By the time the scientists investigating the origins of SARS-CoV-2 in the Chinese city of Wuhan convened on March 10 for a virtual press conference, more than a year had passed since the World Health Organization asked Beijing for permission to admit them.
During this time, as more than 2.6 million people died of COVID-19 and millions more suffered lasting effects of the illness, the mystery of the coronavirus' origins has loomed.
There is broad agreement that the coronavirus is part of a lineage of viruses found in horseshoe bats in South China.
The mystery centers on how the virus came to cause an outbreak in Wuhan, a thousand miles away, without leaving any trace of its journey.
That is essentially the question Peter Daszak, an expert on disease zoology and a member of the WHO team, attempted to address in the press conference.
Daszak, who is head of EcoHealth Alliance, a non-profit that funded research into bat coronaviruses in China and elsewhere, is a vocal critic of the notion that the pandemic's origin involved some kind of laboratory accident; instead, he favors natural zoonosis as an explanation, which holds that the pandemic jumped from animals to people without human intervention.
In guessing how this might have happened, Daszak suggested frozen food imports, and mentioned animal carcasses as a potential "conduit from Wuhan to the provinces in South China, where the closest relative viruses to [the coronavirus] are found in bats.
This was a telling admission for Daszak because it highlights a weakness in the random-zoonosis theory and points toward a more logical—and often overlooked—culprit.
Daszak and many other scientists have pointed out that, as a general rule, a new pandemic is more likely to be triggered by a random zoonosis than by a virus escaping a lab.
What happens generally, though, does not necessarily explain what happened in this particular instance.
The important question is, what is the most likely way that this particular virus could have caused this particular outbreak in Wuhan? As an explanation, random zoonosis does not seem to fit the facts. It has a vector problem.
For the virus to have jumped from a South China horseshoe bat to another animal, and then to have followed a chain of transmission through an unknown series of hosts, possibly as part of the wildlife trade, until it reached Wuhan, where it then exploded into the population and got noticed at last, requires a series of low-probability events.
The Wuhan Institute of Virology scientist at the heart of the controversy, Shi Zhengli, told Scientific American that she herself "never expected this kind of thing to happen in Wuhan."
She'd have expected it to happen in South China, where these viruses are naturally found.
In addition, South China has a much stronger tradition of wildlife consumption than other parts of China.
(One survey found that 83 percent of the residents of the South China megacity of Guangzhou had eaten wildlife in the past year, while only 5 percent of Beijing residents had.
The wildlife trade flows in that direction, and major wet markets are found there.
A South China virus would have been vastly more likely to trigger an outbreak in South China, which is exactly what happened with the first SARS in 2002.
With almost no data, we can't be certain of any theory.
We could concoct a theoretical game of hopscotch in which the virus randomly transmits from host to host, accumulating mutations along the way, threading the needle of cities and villages and South China wet markets, and landing in downtown Wuhan ready for its coming-out party.
It's not impossible. It's just getting-struck-by-lightning-while-being-eaten-by-a-shark unlikely.
And if it did happen, it should have been part of a noticeable pattern. In a long, meandering chain of transmissions, each node branching into new infections, Wuhan would have been just one twig of a full-fledged tree—the kind of tree epidemiologists use to trace viral evolution.
Even if most other branches eventually died out, the entire tree could not have disappeared without leaving a trace. And yet it doesn't exist.
Most of the scientific evidence points toward COVID-19 beginning in Wuhan in October or November 2019.
Despite a year of intense searching, not a single close ancestor of the virus has been found.
As a group of 26 scientists stated in a recent letter calling for a new, objective investigation into the origins of SARS-CoV-2,
"There is as yet no evidence demonstrating a fully natural origin of this virus." It's as if it just teleported from South China to Wuhan.
Daszak is hoping frozen food can explain this teleportation.
The theory is that wild animals being farmed in South China picked up the virus from bats, were killed, frozen, and shipped to Wuhan, and somehow managed to spread the virus through the environs of the market.
If the idea had merit, it would at least provide a possible direct line to Wuhan. Unfortunately, it's both bad science and bad math.
The theory has its roots in a case last fall in the Chinese port city of Qingdao, when two dock workers tested positive for COVID-19.
An exhaustive search found traces of the virus on frozen packages of cod; the Chinese government seized on the case as evidence that the virus could be imported in frozen food and may have started elsewhere.
The notion has been widely derided by experts, who point out that there has never been a documented case of the virus having been transmitted through frozen food.
The U.S. Food and Drug Administration says there is "no credible evidence" to support it.
The Chinese government has promoted it heavily, perhaps as a way of deflecting responsibility.
Under normal circumstances, it might be surprising for WHO scientists to peddle a discredited theory.
But the origins investigation is a joint exercise with China, which closely negotiated every step of the process, retained veto power over the selection of team members, and required that conclusions in the final report be a consensus between the Chinese and international delegations.
The WHO had limited options.
The science is also bad because it isn't based on any on-the-ground evidence. There were many cases of COVID-19 in Wuhan before the market outbreak.
In addition, the only live animals in the market seem to have been snakes, turtles, and frogs, none of whom are potential carriers, and they all tested negative.
Peter Daszak himself has admitted that all of the carcasses examined tested negative for SARS-CoV-2. Yes, you read that right: the theory is based on the fact that they didn't find any infected animals in Wuhan.
In any case, the director of China's Center for Disease Control dismissed the market theory last May, explaining that there was no connection and that the virus "existed long before" it was found in the market, which makes it rather extraordinary that the WHO team chose to spend its limited time poking around the empty building (which had been closed, cleared and disinfected more than a year ago).
In light of these challenging details, Daszak has floated the idea that perhaps another market in Wuhan that was selling frozen farmed wildlife triggered the outbreak, and has implied that some vendors in Wuhan were supplied from South China, but this still doesn't dig the WHO team out of its statistical hole.
Any South China farm whose animals were infected with the coronavirus would have contaminated its own workers, as well as other vendors it was supplying, and again, most of those would most likely have been in South Asia.
The only way the theory would make any sense (assuming it could be proved to be scientifically feasible) is if a farm in South China was exclusively supplying Wuhan with all its animals, which seems unlikely.
In any case, the WHO team's Chinese investigators pretty much scuttled Daszak's theory by announcing that they had tested tens of thousands of wild and farmed animals across China without finding a single trace of SARS-CoV-2.
Can the frozen-food theory be a credible explanation of the conduit conundrum? No.
But it's good the WHO brought it up, because it takes the investigation in an important direction: the need to find conduits that could have delivered the virus to Wuhan.
These conduits would need to meet the WHO's criteria: they'd have to be animals that make good hosts for the virus, they'd need to have had opportunities to interact with the bats known to carry these particular coronaviruses, and they'd need to have traveled directly to Wuhan from the rural areas of the region without ping-ponging all over South China.
That sounds like a tall order. What are the chances of finding documented proof of such vectors?
Remarkably, however, we actually have a wealth of evidence that fits this description perfectly: human beings—the best SARS-CoV-2 hosts of all—who had extremely close contact with horseshoe bats in South China.
We know they made repeated visits over many years to the exact caves where SARS-related viruses were found.
There, they handled bats directly, spent extensive time inside the caves breathing the air, and brought thousands of samples of guano, blood, and other bits of bat (possibly even live bats) back to Wuhan with them.
Those were researchers from the Wuhan Institute of Virology, who had been scouring South China caves for coronaviruses with the ability to infect humans.
They did this as part of a massive virus-hunting project—a collaboration between Peter Daszak's EcoHealth Alliance, the USAID's Predict program, a new organization called the Global Virome Project, and other groups—to track down the world's viruses in their lairs, bring them back to the lab, and study them.
They would pay special attention to the ones with the ability to infect humans, learning how they enter cells, how they mutate, how they jump from host to host, and how they escape the immune system and persist in individuals.
The project has been criticized by the world's top virologists as being useless at best and extremely dangerous at worst. COVID-19 certainly hasn't weighed in its favor:
Despite 15 years of coronavirus hunting and testing by the WIV, it was helpless to prevent a pandemic in its own backyard. If that's a coincidence, it's one of the great ironies of history.
Over that period, the WIV accumulated one of the largest collections of coronaviruses in the world.
According to the Washington Post, it contained the records of 22,000 samples, including the sequences for more than 100 coronaviruses known only to the WIV.
But we don't know exactly which viruses, because in the earliest days of the pandemic the WIV took the databases containing this information offline and has declined to share them.
(When asked about the database at the press conference, Daszak said, "We did not ask to see the data" but said he was familiar with it from his close collaboration with the WIV.)
From the research papers published in the scientific literature by WIV scientists, we know that the lab held many dangerous ones.
The WIV had spent years identifying the caves where the scary ones lurked. Over a four-year period, it visited one cave in Yunnan repeatedly and brought back 1,322 samples, including at least nine of SARS-CoV-2's close relatives.
Overall, the effort brought more than 15,000 bat samples back to Wuhan, which included more than 400 coronaviruses new to science and at least 50 of the variety that can infect people. And we know almost nothing about this cache.
We do know, from online photos, videos, and accounts, that the virus hunters didn't always wear adequate protective gear, and were bitten by bats with some regularity. And this is what they did while the cameras were rolling.
The WIV also had a reputation for shoddy safety. U.S. diplomats who evaluated the institute in 2017 were so alarmed that they sent an urgent cable back to Washington warning that the institute suffered from "a serious shortage of appropriately trained technicians and investigators needed to safely operate this high-containment laboratory."
In other words, you could not build a better conduit for channeling viruses straight from rural South China to downtown Wuhan.
Whether or not the virus made it into the lab and later escaped, this was its best transport option.
This viral monorail ran between a number of caves in South China and the WIV multiple times every year, with a different mix of students and scientists.
And it wasn't limited to the Wuhan Institute of Virology. The Wuhan CDC also extensively collected bat samples and has been accused of inadequate precautions.
One Wuhan University post about a sampling trip to Yunnan in the summer of 2019 (a few months before the pandemic started) shows researchers entering bat caves in street clothes and handling bats with their bare hands, a shocking violation of safety protocols.
The bats collected included horseshoe bats, the kind most likely to carry SARS-like diseases. If you're looking for potential carriers, there's no need for hypothetical frozen ferret badgers. Start with the motherlode of cases right in the scientific record.
Even in tightly controlled lab settings, viruses infect researchers all too often. The first SARS escaped labs in Singapore, Taiwan, and China six times, and it wasn't nearly as infectious as SARS-CoV-2.
As Alison Young has extensively documented in the USA Today, "lab accidents aren't rare."
Over a five-year period, U.S. labs reported more than 450 accidents involving dangerous pathogens—nearly one hundred per year.
Those include only the ones in the U.S. that were reported. As Harvard's Marc Lipsitch points out, lab accidents are often shrouded in secrecy. Worldwide, the number each year must be staggering.
In the chaotic environment of fieldwork, the risk of infection must be even larger. The close air of a bat cave is the ideal place to pick up a bat virus. Accidents are inevitable.
Collectors may have suffered flu-like symptoms (as has been documented for some WIV researchers, three of whom were hospitalized with pneumonia in the fall of 2019, exactly when the outbreak started) and shaken it off without much thought.
But considering how often COVID-19 is asymptomatic, they may never have known.
Technically, of course, this would count as a "natural origin," because the virus would have evolved naturally and jumped to humans of its own accord. (It's only "unnatural" if the virus was altered by scientists in the lab.)
This may even be one of the scenarios experts quietly have in mind when they carefully use phrases like that.
Obviously, the proximity was anything but natural, and there's the rub. Having seen up close how much damage one virus can do, we want to do everything in our power to reduce the chance of the next one.
That means choking off as many conduits as possible. Clamping down on the world wildlife trade (as China has already begun to do) is an obvious step that can be taken. Animal farming probably needs to be rethought in its entirety.
Lab experimentation on dangerous pathogens is another no-brainer, regardless of its role in this pandemic. There are thousands of labs around the planet doing this research, much of it of limited value.
And as Filippa Lentzos, one of the world's leading biosecurity experts, recently told the Washington Post, there is zero international regulation of their activities.
"There's no set international law that they have to follow. There's nobody checking what they're doing. There are no inspectors, no regulators. There's none of that."
The same is true of virus collection. That's a concern, because a titanic amount of it is planned for the coming years.
Proponents of the Global Virome Project want to collect more than a million of the world's unknown viruses over the next ten years, including hundreds of thousands with the potential to infect humans, and bring them back to labs for research.
If this project moves forward, the flow of unknown viruses from remote locations to population centers will be like nothing in history.
That will certainly increase our knowledge about these viruses and how they work, which is useful information. But it will also give many of them a crack at us for the first time.
Is it worth it? That discussion will be important in the coming years. Now that the WHO team has brought up the vector question, this would be an excellent time to begin.
Rigorous regulation of biolabs can certainly reduce civilization's exposure to emerging viruses, but merely regulating what happens within the labs will not be enough if we keep sending so many vulnerable collectors into the hottest of viral hotspots.