Cryptic population genetic structure: the number of inferred clusters depends on sample size

Clustering methods have been used extensively to unravel cryptic population genetic structure. We investigated the effect of the number of individuals sampled in each location on the resulting number of clusters. Our study was motivated by recent results in Arabidopsis thaliana: studies in which more than one individual was sampled per location apparently have led to a much higher number of clusters than studies where only one individual was sampled in each location, as is generally done in this species. We show, using computer simulations and microsatellite data in A. thaliana, that the number of sampled individuals indeed has a strong impact on the number of resulting clusters. This effect is smaller if the sampled populations have a hierarchical structure. In most cases, sampling 5–10 individuals per population should be enough. The results argue for abandoning the concept of ‘accessions’ in partially selfing organisms.     

The program structure does not reliably recover the correct population structure when sampling is uneven: subsampling and new estimators alleviate the problem

Inferences of population structure and more precisely the identification of genetically homogeneous groups of individuals are essential to the fields of ecology, evolutionary biology and conservation biology. Such population structure inferences are routinely investigated via the program structure implementing a Bayesian algorithm to identify groups of individuals at Hardy–Weinberg and linkage equilibrium. While the method is performing relatively well under various population models with even sampling between subpopulations, the robustness of the method to uneven sample size between subpopulations and/or hierarchical levels of population structure has not yet been tested despite being commonly encountered in empirical data sets. In this study, I used simulated and empirical microsatellite data sets to investigate the impact of uneven sample size between subpopulations and/or hierarchical levels of population structure on the detected population structure. The results demonstrated that uneven sampling often leads to wrong inferences on hierarchical structure and downward‐biased estimates of the true number of subpopulations. Distinct subpopulations with reduced sampling tended to be merged together, while at the same time, individuals from extensively sampled subpopulations were generally split, despite belonging to the same panmictic population. Four new supervised methods to detect the number of clusters were developed and tested as part of this study and were found to outperform the existing methods using both evenly and unevenly sampled data sets. Additionally, a subsampling strategy aiming to reduce sampling unevenness between subpopulations is presented and tested. These results altogether demonstrate that when sampling evenness is accounted for, the detection of the correct population structure is greatly improved.     

Recommendations for utilizing and reporting population genetic analyses: the reproducibility of genetic clustering using the program structure

Reproducibility is the benchmark for results and conclusions drawn from scientific studies, but systematic studies on the reproducibility of scientific results are surprisingly rare. Moreover, many modern statistical methods make use of ‘random walk’ model fitting procedures, and these are inherently stochastic in their output. Does the combination of these statistical procedures and current standards of data archiving and method reporting permit the reproduction of the authors' results? To test this, we reanalysed data sets gathered from papers using the software package structure to identify genetically similar clusters of individuals. We find that reproducing structure results can be difficult despite the straightforward requirements of the program. Our results indicate that 30% of analyses were unable to reproduce the same number of population clusters. To improve this, we make recommendations for future use of the software and for reporting structure analyses and results in published works.     

Identifying the number of population clusters with structure: problems and solutions

The program structure has been used extensively to understand and visualize population genetic structure. It is one of the most commonly used clustering algorithms, cited over 11 500 times in Web of Science since its introduction in 2000. The method estimates ancestry proportions to assign individuals to clusters, and post hoc analyses of results may indicate the most likely number of clusters, or populations, on the landscape. However, as has been shown in this issue of Molecular Ecology Resources by Puechmaille ( 2016 ), when sampling is uneven across populations or across hierarchical levels of population structure, these post hoc analyses can be inaccurate and identify an incorrect number of population clusters. To solve this problem, Puechmaille ( 2016 ) presents strategies for subsampling and new analysis methods that are robust to uneven sampling to improve inferences of the number of population clusters.     

Determining population structure and hybridization for two iris species

Identifying processes that promote or limit gene flow can help define the ecological and evolutionary history of a species. Furthermore, defining those factors that make up “species boundaries” can provide a definition of the independent evolutionary trajectories of related taxa. For many species, the historic processes that account for their distribution of genetic variation remain unresolved. In this study, we examine the geographic distribution of genetic diversity for two species of Louisiana Irises, Iris brevicaulis and Iris fulva. Specifically, we asked how populations are structured and if population structure coincides with potential barriers to gene flow. We also asked whether there is evidence of hybridization between these two species outside Louisiana hybrid zones. We used a genotyping‐by‐sequencing approach and sampled a large number of single nucleotide polymorphisms across these species' genomes. Two different population assignment methods were used to resolve population structure in I. brevicaulis; however, there was considerably less population structure in I. fulva. We used a species tree approach to infer phylogenies both within and between populations and species. For I. brevicaulis, the geography of the collection locality was reflected in the phylogeny. The I. fulva phylogeny reflected much less structure than detected for I. brevicaulis. Lastly, combining both species into a phylogenetic analysis resolved two of six populations of I. brevicaulis that shared alleles with I. fulva. Taken together, our results suggest major differences in the level and pattern of connectivity among populations of these two Louisiana Iris species.     

The phylogeographical and management implications of genetic population structure in the imperiled snuffbox mussel, Epioblasma triquetra (Bivalvia: Unionidae)

Seven populations of the imperiled snuffbox mussel, Epioblasma triquetra , were sampled from across the central basin of North America. Samples were genotyped using 15 microsatellite DNA loci, and maternal history was inferred using mitochondrial DNA (mtDNA) cytochrome  c oxidase subunit-I (COI) sequences. Populations in the Clinch and St Francis rivers were quite distinct in their mtDNA. The population in the St Francis River had a unique, fixed haplotype. Among a suite of haplotypes, the population in the Clinch River had two unique haplotypes of common ancestry. The other populations were dominated by a common haplotype, which also occurred in the Clinch River population. Analysis of DNA microsatellites revealed much greater divergences and showed significant genetic structure between populations in the formerly glaciated regions. Divergence has occurred between the populations, as evidenced by moderate to high fixation indices ( F _ST and R _ST values) and nearly perfect assignment tests. These results indicate the occurrence of three glacial refugia for E. triquetra : the Tennessee River, rivers south of the Ozark Crest, and the lower Ohio River drainage near the confluence with the Mississippi. Populations in the lower Ohio River were likely to be responsible for the postglacial reinvasion into formerly glaciated regions, and into the upper Tennessee River drainage. The population of the St Francis River may constitute a distinct taxonomic entity. Conservation efforts, if necessary for this imperiled species, should not mix populations.  © 2008 The Linnean Society of London, Biological Journal of the Linnean Society , 2008, 93 , 371–384.     

Bayesian clustering algorithms ascertaining spatial population structure: a new computer program and a comparison study

On the basis of simulated data, this study compares the relative performances of the Bayesian clustering computer programs structure , geneland , geneclust and a new program named tess . While these four programs can detect population genetic structure from multilocus genotypes, only the last three ones include simultaneous analysis from geographical data. The programs are compared with respect to their abilities to infer the number of populations, to estimate membership probabilities, and to detect genetic discontinuities and clinal variation. The results suggest that combining analyses using tess and structure offers a convenient way to address inference of spatial population structure.     

Sunshine versus gold: The effect of population age on genetic structure of an invasive mosquito

The genetic diversity and structure of invasive species are affected by the time since invasion, but it is not well understood how. We compare likely the oldest populations of Aedes aegypti in continental North America with some of the newest to illuminate the range of genetic diversity and structure that can be found within the invasive range of this important disease vector. Aedes aegypti populations in Florida have probably persisted since the 1600‐1700s, while populations in southern California derive from new invasions that occurred in the last 10 years. For this comparison, we genotyped 1,193 individuals from 28 sites at 12 highly variable microsatellites and a subset of these individuals at 23,961 single nucleotide polymorphisms (SNPs). This is the largest sample analyzed for genetic structure for either region, and it doubles the number of southern California populations previously analyzed. As predicted, the older populations (Florida) showed fewer indicators of recent founder effect and bottlenecks; in particular, these populations have dramatically higher genetic diversity and lower genetic structure. Geographic distance and driving distance were not good predictors of genetic distance in either region, especially southern California. Additionally, southern California had higher levels of genetic differentiation than any comparably sized documented region throughout the worldwide distribution of the species. Although population age and demographic history are likely driving these differences, differences in climate and transportation practices could also play a role.     

Evolution of population genetic structure of the British roe deer by natural and anthropogenic processes (Capreolus capreolus)

Human influence typically impacts on natural populations of conservation interest. These interactions are varied and sometimes complex, and may be negative and unintended or associated with conservation and management strategy. Understanding the details of how these interactions influence and are influenced by natural evolutionary processes is essential to the development of effective conservation strategies. In this study, we investigate a species in Britain that has experienced both negative impact through overhunting in historical times and management efforts through culls and translocations. At the same time, there are regional populations that have been less affected by human influence. We use mtDNA and nuclear microsatellite DNA markers to investigate patterns of connectivity and diversity and find multiple insular populations in Britain that probably evolved within the Holocene (when the habitat was free of ice). We identify three concurrent processes. First, surviving indigenous populations show highly provincial patterns of philopatry, maintaining and generating population structure on a small geographic scale. Second, founder populations into habitat extirpated of native populations have expanded, but remained largely insular. Third, introductions into established populations generate some admixture. We discuss the implications for the evolution of diversity of the integration of natural processes with anthropogenic influences on population size and distribution.     

widgetcon: A website and program for quick conversion among common population genetic data formats

One of the most tedious steps in genetic data analyses is the reformatting data generated with one program for use with other applications. This conversion is necessary because comprehensive evaluation of the data may be based on different algorithms included in diverse software, each requiring a distinct input format. A platform‐independent and freely available program or a web‐based tool dedicated to such reformatting can save time and efforts in data processing. Here, we report widgetcon , a website and a program which has been developed to quickly and easily convert among various molecular data formats commonly used in phylogenetic analysis, population genetics, and other fields. The web‐based service is available at https://www.widgetcon.net . The program and the website convert the major data formats in four basic steps in less than a minute. The resource will be a useful tool for the research community and can be updated to include more formats and features in the future.     

Using AFLP markers and the Geneland program for the inference of population genetic structure

The use of dominant markers such as amplified fragment length polymorphism (AFLP) for population genetics analyses is often impeded by the lack of appropriate computer programs and rarely motivated by objective considerations. The point of the present note is twofold: (i) we describe how the computer program Geneland designed to infer population structure has been adapted to deal with dominant markers; and (ii) we use Geneland for numerical comparison of dominant and codominant markers to perform clustering. AFLP markers lead to less accurate results than bi-allelic codominant markers such as single nucleotide polymorphisms (SNP) markers but this difference becomes negligible for data sets of common size (number of individuals n≥100, number of markers L≥200). The latest Geneland version (3.2.1) handling dominant markers is freely available as an R package with a fully clickable graphical interface. Installation instructions and documentation can be found on http://www2.imm.dtu.dk/~gigu/Geneland.     

Microsatellite DNA analysis of northern pike (Esox lucius L.) populations: insights into the genetic structure and demographic history of a genetically depauperate species

The northern pike Esox lucius L. is a freshwater fish exhibiting pronounced population subdivision and low genetic variability. However, there is limited knowledge on phylogeographical patterns within the species, and it is not known whether the low genetic variability reflects primarily current low effective population sizes or historical bottlenecks. We analysed six microsatellite loci in ten populations from Europe and North America. Genetic variation was low, with the average number of alleles within populations ranging from 2.3 to 4.0 per locus. Genetic differentiation among populations was high (overall θ_ST = 0.51; overall ρ_ST = 0.50). Multidimensional scaling analysis of genetic distances between populations and spatial analysis of molecular variance suggested a single phylogeographical race within the sampled populations from northern Europe, whereas North American and southern European populations were highly distinct. A population from Ireland was monomorphic at all loci, presumably reflecting founder events associated with introduction of the species to the island in the sixteenth century. Bayesian analysis of demographic parameters showed differences in θ (a product of effective population size and mutation rate) among populations from large and small water bodies, but the relative differences in θ were smaller than expected, which could reflect population subdivision within the larger water bodies. Finally, the analyses showed drastic population declines on a time scale of several thousand years within European populations, which we ascribe to either glacial bottlenecks or postglacial founder events.  © 2005 The Linnean Society of London, Biological Journal of the Linnean Society , 2005, 84 , 91–101.     

VNTR loci reveal differentiation between and structure within populations of the eastern barred bandicoot Perameles gunnii

The Eastern Barred Bandicoot Perameles gunnii has declined in abundance within mainland south-eastern Australia, to a relict wild population of less than 100 individuals in Hamilton, Victoria. It is more common, but is also declining in Tasmania. Genomic DN A variability was compared within and between surviving populations of P. gunnii using variable number of tandem repeat (VNTR) markers in one of two ways. First, average percentage differences (APDs) were determined between profiles for two VNTR probe—endonuclease combinations. Secondly, because one of these combinations revealed two multiallelic VNTR loci, genotypes were assigned and analysed for homogeneity of allele frequencies among subpopulations, for deviation of heterozygosity from Hardy-Weinberg equilibrium within populations and for genetic structuring among individuals from different subpopulations. The results of both the APD and defined locus approaches showed consistent trends within and between populations. Genetic variability was higher among mainland P. gunnii than in Tasmanian populations (higher APDs, number of alleles, and heterozygosity at one locus), despite the known decline and subdivision of the Hamilton population. Eleven per cent of the variability detected in Hamilton was attributed to genetic differentiation between east and west subdivisions of the population. Departure from random mating indicating local inbreeding within collecting localities was evident for one locus in both north and south Tasmania, particularly at one locality. AH alleles at both loci were unique to either Hamilton or Tasmanian P. gunnii. The initial captive colony contains high heterozygosity for these loci. It is concluded that VNTR markers can be of benefit for use in studies of population differentiation and for conservation management.     

Detecting populations in the 'ambiguous' zone: kinship-based estimation of population structure at low genetic divergence

Identifying population structure is one of the most common and important objectives of spatial analyses using population genetic data. Population structure is detected either by rejecting the null hypothesis of a homogenous distribution of genetic variation, or by estimating low migration rates. Issues arise with most current population genetic inference methods when the genetic divergence is low among putative populations. Low levels of genetic divergence may be as a result of either high ongoing migration or historic high migration but no current, ongoing migration. We direct attention to recent developments in the use of the tempo-spatial distribution of closely related individuals to detect population structure or estimate current migration rates. These 'kinship-based' approaches complement more traditional population-based genetic inference methods by providing a means to detect population structure and estimate current migration rates when genetic divergence is low. However, for kinship-based methods to become widely adopted, formal estimation procedures applicable to a range of species life histories are needed.     

Environmental and topographic variables shape genetic structure and effective population sizes in the endangered Yosemite toad

Aim Describing the landscape variables that accurately reflect how environmental and topographic variations affect population connectivity and demography is a major goal of landscape genetics and conservation biology. However, few landscape genetics studies have quantified the relationships between landscape variables and effective population size (N_e), although N_e is a key conservation and population genetics parameter. In this study, I estimated genetic structure and effective population sizes in the Yosemite toad (Bufo canorus) and tested for associations with environmental and geographic variables. Location Yosemite National Park, California, USA. Methods I estimated F_ST, D_ps and N_e using 10 microsatellite loci amplified from 781 individuals from 24 populations. I used three landscape variables (environmental variation, topography and slope) to generate geographic distance models and a series of regression analyses to identify the variables that contributed to genetic structure in this species. I also tested for correlations between N_e and a suite of variables, including geographic and genetic isolation, habitat suitability, elevation, temperature and precipitation. Results I found substantial variation in genetic distances between populations (F_ST = 0.004–0.396, D_ps = 0.045–0.839) and in effective population sizes (N_e = 9–52). Environmental variation and slope played important roles in explaining variation in genetic distances, and precipitation variables were significantly correlated with N_e. Main conclusions These results show that environmental and topographic variables are both important for understanding population connectivity in B. canorus and provide some of the first evidence, in any species, for a link between environmental variables and effective population size.     

Experimental habitat fragmentation increases linkage disequilibrium but does not affect genetic diversity or population structure in the Amazonian liverwort Radula flaccida

Habitat fragmentation increases the migration distances among remnant populations, and is predicted to play a significant role in altering both demographic and genetic processes. Nevertheless, few studies have evaluated the genetic consequences of habitat fragmentation in light of information about population dynamics in the same set of organisms. In a 10,000-km(2) experimentally fragmented landscape of rainforest reserves in central Amazonia, we examine patterns of genetic variation (amplified fragment length polymorphisms, AFLPs) in the epiphyllous (e.g. leaf-inhabiting) liverwort Radula flaccida Gott. Previous demographic work indicates that colonization rates in this species are significantly reduced in small forest reserves. We scored 113 polymorphic loci in 86 individuals representing five fragmented and five experimentally unmanipulated populations. Most of the variation (82%) in all populations was harboured at the smallest (400 m(2)) sampling unit. The mean ((+/-) SD) within-population genetic diversity (Nei's), of forest remnants (0.412 +/- 0.2) was indistinguishable from continuous (0.413 +/- 0.2) forests. Similarly, F(ST) was identical among small (1- and 10-ha) and large (> or = 100-ha) reserves (0.19 and 0.18, respectively), but linkage disequilibrium between pairs of loci was significantly elevated in fragmented populations relative to those in continuous forests. These results illustrate that inferences regarding the long-term viability of fragmented populations based on neutral marker data alone must be viewed with caution, and underscore the importance of jointly evaluating information on both genetic structure and demography. Second, multilocus analyses may be more sensitive to the effects of fragmentation in the short term, although the effects of increasing linkage disequilibrium on population viability remain uncertain.     

spagedi: a versatile computer program to analyse spatial genetic structure at the individual or population levels

spag e d i version 1.0 is a software primarily designed to characterize the spatial genetic structure of mapped individuals or populations using genotype data of codominant markers. It computes various statistics describing genetic relatedness or differentiation between individuals or populations by pairwise comparisons and tests their significance by appropriate numerical resampling. spag e d i is useful for: (i) detecting isolation by distance within or among populations and estimating gene dispersal parameters; (ii) assessing genetic relatedness between individuals and its actual variance, a parameter of interest for marker based inferences of quantitative inheritance; (iii) assessing genetic differentiation among populations, including the case of haploids or autopolyploids.     

Contrasting genetic patterns and population histories in three threatened redfin species (Cyprinidae) from the Olifants River system, western South Africa

Differentiation occurred in different areas of the same river system in Pseudobarbus phlegethon and two sister species of the serrated redfin lineage ( Barbus calidus and Barbus erubescens ) of the Western Cape region of South Africa. In an analysis of 27 allozyme loci a deep divergence was found within P. phlegethon between the Olifants and Doring catchments (seven fixed allelic differences; 0·338 <  D  < 0·366; F _ST = 0·925). In contrast, speciation in the serrated redfin lineage occurred within the Doring catchment between B. calidus and B. erubescens (one fixed allelic difference; 0·008 <  D  < 0·052; F _ST = 0·760). Different niche preferences, behaviour and morphology probably played an important role in the formation of these contrasting patterns of genetic structure and suggest that B. calidus may have been better able to disperse between the Olifants and Doring catchments (no fixed allelic differences; 0 <  D  < 0·040; F _ST = 0·281) than P. phlegethon .     

Phylogeography and population genetic structure of the European roe deer in Switzerland following recent recolonization

In the early 1800s, the European roe deer (Capreolus capreolus) was probably extirpated from Switzerland, due to overhunting and deforestation. After a federal law was enacted in 1875 to protect lactating females and young, and limiting the hunting season, the roe deer successfully recovered and recolonized Switzerland. In this study, we use mitochondrial DNA and nuclear DNA markers to investigate the recolonization and assess contemporary genetic structure in relation to broad topographic features, in order to understand underlying ecological processes, inform future roe deer management strategies, and explore the opportunity for development of forensic traceability tools. The results concerning the recolonization origin support natural, multidirectional immigration from neighboring countries. We further demonstrate that there is evidence of weak genetic differentiation within Switzerland among topographic regions. Finally, we conclude that the genetic data support the recognition of a single roe deer management unit within Switzerland, within which there is a potential for broad‐scale geographic origin assignment using nuclear markers to support law enforcement.     

Microsatellite variation and population genetic structure of the red throat emperor on the Great Barrier Reef

Analysis (using three analytical approaches) of eight microsatellite markers from six locations in three geographic regions of the Great Barrier Reef (GBR), including populations that differed in demographic characteristics, showed no evidence of genetic stock structure in the red throat emperor Lethrinus miniatus. Measures of inter‐population differentiation were non‐significant (P ≥ 0·67). Using a Bayesian clustering approach, ‘admixture’ was detected (mean alpha values >1) with allele frequencies for each of the locations sampled being correlated equally with allele frequencies from all locations sampled. The number of populations (K) identified was one, based on the estimates of the probability of the data at various K values (K = 1, 2, 3, … 6). Additionally, alpha values did not stabilize to relatively constant values in any of the Bayesian analyses performed, indicating that there was no real genetic structure between locations. Analysis of genetic variation as detected by analysis of molecular variance (AMOVA) indicated that almost all of the variance in the data (99·74%, P ≤ 0·023) was within populations, rather than among populations (0·15%, P ≤ 0·176) or amongst regions sampled (0·10%, P ≤ 0·247) on the GBR. F_st statistics identified four individual loci having statistically significant differentiation among populations, but these were only related to one out of 12 pair‐wise comparisons where populations differed demographically. Given these results (albeit using neutral markers), together with the capacity of adults and larvae to be mobile between reefs on the inter‐connected GBR, it is considered unlikely that L. miniatus populations exist as distinct genetic stocks in the GBR. It is therefore not possible, using neutral markers, to reject the null hypothesis that the fishery be managed as a single panmictic stock.