Elucidating phytoplankton limiting factors in lakes and reservoirs of the Chinese Eastern Plains ecoregion.

Knowledge of phytoplankton limiting factors is essential for cost-efficient lake eutrophication management. Herein, we propose a statistical framework to explore site-specific phytoplankton limiting factors and their dependence on water depth (WD) in 54 lakes in the Chinese Eastern Plains ecoregion. First, the maximal chlorophyll a (Chla) response to total N (TN) or P (TP), representing a region-specific "standard" model where phytoplankton were primarily N- or P-limited, was quantified using a 95% quantile regression. Second, site-specific limiting factors were identified using analogical residual analysis. N- or P-limitation was inferred if Fraction (i.e. fraction of Chla observed and predicted by the "standard" model for a given TN) > 0.95 or Fraction >0.95; if both Fraction and Fraction <0.95 in a specific environmental condition (e.g. high non-algal turbidity), light limitation was suggested. As a result, 5%, 7%, 4%, 36%, 16%, 2%, and 30% of the sampling sites were limited by N, P, N+P, light availability, rapid flushing, abundant macrophytes, and unmeasured factors, respectively. Bloom control suggestions in the short run are proposed considering these actual limiting factors. Furthermore, the maximal Fraction or Fraction response to WD was explored, reflecting the effect of WD on Fraction (or Fraction) without significant confounders. The results indicated that phytoplankton in the studied freshwaters would be potentially light-limited, N-limited, N+P-co-limited, or P-limited depending on WD (<1.8, 1.8-2.1, 2.1-5.2, or >5.2 m, respectively), because N will gradually become surplus with increasing WD, while at very shallow depths, strong sediment re-suspension induces light limitation. This finding implies that long-term nutrient management strategies in the studied freshwaters that have WDs of 0-2.1, 2.1-5.2, and >5.2 m can entail control of N, N+P, and P, respectively. This study provides essential information for formulating context-dependent bloom control for lakes in our study area and serves as a valuable reference for developing a cost-efficient eutrophication management framework for other regions.