Anglers noticed it first. In 2012, fishermen working Lake Memphremagog, the 32-mile water that straddles northern Vermont and Quebec, began reporting brown bullhead catfish marked with raised black patches. By 2014, close to one in three fish in the lake carried them.
"It looks like black tar on the skin - not very pretty," said Peter Emerson, a fish biologist with the Vermont Fish & Wildlife Department who has worked the lake's annual fish census since 2009. "You can tell it doesn't help the fish cause they're inside the mouth, they're on the fins." Sometimes the marks covered the lips, the gills, even the eyes. "They arrived very suddenly," Emerson told NPR.
A paper published July 22 in the journal Nature has now explained them, and the answer is stranger than anyone on the boat expected. The lesions are melanoma - and the cancer is transmissible, passing from fish to fish as an infectious agent in its own right. It is the first transmissible cancer identified in any fish, and the first in any freshwater species.
The early theories all failed. Because brown bullhead are used as indicators of environmental quality, the Vermont team assumed a pollutant or a virus was behind the tumours. The Coventry Landfill sits near the lake, which made contamination the obvious suspect.
"But it turns out, as we did the studies, they don't have much of that at all," Emerson said.
That is when Julie Dragon, a scientist at the University of Vermont and co-leader of the study, put the tumours through whole-genome sequencing. A tumour normally resembles the host it grew in. These did not.
"These tumors were very different from the host fish that they came from," Dragon said, "which is not what we expected or what anyone would expect." Instead the tumours closely resembled each other - clones of a cancer that arose in one catfish at some point in the past and has been moving between fish ever since. "It is the infectious agent itself," she said. "Those cells can somehow escape the host, maybe live in the environment - the aquatic environment - and then infect a new host."
How it spreads is still unsettled, and the researchers disagree on emphasis rather than substance. Mark Henderson, a fish biologist in UVM's Rubenstein School and a study co-leader, points to the breeding season. "It seems to only happen in larger fish that are of spawning age," he said. "Maybe some part of spawning behavior leads to the spreading of the cancer between animals." Dragon has a more specific mechanism in mind: bullhead crowd into shallow clusters to spawn and carry sharp spines behind their pectoral fins. "We think that they actually scratch accidentally each other, and that is one potential avenue for the cancer cells to invade a new host."
The landfill theory has largely been ruled out. The cancer is spread evenly along the whole 32-mile lake rather than clustered in South Bay near the landfill, and water tests came back comparable to nearby waters including Lake Champlain, which has no known transmissible melanoma. What the team did find was arsenic - naturally abundant in New England soils - at elevated levels in the cancer cells themselves, tracking loosely with where the disease turns up.
The disease is not staying put. "We don't think it started in Lake Memphremagog, and we have already confirmed that it does exist in lakes in New Hampshire and Maine," Dragon told The Canadian Press.
In May, New Brunswick angler Scott Young - who has seen spotted bullheads across four decades of fishing the province - took University of Cambridge researcher Zoe Clarke out on his boat to collect samples. Clarke, a postdoctoral researcher with Cambridge's Transmissible Cancer Group, expects sequencing results within weeks and believes they will match the Memphremagog lineage.
"How the heck did this happen? Why is it here?" Clarke said. "That's the kind of thing that we're going to be devoting a lot of our energy to over the next few years." She plans to return next year to sample the Saint John River.
For anglers, the practical guidance is narrow. Dragon says there is no evidence the cancer can spread to humans, and Memphremagog supplies drinking water to more than 175,000 people with no known risk. But the US Fish and Wildlife Service does not recommend eating fish with obvious lesions or growths, and Henderson put it more bluntly: "And, personally, I wouldn't eat the ones with tumors."
The population effect is the open question. Tasmanian devil facial tumour disease has wiped out 90 percent of some populations; the canine version has coexisted with its hosts for thousands of years. Heavily diseased bullhead in Memphremagog survive for years. Roughly 30 percent of the lake's brown bullhead are affected on both the US and Canadian sides.
Vincent Lynch, an evolutionary biologist at the University at Buffalo who was not involved in the work, framed what makes it remarkable. "Normally, cancers are sort of a dead end," he said. "When the host dies - when the person with the cancer dies - so does the cancer." Escaping into a new host changes that arithmetic. "The cancer wants to survive. Now it's got a life of its own."
Dragon suspects the discovery says more about what science has been looking for than about how rare the phenomenon is. "Our contention is that maybe it's not as rare as everyone thinks. We're just not seeing it," she said. "In the case of transmissible cancers, it may be that we don't see many simply because we aren't looking for them."
There is a footnote that suggests she may be right. In July 1852, Henry David Thoreau wrote of a catfish on the Concord River in Massachusetts with "a very large velvet-black spot, which included the right pectoral fin; a kind of disease I have often observed on them." The team is now sampling Massachusetts ponds and rivers to see whether the cancer started there.