Can dynamic occupancy models improve predictions of species' range dynamics? A test using Swiss birds

Predictions of species' current and future ranges are needed to effectively manage species under environmental change. Species typically estimated using correlative distribution models (SDMs), which have been criticized for their static nature. In contrast, dynamic occupancy (DOMs) explicitily describe temporal changes in species’ via colonization local extinction probabilities, from time series occurrence data. Yet, tests whether these improve predictive accuracy or conditions rare. Using a long-term data set on 69 Swiss birds, we tested DOMs the predictions over compared SDMs. We evaluated spatial ability detect population trends. also explored how differed when accounted imperfect detection parameterized calibration sets different lengths. All model types had high performance assessed across all sites (mean AUC > 0.8), with flexible machine learning SDM algorithms outperforming parametric DOMs. However, none performed well at identifying where range likely occur. terms estimating trends, best, particularly strong fit sufficient data, while SDMs very poorly. Overall, our study highlights importance considering what aspects matter most selecting modelling method particular application need further research utility. While show promise capturing dynamics inferring trends fitted computational constraints variable selection fitting can lead reduced predictions, an area warranting more attention.