Explainable Artificial Intelligence Reveals Spatially Divergent Effects of Global Change on Mammals
Preprint
- 17 September 2025
- preprint
- Published by Springer Nature in Research Square
Abstract
Understanding how global change reshapes species distributions remains a central challenge in ecology, particularly as environmental drivers exert uneven effects across space and time. Most assessments of the hypothesized impacts of climate change and conservation of future biodiversity emphasize climate-induced risks but overlook where environmental changes may also relax constraints and improve habitat suitability. Here, we apply explainable machine learning to assess how climate averages, variability, extremes, and land cover are projected to reshape future distributions of 1,992 terrestrial mammals worldwide. Leveraging Shapley Additive Explanations (SHAP) applied to species distribution models (SDMs), we quantify the directional contribution of 16 environmental drivers and track how these contributions change over space and time. Our work enables strong tests of past hypotheses and shows that climate extremes produce more localized but intense effects than means or variability; temperature-related drivers dominate, with the strongest and most uncertain impacts for endangered species; and individual drivers can simultaneously increase or reduce suitability across regions. These findings reveal that ecological risks and gains are spatially divergent, highlighting the need for driver-specific, regionally tailored conservation strategies under global change.Keywords
All Related Versions
This publication has 53 references indexed in Scilit:
- Post-hoc vs ante-hoc explanations: xAI design guidelines for data scientistsCognitive Systems Research, 2024
- Assessing species vulnerability to climate change, and implementing practical solutionsBiological Conservation, 2023
- Explainable artificial intelligence reveals environmental constraints in seagrass distributionEcological Indicators, 2022
- The role of land use and land cover change in climate change vulnerability assessments of biodiversity: a systematic reviewLandscape Ecology, 2021
- A comprehensive evaluation of predictive performance of 33 species distribution models at species and community levelsEcological Monographs, 2019
- Climate change modifies risk of global biodiversity loss due to land-cover changeBiological Conservation, 2015
- Spatial bias in the GBIF database and its effect on modeling species' geographic distributionsEcological Informatics, 2014
- Explaining prediction models and individual predictions with feature contributionsKnowledge and Information Systems, 2013
- Evaluating effectiveness of down-sampling for stratified designs and unbalanced prevalence in Random Forest models of tree species distributions in NevadaEcological Modelling, 2012
- Animals as indicators of ecosystem responses to air emissionsEnvironmental Management, 1984