The Mystery of the Disappearing Birds

By Lee Sherman Gellatly

THE MYSTERY OF THE DISAPPEARING BIRDS of Panama’s Barro Colorado Island holds all the elements of a detective story.  The “victims” are 45 avian species, once common, now gone. The prime suspects are monkeys, coatis and other midsized furry omnivores whose numbers boomed after the big cats and raptors vanished from the island (another piece of the mystery).

Possible accomplices include cramped habitat, shrinking gene pools, geographic isolation, climatic changes and unforeseen predators. The investigators are Oregon State University wildlife ecologist Douglas Robinson and Ph.D. student Randall Moore, who have painstakingly collected and analyzed a wide array of evidence to solve the scientific whodunit.

To sleuth out reasons for the rapid and widespread species collapse on Barro Colorado (meaning “red clay” for its russet-colored soil), Robinson began making research trips to the island in 1994. Just two years earlier, he couldn’t have imagined his long-term presence there. A brief island stopover with some other scientists had left a bad impression. “I hated it,” he says, sitting in his office in the OSU Department of Fisheries and Wildlife.  “Being in the lowland tropics means you’re hot and sweaty continuously. The trails were steep and slippery; I fell down multiple times. And there were no birds there, it seemed. It was, like, dead. I didn’t really have an interest in Barro Colorado then because it was so depauperate for birds. It just seemed like a really boring place.”

“An ecosystem is a tapestry of species and relationships. Chop away a section, isolate that section, and there arises the problem of unraveling.”

David Quammen, The Song of the Dodo

But that boringness, that emptiness where birds should be bountiful, soon struck him as an intriguing research question: Where were the birds, anyway? What was driving their demise? He dug into the literature. The prevailing idea was the 1980 Terborgh-Winter hypothesis, suggesting that the island’s medium-sized mammals (coatis, howler monkeys, possums, armadillos and sloths, among them) were flourishing after the last pumas and jaguars faded away (big predators need big territories and 3,700 acres wasn’t big enough). This so-called “mesopredator release” means that midsized mammals whose populations once were held in check by bigger carnivores, are feasting on eggs and nestlings in ever-larger numbers. This imbalance in turn spells doom for such species as the great curassow, the marbled wood-quail, the rufous-vented ground cuckoo and the black-faced antthrush, all of them now extinct on Barro Colorado.

The ecological puzzle, however, turns out to have far more intricacies than a single hypothesis can explain. After two decades of collecting and analyzing all sorts of data from population surveys, catch-and-release experiments and time-lapse videotapes, Robinson and his team have turned up some compelling clues that point in unexpected directions.

 

Flooding a Forest

BARRO COLORADO DIDN’T START OUT as an island. Once upon a time it was a mountaintop, rising from the trackless rainforest that carpeted the Isthmus of Panama. Its deep-green slopes hosted pumas and jaguars and more than 200 species of birds — long-tailed birds with comical yellow pantaloons; green-crested birds with blue and purple feathers shimmering with iridescence; diminutive birds with gigantic eyes adapted for foraging in the sunless forest (what Moore calls the understory’s “eternal twilight”). But commerce demanded a shortcut for shipping between the Pacific and the Atlantic. So nearly a century ago, the Panama Canal was dug, the Chagres River was dammed, and Gatun Lake rose up to flood the watershed’s lush valleys. The former hilltop became the largest island in an archipelago of remnant peaks dotting the blue waters, plied day and night by container ships and oil tankers.

Two things happened very quickly on the newborn island. One, species began to go extinct. And two, scientists began flocking there to study tropical flora and fauna. Today, 900 scientists a year conduct studies at the Smithsonian Tropical Research Institute headquartered on Barro Colorado Island. As science writer David Quammen notes in his celebrated 1996 book, The Song of the Dodo: Island Biogeography in an Age of Extinctions: “For most of its insular history, biologists have been on it like flies on a picnic buffet. It has been one of the most closely scrutinized patches of landscape in the American tropics.”

Island biogeography, sometimes called “insular biogeography,” is the study of biodiversity within patches or fragments of land isolated from larger landmasses and habitats. First put forth in the 1960s by eminent Princeton biologists Robert MacArthur and Edward O. Wilson, the theory of island biogeography provides researchers with a model for studying the ways in which island (or patch) size and distance from the mainland can predict the colonization and extinction of organisms — and, ultimately, the biodiversity and species richness of the place.

Because islands and fragments are self-contained, they make perfect study sites for biologists and ecologists like OSU’s Robinson and Moore. It’s like putting a microscope on evolution and watching it in real time. As Quammen explains, “Islands are natural laboratories of extravagant evolutionary experimentation … because their limited area and their inherent isolation combine to make patterns of evolution stand out starkly.” An island gives “clarity” to evolution, he says, because it’s a simplified ecosystem, “almost a caricature of nature’s full complexity.”

Sunken Stumps

WHEN THE CARETAKER at the Barro Colorado field station first saw OSU’s Panamanian research vessel, he chuckled and said, “Parece que la cascara de aguacate” (“It looks like the husk of an avocado”) — in other words, small, thin-walled, as insubstantial as the peel of a tropical fruit. The tiny fiberglass skiff, thereafter dubbed La Cascara de Aguacate, has survived some near-disasters. One day when Moore was at the helm, the boat suddenly plunged into a 3-foot “canyon of water” generated by the “brutal” wake of a passing tugboat. Swamped, the little vessel started to sink. Moore jumped into the lake and grabbed the bow to counterbalance the motor. Then he “frog-kicked” to shore dragging the boat with him. Not until that evening in the dining hall did he learn that Gloria, the island’s resident 12-foot crocodile, was jealously guarding her nest just a few yards down the shore from where he was swimming.

Another time, Moore had motored to one of the smaller islands to survey birds. As he tied up, a troupe of white-faced capuchins rained out of the trees onto the boat. The monkeys were former pets that had been rescued and released on the island. “This is trouble,” Moore thought as the animals scampered around his fragile vessel, chittering excitedly. So he changed his plans. He motored to a nearby islet, tightly secured the boat and equipment against monkey marauders, and returned the next day. After surveying birds inland for several hours, he returned to shore to find that “one of the little rascals” (he didn’t actually use the word “rascals”) had pulled out the boat’s stopcock valve (the drain plug that keeps water from filling the hull). The boat was all but sunk. Moore managed to save it with some furious bailing, stuffing one of his socks into the drain hole for the six-mile chug back to Barro Colorado. Mourning his shredded dry bag and stolen sunglasses, he cast one last glance ashore to see a monkey in a tree chewing on Moore’s red bandana while looking at him with maddening indifference.

Crocs, tugs and monkeys are only some of the hazards the researchers face in the Gatun Lake archipelago. Another one is stumps. Thousands of old-growth trees, the remnants of towering tropical hardwoods now submerged in Lake Gatun, poke up just under the water’s surface. “There’s a tremendous minefield of tree trunks out there,” says Moore.

Besides transporting the researchers among islands, the fiberglass skiff has served as a launch pad for flight experiments. Robinson and Moore wanted to test the idea that certain tropical, understory birds can’t (or won’t) fly across water. So they strung up finely woven nets, called “mist nets,” in the undergrowth to capture low-flying species on the wing. The captives (10 species in all) ranged from probable good fliers to suspected poor fliers.

One at a time, each bird was placed into a carefully designed, shaded box secured in the bow of the boat. Then, at varying distances from shore, the researcher would open the box and watch what happened. Some birds, like the red-capped manakin and the stripe-throated hermit, flew to land, no problem. Others, such as the spotted antbird, made a short aborted flight and plopped into the water, where they were quickly rescued. A few birds even refused to leave the boat, hunkering down in the relative safety of the box.

The experiment showed that “the ability to fly even short distances between habitat fragments varies dramatically and consistently among species of forest birds,” Moore and Robinson write in the journal Ecology Letters. “This variation correlates strongly with species’ extinction histories and current distributions across the archipelago.” In other words, the birds that can’t or won’t fly from the island to the mainland are the same birds that are failing to thrive on their water-bound patch of land.

Caught on Camera

THE GRAINY BLACK AND WHITE VIDEOS, fuzzy as they are, tell an unambiguous and universal story: Ground-nesting birds lose many eggs and nestlings to predators. But Robinson’s research has revealed a surprising twist on the nest raiders of Barro Colorado: They aren’t necessarily the hairy, clawed species the scientists had expected. Instead, the most common culprits in one recent study had scaly skins and forked tongues.

A series of time-lapse videos, captured by cameras hidden in the jungle foliage, show shadowy figures nabbing eggs and chicks from the birds’ carefully camouflaged nests. On their initial viewing of Tape No. 1, Robinson and Moore watched as the skinny arms of a white-faced capuchin reach into the video frame and, with dainty hands, pilfer a pair of pinkish, speckled eggs from a Chestnut-backed antbird’s bowl-like nest. In Tape No. 2, it’s a white-nosed coati (something like a small raccoon) that steals the clutch of eggs. These first two tapes lent credence to the widely held hypothesis that small to medium-sized mammals were the major culprits in the island’s crashing bird populations.

But then came the surprise. The third video revealed a nest robber that wasn’t a mammal at all. It was a snake. The next video, too, showed a snake. So did the next. Eight of the 10 videos, in fact, documented predation by colubrids or “bird-eating” snakes rather than mammals.

“We were the first ones to figure out that snakes were really important predators on the island,” says Robinson. “Since then, a couple of other studies have shown exactly the same thing — 80 percent predation by snakes in Central America. Based on our video work, we didn’t find any evidence that supported the classical view of mesopredator release.”

Designing Deforestation

MORE EVIDENCE IS NEEDED to close the case of Barro Colorado’s disappearing birds. But the investigations of Robinson and Moore (who has finished his Ph.D. and now teaches at OSU), lend further support to worldwide conservation efforts aimed at reconnecting fragmented forests by creating wildlife corridors — or, better yet, avoiding fragmentation in the first place.

“Our work is one piece in a whole body of evidence showing that many organisms — even birds, considered the iconic organisms for mobility — often won’t cross even absurdly small gaps, like a few meters of water or narrow roads and trails,” says Moore. “It’s yet another rock in the massive and growing mountain of evidence that we need to maintain connectivity of fragmented patches of tropical forests.”