An amphibian toxin phenotype is sexually dimorphic and shows seasonal concordant change between sexes

Recently, several authors have challenged the view that anisogamy, the defining feature of the sexes, is an important determinant of the evolution of sex roles. Sex roles are instead suggested to result from chance, or from non-heritable differences in life histories of females and males. Here, we take issue with these ideas. We note that random processes alone cannot cause consistent differences between the sexes, and that those differences between the sexes in life histories that affect the sex roles are themselves the result of sex-specific selection that can ultimately be traced back to anisogamy. To understand sex roles, one should ask how environmental variation and female-male coevolution cause variation in sex-specific selection in the light of anisogamy. Copyright © 2012 Elsevier Ltd. All rights reserved.

Natural selection often produces convergent changes in unrelated lineages, but the degree to which such adaptations occur via predictable genetic paths is unknown. If only a limited subset of possible mutations is fixed in independent lineages, then it is clear that constraint in the production or function of molecular variants is an important determinant of adaptation. We demonstrate remarkably constrained convergence during the evolution of resistance to the lethal poison, tetrodotoxin, in six snake species representing three distinct lineages from around the globe. Resistance-conferring amino acid substitutions in a voltage-gated sodium channel, Na(v)1.4, are clustered in only two regions of the protein, and a majority of the replacements are confined to the same three positions. The observed changes represent only a small fraction of the experimentally validated mutations known to increase Na(v)1.4 resistance to tetrodotoxin. These results suggest that constraints resulting from functional tradeoffs between ion channel function and toxin resistance led to predictable patterns of evolutionary convergence at the molecular level. Our data are consistent with theoretical predictions and recent microcosm work that suggest a predictable path is followed during an adaptive walk along a mutational landscape, and that natural selection may be frequently constrained to produce similar genetic outcomes even when operating on independent lineages.