Generations of anglers fishing Scotland's rivers could have ruined the DNA of wild salmon by killing so many that the gene pool cannot recover, according to a controversial new theory put forward by an advocate of fish-farming.

Angling and fishery groups have long argued that salmon escaped from fish farms have been interbreeding with wild salmon, weakening their DNA.

This, they say, threatens the wild population's ability to survive while at sea and their ability to return to the rivers they came from, changing thousands of years of behaviour.

But fish-farming expert and outspoken supporter of the industry Dr Martin Jaffa, of Manchester-based consultants CallanderMcDowell, has put forward another theory in his submission to the Scottish Government's aquaculture and fisheries bill consultation.

The consultation is a bid to get better management of the fish-farming industry and the wild fish and angling interests, which are frequently at odds but are estimated to be worth more than £650million a year to Scotland.

Dr Jaffa says evidence of widespread interbreeding between wild and farmed salmon is inconclusive, and says instead that anglers themselves have caused damage to the wild salmon's survival chances.

He says the fact that there are different salmon populations between and within different rivers has long been attributed to thousands of years of natural selection that adapts the salmon to the specific conditions of the rivers, and the sections of the rivers, in which they breed.

However, a Scottish Government funded study, FASMOP (Focusing Atlantic Salmon Management on Populations Project), currently under way to produce a genetic map of wild salmon populations in Scotland's rivers, shows that the genetic differences identified between fish from various locations in the rivers Dee, Annan and Deveron were found to be very weak.

Dr Jaffa said the findings could indicate wild salmon have not evolved into such distinct populations in these rivers as expected, and he suspects a form of evolution called genetic drift has cut genetic differences.

He said: "Genetic drift is caused by random events that occur by chance. As a result, some individuals have a greater impact on the population than might be expected.

"This is because other individuals might die suddenly and unexpectedly and cannot contribute to the 'pool of genes' of the whole population. In wild salmon populations, such a sudden and random loss might be associated with the catching and killing of salmon by anglers over the last 150 years. Over the last 10 years about 350,000 potential breeding salmon from Scotland's rivers have been killed by rod anglers.

"This number would have been much higher had it not been for the recent introduction of a catch-and-release policy on many rivers."

But Nick Chisholm, Director of the River Annan Fisheries Board and Trust, said he found Dr Jaffa's thesis "quite bizarre".

He said it was true the FASMOP study had found few differences between salmon, but it was thought this was due to the methodology employed.

"The way we describe it is the difference between having a microscope on 10 times magnification rather than 100, and perhaps it should have been on at 100 times," he said.

He added there was no question that escaped farmed salmon would breed with wild salmon.

"A number of years ago some eminent scientists in Ireland did some work looking at the survival of fish-farm-bred fish in the wild and they discovered their survival rate - was a couple of orders of magnitude below that of wild fish," he said.

"So if farmed fish get into a river and they breed with wild fish, you introduce these survival traits."

He said there was evidence this had happened in Norway, with survival being reduced not in fresh water but at sea. Crucially this would have a huge impact on the number of adult salmon returning to the rivers.