Interactions among plant and pollinator mutualists form networks that often conform to a characteristic structure. The structure of plant-pollinator networks may be neutrally influenced by the number and abundance of different species, and by the composition of species due to trait matching constraints. Both local and landscape scale habitat variation may affect local species richness, composition and visitation rate. However, little is known about how such landscape-scale spatial variation can alter the quantitative structure of mutualistic networks. We examined the effects of habitat fragmentation on the structure of local plant-pollinator networks in a region of Florida longleaf pine forest and ask: (1) How does network structure change with surrounding habitat fragmentation and (2) is the landscape effect on local networks due to changes in abundance, richness, and/or species composition?
We constructed 15 local plant-pollinator networks using visitation rate and duration as quantitative measures of interaction strength. Sites have similar local habitat quality but span a gradient of habitat fragmentation in the surrounding landscape. We used a variance partitioning approach to examine how variation in landscape structure affects local networks: by altering local abundance, richness, or composition of the plant/pollinator community.
Results/Conclusions
Plant-pollinator interactions were well sampled and networks were highly resolved. A standardized sampling procedure along with independent measures of plant and pollinator abundance allowed us to account for sampling effects. Results indicate that surrounding landscape context is important for local network structure. The effects of habitat fragmentation on local network structure are expressed mostly through landscape dependent changes in local species richness and abundance, and to a lesser degree via species composition. The landscape dependent effects of species richness and composition on network structure appear to be independent, which suggests that the spatial heterogeneity in the landscape can influence network structure in multiple and different ways that are consistent with interaction neutrality and trait matching mechanisms. We discuss how landscape structure can influence the relative importance of these two mechanisms of network organization.