Major evolutionary innovations can greatly influence subsequent evolution. While many major transitions occurred in the deep past, male live-bearing fishes (family Poeciliidae) more recently evolved a novel body plan. This group possesses a three-region axial skeleton, with one region—the ano-urogenital region—representing a unique body region accommodating male genitalic structures (gonopodial complex). Here we evaluate several hypotheses for the evolution of diversity in this region and examine its role in the evolution of male body shape. Examining Gambusia fishes, we tested a priori predictions for 1) joint influence of gonopodial-complex traits on mating performance, 2) correlated evolution of gonopodial-complex traits at macro- and microevolutionary scales, and 3) predator-driven evolution of gonopodial-complex traits in a post-Pleistocene radiation of Bahamas mosquitofish. We found the length of the sperm-transfer organ (gonopodium) and its placement along the body (gonopodial anterior transposition) jointly influenced mating success, with correlational selection favoring particular trait combinations. Despite these two traits functionally interacting during mating, we found no evidence for their correlated evolution at macro- or microevolutionary scales. In contrast, we did uncover correlated evolution of modified vertebral hemal spines (part of the novel body region) and gonopodial anterior transposition at both evolutionary scales, matching predictions of developmental connections between these components. Developmental linkages in the ano-urogenital region apparently play key roles in evolutionary trajectories, but multiple selective agents likely act on gonopodium length and cause less predictable evolution. Within Bahamas mosquitofish, evolution of hemal-spine morphology and gonopodial anterior transposition across predation regimes was quite predictable, with populations evolving under high predation risk showing more modified hemal spines with greater modifications and a more anteriorly positioned gonopodium. These changes in the ano-urogenital vertebral region have facilitated adaptive divergence in swimming abilities and body shape between predation regimes. Gonopodium surface area, but not length, evolved as predicted in Bahamas mosquitofish, consistent with a previously suggested tradeoff between natural and sexual selection on gonopodium size. These results provide insight into how restructured body plans offer novel evolutionary solutions. Here, a novel body region—originally evolved to aid sperm transfer—was apparently co-opted to alter whole-organism performance, facilitating phenotypic diversification.

Data for this study comprises several different types of datasets:

1. Laboratory mating experiment measuring the effects of two morphological traits on mating success.

2. Comparative analysis of morphological trait evolution using phylogenetically indepedent contrasts (using separate phylogenies based on either nuclear and mitochondrial DNA). Measurements of traits from digital x-ray images.

3. Tests of morphological differentiation between predation regimes using wild-caught Bahamas mosquitofish from blue holes on Andros Island, The Bahamas (individual dataset and population-mean dataset). Measurements of traits from digital x-ray images.

4. Tests of genetic basis for morphological differentiation between populations using a laboratory common-garden experiment using Bahamas mosquitofish from 8 blue holes on Andros Island, The Bahamas. Measurements of traits from digital x-ray images.

Information for the datasets are provided in the first sheet of the Excel file.