In September 2021, as the debate raged over whether the COVID-19 pandemic was sparked by infected animals at a food market in Wuhan, China, or a leak from a nearby lab, an anonymous post on the Chinese social medium WeChat floated a completely different, bizarre theory: SARS-CoV-2, the virus that causes COVID-19, had arrived in the country on frozen lobsters from Maine that were sold at the Wuhan market. The post was quickly dismissed by scientists and some considered it part of a Chinese effort to shift blame.
But a new analysis suggests the post may hold clues to the pandemic’s origin—and further evidence that China is withholding vital data on the contentious issue.
Evolutionary biologist Florence Débarre of CNRS, France’s national research agency, who firmly believes SARS-CoV-2 jumped into people from live animals at the market, calls the WeChat post “extremely elaborate disinformation.” But when she recently translated the post and compared the detailed maps of the market it contains with other, official ones, Débarre found surprises. The maps identify specific stalls as having live animals infected with SARS-CoV-2 and vendors with antibodies to the virus—data China has never shared.
If animals at the market were infected, it would bolster—but not prove—the natural origin theory, and the antibody evidence could help epidemiologists better chart the early spread of SARS-CoV-2. Débarre finds the unconfirmed data credible because other details on the maps have been verified since they were posted. “The maps contain information that was made public later, which indicates they come from an informed source,” Débarre says. And if they’re accurate, “then they are extremely important.”
Today, Débarre made her analysis public after sharing it with several colleagues and with Science. She has also sent a copy to World Health Organization (WHO) epidemiologist Maria Van Kerkhove, who oversaw an independent scientific advisory group that evaluated different SARS-CoV-2 origin hypotheses. (Its 2025 report concluded that “the weight of available evidence” supported zoonotic transmission of the virus—from an infected animal to a human—not a lab leak.) Van Kerkhove says Débarre’s analysis is “interesting,” but cautions that the provenance of the data is unclear. “The way in which the information is identified is highly suspect,” she says.
Others are more dismissive. Jesse Bloom, an evolutionary biologist at the Fred Hutchinson Cancer Center who remains unconvinced SARS-CoV-2 made a natural jump from animals to humans, says the WeChat post’s assertions are “absurd propaganda” and “deeply flawed.” He doesn’t put much stock in the maps either—or in Débarre’s analysis of them. “None of the details should be taken too seriously,” Bloom says.
Advertisement
The WeChat post, under the pseudonym Kunlun Sword, sets up its frozen lobster theory by arguing that people in Maine who came down with a severe lung disease associated with vaping in mid-2019 were in fact some of the earliest cases of COVID-19. It was the latest variation of the “frozen seafood” or “cold-chain” hypothesis, pushed by China since June 2020. The idea that the virus originated elsewhere in the world has never gained traction outside of Chinese studies and government decrees.
The WeChat post contained two maps of stalls at the Huanan Seafood Wholesale Market, which became world famous after Chinese authorities reported it was linked to a cluster of cases of a mysterious respiratory infection on 31 December 2019. They initially suspected that wildlife illegally sold there was the source of the virus, and the market was closed the next day.
The post’s assertion that Maine lobsters sold at the market were the culprit got widespread attention after a Chinese Consul General based in India tweeted about the theory. Five weeks later, NBC News reported that a University of Oxford disinformation researcher had identified more than 550 Twitter accounts hyping the story. Twitter—today called X—suspended the accounts, and the Maine lobsters joined Norwegian salmon, 5G cell towers, snakes, and international military games held in Wuhan on the trash heap of COVID-19 origin theories.
Prior to the post, only two maps of the marketplace had become public. One was in a report from a WHO-organized probe of COVID-19’s origins by a team of Chinese and foreign scientists who in early 2020 examined available data and visited the market and nearby labs that worked on related viruses. The other map appeared in a media report that included a leaked document from the Chinese Center for Disease Control and Prevention (China CDC). Neither mentioned evidence that SARS-CoV-2 had infected animals at the market; in fact, the WHO report said tests conducted by the Chinese of 188 animals collected from Huanan and other Wuhan markets after the outbreak surfaced all came back negative.
As Débarre documents, three additional market maps have appeared after the 2021 post: an updated version of the map in the WHO report and two in studies done by China CDC and other Chinese institutions. None of the maps ever reported that people at the market had been tested for antibodies—which can indicate someone was infected in the past, even if the virus is no longer present—or positive viral samples from animals. (Testing people for antibodies in outbreak investigations is standard epidemiologic practice, notes Van Kerkhove, who used to do this type of fieldwork herself.)
One of the WeChat maps pinpointed the stall where a seafood merchant sold lobsters from Maine but provided a lot of other information, perhaps inadvertently. It mapped four stalls with the virus confirmed in humans and two in the environment (on floors and in sewers, for example) that were not in the WHO or China CDC reports but appeared in subsequent maps—a sign, Débarre says, that whoever published the maps had detailed information. More provocatively, the maps indicate that 17 merchants had SARS-CoV-2 antibodies. (All of them were named on the map, and Débarre has been able to verify that some worked at the market by checking a business registry.)
The maps also showed that 18 animal samples taken at one stall, number 6-29, tested positive for the virus itself. In 2023, Débarre and colleagues had fingered the exact same stall as a prime suspect in sparking the market outbreak, based on genomic data from Chinese scientists that Débarre had stumbled on in GISAID, a pathogen sequence database. Those data showed that samples from 6-29 had a mix of SARS-CoV-2 genetic material and DNA from mammals known to be susceptible to the virus, including raccoon dogs and civets. This suggested, but did not prove, there were infected animals at the market. The WeChat post unequivocally states there were.
Even if the WeChat maps are accurate, there’s no way to determine whether the virus traveled from animals to humans or in the other direction, as some Chinese researchers have suggested. But the maps do add to earlier evidence that China hasn’t released everything it knows about the market.
Some of the scientists involved in the origin debate take Débarre’s findings more seriously than Bloom. “It’s a perfect mix of a little loopy but definitely something to it, including the fact that there clearly must be more data from the market that hasn’t been shared,” says evolutionary biologist Kristian Andersen of Scripps Research, who co-authored the 2023 report with Débarre.
Edward Holmes, an evolutionary biologist at the University of Sydney and another co-author of the 2023 report, agrees with Andersen and Van Kerkhove that the new analysis is interesting and applauds Débarre’s ongoing detective work. “She has an attention to detail beyond most humans,” he says. But he doubts it will move the debate toward closure. “It keeps the story going, but it’s impossible to test validity,” he says. Without China releasing more information, “it’s pointless going on about it.”
Van Kerkhove is convinced China is still withholding data related to the pandemic’s origin. “There are people out there who have more information: governments, scientists, professionals,” she says. “WHO has been crystal clear about asking for any and all with information to come forward so that a proper and complete analysis can be conducted.” But social media posts won’t answer the origin riddle, Van Kerkhove says: “They might provide clues, but this is no game.”