Scientists Finally Identify a Deadly Toxin That's Been Killing Birds

Thousands of eagles and other fowl have died from a mysterious condition that attacks their nervous systems. Now, after decades of investigation, we know why.
bald eagle
Photograph: Frank Bienewald/Getty Images

For 25 years, a mysterious killer has been on the loose across the American south, responsible for the deaths of over 100 eagles and thousands of other birds. The first victims were found in the fall of 1994 and winter of 1995 when 29 bald eagles died at or near Lake DeGray, Arkansas. At first, the birds seemed to be untouched. But during an autopsy, scientists found lesions on their brains and spinal cords, a condition they named avian vacuolar myelinopathy (AVM). Researchers at the Department of Fish and Wildlife searched for diseases or toxins like DDT that might cause this debilitating disease, but they found nothing.

The mystery went unsolved.

The killer appeared again a few years later in the Carolinas, Georgia, and Texas. In addition to bald eagles, it had started attacking water birds like Canada geese, coots, and Mallard ducks. First it rendered the birds unable to fly. They stumbled around, their wings drooped, they looked catatonic or paralyzed. Then—in as few as five days—they were dead.

Now, in a paper published today in Science, an international team of researchers from Germany, the Czech Republic, and the United States have finally identified the culprit, a previously unknown neurotoxin called aetokthonotoxin, which could be produced by a deadly combination of invasive plants, opportunistic bacteria, and chemical pollution in lakes and reservoirs.

To find this new toxin, scientists had to work together like detectives, assessing the crime scene and interrogating suspects. Susan Wilde, a professor of aquatic science at the University of Georgia, first began investigating the mystery in 2001 when 17 bald eagles died in Lake J. Strom Thurmond, a man-made reservoir on the Georgia-South Carolina border. “I had seen the eagle deaths before in past events, but this one was the reservoir where I had done my dissertation research,” she says. “It was an interesting mystery but kind of hit home. That was the reservoir I had worked on and seen a lot of eagles flying over.”

When Wilde had been collecting data for her dissertation in the mid-1990s, there wasn’t much vegetation growing in the reservoir. But when she returned a few years later, the lake had been overtaken by an invasive plant called hydrilla, which is easy to grow and had become a popular plant for fish tanks. (It’s rumored that hydrilla was initially released in the US in the 1950s when it outgrew an aquarium and someone dumped it out into a Florida waterway. Since then, it’s become one of the most pernicious aquatic weeds in the country, thriving in freshwater lakes from Washington to Wisconsin to the Carolinas.) Wilde began to wonder if the eagle deaths and the presence of this new plant were related.

But Wilde had to interrogate all the potential suspects. She started by sampling the water and lake sediment for bacteria. She came up empty-handed. But when she started examining the hydrilla plant’s leaves, she found colonies of a previously unknown cyanobacteria. She named it Aetokthonos hydrillicola, “the eagle killer that grows on hydrilla.”

Photograph: Getty Images

Cyanobacteria, also known as blue-green algae, are famous for creating the toxic blooms that poison lakes and seafood. Wilde hypothesized that the toxin was produced on the leaves of this plant and then eaten by herbivorous birds swimming around in the lake. When the poison started to work on the birds’ nervous systems, they became catatonic: easy prey for the bald eagles who migrate south every year to nest. When the eagles ate the infected prey, all the toxins stored in the birds’ muscles and stomachs were transferred to the eagles.

But to be sure she was pursuing the right suspect Wilde needed to grow some Aetokthonos hydrillicola in the lab, to find out which toxin it produces. But that’s easier said than done. Bacteria are notoriously difficult to cultivate. Plus, she had to culture them in a setting that mimicked the water in the reservoir. “It’s sort of hard to recreate that environment in the lab,” says Wilde. The cultures kept getting colonized by other bacteria that grew faster and more readily. “We had a lot of trouble with contamination and getting the culture started,” she says.

That’s when Timo Niedermeyer called. Niedermeyer, a scientist at Martin Luther University Halle-Wittenberg in Germany, studies cyanobacteria, and when he stumbled across Wilde’s work he was intrigued by this new species. His team had Wilde send over a few samples of the colonized hydrilla leaves and they figured out a way to get the bacteria cultured in the lab. It still grew incredibly slowly—they had to wait 18 months to get enough bacteria to run any tests—but it seemed like they were headed in the right direction.

After waiting and waiting, Niedermeyer finally had enough bacteria to run an assay to see what toxins they were producing. They found nothing. “This was really frustrating, of course,” says Niedermeyer. “And we had no idea what to do.”

By now, they’d already spent a lot of time on a project that wasn’t panning out. “We worked for five, six years without really any result. Only cultivating for nothing,” he says.

The scientists had to regroup. They didn’t want to be too committed to one theory. “You don’t want to get so attached to your hypothesis that you can’t look honestly at data that says ‘No, that is wrong,’” says Wilde. “But the trick is that negative data doesn’t necessarily mean the hypothesis is wrong. It just means that you didn’t demonstrate it in that trial. So we did try and try again.”

This time, Niedermeyer asked Wilde to send an entire hydrilla leaf and the stems. Instead of scraping the bacteria off of the leaf, he kept the whole thing intact. He and his team examined it using mass spectrometry, an imaging technique that allowed them to see individual molecules on the leaves. Not only did they see the cyanobacteria, but sitting alongside it on top of the leaf they also noticed another compound, which contained five bromine atoms. Bromine is a chemical element that’s highly reactive and isn’t usually out and about in the environment. It does appear naturally in its less reactive negatively-charged ion form: bromide. But even bromide doesn’t usually show up in freshwater environments like J. Strom Thurmond Lake. Where they do make frequent appearances are in man-made products: Humans use bromide in sedatives, fuel additives, and to sanitize water.

The medium his lab used to grow the cyanobacteria hadn’t contained any bromide. Niedermeyer realized that this must be the missing ingredient the cyanobacteria needed to make its deadly toxin. “This was like a ‘Eureka!’ moment,” he says.

They added bromide to the mix and, indeed, the cyanobacteria produced a toxin. Niedermeyer finally got to call Wilde and tell her they’d found the killer. “That was great,” he says.

Robert Sargent, a program manager for the Georgia Department of Natural Resources, describes the discovery as “outstanding news.” He’s particularly excited that the researchers have figured out a way to detect the toxin in the lab. “It is just remarkable for ecology, for us getting a better grasp on understanding this process and perhaps being able to control it,” he says. He points out that while the eagle deaths are alarming, they are a sign of a much bigger problem. “Whenever we see illnesses or deaths of species at the top of the food chain, it’s a red flag for the potential health of the environment,” he says.

After finding the toxin, the research team picked up speed. They isolated the compound containing bromide and confirmed it was present in the dead birds that showed lesions. They looked at the hydrilla plant itself and discovered it’s able to enrich bromide from the environment, making it even more available to the cyanobacteria. “The concentration of bromide in the plant is much higher than in the water or in the sediment where the plant grows,” says Niedermeyer. “This is kind of intriguing, but we don’t know why the plant does it.”

But in this murder mystery, identifying the culprit isn’t quite the same as ending the story. The team still has a lot of questions. Did the cyanobacteria invade with the hydrilla or was it already in the water? Is the bromide naturally occurring, or could it be coming from man-made sources like coal-fired power plants and flame retardants? Hydrilla is such a persistent pest that people have tried using herbicides like diquat dibromide to kill it off; could that herbicide be the source of the ingredient that creates this toxin? Wilde and Niedermeyer think it’s possible.

They’re also very concerned about whether this neurotoxin could affect humans who eat infected fowl. “This could be a real issue, but we don’t know that yet,” says Niedermeyer. Wilde wants to start monitoring in more locations. Not every lake that has hydrilla has had an AVM outbreak, but there are many where the weed has been treated with herbicide, and they could potentially become toxic in the future. Wilde hopes that with more monitoring, scientists can get ahead of possible outbreaks and keep this from spreading even further.

Sargent adds that residents can also play a role in efforts to control AVM outbreaks by not dumping aquarium plants into waterways. Boaters can remove aquatic plants from their propellers and hulls, and if people see oddly behaving aquatic birds or birds of prey, they can report those sightings to their state wildlife agency.

Just managing the outbreaks that have already happened has proven to be tricky. Hydrilla is a tenacious plant. The Army Corps of Engineers has had luck using grass-eating carp to eat back the weed, but even after being chomped on by fish, it will regrow from tubers buried in the lake’s sediment. And even though it grows slowly, Aetokthonos hydricolla is just as hard to get rid of. “They simply survive. You can’t kill them,” says Niedermeyer. He recalls a few cultures in dishes in his lab that had been forgotten and weren’t cared for properly. “We thought, ‘OK it’s dead,’” he says. “But no. If you just add a little bit of fresh medium, it starts growing again.”

Niedermeyer says that now that they know what they’re looking for, scientists have a better chance of finally stopping the killer once and for all. “Now that we are aware of the problem we can screen for the cyanobacterium. We can monitor the toxin. We can start sampling water bodies for bromide,” he says. “Now that we know what we are looking for, we can start finding a solution.”


More Great WIRED Stories