Document: DAV-3-40-45

Plant species richness on the Shoshone National Forest, Wyoming: The continuum and the cosmic pear.

ROBERTS, D.W.* ¹, M.R.WEIDNER ¹, K.J.MCCLOSKEY ¹ and K.E.HOUSTON ²

Utah State University, Logan, UT 84322 ¹
USDA Forest Service, Shoshone National Forest ²

Abstract:
Classical plant community diversity research emphasized the distribution of species richness and alpha diversity along direct environmental gradients. Current work in community diversity has emphasized productivity, speciation, dispersal, and geomorphic processes as determinants of local species richness. Recently, Austin has emphasized the importance of the continuum concept in diversity research. In response, we analyzed the distribution of species richness for 1200 vegetation samples from the Shoshone National Forest, Wyoming, using Generalized Additive Models (GAM) in a non-metric multidimensional scaling (NMDS) ordination, testing specific hypothesized determinants of diversity. Plant community richness shows a bilaterally-symmetric modal distribution in the NMDS (p<0.001, quasi R²=0.36). We tested for methodological artifact resulting from a relationship between plot mean similarity to other plots and richness and found a significant but small effect (p<0.001, quasi R²=0.045); adjusted ordination results were p<0.001 and quasi R²=0.40. The results were inconsistent with the species pool hypothesis (sensu Taylor et al.), with a modal rather than increasing relationship between regional richness and local richness. The relations between richness and three environmental gradients, elevation, moisture, and degree days, were significant but small (p<0.001, quasi R²=0.014; p<0.001, quasi R²=0.021; and p<0.001, quasi R²=0.009 respectively). Incorporating an ecological classification for soil parent material as a categorical variable only increased the NMDS fit a small amount (p<0.001, quasi R²=0.42 vs 0.40). The results suggest plant community richness varies smoothly along a continuum, in agreement with Austin. Tested predictive variables, however, show relatively low explanatory power.

Keywords:

This abstract is being presented at: 11:45 AM in session:
Oral Session #59: Plant Communities: Vegetative Analysis.