Towards functional biodiversity predictions: a hierarchical modelling framework from primary productivity to biomass of upper trophic levels

Full abstract

Context Current shifts in biodiversity are driven by multiple processes of environmental and landscape change. Particularly, land use/land cover (LU/LC) dynamics are among the major drivers of biodiversity loss worldwide. Objectives In this study we aim to explore the applicability of a new modelling framework to predict top predators’ responses to LU/LC changes. Methods The framework integrates remote-sensing based predictors, statistical inference, stochastic-dynamic simulations and spatial projections in a common and interactive approach. From an ecological modelling perspective, the main innovation of our approach lies on the integration of (1) biomass of birds of prey as an upper trophic indicator of the community characteristics that emerge from the habitat quality across multiple scales of organization and (2) fine-scale biophysical attributes to add a new level of understanding about the role of local LU/LC drivers influencing those emergent biodiversity patterns. Results Based on species data from published atlases this approach allowed transposing species biomass to finer resolutions, overcoming the lack of detailed information for the study area. Our demonstrative case study revealed a disruptive effect of ongoing LU/LC changes in the spatio-temporal distribution of top predators’ biomass, suggesting the possibility of an emergent disturbance pattern in habitat suitability and community stability. Comparative analysis between simulations and independent field data revealed a promising model performance. Conclusions Our modelling approach highlights the importance of integrating local LU/LC functional dynamics to predict key trophic’ responses, considered as pertinent ecological indicators for biodiversity management under realistic’ future changing regional scenarios.