FIELD ESTIMATES OF NET TROPHIC TRANSFER OF PCBS FROM PREY FISHES TO LAKE-MICHIGAN SALMONIDS

Dietary exposure is the principal source of polychlorinated biphenlys (PCBs) to Great Lakes salmonids. PCB transfer efficiency has typically been determined with laboratory studies using high food PCB concentra tions and clean animals, making extrapolations to the field uncertain. Here, we present estimates of net PCB transfer efficiencies for Lake Michigan coho salmon, chinook salmon, and lake trout based on 15 years of PCB concentrations in these fish and their prey. Between 1975 and 1990, large changes in Lake Michigan's pelagic food web led to changes in coho and chinook salmon diets and chinook salmon growth efficiency . Early in this period, PCB transfer efficiencies increased in lake tr out but decreased in chinook and coho salmon. These early trends carry relatively high uncertainty due to high variability in prey and preda tor PCB concentrations and are probably not ecologically significant. After 1983, PCB transfer efficiencies remained relatively constant at similar to 0.55 for lake trout and at similar to 0.60 for chinook salm on and increased slightly to similar to 0.50 for coho salmon. PCB tran sfer efficiencies appear to be little affected by changes in prey PCB concentration, shifts in prey type, and shifts in predator gross growt h efficiency. Use of a constant PCB transfer efficiency in models that consider PCB accumulation by piscivorous fishes is therefore appropri ate.