Numbers of beached bird corpses and mortality of seabirds, how do they relate: a North Sea study in a wider context

This paper summarizes the results of several experiments with bird corpses in the southernmost part of the North Sea and tries to fit this information in the current knowledge of how mortality at sea relates to densities of beached birds on the shoreline. During thirteen drift experiments off the Belgian coast between February 1996 and March 1999, a total of 634 sea- and coastal tagged bird corpses were dropped at sea. Bird corpses were found to drift by the wind at 2.2-3.1% of its own velocity, a value in line with earlier published reports of 2.2-4%. Recovery rates averaged 12%, which is smaller than an overall mean of 22% for thirty published drift experiments worldwide but conform an older ‘standard’ to multiply the number of birds found dead on the beach with ten to arrive at an estimate of total mortality. Despite many complicating factors and a large variability (recovery rates ranging from 0 to 75%) some general patterns appear, useful for the interpretation of beached bird numbers in the Southern North Sea. Recovery rates depend primarily on the dropping location (distance to the coast) and on the size of the bird. An evaluation of bird strandings during 1992-99 at the Belgian coast demonstrates a significant correlation with the cumulative onshore wind speeds of the ten days prior to the survey. Offshore species are much more dependent on enough onshore winds to strand and are found more evenly distributed on the shoreline. Temporal patterns of beached birds usually follow those of seabirds at sea with a time-lag of at least one month. Persistence experiments during three winters with 562 tagged corpses confirm the picture of short residence times in the Southern North Sea. A mean persistence time of merely 5.8-13.3 days and the fact that 50-80% of all birds cannot be recovered anymore after 9 days, indicate how strongly persistence must affect the results of beached bird surveys and drift experiments. Small-sized corpses have persistence times of only 0.7 days, medium-sized birds 5.5 days and large birds persist on average 13 days. Heavily oiled birds are more persistent than other corpses. This and other complicating factors are further discussed in the context of the interpretation of beached bird numbers. We conclude that: (1) beached numbers can be useful and indicate interesting patterns if interpreted with care and in an international context; (2) there is a need for more drift-, persistence- and floating experiments in order to fully understand what a bird corpse does from the moment it dies until it is being found on the beach; (3) special attention should go to differences in ‘behaviour’ of oiled versus unoiled corpses and to what happens with corpses once they have sunk; (4) models should be developed to help predicting patterns of seabird strandings, not to replace drift experiments at incidents or mass strandings