To determine the distribution, numbers and densities of seabirds in the BPNS, the data collected between September 1992 and December 2004 were analysed. Additionally, the data from the counts in 2005 were used to determine the species-diversity (see further). Thus, the compiled dataset does comprise data of standardised counts that are well distributed both temporally (both between years and within years) and spatially on the BPNS. Both swimming and flying birds were counted by a standardised strip-transect-method (Tasker et al., 1984). All swimming birds that are within a distance of 300 m and in an angle of 90° forward from the study-vessel were counted in intervals of 10 minutes. Flying birds were counted using a snapshot method (Komdeur et al., 1992). All flying birds within a distance of 300 m and an angle of 90° forward from the study-vessel were counted every minute. In order to compensate for missed small and dark birds, the mean density of swimming birds has been multiplied with an internationally accepted correction factor (Stone et al., 1995). The results of these counts were transformed into densities by taking into account the speed of the research-vessel. All counts were reduced to the spatial mid points of the concerned 10-minute tracks. These midpoints were called position keys or ‘poskeys’ and are displayed in the dataset in degrees northern latitude and eastern longitude and hold the local densities of all species (number per square km). If the ship changed its course within a 10-minute count, the counts relate to a shorter period. To avoid that counts in very short periods of time would bias the calculation of bird densities, all poskeys in which less than 1 km was covered were omitted. Since ferry counts may result in an underestimation of the densities of certain species (e.g. alcidae and divers) because of the higher speed and the height of the observation platform, the data collected from ferries were not retained in the processed dataset. After these selections, data of 10.808 poskeys were retained. For the calculation of the number of species per 3x3 km-square all counts (also counts from ferries and those of 2005) were used (15.908 poskeys)