Adaptive rewiring and temperature tolerance shape the architecture of plant-pollinator networks globally

Abstract

Rising environmental temperatures are rapidly reshaping plant-pollinator communities by altering species traits and interaction patterns. We develop a simple eco-evolutionary model that integrates species-specific temperature tolerance curves with phenotype-based interaction dynamics. Across temperature gradients, species adaptively rewire, that is, they change their interaction partners. This rewiring is an emergent property of our model, driven by temperature-mediated selection and co-evolutionary trait matching. As temperature increases, our model predicts a consistent decline in network-level specialization, alongside increasing connectance and nestedness which are signatures of structural re-organization. These predictions are supported by empirical patterns from 165 plant-pollinator networks worldwide, where mean annual temperature correlates positively with connectance and nestedness, and negatively with network specialisation. Our findings suggest that temperature-driven trait evolution and emergent adaptive rewiring govern the assembly and architecture of mutualistic networks. By bridging dynamical eco-evolutionary theory with global empirical data, this work reveals the central role of trait-based processes in structuring biodiversity under ongoing and accelerating climate warming.