattern, and this was supported by our experimental data. This result relies mostly on the density-independent male-dispersal strategy, which presumably stems from inbreeding avoidance. Our study highlights the importance of incorporating sex-specific dispersal strategies in population genetic models. Sex-biased dispersal may act as a deterministic force counteracting the tendency for stochastic loss of alleles in small and fragmented populations.