Rates of decrease of polychlorinated biphenyl concentrations in five species of Lake Michigan salmonids

Dynamic linear models (DLM) were used to study time trends in annual averag e polychlorinated biphenyl (PCB) concentrations in five species of Lake Mic higan salmonids using data collected from 1972 to 1994 by both the Michigan Department of Natural Resources and the Wisconsin Department of Natural Re sources. DLMs use an adaptive fitting procedure to track changes over time in both the level (mean) of the series and the rate of increase or decline (growth rate), in contrast with other approaches that fit fixed parameters. We used DLMs to provide retrospective time series of estimates of rates of decline in PCB concentrations. Growth parameters indicate that PCB decline s have slowed more than first-order models fit in the mid-1980s would predi ct. Growth parameters for brown trout (Salmo trutta) and rainbow trout (Onc orhynchus mykiss) increased only slightly, indicating the most consistency with first-order dynamics. Coho (Oncorhynchus kisutch) and chinook salmon ( Oncorhynchus tshawytscha) showed a pattern of high rates of decline in the early to mid-1980s followed by a period of slower PCB concentration changes . The temporal pattern of rates of decline for lake trout (Salvelinus namay cush) stood apart from the other species, with a growth parameter that incr eased steadily during the entire period of record.