Toward a morphometry- and probability-informed framework for eutrophication management: Insights from a shallow...

Eutrophication management has long relied on an empirical cascade linking external nutrient loads to in-lake nutrient concentrations and, ultimately, algal biomass. Yet this cascade is often treated as structurally stable across lakes, despite strong differences in morphometry, as well as its associated mixing regime and internal nutrient cycling. Using a paired comparison of the shallow polymictic Lake Taihu and the deep stratified reservoir Lake Qiandaohu, we show that this eutrophication cascade can weaken at different links under contrasting morphometric and mixing settings, with important consequences for nutrient-control strategies and threshold setting. In Lake Taihu, external loading signals were substantially clearer for total nitrogen (TN) than for total phosphorus (TP), whereas the opposite pattern occurred in Lake Qiandaohu, consistent with contrasting nutrient-retention and internal-cycling pathways between the two systems. Nutrient-chlorophyll a (Chla) relationships also diverged: upper-bound Chla in Lake Taihu increased with both TN and TP, supporting dual nutrient control, whereas Chla in Lake Qiandaohu remained primarily responsive to TP, with TN sensitivity emerging only above approximately 0.89 mg/L and mainly in localized high-TN areas. Across both lakes, pronounced wedge-shaped nutrient-Chla relationships further showed that similar nutrient concentrations can produce markedly different algal outcomes under varying environmental conditions. We therefore reinterpret nutrient thresholds not as fixed deterministic cutoffs, but as risk-dependent probability boundaries, and use quantile regression to explicitly link Chla management targets with different levels of management assurance. For Lake Taihu, where load-concentration relationships were sufficiently robust, these probabilistic concentration thresholds were further translated into external nutrient-load targets. Together, these results establish a morphometry-informed diagnostic pathway that identifies where the load-concentration-response cascade remains reliable, determines whether nitrogen (N), phosphorus (P), or dual nutrient control is warranted, and converts ecological targets into probability-based nutrient thresholds and, where feasible, actionable load targets.