Habitat Structure, Patchiness, and Avian Communities in North American Steppe Vegetation: A Multivariate Analysis

We investigate the relationships between vegetational structure, spatial heterogeneity, avian community structure, and the ecological responses of the breeding bird populations in the structurally simple steppe vegetation of North America. We examined the effect of variation in several measures of both horizontal and vertical patchiness in vegetation distribution on bird distribution and abundance, then integrated these measures into a smaller subset of variables using Principal Component Analysis (PCA). We interpreted the new PCA variables as reflective of proximate niche parameters to which bird species responded. Out of 550 structure/bird abundance bivariate correlations, 18.5% achieved a significant level of at least P <.05. Virtually all individual species responses to vegetation habitat measurements were straightforward: all “typical tallgrass prairie” birds attained highest abundances on plots with highest percent coverage of grass, “typical shrubsteppe” birds attained their highest abundances on plots with the highest values of shrub coverage, and so on. In the PCA of vegetational features, the first axis (which accounted for 41.0% of the total variation) represented in horizontal structure and heterogeneity of vegetation, with high positive factor loadings for 9 of 10 horizontal measures. The second component (22.4%) represented increasing vertical heterogeneity, with significant correlations for five of five vertical measures. The third component(10.7%) reflected variation in the density and coverage of low forbs. The segregation of horizontal and vertical measures on orthogonal axes indicates that variation in horizontal vegetation patchiness is largely independent of variation in vertical patchiness in these steppe systems. The abundances of all typically tallgrass prairie bird species were negatively correlated with increasing horizontal heterogeneity and partially correlated with increasing vertical heterogeneity. Several shrubsteppe bird species were also positively associated with increasing vertical heterogeneity, but unlike tallgrass birds their abundances were also positively associated with increasing horizontal patchiness. Shortgrass prairie birds were generally negatively correlated with vertical patchiness. Only bird species associated with montane meadows varied significantly with changes in the third principal component, forb abundance. At the community level, species diversity increased as vertical heterogeneity increased but appeared to vary independently of horizontal patchiness. Temporal variation in overall vegetation structure was not closely mirrored by correlated changes in bird community composition at the shortgrass prairie site for which yearly data were available. Analysis of species' niche overlaps along the synthetic vegetation axes defined by PCA revealed a significant lack of niche complementarity; species with high overlap on one axis were quite likely (P < .001) to have high overlap on a second, independent axis. Analysis of nich metrics across communities indicated that while average niche breadths of co—occurring species were independent of a sample's position on any multidimensional axis, average nich overlap increased at the nonmontane grassland sites. Tests of current theories of "diffuse competition" vs. “niche breadth compensation for low alpha;diversity” were inconclusive. We suggest that the conventional sorts of data collected to test the hypotheses, are inappropriate. Our results suggest that individual bird species responded independently to the environmental gradients we derived, and that higher species interactions were not necessary (and in some cases statistically improbable) to explain the patterns we observed. These results are consistent with our contention that even though such biotic interactions may be taking place, they are unlikely to be detected in our systems because (1) environmental variation obscures relationships between local populations, (2) populations, are not resource limited or equilibrial during large portions of their history, and (3) individuals are not "fine tuning" their responses to local habitat variation.