Wild bees play a major role in the cultivation of crops for human use, in the reproduction of many wild plants and are a key component of biodiversity. Mainly due to human activities, wild bees, like other insects, face a rapid decline in Europe. Understanding species distribution can help to design efficient conservation measures. Species distribution can also be used to estimate pollination ecosystem service potential, which can benefit the production of crops relying on pollination and the reproduction of wild plant communities. The presence of pollinators depends on a combination of environmental and biotic factors, each playing a determining role at different spatial scales. We therefore developed a model composed as a hierarchical framework for environmental predictors: climatic data and Land Use/Land Cover (LULC) variables at the European scale and species-specific habitat information at the local scale. The model combines the advantages of two different existing approaches: pollinator species distribution predictions based on their environmental requirements and knowledge on bee species life-history traits and habitats. This paper presents the predicted distribution of twenty-five wild bee species of the Andrena genus in an agricultural region in Northern Germany. We used oilseed rape pollinators as a case study and compared the potential pollination services to the potential demand in the Case Study Area. The developed framework allows to determine the capacity of landscapes to support pollination ecosystem services from wild bees at the local scale, which can support the identification of vulnerable areas and the design of local scale measures for habitat improvement and for conservation. The hierarchical approach leaves potential for further adaptations in order to improve the prediction of wild bee species dynamics and factors influencing their spatial distribution.

野生蜂在人类种植作物，许多野生植物的繁殖中起着重要作用，并且是生物多样性的关键组成部分。在欧洲，野生蜜蜂与其他昆虫一样，主要是由于人类活动所致。了解物种分布有助于设计有效的保护措施。物种分布还可用于估计授粉生态系统服务潜力，这可以使依赖于授粉的作物生产和野生植物群落的繁殖受益。传粉媒介的存在取决于环境和生物因素的组合，每个因素在不同的空间尺度上起决定性作用。因此，我们开发了一个模型，该模型构成了环境预测器的分层框架：欧洲范围的气候数据和土地利用/土地覆盖率（LULC）变量，地方范围的特定物种栖息地信息。该模型结合了两种不同现有方法的优点：基于传粉媒介对环境的要求以及对蜜蜂物种的生命历史特征和栖息地的了解，对传粉媒介的物种分布进行了预测。本文介绍了25种野生蜂的预测分布。Andrena属在德国北部的一个农业地区。我们以油菜籽油菜传粉者为例，并将潜在授粉服务与案例研究区的潜在需求进行了比较。制定的框架可以确定景观在地方尺度上支持野生蜂授粉生态系统服务的能力，这可以支持对脆弱地区的识别和设计用于改善生境和保护的地方尺度措施。分层方法为进一步改进野生蜂物种动态和影响其空间分布的因素提供了可能。