4.1 Analysis of drivers of water quality parameters over time
Consistent with previous studies, the influence of natural and anthropogenic factors plays an irreplaceable role in the process of changing water quality parameters in reservoirs (Dodds and Cole, 2007; Wang et al., 2020).
Because of differences in natural geographic location, there is some variability in WT changes among the three reservoirs. Huanglishu Reservoir is located in Chuzhou on the border with Hefei City and Maanshan City, as well as geographically situated to the south, with higher temperatures and greater exchange of heat in the atmospheric cycle. As a result, Huanglishu Reservoir has the highest water temperatures. Shahe Reservoir is in a populated area, where air circulation is difficult, heat dissipation is weak, for relatively high water temperatures. Chengxi Reservoir borders Mt. Langya and is surrounded by large vegetation cover, with less solar radiation absorbed by the ground, leading to the lowest water temperatures. As shown by Pearson Correlation Analysis, WT showed significant correlations (P < 0.01 or P < 0.001) with pH, DO, CODMn, TP, and Chl-a (Fig. 2). Lower WT reduces the rate of planktonic algal blooms, thereby slowing their ability to consume DO (Varol et al., 2012; Naveedullah et al., 2016). This is the primary explanation for the higher DO concentrations in the Chengxi Reservoir than in the other two reservoirs. Appropriate WT increases phytoplankton abundance in the reservoir, enhances photosynthesis, and reduces CO2, ultimately resulting in higher water pH (Wu et al., 2014). Moreover, the Chengxi Reservoir has the highest phytoplankton population in its own right and hence with a higher pH than the other two reservoirs. It is well known that the biological oxidative decomposition of CODMn needs to consume a large amount of DO, however, Excessive WT causes planktonic algae to preemptively deplete DO, which in turn alters the CODMn concentration (Yin et al., 2011). Furthermore, ANOVA showed that CODMnwas remarkably different (P < 0.05) between clusters (Tables 3 to 5), which may be related to the occurrence of water bloom in the reservoirs. During the outbreak of blooms, phytoplankton will rapidly multiply and die, generating large amounts of organic matter and causing elevated CODMn concentrations in reservoirs (Klemas, 2012; Smayda, 2008).
Under anthropogenic interference, the elevated discharge of industrial wastewater and the rise in agricultural fertilizers and livestock manure have to a large extent contributed to the enrichment of reservoir waters with large quantities of organic matter and inorganic salts, which has led to an increase in the concentration of BOD (Liu et al., 2022). Of these, the area surrounded by Huanglishu Reservoir is mostly villages and towns, with relatively backward industrial and agricultural technology and slow development. Therefore, the reservoir does not present the characteristic. As everyone knows, the removal of nitrogen sources mainly relies on nitrification under aerobic conditions and denitrification under anaerobic conditions (Ma et al., 2016; Zhou and Hosomi, 2008). However, when the water environment is polluted by humans, the water DO is no longer stabilized and both effects are repulsive, resulting in the accumulation of NH3-N and TN from factory effluents and municipal wastewater discharges (Yao et al., 2011). Owing to the higher DO concentrations in Chengxi Reservoir, the nitrogen source concentrations (TN and NH3-N) for water quality are lower than the other two reservoirs. As for the variation of TP concentrations, all three reservoirs characterized by higher concentrations in spring and summer (March to August) than in fall and winter (September to December). That is primarily attributed to the fact that the former period is a busy farming season, agricultural cultivation requires the use of large quantities of phosphorus fertilizers, which, after being dissolved by rainfall, sink into the reservoirs along with pollutants enriched with highly concentrated sources of phosphorus, ultimately resulting in higher TP concentrations than in the latter period (Kim et al., 2001). Chl-a can reflect the growth of phytoplankton in water (Wurtsbaugh et al., 2019). As revealed by the TLI method, the eutrophication level of the Huanglishu Reservoir and the Shahe Reservoir was higher than Chengxi Reservoir (Fig. 6 and Table 2). The former two reservoirs would receive more inputs of exogenous nutrients, phytoplankton biomass can be dramatically boosted. Hence, it was relatively easy to increase the Chl-a concentration. (Saluja and Garg, 2017).
Certainly, natural and anthropogenic factors can simultaneously affect water quality parameters. The irregularity exhibited by the parameter SD over time is mainly related to the erratic rainfall in the area and the irregular water demand of the population. There is less small-scale vegetation along the reservoir, large amounts of sediment are easily washed into the reservoir by rainwater, impeding light penetration and causing changes in SD. The water demand of the local population can lead to irregular opening and releasing of the reservoir, which greatly affects the stabilization of the phytoplankton population in the reservoir. Excess phytoplankton can green the water and reduce the SD of the water body (Sommaruga and Augustin, 2006). Moreover, the different light intensities under sunny and cloudy days are another important factor contributing to the erratic SD variation. Again, because of the difference in reservoir eutrophication, clean water quality is less likely to experience algal overgrowth and does not easily prevent light from penetrating the water layer (Guo et al., 2022), thus SD changes are more stable in Chengxi Reservoir.