Reciprocal insights from global aquatic stressor maps and local reporting across the Ramsar wetland network

Abstract Aquatic ecosystemsare exposed to a host of anthropogenic stressorswhose combined effect can be synthesized with cumulative stress indices. The reliability of cumulative stress indices depends primarily on: 1) stressorincidence maps derived from remote sensing or modeling but rarely validated against on-the-ground observations, and 2) the weighting scheme used to combine multiple stressorsinto a cumulative index typically based on expert opinion. In this paper, we evaluate the exposure and weights for 13 aquatic stressormaps of the world’s rivers with a comparison against local stress reporting across 1018 inland and coastal sites from the Ramsar Wetlands of International Importance. We found that globally-mappedand locally-reported stressorsare poorly aligned overall (AUC-ROC = 0.50–0.63), and that concordance did not improve when stratifying by ecosystem types or continents. Agreement varied across individual stressors, was highest for hydrological stressorsand lowest for habitat disconnectivity stressors. We estimated stressorweights from the comparison and found them to be remarkably aligned well with expert-generated weights, suggesting there is convergence on a stressorhierarchy across local and global scales. Our comparison illustrates the value of integrating data across scales to inform the calculation of global stressorindices. Continued systematic stressormonitoring across environmental observation networks is central to benchmarking maps of ecosystem stress globally.