Feeding of larval walleye pollock was examined with respect to density and distribution of micro-and mesozooplanktonic prey (<500 pm) at 6 depths in the oceanic domain of the Bering Sea to determine if first-feeding larvae select among available prey and to assess their foraging environment in comparison to other locations where walleye pollock spawn. At 30 m depth, walleye pollock larvae and copepod nauplii occurred at maximum densities of 106.6 ind. 100 m-3 and 26.3 ind. 1-', respectively. First-feeding walleye pollock larvae (3.5 to 6.5 mm) fed exclusively on copepod nauplii and copepod eggs. Among copepod nauplii, larvae selected for Metridia sp. and Microcalanus sp. but against Oithona similis, even though the latter was the most abundant prey taxon in the study area. Of the nauplii ingested by larval walleye pollock, there was evidence of selection for larger nauplii within Metridia sp., Microcalanus sp., and 0. sirnilis. In addition, larvae preferred Stages I11 to V of calanoid nauplii Larvae at 30 m depth had the highest incidence of feeding (97.5%) and number of prey items (4.6 larva-') ingested. Although the 30 m depth stratum prov~ded best physical and foraging conditions, the overall low percentage of feeding larvae and low numbers of prey consumed suggest that foraging conditions for larval walleye pollock at the time of sampling were below saturation feeding levels.

We tested the hypothesis that growth, condition, and mortality of four "trace" cohorts of Pacific herring, Clupea pallasi, larvae in Auke Bay, Alaska, in 1988 were limited by concentration of prey. Hatching of herring larvae coincided with the spring maximum of copepod nauplii production, so prey concentration was high enough to support successful feeding and growth. However, growth and condition differed significantly among cohorts. Low growth and condition were not caused by density-dependent competition for food among herring larvae because larvae were too dilute to affect prey concentration and were unlikely to be caused by competition within the entire plankton community because there is no evidence that nonherring plankters interfered with feeding of herring larvae. Most larvae with low condition were from two cohorts that hatched from the upper zone, which suggests that desiccation and extremes of temperature may have disrupted embryonic development and produced larvae that were not competent to feed successfully. There were no among-cohort differences in instantaneous mortality. We conclude that growth, condition, and mortality of herring larvae in Auke Bay in 1988 were not limited by food, but by other factors such as predation and environmental conditions during egg incubation.

Fish larvae were collected at 6 depths in Auke Bay, southeastern Alaska, USA, on 4 days in May and June 1987. On 3 dates samples were collected between 08:OO and 13:00 h, on the other date samples were collected every 4 h during 24 h. Vertical distributions of light, temperature, salinity, chlorophyll and copepod nauplii were also measured. Wind speed data were available from a nearby station. Smelt (Osmeridae) larvae, the most abundant larval fish taxon, migrated to the surface at midnight, whereas walleye pollock, flathead sole, rock sole, poacher (Agonidae) and northern smoothtongue moved deeper at dusk. During the day, most fish larvae were concentrated from 5 to 15 m, broadly overlapping the highest abundances of copepod nauplii, although poacher and northern smoothtongue were consistently deeper (15 to 25 m ) . Diurnally, smelt larvae were shallow (5 to 10 m ) , and pycnocllne depth accounted for more variabhty in their mean depth than any other biotic or physical vanable. Mean depths of the other 5 most abundant taxa were in the same rank order by depth in all 4 sets of daytime samples, and depth of isotherms below the pycnocline accounted for more vanation in their mean depths than did any other variable. The diurnal depth distributions we observed apparently result from temperature preferences among marine larval fishes, with descent at dusk as the most common form of diel vertical migration