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

Abstract

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.

Keywords: Zhelin Reservoir; acoustics monitoring; environmental gradients; fish distribution; generalized additive models; schooling behavior.

Figure 1

Location of the Zhelin Reservoir and sampling sites’ distribution among different regions. Inset map indicates the location of the Zhelin Reservoir in the Yangtze River Basin, China. The black line at the beginning of the image represents the Xiafang hydropower station on the Xiuhe River. The red dotted line indicates the boundary of different reservoir regions.

Figure 2

Spatiotemporal variations in fish density (ind./1000 m³) in Zhelin Reservoir. Panel (a) and (b) represent the seasonal and regional variations, respectively. Bars represent mean values, and error bars indicate standard error (SE). Bars with the same letter are not significantly different, while different letters indicate significant differences (p < 0.05) based on Dunn’s post hoc tests following Kruskal–Wallis tests.

Figure 3

Spatiotemporal heterogeneity of fish density composition classified by target strength (TS, dB). Gradient colors represent TS groups, d1: (−70, −60), d2: [−60, −55), d3: [−55, −50), d4: [−50, −45), d5: [−45, −40), d6: [−40, −35), and d7: [−35, −30]. Sector areas represent normalized percentages of each density group.

Figure 4

Spatiotemporal distribution of the hot- and cold spots of fish density in Zhelin Reservoir.

Figure 5

Generalized additive model (GAM) partial effect and explained deviance of selected explanatory variables on the exponential of fish density. Panel (a–f) represent season (a), WD (b), DO (c), Cond (d), PO₄ (e), and Z_Bi (f) on fish density, respectively. Data points represent partial residuals. Tick marks on the x-axis indicate observed data points. Shaded areas represent 95% confidence intervals.