Indices of bird-habitat preference from field surveys of birds and remote sensing of land cover: a study of south-eastern England with wider implications for conservation and biodiversity assessment

摘要

Aim This paper describes the development of novel indices of bird-habitat preference to examine bird species' use of habitats and their distributions relative to habitats. It assesses the implications for bird conservation regionally and the scope for biodiversity assessments generally.Location A 200 km by 400 km area of farmland with seminatural and urban areas, covering south-eastern England.Methods Cluster analysis was used to link birds to landscapes. Cluster centroid coordinate values were processed to derive indices of bird-habitat preference. Further developments assessed the relative values of individual habitats for birds.Results Clustering objectively linked birds to landscapes. Maps of the clusters showed strong regional patterns associated with distinctive habitat assemblages. Derived indices related bird species directly to individual habitats and habitats to birds. Even rare species and scarce habitats showed successful linkages, often to each other. Objective corroboration strongly supported the associations of coastal, wetland, urban and woodland birds and habitats; but, it suggested that farmland birds, whose numbers have nearly halved since 1977, may prefer alternative habitats.Main conclusions Land cover maps from remote sensing provide an effective way to link birds to habitats and vice versa. Thus, generalized habitat maps might be used to extrapolate localized or sample-based bird observations or the results of autecological studies, helping to predict and understand bird distributions in the wider countryside. The weak links between farmland birds and farmland habitats in a region dominated by farming, suggests that reasons for the decline in farmland birds may be deep seated and thus hard to reverse. The procedures described are repeatable elsewhere and applicable more generally to evaluate landscapes and biodiversity. It is suggested that remote sensing could rarely be bettered as a means of assessing habitats, comprehensively, over wide areas, in most parts of the world.