Widespread Gene Flow Research Articles

Connectivity in grey reef sharks (Carcharhinus amblyrhynchos) determined using empirical and simulated genetic data.

Grey reef sharks (Carcharhinus amblyrhynchos) can be one of the numerically dominant high order predators on pristine coral reefs, yet their numbers have declined even in the highly regulated Australian Great Barrier Reef (GBR) Marine Park. Knowledge of both large scale and fine scale genetic connectivity of grey reef sharks is essential for their effective management, but no genetic data are yet available. We investigated grey reef shark genetic structure in the GBR across a 1200 km latitudinal gradient, comparing empirical data with models simulating different levels of migration. The empirical data did not reveal any genetic structuring along the entire latitudinal gradient sampled, suggesting regular widespread dispersal and gene flow of the species throughout most of the GBR. Our simulated datasets indicate that even with substantial migrations (up to 25% of individuals migrating between neighboring reefs) both large scale genetic structure and genotypic spatial autocorrelation at the reef scale were maintained. We suggest that present migration rates therefore exceed this level. These findings have important implications regarding the effectiveness of networks of spatially discontinuous Marine Protected Areas to protect reef sharks.

The diversification of Heliconius butterflies: what have we learned in 150 years?

Research into Heliconius butterflies has made a significant contribution to evolutionary biology. Here, we review our understanding of the diversification of these butterflies, covering recent advances and a vast foundation of earlier work. Whereas no single group of organisms can be sufficient for understanding life's diversity, after years of intensive study, research into Heliconiushas addressed a wide variety of evolutionary questions. We first discuss evidence for widespread gene flow betweenHeliconius species and what this reveals about the nature of species. We then address the evolution and diversity of warning patterns, both as the target of selection and with respect to their underlying genetic basis. The identification of major genes involved in mimetic shifts, and homology at these loci between distantly related taxa, has revealed a surprising predictability in the genetic basis of evolution. In the final sections, we consider the evolution of warning patterns, and Heliconius diversity more generally, within a broader context of ecological and sexual selection. We consider how different traits and modes of selection can interact and influence the evolution of reproductive isolation.

Nuclear genetic variation across the range of ponderosa pine (Pinus ponderosa): Phylogeographic, taxonomic and conservation implications

Ponderosa pine (Pinus ponderosa) is among the most broadly distributed conifer species of western North America, where it possesses considerable ecological, esthetic, and commercial value. It exhibits complicated patterns of morphological and genetic variation, suggesting that it may be in the process of differentiating into distinct regional lineages. A robust analysis of genetic variation across the ponderosa pine complex is necessary to ensure the effectiveness of management and conservation efforts given the species’ large distribution, the existence of many isolated disjunct populations, and the potential susceptibility of some populations to climate change and other threats. We used highly polymorphic nuclear microsatellite markers and isozyme markers from 3113 trees in 104 populations to assess genetic variation and structure across the geographic range of ponderosa pine. The results reveal pervasive inbreeding and patterns of genetic diversity consistent with the hypothesis that ponderosa existed in small, as-yet-undetected Pleistocene glacial refugia north of southern Arizona and New Mexico. The substructuring of genetic variation within the species complex was consistent with its division into two varieties, with genetic clusters within varieties generally associated with latitudinal zones. The analyses indicate widespread gene flow and/or recent common ancestry among genetic clusters within varieties, but not between varieties. Isolated disjunct populations had lower genetic variation by some measures and greater genetic differentiation than main-range populations. These results should be useful for decision-making and conservation planning related to this widespread and important species.

An investigation of genetic population structure in blue crabs, Callinectes sapidus, using nuclear gene sequences

Spatial and temporal genetic heterogeneity in the sequences of nuclear protein-coding genes were investigated in populations of the blue crab, Callinectes sapidus. Haplotype frequencies of these novel markers were determined for adult crabs collected along 300 km in the northern Gulf of Mexico in two different years (2010 and 2011), as well as for megalopal recruits collected in 2010. Tests of genetic differentiation among all locations and between locations spanning known genetic breaks in other species were conducted. In addition, samples from distant locations within the range of C. sapidus were used to assess genetic divergence on a broader geographic scale. Significant between-year differences were found for adults at one location and near significant spatial differentiation was found across northern Gulf of Mexico locations in 2010. These results suggest that although the large population sizes and meroplanktonic life history of blue crabs promote widespread gene flow on a regional scale, genetic composition can change over just one year. Substantial divergence between the northern Gulf of Mexico and Venezuela suggests the possibility that temporal shifts in haplotype frequencies could result from variation in the rate of immigration from genetically distinct source populations. The possible effects of the Deepwater Horizon Oil Spill and attendant fisheries closures during the sampling phase of this study are also considered.

The population genetics of the western purple-crowned fairy-wren (Malurus coronatus coronatus), a declining riparian passerine

We investigate the population genetic structure of the declining western subspecies of the purple-crowned fairy-wren (Malurus coronatus coronatus) in order to guide conservation management recommendations for this riparian habitat specialist. Our analysis of multilocus microsatellite data, from 79 individuals sampled from across the species’ range, indicates that M. c. coronatus occurs as genetically differentiated subpopulations that correspond to catchment boundaries or expansive gaps in habitat along waterways. The genetic similarity of large populations of fairy-wrens on four catchments (Fitzroy, Durack, Drysdale and Victoria) indicates widespread recent gene flow, whereas the high genetic distinctiveness of the Bindoola and Isdell catchments may reflect the current geographic isolation of these smaller populations. Genetic differentiation of these smaller geographically isolated populations affirms the negative effect that habitat degradation and fragmentation can have on population connectivity. A regional-scale approach to conservation with a focus on preventing degradation and enhancing connectivity may be critical to safeguard the persistence of M. c. coronatus subpopulations.

Molecular epidemiology ofRabbitHaemorrhagicDiseaseVirus inAustralia: when one became many

Rabbit Haemorrhagic Disease Virus (RHDV) was introduced into Australia in 1995 as a biological control agent against the wild European rabbit (Oryctolagus cuniculus). We evaluated its evolution over a 16-year period (1995-2011) by examining 50 isolates collected throughout Australia, as well as the original inoculum strains. Phylogenetic analysis of capsid protein VP60 sequences of the Australian isolates, compared with those sampled globally, revealed that they form a monophyletic group with the inoculum strains (CAPM V-351 and RHDV351INOC). Strikingly, despite more than 3000 rereleases of RHDV351INOC since 1995, only a single viral lineage has sustained its transmission in the long-term, indicative of a major competitive advantage. In addition, we find evidence for widespread viral gene flow, in which multiple lineages entered individual geographic locations, resulting in a marked turnover of viral lineages with time, as well as a continual increase in viral genetic diversity. The rate of RHDV evolution recorded in Australia -4.0 (3.3-4.7)×10(-3) nucleotide substitutions per site per year - was higher than previously observed in RHDV, and evidence for adaptive evolution was obtained at two VP60 residues. Finally, more intensive study of a single rabbit population (Turretfield) in South Australia provided no evidence for viral persistence between outbreaks, with genetic diversity instead generated by continual strain importation.

Patterns of genetic diversity in the rare Erigeron lemmoni and comparison with its more widespread congener, Erigeron arisolius (Asteraceae)

Analysis of levels and patterns of genetic variation in a rare species is important for determining whether genetic factors associated with small population size, such as genetic drift or inbreeding, may be negatively affecting a species. In this study, we compared estimates of genetic diversity and patterns of population genetic structure in a rare cliff endemic, Erigeron lemmonii, to those of a widespread congener, E. arisolius. Our goals were to assess whether rarity and small population size have negatively affected levels of genetic diversity in E. lemmonii and to identify genetic threats that may limit the ability of E. lemmonii to persist. Levels of observed and expected heterozygosity and allelic richness in E. lemmonii were approximately 60 % of those found in E. arisolius. After correcting for null alleles, inbreeding coefficients in both species of Erigeron were very small, suggesting that both species are highly outcrossing and may demonstrate self-incompatibility. Patterns of genetic structure in both species revealed almost no population substructuring, indicating that widespread gene flow is occurring within each species. Because we found no evidence for inbreeding or a genetic bottleneck in E. lemmonii, it is likely that the species’ lower genetic diversity may be the result of genetic drift. Because E. lemmoni exists in a single population, no other populations exist to bolster population size or genetic diversity in the event of declines; thus, conservation efforts should focus on seed collection from as many individuals as possible to protect against possible future losses of genetic diversity. We also recommend continued monitoring of both population size and genetic diversity in E. lemmonii to ensure the species’ long-term persistence and viability.

How immunogenetically different are domestic pigs from wild boars: a perspective from single-nucleotide polymorphisms of 19 immunity-related candidate genes

The coexistence of wild boars and domestic pigs across Eurasia makes it feasible to conduct comparative genetic or genomic analyses for addressing how genetically different a domestic species is from its wild ancestor. To test whether there are differences in patterns of genetic variability between wild and domestic pigs at immunity-related genes and to detect outlier loci putatively under selection that may underlie differences in immune responses, here we analyzed 54 single-nucleotide polymorphisms (SNPs) of 19 immunity-related candidate genes on 11 autosomes in three pairs of wild boar and domestic pig populations from China, Iberian Peninsula, and Hungary. Our results showed no statistically significant differences in allele frequency and heterozygosity across SNPs between three pairs of wild and domestic populations. This observation was more likely due to the widespread and long-lasting gene flow between wild boars and domestic pigs across Eurasia. In addition, we detected eight coding SNPs from six genes as outliers being under selection consistently by three outlier tests (BayeScan2.1, FDIST2, and Arlequin3.5). Among four non-synonymous outlier SNPs, one from TLR4 gene was identified as being subject to positive (diversifying) selection and three each from CD36, IFNW1, and IL1B genes were suggested as under balancing selection. All of these four non-synonymous variants were predicted as being benign by PolyPhen-2. Our results were supported by other independent lines of evidence for positive selection or balancing selection acting on these four immune genes (CD36, IFNW1, IL1B, and TLR4). Our study showed an example applying a candidate gene approach to identify functionally important mutations (i.e., outlier loci) in wild and domestic pigs for subsequent functional experiments.

Co-Circulation and Persistence of Genetically Distinct Saffold Viruses, Denmark

Co-Circulation and Persistence of Genetically Distinct Saffold Viruses, Denmark

Small geographic range but not panmictic: how forests structure the endangered Point Arena mountain beaver (Aplodontia rufa nigra)

The landscape genetics framework is typically applied to broad regions that occupy only small portions of a species' range. Rarely is the entire range of a taxon the subject of study. We examined the landscape genetic struc- ture of the endangered Point Arena mountain beaver (Apl- odontia rufa nigra), whose isolated geographic range is found in a restricted (85 km 2 ) but heterogenous region in California. Based on its diminutive range we may predict widespread gene flow and a relatively weak role for land- scape variation in defining genetic structure. We used skin, bone, tissue and noninvasively collected hair samples to describe genetic substructure and model gene flow. We examined spatial partitioning of multilocus DNA genotypes and mitochondrial DNA haplotypes. We identified 3 groups from microsatellite data, all of which had low estimates of effective population size consistent with significant tests for historical bottlenecks. We used least-cost-path analysis with the microsatellites to examine how vegetation type affects gene flow in a landscape genetics framework. Gene flow was best predicted by a model with ''Forest'' as the most permeable, followed by ''Riparian''. Agricultural lands demonstrated the highest resistance. MtDNA data revealed only two haplotypes: one was represented in all 57 individ- uals that occurred north of the east-west flowing Garcia River. South of the river, however, both haplotypes occurred, often at the same site suggesting that the river may have affected historical patterns of genetic divergence.

Statistical phylogeographic tests of competing 'Lake Carpentaria hypotheses' in the mouth-brooding freshwater fish, Glossamia aprion (Apogonidae)

Glacial cycles during the Pleistocene reduced sea levels and created new land connections in northern Australia, where many currently isolated rivers also became connected via an extensive paleo-lake system, ‘Lake Carpentaria’. However, the most recent period during which populations of freshwater species were connected by gene flow across Lake Carpentaria is debated: various ‘Lake Carpentaria hypotheses’ have been proposed. Here, we used a statistical phylogeographic approach to assess the timing of past population connectivity across the Carpentaria region in the obligate freshwater fish, Glossamia aprion. Results for this species indicate that the most recent period of genetic exchange across the Carpentaria region coincided with the mid- to late Pleistocene, a result shown previously for other freshwater and diadromous species. Based on these findings and published studies for various freshwater, diadromous and marine species, we propose a set of ‘Lake Carpentaria’ hypotheses to explain past population connectivity in aquatic species: (1) strictly freshwater species had widespread gene flow in the mid- to late Pleistocene before the last glacial maximum; (2) marine species were subdivided into eastern and western populations by land during Pleistocene glacial phases; and (3) past connectivity in diadromous species reflects the relative strength of their marine affinity.

Matrix quality and habitat configuration interactively determine functional connectivity in a widespread bush cricket at a small spatial scale

Unlike rare or specialised species, widespread abundant species have often been neglected when studying effects of habitat fragmentation. However, recently, it was shown that in the widespread abundant bush cricket Pholidoptera griseoaptera gene flow becomes restricted when the share of suitable habitat dropped below a threshold of 20% at the landscape scale. Here, using the same highly fragmented landscape, we studied the impact of habitat configuration and matrix quality on genetic variation and population differentiation of P. griseoaptera at a small spatial scale. We investigated four clusters of three populations that were either disconnected or connected and had either low quality (arable land) or high quality (grassland) matrix. The number of alleles was significantly lower in disconnected than in connected clusters, irrespective of matrix quality. Genetic differentiation was equally high in the two disconnected clusters and in the connected cluster with low quality matrix (G ST ≥ 0.030; D ≥ 0.082), whereas it was significantly reduced when connected habitats were embedded in a high quality grassland matrix (G ST = 0.004; D = 0.011). Analyses of least-cost paths showed that grassy landscape elements in fact represent high quality matrix, but that linear grassy margins are costly for dispersal. The effect of habitat configuration on genetic diversity may be explained by lower effective population sizes in disconnected habitats. The fact that only the connected populations in high quality matrix were not differentiated indicates that landscape management should simultaneously consider habitat configuration and matrix quality to effectively promote small and dispersal-limited species, also at small spatial scales.

Comparative phylogeography of four Indo-Pacific moray eel species (Muraenidae) reveals comparable ocean-wide genetic connectivity despite five-fold differences in available adult habitat

Most coral reef fishes have non-migratory adults and depend on a pelagic larval stage for dispersal. Species with long pelagic larval duration (PLD) can have tremendous dispersal potential and thus display little geographic-genetic differentiation among reef habitats. Restricted adult niche breadth due to habitat specialization can have the opposite effect of fragmenting populations and increasing geographic-genetic differentiation. If long PLD suffices to ensure widespread gene flow among reef populations, we predict similar geographic-genetic homogene- ity within species whose adults differ in niche breadth. We tested this hypothesis using a compar- ative phylogeographic study of 4 sympatric moray eel species that differ in the amount of available habitat within their reported ranges. We generated molecular genetic data for Echidna nebulosa (N = 79) and Gymnomuraena zebra (N = 67) to measure geographic-genetic structure within these species, whose adult habitat is very restricted for moray eels, and compared these results to iden- tical measurements previously published for habitat generalists Gymnothorax undulatus and Gymnothorax flavimarginatus. These 4 species share an ocean-wide distribution with adults occu- pying the same reefs; however, adults of E. nebulosa and G. zebra are restricted to shallow waters and occupy only 20% of the area occupied by the Gymnothorax species. Mitochondrial (632 bp of cytochrome b and 596 bp of cytochrome oxidase I) genomic sequences revealed high genetic vari- ation (h = 0.995 to 0.998) and low geographic-genetic differentiation (pairwise ΦST < 0.07 and not significant) for each species across 22 000 km of the Indo-Pacific. Nuclear genomic sequences (420 bp of RAG-1 and 746 bp of RAG-2) demonstrated 16 to 25 haplotypes per marker within each species with minimal geographic-genetic differentiation among populations. This suggests that in cosmopolitan and highly dispersive species such as morays, larval life history can ensure wide- spread gene flow despite a 5-fold difference in the habitat breadth occupied by adult populations.

Rotavirus A genotype P[4]G2: Genetic diversity and reassortment events among strains circulating in Brazil between 2005 and 2009

Group A rotaviruses (RV-A) are the leading cause of severe gastroenteritis in infants and young children worldwide. Due to the epidemiologic complexity of RV-A, especially in developing countries, it is important to determine which genotypes are circulating, principally after the introduction in March 2006 of the monovalent (P[8]G1) Rotarix® vaccine in Brazil by the National Immunization Program. In Phase III trials with Rotarix®, the prevalence of genotype P[4]G2 was extremely low, and therefore, evaluation of heterotypic immunization against this genotype was performed by meta-analysis statistics tests. Different studies have shown the re-emergence of genotype P[4]G2 in Brazil, since 2005, as well as in other countries, suggesting that it could be a continental phenomenon related to the temporal variability in the genotype's naturally occurring distribution. It is important to note that genotype P[4]G2 does not share VP4 or VP7 antigens with the vaccine strain. Therefore, we performed a phylogenetic analysis based on VP4 (VP8), VP7, VP6, and NSP4 genes of RV-A genotype P[4]G2 samples isolated from the five regions of Brazil between 2005 and 2009. This study revealed that different genetic variants of RV-A genotype P[4]G2 circulated in Brazil between 2005 and 2009, and that this variability is determined mainly by: occurrence of point mutations; reassortment events; and widespread global gene flow. The results obtained in this study are important to our understanding of the epidemiology and evolution of RV-A genotype P[4]G2 and demonstrate the importance of continuous monitoring and molecular characterization of RV-A strains circulating in human and animal populations.

The role of altitude and associated habitat stability in determining patterns of population genetic structure in two species of Atalophlebia (Ephemeroptera: Leptophlebiidae)

Summary 1. Previous studies have identified lowland areas as barriers to gene flow (dispersal) between distinct mountain ranges in montane species of aquatic insects. In this study, we investigated the population genetic structure of two closely related Atalophlebia (mayfly) species inhabiting lowland areas of south-east Queensland, Australia, with the expectation of widespread gene flow throughout the low-altitude environment and associated homogeneous genetic structure. 2. In particular, we asked whether species with lower-altitude distributions demonstrate greater spatial distribution of mtDNA (COI) alleles than the upland species studied previously. This pattern would be expected if good dispersal ability is associated with population persistence in these drought-prone habitats. 3. The two species demonstrated contrasting genetic population structure. Atalophlebia sp. AV13 D revealed strong population structure, with populations on each side of the low-altitude area isolated from each other for a long time (c.350 kya), and the presence of an isolation-by-distance pattern over relatively small geographical distances (<40 km). In contrast, Atalophlebia sp. AV13 A was panmictic at the scale investigated (≤160 km), with no history of past population fragmentation. 4. Examination of sample distribution along the altitudinal gradient reveals that Atalophlebia sp. AV13 D may have a more upland distribution (associated with greater habitat stability) than previously supposed, while Atalophlebia sp. AV13 A inhabits more xeric lowland areas, where freshwater habitats are less stable. We consequently hypothesise that these contrasting genetic population structures result from differences in habitat stability along the altitudinal gradient, only species with good dispersal ability being able to persist in unstable habitats. These findings may be applicable to other regions of the globe where habitat instability is associated with altitudinal gradients.

Anthropogenic and natural disturbance lead to differing patterns of gene flow in the Rocky Mountain tailed frog, Ascaphus montanus

Understanding the influence of habitat alteration on population structure and persistence is critical for effective conservation strategies. Timber harvest and wildfire are two of the most prevalent disturbances across temperate forests, yet the long-term effects of these two forces on population connectivity have rarely been studied. We studied populations of the Rocky Mountain tailed frog (Ascaphus montanus) across landscapes that have experienced either timber harvest or broad scale fires. Rocky Mountain tailed frogs generally require forested habitat, and are a species of concern in managed forests. We used landscape genetic techniques to test the explanatory power of alternative paths of connectivity across both burned and harvested forests and identified topographical and climatic variables that significantly influenced gene flow. We found that timber harvest and fire led to different patterns of genetic connectivity. Widespread terrestrial gene flow was maintained across previously burned areas and connectivity was primarily limited by solar radiation. In contrast, gene flow among populations in managed forests followed riparian corridors, presumably to avoid the loss of cover due to timber harvest. Gene flow along riparian corridors in Rocky Mountain tailed frogs differs from a population of the closely related coastal tailed frog (Ascaphus truei), for which gene flow appears to occur primarily over land. These results suggest management of Rocky Mountain tailed frogs should focus on maintaining riparian buffer zones, which may be less effective for the coastal species. Thus, consideration of landscape differences is essential even for conservation of closely related, morphologically similar species.

Dispersal and gene flow in the habitat-forming kelp, Ecklonia radiata: relative degrees of isolation across an east - west coastline

Characterising patterns of dispersal and gene flow in habitat-forming organisms is becoming a focal concern for conservation and management strategies as anthropogenic impacts drive change in coastal ecosystems. Here, we use six microsatellite markers to characterise dispersal and gene flow across the South Australian distribution of the habitat-forming kelp Ecklonia radiata. Populations of E. radiata on subtidal reefs in South Australia were highly genetically structured on large (100s of km, FST = 0.211) and small (10s of km, FST = 0.042) spatial scales with the extent of differentiation positively correlated with geographic distances among populations. Neither the presence of oceanic currents nor intervening rocky reef habitats appeared to facilitate widespread gene flow. There was a trend for island populations to be more genetically differentiated from those on the mainland and to have slightly greater levels of heterozygosity than mainland populations. Our results show relatively low dispersal and gene flow suggesting that recovery following kelp loss may be slow. Such information not only provides insights into relative rates of recovery, but may also identify which populations may be best used for propagation and restoration efforts.

Longitudinal patterns of genetic diversity and larval density of the riverine caddisfly Hydropsyche orientalis (Trichoptera)

Small local populations may have low levels of genetic diversity, although widespread gene flow may counteract genetic drift and maintain high local diversity. At a larger spatial scale, a relationship between population size and genetic diversity could have important effects on the longitudinal genetic patterns in riverine corridors. We examined the genetic structure of 15 subpopulations of the caddisfly Hydropsyche orientalis in four rivers within the Natori River system in Japan using 78 polymorphic RAPD loci. Six subpopulations were fragmented by reservoirs and thus may be smaller, genetically isolated, or both. Pairwise FST values were generally high throughout the study, and a higher mean value across the reservoirs (mean=0.219) relative to unfragmented streams 0.157) indicated reduced gene flow. Genetic diversity was not consequently reduced in these subpopulations; however, genetic diversity was positively correlated with larval density at all sites (percentage of polymorphic loci, %P = 46.3–59.1; expected heterozygosity Hexp = 0.163–0.225). The longitudinal distribution of genetic diversity exhibited a unimodal pattern, with a peak in middle reaches. Correlation with larval density was a significant factor, although the presence of cryptic species as well as unimodally distributed environmental variables (e.g., habitat heterogeneity) provide alternative hypotheses. We suggest that studies examining longitudinal patterns of genetic diversity will provide a better understanding of the principal mechanisms that create and maintain biodiversity in rivers and streams.

Local genetic structure in a North American epiphytic lichen, Ramalina menziesii (Ramalinaceae)

Epiphytic lichens possess unique life history traits that can have conflicting effects on genetic structure: symbiotic mutualism between a fungus with its algal or cyanobacterial photobiont, association with a host plant, and ability to reproduce sexually and asexually. Our study species, Ramalina menziesii, has small ascospores that can facilitate long-distance gene movement, and it is capable of clonal reproduction. The goals of this study are to test whether different haplotypes were differentially distributed across host plant species, to look for evidence of asexual vs. sexual reproduction, and to assess the local genetic structure of the population. We sampled individuals from multiple trees of three oak species in four lichen subpopulations within a savanna ecosystem. Using DNA sequence data from four fungal nuclear loci, we found no tendency for host specialization. Alleles were randomly distributed across subpopulations. The frequency of multilocus genotypes was consistent with a randomly mating population. Sexual reproduction involving relichenization appeared to be the predominant mode of reproduction of R. menziesii at this study site. We found no significant local genetic structure suggesting widespread gene flow at the local scale. The genetic structure of this lichen is comparable to that of widely distributed epiphytic plants.

Widespread gene flow and high genetic variability in populations of water volesArvicola terrestrisin patchy habitats

Theory predicts that the impact of gene flow on the genetic structure of populations in patchy habitats depends on its scale and the demographic attributes of demes (e.g. local colony sizes and timing of reproduction), but empirical evidence is scarce. We inferred the impact of gene flow on genetic structure among populations of water voles Arvicola terrestris that differed in average colony sizes, population turnover and degree of patchiness. Colonies typically consisted of few reproducing adults and several juveniles. Twelve polymorphic microsatellite DNA loci were examined. Levels of individual genetic variability in all areas were high (H(O) = 0.69-0.78). Assignments of juveniles to parents revealed frequent dispersal over long distances. The populations showed negative F(IS) values among juveniles, F(IS) values around zero among adults, high F(ST) values among colonies for juveniles, and moderate, often insignificant, F(ST) values for parents. We inferred that excess heterozygosity within colonies reflected the few individuals dispersing from a large area to form discrete breeding colonies. Thus pre-breeding dispersal followed by rapid reproduction results in a seasonal increase in differentiation due to local family groups. Genetic variation was as high in low-density populations in patchy habitats as in populations in continuous habitats used for comparison. In contrast to most theoretical predictions, we found that populations living in patchy habitats can maintain high levels of genetic variability when only a few adults contribute to breeding in each colony, when the variance of reproductive success among colonies is likely to be low, and when dispersal between colonies exceeds nearest-neighbour distances.