Mist clouded my glasses, and a breeze lashed the long, coarse grass on the steep bluff where I sat. The knife-edge headland jutted into the gray Bering Sea, its slopes spangled with the late-summer blue of monkshood flowers Fifty or sixty feet below me, waves washed on a cobbled beach of round black rocks, each surge making the smallest pebbles snicker lightly against on, another.
The tide was half out, and looking through gaps in the fog across the half-mile width of Izembek Lagoon, I could see a delicate ruffle to the surface of the waterthe tips of great eelgrass beds were just dimpling the top, and smooth channels of open water branched through them like roads. In one of these channels, a harbor seal's blocky head emerged from the waves for a brief moment, blinked its eyes, then disappeared. Nearby, two sea otters lay sprawled on their backs, wrapped in the sinuous eelgrass to keep from floating away; one was working at dinner, using its paws and teeth to wrench apart a crab.
Through the fog came a sound, a creaky, three-noted honk Claa-ha-aa! Claa-ha-aa! like someone blowing on a clarinet reed. A small flock of emperor geese broke into clear air, eight of them rowing against the winddark, purplish-black bodies and white heads stained with rusty orange from feeding in iron-rich tundra ponds. As the mist lifted, I saw birds in almost every direction: skeins of small, dark geese called brant, in hurried flight squadrons of eiders and scoters and other sea ducks riding the swells, pudgy afootballs; swirls of kittiwakes, long-winged gulls with balletic grace; and smokelike tendrils against the horizon that were thousands of shorebirds, rising and falling on distant mudflats exposed by the dropping tide.
At the very edge of the continent, I was looking off across the Bering Sea toward hidden Siberia and the Old World. Izembek National Wildlife Refuge lies at the tip of the Alaska Peninsula, where this crescent of rugged land dissolves into the thousand-mile arc of the Aleutian Islands. It is a land of fire and ice, of alpine snowfields and active volcanoes, like one of the rugged cones that flank nearby Cold Bay.
I came to Izembek in early September, after weeks of wandering across western Alaska, because this tendril of land is a global crossroads. Fifteen or twenty thousand years ago, during the height of the last ice age, continental glaciers captured much of the world's water and sea levels were hundreds of feet lower. What is now the shallow Bering Sea was then a marshy land bridge, a port of entry between Asia and America. Across this bridge, named Beringia by geologists, humans wandered from Siberia, hunting mammoths, giant bison, and other ice age game, colonizing a new world.
Beringia was inundated roughly eleven thousand years ago, as the glaciers melted and the seas rose. But Asia and America still nearly touch here, a brushing kiss across the 50-mile-wide Bering Strait, and Beringia is still a way station of international significance. Some of the travelers come by sea: gray whales from Baja, salmon returning to their natal streams from the black waters off Japan and Korea, northern fur seals hauling out on the rocky islands of the Pribilofs. But far more journey by air. Many birds whose travels span the globe breed in western Alaska, and now, as summer faded to autumn, they were taking to the wind once more.
They were not leaving because the weather would soon turn coldalthough it would, the raw, bone-deep cold of coastal Alaska, with its sea ice and wet snow and howling winds. Migration is, fundamentally, about food, not temperature; those birds that can continue to find enough to eat during the winter rarely migratewhy bother?while those whose food supplies are seasonal must flee. Almost all of the more than five hundred North American species that migrate depend on weather-sensitive food suppliesthe ducks and wading birds whose marshes are sealed with ice, for instance, or the insectivorous songbirds that can't find bugs in a December snowstorm. Seed-eaters are less likely to migrate than insect-eaters and tend not to go as far when they do; they can find plenty of weed seeds in North Dakota in January, but for flying insects a bird must travel at least to the Gulf States, or all the way to the tropics.
Behind me, a small cove lay in the protective lee of the bluff. Windrows of dead eelgrass formed thick, snaky ropes at the high-water mark, black against the oily gray of the mudflats. Flocks of shorebirds were feeding there with restless energyscurrying, probing, poking into the rich tidal muck for small invertebrates. Most of the birds were dunlin, small sandpipers with drooping bills, many still in breeding plumage, with reddish backs and black bellies, as though they'd squatted in soot. Every few minutes, responding to some silent signal of alarm, they would leap into the air, wings flashing white, then twist and circle to earth again to resume foraging.
Migration is not the simple, north-to-south-and-back-again affair that most of us assume, and the shorebirds feeding on that mudflat were a perfect example. Dunlin breed in much of Alaska and the Canadian Arctic, but they form three distinct populations, each with radically different migration routes. Those that I was watching were probably of the subspecies pacifica, which nests in this part of southwestern Alaska and travels relatively short distances along the coast, stopping anywhere from the southeastern Alaskan panhandle to Baja California. Dunlin that nest in the central Canadian Arctic, on the other hand, migrate overland to the southern Atlantic and Gulf Coasts of the United States. And those that breed in the Northwest Territories and northern Alaska cross the Bering Strait to Siberia, where they join Russian dunlin and migrate on to eastern China, Japan, and Korea.
Among the masses of dunlin were a number of other species of shorebirds, each laying on fat before departing for far-flung destinations. Rock sandpipers, matching the gray volcanic stone, would barely budge; although some would travel to California, many would pass the wet, gloomy winter right here. Least sandpipers no bigger than sparrows, like little, buffy windup toys, would skirt the coast to South America. Among them were two Pacific golden-plovers, the color of hand-worn brass, which might take one of two routes from Alaska. Some cross the Bering Strait and follow the Asian coast, finally veering southeast to Australia or the Pacific islands. Others fly southwest across the open ocean to Hawaii, and then on to the islands of the South Pacific; some apparently make the flight to the Marshall and Line Islandsa thousand miles south of Hawaii, and a trip of nearly 4,000 milesin a single, nonstop flight. Others on the beach and flats that day would make similar migrationsthe wandering tattler, a whimsically named sandpiper, which winters across much of the South Pacific, and the squat, piebald ruddy turnstone, which travels to Southeast Asia, Australia, and the islands of Oceania, as well as to the California coast.
This single acre of mud held examples of a dozen different migration strategies. Not far from the Pacific golden-plovers stood five American golden plovers, close cousins distinguished (with difficulty) by their duller plumage and somewhat longer wings. They also breed in western Alaska and make equally epic trips, but in the opposite direction. First they fly east across Arctic Canada, feasting on its wealth of berries and insects before gathering in the Atlantic Maritime provinces; then they swing south, cutting across the western Atlantic nearly 2,000 miles to South America, eventually ending their trip in the grasslands of Argentina. Joining them on the pampas would be greater yellowlegs, which I had seen earlier in the day along freshwater streams flowing into Izembek Lagoongraceful sandpipers with steely-gray plumage and colorful, ripe-lemon legs. From the muskeg forests of Alaska and Canada, the yellowlegs spill south each autumn across the hemisphere, but they take an overland route, some stopping as far north as the Pacific Northwest and mid-Atlantic coasts, but others pushing on clear to Patagonia.
Simply cataloging the avian wanderers that pass through western Alaska would be a lengthy chore. Virtually every black brant on earth stops at Izembek in autumn, feeding on the world's most expansive eelgrass beds, before moving out in November for destinations as far away as Mexico. Most of the world's emperor geese and Steller's eiders, the latter a threatened species, congregate at Izembek in the fall; the geese move on to the Aleutians for the winter, but the ducks stay put in the sheltered lagoons.
In Beringia, a naturalist may find Hudsonian godwits bound for Tierra del Fuego and bar-tailed godwits headed for New Zealand; the small greenish songbird known as the Arctic warbler, which migrates to the Philippines, and Wilson's warbler, yellow with a black cap, which flies to Central America. There are fox sparrows and golden-crowned sparrows that winter in Pacific coastal woodlands, and gray-cheeked thrushes that travel to the Amazon. All morning, I had been watching wheatears, black-and-white songbirds that look like slim thrushes. They'll join their Siberian brethren and fly to China and India, then continue on to eastern Africa together, while wheatears from the eastern Arctic swing across Greenland and Iceland to reach western Africa by a European route, embracing the world in a wishbone of movement.
Out on the ocean around Izembek were more migratory wonders. The day before, an Aleutian gale had savaged the region, ripping even protected harbors to foam and fury. But from a sheltered nook overlooking Cold Bay, I had watched dark shapes skimming the waves, gliding on stiff, narrow wings that barely cleared the tops of violent whitecaps, to all appearances blithely unconcerned by the storm. These were shearwaters, fittingly named seabirds that spend virtually their entire lives on the open sea and that are among the world's most accomplished migrants.
This particular species, the short-tailed shearwater, nests on small islands off the south coast of Australia and Tasmania. In April and May, at the conclusion of the breeding season, millions of shearwaters pour rapidly north along the western Pacific rim, taking advantage of the prevailing windsacross Oceania, past Japan, and into the Bering Sea a month later, where they enjoy perpetual subarctic daylight and a rich food supply. They remain off Beringia until September, when they head down the American coast and across the central Pacific, again aided by local wind currents. The shearwaters return to Australia by late October or November, when observers along the coast of New South Wales have seen as many as 60,000 an hour going past, a torrent of birds fueled by the bounty of the Bering Sea. A short-tailed shearwater may cover more than 18,000 miles in a single year, carving a vast circuit on the Pacific, aided at each step of the way by the prevailing breezeand yet they arrive back in Australia within the same eleven-day period each year.
Migrations like this leave us staggered; we are such stodgy, rooted creatures. To think of crossing thousands of miles under our own power is as incomprehensible as jumping to the moon. Yet even the tiniest of birds perform such miracles.
Some days before, four hundred miles to the east, I had been camping along the Alagnak River, a crystalline stream that roars out of the Aleutian Mountains of Katmai National Park. The Alagnak flows through open tundra and sparse spruce forests, its banks wrapped in thickets of alder and stunted birch; several times each day, I saw brown bears lunging into the river for forty-pound king salmon as large and red as fireplugs.
The streamside thickets were full of small birds, streaky and tinged with greenblackpoll warblers, five and a half inches long and weighing barely half an ounce each; you could mail two of them for a single first-class stamp. Warblers are a largely tropical family, either as permanent residents or winter migrants, but the blackpoll is the most northerly of the clan, breeding from western Alaska across the midsection of Canada to Hudson Bay, Labrador, and New England. It is also the most southerly in its wintering grounds, migrating as far as the western Amazon, giving it the longest migration of any North American songbird.
That would be remarkable enough even if the blackpoll took the most direct course. But it doesn't. Instead, like the American golden-plovers, they cross first a continent, and then an ocean. As August wanes, most of the Alaskan birds travel east, across the boreal forests of Canada, all the way to the Maritime provinces and the coast of New England. For a blackpoll born on the banks of the Alagnak, that alone is a journey of roughly 3,000 miles. While some of them then hug the shore toward Florida, it appears that many blackpolls strike out south over the open ocean, departing the Northeast coast at dusk, ordinarily picking a night with a brisk, northerly tailwind after the passage of a cold front.
They will need the help. For the next forty or fifty hours, the tiny songbirds will fly over the western Atlantic, wings buzzing at twenty flaps a second, climbing to altitudes of more than 5,000 feet. They will show up on weather radar as they pass Bermuda and the Greater Antilles, glowing green specks that form diffuse blobs on the monitors, like ghosts beneath the moon.
The warblers follow a curving track, steered and abetted by the wind. At first, the northwesterlies carry them out to sea, the wind's push adding to the 20 miles per hour that the warblers can fly on their own. Midway, somewhere around Bermuda, the northwesterlies fail and the migrants come under the influence of the subtropical trade winds, which blow from the northeast. The tiny birds are shepherded back to the southwest, toward South America, finally making landfall along the coast of Venezuela or Guyana, an overwater trip of about 2,000 milesa passage with no rest, no refueling, no water, during which each will have flapped its wings nearly 3 million times. "If a Blackpoll Warbler were burning gasoline instead of its reserves of body fat, it could boast of getting 720,000 miles to the gallon," note two researchers.
Nor are the birds finished; although some blackpolls do winter in the rain forests of northern South America, others continue south as far as northern Bolivia and western Brazil, another 1,500 miles or so. Then, in April and May, they reverse course, making a less-spectacular but still daunting traverse of the Gulf of Mexico or the western Caribbean and returning to their breeding grounds via the interior of North America. In all, the elliptical round trip for an Alaskan blackpoll warbler may cover eleven or twelve thousand miles.
I have seen blackpolls crowding the spruce woodlands of coastal Maine in late September, the gathered multitudes of the northern woods, and heard their slightly buzzy call notes in the dark as they set off over the sea, staking their lives on an exhausting journey through storm-raked skies. Knowing that some have already come from as far as the Alaska Peninsula is humbling. Yet, to my thinking, the most astonishing of all the migrants that leave Beringia each fall are two kinds of long-legged shorebirds, the bristle-thighed curlew and the bar-tailed godwit, some of which cross not just part of the Atlantic but the entire width of the Pacific Ocean.
The bristlethigh has always been something of an enigma. About seventeen inches long, it looks very similar to a common, widespread shorebird called the whimbrelthe same mottled brown-and-buff plumage, the same strongly down-curved bill and striped head. Differences between the two are subtle enough to give most birders fits: a pale base to the curlew's beak and a more cinnamon wash to its body feathers. Only at very close range can one see the unique feature that gives the bristlethigh its nametufts of long, shiny, hairlike feathers that grow around the upper legs and whose function remains unknown.
Bristle-thighed curlews nest only in a small part of western Alaska, in the heart of Beringia from the Seward Peninsula south through the Yukon River delta. It took a long time to discover that fact. The species was recognized by scientists as early as the 1760s, when it was collected by Captain Cook's expedition, but the first nest was not discovered by ornithologists until 1948, in the remote tundra near the mouth of the Yukon. For another forty years its breeding ecology was essentially unknown, until fieldwork by biologists began in the 1980s. Although a sketchy picture of the curlew's life history has emerged, it remains one of the least-understood shorebirds in the world.
The curlews, biologists have found, arrive on the breeding grounds in May, when pairs set up their territories in areas with shrubby cover, placing their small, bowl-like nests beneath the tangled, protective canopy of dwarf willows. The eggslovely things of warm ocher, blotched with chestnutare laid in clutches of four and incubated for nearly a month, during which time the defensive curlews may actually strike intruding predators like eagles and foxes.
Just a week or two after hatching, the parents lead their small, downy chicks overland for miles, joining other flightless broods on windswept hilltops to form creches, or nurseries, with just a few adults watching over them. The reason for such gatherings, which may include up to thirty curlews as well as golden-plovers, whimbrels, and godwits, is obscure, but is probably related to protection against predators, especially because sharing guard duty allows the adults to feed more often, gaining back the weight they lost during the arduous nesting season.
Not long after the chicks can flyabout four weeks of agethe adult females desert them entirely, moving hundreds of miles to staging areas in the central Yukon Delta, where they will feed, gather in flocks, and prepare to migrate, The males follow shortly thereafter, and the adults gain a precious week or two in which, freed of all family responsibilities, they can focus solely on eating before migration.
In late August and early September, the entire species, first the adults, then some weeks later the juveniles, leaves Alaska and flies southwest, across the Aleutians and over the Pacific. Some will land for the winter 3,000 miles later on Laysan, Lisianski, and other tiny islands in the northwestern Hawaiian chain. Others will push on for thousands of miles more, eventually coming down on a crescent of tiny South Sea islets, from the Caroline Islands in Micronesia through Fiji, Tonga, and French Polynesia. (It was on Tahiti that Captain Cook's crew first encountered the curlew in 1769, and for the next century scientists assumed they were permanent residents of the islands.) Lacking waterproof plumage, they cannot rest on the sea, and there is no evidence that most of them stop off in Hawaii along the way. Some of the curlews may make a journey of more than 5,000 miles in a single flight, spreading out across the Pacific archipelagos.
It is extraordinary to consider what a young bristle-thighed curlew must accomplish, what obstacles it must overcome. Only five weeks out of the egg, it sits with a flock of its peers on the treeless, marshy land of Beringia. Its parents have abandoned it, fleeing to the tropics as the nights turned frosty. The curlew feeds voraciously, gobbling down purple-black crowberries plucked from the tundra and insects that swarm in the shallow ponds and across the hummocky ground. Fat builds under its skin, a thick, yellowish layer along the sides of its body, in the hollow of its neck, and under its wings; each gram of the stuff contains eight times the energy of an equal amount of protein, essential fuel for the trial ahead. The rapidly diminishing amount of daylight triggers changes within its brainan instinctive orientation toward the south-southwest and a growing agitation that sends the flock into frequent short flights, like premigratory drills, during which they may test the winds aloft, searching for the right combination of speed and direction.
Then, one day, it is a drill no longer; the birds vault into the sky, find the right winds, and stream inexorably toward their ancestral wintering grounds. There are no adults to lead them; their parents are long gone, already in the South Sea. It is a comfortable myth that experienced elders lead the flocks of migrating birds southfor curlews, as for most species, the novices are on their own. Instead, they rely on genetically programmed cues. They orient themselves using the sun by day and the stars by night, innately compensating for those markers' changing positions in the sky; they also apparently sense the earth's magnetic field, and may use other clues as wellpolarized light at sunrise and sunset that is invisible to human eyes, perhaps even very low-frequency sound waves generated by trade winds and ocean surf.
For most birds, migration is a leap of blind faith, an instinctive urge over which they have no real control. The curlew does not "know, in a conscious sense, that coconut palms and placid coral atolls await it in Tonga or Fijiit can sense only an urgency to fly in a certain direction for a certain length of time, following a path graven in its genes and marked by the stars.
Some will not make it. They have been born with a faulty navigational sense, their internal compass skewed a few degrees; these most often die lonely deaths in the vastness of the Pacific. Others will not survive because they couldn't lay on enough fat for the long, nonstop flighta fault of their physiology, perhaps, or a failing of their environment, a drought or late frost that reduced the berry crop. Storms will claim some, predators others, weakness and exhaustion and disease and all the other implacable agents of natural selection taking still more.
Yet most will survive, to descend on quivering wings to coral sand beaches and shimmering mudflats beside cobalt waters. Banding studies have shown that bristle-thighed curlews are extremely long-lived, with some surviving more than twenty-three years, and they have exceptionally low adult mortality from year to year. The system, impossible as it sounds to us, works very well for the birds. The South Pacific is as different a place from the clammy, permafrostgirded tundra as it is possible to imagine, yet the curlews slip neatly into the local ecosystem. In the islands, bristle-thighed curlews become opportunistic hunters, using their long, probing bills to capture bugs and spiders, crabs, even scorpions; they are not above scavenging carrion or stealing eggs from under the breasts of incubating seabirds. There are even reports of tool use by the curlews, which may throw small stones at especially obdurate eggs.
The Pacific islands were once an idyllic place for a bird; most of the islands had no naturally occurring mammalian predators, and the curlewsalone among the world's shorebirdsevolved the tendency to molt most of their large wing feathers at once after reaching their wintering grounds, rendering them flightless. That became a disadvantage once the Polynesians arrived, a thousand or more years ago; humans started trapping the curlews for food, and introduced animals like pigs and dogs (and later, rats and cats from European ships) took a toll. The curlews are no longer found in any significant numbers on many of the larger islands, like the main Hawaiian chain, and today's population of roughly 10,000 is undoubtedly smaller than in centuries past, though as far as anyone can tell, the species is fairly stable.
But stunning as the curlew's migration is, the bar-tailed godwit's exceeds it. The same size as the curlew, but with a long, straight, ice pick of a bill, the bartail nests in eastern Siberia and western Alaska. In mid-July, the Alaskan birds leave their tundra breeding grounds and congregate by the tens of thousands along the Alaska Peninsula, in places like Nelson Lagoon and the mouth of the Egegik River, where they gorge on tiny clams pulled from the intertidal mud. They eat so much that the fat forms in thick rollsmore, by proportion of body weight, than any other wild bird, up to 55 percent of their total mass. And then, when they can eat no more, the godwits undergo a remarkable internal change. Their kidneys, liver, and intestines atrophy, shrinking to a fraction of their usual size, a trait scientists suspect many long-distance migrants share. Obese with fuel, freed of the baggage of heavy guts, the godwits are ready for the air.
They wait for wind, for one of the big autumn storms that rip through southwest Alaska every few days in September. As soon as the front has passed and the pounding wind shifts out of the north, they take off by the hundreds or the thousands, aiming south over the open ocean.
The godwits will not stop again until they reach New Zealand, as much as 6,800 miles away, a trip that will last at least four or five days, even with the wind at their backs and their flashing wings pushing them at 45 miles per hour. Robert E. Gill, Jr., of the Alaska Biological Sciences Center, who has been studying godwits and curlews for years (and who has documented much of this remarkable story), believes the godwits' transpacific flight is the longest nonstop bird migration in the worldfor unlike seabirds like terns or shearwaters, which can rest and feed along the way, the godwits (like the curlews) will drown if they land on the ocean.
But wait a minute, the skeptical part of the mind protests, how could this be? How does a migration like this evolve? It's easy to imagine how migratory paths might lengthen on a continentperhaps birds pushing farther and farther north each year as ice-age glaciers retreat, returning in winter to a traditional nonbreeding ground that lies farther away with each generation. And that is precisely how ornithologists have long believed some migratory routes did evolve. But how did a godwit first find New Zealand? And how on earth did a bristle-thighed curlew first find a flyspeck like tiny Rangiroa Atoll in the vastness of the Pacific?
Perhaps by accident. A pioneering bird, like a pioneering human seafarer, has no way of knowing what's beyond the horizon. Experiments have proven that migratory birds inherit a genetic urge to travel in particular directions at particular times of the year, thus retracing a path their ancestors successfully followed for generations. But whenever an animal blunders off in new directions, their biological compass awry, any discoveries they make are pure serendipity.
The bristle-thighed curlew's nearest relative is believed to be the whimbrel, which is found in much of the Northern Hemisphere, including Alaska and Siberia. Siberian whimbrels migrate along the western Pacific rim, down through Japan to Southeast Asia and the islands of Australasia. Alaskan whimbrels also take a coastal route to the Pacific shores of Central and South America. Whimbrels do not regularly migrate straight across the Pacificbut they do show up in places like Hawaii occasionally, the odd vagrant victimized by its own bad sense of direction.
Such misfits would ordinarily be culled out of the populationif you fly blindly into an ocean as wide as the Pacific, your chances of finding land are pretty slim. But every once in a while, one will stumble across an island, the dumb-luck payoff for having a mistuned guidance system. Once, before humans found the islands, this would have been like hitting the jackpota warm, safe place with no ground predators. Some of their progeny, inheriting the gene for the "defective" migratory route, would repeat the journey, although only those capable of making the arduous, nonstop flight would survive to breed. If this transoceanic trip offers them a survival advantage (and a lack of competition on the wintering grounds is a very big advantage indeed), then over time, a distinct population of long-distance migrants arises. If there is geographic or behavioral isolation on the breeding grounds, which keeps them from mingling with others of their kind and diluting the genetic changes, they may eventually become a separate speciesthe bristle-thighed curlew.