The European honey bee species (Apis mellifera L.) is under increasing pressure from anthropogenic and other stressors. Winter mortality of entire colonies is generally attributed to biological, environmental, and management conditions. The rates of winter mortality can vary extremely from place to place. A landscape approach is used here to examine the dependency between spatially distributed winter mortality rates, environmental and biological conditions, and apiary management. The analysis was applied to data for the region of Wallonia in Belgium with winter mortality rates obtained from the European project EPILOBEE. Potential explanatory variables were spatially allocated based on GIS analysis, and subjected to binomial linear regression to identify the most predominant variables related to bee winter mortality. The results point to infestation with Varroa, the number of frost days, the potential flying hours, the connectivity of the natural landscape, and the use of plant protection products as most dominant causes for the region of Wallonia. The outcomes of this study will help focus beekeeping and environmental management to improve bee health and the effectiveness of apiary practices. The approach surpasses application to the problem of bee mortality and could be used to compare and rank the causes of other environmental problems by their significance, particularly when these are interdependent and spatially differentiated.

欧洲蜜蜂（Apis mellifera）L.）受到来自人为和其他压力因素的压力越来越大。整个殖民地的冬季死亡率通常归因于生物学，环境和管理条件。各地的冬季死亡率差异很大。这里使用景观方法来检查空间分布的冬季死亡率，环境和生物条件以及养蜂场之间的依赖关系。该分析已应用于比利时瓦隆地区的数据，并从欧洲项目EPILOBEE获得了冬季死亡率。基于GIS分析，对潜在的解释变量进行了空间分配，并进行了二项式线性回归，以确定与蜜蜂冬季死亡率相关的最主要变量。结果表明有Varroa感染，霜冻天数，潜在飞行时间，自然景观的连通性以及使用植物保护产品是瓦隆大区的最主要原因。这项研究的结果将有助于集中养蜂和环境管理，以改善蜜蜂的健康状况和养蜂业的有效性。该方法超越了对蜜蜂死亡率问题的应用，可用于根据其重要性对其他环境问题的原因进行比较和排序，尤其是当这些问题相互依赖且在空间上有所区别时。