COS 31-5 - Which processes drive ant community assembly in experimentally fragmented forest?

Tuesday, August 8, 2017: 9:20 AM
C120-121, Oregon Convention Center
Kendi F. Davies1, Jeffrey L. McClenahan1, Saul Cunningham2, Andrew J. King3, Brett A. Melbourne1, Julian Resasco1 and Kika Tuff1,4, (1)Department of Ecology and Evolutionary Biology, University of Colorado at Boulder, Boulder, CO, (2)Fenner School of Environment and Society, The Australian National University, ACTON, Australia, (3)Land and Water, CSIRO, Canberra, Australia, (4)Founder, Creative Director, Impact Media Lab
Background/Question/Methods

Understanding the processes that determine community structure is a fundamental problem in ecology, with potential to directly affect the management and conservation of ecosystems. When landscapes are fragmented, the roles of dispersal and selection by the environment can be dramatically altered, resulting in significant changes to community assembly. We use a long-term, large-scale fragmentation experiment (Wog Wog, NSW, Australia) to study the relative importance of dispersal and selection in ant communities where the spatial context of communities has been experimentally altered, compared to unaltered controls. We hypothesize that experimental fragmentation will alter community assembly by: 1) changing the environment (the distribution of species’ niches) and thereby altering the distribution and abundance of species on fragments; 2) reducing dispersal between fragments and isolating small populations on fragments, thus reducing the number of species on fragments either deterministically or stochastically.

Results/Conclusions

Fragmentation reduced ant richness but increased beta diversity in fragments. The increased beta diversity of fragments was driven by both increased environmental heterogeneity in large fragments (variance in litter depth: 64% contribution to effect size) and an increase in dispersal between fragments (29% contribution to effect size), with a minor role for other direct effects of fragmentation (6% contribution to effect size). Finally, compared to controls, fragmentation increased the role of dispersal in the assembly of communities in small fragments, and, to a lesser extent, medium fragments, while at the same time increasing the role of selection in the assembly of communities in large fragments. Thus, we show experimentally that the processes that determine community assembly (selection, dispersal) are altered when the spatial context of a community is altered, and that fragment size can change the relative roles of the processes that drive community assembly, increasing the role of dispersal in small fragments and selection in large fragments. Like other temperate systems, selection is an important driver of community assembly at Wog Wog but is subordinate to dispersal on small fragments where there is more dispersal from the matrix and where litter depth is less variable.