The Best American Science and Nature Writing Part 5

The prime hunting ground for goshawks is dense forest, where their jet-fighter flight tactics provide an advantage that keeps out other raptors better suited to open s.p.a.ces. Goshawks prefer habitat with a closed tree canopy, shrubby cover on the forest floor to give prey a false sense of security, and a midway fly zone free of branches. One Minnesota goshawk study showed that they preferred nest trees averaging seventy-two feet tall and forests with 60 percent to 90 percent canopy closure.

This mature woods preference has landed the goshawk in a few spotted owl-type timber harvest controversies. In Arizona and New Mexico, lawsuits were filed in the early 1990s against the U.S. Forest Service to add the goshawk to the endangered species list as a means of halting old-growth logging projects, though the attempts failed due to a lack of evidence that goshawk populations were declining.

In the mid-1990s, the Minnesota offices of the Sierra Club and Audubon filed appeals on timber-harvest proposals in Chippewa National Forest, charging that the USFS hadn't properly considered goshawk habitat needs. Harvest levels in the forest were high in the late 1980s and early 1990s, says the Chippewa wildlife biologist Jim Gallagher, and the harvests were guided by a 1986 forest plan that didn't contain specifics about goshawk management. Gallagher says that the USFS didn't know much about goshawks back then. He says the USFS was aware of only one or two goshawk nests in the entire 1.6-million-acre Chippewa National Forest.

As a result of the Sierra Club and Audubon appeals, the USFS got to work finding out more about goshawks in Chippewa National Foresta"and soon discovered the forest's goshawk population was more abundant than previously thought.

"Researchers at other national forests told me they searched all summer and couldn't find any goshawk nests," recalled Gallagher. But he found a couple of nests just by visiting some proposed timber-harvest sites and being attacked by goshawks. Several more nests were reported to him as eagle or osprey nests but turned out to be goshawk nests. And as Gallagher gained experience, he got better at finding goshawks. After a few field seasons, Gallagher had located twenty-one goshawk nests.

Habitat Challenges.

Maya Hamady's second goshawk nest visit of the day took us to a different nook of Chippewa National Forest. In her DNR truck, we b.u.mped down an active logging road past fifteen-foot-tall stacks of logs that looked like Paul Bunyan's firewood. We stopped at the edge of a pine plantation, where we immediately saw an ominous sign: a red-tailed hawk perched in a tree.

Red-tailed hawks and great horned owls are the archrivals of goshawks; the presence of redtails and great horned owls often results in nest and territory abandonment by goshawks. Redtails and owls are prime predators of goshawks (which lose their acrobatic flight advantage beyond the forest's edge) and their nestlings. More important, they are harbingers of forest fragmentation, which renders goshawk habitat unsuitable.

We walked a trail until red pines mixed with aspens, where sunlight filtered through an enclosed tree canopy and dappled the forest floor. In the fork of an aspen, about fifty feet high, sat a nest of kindling-sized twigs, patched together with mosses and duff. A goshawk emerged from the nest, circled silently overhead, then flew off into the forest, leaving the echo of a fading call, KEE-kee-kee. Hamady said this goshawk might have been a male attempting to draw us away from the nest.

Of the fifteen goshawk territories closely monitored by the DNR Nongame Wildlife Program for habitat requirements, this site has had the most logginga"with 35 percent of the upland mature habitat harvested between 1995 and 2005. Yet this nest has also been one of the most productive, with hatchlings most years since 2003.

Gallagher says the USFS and DNR are smarter about balancing timber harvests with goshawk habitat needs now, compared with the 1990s, when logging activity in federal forests used to take all trees except the actual nest tree. In 2003 the DNR issued a set of state forest management considerations that are sensitive to goshawk breeding territories, including nest areas.

In 2004 the revised Chippewa National Forest plan introduced new rules for timber harvests near goshawk nests on federal land: no logging allowed during the nesting season; no logging allowed in an area of fifty acres around the nest; and only selective logging within a five-hundred-acre area around the nest, such that 60 percent of that area remains in suitable habitat condition (tall trees, partially closed canopy). Since the national forest rules were implemented, Gallagher says timber harvests have not harmed monitored goshawk nests in Chippewa National Forest. And there have been no further appeals from the Sierra Club and Audubon.

"Audubon has been watching to see that [the rules] work," said Mark Martell, director of bird conservation for Audubon Minnesota. "We want to give the system a chance to work. It's premature to call it a success after only a few years. But it seems to be working so far."

Changing Forests.

Despite the success in Chippewa National Forest, there have still been logging-goshawk conflicts in Minnesota. Federal and state guidelines apply only to national and state forests, which hold 60 percent and 15 percent of known goshawk nests, respectively. About 15 percent of known goshawk nests are on private land, and 10 percent are on county forestland. From 2003 to 2007, DNR goshawk nest surveyors found three instances in which nest stands were logged, twice on private land and once on county land. In two of those cases, the goshawks abandoned their territory.

But individual timber harvest-goshawk nest conflicts don't concern Hamady nearly as much as the forecasts for a changing northern forest composition, which could result in less goshawk nesting and foraging habitat over the next two decades.

Goshawks are very particular in choosing nesting areas, seeking mature forest structure with a closed canopy and trees about fifty to one hundred feet tall. In the Black Hills and Arizona, those nesting areas are in stands of ponderosa pine. In British Columbia, goshawks choose Douglas fir or lodgepole pine. And in Minnesota, that prime goshawk nesting habitat tends to consist of the most dominant trees on the northern forest landscape: aspen.

Aspens aren't inherently suited to goshawk nests, but their abundance here makes aspens the most likely trees to provide suitable nesting structure, at least for now. Red and white pines can also provide that structure, but Minnesota currently has far fewer mature pines on the landscape.

Hamady worries that over the next few decades there will not be enough mature aspen forests, or mature upland forests in general, for goshawks. The Chippewa National Forest plan estimates that mature and old forest will decrease from 49 percent of total upland forest to 43 percent within the next ten to twenty years. In state forests, currently over 30 percent of aspen forests are mature, but long-term goals aim for about 10 to 15 percent of aspen stands in a mature or old forest condition. County forests have more early successional forests and could see a steeper decline in mature forest habitat. All federal, state, and county forestlands have a lack of middle-aged aspens (thirty to fifty years old), which would age over the next two decades into mature forest habitat.

"Whether through harvest or natural processes like wind, decay, or diseases, the old aspen trees that currently exist won't last very long," says the DNR forestry policy and planning supervisor Jon Nelson. He notes that aspen is a short-lived species with a life expectancy of about eighty years. "We can't create middle-aged aspen stands out of thin air to replace them," he says.

Hamady says she sees that conifers will replace old aspens in the fifty- to sixty-year forest plans for the Chippewa and Superior national forests. "But old aspens will [be less abundant] in the next twenty years," Hamady says. "What then?" she asks.

To conserve goshawk habitat, the DNR is identifying contiguous, mature forest complexes in the north woods. DNR biologists are working with foresters to coordinate and maintain that habitat on the landscape through extended-rotation forestsa"timber stands left to grow longer than their typical harvest age.

"Coordination among land management agencies to conserve goshawk habitat is a crucial part of goshawk conservation," says Hamady. "It's not about stopping logging, but managing forests so there's a dynamic patchwork of mature forest habitat available. That doesn't just benefit goshawks, but fishers, spruce grouse, boreal owls, and many species of warblers, including ovenbirds, black-throated blue warblers, and Blackburnian warblers."

The Superior and Chippewa national forest plans now contain guidance to manage for larger patches of mature forest. "The goshawk was one of the driving species for this large-patch management," says Gallagher.

Likewise, the DNR's forest management planning process is working to maintain and restore larger patches of older forest, such as long-lived conifers, to provide more mature forest habitat.

Gos in the Hand.

Months after my summer nest visits, I was still mesmerized by the goshawka"its fearless spirit, its regal form. So in the autumn, I went to Hawk Ridge in Duluth to perchance see a goshawk again. Hundreds of goshawks from Minnesota, Canada, and Alaska migrate by Hawk Ridge every fall.

I was in luck. On the October morning of my visit, the biologists at the bird banding station had captured a goshawk. They handed the bird to me so I could admire it for a few moments before its re-lease. A juvenile with a brown-streaked body, this goshawk shifted within my gripa"a bundle of intense power. The eyes were yellow, not yet red, but every bit aflame. The goshawk glared at me like an alpha predator glaring at its prey, like a timber wolf wrapped in feathers.

Upon release, the gos swooped to take a swipe at a bystandera"one last act of pugnacity before winging to wintering grounds somewhere farther south.

DON STAP Flight of the Kuaka.

FROM Living Bird.

IN FEBRUARY, at 37 degrees 12 minutes south lat.i.tude, the sun sets late, but night has fallen and the darkness is thick and close. In the hills to the west I see a few dull globes of light from distant houses. Above, the stars glimmer like chipped ice, but they cast little light, and I would like very much to see where I'm stepping. I am in a shallow pond about a hundred yards west of the Firth of Thames on New Zealand's North Island. And I am up to my shins in mud as slick as lard. Having given up all pretense of grace, I wave my arms about with each step, as if I'm on a tightrope. I hope not to fall into this black goopa"the manure-enriched runoff from a cow pasturea"as two people near me did moments ago.

A few feet away, Nils Warnock, codirector of the Wetlands Ecology Division at California's PRBO Conservation Science, is disentangling a bar-tailed G.o.dwit from a mist net. Hundreds of G.o.dwitsa"large, long-legged, cinnamon-breasted sandpipers with upturned billsa"flew to this roosting site when high tide submerged the mud flats of the firth, where they spent the day feeding and preening. Warnock, who speaks softly and slowly, never seems to be in a hurry, but suddenly he is moving quickly in my direction. He's holding three G.o.dwits, taking them to a site on dry ground where the birds are being separated into holding crates according to s.e.x. But apparently he has taken one too many birds. "Here," he says, handing me a warm bundle with long, kicking legs. "Don't hold him too tightly." I cradle the bird against my body and stumble through the darkness, the G.o.dwit's heart beating like a trapped moth against my chest.

New Zealand is the princ.i.p.al wintering site for the baueri race of the bar-tailed G.o.dwit, a subspecies that breeds in Alaska. The bird's annual journey to these southern lat.i.tudes caught the attention of South Sea Islanders long ago. About 950 AD a group of Polynesians left their homeland, heading out across the South Pacific in seagoing canoes for a land they believed existed somewhere to the south. Their only guide was the flight path of an elegant bird they knew as the kuakaa"the bar-tailed G.o.dwita"which they had observed flying south each year at the same time. Because the kuaka was not a seabird, they reasoned it must go to some land as yet undiscovered.

They were right. The explorers eventually saw a cloud on the horizon that stretched for miles, hanging, they correctly guessed, above a large landma.s.s. So goes a Maori legend describing how their ancestors discovered New Zealand. They were the first humans to set foot on the island they called Aotearoaa""land of the long, white cloud."

Roughly a thousand years later, Nils Warnock and Bob Gill, a wildlife biologist at the Anchorage, Alaska, office of the United States Geological Survey (USGS), followed the flight of a bar-tailed G.o.dwit in a more leisurely fashion. From August 29 to September 7, 2007, they sat in their offices and watched their computers produce a line across a map of the Pacific Oceana"the flight path of E7, a baueri bar-tailed G.o.dwit whose surgically implanted satellite transmitter was sending signals to National Oceanic and Atmospheric Administration satellites 510 miles above Earth.

Prior to her departure, E7 had been feeding voraciously on fingernail-sized clams, worms, and other marine invertebrates on the Yukon-Kuskokwim Delta along with tens of thousands of her kind. On August 29, she probably rose into the air throughout the day with other bar-tails, circling for several minutes, judging the weather, then settling back down. In early evening, a couple of hours before sunset, she rose into the sky and this time kept going, heading southeast. Within a few hours she crossed the Alaska Peninsula and headed out over the North Pacific.

Gill hoped that the battery in her transmitter would last long enough to prove what he had long suspected about baueri's southward journey to New Zealand. Birds that migrate between northern and southern sites in the Pacific Rima"including baueri's close relative, the Siberian-nesting menzbieri race of the bar-tailed G.o.dwita"usually follow coastlines as they travel, stopping at sh.o.r.eline mud flats along the way to rest and feed. To cross the Pacific Ocean, essentially a barren desert to land birds not adapted to landing and feeding on the water, would seem to be a fatal mistake. But rather than fly toward the Asian coast, E7, as Gill expected, turned southa"toward open ocean. Three days later she was still flying above blue waters. Around noon, flying at an alt.i.tude of perhaps two miles or more, she must have had a beautiful view of the Hawaiian archipelago as she pa.s.sed over it four hundred miles west of Honolulu. Two days later, she crossed the international dateline about three hundred miles north-northeast of Fiji. And then on the afternoon of September 7, E7 approached North Cape, the northernmost tip of New Zealand. She adjusted her flight path, and late that evening touched down at the mouth of the Piako River on the Firth of Thames, eight miles from where she'd been captured seven months earlier in the same mud-bottomed ponds where Gill, Warnock, and crew were now trapping more birds. She had flown for eight daysa"nonstopa"covering approximately 7,250 miles at an average speed of nearly 35 miles per hour.

This journeya"the longest nonstop migratory flight doc.u.mented for any birda"seems barely credible. "I just did a talk yesterday for some colleagues at the U.S. Geological Survey," Gill told me not long after E7 had been tracked to New Zealand. "And I showed these graphics of E7's flight and said, 'Okay, the flight is nonstop, no food, no water, no sleep as we know it, flying for eight days,' and there was just this silence in the room, and I could see their minds trying to wrap around thisa"as does mine. I try to be objective as a scientist, but this just..." Gill's sentence trailed off as he seemed unable to summon up the right word to describe his reaction.

Although he may be at a loss for words to describe the wonder of the feat, Gill knows something of what makes such a flight possible. Like sh.o.r.ebirds in general, the G.o.dwit has a sleek, aerodynamic body and long, tapered wings that reduce drag in flight. Its endurance comes from the enormous reserves of fat the bird builds up in the weeks preceding migration, when it gorges itself on marine invertebrates, more than doubling its body weight until, as Gill has said, it looks like the Concorde when it takes off. Burned off dur ing flight, the fat yields more than twice the energy of comparable amounts of carbohydrates or protein. In addition, the G.o.dwit's body undergoes a remarkable change: its intestines and gizzard, which the bird makes little use of during migration, shrink, allowing more s.p.a.ce to store fat.

In his early sixties, with close-cropped white hair, Gill is the senior member of an international team of scientists a.n.a.lyzing the migration of G.o.dwits and curlews as part of a four-year project funded largely by the David and Lucile Packard Foundation. Gill's voice rises with enthusiasm when he speaks of the bar-tails, and here in New Zealand his spirit is all the more buoyant because he's wearing shorts and sandals; when he left Alaska two days ago, it was 10 below zero. Gill, Warnock, and their North American colleagues are working with a number of New Zealand biologists, among them Phil Battley of Ma.s.sey University, as well as a local sh.o.r.ebird expert, Adrian Riegen. It is Riegen's van that Warnock and I are looking for in the darkness as we cross the cow pasture. There I pa.s.s on my G.o.dwit to more experienced hands.

At the side of the van, Lee Tibbitts, one of the USGS Alaska crew, holds each G.o.dwit up into the light from Riegen's headlamp so he can measure the bird's bill with calipers. Female bar-tails are larger than males, and bill length is an easy way to distinguish the s.e.xes. The team hopes to find several males large enough to carry an implanted transmitter so they can track an equal number of each s.e.x. A transmitter weighs 25 grams, and the rule of thumb is that a bird should not carry anything that is more than 3 percent of its body weight.

According to figures gathered over the past thirty years by the International Sh.o.r.ebird Survey, more than half of all sh.o.r.ebird species show evidence of serious decline. Some populations are dwindling slowly, some plummeting at a rate that, unchecked, will lead to their extinction in the not-too-distant future. Of all the taxonomic groups of sh.o.r.ebirds, the Numeniini (G.o.dwits and curlews) are the most threatened. Thirteen species of Numeniini exist worldwide. All but two of thema"the black-tailed G.o.dwit of Europe and Russia and the Eurasian curlew of Europe and Asiaa"have received designations ranging from "critically endangered" to "species of high concern" by one or more world conservation organizations. With this tracking project, Gill and Warnock hope to learn more about the timing and routes of migration for several species of Numeniini, valuable information for future conservation efforts.

Of the species that breed in North Americaa"which also include the bristle-thighed curlew, long-billed curlew, Hudsonian G.o.dwit, marbled G.o.dwit, whimbrel, and upland sandpipera"the bar-tailed G.o.dwit is actually the most populous, with an estimated 1.2 million birds worldwide, 120,000 of which are the baueri race, which breeds mainly on estuaries in western Alaska. More than a million birds seems like a healthy population, but you don't have to look far to find a cautionary tale. One Numeniini, the Eskimo curlew, very likely exists in name only. The Eskimo curlew's population fell sharply in a few decades in the nineteenth century, dropping from hundreds of thousands of birdsa"perhaps millionsa"to a few individuals, the result of indiscriminate hunting and loss of the gra.s.slands that they depended on during migration. Some estimate that anywhere from two dozen to a hundred Eskimo curlews may still be moving between their ancestral wintering sites in the pampas of Argentina and breeding territories in the far north, but this may be wishful thinking. The last confirmed record was a solitary bird shot in Barbados in September 1963.

G.o.dwits and curlews, like most sh.o.r.ebirds, are particularly vulnerable to habitat loss because they congregate in great numbers at a relatively few sites during migration, and many have restricted wintering sites as well. Understanding their migration is vital to any hope of developing effective conservation strategies to halt or reverse population declines. Although studies that began in the 1970s mapped the general migration routes for many sh.o.r.ebirds, it wasn't until birds were captured and outfitted with radio transmittersa"and now satellite transmittersa"that scientists began to see how individual birds used various sites.

Warnock explains one reason this is important: "If you look at two sites, A and B, and site A has one hundred birds on it on day one, and site B also has a hundred birds on day one, and then you go back ten days later and both sites still have a hundred birds, you might think each site is equally important. But you don't know if it's the same one hundred birds or not. Site A may have a different one hundred birds each daya"meaning one thousand birds have used it over a ten-day perioda"while site B has had the same one hundred birds for ten days. Tracking individual birds can indicate how long a bird typically stays at a site. And that can tell you which areas are really most important."

The Numeniini have significantly different migration strategies. Although the bar-tailed G.o.dwit is the ultimate long-distance migrant, the long-billed curlew, which breeds in the western United States, may travel only a few hundred miles from its breeding grounds to wintering sites. The bristle-thighed curlew is an intermediate-distance migrant, moving between Alaska and islands in the Pacific Ocean. Furthermore, the birds do not always follow the same migration paths in spring and fall. The baueri bar-tails, for instance, return to Alaska by a different route, and their northward journey is as interesting to Gill and Warnock as their record-setting southward flight.

When they leave New Zealand in the spring, the birds head to the Yellow Sea. The coastline of this large, relatively shallow body of water between mainland China and the Korean Peninsula has 8,000 square miles of intertidal flats that support more than 5 million migratory sh.o.r.ebirds each year. There the baueri stop for five weeks or so to rest and feed. Then they launch out across the Pacific again on a beeline to their breeding territories in Alaska. Recent surveys have suggested that a great many of the baueri bar-tails stop at one site, the Yalu Jiang National Nature Reservea"450,000 acres at the mouth of the Yalu Jiang River, which separates North Korea from China.

But fast-growing economies in this heavily populated region of the world are in direct conflict with sh.o.r.ebird habitat. Just down the coastline from Yalu Jiang, in South Korea, lies the Saemangeum estuary, a major stopover site for migrating sh.o.r.ebirds, including 30 percent of the world's population of great knots. In 2006, after years of court battles, the South Korean Supreme Court gave the government permission to complete a twenty-mile-long seawall separating the estuary from the Yellow Sea. The area's extensive tidal flats will thereby be "reclaimed" for agricultural land, and the water that remains will become fresh water from the rivers that feed the estuary. In time, the project will drain an estimated 154 square miles of tidal flats. Some of the new land can already be seen on photographs taken from NASA satellites. Where fertile intertidal flats existed, there is now stark white, barren land.

It may get worse. Mark Barter, an Australian biologist who has worked in mainland China, writes that "80 percent of the significant wetlands in east and southeast Asia [are] cla.s.sified as threatened in some way; 51 percent of these are under serious threat." Recently he told Gill of two reclamation projects on the Chinese coast of the Yellow Sea, each larger than Saemangeum. "The coastline of the Yellow Sea is just being a.s.saulted," Gill says. It is tricky for Westerners to express outrage over this, considering that the ground we stand on is often the result of dams, dikes, impoundments, and other contrivances that have destroyed our own wetland ecosystems. "It's like San Francis...o...b..y a century ago," says Gill.

So how safe is Yalu Jiang? How important is it really? How accurate are the surveys? If the need arises to make a case for Yalu Jiang, the bar-tailed G.o.dwits carrying satellite transmitters have provided irrefutable evidence: so far, the majority of satellite-tagged birds have stopped there.

Wetlands are not the only threatened habitat. The windsa"a "habitat" as tangible to the G.o.dwits as mountains, valleys, and wide-open plainsa"will certainly be affected by global climate change. The G.o.dwits "have evolved wind-sensitive migration strategies," Gill says. Presently, the birds must negotiate five different wind systems that rule different regions of the Pacific Ocean. Their departure from Alaska appears to be timed to take advantage of the winds created by storms that the Aleutian Low Pressure System sends to the Alaska coastline on an almost weekly basis during the period when G.o.dwits usually begin their journey. The birds ride the tailwinds from the back side of a storm for the first 600 miles or more, the winds boosting their speed up to 80 to 90 miles per hour.

North of the Hawaiian archipelago the Northeasterly Trades blow toward the southwest. These "quartering" tailwinds (midway between a tailwind and a crosswind) push the birds along, but if they do not compensate for the westward wind flow, they will be blown far off course. The equatorial doldrums, a zone of little or no wind that sailors have always feared, neither help nor hinder the birds. At approximately 20 degrees south lat.i.tude, they face the Southeasterly Trades, quartering headwinds that they must fly against. And then, in the South Pacific, their fat reserves running low, they face more quartering headwinds as they enter a "conver gence zone." In recent years, studies that use computer models to predict how climate change will affect wind systems have come up with different results. But they agree on one thing: the winds will change. Perhaps the G.o.dwits will adapt. Or perhaps, over time, the winds will push them toward new, less agreeable wintering sites.

At one in the morning on the second day of capturing G.o.dwits, I go with Nils Warnock and Jesse Conklin, a Ph.D. student at Ma.s.sey University, to release two G.o.dwits that now carry satellite transmitters. We drive to a deserted beach north of the sh.o.r.ebird center rather than take them back to the crowded ponds. This will be a kind of post-op recovery room, where they will have some peace and quiet as they adjust to their surroundings. When the birds are released, they do not rush away, but stand motionless for a few minutes. Then, slowly, they walk off toward the sound of water lapping the sh.o.r.eline.

With the birds out of sight, we stand for a moment enjoying the night air. I look up, once again drawn to the starry sky. A poor student of astronomy, I'm delighted to recognize a constellationa"Crux Australis, the Southern Cross. It is, of course, one of the most famous formations in the Southern Hemisphere, remarked upon by virtually every early explorer who sailed south of the equator. In a few weeks the G.o.dwits will rise into the air and leave behind Crux Australis for the cold northern skies of Ursa Major. On the tundra of the far north, they will breed and raise young, then move again to their staging grounds to prepare for the long flight southward. The adults will depart first, leaving behind the juveniles. A short while later, the young birds, guided by some deep baueri knowledge of the earth and wind and stars, will set off on a 7,000-mile journey to a place they've never seena"the land of the long white cloud.

MATT RIDLEY Modern Darwins.

FROM National Geographic.

JUST TWO WEEKS BEFORE HE DIED, Charles Darwin wrote a short paper about a tiny clam found clamped to the leg of a water beetle in a pond in the English Midlands. It was his last publication. The man who sent him the beetle was a young shoemaker and amateur naturalist named Walter Drawbridge Crick. The shoemaker eventually married and had a son named Harry, who himself had a son named Francis. In 1953 Francis Crick, together with a young American named James Watson, would make a discovery that has led inexorably to the triumphant vindication of almost everything Darwin deduced about evolution.

The vindication came not from fossils, or from specimens of living creatures, or from dissection of their organs. It came from a book. What Watson and Crick found was that every organism carries a chemical code for its own creation inside its cells, a text written in a language common to all life: the simple, four-letter code of DNA. "All the organic beings which have ever lived on this earth have descended from some one primordial form," wrote Darwin. He was, frankly, guessing. To understand the story of evolutiona"both its narrative and its mechanisma"modern Darwins don't have to guess. They consult genetic scripture.

Consider, for instance, the famous finches of the Galpagos. Darwin could see that their beaks were variously shapeda"some broad and deep, others elongated, still others small and short. He surmised (somewhat belatedly) that in spite of these differences, all the Galpagos finches were close cousins. "Seeing this gradation and diversity of structure in one small, intimately related group of birds," he wrote in The Voyage of the Beagle, "one might re ally fancy that from an original paucity of birds in this archipelago, one species had been taken and modified for different ends."

This, too, was inspired guesswork. But by a.n.a.lyzing the close similarity of their genetic codes, scientists today can confirm that the Galpagos finches did indeed descend from a single ancestral species (a bird whose closest living relative is the dull-colored gra.s.s-quit).

DNA not only confirms the reality of evolution, it also shows, at the most basic level, how it reshapes living things. Recently, Arhat Abzhanov of Harvard University and Cliff Tabin of Harvard Medical School pinned down the very genes responsible for some of those beak shapes. Genes are sequences of DNA letters that when activated by the cell make a particular protein. Abzhanov and Tabin found that when the gene for a protein called BMP4 is activated (scientists use the word "expressed") in the growing jaw of a finch embryo, it makes the beak deeper and wider. This gene is most strongly expressed in the large ground finch (Geospiza magnirostris), which uses its robust beak to crack open large seeds and nuts. In other finches, a gene expresses a protein called calmodulin, which makes a beak long and thin. This gene is most active in the large cactus finch, G. conirostris, which uses its elongated beak to probe for seeds in cactus fruit.

In another set of islands, off the Gulf Coast of Florida, beach mice have paler coats than mice living on the mainland. This camouflages them better on pale sand: owls, hawks, and herons eat more of the poorly disguised mice, leaving the others to breed. Hopi Hoekstra, also at Harvard, and her colleagues traced the color difference to the change of a single letter in a single gene, which cuts down the production of pigment in the fur. The mutation has occurred since the beach islands formed less than 6,000 years ago.

Darwin's greatest idea was that natural selection is largely responsible for the variety of traits one sees among related species. Now, in the beak of the finch and the fur of the mouse, we can actually see the hand of natural selection at work, molding and modifying the DNA of genes and their expression to adapt the organism to its particular circ.u.mstances.

Darwin, who a.s.sumed that evolution plodded along at a glacially slow rate, observable only in the fossil record, would be equally delighted by another discovery. In those same Galpagos finches, modern Darwins can watch evolution occur in real time. In 1973 Peter and Rosemary Grant, now of Princeton University, began annual observations of the finch populations on the tiny Galpagos island of Daphne Major. They soon discovered that the finches in fact evolved from one year to the next, as conditions on the island swung from wet to dry and back again. For instance, Daphne Major initially had only two regularly breeding ground finches, one of which was the medium ground finch (G fortis), which fed on small seeds. When severe drought struck the island in 1977 and small seeds became scarce, the medium finches were forced to switch to eating bigger, harder seeds. Those with larger beaks fared better and survived to pa.s.s on the trait to their offspring.

Another shift took place after a compet.i.tor arrived in 1982: the large ground finch (G magnirostris), which also eats large, tough seeds. For many years the two species coexisted, and in 2002 both became unusually abundant. But then drought struck, and by 2005, only thirteen large and eighty-three medium ground finches remained alive. Remarkably, instead of adjusting to the drought by eating bigger seeds, as they had twenty-eight years before, the surviving medium finches experienced a marked reduction in the size of their beaks, as in compet.i.tion with their larger cousins they struggled to carve out a niche by surviving on very small seeds. A finch with a smaller beak is not a new species of finch, but Peter Grant reckons it might take only a few such episodes before a new species is established that would not choose to reproduce with its parent species.

The variation seen among the Galpagos finches is a cla.s.sic example of "adaptive radiation," each species evolving from a common ancestor to exploit a special kind of food. Another famous radiation took place on a different set of islandsa"islands of water rather than land. The lakes and rivers of Africa's Great Rift Valley contain some 2,000 species of cichlid fish that have evolved from a few ancestors, some in an instant of geologic time. For example, Lake Victoria, the largest of those lakes, was completely dry just 15,000 years ago. Its 500 diverse species of cichlids have all evolved since then from a handful of species of uncertain origin. Like the finches, cichlid fish species have adapted to diets in different habitats, such as rocky or sandy patches of lake beds. Some species eat algae and have densely packed teeth suited to sc.r.a.ping and pulling plant matter, while others feed on snails and have thick, powerful jaws capable of crushing open their sh.e.l.ls. And what gene is responsible for thickening those jaws? The gene for the protein BMP4a"the same gene that makes the Galpagos ground finch's beak deep and wide. What better evidence for Darwin's belief in the commonality of all species than to find the same gene doing the same job in birds and fish, continents apart?

In The Origin of Species, Darwin tactfully left unspoken how his theory would extend that commonality to include humankind. A decade later he confronted the matter head-on in The Descent of Man. He would be delighted to know that a certain gene, called FOXP2, is critical for the normal development of both speech in people and song in birds. In 2001 Simon Fisher and his colleagues at the University of Oxford discovered that a mutation in this gene causes language defects in people. He later demonstrated that in mice, the gene is necessary for learning sequences of rapid movement; without it, the brain does not form the connections that would normally record the learning. In human beings, presumably, FOXP2 is crucial to learning the sophisticated flickers of lips and tongue with which we express our thoughts.

Constance Scharff of the Free University of Berlin then discovered that this very same gene is more active in a part of the brain of a young zebra finch just when the bird learns to sing. With fiendish ingenuity, her group infected finches' brains with a special virus carrying a mirror-image copy of part of the FOXP2 gene, which stifled the gene's natural expression. The result was that birds not only sang more variably than usual but also inaccurately imitated the songs of adultsa"in much the same way as children with mutant FOXP2 genes produce variable and inaccurately copied speech.

Today's Darwins see in detail how pressures such as compet.i.tion and a changing environment can forge new species. But Darwin also proposed another evolutionary driver: s.e.xual selection. In Lake Victoria, cichlid fish have vision adapted to the light in their surrounding environmenta"at greater depths, where available light is shifted toward the red end of the spectrum, their visual receptors are biased toward red light, while closer to the surface they see better in blue. Ole Seehausen of the University of Bern and the Swiss Federal Inst.i.tute for Aquatic Science and Technol ogy has found that male cichlids have evolved conspicuous colors to catch the female eye: typically red nearer the lake bottom and blue at shallower depths. The blue and red populations appear to be genetically diverginga"suggesting they represent two separate species in the making.

If natural selection is survival of the fittest (a phrase coined by the philosopher Herbert Spencer, not by Darwin), then s.e.xual selection is reproduction of the s.e.xiest. It has the delightful effect of generating weapons, ornaments, songs, and colors, especially on male animals. Darwin believed that some such ornaments, such as stags' antlers, helped males fight each other for females; others, such as peac.o.c.ks' tails, helped males "charm" (his word) females into mating. It was, in truth, an idea born of desperation, because useless beauty worried him as an apparent exception to the ruthlessly practical workings of natural selection. He wrote to the American botanist Asa Gray in April 1860 that "the sight of a feather in a peac.o.c.k's tail, whenever I gaze at it, makes me sick!"

His notion of s.e.xual selection was politely ignored by most Victorian opinion, which was mildly scandalized by the thought of females actively choosing a mate rather than submitting coyly to the advances of males. Even biologists dropped the idea for roughly a century, because they became obsessed with arguing that traits evolve to suit the species rather than to suit the individual. But we now know Darwin was right all along. In all sorts of species, from fish and birds to insects and frogs, females approach the males with the most elaborate displays and invite them to mate.

Darwin did not speculate much on why a female would choose an ornamented male. It is a question that still excites biologists, because they have two equally good answers to it. One is simply fashion: when females are choosing gorgeous males, other females must follow suit or risk having sons that do not attract females. The other is more subtle. The tail of a peac.o.c.k is an exhausting and dangerous thing for the bird to grow. It can be done well only by the healthiest males: parasites, starvation, and careless preening will result in duller plumage. So bright plumage const.i.tutes what evolutionary biologists call an "honest indicator of fitness." Substandard peac.o.c.ks cannot fake it. And peahens, by instinctively picking the best males, thereby unknowingly pa.s.s on the best genes to their offspring.

In one of his flights of fancy, Darwin argued that s.e.xual selection might account for human racial differences: "We have seen that each race has its own style of beauty ... The selection of the more attractive women by the more powerful men of each tribe, who would rear on an average a greater number of children, [would] after the lapse of many generations modify to a certain extent the character of the tribe." The jury is still out on that particular idea, but there are hints that Darwin might be at least partly right.

Take blue eyes. Darwin, like many Europeans, had blue eyes. In early 2008, Hans Eiberg and his colleagues at the University of Copenhagen announced that they had found the genetic mutation common to all people having pure blue eyes. The mutation is a single letter change, from A to G, on the long arm of chromosome 15, which dampens the expression of a gene called OCA2, involved in the manufacture of the pigment that darkens the eyes. By comparing the DNA of Danes with that of people from Turkey and Jordan, Eiberg calculated that this mutation happened only about 6,000-10,000 years ago, well after the invention of agriculture, in a particular individual somewhere around the Black Sea. So Darwin may have gotten his blue eyes because of a single misspelled letter in the DNA in the baby of a Neolithic farmer.

Why did this genetic change spread so successfully? There is no evidence that blue eyes help people survive. Perhaps the trait was a.s.sociated with paler skin, which admits more of the sunlight needed for the synthesis of vitamin D. That would be especially important as people in less sunny northern climates became more de pendent on grain as a food source, which is deficient in vitamin D. On the other hand, blue-eyed people may have had more descendants chiefly because they happened to be more attractive to the opposite s.e.x in that geographic region. Either way, the explanation leads straight back to Darwin's two theoriesa"natural and s.e.xual selection.

Intriguingly, the spelling change that causes blue eyes is not in the pigment gene itself but in a nearby snippet of DNA scripture that controls the gene's expression. This lends support to an idea that is rushing through genetics and evolutionary biology: evolution works not just by changing genes but by modifying the way those genes are switched on and off. According to Sean Carroll of the University of Wisconsin at Madison, "The primary fuel for the evolution of anatomy turns out not to be gene changes, but changes in the regulation of genes that control development."

The notion of genetic switches explains the humiliating surprise that human beings appear to have very few, if any, special human genes. Over the past decade, as scientists compared the human genome with that of other creatures, it has emerged that we inherit not just the same number of genes as a mousea"fewer than 21,000a"but in most cases the very same genes. Just as you don't need different words to write different books, so you don't need new genes to make new species: you just change the order and pattern of their use.

Perhaps more scientists should have realized this sooner than they did. After all, bodies are not a.s.sembled like machines in factories; they grow and develop, so evolution was always going to be about changing the process of growth rather than specifying the end product of that growth. In other words, a giraffe doesn't have special genes for a long neck. Its neck-growing genes are the same as a mouse's; they may just be switched on for a longer time, so the giraffe ends up with a longer neck.

Just as Darwin drew lessons from both fossil armadillos and living rheas and finches, his scientific descendants combine insights from genes with insights from fossils to understand the history of life. In 2004 Neil Shubin of the University of Chicago and his colleagues found a 375-million-year-old fossil high in the Canadian Arctica"a creature that fit neatly in the gap between fish and land-living animals. They named it Tiktaalik, which means "large freshwater fish" in the local Inukt.i.tut language. Although it was plainly a fish with scales and fins, Tiktaalik had a flat, amphibian-style head with a distinct neck and bones inside its fins corresponding to the upper and lower arm bones and even the wrists of land animals: a missing link if ever there was one. It may even have been able to live in the shallows or crawl in the mud when escaping predators.

Equally intriguing, however, is what Tiktaalik has taught Shubin and his colleagues in the laboratory. The fossil's genes are lost in the mists of time. But, inspired by the discovery, the researchers studied a living proxya"a primitive bony fish called a paddlefisha"and found that the pattern of gene expression that builds the bones in its fins is much the same as the one that a.s.sembles the limbs in the embryo of a bird, a mammal, or any other land-living animal. The difference is only that it is switched on for a shorter time in fish. The discovery overturned a long-held notion that the acquisition of limbs required a radical evolutionary event.

"It turns out that the genetic machinery needed to make limbs was already present in fins," says Shubin. "It did not involve the origin of new genes and developmental processes. It involved the redeployment of old genetic recipes in new ways."