Walleye pollock in Shelikof Strait, Gulf of Alaska, spawn in a n area with strong interannual differences in the oceanic environment. Feeding conditions and mortality of walleye pollock larvae in Shelikof Strait were compared in 2 consecutive years of markedly contrasting oceanographic conditions. In 1990, winds were relatively calm, and a large eddy was formed in the lower portion of the strait; walleye pollock larvae were found concentrated in this eddy feature. In 1991, winds were very strong and sea surface temperatures were anon~alously cold Flow through the Shelikof region was strong in that year, and larvae were sparse. In 1990, copepod naupliar abundance was high throughout the study area. There were no geographic differences in feeding intensity of larvae, RNA content or larval growth. In 1991 the major differences occurred between ~n s h o r e and offshore stations. Con~paring conditions in 1990 and 1991, naupliar abundance, larval feeding intensity, RNA content and length-atage were all low in the stormy conditions of 1991 In 1991 estimated mortality was significantly higher than that measured in 1990, although part of the loss could have been due to strong advection out of the area. Survival of expatriated larvae is discussed in light of very low juvenile abundances in 1991. This study shows the dramatic effect of environmental conditions on early larval survival rates.

Fish samples collected at the cooling water intake screens of West Thurrock power station located 35.5 km downstream of London Bridge on the Thames estuary during the decade 1980–1989 were analysed. Seasonal and long‐term changes in the abundance of the 15 most numerous fish species and in several fish community parameters were analysed. The majority of species were highly seasonal in their distribution and abundance. Species diversity was lower in the summer (May–August) compared with spring/winter (October–February). Evidence of long‐term changes in species diversity and community structure over the decade is presented. The changes were consistent with a period of relative stability (1980–1984) followed by a period of change (1985–1989) and may reflect a deterioration in water quality in the second quinquennium. It is argued that the monitoring of fish communities in estuaries should be based on a multi‐metric approach as no single indicator alone can describe the complex community structure.