Recently hatched fry of brown trout were exposed to each of eight trace metals (Al, Cd, Cu, Fe, Mn, Ni, Pb, Zn) and to a mixture of all eight, at concentrations typical of soft acid waters, in flowing artificial soft water medium at pH 4.5 and 6.5, for 30 days. At pH 4.5, in the absence of trace metals, net uptake of Ca, Na and K and skeletal calcification were impaired but mortalities were low (10Y0). At pH 4.5, in the presence of Al, Cu, Pb and Zn individually and the mixture, mortalities were high (87-100%). Cd, Fe, Mn and Ni at pH 4.5 each caused some mortalities (1 7 4 3 % ) and impaired skeletal calcification. At pH 6.5, each of Al, Cd, Cu, Fe, Mn, Pb and the mixture of metals impaired net Ca uptake and (except Al, Fe and Pb) skeletal calcification. Net K uptake was impaired by Al and by the mixture, and less severely by Cu and by Fe. Net Na uptake was impaired by A1 and by the mixture, and less severely by Fe and by Mn. The role of trace metals other than A1 in fisheries' decline in soft acid waters is discussed.

Exposure of brown trout, Safmo frurfa, to zinc under continuous flow conditions over 96 h showed that both water hardness and pH exert major influences on the toxicity of the metal. 96-h LC50 values for total zinc ranged from cO.14mg I-' in alkaline soft water (PH 8; lOmg I-' as CaCO,) to 3.20 mg I-' in acidic hard water @H 5; 204 mg I-' as CaCOJ. A variable reduction in zinc toxicity in hard water compared with soft water over the pH range 4 9 was attributed to high external calcium. Zinc toxicity was positively correlated with decreasing acidity over the pH range 5-7, the metal being most toxic at pH 8-9 where metal complexes predominate. Below pH 5 metal toxicity also increased, irrespective of hardness. Water hardness and pH interacted with zinc toxicity in a complex manner, apparently dependent on physical and chemical transformations of the metal, and as changes in uptake, detoxification and excretion by the fish.

The effluent cause of a noninfectious hyperbilirubinaemia. or jaundice, in wild Don salmon was determined by exposing adult captive North Esk fish to environmentally relevant single and combined exposures of Donside discharges with control river water in situ. Adult North Esk salmon were chosen as test animals because they were physiologically pertinent, of different genetic origin or stock to Don fish, from a 'clean' river and successfully used in previous in situ riverine experiments.Experiments were conducted using a control group with 3 test groups of salmon exposed to individual or sequentially-combined industrial stream and untreated paper mill effluent for a test period of 4 wk. Sequential exposure of industrial stream followed by paper mill effluent produced haematological 'clinical profiles', i.e. a hyperbilirubinaemia in test North Esk salmon that was indistinguishable from wild Don fish. A combination of 2 types of effluent therefore appeared to be responsible for the noninfectious hyperbilirubinaemia reported in adult wild Don salmon during the previous decade. Chemical characterisation of the isolated effluents and the river are discussed with respect to known haemolytic agents from the tox~cological l~terature.

The riverine location of haematotoxic discharges was determined by exposing adult captive Atlantic salmon (Salmo salar L.) to River Don water in situ. Experiments were conducted using a control group and 4 test groups at different distances downstream of a paper mill and field drain discharges. Significant changes in whole blood haemoglobin, plasma cholesterol and bilirubin in salmon from polluted sites indicated the induction of a marked bilirubinaemia in test fish. The most marked pathology occurred in salmon exposed at a site below all the major effluent inputs into the river and suggested additivekynergistic effects between discharges. Pathophysiological changes in captive test fish from the polluted area closely resembled those recorded in wild Don salmon suffering from a non-infectious hyperbilirubinaemia. The effluent cause of the pigmented salmon syndrome in wild Don salmon appeared to be located between the reference site and a point 14 km downstream in the River Don.