Loss or redistribution? A better way of estimating regional changes in animal distribution and numbers caused by increased human activities

A differentiated understanding of how regional human activities affect the spatial distribution and numbers of animals within specific areas of interest is of great ecological importance. Estimating these effects from empirical data is challenging however, because human activities can affect animals in qualitatively different ways and on different spatial and temporal scales. In addition, spatio-temporal animal abundance is frequently influenced by factors intrinsic and extrinsic to the area of interest, potentially confounding impact studies, e.g., based on trends. In this study, we synergistically combined regression and mechanistic modelling to separate these different influences. We first used partial differential equations to simulate various potential animal redistribution patterns affected by regional human activities. We then selected appropriate patterns as predictors in regression-based species distribution models, together with additional anthropogenic and natural covariates. The simultaneous consideration of large-scale (number-conserving) animal reorganisation, their regional loss or gain, and the influence of additional environmental covariates eventually allowed the generation of qualitative and quantitative estimates and predictions of human-induced changes. We exemplarily applied our approach to investigate the current and future impact of increasing offshore wind farm (OWF) implementation in the German North Sea on common murres (Uria aalge) during autumn. OWFs constructed up to 2019 reduced common murre numbers in German waters by 18.3%. If the planned OWF priority and reservation areas outlined in the German Marine Spatial Plan are implemented, the predicted loss would increase to 77.7%. Notably, these predictions did not include additional anthropogenic activities or further plans for OWF installation, which could together lead to the almost complete disappearance of common murres from the German North Sea. By directly comparing predicted animal numbers and distributions in hypothetical scenarios with and without human pressures, the presented method allows us to measure and predict the effects of human activities on regional trends and large-scale reorganisation. This in turn helps us to quantify and predict the impact of planned human activities on wildlife, including in the context of the current rapid expansion of alternative energies.