基于克里金插值的白纹伊蚊空间分布研究

SPATIAL DISTRIBUTION OF AEDES ALBOPICTUS BASED ON KRIGING INTERPOLATION

研究使用克里金(Kriging)模型在诱蚊诱卵器法中预测非监测点的白纹伊蚊Aedes albopictus丰度的可行性,用以识别白纹伊蚊高密度区域,为制定蚊虫防控措施提供依据。本次在约60000 m 2研究区域内设置了132个诱蚊诱卵器监测点,于2020年7月28日至10月6日每周监测1次,统计诱蚊诱卵器指数和卵粒数等。通过计算增量空间自相关来评估白纹伊蚊数量是否存在空间自相关;计算局部Moran′s I(Local Moran′s I)用以分析是否存在具有统计意义的聚集、冷点和空间异常值的点。采用ArcGIS 10.8软件的地统计学分析扩展模块的经验贝叶斯克里金模型进行蚊密度的空间插值。本次研究共回收有效诱蚊诱卵器1308只,阳性424只,平均诱蚊诱卵器指数为32.42。每个监测点10次监测合计捕获的卵粒数增量空间自相关分析显示,全局空间自相关峰值半径(Z值最大)为62 m,Moran′s I指数为0.067,Z值为2.995(P<0.05);局部自相关分析结果显示研究区域呈现出西南角的白纹伊蚊密度较低,其中有3个高密度的异常点;东侧蚊密度较高,其中也有数个低密度的异常点。本次研究显示克里金模型可用于诱蚊诱卵器法监测;在本次研究区域内,白纹伊蚊密度东部高于西部,通过经验贝叶斯克里金模型能有效识别白纹伊蚊高密度区域,为制定高效的防控策略提供依据。

Kriging model was attempted to predict Aedes albopictus abundance in non-monitoring sites by mosq-ovitrap method,so as to identify high density areas of Ae.albopictus and provide basis for formulating mosquito control measures.In this study,132 monitoring points were set up in the study area of about 60000 square meters,which were monitored once a week from July 28 to October 6,2020.The spatial autocorrelation of Ae.albopictus was evaluated by calculating the incremental spatial autocorrelation;the local Moran′s I was analyzed to determine statistically significant points of aggregation,cold points and spatial outliers and the empirical Bayes Kriging model of the geographical analysis extension module in ArcGIS 10.8 software was used for spatial interpolation results of mosquito density.A total of 1308 mosq-ovitraps were recorded in this study,and 424 were positive,with an average positive rate of 32.42%.The spatial autocorrelation analysis showed that the global spatial autocorrelation peak(Z value maximum)was 62 m,Moran′s I index was 0.067,Z value was 2.995(P<0.05);the results of local autocorrelation analysis showed that the density of Ae albopictus in the southwest corner was low,but for three outlier points;the density of mosquitoes in the east was higher,but for several outlier points.As a conclusion,the empirical Bayes Kriging model can sufficiently identify the high density areas of Ae.albopictus and suitable for mosq-ovitrap monitoring program.