FISH MERCURY LEVELS IN RELATION TO CHARACTERISTICS OF HYDROELECTRIC RESERVOIRS IN NEWFOUNDLAND, CANADA

Mercury levels in fish have been demonstrated to increase after impoun dment with augmented levels of mercury predicted to decline as the res ervoir ages. Previous research in Newfoundland predicted return rates in the order of 10 to 12 years for landlocked Atlantic salmon or ouana niche (Salmo salar) and 7 years for brook trout (Salvelinus fontinalis ). In order to test the validity of these predictions on a broader spa tial and temporal scale, and develop more generally predictive 'models ', mercury levels in three fish species were studied in 16 older Newfo undland hydroelectric reservoirs of various age (32 to 95 years) and a rea flooded (21 to 13,000 ha). Mercury concentrations were standardize d to fish length and correlated with physical, chemical, and biologica l characteristics of the sampling sites. Standard length mercury level s ranged from 0.23 to 0.86 ppm in ouananiche, 0.13 to 0.59 ppm in broo k trout, and 0.22 to 0.72 in arctic charr (Salvelinus alpinus). Fish i n excess of the Canadian Safety Limit (0.5 ppm) were collected from 14 of 16 sites for ouananiche, 8 of 17 sites for brook trout, and 3 of 7 sites for arctic charr, including control lakes. Standard length fish mercury levels were correlated with reservoir age and (log(10)) area flooded for ouananiche and with pH for arctic charr. A multiple regres sion model was developed relating standard length mercury in ouananich e with reservoir age and log(10) of the flooded area. There were no ap parent relationship between reservoir characteristics and brook trout mercury concentrations. Based on this analysis, it is not possible, at present, to develop generally predictive models for all species found in Newfoundland impoundments.