New Scientist said it had obtained a progress report into a 750,000 pound ($1.1 million) British government study, which suggested "the possibility of a large scale effect that may be related to the use of chemicals on fish farms," although it said its results could be partly explained by "poor sampling".

Researchers found a drop in the number of nematode worms in the sediment close to the salmon farms in Loch Sunart and Loch Craignish on Scotland's Atlantic coast and an almost complete absence of copepods, tiny crustaceans that are an important source of food for young fish.

Experiments at another marine laboratory in western Scotland showed the chemicals either killed or deformed copepods at concentrations as low as 100 nanograms a litre, but so far they had been not been able to find the effect in the wild.

Scotland is one of the top four producers of farmed Atlantic salmon, along with Norway, Chile and Canada.

Critics of the industry have seized upon the findings of the report, suggesting salmon farming had much wider impact than previously realised.

Graham Shimmield, one of the researchers, was quoted as saying: "Laboratory experiments so far suggest that fish-farm chemicals may have an effect. The challenge now is to quantify that effect."