COS 136-6
Changes in grassland plant-pollinator community structure with management and scale

Friday, August 14, 2015: 9:50 AM
318, Baltimore Convention Center
Ellen Welti, Division of Biology, Kansas State Univerisity, Manhattan, KS
Anthony Joern, Division of Biology, Kansas State University, Manhattan, KS
Background/Question/Methods

Species interactions underlie dynamics of the structure and function of ecological communities. Recent research using an ecological network approach emphasizes inherent community-level structural properties among interacting species groups to understand system robustness to extinction cascades. Plant-pollinator communities are a useful system in which to study community persistence because plants and pollinators are directly dependent on one another (i.e., are mutualistic), provide the critical ecosystem service of pollination, and interactions are readily observable in the field. Nestedness of species interactions and degree of specialization are two network structural properties argued to largely affect the robustness of mutualistic networks. In this study, we investigated now the network structural properties of specialization, nestedness and robustness of grassland plant-pollinator communities varied with fire frequency and ungulate grazing regimes and with changes in sampling scale. We hypothesized that nestedness, specialization and robustness would increase with scale and disturbance (more frequent fires and presence of large ungulate grazers). Sampling was conducted at 12 sites on Konza Prairie Biological Station, a tallgrass prairie located in NE Kansas in early June and early July. A Partial Redundancy Analysis was used to assess effects of grassland management regimes on network structural properties.

Results/Conclusions

Sampling of pollinator communities included a total 7,070 individual insects belonging to 353 morphospecies. Insects were collected on 44 flowering plant species and 29,324 individual flowers were counted in surveyed sites. Model selection of the most important habitat variables explaining variance in network structural properties resulted in a top model including grazing, burning, and conditional on sampling date. Grazing was the most important habitat factor, explaining 20.9% of the variance in plant-pollinator network structure. The presence of large ungulate grazers was associated with increases in nestedness but had negligible effects on degree of specialization. Mean robustness of plants and pollinators did not change with grazing treatment but was more variable in ungrazed compared to grazed sites. Nestedness did not change with scale (R2<0.01, F1,84=0.01, P=0.93) nor did specialization (R2=0.03, F1,84=3.2, P=0.08). The robustness of plants did not change with scale (R2=0.04, F1,84=3.1, P=0.08) whereas the robustness of pollinators significantly increased (R2=0.21, F1,84=21.9, P<0.001). This research helps access how the disturbances of fire and grazing affect the robustness of plant-pollinator communities at different scales.