Genital coevolution is a pervasive phenomenon as changes in one sex tend to impose fitness consequences on the other generating sexual conflict. Sexual conflict is often thought to cause stronger selection on males due to the Darwin-Bateman’s anisogamy paradigm. However, recent studies have demonstrated that female genitalia may be equally elaborated and perform diverse extra-copulatory functions. These characteristics suggest that female genitals can also be primary targets of selection, especially where natural selection acts on female-exclusive functions such as oviposition. Here, we test this hypothesis in a statistical phylogenetic framework across the whole beetle (Coleoptera) phylogeny, investigating whether coevolution of specific genital traits may be triggered by changes in females. We focus on traits of the proctiger, which composes part of the male terminalia and the female ovipositor. Our results present a comprehensive case of male-female genital coevolution, and provide solid statistical evidence for a female-initiated coevolutionary process where the vast majority of evolutionary transitions in males have occurred only after changes in females. We corroborate the hypothesis that female traits may change independently and elicit counter-adaptations in males. Furthermore, by showing a consistent pattern across the phylogeny of the most diverse group of animals, our results suggest that this female-driven dynamics may persist through long time scales.

The files herein contain morphological datasets (Baculi.csv and Membranization.csv) used for all MCMC modelling procedures and the phylogenetic tree inferred in the same work (coleopteraTREE.tre). These files allow for replication of all our analyses. Raw morphological information were collected from Lawrence et al. (Annales Zoologici, 2011, 61(1):1-217), then re-coded as two binary traits: (a) proctegier membranization (absence/presence) for males and females and (b) proctigeral baculi (absence/presence) for males and females. In the morphological datasets, "0" denotes absence and "1" denotes presence. Please, see the methods section in our paper for detailed description of the analytical procedures.