How useful are genomic data for predicting maladaptation to future climate?

Methods using genomic information to forecast potential population maladaptation to climate change or new environments are becoming increasingly common, yet the lack of model validation poses serious hurdles toward their incorporation into management and policy. Here, we compare the validation of maladaptation estimates derived from two methods-Gradient Forests (GFoffset) and the risk of non-adaptedness (RONA)-using exome capture pool-seq data from 35 to 39 populations across three conifer taxa: two Douglas-fir varieties and jack pine. We evaluate sensitivity of these algorithms to the source of input loci (markers selected from genotype-environment associations [GEA] or those selected at random). We validate these methods against 2- and 52-year growth and mortality measured in independent transplant experiments. Overall, we find that both methods often better predict transplant performance than climatic or geographic distances. We also find that GFoffset and RONA models are surprisingly not improved using GEA candidates. Even with promising validation results, variation in model projections to future climates makes it difficult to identify the most maladapted populations using either method. Our work advances understanding of the sensitivity and applicability of these approaches, and we discuss recommendations for their future use.

The immediate costs and long-term benefits of assisted gene flow in large populations.

With the genetic health of many plant and animal populations deteriorating due to climate change outpacing adaptation, interventions, such as assisted gene flow (AGF), may provide genetic variation necessary for populations to adapt to climate change. We ran genetic simulations to mimic different AGF scenarios in large populations and measured their outcomes on population-level fitness to determine circumstances in which it is worthwhile to perform AGF. In the absence of inbreeding depression, AGF was beneficial within a few generations only when introduced genotypes had much higher fitness than local individuals and traits affecting fitness were controlled by a few genes of large effect. AGF was harmful over short periods (e.g., first ∼10-20 generations) if there was strong outbreeding depression or introduced deleterious genetic variation. When the adaptive trait was controlled by many loci of small effect, the benefits of AGF took over 10 generations to realize-potentially too long for most climate-related management scenarios. The genomic integrity of the recipient population typically remained intact following AGF; the amount of genetic material from the donor population usually constituted no more of the recipient population's genome than the fraction of the population introduced. Significant genomic turnover (e.g., >50% replacement) only occurred when the selective advantage of the adaptive trait and translocation fraction were extremely high. Our results will be useful when adaptive management is used to maintain the genetic health and productivity of large populations under climate change.

Con el deterioro de la salud genética de muchas poblaciones de plantas y animales debido a la ventaja que le lleva el cambio climático a la adaptación, algunas intervenciones, como el flujo génico asistido (FGA), pueden proporcionar la variación genética necesaria para que las poblaciones se adapten al cambio climático. Simulamos diferentes escenarios de FGA aplicado en poblaciones grandes y medimos los resultados en la aptitud a nivel poblacional para determinar las circunstancias en las que merece la pena realizar FGA. Cuando no hubo depresión endogámica, el FGA produjo un beneficio en pocas generaciones sólo cuando se introdujeron genotipos que tenían una aptitud mucho mayor que los individuos locales y cuando unos cuantos genes de gran efecto controlaron los rasgos que afectaban a la aptitud. El flujo génico asistido fue dañino en periodos cortos (p.ej.: las primeras 10-20 generaciones) si existía una fuerte depresión exogámica o una variación genética deletérea introducida. Cuando muchos loci de pequeño efecto controlaron el rasgo adaptativo, los beneficios del FGA tardaron más de 10 generaciones en aparecer - un tiempo potencialmente muy largo para la mayoría de la gestión relacionada con el clima. La integridad genómica de la población receptora casi siempre permaneció intacta después del FGA; es decir, la cantidad de material genético de la población donante generalmente no constituyó más que la fracción de población introducida en el genoma de la población receptora. La rotación genómica significativa (p.ej.: reemplazos >50%) sólo ocurrió cuando la ventaja selectiva del rasgo adaptativo y la fracción de reubicación fueron extremadamente elevadas. Nuestros resultados serán útiles cuando se use la gestión adaptativa para mantener la salud genética y la productividad de las poblaciones grandes bajo el cambio climático.

Quantifying how constraints limit the diversity of viable routes to adaptation.

Convergent adaptation occurs at the genome scale when independently evolving lineages use the same genes to respond to similar selection pressures. These patterns of genetic repeatability provide insights into the factors that facilitate or constrain the diversity of genetic responses that contribute to adaptive evolution. A first step in studying such factors is to quantify the observed amount of repeatability relative to expectations under a null hypothesis. Here, we formulate a novel index to quantify the constraints driving the observed amount of repeated adaptation in pairwise contrasts based on the hypergeometric distribution, and then generalize this for simultaneous analysis of multiple lineages. This index is explicitly based on the probability of observing a given amount of repeatability by chance under a given null hypothesis and is readily compared among different species and types of trait. We also formulate an index to quantify the effective proportion of genes in the genome that have the potential to contribute to adaptation. As an example of how these indices can be used to draw inferences, we assess the amount of repeatability observed in existing datasets on adaptation to stress in yeast and climate in conifers. This approach provides a method to test a wide range of hypotheses about how different kinds of factors can facilitate or constrain the diversity of genetic responses observed during adaptive evolution.

Variation in recombination rate affects detection of outliers in genome scans under neutrality.

Genome scans can potentially identify genetic loci involved in evolutionary processes such as local adaptation and gene flow. Here, we show that recombination rate variation across a neutrally evolving genome gives rise to mixed sampling distributions of mean FST ( FST^ ), a common population genetic summary statistic. In particular, we show that in regions of low recombination the distribution of FST^ estimates has more variance and a longer tail than in more highly recombining regions. Determining outliers from the genome-wide distribution without taking local recombination rate into consideration may therefore increase the frequency of false positives in low recombination regions and be overly conservative in more highly recombining ones. We perform genome scans on simulated and empirical Drosophila melanogaster data sets and, in both cases, find patterns consistent with this neutral model. Similar patterns are observed for other summary statistics used to capture variation in the coalescent process. Linked selection, particularly background selection, is often invoked to explain heterogeneity in FST^ across the genome, but here we point out that even under neutrality, statistical artefacts can arise due to variation in recombination rate. Our results highlight a flaw in the design of genome-scan studies and suggest that without estimates of local recombination rate, interpreting the genomic landscape of any summary statistic that captures variation in the coalescent process will be very difficult.

No evidence of positive assortative mating for genetic quality in fruit flies.

In sexual populations, the effectiveness of selection will depend on how gametes combine with respect to genetic quality. If gametes with deleterious alleles are likely to combine with one another, deleterious genetic variation can be more easily purged by selection. Assortative mating, where there is a positive correlation between parents in a phenotype of interest such as body size, is often observed in nature, but does not necessarily reveal how gametes ultimately combine with respect to genetic quality itself. We manipulated genetic quality in fruit fly populations using an inbreeding scheme designed to provide an unbiased measure of mating patterns. While inbred flies had substantially reduced reproductive success, their gametes did not combine with those of other inbred flies more often than expected by chance, indicating a lack of positive assortative mating. Instead, we detected a negative correlation in genetic quality between parents, i.e. disassortative mating, which diminished with age. This pattern is expected to reduce the genetic variance for fitness, diminishing the effectiveness of selection. We discuss how mechanisms of sexual selection could produce a pattern of disassortative mating. Our study highlights that sexual selection has the potential to either increase or decrease genetic load.

Background selection and FST : Consequences for detecting local adaptation.

Background selection is a process whereby recurrent deleterious mutations cause a decrease in the effective population size and genetic diversity at linked loci. Several authors have suggested that variation in the intensity of background selection could cause variation in FST across the genome, which could confound signals of local adaptation in genome scans. We performed realistic simulations of DNA sequences, using recombination maps from humans and sticklebacks, to investigate how variation in the intensity of background selection affects FST and other statistics of population differentiation in sexual, outcrossing species. We show that, in populations connected by gene flow, Weir and Cockerham's (1984; Evolution, 38, 1358) estimator of FST is largely insensitive to locus-to-locus variation in the intensity of background selection. Unlike FST , however, dXY is negatively correlated with background selection. Moreover, background selection does not greatly affect the false-positive rate in FST outlier studies in populations connected by gene flow. Overall, our study indicates that background selection will not greatly interfere with finding the variants responsible for local adaptation.

Environmental stress does not increase the mean strength of selection.

A common intuition among evolutionary biologists and ecologists is that environmental stress will increase the strength of selection against deleterious alleles and among alternate genotypes. However, the strength of selection is determined by the relative fitness differences among genotypes, and there is no theoretical reason why these differences should be exaggerated as mean fitness decreases. We update a recent review of the empirical results pertaining to environmental stress and the strength of selection and find that there is no overall trend towards increased selection under stress, in agreement with other recent analyses of existing data. The majority of past studies measure the strength of selection by quantifying the decrease in fitness imposed by single or multiple mutations in different environments. However, selection rarely acts on one locus independently, and the strength of selection will be determined by variation across the whole genome. We used 20 inbred lines of Drosophila melanogaster to make repeated fitness measurements of the same genotypes in four different environments. This framework allowed us to determine the variation in fitness attributable to genotype across stressful environments and to calculate the opportunity for selection among these genotypes in each stress. Although we found significant decreases in mean fitness in our stressful environments, we did not find any significant differences in the strength of selection among any of the four measured environments. Therefore, in agreement with our updated review, we find no evidence for the oft-cited verbal model that stress increases the strength of selection.

Bioinformatically predicted deleterious mutations reveal complementation in the interior spruce hybrid complex.

BACKGROUND: Mutation load is expected to be reduced in hybrids via complementation of deleterious alleles. While local adaptation of hybrids confounds phenotypic tests for reduced mutation load, it may be possible to assess variation in load by analyzing the distribution of putatively deleterious alleles. Here, we use this approach in the interior spruce (Picea glauca x P. engelmannii) hybrid complex, a group likely to suffer from high mutation load and in which hybrids exhibit local adaptation to intermediate conditions. We used PROVEAN to bioinformatically predict whether non-synonymous alleles are deleterious, based on conservation of the position and abnormality of the amino acid change. RESULTS: As expected, we found that predicted deleterious alleles were at lower average allele frequencies than alleles not predicted to be deleterious. We were unable to detect a phenotypic effect on juvenile growth rate of the many rare alleles predicted to be deleterious. Both the proportion of alleles predicted to be deleterious and the proportion of loci homozygous for predicted deleterious alleles were higher in P. engelmannii (Engelmann spruce) than in P. glauca (white spruce), due to higher diversity and frequencies of rare alleles in Engelmann. Relative to parental species, the proportion of alleles predicted to be deleterious was intermediate in hybrids, and the proportion of loci homozygous for predicted deleterious alleles was lowest. CONCLUSION: Given that most deleterious alleles are recessive, this suggests that mutation load is reduced in hybrids due to complementation of deleterious alleles. This effect may enhance the fitness of hybrids.

Local Adaptation Interacts with Expansion Load during Range Expansion: Maladaptation Reduces Expansion Load.

The biotic and abiotic factors that facilitate or hinder species range expansions are many and complex. We examine the impact of two genetic processes and their interaction on fitness at expanding range edges: local maladaptation resulting from the presence of an environmental gradient and expansion load resulting from increased genetic drift at the range edge. Results from spatially explicit simulations indicate that the presence of an environmental gradient during range expansion reduces expansion load; conversely, increasing expansion load allows only locally adapted populations to persist at the range edge. Increased maladaptation reduces the speed of range expansion, resulting in less genetic drift at the expanding front and more immigration from the range center, therefore reducing expansion load at the range edge. These results may have ramifications for species being forced to shift their ranges because of climate change or other anthropogenic changes. If rapidly changing climate leads to faster expansion as populations track their shifting climatic optima, populations may suffer increased expansion load beyond previous expectations.

Finding the Genomic Basis of Local Adaptation: Pitfalls, Practical Solutions, and Future Directions.

Uncovering the genetic and evolutionary basis of local adaptation is a major focus of evolutionary biology. The recent development of cost-effective methods for obtaining high-quality genome-scale data makes it possible to identify some of the loci responsible for adaptive differences among populations. Two basic approaches for identifying putatively locally adaptive loci have been developed and are broadly used: one that identifies loci with unusually high genetic differentiation among populations (differentiation outlier methods) and one that searches for correlations between local population allele frequencies and local environments (genetic-environment association methods). Here, we review the promises and challenges of these genome scan methods, including correcting for the confounding influence of a species' demographic history, biases caused by missing aspects of the genome, matching scales of environmental data with population structure, and other statistical considerations. In each case, we make suggestions for best practices for maximizing the accuracy and efficiency of genome scans to detect the underlying genetic basis of local adaptation. With attention to their current limitations, genome scan methods can be an important tool in finding the genetic basis of adaptive evolutionary change.

Modern Approaches to Local Adaptation.

The study of local adaptation has been very important for the development of evolutionary biology broadly. This tradition continues; through new developments in diverse areas, we have more insight into the nature of the ongoing evolutionary process in a heterogeneous spatial context than ever before. This special issue follows from the 2014 Vice-Presidential Symposium of the American Society of Naturalists, bringing together outstanding researchers doing theoretical and empirical work, laboratory work, and fieldwork, with new data and new nodes of analysis. Through study of local adaptation, we learn about a tremendous source of biodiversity, both genetic and phenotypic, within and among species.

Reliable Detection of Loci Responsible for Local Adaptation: Inference of a Null Model through Trimming the Distribution of F(ST).

Loci responsible for local adaptation are likely to have more genetic differentiation among populations than neutral loci. However, neutral loci can vary widely in their amount of genetic differentiation, even over the same geographic range. Unfortunately, the distribution of differentiation--as measured by an index such as F(ST)--depends on the details of the demographic history of the populations in question, even without spatially heterogeneous selection. Many methods designed to detect F(ST) outliers assume a specific model of demographic history, which can result in extremely high false positive rates for detecting loci under selection. We develop a new method that infers the distribution of F(ST) for loci unlikely to be strongly affected by spatially diversifying selection, using data on a large set of loci with unknown selective properties. Compared to previous methods, this approach, called OutFLANK, has much lower false positive rates and comparable power, as shown by simulation.

A clever solution to a vexing problem.

F(ST) (as well as related measures such as G(ST)) has long been used both as a measure of the relative amount of genetic variation between populations and as an indicator of the amount of gene flow among populations. Unfortunately, F(ST) and its clones are also sensitive to mutation, particularly when the mutation rate per locus is greater than the migration rate among populations. Relatively high mutation rates cause estimates of F(ST) and G(ST) to be much lower than researchers sometimes expect, when migration rates are low in the studied species. Several recent suggestions for dealing with this problem have been unsatisfactory for one reason or another, and no general solution exists (if we are not to abandon these otherwise useful measures of differentiation). In an important article in this issue, Jinliang Wang (2015) shows that it is possible to identify whether the genetic markers in a given study are likely to give estimates of F(ST) that are strongly affected by mutation. The proposed test is simple and elegant, and with it, molecular ecologists can determine whether the F(ST) from their makers can be depended on for further inference about their species' genome and the demographic forces which shaped its patterns.

The relative power of genome scans to detect local adaptation depends on sampling design and statistical method.

Although genome scans have become a popular approach towards understanding the genetic basis of local adaptation, the field still does not have a firm grasp on how sampling design and demographic history affect the performance of genome scans on complex landscapes. To explore these issues, we compared 20 different sampling designs in equilibrium (i.e. island model and isolation by distance) and nonequilibrium (i.e. range expansion from one or two refugia) demographic histories in spatially heterogeneous environments. We simulated spatially complex landscapes, which allowed us to exploit local maxima and minima in the environment in 'pair' and 'transect' sampling strategies. We compared F(ST) outlier and genetic-environment association (GEA) methods for each of two approaches that control for population structure: with a covariance matrix or with latent factors. We show that while the relative power of two methods in the same category (F(ST) or GEA) depended largely on the number of individuals sampled, overall GEA tests had higher power in the island model and F(ST) had higher power under isolation by distance. In the refugia models, however, these methods varied in their power to detect local adaptation at weakly selected loci. At weakly selected loci, paired sampling designs had equal or higher power than transect or random designs to detect local adaptation. Our results can inform sampling designs for studies of local adaptation and have important implications for the interpretation of genome scans based on landscape data.

Evaluation of demographic history and neutral parameterization on the performance of FST outlier tests.

F(ST) outlier tests are a potentially powerful way to detect genetic loci under spatially divergent selection. Unfortunately, the extent to which these tests are robust to nonequilibrium demographic histories has been understudied. We developed a landscape genetics simulator to test the effects of isolation by distance (IBD) and range expansion on FST outlier methods. We evaluated the two most commonly used methods for the identification of F(ST) outliers (FDIST2 and BayeScan, which assume samples are evolutionarily independent) and two recent methods (FLK and Bayenv2, which estimate and account for evolutionary nonindependence). Parameterization with a set of neutral loci ('neutral parameterization') always improved the performance of FLK and Bayenv2, while neutral parameterization caused FDIST2 to actually perform worse in the cases of IBD or range expansion. BayeScan was improved when the prior odds on neutrality was increased, regardless of the true odds in the data. On their best performance, however, the widely used methods had high false-positive rates for IBD and range expansion and were outperformed by methods that accounted for evolutionary nonindependence. In addition, default settings in FDIST2 and BayeScan resulted in many false positives suggesting balancing selection. However, all methods did very well if a large set of neutral loci is available to create empirical P-values. We conclude that in species that exhibit IBD or have undergone range expansion, many of the published FST outliers based on FDIST2 and BayeScan are probably false positives, but FLK and Bayenv2 show great promise for accurately identifying loci under spatially divergent selection.

The WZA: A window-based method for characterizing genotype-environment associations.

Genotype-environment association (GEA) studies have the potential to identify the genetic basis of local adaptation in natural populations. Specifically, GEA approaches look for a correlation between allele frequencies and putatively selective features of the environment. Genetic markers with extreme evidence of correlation with the environment are presumed to be tagging the location of alleles that contribute to local adaptation. In this study, we propose a new method for GEA studies called the Weighted-Z Analysis (WZA) that combines information from closely linked sites into analysis windows in a way that was inspired by methods for calculating FST . Performing GEA methods in analysis windows has the advantage that it takes advantage of the increased linkage disequilibrium expected surrounding sites subject to local adaptation. We analyse simulations modelling local adaptation to heterogeneous environments to compare the WZA with existing methods. In the majority of cases we tested, the WZA either outperformed single-SNP (single nucleotide polymorphism)-based approaches or performed similarly. In particular, the WZA outperformed individual SNP approaches when a small number of individuals or demes were sampled. Particularly troubling, we found that some GEA methods exhibit very high false positive rates. We applied the WZA to previously published data from lodgepole pine and identified candidate loci that were identified in the original study alongside numerous loci that were not found in the original study.

The genetic architecture of repeated local adaptation to climate in distantly related plants.

Closely related species often use the same genes to adapt to similar environments. However, we know little about why such genes possess increased adaptive potential and whether this is conserved across deeper evolutionary lineages. Adaptation to climate presents a natural laboratory to test these ideas, as even distantly related species must contend with similar stresses. Here, we re-analyse genomic data from thousands of individuals from 25 plant species as diverged as lodgepole pine and Arabidopsis (~300 Myr). We test for genetic repeatability based on within-species associations between allele frequencies in genes and variation in 21 climate variables. Our results demonstrate significant statistical evidence for genetic repeatability across deep time that is not expected under randomness, identifying a suite of 108 gene families (orthogroups) and gene functions that repeatedly drive local adaptation to climate. This set includes many orthogroups with well-known functions in abiotic stress response. Using gene co-expression networks to quantify pleiotropy, we find that orthogroups with stronger evidence for repeatability exhibit greater network centrality and broader expression across tissues (higher pleiotropy), contrary to the 'cost of complexity' theory. These gene families may be important in helping wild and crop species cope with future climate change, representing important candidates for future study.

The evolution of genetic covariance and modularity as a result of multigenerational environmental fluctuation.

The genetic covariance between traits can affect the evolution of a population through selection, drift, and migration. Conversely, research has demonstrated the reciprocal effect of evolutionary processes on changing genetic covariances, in part through mutational covariance, correlational selection, and plasticity. In this article, we propose that correlated changes in selective optima over generations can cause the evolution of genetic covariance and the G-matrix in such a way that the population can, in the future, evolve faster. We use individual-based simulations of populations exposed to three types of changing environments that differ in the correlation of the change between selective pressures. Our simulation experiments demonstrate that selection pressures for different traits changing in a correlated pattern over generations can lead to stronger trait correlations compared to the case with independently changing selective optima. Our findings show that correlated selective pressures result in significantly higher genetic trait covariance and that pleiotropy accounts for the majority of the difference in covariance between treatments. We also observe that the mutational variance evolves according to the environment that the populations were exposed to. Moreover, we show that clustered patterns of changes in selection can allow the evolution of genetic modularity. We show that the pattern of change in the selective environment affects the pace at which fitness evolves, with populations experiencing correlated change in optima having on average higher mean fitness than those experiencing uncorrelated environment change.

Using genome scans to identify genes used repeatedly for adaptation.

Adaptation occurring in similar genes or genomic regions in distinct lineages provides evolutionary biologists with a glimpse at the fundamental opportunities for and constraints to diversification. With the widespread availability of high-throughput sequencing technologies and the development of population genetic methods to identify the genetic basis of adaptation, studies have begun to compare the evidence for adaptation at the molecular level among distinct lineages. However, methods to study repeated adaptation are often oriented toward genome-wide testing to identify a set of genes with signatures of repeated use, rather than evaluating the significance at the level of an individual gene. In this study, we propose PicMin, a novel statistical method derived from the theory of order statistics that can test for repeated molecular evolution to estimate significance at the level of an individual gene, using the results of genome scans. This method is generalizable to any number of lineages and, indeed, statistical power to detect repeated adaptation increases with the number of lineages that have signals of repeated adaptation of a given gene in multiple lineages. An implementation of the method written for R can be downloaded from https://github.com/TBooker/PicMin.