Description
Considering the large number of ecological network studies that have examined how species interactions and community structure change through time, it is surprising that seasonal (i.e., intra-annual) dynamics have received so little attention. Intra-annual dynamics appear to be particularly important in plant-pollinator communities, where daily and seasonal shifts in the abundance and behavior of species play a role in structuring both interactions and entire networks. Downscaling analyses from yearly intervals to timeframes that more closely reflect the biology of individuals and species (i.e., intra-annual time scales), may be a step towards separating the influence of phenological coincidence from other constraints, such as host choice, on species interactions and network structure. Here, I investigated the intra-annual dynamics of a plant-pollinator network by examining temporal patterns (1) of interaction turnover between census periods, (2) in network membership and the roles assigned to individual species by modularity algorithms, and (3) in classic metrics describing network topology. I conducted field observations of plant-pollinator interactions across a single season at the Crestridge Ecological Reserve (El Cajon, CA), which supports a mixture of coastal sage scrub and chaparral habitats. Interaction turnover increased with temporal distance between networks, and the magnitude of turnover was substantial (> 55%) between networks at all temporal distances. Week-to-week differences were driven largely by changes in flower-visitor composition. However, simultaneous loss of plant-pollinator pairs was the dominate driver of interaction turnover when considering temporal distances of seven weeks or greater. Host-switching (i.e., interaction rewiring) accounted for less than 28% of turnover at all temporal distances. Among flower-visitors, Apis mellifera consistently occupied important roles in the structure of monthly networks. Surprisingly, flowering-plants rarely occupied important roles in network structure, despite having high species degree(s) relative to the average flower-visitor. Although network membership and topology differed strongly between months, May and June exhibited similar structural complexity, and contained a larger number of species than other monthly networks. These findings highlight some advantages of viewing plant-pollinator networks at the intra-annual scale, rather than the annual scale where many interesting dynamics may be overlooked.
