Phantom refugia for fruit-eating migratory fishes in dammed Neotropical rivers

Introduction: Climate change, river impoundment, and riparian degradation act synergistically to threaten Neotropical freshwater biodiversity, yet conservation approaches rarely integrate these stressors simultaneously. We developed a triple mismatch framework combining species distribution models, dam fragmentation analysis, and vegetation condition assessment to identify priority conservation areas for two fruit-eating fish species in the Parana–Paraguay Basin. Materials and methods: We modelled present and future (2041–2080; Shared Socioeconomic Pathway (SSP) SSP2-4.5 and SSP5-8.5) habitat suitability for Piaractus mesopotamicus and Myloplus tiete using True Skill Statistic (TSS)-weighted ensemble species distribution models across 16,538 river-associated grid cells (10 km resolution, restricted to cells intersecting river segments of Strahler order ≥ 3). Occurrence records were compiled from databases and the scientific literature. We integrated an ensemble of six dam databases, filtering only structures constituting actual river barrages within the basin, resulting in 3582 dams. Cells were classified into four fragmentation categories, and vegetation was classified into three ecologically grounded classes. We then applied a nine-class mismatch diagnostic combining climate refugia status, vegetation condition, and dam fragmentation to derive nine priority conservation actions. Results: Projected habitat loss ranged from 24.0% to 39.6% for P. mesopotamicus and from 32.7% to 50.8% for M. tiete. Of the 2265 climate refugia identified for P. mesopotamicus, only 14.3% were classified as functional refugia (free-flowing with intact vegetation ≥ 80%), whereas compound severe mismatch and impounded degraded refugia accounted for 41.1% and 13.9%, respectively. For M. tiete (1242 refugia), no functional refugia were identified, with compound severe (57.4%) and impounded degraded (23.3%) dominating. All restoration categories combined accounted for 63.0% of P. mesopotamicus and 72.0% of M. tiete refugia. Restoration-demanding refugia represented most conservation priorities for both species, while impounded refugia accounted for 16.3% of P. mesopotamicus and 26.5% of M. tiete refugia. Conclusions: Our triple mismatch framework reveals that identifying climate refugia alone is insufficient for conservation planning when dams and riparian degradation create structural barriers to species persistence. Dam impoundment constitutes an effectively irreversible constraint for migratory fishes, while riparian degradation remains reversible through restoration.