Catch of the Day: Transgenic Salmon? Not So Fast

Today marks the deadline for public comments on a genetically modified salmon currently under review by the Food and Drug Administration. If approved, the fish will be the first transgenic animal ever to enter the human food supply. Some say it's about time. In an op-ed that appeared last month in the New York Times under the title "Don't Be Afraid of Genetic Modification," science writer Emily Anthes explained that the company behind the fish, Massachusetts-based AquaBounty Technologies, has been waiting more than 17 years for approval of its product.

Anthes, author of the just-published Frankenstein's Cat: Cuddling up to Biotech's Brave New Beasts, called the FDA's extension of the comment period on its environmental assessment (the original deadline was back in February) "just one more delay in a process that's dragged on far too long." The tests were in, she suggested, and the diagnosis an obvious one. Couldn't we please put this poor company out of its misery and give its salmon the green light already?

Not so fast. The FDA says the fish is just as safe to eat as conventional salmon -- though you can count me among those who don't look to that troubled agency (see "The FDA is Out to Lunch," Winter 2013) for the last word on food safety. But the potential threats to the oceans are significant, and the FDA doesn't acknowledge them. More troublingly, the agency seems content to leapfrog over the entire process for determining whether those threats exist in the first place. Given the far-reaching impact this salmon could have on our natural ecosystems and global seafood supply, fish scientists and food advocates make a strong case for closer scrutiny.

AquaBounty has created its AquAdvantage fish by modifying farmed Atlantic salmon with a growth-hormone gene from the Chinook salmon and a piece of DNA from the ocean pout. The two work together to allow the fish to grow year-round, cutting down on feed costs and on the time it takes to reach marketable size. AquaBounty says the fish pose "no threat to wild salmon populations." After producing the eggs on Prince Edward Island, in Canada, the company grows the fish to size at an inland facility in Panama, where AquaBounty has a "confinement system" in place involving multiple physical barriers. In addition, the company says that it is breeding these salmon to be sterile, so that even if they were to escape, they would presumably be unable to take hold in any ecosystem or reproduce with any native salmon populations.

It's possible that AquaBounty's various safeguards could succeed in sequestering its GM fish in the highly specific, highly limited scenario it's describing -- but that's ultimately beside the point. Ronald Stotish, the company's CEO, has said that the AquAdvantage salmon are designed for "facilities that can be built closer to consumers" and that the fish "can help us to feed an overpopulated planet." In other words, the company's plan is to scale up. (How would AquaBounty -- which has already danced on the brink of bankruptcy -- get a return on its multi-million-dollar investment otherwise?)

But if fish farmers everywhere will be able to purchase the company's eggs, they'll be growing the salmon in very different environments, and it's hard to imagine that many will have the capacity to put in place the rigorous and costly control system that AquaBounty has installed in Panama. (The FDA can extend an original approval to cover new manufacturing facilities with little public input or environmental review.) Nor is it a leap to imagine that continued expansion will eventually lead to conventional open net-pens, the standard system for the global salmon aquaculture industry, and, significantly, a form of confinement from which millions of farmed salmon have escaped in the past.

AquaBounty acknowledges that its sterilizing procedure isn't 100 percent effective. Even if its failure rate is as low as 0.1 percent, when you're talking about massive farms - and it's not uncommon for today's operations to entail a million fish -- huge numbers of fertile fish could escape and interbreed with the already endangered Atlantic salmon, compromising its fitness and threatening its ultimate survival.

Standard risk assessment generally involves three steps. The first step is to identify the hazard; the second, to determine the environmental consequences should the hazard be realized; and the third, to figure out how best to manage the risk and prevent those consequences from emerging. But the environmental assessment released by the FDA back in December essentially neglected the first two of these steps. It included no in-depth analysis of hazardous weaknesses in AquaBounty's salmon-confinement system and no examination of the consequences should the company's GM salmon escape. Also absent was a formal "uncertainty analysis," which -- given the inevitable unknowns involved with the natural world  (remember when we thought it was a good idea to import Asian carp?) -- is particularly critical in assessments related to the environment. Anne R. Kapuscinski, an environmental studies professor at Dartmouth College and the lead writer of a book about risk-assessment science as applied to genetically modified fish, has called the FDA's environmental assessment "not even close to approaching the standards of state-of-the-art risk assessment."

Part of the problem may be the regulatory framework itself. Currently, applications for transgenic animals are being considered by the FDA under the category of "new animal drugs" (go figure). And while the application process does take into account the potential effects of such "drugs" on human and animal health, it doesn't allow for any thorough consideration of environmental hazards. Nor does it require substantive input from other agencies -- like, for example, the National Oceanic and Atmospheric Administration, the National Marine Fisheries Service, the U.S. Fish and Wildlife Service, or the Environmental Protection Agency -- that have actual expertise in areas such as fisheries and ecological risk.

Even more outrageous, because of concerns over trade secrets, the process mostly takes place in the dark. The public-comments period under the National Environmental Policy Act, for example, takes effect only after a New Animal Drug Application has been approved by the FDA, turning it into something that feels very much like a formality. (AquaBounty, as the first GM-animal applicant, gave the FDA permission to make public some information about its salmon; future applications would carry no such requirements.)

It is precisely because this environmental assessment (by virtue of being the first) will set the precedent for all transgenic animals moving forward -- both in this country and internationally -- that it's important to get it right. Rather than relying on the shoddy version released in December, the agency could complete a new, and comprehensive, environmental impact statement that fully assesses the potential risks associated with broad adoption of these GM fish and others like them. Agencies with deep knowledge of fisheries and the environment could play a much larger role in determining the threats. Until the best available science can affirmatively demonstrate that GM fish pose little or no risk to our wild fish populations, it would be careless not to slam the brakes on their cultivation.

In her Times op-ed, Anthes wrote that the rejection of AquaBounty's salmon application would be "just fine" with groups that "traffic in scare tactics rather than science." In fact, among those calling for a rejection, sound science is all they want.

This story was originally published by OnEarth.