Abstract: Objective This study was intended to elucidate the spatial distribution characteristics of fly density, find correlated environmental factors, and respond to the need of fly control in disaster fields. Methods Respectively, in the study, the spatial statistical analysis was used to detect the spatial aggregation of fly population on the scales of village and town. Furthermore, spatial error regression was performed to uncover the correlated environmental factors with fly density on the same two geographic scales. Results The spatial aggregation of fly population displayed statistical significance in flooding sites on the level of village(Moran's I=0.387, P=0.016)and town(Moran's I=0.381, P=0.039), which means that such spatial pattern was not original from random process. Accordingly, as the spatial regression shown, the fly density correlated with different environmental factor on the scales of village and town; The lighting index values was related to village(Correlation Coefficient=0.514)and town(Correlation Coefficient=1.472)scales, while the slop values related only to the village scales(Correlation Coefficient=-2.398). Conclusion Present paper not only shed light on the fly spatial distribution pattern in flooding area in Fangshan, but also could provide direction for the fly control for any disaster sites being similar to flooding region in Fangshan.

Key words: Flooding disaster, Scales, Fly density, Spatial error analysis

摘要： 目的 研究北京市房山区水灾后现场蝇类密度空间分布特征及其相关的环境因素，为灾后蝇类防控提供技术支持。方法 应用空间统计方法对乡镇与村庄尺度上蝇类密度的空间聚集性进行检测，并应用空间误差回归方法研究两级尺度上蝇类孳生密度相关的环境因素。结果 研究表明，房山区水灾后蝇类密度在乡镇（Moran's I=0.381，P=0.039）与村庄（Moran's I=0.387，P=0.016）两级尺度的空间聚集差异有统计学意义，蝇密度在水灾后现场的空间分布并不是随机过程。另外，空间误差回归分析表明，乡镇与村庄两级尺度上与蝇类密度相关的环境因素并不相同，灯光指数与村庄和乡镇两级尺度上的蝇密度均呈正相关（相关系数为0.514和1.472），而坡度值与村庄尺度上蝇密度呈负相关（相关系数为-2.398）。结论 研究结果不仅阐明了房山地区水灾后不同地理尺度上蝇类的空间分布规律，也可为房山区或类似地区灾后蝇类防控工作提供有针对性的指导。

关键词: 洪灾, 尺度, 蝇密度, 空间误差统计分析