Left: Backyard chickens, Long Island, NY. Credit Erica Cirino. Right: Northern pintail duck (male), Japan. Credit U.S. Geological Survey.
Most people think of bird flu as a poultry disease. But where it starts is inside the bodies of wild migratory birds.

The wild side of avian influenza

It all started on December 1, 2014, when a group of biologists from the Washington Department of Fish and Wildlife set out for Whatcom County to conduct some routine fieldwork.

Trampling through the snow toward the glassy cobalt waters of Wiser Lake, they noticed something amiss. On a typical December day thousands of ducks and geese would be quacking, honking, splashing and swimming. That day, energetic waterfowl still bandied about, but the lake was also shaded with the quiet pallor of death: dozens of lifeless birds bobbing in the shallows by the shore.

Normally, a small bird die-off on a big lake isn’t something to be too worried about. It happens sporadically on large water bodies with large populations of birds, like Wiser Lake, during migration season. If lake or pond water is contaminated with illegal lead hunting shot or a naturally occurring fungal disease called aspergillosis, it’s not unusual for some birds to die.

Yet the biologists were very concerned. Weeks earlier, a deadly bird disease had been detected on two poultry farms just miles across the U.S.-Canada border in Fraser Valley, British Columbia, where it sickened dozens of chickens and turkeys: avian influenza.

Normally, avian influenza in North America isn’t concerning — so long as it’s of the North American low pathogenic variety. This milder form of the virus circulates among ducks and geese throughout the continent each year without posing a high risk of infecting poultry. But, as its name suggests, highly pathogenic avian influenza — the type detected on those poultry farms in British Columbia — is notorious for killing poultry.

Poultry on outdoor farms, which typically have access to the outdoors, can pick up avian influenza through direct contact with wild birds — or more often, their scat. Poultry on commercial farms, on the other hand, are overwhelming kept indoors and thus out of direct contact with wildlife.

Hypothetically, birds on commercial farms shouldn’t get sick. But it does happen, and it’s a result of poor farm hygiene, from cross-contaminated equipment to workers who don’t disinfect their boots before entering a barn from the outdoors. After it enters the closed confines of an airplane-hangar-sized industrial barn where, avian influenza spreads rapidly among the thousands — sometimes hundreds of thousands — of birds inside. And just as it entered the farm through poor farm hygiene, avian influenza can be released from inside a poultry barn to the outdoors, where it can infect wild birds.

Aware that avian influenza — both low and highly pathogenic strains — can be transmitted from poultry to wild birds, and vice versa, wildlife and agricultural experts across the U.S.-Canada border had been on high alert since the Fraser Valley detections were announced. That North American waterfowl migrate from north to south in winter meant the wild birds then in Canada would soon be flying into the U.S. And highly pathogenic avian influenza in America, the biologists knew, could spell disaster for the $50 billion U.S. poultry industry.

All eyes remained on Wiser Lake.

Very concerned after a week of persistent waterfowl deaths, one of the biologists, Paul Debruyn, called a diagnostic virologist he knew in Wisconsin — Dr. Hon Ip, an expert in avian influenza expert at the United States Geological Survey-National Wildlife Health Center in Madison, Wisconsin, one of the nation’s top wildlife disease research facilities. Debruyne told Ip he’d be sending over nine carcasses — eight ducks and one swan.

The carcasses, collected on December 8, were bagged and boxed up with ice packs, then swiftly overnighted to Ip, via UPS. As soon as they arrived, the boxes were dropped from the center’s loading dock to its basement-level necropsy room with a dumbwaiter — garnering the same level of care and caution a confirmed shipment of SARS samples would receive. Ip’s colleagues in the necropsy room examined six of the carcasses closely to help determine a primary cause of death.

Several carcasses showed hemorrhaging and inflammation of organs, especially those of the respiratory system, including lungs and air sacs: all classic signs of aspergillosis. To determine if avian influenza was present in the birds’ bodies would take a little more work. Because waterfowl can carry avian influenza without showing signs of sickness, the only true way to diagnose it is through viral analyses of body tissue, blood and swab samples taken from primary body cavities.

So scientists in the necropsy lab collected tissue, blood and swab samples from all nine birds’ bodies and sent them, via dumbwaiter, back up to Ip’s diagnostic virology lab on the center’s first floor. After a first round of testing, Ip quickly determined that avian influenza was present in five birds’ bodies at their time of death. A second round of testing revealed four of the birds were carrying low pathogenic avian influenza, again, not a major concern.

However, on December 15, the fifth bird, a second round of testing on a lung specimen from Northern pintail duck tested positive for highly pathogenic avian influenza. It was the first highly pathogenic case of the virus diagnosed in the U.S. during what would come to be known as the global 2014–2015 avian influenza outbreak.

What’s more, the type of highly pathogenic avian influenza found in the Northern pintail was precisely the same strain that had hit the poultry farms in British Columbia: highly pathogenic Eurasian/North American H5N2, a hybrid of the North American low pathogenic virus and Eurasian highly pathogenic H5N8. H5N8 itself was a new highly pathogenic strain that had spread rapidly across Asia and Europe throughout 2014.

Along with H5N8 and another new strain, highly pathogenic Eurasian/North American H5N1, highly pathogenic Eurasian/North American H5N2 would be one of three highly pathogenic avian influenza viruses to be detected in the U.S. after December 2014. And for the following six months, it was primarily H5N2 that would go on to cause catastrophe on U.S. poultry farms.

As North American waterfowl continued their migration south, they carried with them highly pathogenic strains of avian influenza. Initially poultry on a handful of backyard and commercial farms in the American Northwest were sickened, presumably from direct contact with wild birds carrying the virus. Then migratory birds carried the virus to major commercial turkey- and egg-producing regions of the west — places like Minnesota, Iowa and South Dakota — where it spread rapidly from farm to farm.

From the Northwest, avian influenza continued east, carried by migratory waterfowl and sickening birds on poultry farms all the way to Indiana. Between December 2014 and June 2015, more than 51 million individual chickens — mostly egg-laying hens — and turkeys on 211 commercial and 21 backyard poultry farms across the American West and Midwest died or were euthanized after being infected with highly pathogenic avian influenza. According to USDA data, the $10 billion U.S. egg industry lost about 10 percent of its flock and the $5 million U.S. turkey industry lost three percent of its flock in those six months. The precipitous drop in supply sent the national average retail price of a dozen eggs to $2.72 in August, a five-year high and twice the price of a year earlier, according to commodities firm Urner Barry.

Similarly, the average national retail price of a pound of a whole frozen turkey went up from a dollar a pound in the spring to $1.26 a pound over the summer, according to an October USDA agricultural supply and demand report. However, this price increase was relatively short-lived: to keep the country’s turkey supply steady, farmers on the East Coast unaffected by avian influenza ramped up production when their western counterparts were hit, preventing any major dip in total U.S. turkey production.

In the weeks leading up to Thanksgiving 2015, the abundant turkey supply provided by eastern farms assuaged any fears of a shortage for the holiday on which Americans gobble down an estimated 46 million turkeys, according to the National Turkey Federation, America’s major turkey industry trade group. In fact, the supplemental turkeys helped to drive down the average cost of a whole frozen turkey in the U.S. to about 90 cents a pound, according to the USDA.

In July, agricultural economist Dr. Tom Elam, president of agribusiness consulting firm FarmEcon LLC, testified before the Senate Committee on Agriculture, Nutrition and Forestry that the disruption in poultry production as a result of the outbreak cost the U.S. economy a “conservatively estimated $3.3 billion.” Excluding the costs of euthanasia and facility cleanup, which is mandated by the USDA when the virus is detected in a poultry facility, Elam estimated the direct hit to producers at $530 million for turkeys and $1.04 billion for eggs.

While such economic devastation was considered a possibility when highly pathogenic avian influenza was detected in wildlife, at least one U.S. agricultural expert I spoke to by phone knew doom to the U.S. poultry industry was practically imminent once the virus hit a U.S. commercial operation.

“When we have highly pathogenic avian influenza [in wild birds] in the United States, consider that to be a major problem because that is a foreign animal disease to the U.S.,” said Dr. Brian McCluskey, chief epidemiologist at USDA’s Animal and Plant Health Inspection Service, the primary agency charged with monitoring and managing the outbreak. “But once it got into a big commercial flock we got even more concerned because the potential spread of that is considered to be of greater likelihood.”

Once a wild bird carries avian influenza to a geographic region, poor farm hygiene can bring it into a closed commercial poultry barn, which appears to be what happened the Stanislaus farm. However, in major poultry-producing regions where farms often share workers and equipment, poor farm hygiene can work to spread the virus from one farm directly into another. The latter is what appears to have happened on most of the commercial farms hit in the country’s major poultry-producing regions in the West and Midwest, according to new USDA research.

The USDA initiative, one of two major federal avian influenza research efforts triggered by the 2014–2015 outbreak, focuses on understanding how farm hygiene and avian influenza are linked so farm management practices can be improved and future outbreaks avoided. The other major effort, in which Ip is involved, entails surveying wild birds — acting as the proverbial canary in a coal mine — to stay ahead of any potential future outbreaks.

“This virus is not a superbug, it’s not too difficult to prevent,” said Ip, “Knowing what the science is, I can tell it’s clear many farms — especially those that produce animals in very high numbers — don’t have adequate biosecurity practices.”

On farms, biosecurity refers to management procedures meant to keep animal diseases at bay. While the USDA and poultry producer associations encourage farms of all types and sizes to adhere to biosecurity best practices such as using bleach booth baths and disinfecting farm vehicles after use, no law mandating or regulating any specific biosecurity standards currently exists.

What’s more, once avian influenza — especially in its highly pathogenic form — begins to spread, it has a high tendency to mutate, especially in poultry.

“‘Will it evolve?’ is always a major question when dealing with avian influenza,” said Ip. “Past and present research indicates that avian influenza’s virulence tends to increase in poultry once it gets onto a farm.”

Increased virulence, combined with changes in a virus’s genetic code, said Ip, could together create an avian influenza virus capable of infecting other animals, like swine, or even humans.

I traveled to Wisconsin in late September to meet Ip and others involved in avian influenza research, and got my first taste of the state after exiting my plane into the Dane County Regional Airport in Madison. The airport, I learned, is like most places in Wisconsin — that is, it’s filled with cheese. You can buy official Green Bay Packers Cheesehead hats, orange cheese wedge key chains and t-shirts with images of cheese. Of course, you can also get the real thing: blocks of cheddar and bags of curds are tantalizingly displayed in myriad refrigerator cases in these shops.

Indeed, it seems cheese is the symbol of Wisconsin: “America’s Dairyland,” its license plates proclaim. But it’s also home to a decently sized poultry industry. During the outbreak, one backyard flock and nine commercial poultry operations — mostly turkey farms — were hit across the state.

But I didn’t come to Wisconsin to talk turkey: I was there to see the wild bird survey effort in action.

First on my agenda was the National Wildlife Health Center. About 15 miles from the heart of Madison and with nothing but a small green-and-white sign marking its entrance, the place is easy to miss. Follow its long rural driveway to the end, and you’re greeted first by a small array of freestanding solar panels. The center itself is just beyond the solar panels, composed of two single-story brown brick buildings straight out of the 1960s, tucked into a luminously lush green-and-gold prairie landscape teeming with songbirds.

I spent my first morning in Madison exploring the center’s research labs and interviewing about a dozen of its employees. My afternoon there was spent shadowing a key player in the wild bird surveillance effort: Dr. Ip.

Ip is a fit and energetic bespectacled man: a runner and a scientist. He met me in the center’s brightly lit lobby, which was decorated with informational displays, framed nature photography and a healthy dose of taxidermy. From there he led me to the access room to his lab, where he and I donned matching teal green lab scrubs. Now we were ready to go into his lab…or so I thought. Ip’s hand unexpectedly paused on the door handle, and he stopped suddenly, spinning around to face me.

“Wait,” he said. “Before we go in, I must ask: Do you have any birds at home?” Ip had asked me this earlier, in the lobby where we first met.

“No,” I said again. “Just a dog.”

“Okay,” he said, still facing me with a hesitant hand on the door handle. “Please remember, you cannot go near any live birds for five days, starting tomorrow. You will not come into contact with any live birds, correct?”

“No,” I said, agreeing to his lab’s quarantine stipulation. “I won’t go near any birds.” A five-day quarantine from live birds, I learned, is mandatory for those working in and visiting Ip’s diagnostic virology lab. Though no new cases of highly pathogenic avian influenza had been detected in the U.S. since late July, it was still possible that the new samples coming into his lab for analysis may test positive. And if one was to pick up the virus and carry it outside, the U.S. poultry industry could face disaster all over again.

Ip smiled. Presumably satisfied with my verbal promises, he swung open the door to his lab and hurried me to an empty chair, placing a sheaf of paperwork and a pen on the table in front of it. Now I had to put my promises in writing.

No contact with live birds for five days. Signature. Acknowledgement I was entering a “biosafety level three” facility, where there was just a very small chance that I may come into contact with viruses that may cause me to develop dangerous or potentially lethal disease(s) after inhalation, such as West Nile virus, Eastern equine encephalitis virus, and, of course, avian influenza. Signature….

Similar to the idea of biosecurity on farms, biosafety measures in laboratories are meant to protect the people who work in the labs — and the outside world — from disease. Unlike farm biosecurity, lab biosafety is established on a scale of one to four, with more stringent, biosafety measures required the higher a lab is rated. At just one step down from the level of biosafety required for ebola lab (level four), biosafety in Ip’s lab (level three) is taken very seriously.

“In this lab we may study wildlife viruses, but many of these viruses have a strong impact on people,” said Ip, pulling up a chair and taking a seat. “West Nile, monkey pox, and of course, [highly pathogenic avian influenza].”

Avian influenza is an influenza A virus. These viruses are named according to the two types of proteins they’re made of. These two proteins, represented by the letters “H” and “N,” come in various subtypes, each indicated by a number. Also usually denoted is the virus’s geographic origin and pathogenicity (either low or highly pathogenic). Influenza A viruses — both animal and human — tend to “mutate readily, and that’s why you need to get a new flu shot every year,” explained Ip.

This isn’t the first time Ip’s lab has been involved in highly pathogenic avian influenza research: The virus has been on his radar since 1996. That year, highly pathogenic Asian H5N1 was first discovered on a goose farm in China, and went on to cause both poultry and human deaths there and in Egypt, Indonesia and Vietnam.

But it was a string of very large highly pathogenic Asian H5N1 outbreaks from 2003 to 2007 — which hit four continents and sickened both birds and people — that set into motion Ip’s first serious highly pathogenic avian influenza research effort. Under directives from the USDA, from 2006 to 2011, Ip and his colleagues tested swab samples from thousands of wild migratory ducks. The hope was that ongoing testing would serve as an early monitoring system, detecting any cases of the deadly virus in wildlife before it infected the nation’s poultry, and possibly humans.

The advent of the 2014–2015 U.S. outbreak sparked a similar monitoring initiative. The USDA’s current effort calls for the collection and subsequent testing of, at minimum, 31,150 samples from wild migratory ducks living in “targeted watersheds.” These include 98 areas across the nation with lakes and ponds popular among waterfowl during migration season, and where avian influenza has been detected in the past.

All this is to be accomplished between April 1, 2015, and March 31, 2016. Each state is responsible for coordinating its natural resources and wildlife management agencies to collect swab samples both live and deceased wild ducks at targeted watersheds, then send those samples to an approved animal health lab for avian influenza testing. Positive samples are then sent to the USDA’s National Veterinary Services Lab in Ames, Iowa, for a final diagnosis. The data from each sample, positive or negative, is collected and added to a USDA database.

Wisconsin’s designated animal health lab is Ip’s diagnostic virology lab at the National Wildlife Health Center. There, Ip receives samples primarily from the state’s USDA Wildlife Services unit.

Wildlife Services is a branch of USDA’s Animal and Plant Health Inspection Service, the USDA agency tasked with safeguarding the health of the nation’s livestock and crops. The original name of USDA Wildlife Services was Wildlife Damage, and it’s precisely that which remains the major focus of the branch’s work. Whereas the Wildlife National Health Center concentrates on wildlife and ecosystem conservation, Wildlife Services seeks to manage wildlife in a way that balances natural conservation with the activities and economic interests of humans. Each U.S. state has its own Wildlife Services team.

A day after meeting Ip I took a short cab ride from my hotel in downtown Madison to USDA Wisconsin Wildlife Services’ headquarters. Located in a plain, pale-blue building nestled in a commercial office complex in Sun Prairie, a quiet Madison suburb, it’s where Wisconsin’s wild waterfowl sample collection efforts are planned.

The inside of the building was almost as plain as its outside: aside from the oversize U.S. government crest that hung on the façade of the front desk, the office’s lobby was decorated sparingly. A smiling receptionist wearing a polished skirt suit greeted me. Later I would learn that, when triggered, the building’s security system summoned not the local police department, but Homeland Security.

I soon found myself in a room adorned abundantly with taxidermy. There, Dan Hirchert, director and biologist at Wisconsin Wildlife Services, and J.D. McComas, a Wisconsin Wildlife Services technician, joined me to explain what it’s like to collect samples to send to the National Wildlife Health Center as part of the national wild bird surveillance plan.

“The best time of year to collect samples in Wisconsin are fall and winter,” said Hirchert. This is when migratory ducks fly south for the winter down a major North American bird migration route called the Mississippi flyway, he explained, periodically landing in the U.S. to rest and refuel. The Mississippi flyway is one of four vertical paths North American migratory birds use to travel. Each of the four flyways overlaps so birds using separate flyways may interact with one another.

During the outbreak, highly pathogenic avian influenza hit three flyways: the Pacific, Central and Mississippi. Due to sheer geographic luck, the virus was not carried into the Atlantic flyway runs down the East Coast. That the virus did not reach the East, home to the country’s $32.7 billion broiler chicken industry was a very good thing, said Hirchert.

In August before wild waterfowl were to begin their fall migration, Wisconsin Wildlife Services capitalized off the Wisconsin Department of Natural Resources’ annual bird banding effort to help meet its surveillance quota. In all, Wisconsin Wildlife Services biologists and techs swabbed the front and back ends of more than 300 live ducks temporarily detained for banding. While interagency cooperation is not unusual, for the fall Wildlife Services is relying on what some may consider a unique ally: hunters.

Wisconsin Wildlife Services coordinated two major hunter-harvested duck sampling events this fall. The first was held the last weekend of September in Wisconsin’s northern hunting zone, while the second was held after my visit to the state on the first weekend of October in Wisconsin’s southern hunting zone.

On such collection weekends, Wisconsin Wildlife Services biologists and techs flock to popular hunting spots, waiting at boat landings armed with avian influenza sample kits: boxes filled with screw-topped test tubes, sterile cotton swabs, barcoded stickers and log sheets. After nabbing a day’s worth of ducks by boat — a popular method of hunting waterfowl — hunters return to these landings. There, the biologists and techs ask for permission to swab each returning hunter’s catch.

Because I hadn’t visited on a major sample collection day, Hirchert and McComas took me outside behind their office to show me what does happen on such days. Together, they carried out their not-so-glamorous task in a quiet grassy area below a pure cerulean sky.

First up was a small duck with a mostly gray-and-brown body with a flashy gleaming green head and curling tail feathers: an adult male mallard, called a drake. Hirchert jotted down that information — the duck’s age, species and sex — onto a log sheet along with a barcode sticker, while McComas pulled out a tube and stuck a matching barcode sticker onto its side.

Hirchert put down the log and picked up the duck, offering McComas its front end first. McComas swabbed the duck’s windpipe, or trachea. After a quick flip of his swab and the duck, he swiped its other end, its anal/reproductive cavity, or cloaca. Once finished sampling he snapped the swab in two and stuck both ends into the tube, which contained about a tablespoon of a cloudy liquid. That liquid, called culture media, helps preserve the swab samples for transport to the lab.

The two repeated this process with the drake’s counterpart: a less-flashy looking duck of about the same size, but with a mottled-brown body, a female mallard.

The process doesn’t take too long. But being a New Yorker I naturally wondered how willing hunters would be to donate their time in the name of science. Hirchert laughed when I brought this up. “Things are slower here in Wisconsin. Hunters are very patient, I think, because they know healthy wildlife is a benefit to them.”

After my trip, I spoke to several Wisconsin hunters about their experiences with Wildlife Services. All whom I reached out said they allowed Wildlife Services to sample their birds because they felt it was important to contribute to the U.S.’s ongoing avian influenza research.

One hunter, Dan Riese, a resident of Columbus, Wisconsin, was approached by Wisconsin Wildlife Services techs while at a popular duck hunting spot in Wisconsin’s southern zone called Horicon Marsh on opening day. He said he allowed the techs to test his three freshly caught wood ducks for two main reasons: “I think it’s important for them to track the present and [future] spread of diseases, plus nothing they were gong to do would have any effect on how I was going to use the birds.”

Back at the office, McComas is tasked with entering information from the collection log sheets into the USDA’s avian influenza database. Once logged samples are tested, the database is updated with their virology results.

This data provides the USDA and other federal agencies involved in the surveillance plan with a detailed picture of what birds tested positive for avian influenza, where they tested positive, and which strain — if any — they carried. This picture can then be used to predict where the next poultry farm may fall prey to avian influenza, and if any circulating strains pose human health risks.

No matter what their logic for contributing to his branch’s research, Hirchert said he is grateful for hunters’ help. “If it wasn’t for their cooperation, we’d have no data.”

The samples collected by Wisconsin Wildlife Services arrive at Ip’s lab via UPS, shipped overnight and on ice. As soon as they arrive, lab techs crosscheck each tube’s barcode with its corresponding log sheet, and assign it a specimen ID number. All of this information is copied into a thick black binder for safekeeping. Once accounted for, samples are ready for what is a decidedly long and somewhat tedious testing process.

Before any testing can happen, Ip explained, each tube must be properly prepped. Techs extract a small amount of each sample from its tube with a very delicate high-tech device called a micropipette that measures out extremely precise tiny quantities of liquid. Most micropipettes measure out liquid one sample at a time. Designed for speed, this one does 12.

The techs then use the micropipette to deliver the samples into smaller tubes containing tiny weak magnetic beads. These small tubes are thoroughly shaken on a vibrating tray the size of a shoebox, and then placed en masse into the hundred-plus rings of a magnetic tray made of rare earth metals. This very expensive and very heavy tray pulls the metal beads toward the tubes’ sides, along with any possible contaminants. The pure liquid left in the middle of each tube contains a trace amount of the genetic material Ip needs to test.

But before this genetic material can be tested, it must be amplified so there is enough to analyze. To do this Ip uses a technique called PCR, or polymerase chain reaction, which copies and duplicates the material with a machine that acts like a molecular photocopier. PCR is carried out in a thermal cycler, a nondescript white machine roughly the same the size and shape of a cash register, into which Ip places his samples.

Once the PCR process is complete, Ip uses another boxy white machine to compare the replicated genetic material from the samples he receives from Wisconsin Wildlife Services with that of confirmed samples of avian influenza.

On the day of my visit, the National Wildlife Health Center had just received a fourth thermal cycler to add to its collection. The new machine, Ip said excitedly, would increase his lab’s avian influenza research output by about 25 percent.

“Understanding the genetics behind the virus is important for understanding possible future mutations that may occur. With mutations, it’s possible highly pathogenic avian influenza could become infectious in people or swine,” Ip said.

But genetic testing has also helped him map out how the current virus spread. This is critical to preventing another outbreak “that would again devastate the U.S. poultry industry,” Ip said.

Genetic analyses allow Ip to determine the relationship between each case of avian influenza he receives. Specifically, he compares information the USDA’s avian influenza database and published results from genetic analyses of avian influenza strains found in other countries. The more closely related the viruses, the more likely they are to have come from the same region — or even the same source.

Based on these data, Ip and other infectious disease experts across the world believe the outbreak happened something like this: Wild waterfowl summering in Siberia in 2013 carried various low pathogenic avian influenza strains, at least one of which was related to highly pathogenic Asian H5N1. Come fall and winter, the waterfowl began to migrate south, and by January 2014 reached their winter feeding grounds in South Korea. There, the wild birds came into direct contact with poultry on outdoor farms, which picked up low pathogenic avian influenza. Once inside the bodies of poultry, the viruses mutated, increasing in virulence and creating a new strain: highly pathogenic Eurasian H5N8. Poultry infected with H5N8 then passed it into the wild birds, which, along with poor farm hygiene, helped to spread the virus on both backyard and commercial poultry farms throughout South Korea and nearby Japan until April.

Come spring, wild waterfowl infected with H5N8 left South Korea and Japan to return to their summer breeding grounds in Siberia and Alaska. As in North America, in Asia flyways overlap with those of other regions. As a result, groups of wild waterfowl that spend the summer in Siberia mingled with birds from several continents other than Asia, including Europe, Australia, Africa and North America.

During the summer of 2014 H5N8 circulated among wild waterfowl in Siberia and Alaska. In September 2014, some of the wild birds in Siberia again began migrating south toward South Korea and Japan. As a result, H5N8 hit poultry in China and South Korea in September, and in Japan and Taiwan in November.

Other wild birds left Siberia for winter feeding grounds west of Eastern Asia. By November, migratory birds that had spent the summer of 2014 in Siberia had carried H5N8 to Germany, where it spread to several backyard and commercial poultry farms. Once in Western Europe, H5N8 continued to spread from wild birds to poultry and vice versa as more poultry became infected, and also from farm-to-farm. By mid-December 2014, H5N8 had hit poultry on farms in The Netherlands, United Kingdom and Italy.

Also during summer 2014, H5N8 circulated among wild waterfowl summering in Alaska, brought there by birds that had picked it up while migrating over Asia. In the fall, waterfowl from Alaska migrated south into Canada, their eventual destination being the southern U.S., Mexico and South America. They reached British Columbia in November, where they stopped over to rest and refuel. There, poor farm hygiene facilitated the spread of H5N8 and several low pathogenic strains of North American avian influenza from the outdoor environment into a commercial poultry farm in Fraser Valley.

Inside poultry, the various viruses merged to create a new highly pathogenic strain: Eurasian/North American H5N2, which became the first highly pathogenic avian influenza virus to be detected on a commercial farm in North America. Around the same time, another new highly pathogenic strain, Eurasian/North American H5N1, was similarly formed when H5N8 and low pathogenic avian influenza strains merged in poultry on a different Canadian farm nearby. Both H5N1 and H5N2 exited the farms back into the surrounding environment — presumably through lapses in farm hygiene — where the same migratory waterfowl that had carried H5N8 and low pathogenic avian influenzas into Canada picked them up.

These wild waterfowl then carried all three highly pathogenic strains — H5N1, H5N2 and H5N8 — across the U.S.-Canada border to Wiser Lake, a major migratory waterfowl stopover. H5N1 and H5N2 were both diagnosed in wild birds harvested from the lake, while H5N8 was diagnosed in a captive falcon that had died after eating a wild bird caught there. From Wiser Lake, some wild birds carried H5N2 and H5N8 to nearby backyard poultry farms in the Pacific Northwest. At least one wild bird carried H5N8 all the way down the Pacific flyway to Southern California by late January, where the virus managed to make its way into the nation’s first commercial poultry farm. By early April, migratory birds brought H5N2 to a backyard poultry farm in the Central flyway. Later that same month migratory birds introduced H5N2 to the Mississippi flyway, after which point farm-to-farm spread of the virus predominated.

“Research suggests that wild birds naturally have only low pathogenic avian influenza,” said Ip. “The evolution and spread of these hybrid viruses in poultry support the hypothesis of spillback. It seems as though poultry is responsible for infecting wild birds with highly pathogenic avian influenza.”

But that should not happen on industrial farms where chickens and turkeys are typically kept indoors and do not encounter wild birds, Ip said. The fact highly pathogenic avian influenza “appears to evolve exclusively in the poultry system is a result of inadequate farm biosecurity.”

“I wish there was an education tool to help farmers really see how their farm biosecurity practices are working,” Ip mused. “Imagine if we could make pathogens fluoresce — if farmer’s boots or truck tires were glowing with avian influenza then maybe they would understand that what they’re doing isn’t working.”

McCluskey acknowledges farm biosecurity as a whole isn’t perfect. But unlike Ip, he said that he did not explicitly implicate U.S. poultry industry as the main offender in the recent outbreak. Instead, McCluskey said both wild birds and farmers appear to be at fault.

For several months McCluskey’s USDA Veterinary Services branch has been working to discern what makes poultry farms so vulnerable to avian influenza. As part of this effort, branch techs and scientists visit farms to conduct surveys — with questions focusing on biosecurity practices — and collect samples (mostly air and bird manure) to be analyzed in the lab. Positive cases of highly pathogenic avian influenza are mapped.

Based on his branch’s research, especially genetic analyses of the viruses found on farms, McCluskey said the outbreak’s earliest cases occurred independently from one another, while its later cases “include potential for human involvement” — in other words, are almost certainly the result of poor farm biosecurity.

During the outbreak, many of the earliest avian influenza cases on U.S. poultry farms, including the first five, were detected on backyard farms, which can be anything from a small backyard coop holding a handful of laying hens to a flock of 100 free-range turkeys. On backyard farms, poultry typically live at least partly outdoors and thus have a greater chance of interacting with wild waterfowl and contracting disease from those birds independent from other farms. This runs in line with Veterinary Services’ findings that the genetic makeups of virus samples taken from these farms were slightly dissimilar, suggesting a different wild bird infected each farm.

By contrast, later avian influenza cases on U.S. poultry farms were detected almost exclusively on commercial farms. There most birds are kept in indoor facilities their whole lives and thus do not interact with wild birds. Veterinary Services reports reveal that the virus samples taken from these farms were very similar, implying a common source of infection.

Major poultry companies contract several farms in order to meet demand for their products. These contracted farms tend to share workers, vehicles and equipment, which can facilitate the spread of disease from one farm to another, or from the outside environment into a farm, McCluskey said.

However, biosecurity practices are supposed to prevent such so-called linear spread of disease from occurring. Though Veterinary Services stops short of identifying any specific cause or causes of this, in a research report published in September it emphasized that “comprehensive and stringent biosecurity practices remain crucial to reducing the risk of [highly pathogenic avian influenza] infection.”

“We found some poultry operations with good biosecurity practices, but also a few places that could use improvement,” said McCluskey. “USDA is now discussing with industry how to monitor biosecurity practices, and is performing audits to ensure those practices are being followed.”

Turkey giant Jennie-O, a major employer in Wisconsin and Minnesota, was contacted by phone and invited to discuss the biosecurity practices employed on their contracted farms, and also the outbreak’s impact on their business. But Pat Solheid, the company’s vice president of human resources and administration said, on behalf of the company, “We are not sure on what level we’d like to contribute,” suggesting there was “plenty of public information” detailing how avian influenza impacted their company, available on the company’s website.

So, I wasn’t able to inquire whether or not the company holds its independently contracted farmers to any specific biosecurity standards and whether or not those standards changed after the outbreak. However, their website does reveal the extent to which the company was impacted with the virus: in all, 58 of Jennie-O’s turkey facilities — both independently contracted farms and farms owned directly by the company in Minnesota and Wisconsin — were hit by highly pathogenic avian influenza.

Similarly, Jennie-O’s contracted farms were unavailable for comment, not because they declined, but because their contact information is kept private. Agricultural states have statutes, usually called “livestock premises registration” laws, which require livestock facilities managers register their personal and business information with their state of residence under the understanding that their identifying information will be kept confidential. This is done to protect the privacy of farm owners, for whom “these farms are their homes,” said Keith Williams, vice president of communications and marketing of the National Turkey Federation.

What is known is this: the 58 Jennie-O’s turkey farms that tested positive for highly pathogenic avian influenza — all of which were hit between March 27 and June 4 — were forced to undergo a federally mandated process of quarantine, depopulation and disinfection, like all poultry farms that test positive for highly pathogenic avian influenza. During the outbreak McComas was deployed to Minnesota to oversee this process on several infected turkey farms (though he wouldn’t disclose the farms’ corporate affiliation) and he explained to me the federal response process he witnessed firsthand.

Once avian influenza is suspected on a farm movement of the farm’s workers, vehicles, equipment and birds off-site restricted, said McComas. If avian influenza is confirmed, all birds are euthanized, composted and buried on site. A farm’s barns are rid of the “three Fs: feathers, feces and feed,” then are sprayed with a disinfectant solution. After three weeks, the farm can begin to repopulate with new birds.

McComas said the farm managers he worked with appeared to embrace a new culture of improved farm hygiene, even though some appeared leery at first of government intervention.

“Most farmers don’t exactly want people in white Tyvek suits with respirators on their property,” said McComas. “It doesn’t look good; it’s upsetting. We had to let them know we were there to help.”

But besides the emotional difficulty of coping with the process of handling a case of avian influenza, farmers, and often their entire communities, are forced to face a harsh economic reality in the virus’s wake. Contracted farms are provided a federal stipend to cover the costs of euthanasia and hazardous material cleanup teams, as well as compensation for the live birds they must destroy. But the government does not reimburse farmers for birds who succumb to avian influenza before the euthanasia process, nor does it pay them salaries for workers who must be let go when farms are shuttered.

“When poultry farms are shut down, even for a relatively short period, it can profoundly effect whole communities,” said Hirchert. “In an area with several huge farms, hundreds of people are out of work both on farms and in processing facilities — no one is getting paid. Some may be laid off.”

With birds destroyed and contracted farms shuttered across two states, in May Jennie-O laid off more than 200 employees at a major processing plant in Minnesota, and cut the hours of many employees it kept. Rembrandt Enterprises, one of the nation’s largest egg producers, implemented similar layoffs and cuts after a positive case found on a major contract farm in Iowa led to the euthanasia of over 250,000 hens.

The experts agree that better farm biosecurity can help prevent a future outbreak. Yet federally mandating better biosecurity on the nation’s 233,770 commercial poultry farms, which would require routine inspections by USDA officials, is no small task. Recognizing this challenge, USDA is now ready to deploy a quicker and less expensive, albeit more reactionary, fix in the case that another avian influenza outbreak does occur: an avian influenza vaccine.

Almost immediately after the outbreak struck, several federal, university and private labs turned their attention to avian influenza vaccines. Such vaccines have been previously developed and tested for use in chickens, but only in labs. Protecting only six out of 10 inoculated hens from highly pathogenic avian influenza, these vaccines haven’t yet hit the efficacy level USDA seeks. Nor have they been tested in turkeys. For these two reasons, the USDA hasn’t approved them for use on poultry farms.

However, in October, the USDA found what it believes may be a vaccine effective enough for farm use. After running efficacy studies on the newly developed drug, the USDA awarded its creator, animal vaccine developer Harrisvaccines, a $6 million contract to create a 48-million dose stockpile of avian influenza vaccines, with 25 million doses ready in 45 days.

Creating a stockpile may seem risky, said Vice President of Harrisvaccines Joel Harris, lest the nation stock up on a treatment for the “wrong” avian influenza strain. But he said this would not be a problem with his company’s vaccine, which, unlike typical vaccines, can be “easily updated” to treat any problematic strain that may arise.

“Our vaccines are made with a synthetic replica of a virus’s genetic code rather than a live virus, which is what most drug makers use to create vaccines,” said Harris. “This technology is more adaptable and can be produced more rapidly than the traditional technology.”

The USDA says it plans to rapidly distribute vaccines from Harrisvaccines’ stockpile if another highly pathogenic avian influenza outbreak occurs in the U.S. in the future.

At this point, the big question looming is whether or not current outbreak is over for good. Wildlife health experts such as Ip are hesitant to declare it over just yet. The key, he and others have said, will be the continuation of the wild bird surveillance and poultry farm research efforts currently underway.

“We are concerned with the possible reappearance of [highly pathogenic avian influenza],” Ip said, pointing out that while the U.S. outbreak played out, new strains of avian influenza appeared in Asia. “These viruses or additional variants could be introduced into North America by the fall migration. So in the words of Mad-Eye Moody in J. K. Rowling’s Harry Potter books: ‘Constant Vigilance’ is what it’s all about.”

Check out my website to learn more about my experience reporting this story and to access maps and a photo gallery.