Temporal variation in plant-pollinator networks from seasonal tropical environments: Higher specialization when resources are scarce

The temporal dynamics of plant phenology and pollinator abundance across seasons should influence the structure of plant–pollinator interaction networks. Nevertheless, such dynamics are seldom considered, especially for diverse tropical networks. Here, we evaluated the temporal variation of four plant–pollinator networks in two seasonal ecosystems in Central Brazil (Cerrado and Pantanal). Data were gathered on a monthly basis over 1 year for each network. We characterized seasonal and temporal shifts in plant–pollinator interactions, using temporally discrete networks. We predicted that the greater floral availability in the rainy season would allow for finer partitioning of the floral niche by the pollinators, i.e. higher specialization patterns as previously described across large spatial gradients. Finally, we also evaluated how sampling restricted to peak flowering period may affect the characterization of the networks. Contrary to our expectations, we found that dry season networks, although characterized by lower floral resource richness and abundance, showed higher levels of network‐wide interaction partitioning (complementary specialization and modularity). For nestedness, though, this between‐seasons difference was not consistent. Reduced resource availability in the dry season may promote higher interspecific competition among pollinators leading to reduced niche overlap, thus explaining the increase in specialization. There were no consistent differences between seasons in species‐level indices, indicating that higher network level specialization is an emergent property only seen when considering the entire network. However, bees presented higher values of specialization and species strength in relation to other groups such as flies and wasps, suggesting that some plant species frequently associated with bees are used only by this group. Our study also indicates that targeted data collection during peak flowering generates higher estimates of network specialization, possibly because species activity spans longer periods than the targeted time frame. Hence, depending on the period of data collection, different structural values for the networks of interactions may be found. Synthesis. Plant–pollinator networks from tropical environments have structural properties that vary according to seasons, which should be taken into account in the description of the complex systems of interactions between plants and their pollinators in these areas.