Geographic Characteristics and Meteorological Factors Dominate the Variation of Chlorophyll‐a in Lakes and Reservoirs With Higher TP Concentrations

Abstract

Nutrients such as phosphorus and nitrogen lead to extensive growth of harmful algae in lakes and reservoirs, which results in eutrophication. The driving mechanism of primary productivity change in lakes and reservoirs at a wide spatial and temporal scale remains largely unknown. We establish a water quality database using a stacking machine learning model, including monthly chlorophyll‐a (Chl‐a), total nitrogen (TN) and phosphorus (TP) concentrations in 255 lakes and 332 reservoirs from 1980 to 2018. We find an increase in the number of lakes and reservoirs at risk of algal blooms, with approximately 2.66% exhibiting Chl‐a concentrations exceeding 30 μg L−1 by 2018. Variations in Chl‐a concentrations were not entirely synchronized at the individual and regional levels with TN and TP concentrations, or their stoichiometric ratios, as estimated by a hierarchical linear model spanning 1980 to 2018. Further, we discovered unexpected effects of geographical features (i.e., latitude, longitude, and slope) and meteorological factors (i.e., air temperature) on Chl‐a concentrations in the lakes and reservoirs with higher TP concentrations (>0.041 mg L−1 for lakes and >0.027 mg L−1 for reservoirs), through the use of multiple regression trees and structural equation model analysis. Our findings underscore the importance of (a) implementing flexible nutrient pollution control approaches based on nutrient ecoregions that consider geographical variations, and (b) developing mitigation and adaptation strategies to address the uncertain risks posed by climate change in the prevention and control of eutrophication in lakes and reservoirs.