Birds in a complex agricultural landscape in Central Greece: the role of landscape elements and the landscape matrix
Agricultural environments play a key role in the conservation of biodiversity in Europe. The vast majority of agricultural lands, however, are located in the non-designated wider countryside embedding protected Natura 2000 sites with little or no protection of habitats. The European Union wants to halt and reverse biodiversity loss within the next decade by focusing on sustainable agriculture as one out of six main drivers of biodiversity loss. Indeed, many farmland bird species have shown pronounced population declines over the last 50 years with ever-increasing habitat modification due to changing farming systems. This process of rapid habitat modification may have stabilized in Western Europe, but it is very likely to continue in Central and Southern Europe. The effects of landscape composition and configuration on bird assemblages have been scarcely studied in the Balkan Peninsula, despite its important contribution to the subcontinent's biodiversity. We assessed breeding bird assemblages-environment relationships by incorporating environmental predictors at two spatial scales. Birds were surveyed in 90 plots by stratified sampling in eight a priori defined landscape types. The objective of this study was to assess the bird composition in the eight landscape types as well as to quantify the independent and confounding influences of local patch level predictors and broader landscape-level predictors using variation partitioning by running multiple CCA analyses. Samples with dense vegetation structures and samples with lesser amounts of vegetation among the eight landscape types appeared to hold significant different species compositions as it is shown by analysis of similarity. These findings were also reflected by the CCA analysis and the environmental predictors were able to distinguish bird communities adequately well. Almost one third (29.6%) of the total variation in species abundance data was explained by both sets of predictor variables together. Confounding influences of the environmental predictors explained 6.5% of the variation, which indicates that there is a remarkable interdependence of the focal scales. This might have major implications for bird conservation planning and land management in the agricultural countryside outside and within protected Natura 2000 sites.