Submitted to: Ecological Society of America Abstracts
Publication Type: Abstract Only
Publication Acceptance Date: 7/20/2003
Publication Date: 8/20/2003
Citation: Goslee, S.C., Tracy, B.F., Sanderson, M.A. 2003. Plant community structure of managed grasslands at a range of spatial scales [abstract]. Ecological Society of America Abstracts. CDROM.
Technical Abstract: The composition and structure of managed grassland communities can have a large impact on their productivity and sustainability. As part of a project studying the role of diversity in grasslands, plant community composition was surveyed at 88 sites between 1998 and 2002. A modified Whittaker plot sampling design was used, allowing the analysis of community structure at different spatial scales. All potential pairwise relationships were quantified using an asymmetric coefficient. A graph-theoretic framework was employed to visualize the structure of these relationships at spatial scales from 1m2 to 1000m2. At the finest scale, only 2% of the possible pairwise relationships were significant, and 75% of those were negative. At the coarsest scale, less than 1% of the possible relationships were significant, and 63% of those were negative. The coarse scale showed a lower proportion of symmetric relationships. Analysis of indegree (association to) and outdegree (association from) in the graphs of these relationships showed a few key species of high degree. Associations with these species were primarily negative at the fine scale, and positive at the coarse scale. Relationships at one scale were poor predictors of relationships at other scales. The coarse-scale patterns respond to environmental conditions and management effects, while the fine-scale patterns are mainly determined by local plant interactions. Development of management practices to affect grassland community composition must take into account processes at both scales. The graph-theoretic approach presented here provides an analytic framework that allows the examination of complex community structure.