Testing the adaptive decoupling hypothesis in a hypermetaphorphic and sexually dimorphic insect

Abstract

Though seemingly bizarre, the dramatic post-embryonic transformation that occurs during metamorphosis is one of the most widespread and successful developmental strategies on the planet. The adaptive decoupling hypothesis (ADH) proposes that metamorphosis is an adaptation for optimizing expression of traits across life stages that experience opposing selection pressures. Similarly, sex-biased expression of traits is thought to evolve in response to sexually antagonistic selection. Both hypotheses predict that traits will be genetically decoupled among developmental stages and sexes, but direct comparisons between stage-specific and sex-specific decoupling are rare. Additionally, tests of the ADH have been hampered by a lack of suitable traits for among-stage comparisons and by uncertainties regarding how much decoupling is to be expected. To fill these voids, we characterize transcriptome-wide patterns of gene-expression decoupling in the hypermetamorphic and sexually dimorphic insect, Neodiprion lecontei. This species has three ecologically and morphologically distinct larval stages separated by molts, as well as a complete metamorphic transition that produces dimorphic adult males and females. Consistent with the ADH, we observe that: (1) the decoupling of gene expression becomes more pronounced as the ecological demands of developmental stages become more dissimilar and (2) gene-expression traits that mediate changing ecological interactions show stronger and more variable decoupling than expression traits that are likely to experience more uniform selection. We also find that gene-expression decoupling is more pronounced among major life stages than between the sexes. Overall, our results demonstrate that patterns of gene-expression decoupling can be predicted based on gene function and organismal ecology.