Persistent eutrophication frequently causes toxic algal blooms, which is a serious threat to drinking water safety, food security, and public health. Nutrient thresholds, the maximum nutrient load that an aquatic ecosystem can absorb while meeting management objective, are key to avoiding and reducing blooms. The determination of thresholds relies on nutrient load-response curves. The spatial heterogeneity of large shallow lakes in terms of lake characteristics results in different curves shape among areas of the lake, which leads to spatial differences in thresholds. However, the spatial heterogeneity of thresholds is typically neglected; there are few methods to analyse the relationship between river loads and lake-specific area thresholds. Here, we proposed the Area Threshold Analysis Framework to analyse the spatial patterns of nutrient loads and thresholds in lakes. We first quantified the flow and load of the rivers entering the lake. The lake was then zoned and modelled to analyse area thresholds. Finally, an intuitive link between the nutrient loads of specific rivers and the thresholds of specific areas of the lake was established. The results showed a nonlinear without hysteresis response in all areas of Chaohu Lake; the nutrient loads and thresholds were highly variable in space, showing a trend of increasing and then decreasing from west to east. Flow density and wind might be important in influencing the spatial distribution of thresholds. The different effects of the wind and flow density on total phosphorus (TP) and chlorophyll a (Chl-a) lead to large differences in the thresholds with TP and Chl-a as management objectives, respectively. The large gap between nutrient loads and thresholds made it important for management to consider appropriate management goals to deal with unrealistic nutrient reductions. Achieving year-round Chl-a ≤ 30 µg/l in Chaohu Lake was easier than TP ≤ 0.05 mg/l, since for the latter, the Shiwuli, Paihe, and Nanfei river loads needed reductions of >80 %. In addition, for Area 1 and 2, it was more practical to first start meeting TP ≤ 0.05 mg/l for 2 or 3 seasons of the year than for 4 seasons. Overall, we developed a new framework for spatial threshold analysis and established an intuitive link between nutrient loads and thresholds in large shallow lakes. These results are valuable for understanding the threshold properties of spatially heterogeneous ecosystems in general and provide a reference for watershed nutrient management and ecological restoration of lakes.

持续的富营养化经常导致有毒藻类大量繁殖，对饮用水安全、粮食安全和公众健康构成严重威胁。营养阈值，即水生生态系统在满足管理目标的同时可以吸收的最大营养负荷，是避免和减少水华的关键。阈值的确定依赖于营养负荷-响应曲线。大型浅水湖泊在湖泊特征方面的空间异质性导致湖泊区域之间曲线形状不同，从而导致阈值的空间差异。然而，阈值的空间异质性通常被忽略；分析河流负荷与湖泊面积阈值之间关系的方法很少。在这里，我们提出了区域阈值分析框架分析湖泊养分负荷和阈值的空间格局。我们首先量化了流入湖泊的河流的流量和负荷。然后对该湖进行分区和建模以分析面积阈值。最后，建立了特定河流的营养负荷与湖泊特定区域的阈值之间的直观联系。结果表明巢湖全区呈非线性无滞后响应；养分负荷和阈值在空间上存在较大差异，呈由西向东先增后减的趋势。流量密度和风可能对影响阈值的空间分布很重要。风和流密度对总磷（TP）和叶绿素a的不同影响（Chl-a）分别导致以 TP 和 Chl-a 为管理目标的阈值差异很大。养分负荷和阈值之间的巨大差距使得管理层必须考虑适当的管理目标来处理不切实际的养分减少。在巢湖实现全年 Chl-a ≤ 30 µg/l 比 TP ≤ 0.05 mg/l 更容易，因为对于后者，十五里、排河和南淝河的负荷需要降低 >80%。此外，对于 1 区和 2 区，首先开始在一年中 2 或 3 个季节达到 TP ≤ 0.05 mg/l 比 4 个季节更实际。总体而言，我们开发了一个新的空间阈值分析框架，并在大型浅水湖泊的养分负荷和阈值之间建立了直观的联系。