Spatiotemporal Dynamics of Fish Density in a Deep-Water Reservoir: Hydroacoustic Assessment of Aggregation Patterns and Key Drivers

Understanding spatiotemporal patterns of fish density and their environmental drivers is critical for managing river-lake ecosystems, yet dynamic interactions in heterogeneous habitats remain poorly quantified. This study combined hydroacoustic surveys, spatial autocorrelation analysis (Moran's I), and generalized additive models (GAMs) to investigate seasonal and spatial fish distribution, aggregation characteristics, and regulatory mechanisms in China's Zhelin Reservoir. The results reveal pronounced seasonal fluctuations, with summer fish density peaking at 13.70 ± 0.91 ind./1000 m³ and declining to 1.95 ± 0.13 ind./1000 m³ in winter. Spatial heterogeneity was evident, with the Xiuhe region sustaining the highest density (15.69 ± 1.09 ind./1000 m³) and persistent hotspots in upstream bays. Transient high-density clusters (90-99% confidence) near the Zhelin Dam during summer suggested thermal or hydrodynamic disturbances. GAM analysis (R²adj = 0.712, 78.5% deviance explained) identified seasonal transitions (12.26% variance), water depth (16.54%), conductivity (13.75%), and dissolved oxygen (13.29%) as dominant drivers, with nonlinear responses to depth and bimodal patterns for conductivity/oxygen. These findings demonstrate that hydrological seasonality and habitat heterogeneity jointly govern fish aggregation, underscoring the ecological priority of Xiuhe and upstream bays as core habitats. This study provides a mechanistic framework for guiding reservoir management, including targeted conservation, dam operation adjustments to mitigate hydrodynamic impacts, and integrated strategies for balancing hydrological and ecological needs in similar ecosystems.