Discerning the impact of human-mediated factors on biodiversity using bioclimatic envelope models and partial regression techniques

Abstract

[Aim]: Human-related pressures are growing in species-rich regions and pose a threat to the conservation of biodiversity. Here, we use the available data for five taxonomic groups (ferns, monocotyledons, dicotyledons, birds and monkeys) to exemplify a procedure directed to discriminate the degree of conflict between human actions and biodiversity. [Location]: Bioko island, Equatorial Guinea. [Methods]: Using bioclimatic envelope modelling techniques devoted to produce estimations of the potential distributions, we generated geographical representations of the variation in the total number of species as well as in the number of endemic and threatened species. We then employed partial regression techniques to determine how and to what extent current environmental, habitat and human-derived variables are associated with these potential species richness values. [Results]: Although the type of associations we looked for was sometimes difficult to discern since the same patterns could be explained by different types of variables, our results show that potential species richness values are generally positively associated with human-related factors (mainly agriculture and bushmeat hunting activities), suggesting that the localities with environmental conditions favourable to higher species richness tend to be those exploited by humans. [Main conclusions]: We propose that the combined use of distribution models and partial regression techniques can support a better understanding of the relationship between species occurrences/preferences and human-related factors and inform future conservation initiatives, particularly in small but hyperdiverse territories, in which dispersal limitations do not play a prominent role.