Masamu Institute for Research, Training, and Advanced Studies

Abstract

Data integration synthesizes information from multiple sources to yield a comprehensive understanding of complex systems. By combining diverse data streams, researchers can leverage the unique strengths of different sampling designs while mitigating the inherent limitations of individual surveys. In ecology, this approach is vital for monitoring biodiversity across scales by reconciling contrasting data, such as accurate low-resolution datasets with spatially extensive presence-absence information.

We propose a methodology for the integration of species abundance and occupancy data, providing a framework for bias-corrected estimates. Our results show this integrated model significantly improves estimates from independent data by reducing uncertainty. Crucially, our approach resolves persistent consistency issues in existing methods—most notably the integration of capture-recapture histories and occupancy-based designs. These findings enable better use of fragmented ecological surveys and inform new survey design approaches to leverage combined abundance and occupancy studies. On a broader level, this work facilitates a unified approach to ecological modelling, bridging the gap between distinct sampling protocols. Ultimately, providing more reliable population metrics ensures that conservation strategies are based on the most accurate evidence available.