A test of the scale-dependence of the species abundance-people correlation for veteran trees in Italy

Background and Aims The spatial correlation of the presence of people and species has been suggested to be scale-dependent. At local scales, large numbers of people often result in species impoverishment. At coarse scales, species-rich regions tend to be densely inhabited. Recently, broad-scale human presence has been shown to be correlated not only with numbers of species but also with their abundance, as predicted by the more-individuals hypothesis. However, it is not known whether the species abundance-human presence correlation could also be scale-dependent. Methods This hypothesis was tested by use of a database of veteran trees in Italy. Veteran tree species richness and number of individuals were modelled as a function of human population size at two grains of analysis (provinces and regions), controlling for variations in area, latitude and spatial autocorrelation. Key Results A positive correlation was found between human presence and veteran tree species. As predicted, this correlation was stronger at a coarser resolution. However, only at the provincial but not regional level was there a positive correlation between human presence and veteran tree abundance when controlling for area and latitude. These results were confirmed for native and exotic trees. Conclusions The present findings rule out the more-individuals hypothesis as an explanation of the scale-dependence of the species-people correlation for veteran trees in Italy. Potential mechanisms behind the observed spatial coincidence of high numbers of people and veteran tree species are discussed and implications for conservation are highlighted.