
Hybridization Outcomes Have Strong Genomic and Environmental Contingencies
Author(s) -
Claire Brice,
Zebin Zhang,
Devin P. Bendixsen,
Rike Stelkens
Publication year - 2021
Publication title -
the american naturalist
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 1.954
H-Index - 205
eISSN - 1537-5323
pISSN - 0003-0147
DOI - 10.1086/715356
Subject(s) - transgressive segregation , biology , transgressive , hybrid , context (archaeology) , evolutionary biology , genetic fitness , genetic algorithm , adaptation (eye) , genetics , ecology , gene , quantitative trait locus , sedimentary depositional environment , paleontology , botany , structural basin , neuroscience
Extreme F 2 phenotypes known as transgressive segregants can cause increased or decreased fitness in hybrids beyond the ranges seen in parental populations. Despite the usefulness of transgression for plant and animal breeding and its potential role in hybrid speciation, the genetic mechanisms and predictors of transgressive segregation remain largely untested. We generated seven hybrid crosses between five widely divergent Saccharomyces yeast species and measured the fitness of the parents and their viable F 1 and F 2 hybrids in seven stressful environments. We found that on average 16.6% of all replicate F 2 hybrids had higher fitness than both parents. Against our predictions, transgression frequency was not a function of parental genetic and phenotypic distances across test environments. Within environments, some relationships were significant, but not in the predicted direction; for example, genetic distance was negatively related to transgression in ethanol and hydrogen peroxide. Significant effects of hybrid cross, test environment, and cross × environment interactions suggest that the amount of transgression produced in a hybrid cross is highly context specific and that outcomes of hybridization differ even among crosses made from the same two parents. If the goal is to reliably predict hybrid fitness and forecast the evolutionary potential of admixed populations, we need more efforts to identify patterns beyond the idiosyncrasies caused by specific genomic or environmental contexts.