Macroecological correlates and spatial patterns of anuran description dates in the Brazilian Cerrado

Aim To quantify the relationship between the description dates of anuran species in the Brazilian Cerrado and some macroecological traits, and to verify the spatial patterns of average description dates and their correlation with human occupation and biodiversity knowledge. Location Brazilian Cerrado (South America). Methods The average date of description of 131 species of anurans found in 181 cells overlaying the Brazilian Cerrado was recorded. Description date was regressed across species on body size and geographical range size. Phylogenetic effects that could bias the significance tests of the multiple regression model of description dates on macroecological traits were taken into account using a phylogenetic subtraction method in which families and genera were classificatory factors in a nested two-way analysis of variance (ANOVA) model. We also conducted a spatial analysis of the average description date that was estimated for each cell. This cell-based metric was regressed on human population size, the year of foundation of the municipalities and the number of inventories undertaken in each cell. The influence of spatial autocorrelation patterns was taken into account by using the geographically effective number of degrees of freedom. Results The number of new species being discovered in the Brazilian Cerrado has been increasing, especially over the last 50 years. Cross-species analyses indicated that description dates were negatively correlated with body size and geographical range size, taking phylogenetic effects into account. Even after controlling for the spatial structures in all variables, average description date was positively correlated with human population in geographical space, but because of multicolinearity structure in the data, it was not possible to quantify the independent influence of human population and number of inventories on description date. Conclusions As found in previous papers, large-bodied and widely distributed species are likely to be described first. Species yet to be discovered are probably small-bodied and with narrow distributions, more restricted to the Cerrado biome. Also, the explicit spatial approach showed that the average description date is spatially correlated with total human population and biodiversity knowledge in the Cerrado region. Our findings suggest that incorporating human population density into the reserve design algorithms, which has usually been done to avoid or minimize conservation conflicts, may also produce good results because this will preserve many places where most of the non-described species will probably be found in the future.