Data gathered over 77 lake-years from 4 coastal British Columbia sockeye salmon nursery lakes suggest that
density-dependent growth reductions are only possible at exceptionally high fry densities. In Great Central Lake (n=33 years) and Sproat Lakes (n=30 years) there was no relationship between smolt weight and mean summer fry
densities ranging from 760-3800 fry ha-1. However, in two years when Sproat Lake fry densities were unusually high (1983=5183 ha-1, 1996=4801 ha-1) smolt weights were among the lowest recorded. In Woss and Vernon lakes (n=14
lake-years), there were significant bottom-up relationships between fall-fry weights and zooplankton biomass, but no relationships between December-fry weights and average fry densities (range 331-1361 ha-1), nor were there significant
top-down relationships between fry densities and average zooplankton biomass. Comparisons of zooplankton production with bioenergetic-based fry consumption, suggested that the carrying capacity for Vernon Lake which had the highest rate
of zooplankton production, was 12,700 fry ha-1, and for Sproat Lake which had the lowest zooplankton production, was 5200 fry ha-1. We conclude that fry densities in the range commonly observed for British Columbia coastal lakes (i.e.
500-4000 ha-1) cannot cause density-dependent reductions in prey biomass or fry growth rates. Further research is necessary.