Predicting habitat suitability of pollution-sensitive macroinvertebrates under climate change using climate and land use predictors in Ethiopia: an ensemble modeling approach

Background The distribution of pollution-sensitive macroinvertebrates, including Plecoptera (Perlidae) and Trichoptera (Philopotamidae), is jointly influenced by climate change and land use change. This study models habitat suitability for these taxa under current climate conditions and future climate scenarios, using current land use as a predictor variable. Species occurrence records were compiled from field surveys, the Global Biodiversity Information Facility, and literature sources. An ensemble species distribution model was developed using statistical models: Generalized Linear Models (GLM) and Generalized Additive Models (GAM) and machine learning algorithms: Boosted Regression Trees (BRT), Random Forest (RF) and Maximum Entropy (MaxEnt). Results The ensemble model showed strong predictive performance, with AUC 0.88 for both taxa and TSS values of 0.64 and 0.67 for Perlidae and Philopotamidae, respectively. The habitats of Perlidae were influenced by the mean temperature of the driest quarter and land use change, whereas Philopotamidae were shaped by temperature variables, particularly annual range. Current suitable habitats were estimated at 77,430.18 km2 for Perlidae and 92,448.14 km2 for Philopotamidae. Future projections indicate moderate habitat expansion for Perlidae in the 2050s (SSP2-4.5), but a sharp decline by 50.55% under the high-emission scenario SSP5-8.5 by the 2070s. Philopotamidae showed consistent declines, with up to 58.87% loss under SSP5-8.5 by the 2070s. Predicted maps highlight the Ethiopian highlands, especially the Dinki watershed, Oromia, Amhara, and southern regions, as core refugia, while Tigray, Somali, and Afar remain unsuitable. Conclusions These findings underscore the vulnerability of sensitive macroinvertebrates to climate and land use pressures and highlight the urgent need for conservation measures. These include riparian buffer restoration, stricter land use regulation, and climate-smart watershed management. In addition, ecosystem-based approaches that enhance hydrological stability, restore degraded catchments, and involve local communities are vital for sustaining freshwater biodiversity. Long-term biomonitoring using macroinvertebrates as bioindicators can help detect ecosystem decline and guide informed conservation actions.