The case for research integration, from genomics to remote sensing, to understand biodiversity change and functional dynamics in the world's lakes.

Freshwater ecosystems are heavily impacted by multiple stressors, and a freshwater biodiversity crisis is underway. This realisation has prompted calls to integrate global freshwater ecosystem data, including traditional taxonomic and newer types of data (e.g. eDNA, remote sensing), to more comprehensively assess change among systems, regions, and organism groups. We argue that data integration should be done, not only with the important purpose of filling gaps in spatial, temporal and organismal representation, but also with a more ambitious goal: to study fundamental cross-scale biological phenomena. Such knowledge is critical for discerning and projecting ecosystem functional dynamics, a realm of study where generalisations may be more tractable than those relying on taxonomic specificity. Integration could take us beyond cataloging biodiversity losses, and toward predicting ecosystem change more broadly. Fundamental biology questions should be central to integrative, interdisciplinary research on causal ecological mechanisms, combining traditional measures and more novel methods at the leading edge of the biological sciences. We propose a conceptual framework supporting this vision, identifying key questions and uncertainties associated with realising this research potential. Our framework includes 5 interdisciplinary "complementarities". First, research approaches may provide comparative complementarity when they offer separate realisations of the same focal phenomenon. Second, for translational complementarity, data from one research approach is used to translate that from another, facilitating new inferences. Thirdly, causal complementarity arises when combining approaches allows us to "fill in" cause-effect relationships. Fourth, contextual complementarity is realised when together research methodologies establish the wider ecological and spatio-temporal context within which focal biological responses occur. Finally, integration may allow us to cross inferential scales through scaling complementarity. Explicitly identifying the modes and purposes of integrating research approaches, and reaching across disciplines to establish appropriate collaboration will allow researchers to address major biological questions that are more than the sum of the parts. This article is protected by copyright. All rights reserved.