A major goal of biology is to understand the origins and maintenance of phenotypic differences between closely related species. Eco-morphology and ornamentation are two phenotypic dimensions along which co-existing species often diverge, yet theory makes contrasting predictions about how these phenotypes diversify relative to each other. Some theory predicts that intense, idiosyncratic sexual selection will cause more pronounced divergence in ornamentation than in eco-morphology. Other theory predicts that the ability of condition-dependent ornaments to signal local specialization will result in different species evolving similar ornaments but very divergent eco-morphology. Here, we evaluated these conflicting predictions in Nearctic Libelluloidea dragonflies by testing if the diversification of a condition-dependent ornament, male wing melanization, was slower and less pronounced between closely related species than the diversification of two key eco-morphological traits, body size and relative wing size. Our results show that male wing melanization evolved much faster than either body size or relative wing size. Furthermore, in contrast to the patterns for either eco-morphological trait, the best-supported models of diversification in male wing melanization indicate that the majority of divergence arose between the most closely related species. These results reveal that the primary axis of divergence between closely related Libelluloidea dragonflies is ornamentation rather than eco-morphology. Our study therefore suggests that evolutionary responses to disparate reproductive demands may be fundamental to the persistence and co-existence of closely related species.

Data on species' traits was gathered from field guides (body size, relative wing size) and measured from citizen-science observations (male wing melanization)

R