The use of approximate Bayesian computation in conservation genetics and its application in a case study on yellow-eyed penguins

Evolution and disappearance of sympatric Coregonus albula in a changing environment-A case study of the only remaining population pair in Sweden

During the past 50 years, Fennoscandian populations of spring-spawning Baltic cisco (Coregonus albula), sympatric to common autumn-spawners, have declined or disappeared; for example, three out of four known spring-spawning populations in Sweden are regarded as extinct. Over the same period, the climate has changed and populations have been subject to other anthropogenic stressors. We compared historic (1960s) and recent (1990-2000s) morphological data from the still-existent sympatric cisco populations in Lake Fegen, Sweden. Phenotypic changes were found for spring-spawners making them more similar to the sympatric autumn-spawners that had remained virtually unchanged. Based on results for other salmoniform fishes, a phenotypically plastic response to increased temperature during early development appears unlikely. The recent material was also analyzed with microsatellite markers; long-term effective population size in spring-spawners was estimated to be about 20 times lower than autumn-spawners, with signs of long-term gene flow in both directions and a recent genetic bottleneck in spring-spawners. We suggest the change toward a less distinct phenotype in spring-spawners to reflect a recent increase in gene flow from autumn-spawners. Time since divergence was estimated to only c. 1,900 years (95% CI: 400-5,900), but still the Fegen populations represent the most morphologically and genetically distinct sympatric populations studied. Consequently, we hypothesize that less distinct population pairs can be even younger and that spring-spawning may have repeatedly evolved and disappeared in several lakes since the end of the last glaciation, concurrent with changed environmental conditions.

Identification and evaluation of a core microsatellite panel for use in white-tailed deer (Odocoileus virginianus)

Background

Microsatellite loci have been used extensively over the past two decades to study the genetic characteristics of non-model species. The ease of microsatellite development and ability to adapt markers from related species has led to the proliferation of available markers for many commonly studied species. Because it is often infeasible to genotype individuals across all available loci, researchers generally rely on subsets of markers. Marker choice can bias inferences made using disparate suites of loci. This has been a primary motivation for efforts to identify uniform marker panels. Here, we use the geographic distribution of previous studies to identify microsatellite loci for white-tailed deer (Odocoileus virginianus) with the potential for widespread use, and we evaluate the effectiveness of this panel in a portion of the range where few previous studies have been conducted. The purpose was to consolidate the numerous genetic resources for this species into a manageable panel and to provide a uniform methodology that improves comparisons between past and future studies.

Results

We reviewed microsatellite panels from 58 previous or ongoing projects and identified 106 candidate loci. We developed a multiplex protocol and evaluated the efficacy of 17 of the most commonly used loci using 720 DNA samples collected from the Mid-Atlantic region of the United States of America. Amplification errors were detected in six of these loci. The 11 remaining loci were highly polymorphic, exhibited low frequencies of null alleles, and were easy to interpret with the aid of allele binning software.

Conclusions

The development of broadly-applicable, core microsatellite panels has the potential to improve repeatability and comparative ability for commonly studied species. The properties of the consolidated 11 microsatellite panel suggest that they are applicable for many common research objectives for white-tailed deer. The geographic distribution of previous studies using these markers provides a greater degree of confidence regarding the robustness to common sources of error related to amplification anomalies, such as null alleles, relative to loci with more limited use. While this does not replace further evaluation of genotyping errors, it does provide a common platform that benefits future research studies.

Electronic supplementary material

The online version of this article (10.1186/s12863-019-0750-z) contains supplementary material, which is available to authorized users.

Reduced representation sequencing detects only subtle regional structure in a heavily exploited and rapidly recolonizing marine mammal species

Next-generation reduced representation sequencing (RRS) approaches show great potential for resolving the structure of wild populations. However, the population structure of species that have shown rapid demographic recovery following severe population bottlenecks may still prove difficult to resolve due to high gene flow between subpopulations. Here, we tested the effectiveness of the RRS method Genotyping-By-Sequencing (GBS) for describing the population structure of the New Zealand fur seal (NZFS, Arctocephalus forsteri), a species that was heavily exploited by the 19th century commercial sealing industry and has since rapidly recolonized most of its former range from a few isolated colonies. Using 26,026 neutral single nucleotide polymorphisms (SNPs), we assessed genetic variation within and between NZFS colonies. We identified low levels of population differentiation across the species range (<1% of variation explained by regional differences) suggesting a state of near panmixia. Nonetheless, we observed subtle population substructure between West Coast and Southern East Coast colonies and a weak, but significant (p = 0.01), isolation-by-distance pattern among the eight colonies studied. Furthermore, our demographic reconstructions supported severe bottlenecks with potential 10-fold and 250-fold declines in response to Polynesian and European hunting, respectively. Finally, we were able to assign individuals treated as unknowns to their regions of origin with high confidence (96%) using our SNP data. Our results indicate that while it may be difficult to detect population structure in species that have experienced rapid recovery, next-generation markers and methods are powerful tools for resolving fine-scale structure and informing conservation and management efforts.

Genetic assessment of the effects of streamscape succession on coho salmon Oncorhynchus kisutch colonization in recently deglaciated streams

Measures of genetic diversity within and among populations and historical geomorphological data on stream landscapes were used in model simulations based on approximate Bayesian computation (ABC) to examine hypotheses of the relative importance of stream features (geomorphology and age) associated with colonization events and gene flow for coho salmon Oncorhynchus kisutch breeding in recently deglaciated streams (50-240 years b.p.) in Glacier Bay National Park (GBNP), Alaska. Population estimates of genetic diversity including heterozygosity and allelic richness declined significantly and monotonically from the oldest and largest to youngest and smallest GBNP streams. Interpopulation variance in allele frequency increased with increasing distance between streams (r = 0·435, P < 0·01) and was inversely related to stream age (r = -0·281, P < 0·01). The most supported model of colonization involved ongoing or recent (<10 generations before sampling) colonization originating from large populations outside Glacier Bay proper into all other GBNP streams sampled. Results here show that sustained gene flow from large source populations is important to recently established O. kisutch metapopulations. Studies that document how genetic and demographic characteristics of newly founded populations vary associated with successional changes in stream habitat are of particular importance to and have significant implications for, restoration of declining or repatriation of extirpated populations in other regions of the species' native range.

Differential influences of local subpopulations on regional diversity and differentiation for greater sage-grouse (Centrocercus urophasianus)

The distribution of spatial genetic variation across a region can shape evolutionary dynamics and impact population persistence. Local population dynamics and among-population dispersal rates are strong drivers of this spatial genetic variation, yet for many species we lack a clear understanding of how these population processes interact in space to shape within-species genetic variation. Here, we used extensive genetic and demographic data from 10 subpopulations of greater sage-grouse to parameterize a simulated approximate Bayesian computation (ABC) model and (i) test for regional differences in population density and dispersal rates for greater sage-grouse subpopulations in Wyoming, and (ii) quantify how these differences impact subpopulation regional influence on genetic variation. We found a close match between observed and simulated data under our parameterized model and strong variation in density and dispersal rates across Wyoming. Sensitivity analyses suggested that changes in dispersal (via landscape resistance) had a greater influence on regional differentiation, whereas changes in density had a greater influence on mean diversity across all subpopulations. Local subpopulations, however, varied in their regional influence on genetic variation. Decreases in the size and dispersal rates of central populations with low overall and net immigration (i.e. population sources) had the greatest negative impact on genetic variation. Overall, our results provide insight into the interactions among demography, dispersal and genetic variation and highlight the potential of ABC to disentangle the complexity of regional population dynamics and project the genetic impact of changing conditions.

An approximate Bayesian computation approach to parameter estimation in a stochastic stage-structured population model

Complex population processes may require equally complex models, which can lead to analytically intractable estimation problems. Approximate Bayesian computation (ABC) is a computational tool for parameter estimation in situations where likelihoods cannot be computed. Instead of using likelihoods, ABC methods quantify the similarities between an observed data set and repeated simulations from a model. A practical obstacle to implementing an ABC algorithm is selecting summary statistics and distance metrics that accurately capture the main features of the data. We demonstrate the application of a sequential Monte Carlo ABC sampler (ABC SMC) to parameter estimation of a general stochastic stage-structured population model with ongoing reproduction and heterogeneity in development and mortality. Individual variation in demographic traits has considerable consequences for population dynamics in many systems, but including it in a population model by explicitly allowing stage durations to follow a realistic distribution creates a complex model. We applied the ABC SMC to fit the model to a simulated representative data set with known underlying parameters to evaluate the performance of the algorithm. We also introduced a systematic method for selecting summary statistics and distance metrics, using simulated data and receiver operating characteristic (ROC) curves from classification theory. Evaluations suggest that the approach is promising for model inference in our example of realistic stage-structured population models.

Demographic factors shaped diversity in the two gene pools of wild common bean Phaseolus vulgaris L

Wild common bean (Phaseolus vulgaris L.) is distributed throughout the Americas from Mexico to northern Argentina. Within this range, the species is divided into two gene pools (Andean and Middle American) along a latitudinal gradient. The diversity of 24 wild common bean genotypes from throughout the geographic range of the species was described by using sequence data from 13 loci. An isolation-migration model was evaluated using a coalescent analysis to estimate multiple demographic parameters. Using a Bayesian approach, Andean and Middle American subpopulations with high percentage of parentages were observed. Over all loci, the Middle American gene pool was more diverse than the Andean gene pool (π(sil)=0.0089 vs 0.0068). The two subpopulations were strongly genetically differentiated over all loci (F(st)=0.29). It is estimated that the two current wild gene pools diverged from a common ancestor ∼111 000 years ago. Subsequently, each gene pool underwent a bottleneck immediately after divergence and lasted ∼40 000 years. The Middle American bottleneck population size was ∼46% of the ancestral population size, whereas the Andean was 26%. Continuous asymmetric gene flow was detected between the two gene pools with a larger number of migrants entering Middle American gene pool from the Andean gene pool. These results suggest that because of the complex population structure associated with the ancestral divergence, subsequent bottlenecks in each gene pool, gene pool-specific domestication and intense selection within each gene pool by breeders; association mapping would best be practised within each common bean gene pool.

Inferring recent historic abundance from current genetic diversity

Recent historic abundance is an elusive parameter of great importance for conserving endangered species and understanding the pre-anthropogenic state of the biosphere. The number of studies that have used population genetic theory to estimate recent historic abundance from contemporary levels of genetic diversity has grown rapidly over the last two decades. Such assessments often yield unexpectedly large estimates of historic abundance. We review the underlying theory and common practices of estimating recent historic abundance from contemporary genetic diversity, and critically evaluate the potential issues at various estimation steps. A general issue of mismatched spatio-temporal scales between the estimation itself and the objective of the estimation emerged from our assessment; genetic diversity-based estimates of recent historic abundance represent long-term averages, whereas the objective typically is an estimate of recent abundance for a specific population. Currently, the most promising approach to estimate the difference between recent historic and contemporary abundance requires that genetic data be collected from samples of similar spatial and temporal duration. Novel genome-enabled inference methods may be able to utilize additional information of dense genome-wide distributions of markers, such as of identity-by-descent tracts, to infer recent historic abundance from contemporary samples only.

Reconstruction of a beech population bottleneck using archival demographic information and Bayesian analysis of genetic data

Range expansion and contraction has occurred in the history of most species and can seriously impact patterns of genetic diversity. Historical data about range change are rare and generally appropriate for studies at large scales, whereas the individual pollen and seed dispersal events that form the basis of geneflow and colonization generally occur at a local scale. In this study, we investigated range change in Fagus sylvatica on Mont Ventoux, France, using historical data from 1838 to the present and approximate Bayesian computation (ABC) analyses of genetic data. From the historical data, we identified a population minimum in 1845 and located remnant populations at least 200 years old. The ABC analysis selected a demographic scenario with three populations, corresponding to two remnant populations and one area of recent expansion. It also identified expansion from a smaller ancestral population but did not find that this expansion followed a population bottleneck, as suggested by the historical data. Despite a strong support to the selected scenario for our data set, the ABC approach showed a low power to discriminate among scenarios on average and a low ability to accurately estimate effective population sizes and divergence dates, probably due to the temporal scale of the study. This study provides an unusual opportunity to test ABC analysis in a system with a well-documented demographic history and identify discrepancies between the results of historical, classical population genetic and ABC analyses. The results also provide valuable insights into genetic processes at work at a fine spatial and temporal scale in range change and colonization.

The complex analytical landscape of gene flow inference

Gene flow estimation is essential for characterizing local adaptation, speciation potential and connectivity among threatened populations. New model-based population genetic methods can resolve complex demographic histories, but many studies in fields such as landscape genetics continue to rely on simple rules of thumb focused on gene flow to explain patterns of spatial differentiation. Here, we show how methods that use gene genealogies can reveal cryptic demographic histories and provide better estimates of gene flow with other parameters that contribute to genetic variation across landscapes and seascapes. We advocate for the expanded use and development of methods that consider spatial differentiation as the product of multiple forces interacting over time, and caution against a routine reliance on post-hoc gene flow interpretations.

Reconstructing demographic events from population genetic data: the introduction of bumblebees to New Zealand

Four British bumblebee species (Bombus terrestris, Bombus hortorum, Bombus ruderatus and Bombus subterraneus) became established in New Zealand following their introduction at the turn of the last century. Of these, two remain common in the United Kingdom (B. terrestris and B. hortorum), whilst two (B. ruderatus and B. subterraneus) have undergone marked declines, the latter being declared extinct in 2000. The presence of these bumblebees in New Zealand provides an unique system in which four related species have been isolated from their source population for over 100 years, providing a rare opportunity to examine the impacts of an initial bottleneck and introduction to a novel environment on their population genetics. We used microsatellite markers to compare modern populations of B. terrestris, B. hortorum and B. ruderatus in the United Kingdom and New Zealand and to compare museum specimens of British B. subterraneus with the current New Zealand population. We used approximate Bayesian computation to estimate demographic parameters of the introduction history, notably to estimate the number of founders involved in the initial introduction. Species-specific patterns derived from genetic analysis were consistent with the predictions based on the presumed history of these populations; demographic events have left a marked genetic signature on all four species. Approximate Bayesian analyses suggest that the New Zealand population of B. subterraneus may have been founded by as few as two individuals, giving rise to low genetic diversity and marked genetic divergence from the (now extinct) UK population.

ABC as a flexible framework to estimate demography over space and time: some cons, many pros

The analysis of genetic variation to estimate demographic and historical parameters and to quantitatively compare alternative scenarios recently gained a powerful and flexible approach: the Approximate Bayesian Computation (ABC). The likelihood functions does not need to be theoretically specified, but posterior distributions can be approximated by simulation even assuming very complex population models including both natural and human-induced processes. Prior information can be easily incorporated and the quality of the results can be analysed with rather limited additional effort. ABC is not a statistical analysis per se, but rather a statistical framework and any specific application is a sort of hybrid between a simulation and a data-analysis study. Complete software packages performing the necessary steps under a set of models and for specific genetic markers are already available, but the flexibility of the method is better exploited combining different programs. Many questions relevant in ecology can be addressed using ABC, but adequate amount of time should be dedicated to decide among alternative options and to evaluate the results. In this paper we will describe and critically comment on the different steps of an ABC analysis, analyse some of the published applications of ABC and provide user guidelines.

Approximate Bayesian Computation (ABC) in practice

Understanding the forces that influence natural variation within and among populations has been a major objective of evolutionary biologists for decades. Motivated by the growth in computational power and data complexity, modern approaches to this question make intensive use of simulation methods. Approximate Bayesian Computation (ABC) is one of these methods. Here we review the foundations of ABC, its recent algorithmic developments, and its applications in evolutionary biology and ecology. We argue that the use of ABC should incorporate all aspects of Bayesian data analysis: formulation, fitting, and improvement of a model. ABC can be a powerful tool to make inferences with complex models if these principles are carefully applied.