The phoenix hypothesis of speciation

Abstract

Genetic divergence among allopatric populations builds reproductive isolation over time and is thought to be the major mechanism underlying the formation of new species. This process is accelerated when populations face a changing environment, but abrupt change also places populations at risk of extinction. Here we use simulations of Fishers geometric model with explicit population dynamics to explore the genetic changes that occur in the face of extreme environmental changes to which populations must adapt or go extinct. We show that evolutionary rescue leads to the fixation of mutations whose effects are larger on average and that these mutations are more likely to lead to reproductive isolation, compared with populations not at risk of extinction. We refer to the formation of new species from the ashes of populations in decline as the phoenix hypothesis of speciation. The phoenix hypothesis predicts more substantial hybrid fitness breakdown among populations surviving a higher extinction risk. The hypothesis was supported when many loci underlie adaptation. When, however, there was only a small number of potential rescue mutations, we found that mutations fixed in different populations were more likely to be identical, with parallel changes reducing isolation. With a limited genomic potential for adaptation, we find support for a modified version of the phoenix hypothesis where reproductive isolation builds fastest in populations subject to an intermediate extinction risk. While processes driving extinction lead to the loss of lineages with deep evolutionary histories, they may also generate new taxa, albeit taxa with minimal genetic differences.