COS 7-6 - Spatial variability in granivory determines the strength of stochastic community assembly

Monday, August 8, 2011: 3:20 PM
8, Austin Convention Center
Rachel M. Germain1, Laura Johnson2, Andrew S. MacDougall2, Karl Cottenie3 and Elizabeth Gillis4, (1)Ecology & Evolutionary Biology, University of Toronto, Toronto, ON, Canada, (2)Integrative Biology, University of Guelph, Guelph, ON, Canada, (3)University of Guelph, ON, Canada, (4)Vancouver Island University, Nanaimo, BC, Canada

High diversity in many biological communities is not well explained by trait-based deterministic models, in part because stochastic processes also influence community assembly.  Testing how deterministic and stochastic processes combine to regulate diversity, however, has been limited by the spatial complexity of these interactions.  Granivory effects by small mammals often decrease with distance from nesting sites and under perceived predation risk, allowing plant assembly to unfold differently depending on whether small mammals are frequent, rare, or absent.  Here, we ask whether spatial variability in the impacts of granivory results in fine-scale switching between deterministically and stochastically regulated plant community assembly in an otherwise environmentally homogeneous 18-hectare tallgrass prairie.  To address this question, we combined experimental and survey based approaches through cafeteria trials, habitat manipulations of predation risk, small mammal tracking, and plant diversity surveys across our study site.  We predicted that small mammal impacts would decrease in high-risk open cover plots, as well as along a prairie edge-to-interior gradient, with corresponding increases in β-diversity across the plant community.


We determined that small mammals preferred the seeds of some species over others, indicating that palatability is a mechanism for deterministic assembly in this system.  Similarly, granivores were over 9x more abundant in closed compared to open cover plots.  Contrary to our predictions, small mammal impacts increased with distance from the prairie edge even though β-diversity increased as expected along the same gradient.  We attribute this discrepancy to the time at which small mammal abundances were surveyed.  In fall when we surveyed the small mammal community, the site was dominated by ruderal weeds that likely provided a safe haven for small mammals to invade the entire field.  In spring when the prairie was seeded, the site was devoid of all vegetation, resulting in a high-risk environment which likely restricted seed removal to the prairie edges.  As such, as granivory impacts decline in more open areas in association with perceived predator risk, assembly unfolds stochastically based on which species happen to land in a given location (high β-diversity).  The net effect across the study area is a heterogeneous and species-rich array of co-occurring species which is explained by cryptic but predictable fine-scale differences in the strength of limiting factors.

Copyright © . All rights reserved.
Banner photo by Flickr user greg westfall.