Some blackpoll warblers are born to be different. Each fall, instead of migrating southeast with their brethren on the long journey from Canada’s boreal forest to South America, they head southwest across the Rocky Mountains and toward California. Traveling at night, a few overshoot the coast and find themselves, in the morning, far out over the Pacific Ocean. 

If they can still see the mainland, these diminutive birds, which weigh barely more than two quarters, are thought to backtrack there. But as every San Franciscan knows, more often than not it’s too foggy or overcast to see much of anything. So, when Karl envelops them, some warblers descend to a few shrubby trees on the only specks of land in sight: the Farallon Islands.

The Farallones, located some 27 miles west of San Francisco and closed to the public to protect the wildlife, are best known for hosting the largest seabird colony in the Lower 48. But they’re also famous for attracting vagrant birds that have wandered far outside their normal ranges—not just blackpolls and other warblers, which are beloved by birders for their splashy colors and complex songs, but all types of birds from all corners of the globe. Thanks to the large number of vagrants, 446 bird species have been spotted so far on the 70-acre Southeast Farallon Island, where researchers have their home base. That’s more species than just about any other location in the United States. 

The Farallon Islands don't have many decent perches for a traveling songbird, except for a few scrubby trees on Southeast Farallon Island, shown here in 2021.
The Farallon Islands don’t have many decent perches for a traveling songbird, except for a few scrubby trees on Southeast Farallon Island, shown here in 2021. Matt Brady

Researchers have been watching and puzzling over the vagrants there for decades. Studying vagrants—whether it’s birds, seals, butterflies, dragonflies, bats, sea turtles, whales, or mountain lions—is inherently difficult. How do you study something that by definition shows up where it’s least expected? The Farallones are one of the few places on Earth that reliably attract enough birds to make a decent enough data set to try. From 2016 to 2022, about 9 percent of the birds banded on Southeast Farallon Island were species that don’t regularly occur west of the Sierra Nevada. On the California mainland, even at noted vagrant hotspots like Point Reyes, that number is typically less than 1 percent. “I can’t think of any place better to study vagrants. [It’s] the holy grail,” Benjamin Van Doren, a migration ecologist at the University of Illinois Urbana-Champaign, says of the Farallones.

Now, Van Doren has co-authored a new paper, published in the journal Ornithology, that  suggests we may have been underestimating some of these visitors from afar. The paper presents “some really intriguing results,” says Alexander Lees, an ecologist at Manchester Metropolitan University in England and co-author of the book Vagrancy in Birds, who was not involved in the research. There’s still much we don’t know about why a select few birds go so far off the (wing-)beaten path. But the Farallones are helping to provide answers.


Vagrancy in birds dates back millions of years; that’s how, for instance, finches found their way to Hawaii and evolved into honeycreepers. The travelers bring not only themselves but seeds and other biological material, sometimes changing the ecology of the places they land. “Insect eggs and even fish eggs can pass through a bird and come out the other side and be viable,” Lees says. Scientists used to believe this was all accidental. Vagrant birds were, probably, blown off course by nasty weather. Or they were essentially flying about haphazardly. 

Diagram of an Emlen funnel, invented in the 1960s, which has been used to see in what direction vagrant birds want to take off.
Diagram of an Emlen funnel, invented in the 1960s, which has been used to see in what direction vagrant birds want to take off. L. Shyamal via Wikimedia Commons, CC0

Dave DeSante, a bird researcher working on the Farallones in the 1970s, was among the first to realize there might be more to it. After capturing “lost” blackpoll warblers in mist nets, he transferred each one to a funnel-shaped cage, with an inkpad at the base, so that the bird would leave footprints when it tried to fly off. The footprints revealed the blackpolls’ desire to head southwest—toward the open ocean. As a result, DeSante theorized that they must have a “mirror-image” abnormality—possibly a genetic mutation—that flipped their internal compass. In other words, these vagrant birds were misoriented in a specific direction, not randomly disoriented. For birders who delight in spotting rare East Coast warblers in coastal California, this would mean that every one they witness may be about to take its last flight before plummeting of exhaustion and drowning in the Pacific.


Get Bay Nature’s Free Weekly Newsletter


Dave De Sante put Farallon vagrant birds in funnels to see which direction they would go upon departing, as this diagram from his 1973 Ph.D. thesis shows.
Dave De Sante put Farallon vagrant birds in funnels to see which direction they would go upon departing, as this diagram from his 1973 Ph.D. thesis shows. Dave DeSante, 1973
Blackpoll feet made these data.
Blackpoll feet made these data. Dave DeSante, 1973

DeSante’s “mirror-image” hypothesis remains plausible, if not proven, Lees says. But it doesn’t explain all the vagrancy. Solar storms also seem to impact bird migration, likely by cutting off birds’ ability to navigate via the Earth’s magnetic field. The agonic line—where true north and magnetic north intersect—may play a similarly disruptive role. Lees points out that the warblers analyzed for the new Ornithology paper travelled to the Farallones from a part of Canada corresponding with the agonic line. Other possibilities for avian vagrancy include exposure to toxins and tumors and other diseases. “We may be closing in on one or more unifying vagrancy mechanisms,” Lees says.

Then, of course, there’s the weather: some birds clearly do get blown away, and clouds interfere with their ability to navigate via the stars. Eastern U.S. warblers show up regularly in Europe, while European songbirds almost never cross the Atlantic to North America. These errant U.S. warblers are believed to be getting caught in storms and following the prevailing winds—essentially the same winds that make a plane flight to Europe faster than a plane flight home.


Warblers are too small to tag with GPS devices. So Farallones scientists decided to investigate warbler vagrancy with their feathers instead. In the keratin of feathers, birds carry a chemical signature of their birthplaces. A hydrogen atom—a key building block in all living creatures—can have one neutron (that’s the “heavy” kind), or zero neutrons. Rainwater in different parts of the world has different ratios of these two isotopes; the ratio is the chemical signature. Then, that rainwater is taken up by plants and the rest of the food web, including birds, whose feathers bear that same signature. Researchers can measure the ratio using mass spectroscopy—it’s called “stable isotope analysis.” The method is commonly used to understand the movements of lightweight animals from butterflies to birds. 

The Connecticut warbler, says Jim Tietz, is very hard to see on the mainland because it is "very skulky" and hides in vegetation. On the Farallones, it's a lot easier to spot.
The Connecticut warbler, says Jim Tietz, is very hard to see on the mainland because it is “very skulky” and hides in vegetation. On the Farallones, it’s a lot easier to spot. Matt Brady
Chestnut-sided Warbler.
Chestnut-sided warbler. Jim Tietz / Point Blue Conservation Science
A black-throated gray warbler in the Farallones. It would more usually be found in pine forests west of the Rockies, or wintering farther south.
A black-throated gray warbler in the Farallones. It would more usually be found in pine forests west of the Rockies, or wintering farther south. Kurt Ongman / Point Blue Conservation Science

Farallones scientists collected feathers from 127 blackpoll, Blackburnian, palm, Tennessee, chestnut-sided, and magnolia warblers that were netted on Southeast Farallon Island, and got additional feathers from 28 museum specimens. All six of these warbler species typically nest in Canada’s boreal forest and migrate through the eastern or central U.S. on their way to and from the tropics. 

If vagrancy was just random, scientists figured, most of the vagrants would come from the East, where warblers are most abundant. But the feather isotopes revealed that most of the warblers had come from the western end of their breeding ranges instead. Other studies have similarly uncovered evidence of “increased vagrancy at range edges independent of population sizes,” Lees says. Perhaps these western warblers have been naturally selected to seek out new suitable habitat. If that’s the case, some vagrants arriving on the Farallones may not be hopelessly adrift so much as they are advanced scouts on dangerous range-expansion missions. “You could see them, in an evolutionary sense, as explorers for their species,” Van Doren says.

In an evolutionary sense, that is, because if some vagrants are little Vasco da Gamas and Amerigo Vespuccis (without the colonialist undertones), they could be altering the trajectory of their species. Rampant climate change and habitat destruction are making once core areas uninhabitable. “If birds can’t rapidly adapt, then they’re in big trouble,” Lees says. “They have to be able to shift as the world changes.” At least some species are doing just that. Lees notes that migratory orientation is heritable, meaning the offspring of vagrant explorers could follow them to new lands. In recent decades, palm warblers have established a small wintering population in California, presumably derived from vagrants, as have Richard’s pipits in southern Europe and lesser black-backed gulls on the East Coast.


Even on the Farallones, vagrants aren’t exactly an everyday occurrence, especially outside of migration season. “It’s very hit or miss,” says Pete Warzybok, Farallon program leader at Point Blue Conservation Science, which, in partnership with the U.S. Fish and Wildlife Service, has been conducting year-round research on the Farallones since the 1960s. “You could be out there for a month and a half and not see anything unusual. And then one day you get five new species that are crazy.”

Cedar waxwing on the Farallons. These birds are found across North America, but they're not exactly seagoing birds. Normally.
Cedar waxwing on the Farallons. These birds are found across North America, but they’re not exactly seagoing birds. Normally. Boo Price / Point Blue Conservation Science
Blue-gray gnatcatcher, catching. Its breeding range across North America includes the Bay Area, but not so much the Farallones, where this one turned up.
Blue-gray gnatcatcher, catching. Its breeding range across North America includes the Bay Area, but not so much the Farallones, where this one turned up. John Garrett / Point Blue Conservation Science
A dickcissel, whose normal range spans from the upper Midwest to northern South America.
A dickcissel, whose normal range spans from the upper Midwest to northern South America. Jim Tietz / Point Blue Conservation Science

On April 25, 2024, for example, a swallow-tailed gull (normally found in the Galápagos) and a blue rock-thrush (normally found in Eurasia) both showed up, and a swallow-tailed kite (which normally ranges from the southeast U.S. to South America) came in the following week. Two of these birds were from the “wrong continent” and “wrong hemisphere,” and one was on the “wrong side of North America,” says Warzybok, who has spent over 2,500 nights on the islands, more than any other living person. Other recent rarities include Nazca booby, Oriental greenfinch, yellow-green vireo, brambling, Cassin’s sparrow, bobolink, gray-crowned rosy-finch, song thrush, crested auklet and yellow-footed gull, none of which generally appear anywhere near the Bay Area. 

A Blue Rock-Thrush on the Farallones, the first record for California. There were only two prior records in North America. This bird’s native range is in Asia extending from northern Japan to Indonesia.
A Blue Rock-Thrush on the Farallones, the first record for California. There were only two prior records in North America. This bird’s native range is in Asia extending from northern Japan to Indonesia. Jason Talbott

Many of the landbirds make their way to Southeast Farallon Island’s four trees: three Monterey cypresses that span the gamut from partly to fully dead and one healthy (though short) Monterey pine. But birds can show up anywhere, including a yellow-billed cuckoo once found perched in a cave. “We’re pretty diligent about searching the island,” Warzybok says. “Of course, there’s no guarantee you’re going to see every tiny bird.”

Still, one reason Southeast Farallon has such a voluminous bird list is that trained scientists are out there every day searching for them. “After we have dinner, we sit around the table and write down all the day’s events,” says Jim Tietz, a Point Blue biologist and another co-author of the Ornithology paper who has been working on the Farallones every fall since 2006. (In addition to observing vagrants, Farallones researchers study and help conserve the islands’ plentiful seabirds, seals and sea lions, and other marine life.)

 Southeast Farallon research station and its songbird-friendly tree in 1968, with a mist net nearby. Researchers have been puzzling for decades over why vagrant birds land there, where they come from, and where they are going.
Southeast Farallon research station and its songbird-friendly tree in 1968, with a mist net nearby. Researchers have been puzzling for decades over why vagrant birds land there, where they come from, and where they are going. Ron LeValley / Point Blue Conservation Science
Another view of the songbird-friendly tree, from 2021.
Another view of the songbird-friendly tree, from 2021. Matt Brady

Where do vagrant birds go after their Farallon adventures? Despite the macabre implications of the “mirror-image” hypothesis, some do make it back to the mainland, like a blackpoll warbler that was banded on Southeast Farallon in September 2025 and spotted 10 days later in Los Angeles. But mostly, their fates are unknown. Tietz says a future paper will examine the direction captured Farallones vagrants head once they’re released. He’d also like to get approval from Fish and Wildlife to install a Motus radio telemetry station for tracking tagged animals. 

Tietz notes that since he started working on the Farallones, the number of vagrants and other migrant landbirds has dropped. Changes to the lighthouse beacon atop Southeast Farallon could have played a role—it now flashes, rather than projecting a rotating beam. But it could also just be that birds are less numerous in general. If certain blackpoll warblers are out there, inadvertently searching for new refuges from us humans, it comes none too soon for a species in rapid decline


Jesse Greenspan is a Berkeley-based freelance journalist who writes about history, science, and the environment.