Data from: Testing scale-dependent effects of semi-natural habitats on farmland biodiversity

The effectiveness of conservation interventions for maximizing biodiversity benefits from agri-environment schemes (AESs) is expected to depend on the quantity of semi-natural habitats in the surrounding landscape. To verify this hypothesis, we developed a hierarchical sampling design to assess the effects of field boundary type and cover of semi-natural habitats in the landscape at two nested spatial scales. We sampled three types of field boundaries with increasing structural complexity (grass margin - simple hedgerow - complex hedgerow) in paired landscapes with presence or absence of semi-natural habitats (radius 0.5 km), that in turn, were nested within 15 areas with different proportions of semi-natural habitats at a larger spatial scale (10 × 10 km). Overall, 90 field boundaries were sampled across a Mediterranean region (NE Italy). We considered species richness response across three different taxonomic groups: vascular plants, butterflies, and tachinid flies. No interactions between type of field boundary and surrounding landscape were found at either 0.5 and 10 km indicating that the quality of field boundary had the same effect irrespective of the cover of semi-natural habitats. At the local scale, extended-width grass margins yielded higher plant species richness, while hedgerows yielded higher species richness of butterflies and tachinids. At the 0.5 km landscape scale, the effect of the proportion of semi-natural habitats was neutral for plants and tachinids, while butterflies were positively related to the proportion of forest. At the 10 km landscape scale, only butterflies responded positively to the proportion of semi-natural habitats. Our study confirmed the importance of testing multiple scales when considering species from different taxa and with different mobility. We showed that the quality of field boundaries at the local scale was an important factor in enhancing farmland biodiversity. For butterflies, AESs should focus particular attention on preservation of forest patches in agricultural landscapes within 0.5 km as well as the conservation of semi-natural habitats at a wider landscape scale.

旨在通过农业环境计划（Agri-environment Schemes, AESs）实现生物多样性收益最大化的保护干预措施，其成效预计取决于周边景观内半自然生境（semi-natural habitats）的数量。为验证该假说，我们构建了分层抽样设计（hierarchical sampling design），以两个嵌套空间尺度（nested spatial scales）评估田块边界（field boundary）类型与景观内半自然生境覆盖度的影响。我们在存在或不存在半自然生境（半径0.5km范围）的成对景观中，选取了结构复杂度依次递增的三类田块边界：草本边缘带、简单绿篱与复杂绿篱；而这些成对景观又嵌套于15个更大空间尺度（10×10km）、半自然生境占比各异的区域内。本研究共在意大利东北部地中海区域完成了90处田块边界的采样工作。我们针对维管植物（vascular plants）、蝴蝶及寄蝇科蝇类（tachinid flies）三类分类群（taxonomic group）的物种丰富度（species richness）响应展开了分析。在0.5km与10km两个尺度下，均未发现田块边界类型与周边景观间存在交互效应，这表明田块边界的质量对生物多样性的影响不受半自然生境覆盖度的制约。在本地尺度下，加宽型草本边缘带可带来更高的维管植物物种丰富度，而绿篱则能提升蝴蝶与寄蝇科蝇类的物种丰富度。在0.5km景观尺度下，半自然生境占比对维管植物与寄蝇科蝇类的影响呈中性，而蝴蝶物种丰富度与森林占比呈正相关关系。在10km景观尺度下，仅蝴蝶物种丰富度与半自然生境占比呈正相关关系。本研究证实，在研究不同分类群、具备不同迁移能力的物种时，开展多尺度测试的必要性。研究表明，本地尺度下的田块边界质量是提升农田生物多样性的关键因素。针对蝴蝶而言，农业环境计划应重点关注0.5km范围内农业景观中森林斑块的保护，以及更大尺度景观下的半自然生境保育工作。