Herrando S., Keller V., Voříšek P., Kipson M., Franch M., Anton M., Pla M., Villero D., Sierdsema H., Kampichler C., Telenský T., Gillings S., Johnston A., Gottschalk T., Guélat J., Sattler T., Brotons L., Titeux N., Jiguet F., Kéry M. & Milanesi P. High resolution maps for the new European Breeding Bird Atlas: A first provision of standardised data and pilot modelled maps. Vogelwelt, 137, in press.
One of the main objectives of the Second European Breeding Bird Atlas (EBBA2) is to provide maps showing bird occurrence at a 10×10 km resolution for as many of the European breeding bird species as possible. With almost 120,000 10×10 km squares across Europe, it would be practically impossible to carry out comprehensive surveys in each spatial unit. Consequently the only feasible approach to achieve this goal is modelling the probability of bird occurrence by means of i) gathering a sample of standardised bird occurrence data, ii) using these data to model the relationships between bird occurrence and the environment (e.g. habitat, climate) and iii) projecting these relationships across the whole set of 10×10 km squares in Europe. In the middle of the fieldwork period of EBBA2 a first collation of standardised data was carried out. Almost 100,000 timed surveys from a total of 12,057 10×10 km squares distributed among almost all countries were gathered and the Spatial Modelling Group of the EBCC explored a large range of possible models for inference about distribution and mapping based on these available data. One of the main difficulties of such modelling was to cope with the huge differences in coverage across Europe: data are abundant and comparatively evenly distributed in many regions of Western Europe but scarce and aggregated in a few areas in the East. This heterogeneity in coverage is higher than it was in any of the previous modelling experiences by EBCC partners at the national level. Another particular challenge to face is the coexistence of diverse field methodologies. Fortunately, the time spent in each survey was recorded providing the analyses with a robust surrogate of survey effort to standardise the original data and generate a first set of modelled maps for a few species. These pilot maps will certainly be improved in the final product thanks to new timed surveys and additional work developing the modelling approach.