An environmentally defined biogeographical zonation of Scotland designed to reflect species distributions

1 We present a method, worked out in detail for Scotland, that classifies the environment in a way that maximizes information about species distributions. Biogeographical zones defined by the method are based on permanent features of the environment, while at the same time reflecting species distributions and vegetation cover. 2 Data on the occurrence of native species in six taxonomic groups (breeding birds, diurnal insects, terrestrial molluscs, vascular plants, mosses and liverworts) in 10-km squares in Scotland were extracted from national databases. The environment of each square was specified by 12 climatic variables and four topographic variables. 3 A measure of the environmental difference between 10-km squares was derived by relating their environmental attributes to the species that occur in them, using detrended canonical correspondence analysis (DCCA), a method of ordination. DCCA defines a metric in environmental space such that unit difference corresponds to a given degree of species difference. 4 For each of the six taxonomic groups, DCCA was used to generate a four-axis ordination of the 10-km squares, which were then clustered into 10 groups by a minimum-variance clustering algorithm. Ten clusters were also derived using the full set of 24 axes resulting from all six ordinations. Characteristic species for each of the clusters were identified using a numerical preference index. 5 Most clusters could be mapped as discrete zones. There was substantial similarity between the clusters for differing taxonomic groups. The 10 clusters created by combining all six groups provided a succinct general summary of the natural zonation of Scotland. The characteristic species of these clusters were used to contrast the high conservation value of some areas with the lesser value of others. 6 Our biogeographical zones compared well with certain previous classifications, but have the additional strength of being based on a combination of environmental data and species data at a national scale. There is considerable scope for developing the methodology and for extending its application for conservation purposes.