Abstract
Reservoir bays, the boundary of terrestrial and water where water fluidity slows down and self-purification ability turn weak, hence they are especially sensitive to terrestrial exogenous pollutants, even resulting in eutrophication. According to N:P, water nutrients types can be divided into N limited, P limited and N + P limited classes. Phytoplankton biomass is represented by chlorophyll a, which is one of the sensitive indicators of water eutrophication. Comprehensively traced non-point pollution from terrestrial exogenous pollutants (fertilizer, soil release, anthropogenic discharge) to water nutrients that happen in reservoir bays is of great significance. This paper identified the dominant environmental variables and nutrients limited types of reservoir bays at storage and discharge periods, constructed partial least squares structural equation model (PLS-SEM) to explore the impacts of terrestrial exogenous pollutants. Results showed that in storage period water contamination mainly came from residential discharge and soil endogenous release, the total contribution rate reached 61%. In discharge period, with the increase of rainfall – runoff erosion, the explanatory ability of land use, topography and landscape pattern to water quality increased, up to 58%. The dominant nutrients limited types of reservoir bays were P limited (35%–47%) and N + P limited (35%–59%) at both stages, N limited situations less than 20% and generally appeared in storage period. Whatever the nutrients limited type was, phosphorus always had a higher effect on phytoplankton biomass. In N limited situation, nitrogen mainly from soil release (total effect = 0.6) and phosphorus from fertilizer (total effect = 0.22) and soil release (total effect = 0.17). In P limited situation, all three sources had almost high effects on nitrogen, phosphorus, and phytoplankton biomass. In N + P limited situation, the anthropogenic discharge was the main source of nutrients and the primary threaten factor for phytoplankton biomass. The approaches employed in this study could be generalized to the other basin and the results were significant to early warning and controlling water eutrophication.