Using simulations to evaluate Mantel‐based methods for assessing landscape resistance to gene flow

Mantel‐based tests have been the primary analytical methods for understanding how landscape features influence observed spatial genetic structure. Simulation studies examining Mantel‐based approaches have highlighted major challenges associated with the use of such tests and fueled debate on when the Mantel test is appropriate for landscape genetics studies. We aim to provide some clarity in this debate using spatially explicit, individual‐based, genetic simulations to examine the effects of the following on the performance of Mantel‐based methods: (1) landscape configuration, (2) spatial genetic nonequilibrium, (3) nonlinear relationships between genetic and cost distances, and (4) correlation among cost distances derived from competing resistance models. Under most conditions, Mantel‐based methods performed poorly. Causal modeling identified the true model only 22% of the time. Using relative support and simple Mantel r values boosted performance to approximately 50%. Across all methods, performance increased when landscapes were more fragmented, spatial genetic equilibrium was reached, and the relationship between cost distance and genetic distance was linearized. Performance depended on cost distance correlations among resistance models rather than cell‐wise resistance correlations. Given these results, we suggest that the use of Mantel tests with linearized relationships is appropriate for discriminating among resistance models that have cost distance correlations <0.85 with each other for causal modeling, or <0.95 for relative support or simple Mantel r. Because most alternative parameterizations of resistance for the same landscape variable will result in highly correlated cost distances, the use of Mantel test‐based methods to fine‐tune resistance values will often not be effective.     

A statistical procedure to assess the significance level of barriers to gene flow

Although several methods are available to study the extent of isolation by distance （IBD） among natural populations, comparatively few exist to detect the presence of sharp genetic breaks in genetic distance datasets. In recent years, Monmonier＇s maximum-difference algorithm has been increasingly used by population geneticists. However, this method does not provide means to measure the statistical significance of such barriers, nor to determine their relative contribution to population differentiation with respect to IBD. Here, we propose an approach to assess the significance of genetic boundaries. The method is based on the calculation of a multiple regression from distance matrices, where binary matrices represent putative genetic barriers to test, in addition to geographic and genetic distances. Simulation results suggest that this method reliably detects the presence of genetic barriers, even in situations where IBD is also significant. We also illustrate the methodology by analyzing previously published datasets. Conclusions about the importance of genetic barriers can be misleading if one does not take into consideration their relative contribution to the overall genetic structure of species.     

Performance of individual vs. group sampling for inferring dispersal under isolation‐by‐distance

Models of isolation‐by‐distance formalize the effects of genetic drift and gene flow in a spatial context where gene dispersal is spatially limited. These models have been used to show that, at an appropriate spatial scale, dispersal parameters can be inferred from the regression of genetic differentiation against geographic distance between sampling locations. This approach is compelling because it is relatively simple and robust and has rather low sampling requirements. In continuous populations, dispersal can be inferred from isolation‐by‐distance patterns using either individuals or groups as sampling units. Intrigued by empirical findings where individual samples seemed to provide more power, we used simulations to compare the performances of the two methods in a range of situations with different dispersal distributions. We found that sampling individuals provide more power in a range of dispersal conditions that is narrow but fits many realistic situations. These situations were characterized not only by the general steepness of isolation‐by‐distance but also by the intrinsic shape of the dispersal kernel. The performances of the two approaches are otherwise similar, suggesting that the choice of a sampling unit is globally less important than other settings such as a study's spatial scale.     

DNA polymorphisms detect ancient barriers to gene flow in Basques

This work features the first district-by-district analysis of all provinces in the Iberian Peninsula with an autochthonous Basque population, and indicates the existence of genetic heterogeneity. The populations cluster in three groups arising from processes of genetic drift which probably occurred in pre-Mesolithic times, and were probably those which repopulated the southern areas of the Basque Country after the Last Glacial Maximum. It seems that from that period onwards, the population settled in three major groups (West Basques, Central Basques, and East Basques), along geographical axes which appear substantial in the maintaining of each population unit. This genetic structure is probably reflected in other aspects such as the existence of ancient tribes and the dialects of the Basque language, the boundaries of which may be related at origin and which are quite similar to those detected in this work. Our results indicate that the populations of the Basque Country are genetically close to other neighboring populations, such as that of Aragon, which may indicate an outgoing gene flow from the Basque area down the River Ebro towards the Mediterranean seaboard. While our short tandem repeat data suggest that population structure within the Basques dates back to the Mesolithic, our findings are also consistent with the hypothesis that patterns of modern European genetic diversity have been shaped mainly during the Neolithic.     

Isolation and gene flow: inferring the speciation history of European house mice

Inferring the history of isolation and gene flow during species differentiation can inform us on the processes underlying their formation. Following their recent expansion in Europe, two subspecies of the house mouse (Mus musculus domesticus and Mus musculus musculus) have formed a hybrid zone maintained by hybrid incompatibilities and possibly behavioural reinforcement, offering a good model of incipient speciation. We reconstruct the history of their divergence using an approximate Bayesian computation framework and sequence variation at 57 autosomal loci. We find support for a long isolation period preceding the advent of gene flow around 200,000 generations ago, much before the formation of the European hybrid zone a few thousand years ago. The duration of the allopatric episode appears long enough (74% of divergence time) to explain the accumulation of many post-zygotic incompatibilities expressed in the present hybrid zone. The ancient contact inferred could have played a role in mating behaviour divergence and laid the ground for further reinforcement. We suggest that both subspecies originally colonized the Middle East from the northern Indian subcontinent, domesticus settling on the shores of the Persian Gulf and musculus on those of the Caspian Sea. Range expansions during interglacials would have induced secondary contacts, presumably in Iran, where they must have also interacted with Mus musculus castaneus. Future studies should incorporate this possibility, and we point to Iran and its surroundings as a hot spot for house mouse diversity and speciation studies.     

Varying levels of clonality and ploidy create barriers to gene flow and challenges for conservation of an Australian arid‐zone ecosystem engineer,Acacia loderi

Acacia loderi, the ecosystem engineer of the endangered Acacia loderi Shrublands in arid eastern Australia, spans a persistent (> 15 000 year) but poorly studied landscape feature, the Darling River. We investigated the genetic structure of 19 stands of eight to > 1000 plants separated by < 300 km to test for variation in life histories between semi‐arid and arid stands to the east and west of the Darling River, respectively. Eight of nine stands east of the Darling were exclusively sexual, whereas most of those to the west were clonal. Three western stands were monoclonal, two were polyploid, and one was a diverse mix of diploid and triploid phenotypes. Bayesian analysis revealed a complex genetic structure within the western stands, whereas the eastern stands formed only two genetic clusters. Conservation of small stands may require augmentation of genotypic diversity. However, most genotypic diversity resides within the eastern stands. Although arid zone stands of A. loderi are not always clonal, clonality and polyploidy are more common in the arid west. Clear demarcation of life histories either side of the Darling River may reflect ancient or contemporary effects of physical disturbance associated with the river channel, or cryptic environmental differences, with sexual and asexual reproduction, respectively, at a selective premium in the semi‐arid east and arid west. The restricted distribution of clones and variation in clonality and polyploidy suggests that smaller stands may be vulnerable and warrant individual management.     

Connectivity of Caribbean coral populations: complementary insights from empirical and modelled gene flow

Understanding patterns of connectivity among populations of marine organisms is essential for the development of realistic, spatially explicit models of population dynamics. Two approaches, empirical genetic patterns and oceanographic dispersal modelling, have been used to estimate levels of evolutionary connectivity among marine populations but rarely have their potentially complementary insights been combined. Here, a spatially realistic Lagrangian model of larval dispersal and a theoretical genetic model are integrated with the most extensive study of gene flow in a Caribbean marine organism. The 871 genets collected from 26 sites spread over the wider Caribbean subsampled 45.8% of the 1900 potential unique genets in the model. At a coarse scale, significant consensus between modelled estimates of genetic structure and empirical genetic data for populations of the reef-building coral Montastraea annularis is observed. However, modelled and empirical data differ in their estimates of connectivity among northern Mesoamerican reefs indicating that processes other than dispersal may dominate here. Further, the geographic location and porosity of the previously described east-west barrier to gene flow in the Caribbean is refined. A multi-prong approach, integrating genetic data and spatially realistic models of larval dispersal and genetic projection, provides complementary insights into the processes underpinning population connectivity in marine invertebrates on evolutionary timescales.     

Sex chromosome turnovers and genetic drift: a simulation study

The recent advances of new genomic technologies have enabled the identification and characterization of sex chromosomes in an increasing number of nonmodel species, revealing that many plants and animals undergo frequent sex chromosome turnovers. What evolutionary forces drive these turnovers remains poorly understood, but it was recently proposed that drift might play a more important role than generally assumed. We analysed the dynamics of different types of turnovers using individual‐based simulations and show that when mediated by genetic drift, turnovers are usually easier to achieve than substitutions at neutral markers, but that their dynamics and relative likelihoods vary with the type of the resident and emergent sex chromosome system (XY and/or ZW) and the dominance relationships among the sex‐determining factors. Focusing on turnovers driven by epistatically dominant mutations, we find that drift‐mediated turnovers that preserve the heterogamety pattern are 2–4× more likely than those along which the heterogametic sex changes. This ratio nevertheless decreases along with effective population size and can even reverse in case of extreme polygyny. This can be attributed to a ‘drift‐induced’ selective force, known to influence transitions between male and female heterogamety, but which according to our study does not affect turnovers that preserve the heterogametic sex.     

A reassessment of explanations for discordant introgressions of mitochondrial and nuclear genomes

Hybridization is increasingly recognized as a significant evolutionary process, in particular because it can lead to introgression of genes from one species to another. A striking pattern of discordance in the amount of introgression between mitochondrial and nuclear markers exists such that substantial mitochondrial introgression is often found in combination with no or little nuclear introgression. Multiple mechanisms have been proposed to explain this discordance, including positive selection for introgressing mitochondrial variants, several types of sex‐biases, drift, negative selection against introgression in the nuclear genome, and spatial expansion. Most of these hypotheses are verbal, and have not been quantitatively evaluated so far. We use individual‐based, multilocus, computer simulations of secondary contact under a wide range of demographic and genetic scenarios to evaluate the ability of the different mechanisms to produce discordant introgression. Sex‐biases and spatial expansions fail to produce substantial mito‐nuclear discordance. Drift and nuclear selection can produce strong discordance, but only under a limited range of conditions. In contrast, selection on the mitochondrial genome produces strong discordance, particularly when dispersal rates are low. However, commonly used statistical tests have little power to detect this selection. Altogether, these results dismiss several popular hypotheses, and provide support for adaptive mitochondrial introgression.     

A Spatial Dirichlet Process Mixture Model for Clustering Population Genetics Data

Summary Identifying homogeneous groups of individuals is an important problem in population genetics. Recently, several methods have been proposed that exploit spatial information to improve clustering algorithms. In this article, we develop a Bayesian clustering algorithm based on the Dirichlet process prior that uses both genetic and spatial information to classify individuals into homogeneous clusters for further study. We study the performance of our method using a simulation study and use our model to cluster wolverines in Western Montana using microsatellite data.     

How mechanisms of habitat preference evolve and promote divergence with gene flow

Habitat preference may promote adaptive divergence and speciation, yet the conditions under which this is likely are insufficiently explored. We use individual‐based simulations to study the evolution and consequence of habitat preference during divergence with gene flow, considering four different underlying genetically based behavioural mechanisms: natal habitat imprinting, phenotype‐dependent, competition‐dependent and direct genetic habitat preference. We find that the evolution of habitat preference generally requires initially high dispersal, is facilitated by asymmetry in population sizes between habitats, and is hindered by an increasing number of underlying genetic loci. Moreover, the probability of habitat preference to emerge and promote divergence differs greatly among the underlying mechanisms. Natal habitat imprinting evolves most easily and can allow full divergence in parameter ranges where no divergence is possible in the absence of habitat preference. The reason is that imprinting represents a one‐allele mechanism of assortative mating linking dispersal behaviour very effectively to local selection. At the other extreme, direct genetic habitat preference, a two‐allele mechanism, evolves under restricted conditions only, and even then facilitates divergence weakly. Overall, our results indicate that habitat preference can be a strong reproductive barrier promoting divergence with gene flow, but that this is highly contingent on the underlying preference mechanism.     

Montane barriers to gene flow of Melanoplus bivittatus (Orthoptera: Acrididae)*

To determine the effectiveness of montane barriers to gene flow of the two-striped grasshopper, Melanoplus bivittatus, collections were made at four sites at the corners of a square location, each bisected by one of five Colorado mountains (total: five locations, 18 sites). The premise of this design is that samples from opposite sides of an effective barrier should have more genotypic differences than samples from the same side. Genotypic differences were measured for each grasshopper as 18 morphological measurements and six enzyme loci. Dissimilarities in morphology or enzyme loci were expressed as significant estimates of Mahalanobis squared distance (respectively, D² or G²) between two sample sites. The presence of foothills had no influence on the number of significant G² (age and sex classes combined) or significant D² (age and sex classes separated). For samples near the foothills, adult females had more significant D² than adult males. For samples near the two higher mountains, adult males had more significant D² for site samples from opposite sides. A combination of gene flow and selection is the most plausible explanation for most of these results.     

neogen: A tool to predict genetic effective population size (Ne) for species with generational overlap and to assist empirical Ne study design

Molecular genetic estimates of population effective size (N_e) lose accuracy and precision when insufficient numbers of samples or loci are used. Ideally, researchers would like to forecast the necessary power when planning their project. neogen (genetic N_e for Overlapping Generations) enables estimates of precision and accuracy in advance of empirical investigation and allows exploration of the power available in different user‐specified age‐structured sampling schemes. neogen provides a population simulation and genetic power analysis framework that simulates the demographics, genetic composition, and N_e, from species‐specific life history, mortality, population size, and genetic priors. neogen guides the user to establish a tractable sampling regime and to determine the numbers of samples and microsatellite or SNP loci required for accurate and precise genetic N_e estimates when sampling a natural population. neogen is useful at multiple stages of a study's life cycle: when budgeting, as sampling and locus development progresses, and for corroboration when empirical N_e estimates are available. The underlying model is applicable to a wide variety of iteroparous species with overlapping generations (e.g., mammals, birds, reptiles, long‐lived fishes). In this paper, we describe the neogen model, detail the workflow for the point‐and‐click software, and explain the graphical results. We demonstrate the use of neogen with empirical Australian east coast zebra shark (Stegostoma fasciatum) data. For researchers wishing to make accurate and precise genetic N_e estimates for overlapping generations species, neogen facilitates planning for sample and locus acquisition, and with existing empirical genetic N_e estimates neogen can corroborate population demographic and life history properties.     

Comparative landscape genetic analyses show a Belgian motorway to be a gene flow barrier for red deer (Cervus elaphus), but not wild boars (Sus scrofa)

While motorways are often assumed to influence the movement behaviour of large mammals, there are surprisingly few studies that show an influence of these linear structures on the genetic make-up of wild ungulate populations. Here, we analyse the spatial genetic structure of red deer (Cervus elaphus) and wild boars (Sus scrofa) along a stretch of motorway in the Walloon part of Belgium. Altogether, 876 red deer were genotyped at 13 microsatellite loci, and 325 wild boars at 14 loci. In the case of the red deer, different genetic clustering tools identified two genetic subpopulations whose borders matched the motorway well. Conversely, no genetic structure was identified in the case of the wild boar. Analysis of isolation-by-distance patterns of pairs of individuals on the same side and on different sides of the motorway also suggested that the road was a barrier to red deer, but not to wild boar movement. While telemetry studies seem to confirm that red deer are more affected by motorways than wild boar, the red deer sample size was also much larger than that of the wild boars. We therefore repeated the analysis of genetic structure in the red deer with randomly sub-sampled data sets of decreasing size. The power to detect the genetic structure using clustering methods decreased with decreasing sample size.     

Weak genetic structuring indicates ongoing gene flow across White-ruffed Manakin (<Emphasis Type="Italic">Corapipo altera</Emphasis>) populations in a highly fragmented Costa Rica landscape

We explored the effects of recent forest fragmentation on fine-scale patterns of population structuring and genetic diversity in populations of White-ruffed Manakins (Corapipo altera) inhabiting premontane forest fragments of varying size in southwestern Costa Rica. Habitat fragmentation is a major conservation concern for avian populations worldwide, but studies of the genetic effects of fragmentation on Neotropical birds are limited. We sampled 159 manakins from nine forest fragments of varying size within an 18 km radius, and genotyped these birds at 13 microsatellite loci. Bayesian clustering methods revealed that birds from all fragments comprised a single genetic population, and an MCMC approach showed that the fragments were likely to be at migration-drift equilibrium. F-statistics showed only modest levels of differentiation between forest fragments. We calculated allelic diversity indices for each fragment but found no correlation between genetic diversity and fragment size. These results suggest that manakins may retain substantial connectivity via inter-fragment dispersal despite habitat fragmentation.     

Genetic structure of Pinus henryi and Pinus tabuliformis: Natural landscapes as significant barriers to gene flow among populations

Mountains as natural barriers often have important effects on intraspecific genetic structure through restraining gene flow and enhancing differentiation among populations. While the Qinling and Daba mountains are considered significant geographic barriers, dividing China into temperate and subtropical regions, little is known about how this barrier influences the genetic patterns of sister species represented in distinct habitats. In this study, we analyzed genetic differentiation and the geographic boundary between Pinus henryi and Pinus tabuliformis using chloroplast microsatellite markers. Our data show high levels of among-population differentiation, consistent with the effects of historical demographic bottlenecks, local adaptation and climate effects. Three main geographic boundaries coinciding with mountain systems indicate natural landscapes, such as large rivers, and habitat loss caused by anthropogenic deforestation, are significant barriers to genetic exchange among populations. The divergence between populations in the eastern and western Qinling Mountains populations may possibly be ascribed to fragmentation driven by climate change. The genetic boundary of P. henryi and P. tabuliformis generally coincides with the previous morphological dividing line based on the unweighted pair group method using arithmetic averages and on spatial analysis of molecular variance.     

locuseater and shadowboxer: programs to optimize experimental design and multiplexing strategies for genetic mark–recapture

Genetic mark–recapture requires efficient methods of uniquely identifying individuals. ‘Shadows’ (individuals with the same genotype at the selected loci) become more likely with increasing sample size, and bias harvest rate estimates. Finding loci is costly, but better loci reduce analysis costs and improve power. Optimal microsatellite panels minimize shadows, but panel design is a complex optimization process. locuseater and shadowboxer permit power and cost analysis of this process and automate some aspects, by simulating the entire experiment from panel design to harvest rate estimation.     

A bivariate quantitative genetic model for a linear Gaussian trait and a survival trait

With the increasing use of survival models in animal breeding to address the genetic aspects of mainly longevity of livestock but also disease traits, the need for methods to infer genetic correlations and to do multivariate evaluations of survival traits and other types of traits has become increasingly important. In this study we derived and implemented a bivariate quantitative genetic model for a linear Gaussian and a survival trait that are genetically and environmentally correlated. For the survival trait, we considered the Weibull log-normal animal frailty model. A Bayesian approach using Gibbs sampling was adopted. Model parameters were inferred from their marginal posterior distributions. The required fully conditional posterior distributions were derived and issues on implementation are discussed. The two Weibull baseline parameters were updated jointly using a Metropolis-Hasting step. The remaining model parameters with non-normalized fully conditional distributions were updated univariately using adaptive rejection sampling. Simulation results showed that the estimated marginal posterior distributions covered well and placed high density to the true parameter values used in the simulation of data. In conclusion, the proposed method allows inferring additive genetic and environmental correlations, and doing multivariate genetic evaluation of a linear Gaussian trait and a survival trait.     

A general approach to mixed effects modeling of residual variances in generalized linear mixed models

We propose a general Bayesian approach to heteroskedastic error modeling for generalized linear mixed models (GLMM) in which linked functions of conditional means and residual variances are specified as separate linear combinations of fixed and random effects. We focus on the linear mixed model (LMM) analysis of birth weight (BW) and the cumulative probit mixed model (CPMM) analysis of calving ease (CE). The deviance information criterion (DIC) was demonstrated to be useful in correctly choosing between homoskedastic and heteroskedastic error GLMM for both traits when data was generated according to a mixed model specification for both location parameters and residual variances. Heteroskedastic error LMM and CPMM were fitted, respectively, to BW and CE data on 8847 Italian Piemontese first parity dams in which residual variances were modeled as functions of fixed calf sex and random herd effects. The posterior mean residual variance for male calves was over 40% greater than that for female calves for both traits. Also, the posterior means of the standard deviation of the herd-specific variance ratios (relative to a unitary baseline) were estimated to be 0.60 ± 0.09 for BW and 0.74 ± 0.14 for CE. For both traits, the heteroskedastic error LMM and CPMM were chosen over their homoskedastic error counterparts based on DIC values.