Breeding bird species diversity across gradients of land use from forest to agriculture in Europe

Loss, fragmentation and decreasing quality of habitats have been proposed as major threats to biodiversity world-wide, but relatively little is known about biodiversity responses to multiple pressures, particularly at very large spatial scales. We evaluated the relative contributions of four landscape variables (habitat cover, diversity, fragmentation and productivity) in determining different components of avian diversity across Europe. We sampled breeding birds in multiple 1-km(2) landscapes, from high forest cover to intensive agricultural land, in eight countries during 2001-2002. We predicted that the total diversity would peak at intermediate levels of forest cover and fragmentation, and respond positively to increasing habitat diversity and productivity; forest and open-habitat specialists would show threshold conditions along gradients of forest cover and fragmentation, and respond positively to increasing habitat diversity and productivity; resident species would be more strongly impacted by forest cover and fragmentation than migratory species; and generalists and urban species would show weak responses. Measures of total diversity did not peak at intermediate levels of forest cover or fragmentation. Rarefaction-standardized species richness decreased marginally and linearly with increasing forest cover and increased non-linearly with productivity, whereas all measures increased linearly with increasing fragmentation and landscape diversity. Forest and open-habitat specialists responded approximately linearly to forest cover and also weakly to habitat diversity, fragmentation and productivity. Generalists and urban species responded weakly to the landscape variables, but some groups responded non-linearly to productivity and marginally to habitat diversity. Resident species were not consistently more sensitive than migratory species to any of the landscape variables. These findings are relevant to landscapes with relatively long histories of human land-use, and they highlight that habitat loss, fragmentation and habitat-type diversity must all be considered in land-use planning and landscape modeling of avian communities.