All About Birds Blog

Home Page



When The Nature Conservancy wanted to put a big bet down on thousands of acres of shorebird habitat in California’s Central Valley, it turned to eBird and its big-data analytics. And the bet paid off—with more than 200,000 birds.

By Gustave Axelson. Originally published in the Autumn 2014 issue of the Cornell Lab’s Living Bird magazine.

Two rice fields in California’s Central Valley—one so dry, the parched ground was cracked in patterns like a shattered windshield. On the other, a couple inches of still water was disturbed by the march of Long-billed Dowitchers, rhythmically pumping and probing like an advancing army of sewing machines.

The dry field is what most of the 500,000 acres of rice country in between the Sierra Nevada and the Coast Range looks like in late winter, as farmers dry out their beds to prepare for seeding in spring. The latter is what just 2 percent of the Central Valley’s rice fields looked like last February, an anomaly holding 2 to 4 inches of water. But it was no accident. The farmer was paid to put the water there, in a wager that the birds would come. And eBirders helped place the bet.


One of the thousands of Dunlins that took refuge in rice fields flooded this past spring by the Bird Returns project. Photo by David Speiser.

In the midst of an epic drought in California, the worst in decades, rice farmer Amelia Harter flooded 170 acres—more than two-thirds of her farm—with a few inches of water, and in the process provided an oasis for these dowitchers and the other 24 species of shorebirds that make great intercontinental voyages along the Pacific Flyway. Before the California Gold Rush of 1849, this valley was all soggy with rivers, creeks, and sloughs that hosted birds by the tens of millions. Now more than 95 percent of the Central Valley’s natural wetlands have been lost. Rice fields represent the next-best surrogate habitat. During duck migration in fall and winter, these rice fields host more than 7 million waterfowl, one of the highest densities anywhere on earth.

But by the end of January, duck-hunting season is over in California, and most farmers pull the plug and dry out their fields. That’s bad news for shorebirds, because their migration peaks from February into spring. Many shorebird species are of conservation concern and appear on the latest State of the Birds Watch List.

“If I go to visit a city on vacation, I don’t want to buy a house. I want to find a hotel room. It’s the same thing here.”

“When you look at the eBird maps, there’s clearly a mismatch,” said TNC California scientist Mark Reynolds. “There’s a lot of shorebird occurrence throughout the valley, but then the satellite imagery shows there’s not a lot of water availability on the ground.”

Last winter Reynolds and TNC California launched a first-of-its-kind conservation partnership with Cornell Lab of Ornithology information science director Steve Kelling and the eBird program to rectify that mismatch. The project combined precision big-data analytics from eBird—fueled by the more than 230,000 birder checklists submitted to eBird from California—with NASA satellite technology and a market-based mechanism to pay farmers to provide shorebird habitat, developed by TNC economists.

How the ebird/Nature Conservancy BirdReturns program works

Click image for a larger version. Graphic by Joanne Avila.

That’s how Harter came to flood her fields for the month of February, as part of a temporary contract with TNC California. “I always see bird watchers driving up and down the roads by the rice fields, looking at cranes and such,” she said. “We’ve always known our farm was good for birds.”

Harter was one of 40 rice farmers enrolled in the pilot program, called BirdReturns. More than 10,000 acres of rice fields were flooded in 4-, 6-, and 8-week contracts. It’s part of an emerging discipline in ecological science called dynamic conservation, where habitat is created over shifting periods of time and ranges of places to meet the changing needs of migratory wildlife. TNC called their flooded fields “pop-up wetlands,” like the trendy ephemeral restaurants that set the culinary scene abuzz.

“If I go to visit a city on vacation, I don’t want to buy a house. I want to find a hotel room,” explained TNC economist Eric Hallstein. “It’s the same thing here. We want to temporarily rent out habitat for these shorebirds, for just a few weeks of the year, when they’re passing through.”

The pop-ups produced some eye-popping results in their first year—surveys on the BirdReturns fields showed more than 220,000 birds representing 57 species, including every migratory shorebird species in the Central Valley. Recorded shorebird densities averaged well over 100 birds per acre in March, 10 times the number of shorebirds found in other areas outside the project.

“A little bit of water goes a long way for shorebirds,” said the TNC’s Reynolds.

A Good Deal for Farmers

Doug Thomas is mighty proud of his family farm at the doorstep of the Sutter Buttes. He watched the day’s first rays of sun bounce off the buttes as he gazed across the glistening water on his rice fields and reflected on how his family makes a living.

“We grow a crop that feeds people from the Pacific Rim to right here at home,” Thomas said, his hands tucked into the pockets of his Carhartt jacket. “If you eat a sushi roll in the United States, you’re most likely eating California rice.” (Buying California rice is one way to support bird-friendly habitat in the United States.)

When a Long-billed Curlew glided overhead, Thomas stopped midsentence to watch. Birds were a big reason he got into rice farming.

An Iraq war vet, Thomas served a tour of duty and then attended the University of California at Davis to earn a master’s degree in avian ecology. He was all set for a career as a waterfowl biologist, but when he and his wife had twins, he suddenly found himself looking for an immediate way to support his new family. His in-laws owned a rice-farming operation and offered Thomas and his wife an opportunity to join the family business.

“The way I looked at it, this was a way to manage 3,000 acres of wetland habitat exactly how I wanted, with no restrictions or government mandates, and make a living at it, too,” Thomas recalls.

Thomas was eager to participate in BirdReturns. On this winter morning in California, he patrolled the dirt embankments separating rice beds to check out the birds using the 170 acres he had flooded for the program.


A Greater Yellowlegs works the mudflats. Photo by Drew Kelly.

“This drought is something else. This place usually looks like Ireland right now, with all the lush, ankle-high grass,” he shouted over his ATV as it kicked up a dusty trail. “But these guys are doing all right.” He stopped beside a shallow pool where about 50 Greater Yellowlegs worked the mudflats, heads down and highstepping on their mustard-colored stilts.

Thomas likes the BirdReturns program because it’s easy. “There’s no fines, no regulatory red tape,” he said. Hallstein, the TNC economist, says the model was designed to be friendly to farmers. Contracts are bid out in a reverse auction, which allows farmers to set their own price.

Thomas says that his affinity for birds aside, the program made fiscal sense for his farm.

“We’re being compensated for taking a risk, sure. By delaying our planting, we’re squeezing our seeding into four weeks instead of eight,” he said. “But from purely a business standpoint, you could not give a rat’s bottom about birds, and you’ll make enough money in BirdReturns to make it worth it.”

It was a good deal for TNC. To preserve shorebird habitat via the traditional, land conservation tools—outright purchase or conservation easement—would cost $3,000 to $4,000 an acre. Through BirdReturns, TNC rented the habitat for a small fraction of that cost.

But there was no guarantee the shorebirds would show up. In that respect, TNC was buying futures on shorebirds, betting that they had rented the rice fields in the right places during shorebird migration. It wasn’t a blind bet, though. They had insider information.

“Something like this has never been done before, where citizen science joins with NASA imagery to provide habitat conservation—at the scale of a rice field.”

“It’s a little like a ‘Moneyball’ approach to conservation,” Hallstein explains, referring to the Michael Lewis bestseller (and subsequent blockbuster movie) Moneyball, about a high-tech new way to choose baseball players. “The same way the Oakland A’s transformed baseball through precision analytics and sabermetrics, we used big-data sets from eBird to apply analytics to selecting farms in locations with a high probability of bird occurrence. The better data allows us to be more efficient and precise in conservation.”

Conservation Meets Cosmos

The data analytics side of the BirdReturns equation resembled something like conservation meets cosmos.

“Something like this has never been done before,” said the Cornell Lab’s Steve Kelling, “where citizen science is joined with high-performance computing and NASA satellite imagery to provide habitat conservation at a very fine scale—at the scale of a rice field.”

animated abundance map for Dunlin in California Central Valley, Pacific Flyway

Orange to white colors show projected abundances of Dunlin as they fill California’s Central Valley in spring and again in fall. By flooding rice fields in spring, farmers north of Sacramento can create extra migratory stopover habitat—in effect, turn their part of the map orange—at just the right time of year. Map by Alison Johnston.

The supercomputing on the eBird side was partially funded by NASA, along with the Leon Levy Foundation and the Seaver Institute. All three organizations saw this project as a way to connect space-age technology with on-the-ground conservation. For eBird, BirdReturns was a golden opportunity to flex its big-data muscles with tangible habitat benefits for birds.

eBird statisticians and computer scientists used the checklists in the database to build predictive models of where shorebirds would be present throughout the Central Valley in February and March. The results were like a weather map for birds, showing where clusters of shorebirds would congregate. These shorebird forecasts, paired with NASA imagery of surface water availability across the Central Valley, made it possible to see where wet habitat was needed.

”This project couldn’t have been situated in a better area for eBird, because the whole region is full of eBird checklists,” said Kelling.

Two of the eBirders sending in those checklists were Linda Pittman and Karen Zumwalt. They were birding some pools just west of Sacramento on a February evening, catching the last volleys of waterbirds over a gravel road before sunset. Both retired, they now bird fulltime. “They call us the ‘valley girls’,” chuckled Pittman, nudging Zumwalt with an elbow.

On this night they had already counted 50 Northern Pintails, 20 Great Egrets, 300 Tundra Swans, and eight Long-billed Curlews when something caught the corner of Zumwalt’s eye.

“Oh, there go some coots!” she said, smiling at the stocky little fellows flip-flopping across the road like snorkelers in flippers. Turns out, of all the 667 species on her life list, American Coots are Zumwalt’s favorite. “I don’t know; there’s something about them. They have cute feet.”


eBirders such as Linda Pittman (at right) and Karen Zumwalt played a vital role, supplying data through their observations that made it possible to put the water where the birds needed it most. Photo by Gustave Axelson.

Together, Pittman and Zumwalt put up nearly 2,000 checklists on eBird last year. When they heard through the Central Valley Birding Club that their sightings were going to be used for a ricefields conservation project, they began plotting daytrips to places that hadn’t received much eBird coverage.

“Suddenly, it’s not just another birding trip. You know you’re contributing to scientific data collection and to actually creating habitat for the birds you see,” said Pittman. “It makes you bird a littler harder. It gives you the urge to cover everything.”

The Central Valley Birding Club organized several birding blitzes to get their 500 members out to target areas in an organized fashion, much like a formal bird survey might be planned. But unlike a professional monitoring program, which might cost several thousand dollars, the eBirders donated their data for free.

“This was the original vision for eBird,” says the Cornell Lab’s Kelling, “Directly applying bird-watcher checklists to the conservation of birds.”

In this short video, TNC staff, Cornell Lab scientists, and farmers talk about how BirdReturns works from each of their perspectives:

Waves of Shorebirds

BirdReturns will be back this fall, but the market environment could be even more challenging. As the drought drags on, farmers may not receive their allotments from local and state water projects. The U. S. Department of Agriculture has forecast a 20-percent decline in California’s rice crop.

Hallstein, the TNC economist, says the market-based mechanism will still be able to put flooded rice fields in the Central Valley.

“Even in a drought, creating wetland habitat is still cheaper for some farmers than others. The reverse auction mechanism allows us to identify those farmers who can create habitat most economically,” he said. “We expect to pay farmers more on average for wetland habitat this year. Higher prices mean we’ll likely be able to pay for fewer acres total. But that’s okay, because we know that any habitat created during a drought year like this one is incredibly valuable for shorebirds, because it may be some of the only habitat available.”

Kelling and the eBird team are preparing new models of bird occurrence and abundance for the Central Valley this fall, as well as all along the Pacific Flyway. In the future, the Cornell Lab plans to reach out to regional and local conservation groups to get timely habitat on the ground for full-life-cycle conservation of species such as Dunlin, which nest in the Alaskan Arctic and winter as far south as Baja California.

Altogether, the Dunlin flocks recorded this spring in BirdReturns numbered more than 20,000 individuals, which represented about 20 percent of the entire Central Valley wintering population taking refuge in the project’s flooded rice fields. About 300 of those Dunlins were pecking around the muddy clumps in one of Amelia Harter’s fields one February afternoon, soon to be joined by a flock of incoming Long-billed Dowitchers, and then Black-necked Stilts.

So it went on throughout the day, wave after wave, an eBird predictive migration model come to life. The shorebirds didn’t know this water was put there for them. But Harter did.

“It feels like such a personal accomplishment to be able to provide this habitat for these shorebirds,” she said. “We’re part of their world here in the Pacific Flyway. I’m just glad we’re able to feed them as they come through.”

(Top image by Drew Kelly: Massive flocks of Dunlins were a welcome sign of success on the BirdReturns fields in spring 2014.)

Posted in conservation | Tagged , , , , , , , , , | Leave a comment

Join Irby Lovette, director of the Fuller Evolutionary Biology Program, Cornell Lab Ornithology, and Fausto Rodriguez, Galapagos Park Naturalist and founder of Galapagos Best, and enjoy their stories of some the wonders they have witnessed on their trips through the Galapagos archipelago. They also present new research findings from their own projects and those of their colleagues, and discuss some of the challenging conservation issues that may change the Galapagos forever.

The talk took place on October 20, 2014. It was part of the Cornell Lab’s long-running Monday Night Seminar series, a tradition established decades ago by Lab founder Dr. Arthur Allen. If you enjoyed this seminar, check this page for our list of future speakers—we’ll note which upcoming talks will be livestreamed—or come visit us in person! If you missed any talks, please see our index of archived livestreamed seminars.

Posted in Uncategorized | Leave a comment

Previous Image
Next Image

By Pat Leonard

The family of cotingas includes some of the brightest, most ear-splitting, oddest-looking, least-understood birds on the planet. Some have bulbous crests, long fleshy wattles, or Elvis-worthy pompadours. Others have shockingly bright feathers: electric blue, deep purple, or screaming orange. How did they get that way? What twisting evolutionary path resulted in the origin of more than 60 species that are so variable you’d be hard pressed to understand how they could possibly be related each other?

Some of these questions can now be answered, thanks to a new evolutionary “tree of life” (technically known as a phylogeny) for the cotingas, which are native to Central and South America. The door is now open to new discoveries about this amazingly diverse group of birds.

“They are so variable that even defining just what a cotinga is has been a difficult question,” says lead author Jake Berv, a Ph.D. student in the Fuller Evolutionary Biology Lab at the Cornell Lab of Ornithology. “Our study provides comprehensive insight into how nearly all the cotinga species are related to each other going all the way back to their common ancestor. No previous attempts to understand the evolutionary history of this group have included genetic samples from nearly all the existing species.”

Berv began sequencing DNA samples and compiling data in late 2010 while working as a lab technician at Yale University with coauthor Rick Prum, a leading expert on cotingas. For species whose DNA had never been sequenced before, the researchers took small tissue samples from the toe pads of museum specimens.


This circular rendition of the “tree of life” shows nearly all the sizes, colors, and adaptations for the continga family of birds radiating outward from the center and its common ancestor. Both males and females are shown except where plumage for both is the same. Illustrations are reproduced with permission from the Handbook of the Birds of the World. Vol. 9. Cotingas to Pipits and Wagtails, Lynx Edicions 2004. Click on the image for a full-screen view.

Understanding how one species is related to another within this group allows scientists to trace the evolution of specific traits and behaviors. For example, Berv and Prum wanted to learn if the evolution of differently colored males and females in this bird group (known as sexual dimorphism) is directly linked to a breeding system in which males have multiple mates (known as polygyny—a fairly widespread practice in some birds, including the familiar Red-winged Blackbird).

Darwin first theorized that the increased pressure of sexual selection in polygynous birds spurred the development of color differences between the sexes. That is to say, in situations where males have the chance to mate with many female birds, there’s pressure for them to step up their game by adopting brilliant colors. This appears to be true for many species—but not the cotingas. When Berv and Prum examined patterns of evolution for these two traits across their new tree of life, there was no significant match. There have been many more evolutionary changes in color than there have been for breeding system. Berv says these traits may be evolutionarily “decoupled” in the cotingas.

However, sexual selection appears to have played a role in the evolution of sexual differences other than color in some cotingas.

“In one case, the Screaming Piha, the males and females look alike but the male sings one of the loudest songs on the planet,” says Yale’s Rick Prum. “That means male-female plumage difference alone is not evidence for sexual selection because sexual selection is also driving other traits such as voice and behavior.”

Though the researchers didn’t find a link between the coloration of males and females and the evolution of polygynous breeding systems, Prum says some form of sexual selection almost certainly did play a role in the eye-popping colors some cotingas display. Some species have evolved a variety of colors seen only in this group of birds, such as the glowing orange cowl of the Andean Cock-of-the-Rock, the fluorescence of the Turquoise Cotinga, or the blue-and-maroon combo of the Banded Cotinga.

“Some cotinga colors are not produced by pigments,” Berv explains. “Some of these birds have evolved cool nanoscale feather structures made with feather protein molecules. The nanostructures scatter light and produce visible color the birds use in sexual signaling.”

Figuring out how these feather structures evolved is an area of future study. In fact, “the sky’s the limit,” according to Prum, who notes other areas of study could include the evolution of elaborate courtship displays, clutch sizes, or the vocal organs that produce the ear-splitting sounds of the raucous bellbirds and pihas. Even the color question can be studied on a much more refined scale.

“Rather than just asking whether males and females of a species are dimorphic, yes or no, we will next measure various color patches and quantify brilliance,” says Prum. “For example, male and female robins look different, so that would be a ‘yes’ for dimorphism. But they’re not as different as the male and female Scarlet Tanager. On the ‘yes-no’ scale they would be scored the same. If we actually measure the colors we’d be able to say ‘Wow, males and females of one species are really a lot more different than the other.’ That would reflect the impact of sexual selection a lot better.”

“One of the biggest analytical differences between what we’ve done and past work is that we used a ‘species tree’ approach, which is potentially more accurate than what is typically applied to genetic data,” Berv says. “We ran our data through more traditional types of analyses as well, and all of them strongly supported a consistent evolutionary ‘tree of life.’ We hope other scientists who are interested in these birds take our phylogeny and do all sorts of great things with it.”

The paper, A Comprehensive Multilocus Phylogeny of the Neotropical Cotingas (Cotingidae, Aves) with a Comparative Evolutionary Analysis of Breeding System and Plumage Dimorphism and a Revised Phylogenetic Classification was published in the journal Molecular Phylogenetics and Evolution.

The research was paid for by W. R. Coe funds from Yale University and supported in part by the facilities and staff of the Yale University Faculty of Arts and Sciences High Performance Computing Center. The DNA work was performed at Yale; Berv did the data analysis and writing at the Cornell Lab, where he is a Ph.D. student in the Fuller Evolutionary Biology Lab.

For more information on bird evolution:

(Photos: Andean Cock-of-the-Rock © Lorraine Minns; Turquoise Cotinga © Juan Carlos Vindas; Three-wattled Bellbird © Andrew Spencer; Red-ruffed Fruitcrow © Daniel Field; Red-crested Cotinga © Glenn Bartley; White-tipped Plantcutter © Daniel Field; Bare-throated Bellbird © Ben Tavener; Spangled Cotinga © Greg Hume; Bearded Bellbird © Steve Garvie; Purple-throated Fruitcrow © Eduardo Iñigo-Elias; Purple-throated Fruitcrow © Daniel Field)

Posted in science | Tagged , , | 1 Comment


By Pat Leonard

If you have an ear for dialects, you can probably take a good guess as to a person’s origins. The same words spoken by people raised in Brooklyn, Boston, and Bakersfield tend to sound different (though still recognizable). The tone and pitch of words are often influenced by the sounds heard and learned from within each individual’s neighborhood and family.

Family and neighbors also play a key role in the songs sung by Western Bluebirds, according to a two-year study published in the journal Animal Behaviour. Conducted at the Hastings Natural History Reservation in Carmel Valley, California, the study found that Western Bluebirds not only share their songs with relatives but with unrelated bluebirds that live nearby—i.e., the neighbors.

The researchers found that, no matter where they were raised, when male Western Bluebirds move to their own territories they share songs with their closest neighbors. That’s true whether the other birds are related or not, though they do a bit more sharing with nearby relatives. On the flip side, they rarely share notes with non-neighboring birds, even if they are related.

“It’s like speaking the dialect of the neighborhood you move into,” says study author Çağlar Akçay who led the research during his postdoctoral work at the Cornell Lab of Ornithology. “We wanted to find out if the birds learn from and share songs only with relatives—in which case the songs would function as a way to identify family—a ‘family signature.’ But if they learned and shared songs with nonrelatives too, then the songs could not stand alone as a means of recognizing kin.”

As a cooperatively breeding species, it’s important for a Western Bluebird to recognize close family. Male Western Bluebirds tend to breed next door to their parents, brothers, and sometimes grandparents, often sharing a boundary with a male relative’s territory. If dispersed males are unable to breed on their own, they will help raise the young in the nest of their parents, a parent and step-parent, a brother, or even a grandfather. Western Bluebird territories are variable but the longest song-sharing distance measured to date at the Hastings Reservation is 2.2 miles (3.6 km). A maximum sharing distance has not been determined.

Though females of the species also sing, they disperse farther away and only one-fifth as many stay in the study area. Also, females don’t sing during the pre-dawn chorus, and consequently they are harder to record. For these reasons, female Western Bluebirds were not part of this study.

Male Western Bluebird songs were recorded for the study during the song-filled early dawn. Because the human ear isn’t up to the job when it comes to noting the subtle differences among the bluebird songs, the recordings were processed through sound analysis software to produce images, or spectrograms, of each note. Three independent judges compared spectrograms visually and agreed 98% of the time on which songs were shared and which were not. (More on how to read spectrograms.)

“On average, Western Bluebirds have eight or nine ways of making their pew sound and a couple variations for their chuck notes,” Akçay says. “These are the building blocks of their song.”

Akçay says it is the “acoustic signature” of each note type that matters for identification—meaning how the sound frequency is modulated. One note might have a very steep increase in frequency while another might be more gradual.


Spectrograms gave the researchers a visual way to compare bluebirds’ songs. Click the player below to hear the song and follow along on the spectrogram. Source: Macaulay Library; recordist: William R. Fish. (Recordings in this blog post were not used in the research.)

Hear the song:



By comparing this recording with the previous one, you can see subtle differences between the songs of Western Bluebirds. The differences are harder to pick out without the help of a spectrogram. Click the player below to hear the recording. Source: Macaulay Library; recordist: Thomas G. Sander.

Hear the song:


It’s not clear what advantage, if any, the bluebirds might gain by sharing songs with nonrelatives but Akçay says a number of theories could be investigated. Studies of other bird species have suggested that shared songs can be used to signal aggression between males, to indicate a male’s fitness for breeding, or as a way to reduce or avoid aggression among birds that are known to each other.

So what happens if a bluebird changes its territory or its neighbors change when new birds move in next door? Preliminary findings suggest the birds don’t completely change their repertoire but they do change.

“We actually had one individual that made a really big move from one year to the next,” Akçay says. “We had recordings of him both years and he kept 80 percent of his songs after the move. The other 20 percent were new songs adopted after the move.”

Akçay concludes, “The major takeaway here is that Western Bluebirds probably identify kin in a linear process—first noting through song that the other bird is a specific individual such as ‘Joe,’ or ‘Fred,’ then further noting, ‘Joe is my brother’ or ‘it’s Fred the neighbor who is not related.’ This second stage of classification may be helped by other cues such as behavior, appearance, and even smell in addition to shared song.”

The study is part of a 30-year research program conducted at Hastings Reservation by Janis Dickinson, who directs the Cornell Lab’s Citizen Science program and is a Cornell University professor. This long-term study, funded in part by the National Science Foundation, has provided a wealth of knowledge about social behavior on a large sample of birds with known social relationships, making these kinds of discoveries possible.

The study, Song sharing with neighbors and relatives in a cooperatively breeding songbird, was coauthored by Akçay and Dickinson with students Katherine L. Hambury (Cornell Lab), J. Andrew Arnold (Cornell Lab and Old Dominion University), Alison M. Nevins (Cornell Lab), all of whom contributed to recording, field experiments, and classification of vocalizations.

More about bird song:

(Top photo: Western Bluebird by Stephen Parsons via Birdshare)

Posted in science | Tagged , , , , , , , | 4 Comments
Older Entries |