Adaptation in sulfide spring fishes: experimental approaches

Abstract

Poecilia mexicana is a live-bearing fish known for its extremophilic populations. The species has been studied for their high tolerance to hydrogen sulfide (H₂S), an environmental toxicant that occurs naturally in some freshwater springs in southern Mexico. Though studies often reduce this system to a presence or absence of H₂S, focusing entirely on H₂S is likely to generate an incomplete picture. H₂S springs are complex habitats, differing in biotic and abiotic conditions besides the presence of H₂S. For example, sulfidic habitats are characterized by hypoxia, low pH, different trophic regimes and predators, and strong energy constraints. How these other sources of selection impact the evolution of extremophile fish populations, and how they might interact with H₂S, remains largely unstudied. My research focused on two questions within the P. mexicana system, where one question acknowledged the complexity of environment and organism, and the other consciously stripped away complexity to isolate the direct impacts of H₂S. My first chapter embraced environmental and organismal complexity to investigate phenotypic differences between sulfidic and nonsulfidic populations of P. mexicana. Though H₂S is commonly identified as the driving selective force between the two populations, not all divergent traits exhibit clear adaptive hypotheses related to this toxicant. I compared phenotypic traits between populations in the wild, then compared those same traits in first-generation individuals from a common-garden environment. Almost every trait showed significant divergence between wild sulfidic and nonsulfidic populations. Most traits remained different between populations reared in the lab, though many simultaneously presented evidence of plasticity. My results suggested that trait divergences between populations in different habitats were largely shaped by evolutionary change, and I developed hypotheses about the potential adaptive function of divergent traits in the context of the complex selective regimes in sulfide springs. My second chapter examined H₂S responses outside of environmental and organismal context. To test what organismal responses are a direct response of H₂S alone, I analyzed the gene expression response of P. mexicana to an H₂S-releasing pharmaceutical compound (AP39) that targets the mitochondria. I exposed both sulfidic and nonsulfidic populations to AP39, then compared their gene expression in the liver and the gills to individuals not exposed to AP39. I found no gene expression differences between the gill tissue of AP39-exposed and unexposed individuals. While there were significant differences in gene expression in livers, none of the expressed genes were related to H₂S tolerance or detoxification. I concluded that AP39 makes a poor model for H₂S exposure in P. mexicana. My findings underscore the importance of context. Exposing just the mitochondria to H₂S via a targeted compound did not evoke the same responses as exposing the whole organism via H₂S dissolved in water, as it is found in nature. Overall, my research attests that evolutionary biologists cannot ignore environmental and organismal complexity, even in systems with strong selective pressures.