Fine-scale spatial organization of tallgrass prairie vegetation along a topographic gradient

Fine-scale spatial patterns of native tallgrass prairie vegetation were studied on Konza Prairie, Kansas, USA. Three sites, upland, slope, and lowland, were sampled in an ungrazed watershed. Presence of vascular plant species was recorded in two 25.6 m long transects of contiguous 5×5 cm micro-quadrats on each topographic position. Spatial patterns of species and functional groups were analyzed by information theory models ofJuhász-Nagy. Within-community variability of coexistence was expressed by the diversity and spatial dependence of local species combinations. Considerable diversity in the local coexistence of species was found on each site. Upland and hillside communities were richer and more diverse in species combinations than lowland. Spatial scale effected coexistence relationships. The maxima of information theory estimates varied between 15 and 30 cm. There was no trend in the variation of characteristic scales along the topographical gradient. Above 10 m, all sites tended to be homogeneous. The analysis of spatial associations revealed that variability in the local coexistence of species was strongly constrained in all topographic positions. Overall spatial association of species was the lowest on lowland. The characteristic scales of maximum association were between 1.2 m and 3 m at all sites. The maxima of information theory estimates for the functional group-based data appeared at smaller plot sizes than for the species based analyses. Only weak spatial associations were detected among the functional groups indicating that individuals of functional groups coexist well at small scales, and form combinations close to random expectations. The length of transects did not effect the relative associations. Strong positive correlations were found between the number of components (species or functional groups) and the maxima, of information theory models suggesting that richness is a good predictor of within-community coexistence relations. However, there was no relationship between richness and the characteristic scales of community patterns.