Evolution

When populations experience different sensory conditions, natural selection may favor whole sensory system divergence, from the peripheral structures to the brain. We characterized the outer eye morphology of sympatric Heliconius species from different forest types, and their first-generation reciprocal hybrids to test for adaptive visual system divergence and hybrid disruption. In Panama, Heliconius cydno occurs in closed forests, whereas Heliconius melpomene resides in more open areas. Previous work has shown that, among wild individuals, H. cydno has larger eyes than H. melpomene, and there are heritable, habitat-associated differences in the visual brain structures that exceed neutral divergence expectations. Notably, hybrids have intermediate neural phenotypes, suggesting disruption. To test for similar effects in the visual periphery, we reared both species and their hybrids in common garden conditions. We confirm that H. cydno has larger eyes and provide new evidence that this is driven by selection. Hybrid eye morphology is more H. melpomene-like despite body size being intermediate, contrasting with neural trait intermediacy. Thus, eye morphology differences between H. cydno and H. melpomene are consistent with adaptive divergence, and when combined with previous neuroanatomy data, suggest hybrid visual system disruption due to mismatched patterns of intermediacy and dominance in the visual pathway.

While traits that contribute to premating sexual interactions are known to be wildly diverse, much less is known about the diversity of postmating (especially female) reproductive traits and the mechanisms shaping this diversity. To assess the rate, pattern, and potential drivers of postmating reproductive trait evolution, we analyzed male and female traits across up to 30 Drosophila species within a phylogenetic comparative framework. In addition to postmating reproductive morphology (e.g., sperm length, reproductive tract length and mass), we also quantified mating behaviors including female remating rate--a common proxy for the strength of postmating sexual selection. We found evidence for strong coevolution between male and female postmating traits (specifically sperm length and sperm storage organ size). However, remating rate was not associated with the rate of evolution or exaggeration of either male or female postmating reproductive morphology, once phylogenetic relatedness was accounted for. We infer that female-mediated and intersexual selection predominantly drive the evolution of our postmating morphological traits, including via divergent male and female interests in controlling paternity. In comparison, remating rate has a complex and likely secondary role in shaping this evolution, in part because this trait can be both a driver and a product of postmating selection.

The between sex genetic correlation for traits has long been hypothesized as a constraint to the evolution of sexual dimorphism. Both empirical and theoretical work has suggested that this constraint is influenced by genotype-sex-environment interactions. We examine genotype-sex-environment interactions in both sexually exaggerated and non-exaggerated legs of Drosophila prolongata, to examine the role of organismal condition on evolvability of an extreme trait. We employed a nested full-sib half-sib crossing design, providing food either ad libitum, or restricting food during larval growth, to each brood. When provided food ad libitum, inter-sex genetic correlations between traits is high and positive, whereas under food restriction this correlation substantially weakens, with a modest negative sign. Similarly, comparisons of the G matrix across sexes becomes less associated under food restriction. We discuss these results in the context of the growing appreciation of the factors that facilitate sex-specific evolutionary change.

High investment in current reproduction can limit future reproductive opportunities, but how selection shapes these hormone-mediated traits remains poorly understood. Androgens can mediate male reproductive investment, and in three-spined stickleback (Gasterosteus aculeatus), exert antagonistic effects on breeding effort versus spermatogenesis. To understand how shifts in reproductive strategy shape this tradeoff, we compared testes transcriptomes and androgen production between two recently diverged stickleback ecotypes that differ in reproductive strategy: the ancestral "common" ecotype, which provides paternal care, and the non-parental "white" ecotype, which has lost paternal care and prioritizes mating effort. During typical breeding, testes gene expression differed little between ecotypes. However, under prolonged summer-like conditions, testes gene expression diverged substantially. Common-biased genes were enriched for meiotic functions and spermatogenic cell type markers, suggesting commons had initiated spermatogenesis while whites had not. Instead, whites expressed higher levels of steroidogenic candidate genes and released significantly more 11-ketotestosterone than commons, indicating sustained investment in current reproduction. F1 hybrids released 11-ketosterone at intermediate rates, suggesting a genetic basis for this divergence. Sustained androgen production in whites may possibly delay the transition into spermatogenesis, limiting investment in future reproduction. These results illustrate how selection on hormonally-integrated traits can drive rapid divergence in life history strategy.

Phenotypic traits are often constrained in their evolution by the genetic and developmental map underlying them. However, the extent to which these constraints themselves evolve has not been well-characterized. Traits that are integrated are generally thought to be constrained in their ability to evolve because their shared developmental and genetic architecture causes any change in one to be mirrored in the other. Nevertheless, it is not yet clear whether correlations can constrain major episodes of phenotypic diversification over longer timescales. We reconstruct patterns in integration within nine primate lineages and model the evolution of the resulting modules in a phylogenetic context. We find that patterns in integration are generally evolvable across lineages, with apes displaying particular lability in the composition of morphological modules. This increased turnover in module composition corresponds to both divergent evolution of skeletal shape and the formation of novel complexes of locomotor behaviors and postures. While integration may play a role in constraining evolutionary innovation, its effects are dynamic, shifting as the structure of integration itself evolves during episodes of exceptional ecomorphological diversification.

Colonisation of new habitats is typically followed by divergent selection acting on traits that are immediately important for fitness in the new habitat. For example, shifting sensory environments are often associated with variation in sensory traits critical for navigation and foraging. However, the extent to which the initial response to novel sensory conditions is mediated by phenotypic plasticity, and its contribution to early species divergence remains unclear. We took advantage of repeated cases of speciation in Heliconius butterflies with independent allopatric distributions in the west of the Colombian and Ecuadorian Andes. Using volumetric brain measurements, we analysed patterns of investment in sensory processing in brain components across different localities and habitats. We find that a higher-altitude species, H. chestertonii, differs in levels of investment in visual and olfactory brain centres compared to its lower altitude relative H. erato venus, mainly attributable to heritable variation as inferred from comparisons between wild and common-garden reared individuals. We compared these shifts with those reported for another high-altitude species, H. himera, and its parapatric lowland counterpart, H. erato cyrbia, and demonstrate parallel reductions in the size of specific optic lobe neuropils. Conversely, for the antennal lobe, we detected disparate trait shifts in H. himera and H. chestertonii in respect to their lowland erato neighbours. Overall, our findings add weight to the adaptive potential for neuroanatomical divergence related to sensory processing during early species formation. Lay summaryRepeated associations between trait variation and environmental shifts may indicate adaptation to local sources of natural selection. For instance, in fish, the presence of certain morphological traits in specific ecological conditions across independent populations is well documented, suggesting equivalent phenotypic responses to shared sources of natural selection. We compared independent cases of ecological divergence in Heliconius butterflies distributed along altitude gradients from sea level to mid mountain in the west of the Colombian and Ecuadorian Andes. Shifts in altitude involve repeated, abrupt transitions from wet, large-leaved, warm forests to higher dry, open, cold scrubs. We tested hypotheses about the role of these ecological shifts in driving adaptive evolution in neuroanatomical traits during early speciation. We showed that in Heliconius, independent changes in forest-type have been accompanied by heritable parallel patterns of divergence in sensory investment in visual processing in the brain. We propose these differences likely facilitate species divergence in the face of ongoing geneflow.

Hybridization between distinct populations injects genetic variation, which can bring fitness benefits. However, these benefits often appear as F1 heterosis, and might not persist into later generations; especially since, as emphasized by classical theories, heterosis can be caused in several different ways. Here, we study the long-term outcomes of hybridization, using a model that allows us to tune several properties of the genetic variation, including the strength and architecture of heterosis, thereby unifying the classical theories. Results suggest that long-term outcomes depend mainly on the variance in epistasis, which determines the ruggedness of the fitness landscape, but without affecting the heterosis. Together, results suggest that the study of heterosis may tell us relatively little about the long-term outcomes of hybridization, and that hybridization might bring benefits more often than has been assumed.

Populations no longer experiencing a sufficient rate of gene flow will accumulate genetic differences over time. One potential consequence of divergence between natural populations is hybrid breakdown, which can occur during secondary contact when untested allelic combinations in hybrids beyond the F1 generation are maladaptive and restrict gene flow. Hybrid breakdown is an important process in the development and maintenance of species boundaries, and has largely been studied between populations that are completely or nearly completely isolated. Here, we leverage the recent worldwide expansion of Drosophila melanogaster to investigate signatures of hybrid breakdown between populations that diverged within approximately the last 13,000 years. We did not find clear evidence for hybrid breakdown in viability or female reproductive performance. In contrast, we found that many but not all between-population crosses yielded an elevated fraction of second generation male offspring that were unable to reproduce. The frequency of non-reproducing F2 males varied among different crosses involving the same southern African and European populations, as did the qualitative effect of cross direction, implying a genetically variable basis of hybrid breakdown and a role for uniparentally inherited factors. The levels of male reproductive failure observed in F2 hybrids were not recapitulated in backcrossed individuals, suggesting the existence of incompatibilities with at least three partners. These results suggest that some of the very first steps toward reproductive isolation may involve incompatibilities with complex and variable genetic architectures, and they support the prediction that hybrid breakdown affects the heterogametic sex first. Collectively, our findings on polymorphic incompatibilities within D. melanogaster emphasize this systems potential for future studies on the genetic and organismal basis of early-stage reproductive isolation. IMPACT SUMMARYThe biological diversity that exists around the world is an emergent property of the generation of forms, which are commonly grouped into units we call species. The rate at which new species form can be influenced by the evolution of reproductive isolation, the inability of groups to interbreed. When reproductive isolation is studied in its nascent stages, researchers can gain critical insights into the genetic architectures and evolutionary forces underlying the earliest steps toward speciation. One process that may contribute to early-stage reproductive isolation is hybrid breakdown, when genetic incompatibilities in the offspring of hybrid individuals reduce their fitness. Here, we illuminate a complex pattern of hybrid breakdown among natural populations of Drosophila flies that diverged within the past 13,000 years. We find signals of hybrid breakdown involving male reproduction, between some but not all population pairs, whereas we find no clear evidence for hybrid breakdown impacting female reproduction or developmental survival. These findings are in agreement with Haldanes Rule, which posits that hybrid incompatibilities are more likely to affect the sex that carries distinct sex chromosomes (here, XY males). From certain crosses between African and European fly strains, we find strongly elevated rates of reproductive failure in second generation hybrid males, but outcomes vary dramatically depending on the individual strains crossed. We also provide evidence of incompatibilities underlying male reproductive failure that involve three or more genes, including uniparental factors such as the Y chromosome or mitochondrial genome. Our results highlight a complex and variable basis of hybrid breakdown during the earliest stages of reproductive isolation, in contrast to commonly envisioned scenarios that focus on two-locus incompatibilities caused by fixed genetic differences between groups. These findings also suggest that recently diverged populations of D. melanogaster provide notable opportunities for future studies of the genetic basis of early-stage reproductive isolation.

Evolutionary biologists have long been interested in understanding the mechanisms underlying Haldanes rule. The explanatory theories of dominance and faster-X, which are based on recessive alleles being expressed in the heterogametic sex, have been proposed as common mechanisms. These mechanisms predict that greater hemizygosity leads to both faster evolution and greater expression of intrinsic postzygotic isolation. Under these mechanisms, haplodiploids should evolve and express intrinsic postzygotic isolation faster than diploids because the entire genome is analogous to a sex chromosome. Here, we measure sterility and inviability in hybrids between Neodiprion pinetum and N. lecontei, a pair of haplodiplopids that differ morphologically, behaviorally, and genetically. We compare the observed isolation to that expected from published estimates of isolation in diploids at comparable levels of genetic divergence. We find that both male and female hybrids are viable and fertile, which is less isolation than expected. We then discuss several potential explanations for this surprising lack of isolation, including alternative mechanisms for Haldanes rule and a frequently overlooked quirk of haplodiploid genetics that may slow the emergence of complete intrinsic postzygotic isolation in hybrid males. Finally, we describe how haplodiploids, an underutilized resource, can be used to differentiate between mechanisms of Haldanes rule.

To predict adaptive evolution, we need to understand the degree to which selection on one trait can constrain or redirect evolutionary responses in other traits. We used artificial selection on cuticle color in Drosophila melanogaster and D. simulans to investigate whether color and behavior evolve in tandem. We selected for light and dark thoracic colors for 16 generations in two populations per species and measured correlated responses in a suite of behavioral traits, including aggression, basal activity, total activity, sleep, and geotaxis. Dark selected individuals consistently showed higher aggression and basal activity than light selected or control flies across both species and sexes, pointing to a modest but repeatable correlated response that must be interpreted in light of low and variable aggression in our assay. In contrast, patterns for geotaxis, sleep, and total activity evolved unpredictably, often varying across species and populations, and showed no clear or uniform association with color. Taken together, our results suggest that some behaviors may share a predictable relationship with color, whereas others behave largely independently under the conditions we examined. The persistence of correlated responses in aggression and basal activity in response to selection on color is consistent with a conserved genetic basis, such as pleiotropy or tight linkage, although our data do not, by themselves, distinguish among these mechanisms. Future work that identifies the loci underlying variation in both color and behavior and tests their effects experimentally will be important for clarifying how multivariate genetic covariance shapes the direction and pace of adaptive evolution.

The study of trait evolution in modular animals is more complicated than that in solitary animals, because a single genotype of a modular colony can express an enormous range of phenotypic variation. Furthermore, traits can occur either at the module level or at the colony level. However, it is unclear how the traits at the colony level evolve. We test whether colony-level aggregate traits, defined as the summary statistics of a phenotypic distribution, can evolve. To quantify this evolutionary potential, we use parent-offspring pairs in two sister species of the bryozoan Stylopoma, grown and bred in a common garden breeding experiment. We find that the medians of phenotypic distributions are evolvable between generations of colonies. We also find that the structure of this evolutionary potential differs between these two species. Ancestral species align more closely with the direction of species divergence than the descendent species. This result indicates that aggregate trait evolvability can itself evolve.

Hybrid zones provide a window into the evolutionary processes governing species divergence. While the role of postzygotic isolation has been extensively characterized in the context of hybrid zones, the contribution of prezygotic isolation is less well explored. Here, we investigate the effects of assortative mate choice, the underlying preference function and mating-trait architecture, and the strength of sexual selection on hybrid zone dynamics. We explore this question by means of a mathematical model parameterized with phenotype and genotype data from the hybrid zone between all-black carrion and grey-coated hooded crows. The best-fit model resulted in narrow clines for two mating-trait loci coding for colour phenotype maintained by a moderate degree of assortative mating. Epistasis between the two loci induced hybrid-zone movement in favor of alleles conveying dark plumage followed by a shift in the opposite direction favouring grey-coated phenotypes [~]1,200 generations after secondary contact. Unlinked neutral loci diffused near-unimpeded across the zone. These results were generally robust to the choice of matching rule (self-referencing or parental imprinting) and effects of genetic drift. Overall, this study illustrates under which conditions assortative mating can maintain steep clines in mating-trait loci without generalizing to genome-wide reproductive isolation. It further emphasizes the importance of mating-trait architecture for spatio-temporal hybrid-zone dynamics.

Sexual selection can promote speciation in theory, but macroevolutionary studies have reported mixed results. A possible explanation for this inconsistency is the use of inappropriate proxies of sexual selection. Here, by focusing on swallows and allies (Aves: Hirundinidae), I examined whether or not a classic example of sexually selected trait, long outermost tail feathers, explains speciation rate in this clade. Long tails have been repeatedly shown to be intersexually selected in manipulative experiments, which is further corroborated by a series of macroevolutionary studies, thereby validating their use as a target of sexual selection. I found that hirundines with sexually dimorphic tail length have a significantly higher speciation rate than those with sexually monomorphic tail length. Furthermore, evolutionary changes in the extent of sexual tail dimorphism (and those in the extent of male outermost tail exaggeration) was significantly positively associated with speciation rate. Sexual plumage dichromatism and wing dimorphism are irrelevant to speciation rate. Together, the current study demonstrated the importance of using valid targets of sexual selection in studies on the macroevolutionary patterns of speciation. Using a classic, unidimensional sexual trait (i.e., long tail), I demonstrated a strong macroevolutionary support that divergence of intersexually selected traits promotes speciation.

A scaling model named the "geometric-similarity-first model" is developed to explain allometric trait divergence over recent evolutionary time. In the models first step, traits undergo geometric scaling with a populations change in body size. In step 2, directional natural selection re-optimizes trait shapes such that traits showing positive ontogenetic allometry undergo positive evolutionary allometric scaling while traits that exhibit negative ontogenetic allometry go through negative evolutionary allometric scaling. Five predictions of the model were tested using morphological data for three locomotor-relevant traits in pygopodid lizards. The dataset, which was based on 1,756 museum specimens representing 31 species, supported all of these predictions. An implication of these results is that geometric scaling, driven by natural or sexual selection, may be a mechanism for peak shifts on an adaptive landscape. Given the ubiquity of body size variation in nature, this hypothetical process, termed "niche scaling," may be important to ecological diversification. Applications of the model to some other well-studied species are discussed.

Sexual dimorphism is common despite factors such as inter-sex genetic correlations and sex-specific patterns of selection that might limit its evolution. Sexual dimorphism can be phenotypically plastic and condition dependent, that themselves may be targets of selection. However, it remains unclear how sexual dimorphism, its plasticity and condition dependence evolves, in particular during rapid adaptation to a new environment. Furthermore, the interplay between SSD and other forms of dimorphism, such as shape dimorphism co-evolves. Using Sub-Saharan populations of Drosophila melanogaster that vary for size and shape as a result of adaptation to high altitude environments, we examined sex specific patterns of developmental plasticity. We raised strains of Drosophila from low (Zambia) and high (Ethiopia) altitude populations varying for food quality or rearing temperature. We observed expected differences in wing size and shape due to population, sex and plasticity. While larval mass showed substantial evolved changes for sex specific condition dependence, effects on wing size and shape were modest. We examined shape-size allometric effects between groups. Allometric effects were generally similar across sexes, but differed substantially due to population of origin and plasticity. We discuss findings within the context of the evolution of plasticity for SSD, condition dependence and allometric relationships.

The shared genome prevents each sex from independently responding to the selection experienced by that sex. We used experimental evolution in Drosophila melanogaster with separate pools of Chromosome 3s for males (male-limited chromosomes) and females (female-limited chromosomes) for 15 generations. Viewing each sex as a separate environment, we performed a reciprocal transplant between the sexes to quantify the strength of local adaptation to each sex environment. Each chromosome type was more beneficial in the sex it had been selected for (i.e., local adaptation to sex). Because it has been postulated that sex differences in selection may depend on how well adapted a population is to the abiotic environment, we performed experimental evolution at two thermal regimes: one benign temperature to which the populations were well-adapted and one novel temperature. Female-specific adaptation was stronger at the benign temperature whereas male-specific adaptation was stronger in the novel temperature. Within chromosome pools, male and female fitness were more positively correlated in the novel compared to the benign temperature. Though males carrying male-limited chromosomes were typically more fit than males carrying female-limited chromosomes, they were also more harmful to their female mating partners.

Hosts can often evolve resistance to parasites (and other stressors), but such resistance is generally thought to be constrained by trade-offs with other traits. These trade-offs determine the hosts optimal resistance strategy and whether resistance cycles, diversifies, and/or is maintained in the absence of parasite. However, trade-offs are often inconsistently measured across experiments and can depend on environmental conditions. Here, we extend a selection experiment evolving resistance to viral infection under variable resource quality in the Plodia interpunctella model system to explore the evolutionary conditions leading to an incongruent earlier measurement of costless resistance. We find that environmental resource quality, historical contingency, and the time scale of selection all affect trade-offs in our long-term selection experiment. Specifically, populations selected for resistance with the dual stressor of low resource quality are slowed, but not prevented, from evolving resistance. Second, variation in starting populations or early sampled adaptations led to contingency towards context-dependent resistance. Finally, some costs to resistance observed at early time points were compensated over longer evolutionary time scales. Our work therefore informs perspectives for the predictability of adaptation and how variation in specific evolutionary conditions can alter the evolutionary trajectories of a population towards costly or costless resistance strategies.

The body size of adults and immature stages are fundamental animal traits that influence animal physiology, ecology, and range distribution. While the importance of egg size has been acknowledged as a proxy of parental investment in animals, little work has addressed the tempo and mode of evolution of egg size and shape. Here, we present a comparative study of this trait using a phylogeny based on genome-wide markers together with measurements of egg size and adult body size from 29 drosophilid species. Our analyses revisit the allometric relationship between egg size and body size and show that egg size scales negatively with respect to adult size, even after accounting for shared evolutionary history. In other words, larger species tend to produce proportionally smaller eggs. We also detect a moderate phylogenetic signal in both egg size and egg shape, indicating that closely related species resemble each other in these traits. Model comparisons show that the evolution of egg morphology in drosophilids is best described by gradual divergence through time driven by stochastic evolutionary change. This pattern contrasts with findings from other animal groups, including birds, cephalopods, and reptiles, where alternative evolutionary models better explain trait evolution. Together, these results suggest that the evolutionary dynamics shaping egg morphology in drosophilids differ from those operating in other major lineages and underscore the importance of comparative analyses of early developmental traits across taxa.

Phenotypic plasticity plays a key role in adaptation to changing environments. However, plasticity is neither perfect nor ubiquitous, implying that fitness costs must limit the evolution of phenotypic plasticity in nature. The measurement of such costs of plasticity has proved elusive; decades of experiments show that fitness costs of plasticity are often weak or nonexistent. Here, we show that this paradox can be at least partially explained by condition-dependence. We develop two models differing in their assumptions about how condition-dependence arises; both models show that variation in condition can readily mask costs of plasticity even when such costs are substantial. This can be shown simply in a model where costly plasticity itself evolves condition-dependence. Yet similar effects emerge from an alternative model where trait expression is condition-dependent. In this more complex model, average condition in each environment and genetic covariance in condition across environments both determine when costs of plasticity can be revealed. Analogous to the paradox of missing trade-offs between life history traits, our models show that variation in condition masks costs of plasticity even when costs exist, and suggests this conclusion may be robust to the details of how condition affects trait expression. Our models demonstrate that condition dependence can also account for the often-observed pattern of elevated plasticity costs inferred in stressful environments, the maintenance of genetic variance in plasticity, and provides insight into experimental and biological scenarios ideal for revealing a cost of phenotypic plasticity.

Many species pairs form F1 hybrids that are fitter than their parents. Such heterosis can arise if the parents carry recessive deleterious mutations; and in this case, the heterosis should be fixable, because selecting out the deleterious mutations yields a high-fitness homozygous hybrid. However, heterosis might not be fixable if caused by overdominance (an intrinisic advantage to heterozygosity) or if the parents contain coadapted gene complexes. These alternatives have been tested with introgression lines, where small regions of genome are scored in the heterospecific background. We develop predictions for introgression line data under a simple model of phenotypic selection, where parents diverge by fixing deleterious mutations via genetic drift. We show that this simple process can generate complex patterns in the data, misleading tests for both overdominance and coadaptation. We also suggest new ways to analyse the data to overcome these difficulties. Reanalyses of published data from Solanum and Gossypium suggest that the model can account for the qualitative patterns observed, though not the extent of apparent overdominance.