A case study of factors controlling water quality in two warm monomictic tropical reservoirs located in contrasting agricultural watersheds

Abstract

The integration of internal (e.g., stratification) and external (e.g., pollution) factors on a comprehensive assessment of reservoir water quality determines the success of ecosystem restoration initiatives and aids watershed management. However, integrated analyses are scarcer than studies addressing factors separately. Integration is likely more efficient in studies of small well-characterized (experimental) reservoir watersheds, because the isolation of factor contributions is presumably clearer. But those studies are uncommon. This work describes the water quality of two small 5.5 m-deep reservoirs (MD-Main and VD-Voçoroca dams) located in Pindorama Experimental Center, state of São Paulo, Brazil, considering the interplay between reservoir dimension, seasonal thermal stratification, chemical gradients, erosive rainfall events, presence of natural biofilters, and land uses and landscape patterns around the reservoirs and within the contributing watersheds. The monitoring of agricultural activities and water quality parameters occurred in October 2018–July 2019. A 4 °C thermal stratification occurred in October (difference between surface and bottom water temperature), which decreased until disappearance in January (VD) or April (MD). The longer stratification period of MD was justified by its larger area relative to VD (≈10×). Thermal stratification triggered hypoxia at the bottom of both reservoirs (DO ≈ 1 mg/L), more prolonged and severe in MD. Hypoxia activated Ec and TDS peaks in January likely explained by bottom-sediment nutrient releases, presumably phosphorus. The Ec peak reached 560 μS/cm in MD and 290 μS/cm in VD. The smaller VD peak was probably explained by the action of macrophytes. In March, a 240 NTU turbidity peak occurred in MD, caused by precedent erosive rainfall and the lack of vegetation protection alongside the south border. As expected, the study accomplished clear isolation of factor contributions, verified by Factor and Cluster analyses. Our results can subsidize studies on larger reservoir watersheds requiring restoration, where the isolation of factors is more challenging.