Integrated community occupancy models: A framework to assess occurrence and biodiversity dynamics using multiple data sources

1. The occurrence and distributions of wildlife populations communities are shifting as a result global changes. To evaluate whether these shifts negatively impacting biodiversity processes, it is critical to monitor the status, trends, effects environmental variables on entire communities. However, modeling dynamics multiple species simultaneously can require large amounts diverse data, few approaches exist provide community level inferences. 2. We present an “integrated occupancy model” (ICOM) that unites principles data integration hierarchical in single framework inferences species-specific using sources. ICOM combines replicated nonreplicated detection-nondetection sources explicitly accounts for different detection sampling processes across use simulations compare previously developed models integrated distribution models. then apply our model assess foliage-gleaning birds White Mountain National Forest northeastern USA from 2010-2018 three independent 3. Simulations reveal integrating increased precision accuracy compared source models, although benefits were dependent information content individual (e.g., amount replication). Compared yielded more precise species-level estimates. Within case study, had highest out-of-sample predictive performance used only subset 4. provides estimates multi-species or single-species further found improved broad region interest with empirical study forest birds. offers attractive approach estimate dynamics, which additionally valuable inform management objectives both their broader