Methods on quantifying ecological operation target based on bio-hydrological response model for stimulating the reproduction of four major Chinese carps in the Three Gorges Reservoir

为了改善生态调度效果,提高生态调度实践成功率,采用医学领域适用性广、准确性高、直观性强的受试者工作特征曲线法(receiver operating characteristic, ROC),以三峡水库为例,提出了一套基于生物水文响应模型的刺激四大家鱼繁殖生态调度目标量化方法,包括涨水事件界定及生态水文指标选择、鱼类自然繁殖事件及有效繁殖响应界定、生物-水文响应模型构建与多指标判别优选等多个环节。运用三峡水库运行后2013—2019年(除2016年)6年47场次宜昌江段的涨水事件和24次四大家鱼有效繁殖响应的实际观测数据构建模型并量化三峡水库生态调度目标。结果表明:(1)依据初始流量Q₀、洪峰流量Q_peak、流量总增长量Q_sum和流量日增长率dQ这4个指标所建立的单指标生物-水文响应模型具有较好判别四大家鱼自然繁殖对水文条件是否响应的能力;(2)各评价指标Q₀、Q_peak、Q_sum、dQ的最佳阈值分别为14960 m³/s、19610 m³/s、4050 m³/s及1405 m³/(s·d);单指标模型性能排序为dQ＞Q_peak＞Q_sum＞Q₀;(3)依据dQ或Q_peak的指标组合方式建立的多指标生物-水文响应模型综合性能最佳。结果表明,ROC模型方法结果形式简单、预见性强,据此量化生态调度目标,可为其他流域、水库的鱼类繁殖响应判别和水库生态调度方案制定提供参考借鉴。;In order to improve the ecological operation effect and increase the success rate of ecological operation practice, a set of ecological operation target quantification methods based on the bio-hydrological response model for stimulating reproduction of four major Chinese carps is proposed. The method includes the definition of upwelling event, the selection of eco-hydrological indicators, the definition of natural fish reproduction event and effective reproduction response, the construction of bio-hydrological response model and multi-indicator discriminative preference. The actual observed data of 47 upwelling events and 24 effective reproduction responses of four fish in the Yichang River section for 6 years after the operation of the Three Gorges Reservoir from 2013 to 2019 (except 2016), were used to construct the model and quantify the ecological planning goals of the reservoir. The results showed that (1) the single-indicator bio-hydrological response model based on the four indicators of initial flow (Q₀), peak flood flow (Q_peak), total flow growth (Q_sum) and daily flow growth rate (dQ) had the ability to discriminate whether the natural reproduction of the four major carp was responsive to hydrological conditions; (2) the optimal thresholds for the above evaluation indicators were 14960 m³/s, 19610 m³/s, 4050 m³/s and 1405 m³/(s·d), respectively; the performance of the single indicator model was ranked as dQ>Q_peak>Q_sum>Q₀; (3) the comprehensive performance of the multi-indicator bio-hydrological response model based on the combination of dQ or Q_peak was the best. The model results are simple and predictable, and can be used to quantify the ecological operating targets, which can provide a reference for discriminating fish reproduction response and formulating reservoir ecological planning in different watersheds and reservoirs.