Latitudinal gradients in taxonomic overdescription rate affect macroecological inferences using species list data
Ecography, 35:4, 333–340 (2012)
Species lists for particular geographic areas are often used in macroecology and conservation; for example, they have been used to identify hotspots of biological diversity, and for the study of latitudinal species diversity gradients. However, there is uncertainty over the accuracy of species lists due to undiscovered species and synonymy of described species. This paper concentrates on taxonomic overdescription caused by the latter. Where bias in the amount of taxonomic overdescription occurs along a variable of interest (e.g. latitude, or body size), inferences from macroecological studies may be affected. This form of taxonomic overdescription is likely to be most pronounced in speciose groups, where many species have relatively small geographic range sizes and have low numerical abundance. A good example of such a group is the wasp family Ichneumonidae. We first use taxonomic and region-specific species list data for the Ichneumonidae to estimate the probability of species validity. Then we use this estimated probability to statistically correct the region’s species richness estimate using a Monte Carlo simulation approach, and examine the effect the correction has on three major macroecological patterns: the relative species richness of geographic regions, latitudinal species richness pattern, and body size. Our results indicate that although there is significant geographic variation in overdescription, the bias is not sufficient to qualitatively alter broad-scale macroecological conclusions such as hotspot identity, and the qualitative global patterns of diversity and mean body size.