Temporal dynamics override spatial gradients in Afrotropical reservoir water quality

Freshwater systems in tropical Africa face growing pressures from dams, climate change, and land-use expansion, yet we still do not fully understand which factors most significantly affect water quality. Knowing this is crucial for designing effective monitoring and management, especially when resources are limited. We conducted monthly water-quality surveys at six sites across the Kainji–Jebba reservoir cascade in Nigeria from April 2024 to March 2025, encompassing two dams and three distinct habitat types (riverine, ecotonal, and lacustrine). Using a modified NSF Water Quality Index (WQI), variance partitioning (RDA), Granger causality, and six changepoint detection methods, we untangled the effects of space, time, land use, and climate, and identified early-warning indicators of water quality change. Our findings show that when changes happen matters more than where. Seasonal and monthly variations explained 28.3% of water quality variance, while spatial differences accounted for only 1.8%, and direct land-use effects were minimal (0.4%). Agricultural impacts were expressed primarily through short, wet-season runoff pulses that affected all sites simultaneously, rather than through persistent spatial differences. After false discovery rate correction, no relationships between climate variables and water quality remained significant. WQI, electrical conductivity (EC), and relative humidity (RH) emerged as reliable early-warning sentinel indicators, detecting regime shifts in 71.4% of cases. These findings suggest that monitoring programmes in tropical reservoir systems should prioritise targeted seasonal windows: the wet-season runoff peak (June–July) and the arrival of sediment-rich black floods (January–February), rather than expanding spatial coverage.