Parallel genetic evolution and speciation from standing variation

Abstract Adaptation often proceeds via the sorting of standing variation, and natural selection acting on pairs of populations is a quantitative continuum ranging from parallel to divergent. Yet, it is unclear how the extent of parallel genetic evolution during adaptation from standing variation is affected by the difference in the direction of selection between populations. Nor is it clear whether the availability of standing variation for adaptation affects progress toward speciation in a manner that depends on the difference in the direction of selection. We conducted a theoretical study investigating these questions and have two primary findings. First, the extent of parallel genetic evolution between two populations is expected to rapidly decline as the difference in their directions of selection increases from fully parallel toward divergent, and this decline occurs more rapidly in organisms with greater trait ‘dimensionality’. This rapid decline results because seemingly small differences in the direction of selection cause steep reductions in the fraction of alleles that are beneficial in both populations. For example, populations adapting to optima separated by an angle of 33° have only 50% of potentially beneficial alleles in common (for a case of five trait ‘dimensions’). Second, we find that adaptation from standing variation leads to higher ecologically-dependent hybrid fitness under parallel selection, relative to when adaptation is from new mutation only. This occurs because genetic parallelism based on standing variation reduces the phenotypic segregation variance in hybrids when parents adapt to similar environments. In contrast, under divergent selection, the pleiotropic effects of alternative alleles fixed from standing variation change the major axes of phenotypic variation in hybrids and reduce their fitness in parental habitats. We conclude that adaptation from standing genetic variation is expected to slow progress toward speciation via parallel natural selection and can facilitate progress toward speciation via divergent natural selection. Impact summary It is increasingly clear that much of adaptation, especially that which occurs rapidly, proceeds from the sorting of ancestral standing variation rather than complete reliance on de novo mutation. In addition, evolutionary biologists are increasingly embracing the fact that the difference in the direction of natural selection on pairs of populations is a quantitative continuum ranging from completely parallel to completely divergent. In this article, we ask two questions. First, how does the degree of genetic parallelism—here, adaptation using the same alleles in allopatric populations—depend on the differences in the direction of natural selection acting on two populations, from parallel (0°) to divergent (180°)? And second, how does adaptation from standing variation affect progress toward speciation, and does its effect depend on the direction of natural selection? We develop theory to address these questions. We first find that very small differences in the direction of selection (angle) can largely preclude genetic parallelism. Second, we find that adaptation from standing variation has implications for speciation that change along the continuum from parallel to divergent selection. Under parallel selection, high genetic parallelism causes inter-population hybrids to have high mean fitness when their parents adapt from standing variation. As selection tends toward divergent, adaptation from standing variation is less beneficial for hybrid fitness and under completely divergent selection causes inter-population hybrids to have lower mean fitness than when adaptation was from new mutation alone. In sum, our results provide general insight into patterns of genetic parallelism and speciation along the continuum of parallel to divergent natural selection when adaptation is from standing variation.

Récupéré en direct depuis OpenAlex et désinversé. Les résumés ne sont pas conservés dans cette base de données : les index inversés représentent 8,6 Go des 9,3 Go de texte de la base, et le serveur dispose de 13 Go libres.