Addressing data integration challenges to link ecological processes across scales

Abstract

Data integration is a statistical modeling approach that incorporates multiple data sources within a unified analytical framework. Macrosystems ecology-the study of ecological phenomena at broad scales, including interactions across scales-increasingly employs data integration techniques to expand the spatiotemporal scope of research and inferences, increase the precision of parameter estimates, and account for multiple sources of uncertainty in estimates of multiscale processes. We highlight four common analytical challenges to data integration in macrosystems ecology research: data scale mismatches, unbalanced data, sampling biases, and model development and assessment. We explain each problem, discuss current approaches to address the issue, and describe potential areas of research to overcome these hurdles. Use of data integration techniques has increased rapidly in recent years, and given the inferential value of such approaches, we expect continued development and wider application across ecological disciplines, especially in macrosystems ecology.

Working across space and time: nonstationarity in ecological research and application

“…The use of specific management practices (eg fire) is known to have mixed results among locations due to spatial variability in site and environmental conditions (McEwan et al 2011). The rise of macrosystems biology via "big data", new technologies in fields like remote sensing, and advances in data integration have led to novel discoveries in global-scale and emergent phenomena that are not possible from smaller scale studies based on a few sites or short temporal extent alone (Zipkin et al 2021). Translating such advances into reliable and actionable information for practitioners and policy makers requires that spatial and temporal variability in ecological patterns and processes be accounted for in both inference and prediction (Rodo et al 2002;Gouveia et al 2013).…”

Linking Terrestrial and Aquatic Biodiversity to Ecosystem Function Across Scales, Trophic Levels, and Realms

“…Recent developments in ecology and environmental science can help address these three constraints. First, there is increased data availability of relatively fine-grain (tens of meters) measures of ecosystem and biotic properties across broad spatial extents (Zipkin et al, 2021), often obtained from remote sensing platforms in both terrestrial and aquatic realms (Jetz et al, 2016;Yang et al, 2020). Second, networks of ecologists working in different realms and in different geographic areas provide opportunities for cross-realm study.…”

Macrosystems revisited: challenges and successes in a new subdiscipline of ecology

“…Statistical approaches that are insensitive to nonstationarity are important to macrosystems biologists as well as to the broader community of ecologists faced with similar statistical problems, and as such, Rollinson et al (2021) summarize challenges and solutions relevant to all ecologists. Similarly, Zipkin et al (2021) confront the issue of combining data collected at different scales. Large spatial scale research that seeks to combine disparate datasets is especially subject to such challenges.…”

“…Similarly, Zipkin et al . (2021) confront the issue of combining data collected at different scales. Large spatial scale research that seeks to combine disparate datasets is especially subject to such challenges.…”