Abstract: Objective To establish a mosquito lamp sampling model based on Taylor's power law. Methods From April to November 2019, a total of 229 surveillance points of carbon dioxide trapping lamps were set up in 15 districts of Shanghai, China during every period of ten days to monitor the densities of Culex pipiens pallens and Aedes albopictus. The data were used to fit the Taylor's power law function equation:s²=a×x^b, which described the relationship between the mean and variance of the density of every mosquito species in each period of ten days. The results derived from the equation were substituted into the sample size formula to establish the sampling model:n=t²×a×x^b-2×D^-2. The number of sample units needed for Cx. pipiens pallens and Ae. albopictus density surveillance was calculated at the 95% confidence level according to the sampling model. Results The fitting results of Taylor's power equation were as follows:a=5.847 8, b=1.525 4, and R²=0.911 1 (P<0.001) for Cx. pipiens pallens; and a=3.668 2, b=1.302 6, and R²=0.962 0 (P<0.001) for Ae. albopictus. The fitting results were entered into the sampling model, along with the value of t distribution and the D value of relative precision. The D value at the 95% confidence level for Cx. pipiens pallens was <0.35 during the middle ten days of April and during the first ten days of May to the middle ten days of November, and was <0.25 during the last ten days of May to the last ten days of August. The D value for Ae. albopictus was between 0.25 and 0.35 from the first ten days of July to the first ten days of October and during the last ten days of October, and was >0.35 during other periods. Conclusion This sampling model has practical significance and can be used to estimate the optimal sample size for light trap monitoring. The current mosquito surveillance method by carbon dioxide trapping lamps in Shanghai shows higher relative precision for Cx. pipiens pallens than Ae. albopictus. The relative precision for Ae. albopictus can be improved by increasing the number of light traps.

Key words: Mosquito lamp, Taylor's power law, Sampling, Aedes albopictus, Culex pipiens pallens

摘要： 目的 建立基于泰勒幂法则（Taylor’s power law）的诱蚊灯抽样模型。方法 2019年4－11月在上海市15个区，每旬共设置229个CO₂诱蚊灯监测点对淡色库蚊和白纹伊蚊密度进行监测，以此数据建立每旬每个蚊种的蚊密度样本均数（x）与方差（s²）的泰勒幂法则幂函数方程s²=a×x^b，将结果代入样本量计算公式，建立n=t²×a×x^b-2×D^-2的抽样模型，计算在95%可信区间（95%CI）条件下，上海市淡色库蚊和白纹伊蚊种群开展诱蚊灯密度监测或调查研究所需的抽样单元数。结果 淡色库蚊的泰勒幂方程拟合结果：a=5.847 8，b=1.525 4，R²=0.911 1（P<0.001）；白纹伊蚊泰勒幂方程拟合结果：a=3.668 2，b=1.302 6，R²=0.962 0（P<0.001）。拟合结果与t分布概率值及相对精度D值组合，构成抽样模型。结果显示，在2019年5月上旬至11月中旬间，95%CI条件下，除4月上旬和下旬外，其余时间淡色库蚊的抽样相对精度D值均<0.35，其中5月下旬至8月下旬均<0.25；7月上旬至10月上旬，以及10月下旬白纹伊蚊抽样相对精度D值在0.25～0.35之间，其余时间D值均>0.35。结论 该抽样模型具备实用意义，可以据此估算诱蚊灯监测的最佳样本含量。目前上海市的CO₂诱蚊灯监测方法，淡色库蚊的相对精度高于白纹伊蚊，若要提高白纹伊蚊抽样相对精度需增加诱蚊灯数量。

关键词: 诱蚊灯, 泰勒幂法则, 抽样, 白纹伊蚊, 淡色库蚊