Water quality dynamics in a multi-purpose reservoir in the Brazilian semiarid: responses to hydro-climatic variability and external and internal nutrient loadings

Abstract

The Castanhão reservoir holds great importance for the Brazilian semiarid region due to its diverse uses: irrigation, aquaculture, and human water supply. Situated in a semiarid zone, its volume undergoes significant variations over time, especially during dry periods. Between 2012 and 2018, a prolonged drought led to a substantial decrease in the reservoir's volume, reaching as low as 2% of its capacity. Consequently, the water quality deteriorated, transforming the reservoir into a eutrophic state, complicating its management. As phosphorus plays a crucial role in eutrophication, this study aims to evaluate the spatial and temporal variability of water quality parameters. It also seeks to develop and compare different models for estimating nutrient concentration in the water inflow into the reservoir based on easily obtainable data. Additionally, it seeks to develop a mass balance model simulating the total phosphorus (TP) dynamics in the reservoir, considering its various sources such as inflow, aquaculture, and internal phosphorus (P) release from sediment-water interactions. The goal is to use this model to simulate different scenarios and assess the impact of those P sources. Monthly assessments conducted in this study confirm complete nutrient mixing in the reservoir, with temporal

variations outweighing spatial ones, supporting the applicability of the proposed complete- mixing model. The study also emphasizes the prevalence of chemical stratification in dissolved

oxygen levels over thermal stratification, with dissolved oxygen (DO) decreasing downstream, particularly during the rainy season. Diel assessments reveal a diel cycle of thermal stratification, primarily during the dry season, driven by higher wind speeds. This is corroborated by a significant negative correlation between wind speed and the relative water column stability index. In contrast, during the rainy season, the reservoir experiences continuous thermal stratification due to inflowing water being warmer than the reservoir’s water temperature. Notably, a twofold increase in P throughout the day during the rainy season underscores the influence of phytoplankton community dynamics on diel nutrient variation. Additionally, chemical stratification of DO occurs during both dry and rainy seasons, indicating that even during the dry season, when there is no significant inflow, internal nutrient loading can significantly impact the reservoir's water quality. An analysis of the influence of inflow on water quality variables was performed and several predictive TP concentration models were developed and compared. While Ridge and Lasso regressions demonstrated superior performance, the observation that the test performance surpassed that of the training data suggests overfitting. This discrepancy led to the decision to adopt the non-linear model for application in the TP complete mixing model. In addition, a regression tree model showed that the flow rate is the explanatory variable that better explained the TP variation, whereas DO and temperature were the ones that better explained the total nitrogen concentration variation. The analysis of the bed sediment of Castanhão reservoir showed that the majority of P fraction is the inorganic. In addition, the sampling point closest to the dam is the one with more potential for internal P loading. The proposed model of P dynamics in the reservoir achieved good calibration performance (NSE = 0.62; PBIAS = -0.03%). Scenario simulations indicate that an 70% reduction in external load, a 30% reduction in internal load, and a 30% reduction in aquaculture load have the potential to bring the TP concentration within the class two limit by the end of the simulation period. This designation ensures that the water is suitable for human consumption after conventional treatment according to Brazilian legislation. These reductions could be achieved through improvements in sanitation, erosion control, upstream agricultural practices, aeration, sediment management, and enhancement of fish feed conversion efficiency.