Floral resource diversity drives spatiotemporal variation in plant–pollinator network structure

Mechanisms underlying community assembly, including those related to species interactions, vary across space and time. Plant–pollinator networks exemplify these dynamics, where link rewiring and turnover mediate adaptations to environmental changes. Bees rely on diverse floral resources (e.g. nectar, oil, pollen) for distinct life‐history aspects. However, how spatial and temporal variation in floral resources influences bee preferences, and the structure and assembly of plant–bee networks, remains understudied. This limits our ability to predict plant–bee pairings and, consequently, to understand interaction dynamics. Using finely resolved plant–bee interaction networks, we investigated how floral resource diversity modulates plant–bee network structure (e.g. modularity and nestedness), and plant network roles (i.e. participation in motifs of different classes, and contribution to nestedness and modularity). We found that spatiotemporal variation in floral resource diversity shapes plant–bee networks in complex ways across different network scales (i.e. from species roles and motifs to network structure). Specifically, while lower floral resource diversity increased network nestedness, higher diversity of floral resources increased network modularity. The effects of floral resource diversity across network scales (i.e. plant network roles and network structure) did not follow a consistent pattern, suggesting that floral resource diversity affected species roles and participation in motifs differently from the broader network structure. For example, plants offering specialised floral resources (e.g. oil) consistently increased modularity with greater floral diversity, as they interacted more within their modules and were associated with motif classes that promote modularity. In contrast, plants offering generalised resources (e.g. nectar) showed variable responses to increasing floral resource diversity across network scales. Our findings emphasise the importance of accounting for spatiotemporal changes in floral resources when studying plant–bee networks, and illustrate how including resource‐level information adds to our understanding of plant–pollinator network configuration.