Relating Mammal Species Richness to Landscape Patterns Across Multiple Spatial Scales

Understanding the relationships between spatial pattern, spatial scale and biodiversity can help ecologists to assess the impacts of environmental change and inform management plans. Spatial pattern research has often focussed on the effect of modified landscapes on species diversity. However, few studies have examined species responses to spatial pattern from other sources, including those which vary over time, such as fire. We investigated the effect of composition and configuration for topographic, ecological and disturbance factors on mammal species richness. In addition, we examined whether the magnitude and predictive strength of the relationship with richness varied with spatial scale. We sampled ground-dwelling mammals at 187 sites in the Otway Ranges of south-eastern Australia. A gradient modelling approach was used to characterise landscape composition and configuration for each predictor. Relationships with mammal richness were modelled using Bayesian Networks at ten different spatial scales (7–1165 ha). Composition and configuration were both important to species richness, although the strength and presence of relationships varied across the ten scales. Patterns in NDVI, time since fire, habitat complexity and elevation had the strongest effects on mammal species richness. Our findings highlight the importance of measuring both the composition and configuration of environmental measures at different spatial scales to assess their effect on species richness. Further, studies focusing on just one environmental measure of spatial pattern or one spatial scale will miss important relationships between environmental variables and species richness.