Multi-factor modeling of chlorophyll-a in South China's subtropical reservoirs using long-term monitoring data for quantitative analysis

Abstract

Eutrophication and harmful algal blooms, driven by complex interactions among nutrients and climate, threaten freshwater ecosystems globally, particularly in densely populated Asian regions where rapid urbanization and agricultural intensification exacerbate nutrient pollution. Understanding the non-linear interactions among water temperature, nutrient levels, and chlorophyll-a (Chl-a) dynamics is crucial for addressing eutrophication in freshwater ecosystems. Many existing studies, however, tend to oversimplify these relationships and lack validation with long-term field data. Here, we conducted multi-year field monitoring (2020-2024) of key environmental factors, including total nitrogen (TN), total phosphorus (TP), water temperature, and Chl-a, across three reservoirs in Guangdong Province, China: Tiantangshan (S1), Baisha River (S2), and Meizhou (S3). Strong positive correlations were found between Chl-a and TN, TP, and temperature. Numerical analysis of the long-term data revealed TN as a more influential driver than TP for Chl-a proliferation in these systems, with Chl-a increasing by an average of 4.2 ug/L per unit increase in TN, compared to 2.8 ug/L per unit increase in TP. Based on the collected data, we developed and calibrated a dynamic multi-factor hydro-ecological model. The model accurately reproduced the observed Chl-a patterns, identifying synergistic effects between temperature and nutrients, particularly a 15% enhancement in Chl-a growth rate when temperature exceeded 25 concurrent with high TN.

Figures (7)

• Figure 1: Locations of the three studied reservoirs (S1: Tiantangshan; S2: Baisha River; S3: Meizhou) in Guangdong Province, China. The inset map shows their geographical context within the Pearl River Basin.
• Figure 2: Numerical demonstration of local asymptotic stability for the dynamic model (Eq. \ref{['model']}). (a) Time series showing the convergence of TN ($N$), TP ($P$), and Chl-a ($C$) concentrations to the stable equilibrium $E$. (b) Phase portrait in the $N$-$P$ plane, with arrows indicating the direction of system trajectories towards the stable focus at $E$. The simulation parameters are set as follows: $k_{10} = 0.10$, $k_{20} = 0.6$, $\gamma_{1} = 0.3$, $\gamma_{2} = 0.1$, $\alpha_{1} = 0.3$, $\alpha_{2} = 0.1$, $Q_{N} = 0.01$, $Q_{P} = 0.01$, $N_{in} = 0.5$, $P_{in} = 0.02$, $S_{N} = 0.5$, $S_{P} = 0.1$, $K_{N} = 0.5$, $K_{P} = 0.5$, $T = 25$, $d = 0.002$, $\theta = 1.04$, $K = 0.02$, $w = 0.3$.
• Figure 3: Interannual variations (2020-2024) of total nitrogen (TN), total phosphorus (TP), and Chl-a concentrations in Tiantangshan (S1), Baisha River (S2), and Meizhou (S3) reservoirs. Note the distinct seasonal temperature cycle, the high variability in TN, and the overall increasing trend in Chl-a.
• Figure 4: Detailed temporal trends of (a) water temperature, (b) TN, (c) TP, and (d) Chl-a from 2020 to 2024. The plots highlight the seasonal consistency of temperature, episodic nutrient peaks (particularly in S2), and the synchronous rise in Chl-a and TN in S1 and S2 during 2022.
• Figure 5: PCA of environmental variables in Tiantangshan (S1), Baisha River (S2), and Meizhou (S3) reservoirs (2020–2024). (a) Scree plot showing the variance explained by the first five principal components. The first two PCs (PC1 and PC2) explain >70% of the total variance. (b) Biplot of sample scores (points) and variable loadings (arrows) in the PC1-PC2 space. The trajectory of points from 2020 to 2024 shows a shift along the PC1 axis, correlated with increasing Chl-a. Vectors for TN and TP align with PC2.