The great Alaska earthquake lasted four minutes and 38 seconds when it struck on March 27, 1964. The outbreak it may have seeded wouldn’t strike for another 35 years.

In 2013, I wrote in Scientific American about a subtropical fungus called Cryptococcus gattii that appeared unexpectedly in 1999 in the lungs of hundreds of humans, pets and porpoises in the Pacific Northwest. Although rare, it could be picked up from something as simple as a walk in the woods and prove fatal in otherwise healthy individuals.

One of the most surprising and puzzling twists of the C. gattii story was that what appeared to be one outbreak was actually at least two and maybe three. Two unrelated strains of C. gattii appeared around 1999 on Vancouver Island, while a third emerged six years later in Oregon’s Willamette Valley. Today we know the three are so different they may be separate species. At the time, experts were puzzled about the origins of all.

Many ideas were floated, including chance introduction by wind, ocean, animals, eucalyptus trees, tires, crates or tennis shoes. Most scientists agreed that the fungi seemed to have made their way to the Pacific Northwest many decades prior, and some subsequent disturbance—perhaps climate change—generated a burst of infections.

Now David Engelthaler and Arturo Casadevall, infectious disease scientists at the Translational Genomics Research Institute in Flagstaff and Johns Hopkins University (I interviewed Casadevall for my 2013 story), have suggested a surprising hypothesis: that the fungi not only hitchhiked on ships from South America to the Pacific Northwest, but then surfed a tsunami to reach land. If so, why would infections not strike mammals for another 35 years?

Describing their hypothesis in the journal mBio last year, the pair stitch together a circumstantial case. DNA analyses of all three fungi suggest a burst of evolution when they arrived in the Pacific Northwest around 70 to 90 years ago, hinting at a common origin.

One candidate for that origin, Engelthaler and Casadevall suggest, is the 1914 opening of the Panama Canal. Empty cargo ships pump water into their hull as stabilizing ballast. The water—and any hitchhiking life—is often dumped in the next port. Cryptococcus species survive in seawater, and C. gattii can survive for at least a year. The burst of shipping through the new canal may have brought C. gattii repeatedly from a place like Brazil to the waters off Seattle, Portland and Vancouver.

If so, the fungus still needed to make it on to land. The 1964 earthquake—which generated a tsunami so large it killed people on beaches as far south as California—seems like it could have done the job, they say.

Natural disasters are well-documented vectors. A burst of fungal lung infections followed the 2011 Joplin, Mo., tornado, as I documented here. The 1994 Northridge earthquake in California sparked a mini-outbreak of Valley fever, another inhaled fungal disease. People roughed up by tsunami waves may go on to suffer invasive skin and lung infections, a condition called “tsunami lung,” and such waterborne infections from ocean flooding occurred after both the 2004 Indian Ocean tsunami and the 2011 Japanese tsunami. A survivor of the 2004 tsunami even suffered a skin infection from C. gattii.

But could a natural disaster introduce a pathogen to a new place, resulting in the outbreak of a new disease decades later?

Several lines of evidence suggest so, the pair argue. The forests and soils most heavily contaminated with C. gattii in the Pacific Northwest are those most affected by the tsunami: low-lying and close to the ocean. One exception—the area around Port Alberni in interior Vancouver Island—was nonetheless hard hit by the tsunami. A surge of water traveled down an inlet where it reached 26 feet high and washed away 55 homes. Today the fungus is found abundantly there, even though the town is relatively far from the coast.

The genetic data also reveal a second burst of evolution midcentury followed by another period of stability. After decades at sea, newly marooned fungi may have been forced to evolve quickly to survive in a place not only vastly different from the ocean but also dissimilar to their original home. Wild amoebas—amorphous single-celled microbes—prey on C. gattii. Learning to outsmart their new North American predators may have taken several decades. It may also have inadvertently trained the fungi to evade the amoeba-like immune cells called macrophages that travel our bodies doing essentially the same thing. This learning period could explain the decades-long delay between the tsunami and outbreak, Engelthaler and Casadevall suggest.

The earliest known case of Pacific Northwest C. gattii occurred in 1971 in Seattle. Nothing else is known about this case, but the tsunami hypothesis would help explain this outlier, since the fungi would have already been ashore for several years. Other scattered infections may have occurred between 1971 and 1999 and simply escaped detection, as the Cryptococcus can go dormant in hosts.

Finally, and perhaps most importantly, this hypothesis would help account for both the potpourri of apparently unrelated C. gattii in the Pacific Northwest and their varied emergence times. If several strains had established themselves in the ocean as a result of years of shipping, the tsunami could have washed them ashore simultaneously across hundreds of miles of coast. The corollary, of course, is that there could be still more “surprises” in store for us, ones perhaps even more efficient at attacking mammals. Further environmental testing both in the Pacific Northwest and in ports and nearby land unaffected by tsunamis could help support or refute their hypothesis and would be relatively easy to do, they suggest.

The mBio paper was published in October 2019, but it has implications for subsequent events. Engelthaler and Casadevall propose the Pacific Northwest C. gattii outbreak may be a “black swan”: an unpredictable event of extreme consequence. Indeed, it may be that many or even most outbreaks defy prediction.

The 2014 outbreak of Ebola in West Africa was probably inevitable given the conditions, many scientists believe, but the actual cause was the chance meeting of a group of sick migratory bats with children playing in a hollow tree. No one predicted that a flu pandemic would start in Mexico, but that happened in 2009. Similarly unexpected and unpredicted were the appearance of HIV, the SARS-CoV-1 and MERS coronaviruses, the Nipa and Hendra viruses and the monkeypox virus in the United States; the suddenly severe prenatal effects of Zika virus and the recent polio-like attacks in children suspected to be caused by a previously benign enterovirus D68 were also unforeseen, as was the appearance of C. Gattii in the temperate zone. Our present predicament is probably the biggest swan since the 1918 flu pandemic, which itself may have originated unexpectedly in Kansas.

Huge amounts of money, computing power and investigation resources have been thrown at the problem of predicting outbreaks of new disease. Those efforts failed us spectacularly this year. Financial philosopher Nassim Taleb, who coined the term black swan, argues that the proper response to such events is not to try to predict them; it’s to prepare for them. Although, in my view, it’s worthwhile to plumb their origins so we can try to avert future disasters (outlawing and aggressively prosecuting the sale of wildlife and reducing deforestation seem like obvious and humane choices), governments should just assume pandemics and outbreaks are inevitable and take appropriate action.

It’s not as though we don’t have precedent for expensive, defensive investments. In California, city planners and engineers know giant earthquakes will strike, but they don’t worry too much about the particulars. After all, even after a century or more of studying Golden State seismology and geology, the destructive 1994 Northridge earthquake occurred on a fault that didn’t even appear on seismic maps. Instead, they simply build accordingly.