Complex effects of scale on the relationships of landscape pattern versus avian species richness and community structure in a woodland savanna mosaic


A. Bar-Massada, Dept of Forest and Wildlife Ecology, Univ. of Wisconsin – Madison, 1630 Linden Drive, Madison, WI 53706, USA. E-mail:


Landscape pattern metrics are widely used for predicting habitat and species diversity. However, the relationship between landscape pattern and species diversity is typically measured at a single spatial scale, even though both landscape pattern, and species occurrence and community composition are scale-dependent. While the effects of scale on landscape pattern are well documented, the effects of scale on the relationships between spatial pattern and species richness and composition are not well known. Here, our main goal was to quantify the effects of cartographic scale (spatial resolution and extent) on the relationships between spatial pattern and avian richness and community structure in a mosaic of grassland, woodland, and savanna in central Wisconsin. Our secondary goal was to evaluate the effectiveness of a newly developed tool for spatial pattern analysis, multiscale contextual spatial pattern analysis (MCSPA), compared to existing landscape metrics. Landscape metrics and avian species richness had quadratic, exponential, or logarithmic relationships, and these patterns were generally consistent across two spatial resolutions and six spatial extents. However, the magnitude of the relationships was affected by both resolution and extent. At the finer resolution (10-m), edge density was consistently the best predictor of species richness, followed by an MCSPA metric that measures the standard deviation of woody cover across extents. At the coarser resolution (30-m), NDVI was the best predictor of species richness by far, regardless of spatial extent. Another MCSPA metric that denotes the average woody cover across extents, together with percent of woody cover, were always the best predictors of variation in avian community structure. Spatial resolution and extent had varying effects on the relationships between spatial pattern and avian community structure. We therefore conclude that cartographic scale not only affects measures of landscape pattern per se, but also the relationships among spatial pattern, species richness, and community structure, often in complex ways, which reduces the efficacy of landscape metrics for predicting the richness and diversity of organisms.