Evolution

Article summaryGene interactions are common, yet additive genetic models often predict short-term evolution and breeding response. This study argues that additivity can arise because populations sample only a small neighbourhood of a curved fitness landscape. In additive channels, genetic variation is small enough that local curvature contributes little to heritable fitness differences. The study defines an additivity index ([A]g) that compares variance from the local slope of log-fitness with variance from curvature, and links this ratio to expected prediction accuracy under Gaussian assumptions. A selection-inheritance framework shows when additive channels persist and when populations leave them. It yields testable predictions.

In many systems, mutations can have background-dependent fitness effects due to genetic interactions between loci within a genome (intragenomic epistasis). In some cases, such as when species are coevolving, genetic interactions between loci can span across species; this is described as intergenomic epistasis. It is known that intragenomic epistasis can make adaptation more repeatable by constraining accessible mutational paths. Here, we investigate whether intergenomic epistasis leads to the same pattern of increased repeatability and how repeatability is influenced by the interplay of intra- and intergenomic epistasis. For this, we model a two-species system in which the fitness of a species depends on the combination of genotypes that are present in both species. We implement this system using an NKC model, which allows us to construct coevolutionary fitness landscapes on which we simulate adaptation by means of mutations in both species. To quantify the repeatability of adaptation, we track the realised endpoints of adaptive walks and record the distribution of fitnesses of the focal and partner species at these evolutionary endpoints. We find that intergenomic epistasis creates highly repeatable patterns of adaptation that depend on the underlying shape of the coevolutionary landscapes. The patterns of repeatability deviate from expectations based on intragenomic epistasis due to fitness trade-offs between species, which can lead to cycling and large co-evolutionary fitness loads.

How evolutionary and developmental processes interact to determine axes of neural variation that produce behavioural diversity has been debated for many decades, with alternative hypotheses giving differential emphasis to functional coupling, which favours co-evolution, and developmental constraint, which enforces it. A critical omission is data on the genetic architecture of brain size and structure, which more closely illuminates the shared developmental dependencies between components of an integrated system. Here, we exploit ecological divergence between Astatotilapia calliptera and Aulonocara stuartgranti, two closely related cichlid species from Lake Malawi, to explore the genetic architecture of brain evolution. Using computer vision and machine learning techniques to extract volumetric data from micro-tomographic images, we first demonstrate significant divergence in brain composition between these species. Genomic and micro-tomographic imaging data from a population of hybrids generated between the two species were used to investigate genetic factors shaping this differentiation. We show that the majority of brain components are integrated phenotypically in hybrids, but genetic correlations between them are generally weaker. We further show that variation in multiple brain components is associated with variation in largely structure-specific quantitative trait loci, rather than determined by genetic factors with broad effects across the entire brain. These results suggest a genetic architecture that can facilitate modular changes in brain structure, and imply that individual components are independently evolvable.

Population genetic models excel at identifying the conditions for polymorphisms based on balancing selection but typically disregard the ecological processes that yield particular values of selection coefficients. We model a system that combines antagonistic pleiotropy, dominance reversal and heterozygote advantage: the wood tiger moth Arctia plantaginis, where alternative haplotypes at a major-effect locus determine male hindwing coloration. Yellow offers better protection against predators, while white is often associated with better mating success. The effects of mortality and reproductive success overlap in time because protandrous males can mate as long as they are alive, but they need to avoid predation for several days before the bulk of females emerge. We show that protandry aids polymorphism maintenance whenever the second-fittest genotype (after the heterozygote) is the poorly surviving but mating advantaged homozygote, while increased protandry harms polymorphism when the second-best fitness is that of the survival advantaged morph. Ecologically plausible protandry times predict that dominance reversal does not have to be strong for polymorphism to be maintained. Our study highlights the importance of timing traits in maintaining polymorphisms in Lepidoptera and showcases the benefits of deriving fitness explicitly in place of abstract selection coefficients that lack temporal components within the life cycle.

Phenotypic plasticity allows plants to rapidly respond to changing environments without the need for evolutionary change or migration. While selection can create variation in plasticity across natural populations, these responses are not adaptive in all environments. To predict whether plasticity will be adaptive requires evaluation of its fitness effects across a range of environments, including novel ones. Here, we test how traits and their plasticity vary for genotypes collected across a natural hybrid zone between two tree species with contrasting climatic niches. Fast-growing Populus trichocarpa inhabits maritime environments with relatively warm and stable temperatures, while P. balsamifera inhabits continental environments with cold winters and large temperature variance throughout the year. We planted 44 clonally replicated genotypes into thirteen common gardens and measured vegetative phenology, leaf morphology, stomata morphology and conductance, and photochemistry. Overall, genotypes from colder, more continental environments exhibited higher plasticity. P. balsamifera ancestry was associated with increased plasticity in timing of fall phenology, stomatal conductance, and leaf mass per unit area. We assessed the effects of trait plasticity on fitness estimated as yearly growth across common gardens and found that the plasticity-fitness relationship was often garden-specific, indicating that the planting environment did not consistently mediate plasticity-fitness relationships. When the effects of trait plasticity on growth varied by garden temperature, higher plasticity generally had neutral or negative associations with growth in warmer environments. These results suggest that elevated plasticity evolved in a P. balsamifera genomic background as part of a climate generalist strategy to seasonal temperature variability, but that there is a trade-off between plasticity and growth in warmer environments. Consequently, less-plastic but warm-adapted P. trichocarpa genotypes are likely to have a fitness advantage under warming climates. These results demonstrate that plasticity may sometimes be maladaptive and will not be universally beneficial in a warming world.

Sarita Lake, British Columbia houses a distinctive population of threespine stickleback (Gastrosteus aculeatus L.) with a phenotype characterized by unusually large individuals relative to nearby conspecifics. We tested the hypothesis that members of this population are not isometrically larger but rather exhibit variation in allometric trajectories that reflect changes in developmental timing impacting the developmental-genetic architecture of the phenotype. We used 3D geometric morphometrics to characterize the size and shape of skulls, pectoral girdles and pelvic girdles from a sample of individuals from nearby freshwater and marine populations and compare them to a sample from Sarita Lake. We showed that individuals from the Sarita Lake population are larger in each body region compared to most other populations examined. Further, these individuals have dorsally expanded skulls and relatively robust pelvic armour. We also showed that the relationship between size and shape is differently structured among body regions and is heavily influenced by non-uniform sexually-mediated variation across populations sampled. Our results reflect complex underlying developmental trajectories, and we suggest that the large phenotype observed may be driven by fecundity selection on female size in combination with a limnetic trophic niche and relatively increased predation pressure in Sarita Lake.

Additive fitness landscapes--also called Mount Fuji landscapes--are the simplest and most widely used models of sequence-function relationships. As such, they play essential roles across multiple areas of biology, including evolutionary theory, quantitative genetics, gene regulation, and protein science. One of the most basic properties of any fitness landscape is its genotypic density--the number of sequences near a given fitness value. Understanding this density is especially important near fitness peaks, as it quantifies the supply of high-fitness genotypes. Here I study the genotypic density of additive landscapes near fitness peaks. Although this density is well known to be approximately Gaussian near the middle of the fitness range, its behavior near maximal fitness has not been reported. I begin by deriving a saddle-point approximation that accurately describes the genotypic density of additive landscapes over virtually the entire fitness range. I then show that the log density follows a power law near maximal fitness, with an exponent determined by how much the best allele at each position outperforms its nearest competitor. This power-law behavior holds over a substantial fraction of fitness values, besting the Gaussian approximation on both simulated and empirical landscapes across roughly a quarter to a third of the fitness range. Under certain conditions this behavior also extends to globally epistatic landscapes (defined as nonlinear functions over one or more additive traits), though with a reduced range of validity. These findings advance our understanding of one of the most fundamental models of sequence-function relationships. In particular, they reveal that the uppermost reaches of Mount Fuji landscapes, rather than being sharply peaked, are actually quite stubby.

The evolution of plant resistance naturally occurs in complex, multifaceted environments that consist of multiple simultaneous stressors. Understanding how shifting environmental contexts may shape resistance evolution requires empirical studies that consider the combined effects of interacting stressors on fitness and selection. Here, we examined how exposure to an insecticide impacts the evolution of resistance to the herbicide glyphosate in Ipomoea purpurea (common morning glory). Through a factorial field experiment, we manipulated glyphosate and an insecticide to estimate selection on glyphosate and herbivory resistance. We found that glyphosate acted as the primary agent of selection, favoring higher levels of glyphosate resistance. In the presence of glyphosate alone, positive correlational selection favored a combination of higher glyphosate and herbivory resistance, supporting prior work that suggested these traits may be linked. Importantly, insecticide exposure modified both glyphosate resistance and the strength of selection acting upon the trait by increasing resistance and weakening selection. Together, our results indicate that the evolution of herbicide resistance is context-dependent and that secondary stressors like insecticide can alter the evolutionary trajectories of plant defense.

Understanding how mutations evolve on Y chromosomes is central to explaining the origin, diversity and persistence of sex chromosomes. Mutations occurring on the Y chromosome in sexual populations experience selective dynamics that differ markedly from those on autosomes, due to a reduced effective population size and the presence of large non-recombining regions containing alleles maintained in a permanently heterozygous state. These specific features alter gene transmission in the Y chromosome population compared to autosomes, even within the same pedigree. Here, we provide a two-sex diploid Wright-Fisher model that explicitly incorporates both sex chromosomes and autosomes within a unified population framework, in order to capture the influence of these specificities on the fate of mutations, not only considering fixation probabilities but also segregation times. We use diffusion approximations and provide analytical and numerical tools to compute these quantities across a wide range of parameters and selection regimes. We recover classical results on fixation probabilities in various scenarios, including purely beneficial, deleterious or overdominant mutations, and extend them in the light of mean segregation time, a key but often overlooked determinant of evolutionary outcomes over finite timescales. In particular, our analyses show that overdominant mutations are overall more likely to fix in observable time windows on the Y chromosome than on autosomes. Individual-based simulations corroborate our approximations and highlight parameter regimes where the theoretical approach is particularly useful, especially for parameter values inducing long segregation times or small fixation probabilities, for which simulations are impractical. Our results provide a comprehensive and tractable framework for clarifying how chromosome-specific features shape evolutionary dynamics beyond fixation probabilities alone.

Alleles with opposing effects on fitness characters are said to exhibit selectional antagonistic pleiotropy (broadly construed so that effects are not necessarily confined to the same individual). A number of theoretical investigations considered the case where a pair of alleles at a locus influences two fitness components and derived the conditions giving rise to stable polymorphism under various assumptions about the mode of trait-interaction. Strikingly, many of these analyses concluded that the potential for maintaining polymorphism is strongly constrained by the joint influence of two factors: (1) the prevalence of weak selection coefficients over coefficients of large magnitude, and (2) the absence of beneficial dominance reversals (where the deleterious effects of each allele are partially or completely masked in the heterozygous genotype). Consequently, the conclusion that selective polymorphism is unlikely to be maintained by intralocus mechanisms of antagonistic pleiotropy has achieved widespread acceptance. Here we argue that such conclusions do not apply to any of the following models of antagonism: (i) additive trait-interaction, (ii) multiplicative trait-interaction, (iii) bivoltine selection, (iv) soft selection, (v) hard selection, and (vi) sexual antagonism. We demonstrate that the parameter space giving rise to stable allelic variation is quite large throughout, and moreover, the plenitude of suitable parameters neither depends on the strength of selection nor requires dominance reversal. Dominance coefficients associated with stringent conditions for stable polymorphism are shown to be atypical as compared to all feasible parameters, and best regarded as an outcome of adherence to a special relation: dominance with a constant magnitude and direction, which includes the case of additive allelic effects at a locus. Properties of single-locus equilibria (heterozygosity, allele frequency differentiation) are investigated, as well as the contribution of dominance schemes to the genetic variance in fitness characters in populations at multilocus linkage equilibrium. Author summaryAllelic variants at a locus with opposing effects on multiple fitness components (antagonistic fitness pleiotropy) have long been appreciated as a possible source of balancing selection. The prevalence of polymorphism owing to this form of natural selection, however, has been doubted on theoretical grounds due to the fact that standard assumptions of genetic models (namely, constant magnitudes for the dominance coefficients) are hardly conducive to the maintenance of polymorphism. The major exception to this conclusion lies with schemes that exhibit dominance reversal (where the direction of dominance for antagonistic alleles flips across fitness components). Here we conduct a geometric analysis of the space of polymorphism-promoting dominance parameters and conclude that the conditions for maintaining balanced alleles is unrestrictive, with non-reversals playing an underappreciated role.

Parthenogenesis is relatively rare and often regarded as an evolutionary dead end. Despite this, certain parthenogenetic animal species have endured for millions of years, but it is unclear what enables the persistence of some parthenogenetic lineages. Transitions from sexual to parthenogenetic reproduction can occur through different evolutionary processes that give rise to diverse cytological reproductive mechanisms. These mechanisms are likely to influence genetic diversity, especially in the early stages after the transition to parthenogenesis and may thus affect lineage persistence. To understand such evolutionary transitions, we used experimental crosses to investigate the mechanism of parthenogenesis and the immediate genetic consequences of switching from sexual to parthenogenetic reproduction in the facultatively parthenogenetic phasmid Megacrania batesii. We obtained DNA sequence data from multiple lineages propagated over three generations via sex, parthenogenesis, or transitions between reproductive modes. We quantified heterozygosity and within-family genetic variation and compared the genetic patterns with predictions for known mechanisms of parthenogenesis. We found that a single generation of parthenogenesis typically resulted in (near-)complete loss of heterozygosity and an absence of within-family genetic variation, consistent with automixis with gamete duplication or terminal fusion and little/no recombination. However, we also found evidence of variation in the mechanism of parthenogenesis among lineages and even within the same individual, associated with drastic differences in the amount of heterozygosity and within-family genetic variation maintained across generations. Our findings show that considerable variation in parthenogenetic mechanisms can exist within populations and suggest that such variation could influence the persistence and evolution of parthenogenetic lineages.

Genetic evidence for fluctuating selection has begun to accumulate for different species over the past few decades, especially for the Drosophila genus where studies have reported hundreds of loci undergoing putatively adaptive oscillations across successive seasons. However, most theoretical and simulation studies of fluctuating selection have relied on abstract or weakly parameterized models, making it difficult to assess their relevance for natural populations. In this study, we simulate multilocus seasonally fluctuating selection under a recently developed model and examine its effect on the variance effective population size (Ne) at a genome-wide scale. By recapitulating genomic, demographic, and evolutionary parameters from natural Drosophila populations in our simulations, we were able to reproduce allele frequency oscillations reported in recent studies and show that these lead to [~]50% genome-wide reductions in Ne. We also demonstrate that Ne reductions are well predicted by the maximum frequency amplitude among all adaptively fluctuating loci, and that the frequency amplitudes are largely determined by the number of adaptively fluctuating loci and the strength of their epistatic interactions. Our results demonstrate that fluctuating selection can substantially reduce effective population size and underscore the importance of temporally variable selection in shaping genome-wide patterns of variation beyond classical models. Article SummaryGenetic studies of fluctuating selection in natural populations have grown steadily over the past decade, with reports suggesting that hundreds of loci undergo adaptive oscillations over seasonal timescales in cosmopolitan Drosophila populations. By simulating seasonally fluctuating selection under a recently developed model and ecological scenarios informed by published studies, the authors show that this mode of selection can reduce effective population size by [~]50%, with the magnitude of the reduction correlated with the locus exhibiting the largest allele frequency fluctuations. These findings highlight fluctuating selection as an important factor shaping genome-wide patterns of genetic variation and effective population size.

Life history traits are often correlated, creating trade-offs that may impede the response to natural selection and be responsible for the evolution of senescence. These trade-offs may arise through pleiotropic effects, which can affect the response to selection in ways that resemble intra-locus sexual antagonism. Despite these hypothesized relationships, we lack clear connections between pleiotropy, sexual antagonism, and the evolution of life histories. Empirical tests for inter-sexual differences in life-history traits, including sex-specific aging, can be used to evaluate hypotheses about how pleiotropy and sexual conflict affect evolutionary trade-offs. To those ends, we measured lifespan, development time, and body size in Drosophila pseudoobscura males and females, each of which carried one of six third chromosome inversion genotypes. Temperature affected lifespan and development more than any other factor; higher temperatures increased mortality rate, decreased lifespan, and accelerated development. However, we also observed sex differences in mortality rates and development times that depended on genotype and temperature. Notably, temperature elevated the initial mortality rate across all flies, yet increasing temperatures reduced the rate of aging in some genotype-sex combinations. Similarly, direct effects of genotype on mortality rate and development time depended greatly on sex and temperature, but there was no genotype effect on body size. Despite these context-dependent genotype effects on life history traits, we failed to identify any correlations that would serve as clear evidence for sexual conflict or trade-offs. Our results therefore suggest that either historical conflicts have been resolved or any conflicts that may exist do not result in the correlations predicted by existing models.

Evolving populations, especially in the strong-selection-weak-mutation limit, can be modelled as adaptive walks on fitness landscapes, moving in fitness-increasing mutational steps until reaching a fitness peak--a local optimum. Simulations of such adaptive walks--on a multi-peaked empirical landscape of the folA gene and on landscapes generated by the Rough Mount Fuji (RMF) model-- have shown that some landscapes are highly navigable, meaning that the highest x% of peaks are reached by >> x% of adaptive walks. This prompts the question of how adaptive walks can be so successful despite the local, myopic rules behind each adaptive step. Here, we investigate this question using simulations and mathematical approximations of random adaptive walks on a simplified RMF landscape. The landscape has a low-to-intermediate fitness region, whose size reconciles a low peak density with a high peak number. Despite the high number of peaks, walkers are likely to exit this region without terminating at a peak because the probability of a peak transition at each step is low and a fitness gradient guides walkers to the high-fitness region in few steps. Thus, three features are sufficient to explain why adaptive walks in the simplified RMF landscape are likely to reach a small fraction of top-ranking peaks: a low-to-intermediate fitness region with a high number of peaks, a low peak-transition probability, and which is crossed in few steps. We find that these three features are also present in the empirical folA landscape, suggesting that similar principles may apply.

Predator-prey interactions are key drivers of behavioural and life-history evolution, yet their mechanisms remain difficult to study in natural contexts. The nematode Pristionchus pacificus is a model predator, but most studies exploring its behaviours use Caenorhabditis elegans as prey, a species that it likely only rarely encountered in nature. Here, we examine predation within nematode communities associated with beetle carcasses, the native necromenic habitat of P. pacificus. We identify Oscheius myriophilus as a cohabiting species, likely representing natural prey. Using predatory assays, automated tracking, and machine-learning-based behavioural analysis, we show that P. pacificus actively kills and consumes O. myriophilus. Strikingly, predation rates are lower than those observed for C. elegans, suggesting partial resistance or reciprocal adaptation in O. myriophilus. Consistent with this, O. myriophilus exhibits a mixed reproductive strategy, with early oviposition followed by ovoviviparity and matricide. As later developmental stages are more resistant to predation, internal hatching may protect offspring while providing maternal resources for development. These findings establish these nematodes as a tractable model for investigating predator-prey interactions and their evolutionary consequences, highlighting how behavioural strategies and life-history traits can co-evolve in natural communities.

The phenotypic effects of mutations often depend on the genetic background, yet general patterns remain poorly resolved. Here, we tested whether genotypes drawn from the same natural population, but differing in their breeding values for a polygenic trait, differed in their contribution of new mutational variation to that trait. We established >200 mutation-accumulation (MA) lines from four Drosophila serrata genotypes. Analysing >44,000 wing-size measurements, collected over 30 generations, we quantified mutational variance and mutational bias for size. Genotypes with the smallest and largest breeding values for size contributed similar (statistically indistinguishable) amounts of mutational variance. In contrast, the genotype with an intermediate breeding value exhibited remarkably low (statistically undetectable) mutational variance, low micro-environmental variance, and high line survival over time, consistent with limited mutational decay in fitness. The three genotypes with detectable mutational input showed declines in mean size over time, indicating a consistent mutational bias toward smaller size, as reported in other taxa. The magnitude of this bias appeared genotype dependent, with the MA populations founded from the larger ancestors declining nearly twice as fast as that founded from the smallest ancestor. Together, these results demonstrate substantial heterogeneity in mutational properties among genotypes within a single natural population where the trait value spans a relatively narrow range. Such genotype-specific mutational input is expected to shape both the standing genetic variance and the evolutionary trajectory of polygenic traits.

Hybridisation outcomes often vary across space and time, yet the relative roles of demographic history and genomic architecture in shaping introgression remain unclear. Here, we investigate three replicated hybrid zones between the damselflies Ischnura elegans and Ischnura graellsii across Spain to test whether genomic introgression patterns are repeatable across independently formed zones. Using genome-wide data, we combined demographic modelling, genomic cline, and functional annotation of introgressed loci. Demographic inference supported three independent secondary contact events of different ages: The South-east zone forming first ([~]207 years ago), followed by the North-west ([~]73.5 years) and North-central ([~]33 years) hybrid zones. Despite similar cline steepness across autosomes, asymmetric gene flow from I. graellsii into I. elegans was observed, with a low overlap of introgressed loci between zones. These loci were mainly associated with broad regulatory and transport-related functions in both hybrid zones, indicating repeatability at the level of gene function rather than gene identity. In contrast, the X chromosome showed steeper clines, suggesting strong intrinsic genomic constraints. Together, demographic history explains geographic heterogeneity in introgression, whereas chromosome architecture imposes consistent constraints. These findings highlight how replicated hybrid zones can disentangle contingent versus repeatable genomic responses during early stages of speciation with gene flow.

Evolution of drug resistance to one drug can alter the minimum inhibitory concentration to another drug. This phenomenon, known as a collateral effect, can manifest as either cross-resistance or collateral sensitivity. Various patterns of collateral effects have been observed experimentally. Repeated adaptation from the same parental strain may result in variable collateral effects; this is non-repeatability. Additionally, adaptation of a pathogen to one drug may produce a specific collateral effect to a second drug, while altering the order of drug exposure can result in a different, or even absent, collateral effect. This phenomenon is termed unidirectionality. The genetic and evolutionary mechanisms underlying these patterns remain incompletely characterised. Here, we propose a frame-work that integrates pharmacodynamics and population genetics and provide minimal examples to explain these patterns and their combinations. Furthermore, we demonstrate that drug concentration and selection regime strongly influence patterns of collateral effects, including repeatability, directionality, and their temporal dynamics.

The distribution of fitness effects (DFE) of mutations is a key determinant of both the efficacy of natural selection and the genetic load of populations. It also provides an indirect summary of the underlying fitness landscape, so the extent to which the DFE is conserved across species can offer insights into the invariance of these landscapes. Here, we infer the DFE of amino acid-changing mutations in 38 catarrhine primates using site frequency spectrum (SFS)-based methods. Our results are consistent with the nearly neutral theory, with interspecific differences primarily driven by variation in effective population size (Ne) and thus the efficacy of selection. Specifically, a one-order-of-magnitude increase in Ne is associated with an approximately 10% increase in the fraction of strongly deleterious mutations. Although DFE estimates exhibit clade-level clustering, this pattern largely reflects shared Ne rather than clade-specific effects, and in unscaled units, deleterious DFEs are broadly similar across species. These conclusions are robust to the choice of DFE parameterization, phylogenetic regression framework, and ancestral misidentification. Extending the analysis to non-additive dominance effects further shows that dominance is only weakly identifiable from the SFS and has minimal impact on comparative DFE inference. Finally, simulations demonstrate that demographic correction via nuisance parameters enables robust inference across a range of demographic scenarios.

We study one of the simplest scenarios of polygenic selection that can be imagined: a subdivided population of diploid individuals expressing an additive trait under spatially homogeneous stabilizing selection. We are interested in the amounts of variation that can be maintained at mutation-selection-migration-drift equilibrium, at individual loci and at the level of the trait, within and among subpopulations. We derive analytical approximations for variance components and summary statistics such as FST and QST under the assumptions of the infinite-island model and compare these with individual-based simulations. We find that: (i) There is a critical migration threshold (which depends on effect sizes of trait loci) below which population structure strongly inflates genic variance in the subdivided population to levels well above those in a panmictic population. Variation within each subpopulation is maximized close to the critical migration rate. (ii) The genetic basis of trait variation across subpopulations is most similar close to this migration threshold and (counter-intuitively) decreases for higher migration rates. This has consequences for the portability of Genome-Wide Association Studies (GWAS) between subpopulations, i.e, the extent to which loci with large contributions to variance in one subpopulation explain variance in other subpopulations. (iii) An analytical mean-field approach based on the single-locus diffusion approximation, together with effective migration and selection parameters (to account for associations between loci), very accurately predicts various quantities.