Measuring range sizes of South-East Asian hawkmoths (Lepidoptera: Sphingidae):: effects of scale, resolution and phylogeny

Aim The effects of resolution and spatial extent on range measures were explored in estimates of the geographic distribution of tropical hawkmoths. Furthermore, data were tested for phylogenetic autocorrelation. Location South-East Asia. Methods Various range measures, such as geographic information system (GIS)-supported range estimates, minimum convex polygons, latitudinal and longitudinal extents, and their products, were derived from original distribution records and compared to each other. A taxonomic classification of the species was used to analyse phylogenetic effects on range sizes. Results Range size measures exhibit a strongly right-skewed frequency distribution with many geographically restricted species and few widespread taxa. Rankings from GIS-supported, comprehensive range size estimates do not deviate greatly from more crude measurements of lower resolution. Comprehensive ranges and ranges within South-East Asia are correlated strongly, but already at this rather large scale the ranking of species changes considerably. Other measures of occupancy with an increasingly more localized consideration of 'range' show decreasing strengths of correlation. We found a weak, but significant, autocorrelation in range area data: related groups of species have ranges of similar size. Main conclusions Spatial resolution did not affect range ranking greatly in our data. However, macroecological studies based only on parts of species' ranges must be viewed critically, particularly if their extent is small compared to comprehensive ranges. Phylogenetic non-independence of range size data must be considered in comparative analyses.