The evolution of floral scent and olfactory preferences in pollinators: coevolution or pre-existing bias?

Coevolution is thought to be a major factor in shaping plant–pollinator interactions. Alternatively, plants may have evolved traits that fitted pre-existing preferences or morphologies in the pollinators. Here, we test these two scenarios in the plant family of Araceae and scarab beetles (Coleoptera, Scarabaeidae) as pollinators. We focused on floral volatile organic compounds (VOCs) and production/detection of VOCs by scarab beetles. We found phylogenetic structure in the production/detection of methoxylated aromatics in scarabs, but not plants. Within the plants, most of the compounds showed a well-supported pattern of correlated evolution with scarab-beetle pollination. In contrast, the scarabs showed no correlation between VOC production/detection and visitation to Araceae flowers, with the exception of the VOC skatole. Moreover, many VOCs were found in nonpollinating beetle groups (e.g., Melolonthinae) that are ancestors of pollinating scarabs. Importantly, none of the tested VOCs were found to have originated in pollinating taxa. Our analysis indicates a Jurassic origin of VOC production/detection in scarabs, but a Cretaceous/Paleocene origin of floral VOCs in plants. Therefore, we argue against coevolution, instead supporting the scenario of sequential evolution of floral VOCs in Araceae driven by pre-existing bias of pollinators