Counter-gradient variation in gene expression between fish populations facilitates colonization of low-dissolved oxygen environments.

Abstract

The role of phenotypic plasticity during colonization remains unclear due to the shifting importance of plasticity across timescales. In the early stages of colonization, plasticity can facilitate persistence in a novel environment; but over evolutionary time, processes such as genetic assimilation may reduce variation in plastic traits such that species with a longer evolutionary history in an environment can show lower levels of plasticity than recent invaders. Therefore, comparing species in the early stages of colonization to long-established species provides a powerful approach for uncovering the role of phenotypic plasticity during different stages of colonization. We compared gene expression between low-dissolved oxygen (DO) and high-DO populations of two cyprinid fish: Enteromius apleurogramma, a species that has undergone a recent range expansion, and E. neumayeri, a long-established native species in the same region. We sampled tissue either immediately after capture from the field or after a 2-week acclimation under high-DO conditions, allowing us to test for both evolved and plastic differences in low-DO vs high-DO populations of each species. We found that most genes showing candidate-evolved differences in gene expression did not overlap with those showing plastic differences in gene expression. However, in the genes that did overlap, there was counter-gradient variation such that plastic and evolved gene expression responses were in opposite directions in both species. Additionally, E. apleurogramma had higher levels of plasticity and evolved divergence in gene expression between field populations. We suggest that the higher level of plasticity and counter-gradient variation may have allowed rapid genetic adaptation in E. apleurogramma and facilitated colonization. This study shows how counter-gradient variation may impact the colonization of divergent oxygen environments.