Plant species composition in a temperate forest: Multi-scale patterns and determinants

We examine the spatial patterns of plant species composition at different scales in a hierarchical sampling design with two surveys of contrasting scale. Additionally, environmental and spatial variables are used to explain the observed patterns. atasets were analyzed in this study. The first was obtained as a result of a large-scale spatial survey, in which the study site (132 ha) was divided into 102 large plots of 20 × 20 m. The remaining four datasets were obtained from a small-scale spatial survey, in which four of the former plots were also divided into 100 small plots of 2 × 2 m. Spatial patterns of plant species composition in both spatial surveys were quantified and the factors that influenced them were assessed using multi-scale pattern analysis (MSPA). he large-scale survey the topographic structure of the study site created a spatially structured environment, influencing species composition, and the spatial variables indicated that the environment was structured at a broad scale (relative to grain size and extent of the survey). In the small-scale survey the microenvironmental variables that influenced species composition were also spatially structured at a broad scale (relative to grain size and extent of the survey). However, the analyses point to the existence of spatial autocorrelation that seems to be structured at finer scales than the environmental heterogeneity in both study surveys. tudy indicates that species composition in this temperate forest is not only determined by the environmental variables studied at either of the two spatial scales considered (large- and small-scale surveys). In both scales, the pure spatial component present in the analyses may be indicating the influence of unmeasured environmental variables and/or biotic processes on species composition patterns. However, while environmental heterogeneity has a broad-scale domain, biotic processes seem to work at finer scales, as is indicated by the spatial predictors.