The Surprising Connection Between Spruce Trees, Pine Siskins, and Salmonella Outbreaks

In the winter of 2020 and 2021, Ben Tonelli was a new Ph.D. student at UCLA—but the COVID-19 pandemic meant that instead of moving to Los Angeles, he was starting his doctoral program remotely from his home in Seattle. While he worked, he enjoyed watching through his office window as flocks of Pine Siskins visited his bird feeder. “It was very entertaining,” he says.

But then he started noticing sick birds, siskins that sat motionless and lethargic while their flockmates fed and flew back and forth. Eventually he found multiple dead siskins in his yard. The culprit was a salmonella outbreak that would ultimately send 14 people in Washington, Oregon, and California to the hospital that winter, the majority of whom could trace their illness to backyard bird feeders.

The siskin flocks Tonelli observed, which helped spread the bacteria, were part of an irruption—an irregular, large-scale movement of birds outside their typical range. So he began to wonder: Could salmonella outbreaks have any connection to siskins’ irruptive habits? A new study led by Tonelli and published in January shows that the answer is yes, via a multi-year chain of events that begins in mountain forests far from the cities where birds and people were sickened.

The sequence starts with a warmer-than-average summer in a high-elevation forest of spruce, fir, and hemlock trees. When a warm summer like this is followed by an unusually cool summer, the trees respond by scaling back their cone production the next spring.

This means a shortage of food for siskins and some other boreal birds the following winter, prompting them to wander widely at lower elevations and latitudes in search of sustenance. The result is a flurry of special guests, sometimes in dense congregations, at feeders across the United States. Whether siskins bring salmonella with them or pick up the bacteria at already-contaminated feeders is unclear, but either way, the huge numbers of birds create the perfect conditions for an explosion of salmonella. In total, it’s a two-and-a-half-year time lag between that first warm summer and the appearance of sick birds at feeders in the United States.

Ornithologists noted an apparent link between a lack of conifer seeds and bird irruptions decades ago, and meanwhile tree scientists have begun to unravel the temperature cues that lead to these ups and downs in cone production. More recently, the Finch Research Network has begun rigorously tracking seed crop production in the boreal to predict irruption years for certain species. 

Tonelli’s study, however, connected all of these dots and clearly linked irruptions to salmonella outbreaks for the first time. Drawing from data sources including Christmas Bird Counts, large databases of plant reproductive effort and wildlife disease reports, and temperature-monitoring satellites operated by NASA, he created a series of statistical models showing how one event leads to the next. (He even received funding from NASA for the project.)

Although he had a hunch about how the analysis would turn out, Tonelli was still surprised by the strength of the relationships he found. “They’re very clear,” he says. “If you have a big irruption, the likelihood of getting one of these big [salmonella] outbreaks is just so, so much greater.”

Heather Watts, a Washington State University biologist who’s studied siskins’ irruptive behavior, found the cascade of events linked by the study to be very compelling. “Going from climate to the tree ecology to bird behavior to disease outbreak and making those connections was really exciting to me,” she says.

According to Watts, while disease ecologists have long been aware of the connections between animals’ movements and disease outbreaks, irruptions have been overlooked. “This type of irregular movement in particular has gotten very little attention in the context of disease dynamics,” she says. “I think, as this paper highlights, it could be really important.”

Tonelli hopes this project could turn into a tool for forecasting salmonella outbreaks before they occur and warning people to take precautions, such as bringing in their feeders before huge flocks of siskins arrive. “I think that could go a long way in helping protect backyard birds, and also hopefully reduce the likelihood that people get sick,” he says, noting that it would take an organization with national reach to effectively communicate such alerts.

Aside from its practical applications, however, the research provides a fascinating glimpse into hidden ecological links across time and space. Tonelli, now a postdoctoral researcher at Clemson University, admits that for him this is the real draw.

“When I talk to people [about this], I usually frame it in that way, like a crazy chain of events spread out across years,” he says. “It’s really neat to see these connections that you wouldn’t realize were there unless you really dug into it, between things happening in different years and different places.”