A macroevolutionary gene network reveals diapause evolutionary dynamics beyond the circadian clock and predicts microevolution

Diapause is an alternative developmental pathway evolved independently in many insects to synchronize life cycles with resource abundance. While subsets of this essential phenotype have long been studied at a single species level, the genomic basis of the full diapause syndrome remains poorly understood. Remaining unknown is whether convergent diapause syndromes employ shared mechanisms. This paucity of insights has fueled a long-standing debate about how life cycle synchronization evolves. Using a comparative genomic analysis spanning diverse diapause transitions in butterflies, we identified a large network of coevolving genes unique to diapausing species. The network is composed of functional modules spanning circadian regulation, metabolism, and cell cycle control. We tested whether this macroevolutionary scale network predicts microevolutionary dynamics, hypothesizing that this network is the polygenic architecture underlying the diapause syndrome. Analyses revealed that allelic variation in the diapause network is significantly enriched in signatures of local adaptation across latitudes, but only in diapausing species. Thus, we empirically show that diapause evolves through modular, coevolving gene networks, components of which regulate species/population specific diapause phenotypes. This novel perspective on this complex phenotype opens a new horizon of inquiry for understanding seasonal adaptation across evolutionary scales, while demonstrating the power of using comparative genomics to dissect polygenic phenotypes.