Scale-dependence of the correlation between human population and the species richness of stream macro-invertebrates

Recent biogeographical studies have shown positive correlations between plant/vertebrate species richness and human population presence. The same pattern has been reported for Ephemeroptera (mayflies), Plecoptera (stoneflies) and Trichoptera (caddisflies) (EPT) amongst European countries. This is surprising as EPT are bio-indicators of stream pollution and most local studies report higher species richness of these macro-invertebrates where human influences on water quality are lower. Using a newly collated taxonomic dataset, we studied whether the species richness of EPT is related to human population size at finer resolutions (Italy's regions, provinces and 10×10 km2 UTM cells) controlling for sampling effort, variations in area and for spatial autocorrelation. At all study grains, observed EPT species richness was strongly correlated to the number of records available for the same taxon. At the regional level, the observed number of Ephemeroptera and Plecoptera species significantly increased with increasing human population size. At the provincial level, observed species richness decreased significantly with increasing human population size for Ephemeroptera and did not vary significantly for Plecoptera and Trichoptera. At the finest grain scale, there were significant negative correlations of observed Ephemeroptera and Trichoptera species richness with human population size, although the proportion of variance explained was very low. These results were broadly confirmed when analyzing the estimated number of species using the formula of Chao2. Our analysis confirms the scale-dependence of the human population–biodiversity correlation. Over broad scales more populated regions tend to have more species than less populated ones. Restricting the study grain, the positive EPT species–people relationship disappears and turns into a negative one. Our findings suggest a challenge also for the conservation of regional EPT diversity.