Pupal remodeling and the development and evolution of sexual dimorphism in horned beetles

Horns or hornlike structures in beetles have become an increasingly popular study system for exploring the evolution and development of secondary sexual trait diversity and sexual dimorphisms. The horns of adult beetles originate during a rapid growth phase during the prepupal stage of larval development, and differential activation of growth during this time is either implicitly or explicitly assumed to be the sole mechanism underlying intra- and interspecific differences in adult horn expression. Here I show that this assumption is not based on developmental reality. Instead, after their initial prepupal growth phase, beetle horns are extensively remodeled during the subsequent pupal stage via sex- and size-dependent resorption of horn tissue. I show that adult sexual dimorphism in four Onthophagus species is shaped partly or entirely by such pupal remodeling rather than by differential growth. Specifically, I show that after a sexually monomorphic growth phase, differential pupal horn resorption can generate both regular and reversed sexual dimorphism. Furthermore, I show that in cases in which initial growth is already dimorphic, pupal horn resorption can both magnify and reverse initial dimorphism resulting from differential growth. Finally, I show that complete resorption of pupal horns in both sexes can remove any trace of horn expression from all resulting adults. In such species, examination of adults only would result in the false conclusion that this species lacks the ability to develop a horn. Instead, such species appear to differ from those with sexually dimorphic adults merely in that they activate pupal horn resorption in both sexes rather than in just one. Combined, these results suggest that pupal remodeling of secondary trait expression is taxonomically widespread, at least among Onthophagus species, and is developmentally extensive and remarkably evolutionarily labile. These results have immediate implications for reconstructing the evolutionary history of horned beetles and the role of developmental processes in guiding evolutionary trajectories. I use these results to revise current understanding of the evolutionary developmental biology of secondary sexual traits in horned beetles in particular and holometabolous insects in general. The results presented here seriously call into question whether descriptions of adult diversity patterns alone suffice for meaningful inferences toward understanding the developmental and evolutionary origin of these patterns. These results illustrate that a lasting integration of development into an evolutionary framework must integrate development as a process rather than define it solely by some of its products.