Application limitations and improvement recommendations of trophic state indices in the eutrophication level assessment of small shallow lakes along the lower reach of the Yangtze River

基于多个环境变量、以等权或者加权平均法获得的综合营养状态指数(Carlson指数:TSIc;修正后的Carlson指数:TSImc;综合营养状态指数:TLIc;湖库营养状态指数:EIc)得分是当前评价湖泊营养水平的主要依据,其值计算时先要获得基于单个环境变量的营养状态指数分值。若基于总氮(TN)、总磷(TP)和透明度(SD)等的单个理化指标与基于叶绿素a(Chl.a)的营养状态指数得分间存在显著差异,表明依据理化指标的评估结果低估或者高估湖泊的实际营养水平。长江下游分布大量的小型浅水富营养化湖泊,但基于理化指标的评估结果是否会误判湖泊的实际富营养化水平的研究很少。本研究以长江下游的14个浅水富营养化湖泊为对象,于2019-2020年按照季度采集水样并测定水环境因子,分析基于理化指标和基于Chl.a的各营养状态指数(TSI、TSIm、TLI和EI)得分间的差异。结果表明,基于SD的TSI、TSIm、TLI和EI的年均得分均显著高于基于Chl.a的相应营养状态指数的年均值;除TSI(TP)vs. TSI(Chl.a)外,基于TP与基于Chl.a的其他营养指数的年均得分间均有显著性差异;仅TLI(TN)与TLI(Chl.a)的年均得分间无显著性差异。这表明,除TSI(TP)和TLI(TN)外,基于理化指标与基于Chl.a的其他营养状态指数的评估结果间均不能相互匹配。针对基于理化指标的营养状态指数在长江下游小型湖泊富营养化水平评估中的问题,提出如下改进建议:1)依据lnSD-lnChl.a-lnTP间的回归关系重构了营养状态指数(TSIr、TSImr和TLIr)方程,结合基于单个理化指标与基于Chl.a的各指数得分间的匹配性、基于多个环境变量的综合营养状态指数TSIrc、TSImrc和TLIrc得分间的差异性及其与水体实测Chl.a浓度对应的营养状态指数赋分区间的一致性,建议在长江下游小型浅水湖泊的富营养化水平评估中使用TSImr和TLIr指数;2)以加权平均法计算综合营养状态指数TLIrc得分时,发现基于理化指标实测值的归一化相关权重(Wj1)与基于直接引用文献数据的相关权重(Wj2)的TLIrc得分间无显著差异。从简便性角度出发,建议直接以Wj2计算TLIrc指数得分。;The comprehensive trophic state indices based on multiple environment variables, including the Carlson's trophic state index (TSIc), the modified Carlson's trophic state index (TSImc), the comprehensive trophic level index (TLIc) and the lake and reservoir trophic state index (EIc), are widely used in the trophic level evaluation of lakes. The lake trophic level is generally assessed according to the scores of TSIc, TSImc, TLIc and EIc, which are based on the scores of trophic level indices (TSI, TSIm, TLI and EI) derived from single environment variables, and thereby are calculated with the method of equal weight averages or by weighted averages method. If the scores of trophic level indices based on single physio-chemical indicators such as total nitrogen (TN), total phosphorus (TP) and secchi disk (SD) are significantly different from those based on chlorophyll-a (Chl.a), the assessment results based on the physio-chemical indicators could have underestimated or overestimated the actual trophic levels of the lakes. A large number of small shallow eutrophic lakes are distributed in the lower reach of the Yangtze River. However, little research has been undertaken to determine whether the assessment results based on physio-chemical indicators could misjudge the actual eutrophication levels of these lakes. Here, 14 small shallow eutrophic lakes were selected as the research waterbodies. The water samples in April, July, and October 2019 and January 2020 were collected, and Chl.a, TN and TP concentrations as well as SD were measured simultaneously. Subsequently, the differences between each of the trophic state indices scores (TSI, TSIm, TLI and EI) based on SD, TN, TP and the scores based on Chl.a were detected. The results showed that the annual scores of TSI, TSIm, TLI and EI based on SD were all significantly higher than those based on Chl.a. The score of TSI (TP) was not significantly different from that of TSI (Chl.a), but the scores of TSIm, TLI and EI based on TP were remarkably different from those based on Chl.a. Among TN-based trophic level indices, only the annual score of TLI (TN) was not significantly different from that of TLI (Chl.a). The abovementioned results indicated that except for TSI (TP) and TLI (TN), the assessment results based on physio-chemical indicators and that based on Chl.a could not match each other. For resolving the problem in the eutrophication level assessment of these small lakes based on physio-chemical indicators, the following improvement suggestions were proposed: 1) based on the lnSD-lnChl.a-lnTP linear relationships, the trophic state index equations of TSIr, TSImr and TLIr were reconstructed. TSImr and TLIr indices were recommended to be used in the eutrophication level assessment of the small shallow lakes according to the match degree between the scores of based on single environment variables and that of the Chl.a-based indices, the differences among the scores of TSIrc, TSImrc and TLIrc based on multiple physio-chemical indicators, and the consistency between the scores of TSIrc, TSImrc and TLIrc and the score zones of trophic level indices corresponding to the measured Chl.a concentrations; 2) the score of TLIrc was calculated by weighted averages methods. TLIrc score that was based on normalized correlation weight (Wj1) derived from measured physio-chemical indicators was not significantly different from that based on the weights (Wj2) originated directly from literature data. Therefore, it is recommended to calculate the score of TLIrc index with Wj2 for convenience.