A FIERCE debate still rages about the effects of releasing various genetically modified organisms. But most scientists and campaigners agree on one thing: GM fish could create havoc if they escaped and interbred with their wild cousins, or outcompeted native species.
So US company Aqua Bounty, which is eager to start selling millions of GM Atlantic salmon to fish farmers around the world, plans to make them all sterile females. The question facing regulators is, will that be enough?
The company’s AquAdvantage Bred salmon have an extra gene for a growth hormone, making them grow up to six times as fast as normal, though adults are no larger. Other groups in the US, Australia, Cuba and China are also creating fast-growing super fish. There are also plans to make fish more disease-resistant, more nutritious and able to withstand icy waters without freezing.
Advertisement
If the potential benefits – at least for producers – are great, so are the risks. Last week, Britain’s Agriculture and Environment Biotechnology Commission (AEBC), which advises the government, called for a complete ban on GM fish farming in pens open to natural waterways until there are “watertight” technologies for preventing fish from escaping and breeding.
Such escapes are inevitable if Aqua Bounty’s salmon are grown in nets floating in coastal waters, like most farmed salmon. Hundreds of thousands of these get away each year.
Critics are also concerned about the leakiness of US regulations. The Food and Drug Administration is the main agency involved in approving AquAdvantage fish for sale, on the basis that food quality could be affected. But since the added gene comes from the Chinook salmon, few are worried about food safety.
“The environmental concerns here are greater than food safety concerns,” says Rebecca Goldburg, a senior scientist at the Environmental Defense Fund. “And I’m concerned the FDA doesn’t have the framework in place to evaluate this on environmental grounds.”
Advocates of the technology have long argued that the damage such fish could cause would be limited. Transgenic animals tend to be less hardy, they argue, so natural selection would weed out them and their descendants if they escaped.
While some GM fish are less hardy, though, they can be better at reproducing. According to the “Trojan gene” theory of William Muir and Richard Howard at Purdue University in Indiana, a population of GM fish could drive a natural population a thousand times larger to extinction in just 40 generations
(91av, 4 December 1999, p 4).
Even making fish sterile doesn’t solve all the problems. Sterile males can still lure females into unproductive matings that could threaten endangered species.
Aqua Bounty wants to use two types of biological controls to tackle these concerns. First, all their fish will be produced by mating females with other females masculinised by exposure to hormones. These fish can produce sperm, but all their offspring are female.
The freshly fertilised eggs will then be subjected to 20 times atmospheric pressure, which stops them from ejecting an extra set of chromosomes. The resulting fish are “triploid”: they have three sets of chromosomes instead of two. “That makes the fish sterile,” says Aqua Bounty vice-president Joe McGonigle. Triploid female salmon have highly atrophied ovaries and, unlike triploid males, don’t attempt to mate.
The process of making only females is considered infallible. And McGonigle argues that the pressure technique for triploidisation has proved its worth in the lab. “In our experiments with Atlantic salmon the process is 100 per cent,” he says. “But I’m not asking the FDA to take my word on this.” McGonigle says his company will verify the technique in larger batches and regularly sample fish to assure their sterility.
But Muir isn’t convinced. “They haven’t looked at numbers high enough to make the bold statement that their method is foolproof,” he points out. “Every process is imperfect and occasionally fails. I don’t see why this should be different.”
One way to guarantee that all the fish are sterile would be to check every single one. It sounds impractical, but it is already done in Florida, where triploid grass carp, voracious vegetarians from Asia, are released to control exotic water weeds. A blood sample from each fish is tested.
McGonigle agrees that this would be possible, but he questions the reasonableness of testing every single fish. “Every industry does quality control by sampling,” he says. “If it is good enough for producing drugs for humans, why wouldn’t it be good enough for producing fish?”
Adopting universal testing would make it difficult for Aqua Bounty to sell fertilised eggs, since these cannot be tested without destroying them. Yet in US states where universal testing of grass carp isn’t enforced, experts believe a few fertile fish have slipped through and begun breeding.
There are other worries, too. It is well known that some triploid oysters revert to their normal diploid state. And there have been reports of some “triploid” fish that are actually mosaics. While their blood may be triploid, their genitals may not be. Neither phenomenon has been reported in salmon, but the issues haven’t been extensively studied.
Then there’s the possibility of human error, or just sheer stupidity. In Washington state and Oregon, for example, some breeding stations for grass carp were built in flood-prone regions, allowing fertile grass carp to escape during floods.
With such lingering doubts about triploidisation, the issue for regulators is whether even a low risk is acceptable. “Once the [fish] has escaped, there’s virtually nothing that can be done to recall it,” points out Malcolm Grant, the chairman of the AEBC.
Thomas Chen of the University of Connecticut in Storrs, who works on GM crayfish and other species, thinks such animals should not only be made sterile, but also kept in a closed system, with a gauntlet of barriers separating them from any natural waterways. “There just hasn’t been enough study of what happens when any transgenic mixes with a natural system to rely on one technique,” he says.
In the past, such land-based systems have never been feasible for salmon farming. They cost about a third more to build and run. McGonigle thinks there is only an argument for them where native species could interbreed with GM salmon, such as in the US and Canada. “[But] that would make it hard to compete with cheaper ocean pens in Chile, Tasmania and Argentina,” he says. “North America would be out of the business and we wouldn’t want to do that.”
Still, ocean pens are themselves controversial, because of the streams of waste they produce. Land-based systems using recirculated water would allow farms to optimise water conditions for growth, and reduce losses from predation and disease. By lowering costs, transgenic fish could help make such systems viable.
“That transgenic fish could end up moving salmon farms out of coastal waters is quite intriguing,” says Goldburg of the Environmental Defense Fund. “In that case, my opinion of the technology may begin to change.”
