© Marcel Burkhardt
Milanesi, P., S. Herrando, M. Pla, D. Villero & V. Keller (2017)
Vogelwelt 137: 53–60
The first EBCC atlas of European Breeding Birds was a milestone in European ornithology and after 30 years, fieldwork for the second European Breeding Bird Atlas (EBBA2) is underway. It not only aims at providing up-to-date distribution maps at a broad scale but also modelled maps showing relative occurrence at a finer resolution.
Bird data collected across Europe from 2013 to 2015 have been compiled by the national atlas coordinators and provided to EBBA2 as part of a pilot data provision in 2015. These data serve as a basis for first analyses by the EBCC Spatial Modelling Group (SMOG).
Some preliminary steps are necessary to develop robust species distribution models (SDMs). Moreover, these steps could also provide information about areas and environmental variable combinations under-represented in the squares sampled so far.
First, we collected a set of predictor variables potentially relevant for bird distribution. Second, we combined predictor variables and information on sampled cells to identify poorly sampled areas in Europe (using a modified version of the Multivariate Environmental Similarity Surfaces, mMESS).
A total of 39 predictor variables (geography, climate, topography, soil types, land cover, habitat structure, and anthropogenic factors) were collected. The pilot data provision provided data for 12,057 cells (10x10 km) across Europe. Considering mMESS map, we also identified parts of our study area where extrapolations based on SDMs have to be considered carefully. In the same way, we also point at environmental variable combinations for which more sampling effort is required. Additional efforts regarding data collection for EBBA2 should focus on these areas/ environmental space in order to minimize the space for which no information will be available for final SDMs.
keywords: breeding bird atlas, Europe, species distribution models, Multivariate Environmental Similarity Surfaces