| Summary: | Short-tailed Shearwaters (Puffinus tenuirostris) migrate between breeding areas in Australia and wintering areas in the Bering Sea. These extreme movements allow them to feed on swarms of euphausiids (krill) that occur seasonally in different regions, but they occasionally experience die-offs when availability of euphausiids or other prey is inadequate. During a coccolithophore bloom in the Bering Sea in 1997, hundreds of thousands of Short-tailed Shearwaters starved to death. One proposed explanation was that the calcareous shells of phytoplanktonic coccolithophores reduced light transmission, thus impairing visual foraging underwater. This hypothesis assumes that shearwaters feed entirely by vision (bite-feeding), but their unique bill and tongue morphology might allow nonvisual filter-feeding within euphausiid swarms. To investigate these issues, we developed simulation models of Short-tailed Shearwaters bite-feeding and filter-feeding underwater on the euphausiid Thysanoessa raschii. The visual (bite-feeding) model considered profiles of diffuse and beam attenuation of light in the Bering Sea among seasons, sites, and years with varying influence by diatom and coccolithophore blooms. The visual model indicated that over the huge range of densities in euphausiid swarms (tens to tens of thousands per cubic meter), neither light level nor prey density had appreciable effects on intake rate; instead, intake was severely limited by capture time and capture probability after prey were detected. Thus, for shearwaters there are strong advantages of feeding on dense swarms near the surface, where dive costs are low relative to fixed intake rate, and intake might be increased by filter-feeding. With minimal effects of light conditions, starvation of shearwaters during the coccolithophore bloom probably did not result from reduced visibility underwater after prey patches were found. Alternatively; turbidity from coccolithophores might have hindered detection of euphausiid swarms from the air.
|
|---|