# News this Week

Science  11 Jun 2004:
Vol. 304, Issue 5677, pp. 1576
1. ANTIDEPRESSANTS AND CHILDREN

# Buried Data Can Be Hazardous to a Company's Health

1. Eliot Marshall

A time-honored way to deal with negative results is to sweep them under the rug. Few ever get published. But New York State Attorney General Eliot Spitzer gave notice on 2 June that he may punish companies suspected of burying clinical data. This drew cheers from researchers who have been campaigning for a public registry of all clinical trial results —whether positive or negative—as well as those opposed to the use of antidepressants in children.

Spitzer filed suit in a New York court charging the U.K.-based drug firm GlaxoSmithKline (GSK) with “repeated and persistent fraud,” alleging that it had promoted positive findings but hadn't publicized unfavorable data on children and adolescents who were treated for depression with its drug paroxetine, also known as Paxil. The company engaged in “illegal and deceptive” reporting, according to Spitzer, by minimizing reports of suicidal thinking among patients and misleading doctors into overprescribing the drug. The suit asks GSK to “disgorge” millions of dollars.

GSK responded with a one-page note saying that it had “acted responsibly” and that “all pediatric studies have been made available … to regulatory agencies worldwide.” The company also circulated a detailed letter it sent out this spring updating physicians on potential suicide risks, a follow-up to a public review organized by the U.S. Food and Drug Administration (FDA) (Science, 6 February, p. 745). GSK's letter says that “it is not yet clear whether antidepressants contribute to the emergence of suicidal thinking and behavior.”

Spitzer's allegations fell like a live grenade among other drug companies, but they had little to say. Industry spokesperson Jeffrey Trewhitt of the Pharmaceutical Research and Manufacturers of America (PhRMA) in Washington, D.C., declined to discuss the potential impact on clinical reporting. But PhRMA is on record, Trewhitt said, in support of the proposal to report all clinical trial results.

The suit has energized the campaign to create a clinical trial registry. “I'm just thrilled with Spitzer,” said one champion of this effort, Drummond Rennie, an editor at The Journal of the American Medical Association. Members of the American Medical Association (AMA) will consider a proposal endorsing a national clinical trials register at their annual meeting 14 to 16 June in Chicago.*

Spitzer's suit also drew praise from critics of psychoactive drug use in children. “I hope this will clean up the mess in psychiatric drugs,” says Vera Sharav of the Alliance for Human Research Protection in New York City.

But some psychiatrists who have prescribed antidepressant drugs in the category that includes paroxetine—the selective serotonin reuptake inhibitors (SSRIs)—fear that this punitive action will muddy the waters. Harold Koplewicz, director of New York University's Child Study Center, says, “I have great respect for Eliot Spitzer … but I'm concerned about this lawsuit” because it will further discourage use of SSRIs to treat depression in children. He's not persuaded that paroxetine is significantly riskier than other SSRIs, such as fluoxetine (Prozac), which is widely viewed as safe. The big issue, as Koplewicz sees it, is that all SSRIs must be carefully monitored in the first weeks of use in all ages because they have a “disinhibitory” effect.

Spitzer's lawsuit focuses mainly on three trials of paroxetine to treat major depression in children and adolescents, all funded by GSK (see table). Only one was published: study 329, led by psychiatrist Martin Keller of Brown University. Keller could not be reached for comment, but Koplewicz, who participated in the trial, said Keller organized it and solicited funding from GSK because he wanted to fill a gap in information on how children respond to this type of drug therapy. Keller's group reported in the Journal of the American Academy of Child and Adolescent Psychiatry in 2001 that, based on a randomized, placebo-controlled trial in 275 people, “paroxetine is generally well tolerated and effective for major depression in adolescents.” GSK circulated these words far and wide.

View this table:

Spitzer's complaint, however, charges that none of the studies produced significant evidence that paroxetine is “efficacious” in treating depression in this age group and that the two unpublished studies “failed to show that paroxetine was more effective than placebo.”

Keller's study identified one serious adverse event (headache) that might have been caused by the drug. But Spitzer's complaint alleges that there were more significant problems. Citing study data, it states that 6.5% of patients on paroxetine showed “emotional lability”—the term that covers suicidal thinking—versus 1.4% for those on placebo. The percentages were also higher for patients on paroxetine in the two unpublished studies, Spitzer charges. And his complaint says that, “combined, studies 329, 377, and 701 showed that certain possibly suicide-related behaviors were approximately two times more likely in the paroxetine group than the placebo group.”

Spitzer's complaint alleges that GSK “repeatedly misrepresented the safety and efficacy outcomes” from these studies in internal memos to its sales force and in medical letters to physicians mailed between November 2001 and January 2003. In addition to holding back negative findings, GSK failed to point out that hints of efficacy in study 329 lacked statistical significance, Spitzer claims.

The apparent conflicts in GSK's data records came to light as GSK applied to drug authorities to win approval for pediatric use of paroxetine. As British, then Canadian, experts pored over the data, they challenged GSK's claims in 2003. The U.S. FDA held a review earlier this year and on 22 March issued a “talk paper” asking SSRI manufacturers to include new warnings that patients on SSRIs should be closely monitored.

In its 2004 medical letter, GSK clarifies two points: Combined data on paroxetine “did not show a benefit for the treatment [of depression] in pediatric patients,” and “the incidence of adverse events possibly related to suicidal behavior” in pooled data on 1100 patients was 2.4% for those on paroxetine versus 1.2% on placebo. But, GSK adds, “no patients committed suicide.”

• *AMA Council of Scientific Affairs Report 10-A-04.

2. IMMUNOLOGY

# Unexpectedly, Ancient Molecule Tied to Asthma

1. Jennifer Couzin

A surprise discovery in mice has linked a mysterious, largely unexplored class of molecules to asthma and may bolster the theory that the respiratory disease is a misplaced reaction to parasites. The molecules, called chitinases, were long considered a primordial response to certain parasites and insects; chitinase breaks down the compound chitin, which is produced in the shells and outer surfaces of these animals.

Because humans don't produce chitin, their half-dozen or so chitinase genes have often been dismissed as relics of evolution, although one has been linked to an inherited disease. Now, on page 1678, Yale University School of Medicine pulmonary specialist Jack Elias and his colleagues have tied a second chitinase, an enzyme called acidic mammalian chitinase, to a classic inflammatory response in asthma.

“I think it's going to open up a whole new area of exploration,” says William Busse, an asthma specialist at the University of Wisconsin Medical School in Madison. “A lot of people would have given up and said, ‘This [chitinase] has nothing to do with anything.’ They pursued it.”

The pursuit, however, came 2 years after the scientists mistook early evidence for something more mundane. Elias, pulmonary scientist Zhou Zhu, and others had noticed crystals in the lungs of mice bred to have an asthmalike disease. They assumed that the crystals were the same as those commonly seen in the sputum of asthma patients. Finally, “someone in my lab said, ‘Let's make sure,’” says Elias. After purifying the crystals, the researchers were astonished to learn that they'd hit on a chitinase.

Intrigued, they began tracing the chitinase path in asthma. First, they induced asthma attacks in their mice and then examined their lung tissue. The scientists found far more of the chitinase there than in the lungs of healthy animals. Then they focused on a class of T cells, T helper type 2 (Th2) cells, that many consider critical to triggering asthma. Th2 cells, Elias's team found, prompted levels of the chitinase enzyme to soar. Interestingly, overproduction of the chitinase seemed to depend on a key Th2 cell protein, interleukin-13 (IL-13). Extra IL-13, common in the lungs of asthmatics, is thought to help spark asthma attacks.

When Elias and his colleagues gave the sick mice a serum that blocked the chitinase, the animals' lung inflammation eased. The drug, however, didn't alter levels of IL-13 and other Th2 cell proteins. This suggests that the chitinase is a domino that falls later than IL-13 and other Th2 molecules when they precipitate asthma. It also suggests, says Elias, that the chitinase may be a new drug target for asthma; Yale has licensed patents on the discovery to MedImmune, a biotechnology company in Gaithersburg, Maryland.

The scientists also studied lung tissue from humans with asthma and from healthy controls. Those with asthma had high levels of the chitinase in their lungs; the chitinase was undetectable in those without the disease.

That pattern reminds Rolf Boot, a biochemist at the University of Amsterdam, the Netherlands, of what he's seen of another chitinase in Gaucher disease, a rare enzyme disorder. Boot and his colleagues, who were also discoverers of the acidic mammalian chitinase that surfaced several years later in Elias's lab, have found that in Gaucher patients, a second human chitinase can reach levels 10,000 times higher than normal. Like acidic mammalian chitinase, the function of this other chitinase is a mystery.

Although the Elias work sheds light on acidic mammalian chitinase, it also “raises a lot of questions,” says Marsha Wills-Karp, who directs immunobiology at Cincinnati Children's Hospital Medical Center in Ohio. It's still uncertain whether the discovery that asthmatics overproduce a chitinase bolsters the popular theory that in asthma patients, the body senses parasites where there aren't any and sends the immune system into overdrive. But if that's the case, says Wills-Karp, chitinase should be among the first dominoes to fall in reaction to something in the environment, followed by overproduction of various Th2 molecules and then an asthma attack. “There may be more to the story, something we're missing,” she says, and she hopes that further study of the chitinase will fill in any gaps.

3. AGBIOTECH

# Report Says India Needs Stronger, Independent Regulatory Body

1. Pallava Bagla

NEW DELHI—A blue-ribbon panel has recommended that India spend $300 million on an autonomous, expert body that would regulate agricultural biotechnology. Such an independent authority would both speed up the approval process and make it more transparent, according to a report delivered last week to the Ministry of Agriculture. But critics say that the small number of genetically modified crops in the pipeline doesn't warrant such a major change in the current system and that the money could be better spent on research to improve existing crops. “Public regard and satisfaction for the regulatory systems currently in place are, to say the least, low,” asserts the task force, which was chaired by eminent agriculture scientist M. S. Swaminathan. The present oversight body has had six chairs in the past 2 years and is bogged down in bureaucratic infighting. A National Biotechnology Regulatory Authority would help restore credibility to the process, the task force argues, as well as spur investment in the field by overseeing a venture fund for new technologies. (By comparison, the biotechnology ministry's annual budget is less than$75 million.) In the interim, Swaminathan says, the government should appoint “an outstanding biosafety and technical expert” to handle genetically modified (GM) organisms.

The task force recommends that transgenic research should not be pursued on high-profile domestic crops and commodities such as basmati rice, soybeans, and Darjeeling tea. The report also proposes a ban on genetically engineered crops from designated biodiversity hot spots. Breeding for herbicide tolerance should be given low priority, it adds, because deweeding provides employment for a large number of landless families. At the same time, the panel believes that derivatives of transgenic crops that have passed muster “need not always be evaluated for biosafety to the same extent again.” The 50-page report also suggests creating a mechanism to segregate, certify, and label GM and non-GM products

The only GM crop now in the hands of Indian farmers is a Monsanto variety of Bt cotton, although research is under way on more than a dozen plants. That modest level of activity suggests that “this is not the opportune time to tinker with the regulatory system,” says Sushil Kumar, a geneticist at the National Centre for Plant Genome Research in New Delhi and a former co-chair of the existing Genetic Engineering Approval Committee. “Do not upset the apple cart.”

The report has drawn criticism from both industry and citizen groups. Although seed companies like the idea of regulating the gene, not the crop, they were hoping the task force would describe each step in the regulatory process. Devinder Sharma, chief of the Forum for Biotechnology and Food Security in New Delhi, worries that giving all authority to one body invites corruption. (The current system has three tiers, with approval required from a different group at each level.)

The new government of Prime Minister Manmohan Singh should be favorably disposed to key recommendations in the task force, which was set up in 2003 by the previous government. Science minister Kapil Sibal has already talked about the need for regulatory reform to attract greater foreign investment (Science, 28 May, p. 1227), and agriculture minister Sharad Pawar has said that the country's agbiotech policy must ensure food security. That is code for increased productivity through genetic engineering.

4. ASTRONOMY

# Spitzer Telescope Weighs and Dates a Primeval Galaxy

1. Robert Irion

DENVER, COLORADO—Astronomers have taken the first true measure of a building block of the modern universe. Combined images from the Hubble Space Telescope and NASA's new Spitzer Space Telescope reveal the mass of one of the most distant galaxies known, as well as the ages of its stars. This cosmic physical exam, announced here last week at a meeting of the American Astronomical Society, means that substantial assemblies of stars had coalesced within just 600 million years of the big bang. That's the earliest galactic benchmark yet seen.

The study is a coup for Spitzer, launched in August as the last of NASA's four “great observatories.” The satellite's ultracold detectors are sensitive to faint wafts of infrared radiation, or heat (Science, 6 December 2002, p. 1870). But most astronomers doubted that Spitzer, with its tiny 0.85-meter-diameter mirror, had enough power to see faint galaxies that existed during the universe's first 2 billion years. “The fact that Spitzer detects this object is a real triumph,” says astronomer Richard Ellis of the California Institute of Technology (Caltech) in Pasadena.

Ellis and his colleagues, led by Caltech astronomer Jean-Paul Kneib, found the galaxy in February by using a natural magnifying glass in space: a massive cluster of galaxies called Abell 2218. The cluster's gravity stretched the more- distant object's light into three red blotches and magnified the image at least 25 times, allowing Hubble to see it clearly.

Features in the galaxy's light showed that it shone when the universe was 750 million years old. But Hubble could see only the ultraviolet blaze of the galaxy's newborn stars, stretched into optical light by the expansion of space. To tell how big the galaxy was and how long its stars had lived, the astronomers needed to detect the rest of the stars.

That's where Spitzer came in. Its eye sees the optical light of mature, sunlike stars, shifted to the infrared as the radiation straggles across the cosmos. In two 40-minute exposures with the infrared observatory, a team led by astronomer Eiichi Egami of the University of Arizona in Tucson resolved the same patches of light Hubble had seen. The red colors suggest that the galaxy's oldest stars—composing most of its mass—are 125 million to 200 million years old, Egami says. That information places the galaxy's birth near when the first organized protogalaxies started to shine, astronomers believe.

The combined amount of light detected by Hubble and Spitzer reveals that the galaxy is 1/200 the size of our Milky Way, at most—making it a respectable dwarf galaxy but a small fry by modern standards. “This is the kind of object that assembled into bigger and bigger galaxies” by merging with similar collections of stars, Egami says.

Spitzer's unexpectedly deep vision bodes well for one of astronomy's main goals, says Caltech astronomer Christopher Conselice, who was not part of the team: “Together, Spitzer and Hubble will be incredible tools to help us figure out all phases of galaxy evolution.”

5. PARTICLE PHYSICS

# Top Quark Tips the Scale for a Heavier Higgs Boson

1. Charles Seife

The mass of the top quark just got a top-off. A new analysis of the fundamental particle implies that it is a little heavier than previously thought. This is heartening news for particle physicists, who are awaiting the discovery of several new particles by the end of the decade.

The top quark is the heaviest of the six indivisible quarks that make up composite particles such as the protons and neutrons at the center of every atom. Scientists can “weigh” the top quark by using particle colliders to smash protons and antiprotons together at enormous speeds and studying the resulting sprays of particles, including those that are generated by top quarks created by the energy of the collision. Earlier data collected mainly at the Tevatron collider at the Fermi National Accelerator Laboratory in Batavia, Illinois, pegged the particle's mass at about 174 billion electron volts (GeV), the energy units that particle physicists prefer.

But this week in Nature, physicists with the D0 experiment at the Tevatron report that a new mathematical technique for analyzing collision data increases the mass of the top quark to 178 GeV. That is less than one standard deviation higher than the previous world-average value, but even such a small difference in mass can have a large effect on the expected heft of as-yet-undiscovered particles, such as the Higgs boson, which is thought to imbue particles with mass. “If you measure the top, you can set indirect limits on where to look for Mr. Higgs,” says Greg Landsberg, co-spokesperson for the D0 experiment. “A 1-GeV shift in the top mass means about a 5-GeV shift in the Higgs mass.”

The new estimate might lessen the anxiety among Higgs aficionados, who were worried because experimentalists had failed to find the Higgs at lower energies. (The higher the mass, the higher the energy of collisions needed to find the particle.) The new analysis raises the favored mass of the Higgs to about 117 GeV, says Landsberg, a mass just barely out of reach of previous experiments.

However, other scientists downplay the importance of such a small shift in value. “From a statistical point of view, it's not really very different,” says physicist Gordon Kane of the University of Michigan, Ann Arbor. Kane says that the new technique “in principle should work better” than previous analyses but that systematic errors might come from unwarranted confidence in underlying computer simulations.

In addition to the new analysis, new data from the Tevatron also indicate a top-quark mass a few GeV higher than previously thought. Despite the Higgs's added heft, theorists expect the Large Hadron Collider, now under construction at CERN near Geneva, Switzerland, to spot it later this decade.

6. ASTRONOMY

# Giant Black Holes Shed Their Dusty Veils

1. Charles Q. Choi*
1. Charles Q. Choi is a freelance writer in New York City. With additional reporting by Robert Irion.

Some of the most dazzling objects in the universe are forever hidden from view, teams of astronomers in Europe and the United States have concluded. Working separately with data from orbiting observatories, the astronomers deduced that fiercely energetic glowing cores of distant galaxies—beacons of energy powered by black holes more massive than a billion stars—may be four or five times more numerous than astronomers had recognized.

“We had not identified these supermassive black holes even in deep x-ray surveys, because they are likely shrouded by gas and dust,” says astronomer C. Megan Urry of Yale University in New Haven, Connecticut. But in new infrared images from the Spitzer Space Telescope, says Urry, “these very distant, luminous objects are easily visible.”

Supermassive black holes nestle at the hearts of active galaxies, where their intense gravity sucks in matter and converts it into radiation. Astronomers study them for clues to galactic evolution. Unfortunately, the black holes also attract spiraling rings of dust that, when seen edge-on, block astronomers' view of the galactic core. Such bedimmed galactic nuclei are known as type 2 sources.

Many type 2 sources have been found in our celestial neighborhood, but most give off relatively modest amounts of power. Theorists believed more intense type 2 sources must exist in distant, younger reaches of the universe, but only a few had been spotted.

European researchers led by Paolo Padovani of the European Southern Observatory (ESO) in Munich probed the so-called GOODS fields: patches of sky that Spitzer, the Hubble Space Telescope, and the Chandra X-ray Observatory had studied in a joint effort known as the Great Observatories Origins Deep Survey (Science, 19 March, p. 1750). The team took its data from Europe's 2-year-old Astrophysical Virtual Observatory, which gathers and analyzes public data from astronomical facilities worldwide and makes them available to astronomers at the click of a mouse (Science, 14 July 2000, p. 238). Each telescope had spotted features of high- powered type 2 sources, such as high x-ray output or scant emissions of visible light. Pooling the observations revealed what appear to be 30 new high-powered type 2 sources. Until now, only nine had been seen in GOODS fields and at most 10 to 20 elsewhere, says Padovani, who is a co-investigator on GOODS. The U.S.-led study found the same sorts of objects about a month earlier, Urry said last week in Denver at a meeting of the American Astronomical Society.

The results help “balance the books” between the projected population of active, star-forming galaxies in the distant universe and the population of black holes, says astronomer Alan Dressler of the Carnegie Observatories of Washington in Pasadena, California.

7. EDUCATION RESEARCH

# Meager Evaluations Make It Hard to Find Out What Works

1. Jeffrey Mervis

A popular buzzword in U.S. education these days is discovering “what works.” The Education Department even funds a “What Works Clearinghouse” on programs ranging from teaching math to reducing schoolyard violence. This heightened interest in assessment stems from the massive 2001 education reform bill—known as the No Child Left Behind Act—which requires school districts to offer programs shown to be effective through “scientifically based research.” But there's a dirty little secret behind that requirement: No program has yet met that rigorous standard, because none has been scientifically evaluated and shown to be effective. (A related secret is that there's no consensus on the type of evaluation studies that are needed.)

Two reports issued last month highlight the problem—and suggest ways to bolster the embryonic field of evaluation research. One, by the National Research Council (NRC), examined evaluations of 19 elementary and secondary school mathematics curricula and found them wanting.* The second, from a public-private consortium known as Building Engineering and Science Talent (BEST), did the same for programs aimed at increasing the number of minorities, women, and low-income students studying science and math (bestworkforce.org). Again, none of the programs could claim to be successful based on objective assessments.

“The sad reality is that we really don't know what works. And that leaves people in the lurch,” says William Schmidt, a member of the NRC panel and a mathematics educator at Michigan State University in East Lansing. Schmidt, who is also U.S. coordinator for an ongoing study comparing student math and science achievement in 40-plus countries, says that conducting high-quality evaluations is possible, “but it takes a real national commitment. And money.”

Both reports emphasize that many programs may be doing a terrific job of helping children. It's just that there's no way to tell, scientifically. “Evidence matters, … [but] reliable empirical evidence is hard to find,” notes the BEST report, which compiled material on some 200 programs before deciding that only 34 had been studied with sufficient rigor to justify further scrutiny. Of the 20 that it was able to rate, none met BEST's gold standard: five studies conducted by independent evaluators that showed substantially positive results.

The NRC study took a different approach but reached similar conclusions. Its survey of the literature identified 698 studies on 13 math curricula developed by the National Science Foundation (NSF) and six by commercial publishers. But only 147 (21%) met the panel's criteria for weighing effectiveness. It divided those studies into four types of evaluations before concluding that “the corpus of studies does not permit one to determine the effectiveness of individual programs.” In other words, there weren't enough good data to draw definitive conclusions about their value in the classroom.

The most surprising message from the two studies may be that evaluation experts aren't surprised by the results. “The entire discipline of rigorous evaluation is just emerging,” says Judith Ramaley, head of the education directorate at NSF, which funded the NRC study and helped launch BEST in 2001. “But the good news is that the study provides us with a wonderful user's manual for how to go about our business. And it comes at a time when we are finally capable of doing this type of evaluation.”

Before that happens, however, experts need to agree on what constitutes a rigorous evaluation. The NRC panel devoted most of its 212-page report to that topic. The problem is complicated by the many factors that influence student achievement: students' previous knowledge, their teachers' quality of training, the level of resources available, the degree of parental and community support, and so on. “This has never been done before,” says panel chair Jere Confrey of Washington University in St. Louis, Missouri. “People may think that it's a simple problem, but it's really quite difficult.” The increased reliance on tests to determine the fate of students and their schools has increased pressure on educators and evaluators to get it right, she notes.

Evaluation researchers are moving slowly toward the type of evidence-based standards used in biomedical research, says Carlos Rodriguez of the American Institutes for Research, a Washington, D.C.-based nonprofit that worked with BEST to define principles for both designing an effective program and assessing its performance. But there are limits. “NSF is now asking for multivariate and controlled studies,” he says. “But you have to remember that we are measuring human behavior, and that's hard to quantify.”

Controlled studies are especially difficult in an educational setting, points out Mary Catherine Swanson, whose Advancement via Individual Determination (avidonline.org), a national program for at-risk students, was praised by BEST for “notable effectiveness” but faulted for the absence of studies involving students not in the program. “It's hard to exclude people from a program that they think is working,” she says. Swanson thinks that BEST is correct in setting the bar high, however, and she hopes that a recent expansion of the 24-year-old program to Canada—where educators plan to follow students eligible for AVID but unable to work it into their schedules—will provide clearer signs of the program's effectiveness.

Once more rigorous evaluations are in place, says Rodriguez, educators and the public must be willing to resist the temptation to oversimplify the results. “There are no magic bullets,” he says. “We want to find what works most of the time for most students. But we still have to adapt programs to fit the population being served.”

• *On Evaluating Curricular Effectiveness, NRC, 2004.

8. ROGER PAYNE PROFILE

# A Toxic Odyssey

1. John Bohannon*
1. John Bohannon is a writer based in Berlin.

Roger Payne's discovery of whale song helped make the animals icons of conservation; now he's helping turn them into symbols of how humans are poisoning the oceans

THE INDIAN OCEAN, 2°N, 72°E—Seven days have rolled by without a sighting. Although the waters over these deep ocean trenches east of the Maldives are a well-known feeding ground for sperm whales, the crew of the Odyssey has seen nothing larger than a pod of playful dolphins, riding the ship's bow wave and flinging themselves into the air like Chinese acrobats. A man with silver flyaway hair steps out from the pilothouse and squints up at the observation deck, where two crew members are scanning the horizon. “A gold doubloon for the first man to spot a whale!” he booms, imitating Captain Ahab from Moby-Dick.

Like Ahab, Roger Payne has been plying the seas in a tireless search for sperm whales, the largest of the toothed whales. Unlike Ahab's ship the Pequod, however, the Odyssey is equipped with tissue-sampling crossbows instead of harpoons and a toxicology lab instead of blubber-boiling tryworks.

Payne is best known for revealing that the unearthly vocalizations of humpback whales are structured like songs, a discovery that made the cover of Science in 1971 (16 August), and for his hypothesis that the sounds of fin and blue whales carry information clear across the oceans. But the focus of his research out here is pollution—specifically, the class of humanmade chemicals known as persistent organic pollutants (POPs), which can sabotage biochemical processes by mimicking hormones. Some scientists fear that these compounds would become so concentrated in marine ecosystems that fish stocks would be rendered too toxic for human consumption. “No one knows how polluted the oceans are,” says Payne, “because no systematic, global study has been made.”

To try to plug that gap, Payne has assembled a 12-person crew of scientists and educators to circumnavigate the globe for 5 years aboard the Odyssey, a 28-meter floating laboratory. Because POPs become increasingly concentrated at higher levels of the food chain, Payne's strategy for determining the extent of pollution is to measure contamination in the blubber of sperm whales—a top predator found in all oceans. After 4 years at sea, his team's early results are looking grim: The chemicals have accumulated in the fat of every sample analyzed so far, even from places farther from land than anywhere else on Earth. The presence of POPs, particularly in coastal environments, has long been known, says John Stegeman, a toxicologist at the Woods Hole Oceanographic Institution in Massachusetts, but Payne's study is crucial because it will provide a global snapshot of the extent of contamination.

To take such samples, of course, the intrepid researchers must first find whales. After this unusually long hiatus of 7 days staring down the blue rim of the horizon, the crew is getting antsy. But just when it seems that all the whales have fled for the poles, the hydrophone speakers erupt with clicks. “We've got whales!” says Payne with a grin. The glowing dots on the computer screen show a group of 20 sperm whales feeding just ahead. Bowls of cereal and cups of coffee are left half-full as the crew springs into action and the Odyssey surges forward.

## A life among whales

For Payne, this voyage is part of a personal odyssey that began exactly 50 years ago when, as a “green” Harvard freshman, he was asked to baby-sit some bats. These were the furry research subjects of the late Donald Griffin, the renowned biologist who discovered echolocation. Up to this point, Payne says he “had no idea you could make a living doing something like biology.” After Griffin became his undergraduate research mentor, Payne became “obsessed” with bioacoustics. He went on to do his Ph.D. research on how owls use sound to locate prey, and during his postdoc he showed how moths use sound to evade predators. “The secret of Roger's success,” says Thomas Eisner, an entomologist at Cornell University in Ithaca, New York, who helped supervise Payne's Ph.D., “is his creative, playful mind and infectious passion.” Eisner recalls Payne constantly zipping around campus on his bicycle wearing a cape and a cello on his back.

Then, Griffin called on Payne once again. Griffin had moved to Rockefeller University in New York City, and he persuaded Payne to join him there to pursue something new. It was the late 1960s, recalls Payne, and “I felt that what I was doing was not relevant to the problems that assailed the world around me. The wild world I loved above all else was being destroyed. Then I thought of whales.” Commercial whaling was then in its heyday, with tens of thousands of whales slaughtered each year. “I thought, If I studied whales, maybe I could find a way to change their fate.”

A big question at the time was “just what whales are doing with such big brains.” Payne suspected that they were using them to process and communicate complex sounds, but “I had never even seen a whale.” So, he bought a ticket for Bermuda, where humpback whales were known to pass in their migrations. There he met Frank Watlington, a Bermudian engineer studying underwater sound for the U.S. Navy. The Navy was interested in listening for Soviet submarines, but in his spare time, Watlington had also recorded hours of bizarre underwater sounds that Payne later confirmed were coming from whales.

Payne spent weeks wearing a pair of headphones, listening over and over to one tape in particular. The whales were producing very complex vocalizations—traversing eight octaves in pitch from deep, organlike rumbles to high-frequency, flutelike glissandos—but there just wasn't any obvious structure. Then it suddenly came together. He noticed that the entire performance repeated after about 15 minutes. Payne was ecstatic. It isn't just random, he thought. “These are repeated, rhythmic sequences: They're songs!”

Payne and whale advocate Scott McVay, then an administrator at Princeton University in New Jersey, later demonstrated the song structure together. This was long before the days of home computers. McVay had access to a machine that transcribed sounds onto paper with a stylus. Sure enough, the repeating structure was plain to see. The “pioneering, detailed studies” were among the first to show “the significance of sound to marine mammals,” says John Hildebrand, a bio-acoustics researcher at the Scripps Institution of Oceanography in La Jolla, California.

Just what exactly the whales are saying to each other is another question, one that Payne and others have pursued ever since. He later proposed that the sounds made by two other very loud-voiced whales, the blue and fin whales, can carry clear across deep ocean basins. Payne's working hypothesis is that the songs of humpback whales are sung by males to attract females and threaten other males; the same may be the case with the songs of blue and fin whales. But he thinks that evolution has also endowed these whales with the ability to share information across entire oceans, perhaps to clue each other in to the whereabouts of prey in the ocean's ever-shifting feeding grounds.

Since then, says Christopher Clark, a marine biologist at Cornell, Payne has been an “inspiration” to a generation of whale researchers and has invented many standard techniques, such as identifying individual whales by photos of their natural markings.

The discovery of whale song not only propelled Payne's career but lifted whales into the public gaze: The “Save the Whales” movement was born. “Most people's idea of whales was limited to Moby-Dick,” recalls Payne. So he teamed up with poets, musicians, and “anyone he could get hold of” to generate sympathy. The haunting songs beguiled the public: A vinyl record of Payne's humpback tapes, included in the December 1976 issue of National Geographic, is still the largest single print order in the history of the recording industry. Payne went on to host, write, or direct several award-winning documentaries about whales for television as well as an IMAX film.

Governments responded to the surging interest by creating the largest animal sanctuaries on Earth—the Indian Ocean north of 55° south latitude in 1979 and later most of the Antarctic Ocean, both of which remain no-whaling zones. Then in 1982, a worldwide moratorium on commercial whaling came into effect.

For his championing of whales, Payne has been showered with honors, including a MacArthur Fellowship, appointment as one of the “Global 500” by the United Nations, and even a knighthood from the prince of the Netherlands. But saving the whales has also come at a price. In 1985 he left Rockefeller to devote himself to the Ocean Alliance, which he had incorporated in 1981 to promote the preservation of the ocean. The move later seemed like “one of the bigger mistakes of my life,” he says. Keeping the nonprofit afloat “has left me no more than 10% of my time for basic research, which is the love that got me into all of this in the first place.”

But here aboard the Odyssey, among the whales again, Payne is in his element. He first had the idea for the voyage 26 years ago. To make this possible, Payne and the Ocean Alliance raised $3 million from private donors and foundations, about$750,000 short of what is needed to complete the voyage. “It's taken a long time to realize this dream,” he says, “but it's finally come true. It's so good to be doing research so far from land.”

## Canaries of the sea

“We've got a blow at 1 o'clock, about 300 meters!” crackles a voice over the radio from the observation deck. In the distance, a white plume floats over the water like a puff of steam. Moments later, a second plume, smaller than the first, jets into the air. “Make that two blows.” Bob Wallace, the ship's engineer at the helm, closes in and then cuts the engines. Odyssey glides quietly closer to the pair.

Payne may be 69 years old, but he's the first to run out to the very front of the boat and clamber onto the bowsprit. This puts him in a precarious spot far beyond the safety of the deck, rising and dipping over the waves, but it affords the best view. At first they look like a couple of gray barrels floating in the distance, but once the Odyssey is less than a boat-length away, the whales' full, submerged forms take shape beneath the chalky blue surface: The adult is as big as a school bus, while the juvenile is about half that size. And these are small by sperm whale standards. A full-grown male can reach 18 meters in length and weigh over 50 tons. The only part of each whale above the water line is the tip of its massive, squared-off head, exposing the single S-shaped blow hole offset to the left side. The two whales seem supremely calm, taking a breather before swimming down for more squid.

Just when the whales loom close enough for the crew to see the sky reflected off their glistening skin, they begin to roll into a dive. Their heads disappear as their enormous backs, covered with crenulations like giant sea prunes, breach the surface, followed by their tails in a graceful arc. Before the whales' door-sized flukes slap the water for the plunge, Rebecca Clark, a Canadian research assistant who has been aboard Odyssey for the whole voyage, takes aim with a crossbow from her perch on the bow. The orange-feathered arrow sails down and with a sharp “pock!” bounces off the adult's flank. This is no ordinary arrow: Its tip forms a tiny cylinder like an apple corer, gouging a plug of skin and blubber the size of a pencil eraser before springing back out.

The arrow is scooped up with a net and received with a cheerful “Thank you!” by Veritee Steptoe, an Australian research assistant leaning out of the pilothouse and wearing a white lab coat over her pink and neon-blue surfer shorts and tank top. Payne spots the next one: “Another blow at 3 o'clock!” he calls from the bowsprit, and Odyssey's engine rumbles to life.

Long after the last whale is sampled and the crew has settled in for the night, Clark and Steptoe's work is just beginning. After dissecting the samples, they must parcel them into a range of solutions and file them in the freezer. All the samples will be shipped back for analysis and archiving to Celine Godard, a toxicologist at Woods Hole who directs Odyssey's pollution research.

A picture of the extent of pollution is beginning to emerge from Odyssey's “mountain” of raw data, says Godard. Every whale sampled and analyzed from the Pacific—even those in the most remote waters—has polychlorinated biphenyls and DDT pesticide in its blubber. The whales' meals of fish and squid are almost certainly the source of the contamination, Godard says. “It's sobering to discover that these toxicants are distributed globally,” says Payne, “although how dangerous a threat they pose at low concentrations is not yet known.” “People are largely unaware of POPs,” adds Sylvia Earle, an oceanographer and former chief scientist of the U.S. National Oceanic and Atmospheric Administration, who says that Payne's work could be a well-timed “wake-up call.”

Measuring the chemicals in whale blubber is only the first step, because although the toxicants accumulate during the lifetimes of whales, they do undergo some metabolism. We need a way, explains Godard, to measure lifetime exposure to contamination, because this is the best way to use whales as measuring sticks of global pollution. To do that, the Odyssey team is fine-tuning a molecular test that measures the amount of an antitoxicant protein called CYP1A1 that accumulates in response to contamination. Because this test requires fresh samples, it has to be performed immediately. So as the small hours of the morning roll by, the only light on Odyssey comes from the lab where Clark and Steptoe are toiling away. Samples are carefully sectioned into minuscule chunks and soaked in solution. Eventually the CYP1A1 data should add up to a tool for estimating exposure, turning whales into pollution detectors.

While they're toiling away, Steptoe and Clark must attend to their share of the ship's “glamorous” duties: keeping watch to avoid collisions, preparing meals for the 12 people onboard, and cleaning the toilets. “We've been very lucky,” says Steptoe, because the weather has been clear and the boat's rocking gentle. But when the seas are rough, “you need five hands” to juggle chemicals and samples. “Plus you worry about things taken for granted in labs back on land,” adds Clark, such as making sure the generator is still powering the freezers and checking that none of the data sheets has blown overboard. “But as long as everyone knows what job they're supposed to do, it all works like a well-oiled machine,” says Steptoe. “Well, most of the time.”

## Wired for whales

As the whales vanish into the dark depths, Payne turns to Chris Johnson, the Odyssey's media producer, who is leaning over the railing with a video camera like a paparazzi. “Wow. Did you get that?” he asks. In the coming days, the images will be seen by thousands of people tuned in to Payne's other mission: to “share what we learn out here in real time.”

Every day Johnson has kept fans abreast of the voyage—posting everything from scientific reports and videos on conservation threats to interviews with local scientists (see www.pbs.org/odyssey). “Following the whales around the equator has brought us to the most isolated island nations on the planet,” says Payne, “and what we're finding is that their unique, fragile ecosystems are under constant threat from industrialized countries that want the right to fish, whale, and develop luxury resorts.” Most islanders “don't realize what they have to lose by granting these rights,” he says. “Since they rarely have Internet access,” says Genevieve Johnson, the ship's education officer and the other half of this husband-and-wife team, “we have to take this information to them.”

The Johnsons practice what they call “guerrilla education,” having so far come ashore in 14 countries to date and given multimedia presentations about marine resources to more than 20,000 local people. They leave free educational packets with schools and even bring children and teachers aboard for days at a time to show them how marine science is done.

Odyssey's educational mission is leaving significant changes in its wake. After a workshop organized by Godard and presented to the environmental minister of Papua New Guinea in 2001, the government designated its entire exclusive economic zone—a 2.8 million square kilometer area—as a marine sanctuary. Odyssey's next whistle stop is the Mediterranean this summer, one of the most polluted and populous bodies of water in the world. “We want to light a fire there,” says Chris Johnson, “and really get people thinking about the environment that's right on their doorstep.”

Odyssey cuts an arc through the water, guided by a new cluster of underwater clicks from feeding whales. Payne watches the members of his crew with obvious pride as they scramble to their stations. “All this hard work is going to pay off when the results are published,” he says. “I think that whales will focus the world's attention once more, and hopefully we'll take action to reverse or at least slow down whatever environmental damage we're doing.” There's still a year left in the voyage as planned, although Payne worries that there isn't enough money to cover it. He hopes the whales will help reel in research partners and donors.

“It's so ironic,” Payne says, scanning the horizon for white plumes. “We've hunted whales to the point of extinction. And now, the whales are helping us to save ourselves.”

9. ASTRONOMY

# Astronomers Shine a Light Upon Dim Nearby Stars

1. Robert Irion

Most of the sun's close neighbors seem faint and boring, but a closer look shows that small stars have big science to offer

When Todd Henry talks about nearby stars, he puts on quite a show. At a meeting* this past January, the Georgia State University astronomer, dressed in black, dashed through a ballroom to the Mission: Impossible theme, sweeping a small handheld telescope across the rows of onlookers. At the end of the talk, audience members held up color-coded cards that Henry had planted earlier. The resulting sea of red represented most of the objects closest to our sun: small stars, most of them too faint to be seen by the naked eye, whose relatively cool, slow-burning nuclear reactors will operate for hundreds of billions of years before running out of fuel.

Until recently, astronomers ignored these celestial plodders in favor of bright stars that flare or pulsate, exotic compact remnants such as neutron stars and white dwarfs, and sunlike stars that they consider the best havens for life-friendly planets. But now, red dwarf stars are coming into their own. In an ongoing tally of all nearby stars—loosely defined as objects within 10 parsecs (32.6 light-years) of the sun—astronomers such as Henry have found that gas and dust in stellar nurseries condensed into a few big stars and swarms of little ones.

If our part of the Milky Way is typical, then nailing down the local roster of red dwarfs will give astronomers a good sense of the galaxy as a whole—including billions of stars too dim and distant to be seen directly. Moreover, young red dwarfs may prove the best targets for seeing how disks of dusty debris turn into planets, perhaps even worlds like Earth. As a result, says astronomer Dana Backman of NASA's Ames Research Center in Mountain View, California, the humble red stars in our own neighborhood could hold an important key to understanding galactic “ecology”: how the galaxy processes material into stars of various sizes that live and die and into the planets they may foster. Or, as Henry puts it, “The bright stars are what you see if you're a biased astronomer. But the little red guys are running the show.”

## Cartography and reconnaissance

Henry often describes his work in grand terms. “I'm a cartographer of nearby space,” he says. “I think 500 years from now we will be leaving this planet. Somebody needs to make a map to see where we're going.” Meanwhile, he says, tallying nearby stars helps him pursue his own main research interests: whether there's life on other planets and what the universe is made of.

For the past decade, Henry has directed the Research Consortium on Nearby Stars (RECONS). Team members scour the archives of telescopic atlases such as the Two Micron All-Sky Survey, which contains about 162 million sources that shine in infrared light. The SuperCOSMOS Sky Survey, an archive of star atlases based at the Royal Observatory in Edinburgh, U.K., then uses computer algorithms to determine which faint points have drifted noticeably within the last few decades—a hallmark of nearby stars. Using small telescopes in Chile, the team measures the positions of the most promising candidates for a few years to verify their distances.

In this way, Henry's group has assigned dozens of new neighbors to the sun's galactic suburb. In particular, RECONS has quintupled the number of well-characterized stars with less than one-fifth the sun's mass. These stars are similar to Proxima Centauri, our closest neighbor at 4.22 light-years away. Proxima is an “M dwarf,” the smallest star that can sustain nuclear fusion. The statistics from RECONS suggest that lowly M dwarfs make up a startling 85% of all stars in the Milky Way—and perhaps 40% of the galaxy's stellar mass.

“Until RECONS started, no one had counted the M dwarfs accurately. That's the biggest surprise,” Henry says. “There are so many stars, so close to us, and nobody had a clue about them.” Henry even lays 50-50 odds that RECONS will unveil a star closer than Proxima.

Completing the laborious census will take years. “We're like the tortoise in Greek mythology, asymptotically approaching our goal,” says astronomer Neill Reid of the Space Telescope Science Institute in Baltimore, Maryland. Reid's team has raised the tally of objects within a larger volume of space stretching to 20 parsecs (65 light-years) away by about 300 stars. Hundreds more may remain hidden, especially faint objects in binary systems.

Such dot-counting exercises, their proponents argue, are the only way to determine which sorts of objects actually arise from protostellar clouds. For instance, many astronomers expect that brown dwarfs—“failed” stars that didn't quite gather enough gas to ignite steady fusion—should far outnumber their stellar siblings. But, so far, RECONS and other surveys don't bear out that idea. “There are 238 M dwarfs and just 10 likely brown dwarfs within 10 parsecs,” Henry notes. “I know a lot of them are cooling off [beyond the point where telescopes can spot them], but they're just not popping up. That puzzles me.”

Henry gets most animated when he talks about potential abodes for life among the stars in his sample. Future satellites hunting for rocky planets elsewhere will focus on stars like our sun—stable, warm, and supporting a large “habitable zone” within which liquid water could exist. Red dwarf stars are cooler, so any watery planets would have to exist within a small zone fairly close to the star. However, Henry notes, there are at least 10 times more red dwarfs in our galaxy than there are sunlike stars. Added up, the total volume of life-friendly space is about the same for each category of star.

“I contend that the first Earth-like planet will be found around an M dwarf,” Henry says. “My job is to convince [mission planners] that red dwarfs are worth searching.”

## Red dwarf in the crosshairs

Henry's job may get a little easier in light of a recent discovery by another team, reported online in Science on 26 February. Astronomers detected a visible disk of dusty debris around a young nearby star—the closest obvious womb within which rocky bodies might grow. And the star happens to be a red dwarf.

Called AU Microscopii, the star is 33 light-years from the sun and half the sun's mass. It formed about 12 million years ago from the same cloud as Beta Pictoris, a massive blue-white star now about twice as far away that also displays a dramatic disk of dust. The Beta Pictoris disk, first seen 2 decades ago, shows warps, offset rings, and other patterns that point to newborn worlds orbiting within (Science, 16 January 1998, p. 322). More-distant stars have disks sculpted in suspicious ways as well. “These features could not be sustained without the gravitational influences of unseen companions,” says astronomer David Koerner of Northern Arizona University in Flagstaff, who was not part of the new study.

For the first time, AU Microscopii will let astronomers search for similar features around the Milky Way's most common type of star. Telescopes will have a crisp view of the material orbiting AU Microscopii. “Basically, it's in our face, so we'll see an astounding amount of physical detail,” says study co-author Michael Liu of the University of Hawaii (UH), Manoa.

Those details aren't yet evident in the team's initial image. The astronomers used the UH 2.2-meter telescope atop Mauna Kea to block the star's glare with a coronagraph and expose the disk. Even with that small mirror, the disk was obvious after 10 seconds, says astronomer Paul Kalas of the University of California (UC), Berkeley. When large telescopes equipped with adaptive optics view the star again later this year, he notes, their vision of the system should be up to 20 times sharper.

Kalas and Liu also will use the Hubble Space Telescope to expose the disk's internal structure. The disk's orientation—nearly edge-on to our line of sight—means that planetary perturbations could stand out clearly. “The scattered light is evidence for solid material,” Kalas notes. “It's exciting, because we're observing this system right after its natal cloud has dissipated.”

Already, Liu sees indirect evidence of a “hole” in the disk close to the star, based on an apparent absence of warmer dust in the disk's pattern of infrared radiation. The putative hole, a bit smaller than the orbit of Uranus around the sun, could contain one or more planets sweeping away dust, Liu speculates.

If a planet is indeed there, he adds, it would still glow in infrared light from the residual heat of its formation. The most powerful telescopes just might be able to spot such a planet, with help from that extra imprint of warmth. “It's technically very challenging, and it's not yet a settled question whether we can [see planets directly],” Liu says. “But it's one of the reasons why there is such a significant push to find stars that are young and nearby.”

Red dwarfs, it now seems, will help extrasolar planetary science move from an indirect science to a direct one—a transition that is nearly upon us, in Koerner's mind. “It feels like exploring the outer solar system anew,” he says. “We can think of them as worlds and imagine them as places.”

• *American Astronomical Society, 203rd national meeting, Atlanta, Georgia, 4 to 8 January.

10. ASTRONOMY

# Aliens in the Neighborhood?

1. Robert Irion

A thorough canvassing of stars near the sun should reflect the overall properties of the Milky Way's grand disk of stars—or so astronomers thought. But now, it appears that a few nearby stars could be intruders from alien galaxies.

The Milky Way has eaten many smaller galaxies, according to models of how galaxies assemble. In recent years, astronomers realized that traces of this process linger in the form of long “tidal streams” of stars, which retain the motions of their parent galaxy for billions of years. One such merger, still in process, has strewn ribbons of stars that wrap nearly twice around the Milky Way.

Last year, a team showed that one ribbon dives near the sun's position in the galaxy (Science, 3 October 2003, p. 32). Now, astronomers Steven Majewski of the University of Virginia, Charlottesville, and Verne Smith of the University of Texas, El Paso, are taking the next step. By acquiring detailed spectral patterns of the doomed galaxy's stars, Majewski and Smith are taking their chemical “fingerprints.” The patterns include unusual amounts of lanthanum oxides and other oddities. The team will then test nearby stars for matching prints.

“The grand strategy of this field is to come up with a way to sort stars by their birthplace,” says Majewski. “We want to know how much of the Milky Way formed elsewhere.”

One prominent immigrant already may have been exposed. Arcturus, just 37 light-years away and one of the brightest stars in the northern sky, has a strange orbit around the galaxy and distinct concentrations of metals. Astronomer Julio Navarro of the University of Victoria, British Columbia, and colleagues suggest that Arcturus is an interloper from a galaxy shredded by the Milky Way about 8 billions years ago. Be careful which star you wish upon—it may not be our own.

11. THE BIOLOGY OF GENOMES MEETING

# The Case of the Disappearing DNA Hotspots

1. Elizabeth Pennisi

COLD SPRING HARBOR, NEW YORK—New approaches to genome studies are showing that DNA is full of surprises, researchers learned here 12 to 16 May at the Biology of Genomes meeting.

DNAwise, chimps and humans are virtually identical. But two research teams have found, to their surprise, key differences among these close cousins in the locations of DNA recombination hotspots: places where matching chromosomes exchange DNA much more frequently than normal. Evidence is preliminary, but in at least some cases, chimps don't have hotspots in the same places as humans. And it's now unclear whether chimps have as many hotspots as people.

More than just a curious phenomenon, hotspots contribute to evolution and perhaps to human disease. Recombination, in or out of hotspots, introduces variety into a genome and can eliminate bad genes—or, possibly, introduce deleterious mutations. For example, when chromosomes recombine, one sometimes loses a gene and the other gains an extra copy, a problem that can lead to diseases such as the blood disorder thalassemia. Researchers also want to understand hotspots to make best use of a multimillion dollar project, the HapMap, that is cataloging variation in the human genome as a tool for tracking down disease genes (Science, 30 April, p. 671).

Until recently, researchers were unsure whether hotspots—typically short regions less than 2000 bases long—are common. They assumed that recombination occurs at a fairly constant rate along much of the genome. To try to learn more about what makes a hotspot hot, Molly Przeworski, a population geneticist at Brown University in Providence, Rhode Island, compared them in humans and chimps.

She and her colleagues collected DNA samples from two dozen western chimps, a subspecies that lives in western Africa. They looked at the comparable 10,000-base stretch of DNA where a human hotspot is known to exist. After sequencing that piece of DNA in each chimp, they looked for high recombination rates there. This stretch of chimp chromosome turned out to be not so hot after all, she reported. She couldn't tell whether an equivalent chimp hotspot had been shifted to a different place along that stretch of DNA or whether it didn't exist at all. Either way, the result was puzzling: “You would predict that the human and chimp would be the same [in terms of hotspot location] because the sequence is the same,” says Edward Rubin, director of the Department of Energy Joint Genome Institute in Walnut Creek, California.

After a more extensive comparative study, Harvard graduate student Wendy Winckler found similar results, she reported at the meeting. Independent of Przeworski, she and her colleagues looked at equivalent places in the chimp genome where hotspots turn up in human DNA. First Winckler, working with David Altshuler of Boston's Massachusetts General Hospital and others, sequenced presumed hotspots in 22 chimps and 24 humans to catalog single-base changes, or SNPs, in each. They then looked for those same SNPs in another 38 western chimps and 94 humans, half Africans and half Caucasians.

As expected, both human populations shared the same six hotspots. But in the chimp, all the hotspots were missing. Much more work needs to be done to determine whether chimps have fewer hotspots overall.

The implication is unclear, but at the very least, the work points to a previously unrecognized genomic trait that distinguishes the two primates and raises questions about the role hotspots play in evolution. It also drives home how much we have left to learn about genomes. “I find it amazing that both groups find that human recombination hotspots are totally gone when they check in chimps,” says Svante Pääbo, a geneticist at the Max Planck Institute for Evolutionary Anthropology in Leipzig, Germany. “Since there are no dramatic sequence differences between the species in these areas, this suggests to me that there must be some mechanism that we do not yet understand that is responsible” for creating hotspots.

“It's not sequence,” agrees Arend Sidow of Stanford University. He wonders whether the particular arrangement of the proteins surrounding those DNA regions sets the stage for recombination. Pääbo suggests that perhaps hotspots are chemically marked in some way.

Whatever the cause, both Winckler and Przeworski are trying to determine the history of these and other hotspots. It could be that at one time, both species shared hotspots, but that some disappeared in the chimp. Or it could be that they arose separately in humans and chimps after the species started evolving their separate ways. Or there could be another explanation altogether.

12. THE BIOLOGY OF GENOMES MEETING

# Disposable DNA Puzzles Researchers

1. Elizabeth Pennisi

COLD SPRING HARBOR, NEW YORK—New approaches to genome studies are showing that DNA is full of surprises, researchers learned here 12 to 16 May at the Biology of Genomes meeting.

For a long time, any DNA that didn't make up genes was considered junk, even though it constituted the bulk of the human genome. Gradually, though, genome biologists have found gems among this non-protein-coding sequence, suggesting that “junk” was a serious misnomer. But new research suggests that vast tracts of this sequence may be disposable after all: Marcelo Nóbrega, a geneticist at Lawrence Berkeley National Laboratory (LBNL) in Berkeley, California, finds that mice can do just fine with millions of these bases deleted from their genomes.

About 2 years ago, Edward Rubin, director of the Department of Energy Joint Genome Institute in nearby Walnut Creek, and his colleagues discovered that some DNA sequence in human gene deserts—long stretches of DNA between genes—was almost the same as the sequence in comparable mouse deserts (Science, 31 May 2002, p. 1601). This conservation across species that shared a common ancestor more than 80 million years ago seemed highly unlikely —unless those regions served a purpose. Since then, “we've had the assumption that all these regions are doing something,” says Michael Zody, a computational biologist at the Broad Institute in Cambridge, Massachusetts; that something is probably gene regulation, he says.

Nóbrega and his colleagues found support for this idea when they compared desert regions conserved between fish and humans, quite distant relations. “Most sequences conserved between fish and humans are not only functional, but they help regulate genes,” Nóbrega reported. But their mouse-human comparison told a different story. They analyzed 15 comparable desert sequences and found that only one was a regulatory region in both species.

Puzzled, Nóbrega, Rubin, and their colleagues decided to delete deserts from mouse genomes, hoping to learn what other function these regions might serve. His LBNL colleague, geneticist Yiwen Zhu, knocked out two regions, one about 2 million bases long and the other about 1 million, both of which were conserved in humans and mice but not fish. After inserting the altered genome into embryonic mouse stem cells, Zhu added the cells to mouse embryos and looked for abnormalities in their descendents. “There was no sign of any difference in survival” between the genetically altered and normal mice, Nóbrega reported at the meeting. “There's no sign of overt pathology.”

Other researchers are dumbfounded. “To knock out 2 megabases and not have an effect—that's remarkable,” says Jim Hudson, a geneticist at Open Biosystems in Huntsville, Alabama. “It can't be true,” says a skeptical Arend Sidow of Stanford University. Both Hudson and Sidow wonder whether these noncoding regions have a function that just doesn't show up in the tests Nóbrega did. So the question remains open, says Rubin: “Is the genome like a trash novel, from which you can remove 100 pages and it doesn't matter, or is it like a Hemingway, where if you remove a page, the story line is lost?”

13. THE BIOLOGY OF GENOMES MEETING

# Surveys Reveal Vast Numbers of Genes

1. Elizabeth Pennisi

COLD SPRING HARBOR, NEW YORK—New approaches to genome studies are showing that DNA is full of surprises, researchers learned here 12 to 16 May at the Biology of Genomes meeting.

From the inside of the mouth to farmland to the surface of the ocean, microbes reign supreme. We know there are lots, but not how many; we also know they live remarkable lives, but we don't fully understand how they survive in often-adverse conditions. The enigmas remain because most of these organisms can't be grown in the lab. But now, their genes at least are within reach.

At the meeting, three research teams described successes in surveying various microbial communities by sequencing the organisms' DNA collected in environmental samples. And everywhere these researchers look, they are finding not only more kinds of microbes than expected but a staggering number of new genes. It's becoming clear, says David Relman, a molecular microbiologist at Stanford University, that with this new approach, called community or environmental genomics, “you can discover far more novel genes and novel biology” than was possible before.

Already, “we are amazed by the degree of gene discovery,” adds J. Craig Venter, who runs The Center for the Advancement of Genomics in Rockville, Maryland. Venter is leading a team that is sampling the microbial diversity in oceans around the world.

Until recently, researchers have tended to sequence the genome of a particular organism—be it yeast, mouse, or mold—then analyze its genetic makeup. But now a few genome biologists are sequencing whatever DNA they can get from a test site and then piecing together these DNA sequences into genomes, or at least bits of genomes. Each set of combined sequence represents a different organism. “The approach is to understand the complex ecology [of microbes],” says geneticist Stefan Schreiber of Christian Albrecht University in Kiel, Germany. “The potential is enormous.”

Relman has been looking at the ecology of the mouth. He's surveyed the DNA of microbes that live in the pockets between the teeth and the surrounding gum—the site of gum disease. Not much has been known about that community because so few of its members can be grown in the lab. But by focusing on DNA, Relman has found what could be a genetic signature of gum disease. One type of bacterium, called TM7, thrives in gums with mild disease. But when gums disintegrate, another group—methane- producing archaea—may take over.

Relman, along with Steven Gill and Karen Nelson of The Institute for Genomic Research in Rockville, has yet to look at all the genes of these oral microbes, but other work is showing just how fruitful gene searches can be. Edward Rubin, director of the Department of Energy Joint Genome Institute in Walnut Creek, California, has preliminary data on soil samples collected from a Minnesota farm. He and his colleagues have sequenced 100 million bases but can't seem to fit them together because “there are so many different kinds of organisms,” he explains. But they can pick out as many as 150,000 previously undescribed genes hidden in that DNA, just from one site.

Venter is also finding incredible diversity. At first he focused on the Sargasso Sea, sequencing more than 1 billion bases and grouping them into genomes representing 1800 species, including 148 groups of unknown bacteria. The DNA includes 1.2 million new genes, he reported in April (Science, 2 April, p. 66).

Now, Venter's group has much more data, and off Long Island, New York, alone, only 950 putative organisms—1%—match those found in the Sargasso Sea. Eventually, the team expects to find as many as a billion new genes, he reported at the meeting. All these genes “will provide a different view of evolution,” Venter points out, as they suggest that new genes or new microbes arise much more often than had been previously thought.

14. # Soil and Trouble

When people intensively till fields and clear-cut forests, they can damage or destroy topsoil that took centuries to accumulate. Just how vulnerable soils are depends on underlying conditions. Mismanaged soils in windswept lands can easily turn into desert, for example, and saline soils can become salt-encrusted wastelands.

This map shows the main barriers to productive farming, along with erosion risk, derived from climatic and soil conditions. Overlaid as cross-hatching are regions reported to be highly or very highly degraded according to a global survey of soil experts published in 1990. The hot spots illustrate examples of the worst soil degradation, from the most common physical type—water erosion—to chemical forms, such as that caused by pollution from industrial chemicals and war.

15. # Wounding Earth's Fragile Skin

1. Jocelyn Kaiser

Soil degradation in all its nefarious forms is not a prelude to mass starvation, as analysts once feared. Nevertheless, it is eroding crop yields and contributing to malnourishment in many corners of the globe

The rebel insurgency that toppled Haiti's government earlier this year was, on the surface, simply the latest turn in a bloody cycle of repression and reaction. But deeper down, the uprising exposed a root cause of the country's misery: some of the most denuded land on the planet. Only 3% of the once lushly forested terrain still has tree cover, and up to one-third, some 900,000 hectares, has lost so much topsoil that it is no longer arable, or barely so. Today the landscape is crisscrossed by gully-scarred, deforested hills and “rocks where there used to be dirt,” says Andy White, an economist at Forest Trends, a nonprofit organization in Washington, D.C. Erosion has brought the Caribbean country's agriculture to its knees, he says, and that in turn “has driven rural poverty,” impelling desperate Haitians toward city slums where unrest continues to simmer.

The tremendous hemorrhaging of fertile soil makes Haiti one of the global hot spots for degradation, as illustrated on the preceding pages. For decades soil scientists have deplored the loss of land to erosion, nutrient depletion, salinization, and other insults. But policymakers have been slow to respond. One reason is that dire warnings in the 1980s that soil erosion would doom the world to chronic food shortages have failed to pan out.

But in the past few years, new studies have yielded better data linking soil degradation to a slump in the growth of global crop yields, uniting dueling camps. After years of “pecking away at this problem,” soil scientists, geographers, and economists “are coming toward the middle,” says Keith Wiebe, a resource economist at the U.S. Department of Agriculture (USDA). The bottom line: “As a global problem, soil loss is not likely to be a major constraint to food security,” says Stanley Wood, a senior scientist at the International Food Policy Research Institute in Washington, D.C. But as Haitians know all too well, degradation can be a severe and destabilizing threat in places where farmers are too poor to curb or overcome the damage.

Moreover, global warming is expected to exacerbate regional crises. As the ground heats up, organic matter decomposes more readily, reducing soil fertility and, asserts Duke University ecologist William Schlesinger, releasing carbon dioxide into the air to fuel warming. Deserts are also expected to expand and erosion worsen if violent storms occur more frequently. But there's a flip side, says soil scientist Rattan Lal of Ohio State University, Columbus: “We can do something about it” by managing soils to stem erosion and retain more carbon.

## A litany of loss

The first real eye-opener to the consequences of soil abuse came during America's Great Depression in the 1930s. For 8 years, “black blizzards” of topsoil would intermittently blot out the sun in the Great Plains, burying farms in dust and driving many families from their land. The Dust Bowl was triggered by a combination of drought and shoddy agricultural practices: Farmers would plow grasslands, leaving little vegetation to hold onto water and topsoil when fields lay fallow. The crisis led to the creation in 1935 of USDA's Soil Conservation Service, which developed new practices to curb erosion. These included using plowing methods that follow the land's contours—across slopes, for example—and growing cover crops to retain soil between planting seasons.

This soil revolution of sorts prompted scientists to start delving more deeply into the mechanisms of degradation. One type is purely physical, mainly the result of wind and water erosion. On China's windswept Loess Plateau, northwest of the Yellow River Basin, soil is carried away faster than anywhere else in the world: Each year the plateau loses 1.6 billion tons of loess, a powdery soil deposited over millennia by wind storms. Likewise, mountain ranges are prone to erosion caused by heavy rains that strip topsoil from steep slopes. In the lower Himalayas, for example, “the land has lost its capacity for any productive purposes,” says ecologist P. R. Ramakrishnan of Jawaharlal Nehru University in New Delhi.

In a more vampyric form of degradation, sowing crops without fertilizers season after season drains the soil of its lifeblood: nutrients. Soils in sub-Saharan Africa are rapidly losing fertility because farmers generally can't afford fertilizers and are salvaging postharvest stubble and even livestock dung for fuel. Before the 1950s, says USDA soil scientist Hari Eswaran, African farmers routinely would leave less productive fields fallow for a generation. Then the population exploded, and now “every piece of land is taken up,” Eswaran says.

Another form of chemical degradation occurs when poor irrigation practices saturate land with salt—an acute woe in the Middle East and India. And in Australia, the removal of water-hoarding native plants has led to rising water tables that suffuse topsoil with salt, says hydrologist Mirko Stauffacher of the Commonwealth Scientific and Industrial Research Organisation. The agency estimates that 10% of Australia's agricultural lands are now affected, a figure expected to triple by 2050. In some areas, Stauffacher says, “nothing can grow.”

Desertification, meanwhile, plagues the Sahel in Africa as well as Kazakhstan, Uzbekistan, and northern China, where desert is growing by 3600 square kilometers a year—an area larger than Rhode Island—fueling the massive dust storms that choke Beijing and other cities. New farms encroaching on grasslands, abetted by wind erosion, are driving the desert's expansion, says Tao Wang, director of the Cold and Arid Regions Environmental and Engineering Research Institute in Lanzhou. “It looks like the same story as the Dust Bowl,” he says.

## Battle cry

To help wage a war on so many fronts, soil scientists in the late 1980s embarked on a rapid effort to form a global picture of degradation. The assessment, conducted by the International Soil and Reference Information Centre (ISRIC) in the Netherlands, surveyed more than 250 experts on six continents and mapped out where soils are hurting the most.

The results were startling: Of the 11.5 billion hectares of vegetated land on Earth, 17% was degraded, largely through erosion; of this land, 1 in 6 hectares could no longer support crops. The main causes, according to the survey, were deforestation and farming practices such as overgrazing. The assessment has shortcomings: It relied on expert opinion rather than direct field data, for example, and the map's coarseness means that a region can be labeled degraded even if it is largely intact. Nevertheless, the survey helped “put soil in the picture as an important resource,” says ISRIC geographer Godert van Lynden.

Ratcheting up the alarm, in the mid-1980s agricultural economist Lester Brown, then president of the Worldwatch Institute, a think tank in Washington, D.C., warned that civilization risked running out of soil before oil. He estimated that human activity was responsible for the loss of 26 billion tons of topsoil per year, 2.6 times the natural rate. In an influential study a decade later, Cornell University ecologist David Pimentel concluded that in the United States alone erosion inflicted an eye-popping \$44 billion a year in damage to farmland, waterways, infrastructure, and health (Science, 24 February 1995, p. 1121). He predicted that if farmers failed to replace lost nutrients and water, U.S. crop yields would drop 8% per year.

Both conclusions have been challenged. Brown arrived at his global erosion figure by extrapolating a shaky U.S. estimate to the rest of the world, economist Pierre Crosson later noted. Crosson, who recently retired from the nonprofit organization Resources for the Future, also took aim at Pimentel's calculations, pointing out that the numbers were based largely on models, not field data (Science, 28 July 1995, p. 461). Ohio State's Lal agrees that erosion estimates are “really very crude.” Based on river sedimentation rates, Lal says, human-induced erosion averages 11 billion tons per year—roughly equal to natural erosion.

Sparring aside, a recent set of analyses* has helped clarify the link between soil degradation and tapered growth in crop yields. One study, led by Christoffel den Biggelaar of Appalachian State University in Boone, North Carolina, combined erosion rates, estimated by soil type and climate, with data from hundreds of field studies for various crops. The team estimated global average potential yield losses at 0.3% a year, compared to Pimentel's 8% for the United States. Recognizing that most farmers have incentives to counter these losses—for example, by using better tilling methods—the actual decline may be as low as 0.1%, USDA's Wiebe says. But that figure is not trivial, he cautions. Growth rates of global cereal yields, which have grown at a brisk 2% per year since the 1960s, are expected to rise at a slower rate in coming decades. As that happens, Wiebe says, 0.1% “gets to be a more important number.”

Moreover, the reasonably rosy global picture glosses over disturbing regional trends. Degradation is clearly cutting into yields in parts of Africa, South Asia, and Latin America, notes Forest Trends analyst Sara Scherr. “The critical issue,” she says, “is where are the places in the world” where soil degradation matters most. A USDA analysis found that if struggling regions could reduce degradation and thereby boost crop-yield growth by a mere 0.1% per year, the number of malnourished people would fall by 5%, or 37 million, over a decade.

## A fertile future?

There's no lack of technological fixes for reversing soil degradation. In the United States, for instance, no-till farming has helped reduce water erosion by over 40% since 1982, Pimentel notes. And in developing countries, strategies such as sowing cover crops and planting trees have been shown to restore soil fertility and stem erosion. But these solutions can't be imposed top-down (Science, 21 November 2003, p. 1356). Terracing, for example, has failed in many places, including Haiti, because it is too expensive for most farmers.

Although the United Nations funds some efforts to disseminate approaches that work, major aid organizations, such as the World Bank, have slashed agricultural budgets over the past 2 decades in favor of addressing urban projects, Scherr says. But the pendulum appears to be swinging back to rural development. For example, soil health tops a list of priorities being assembled by the U.N. Millennium Project's hunger task force, which aims to halve the number of starving people by 2015, says co-chair Pedro Sanchez of Columbia University's Earth Institute. Soil degradation, he says, is “the main constraint to reducing hunger” in Africa.

The biggest looming issue may be global warming. Lal notes that erosion already contributes to warming, because some of the carbon in soil-laden water running off fields wafts into the atmosphere. Yet fields could sop up some of this carbon, Lal says, if farmers adopt practices to reduce erosion and retain nutrients, as is encouraged by the Kyoto Protocol on climate change (see p. 1623). Inevitably, “a hotter world is likely to have less organic matter” in its soils, Duke's Schlesinger notes. Such vital nutrients decompose as temperatures rise, releasing carbon. Deserts will also expand as the interiors of continents become drier. Erosion rates could rise if soils dry out and storms increase, Lal says. “The risks of soil degradation are going to go up, but how much, we don't know,” he says.

Soil degradation is no longer seen as a matter of “global survival,” says Wiebe. But “it's still an issue. It will keep coming back.” In a warming world, it could come back to haunt us.

• *Land Quality, Agricultural Productivity, and Food Security: Biophysical Processes and Economic Choices at Local, Regional, and Global Levels, edited by Keith Wiebe (Edward Elgar Publishing, 2003).

16. # From Alaska to Yucatan, a Long-Awaited Soil Survey Takes Shape

1. Fiona Proffitt

Soils are a sponge for pesticides and other nasty compounds filtering down from the surface. Yet experts have only a sketchy idea of how the ground copes with this toxic trickle—critical information for any and all creatures living off the land. That frustrating blind spot may soon change. The U.S. Geological Survey (USGS) and a handful of other agencies are planning an ambitious mission: a soil geochemical survey of all of North America.

The impetus for the survey, which USGS expects to cost tens of millions of dollars, grew out of the need for more comprehensive data on soil contaminants. Regulatory agencies are crying out for such data to improve assessments of environmental and health risks. For example, information on background levels of heavy metals such as arsenic and lead would help determine whether human activities are raising concentrations to hazardous levels, says David Mellard, a toxicologist at the U.S. Agency for Toxic Substances and Disease Registry. “There are literally tens of millions of chemicals in use out there,” adds USGS soil scientist David Smith, who with colleague Martin Goldhaber is leading the survey's feasibility study. But “if you don't know what's currently in the soil,” Smith asks, “how do you know if it's increased?”

Geochemical data in North America are surprisingly spotty. The most widely used resource, the Shacklette data set, draws on soil samples from 1323 noncultivated sites (about one per 6000 square kilometers, an area roughly the size of the state of Delaware) across the United States that were collected in the 1960s and '70s. The U.S. Department of Agriculture's Natural Resources Conservation Service (NRCS) followed with a separate limited survey, sampling 3045 agricultural sites for levels of five metals important for crops: cadmium, copper, lead, nickel, and zinc. Nationwide data are largely lacking on organic contaminants and microbial communities.

USGS, along with NRCS and counterparts in Canada and Mexico, hopes to begin the decade-long continental survey in 2006. It would analyze soil samples from 10,000 sites across North America for more than 40 elements. If feasible, the agencies would also assay for pesticides and other organic compounds as well as screen for pathogens and other soil microbes.

Possibly the biggest challenge is designing a sampling scheme that serves different interests without blowing the budget. Most soil scientists would prefer samples to be taken from each distinct layer of every soil type, whereas researchers assessing health risks are mostly interested only in what's going on in the top several centimeters of soil in populated areas.

Differences in philosophy have caused a rift that could undermine the project. USGS geochemists advocate a random sampling approach, whereas NRCS researchers would prefer to sample soils representing a particular landscape and origin, says an NRCS soil scientist. NRCS is already collecting geochemical data as part of the National Cooperative Soil Survey—primarily aimed at characterizing soils—and is reluctant to divert resources to an effort that doesn't meet its requirements. National Soil Survey Center director Robert Ahrens says he's optimistic that the agencies can resolve their differences and at least share data.

Over the next few years, the agencies plan to test survey design, sampling techniques, and analytical methods, although funding is only set for 2004. Smith says he hopes that successful pilot studies will win funding for the full-fledged survey. People should care about what's in the soil, he says: “It literally sustains our lives.”

17. # Defrosting the Carbon Freezer of the North

The perpetually frozen soils of the Arctic and boreal regions are thawing at unprecedented rates. It's unclear what this bodes for global warming

When Phil Camill goes looking for frozen soil in the spruce forests of northern Manitoba each summer, he finds less and less each year. Instead of rock-hard permafrost, he discovers drenched soil, dying trees, and ever-expanding, mossy bogs. The ground is so saturated that it quakes underfoot, says Camill, a plant ecologist at Carleton College in Northfield, Minnesota: “It's like walking on a trampoline.”

Across huge swaths of the Arctic, permafrost is warming to record high temperatures. “It's really happening almost everywhere,” says Vladimir Romanovsky, a geophysicist at the University of Alaska, Fairbanks. The pace has shocked researchers—and it's accelerating. In Manitoba, at the southern edge of Canada's permafrost, the thaw rate has nearly tripled over 4 decades; this patchy permafrost is now receding up to 31 centimeters per year, and in a forthcoming paper in Climatic Change, Camill predicts that Manitoba will lose most of its permafrost within a century. Even in the far north, where soil deposits are thicker and colder, much permafrost has warmed to the brink of meltdown. That's a major concern for residents: Settling of the ground has already damaged buildings, pipelines, and other infrastructure in Alaska and Siberia (Science, 30 August 2002, p. 1493).

But widespread permafrost melting could have grave consequences well beyond the far north. No one knows exactly how much carbon is locked up in boreal and alpine permafrost, but estimates range from 350 to 450 gigatons—perhaps a quarter to a third of all soil carbon. The big question is what will happen if even a fraction of this massive carbon store is liberated.

Many parts of the Arctic are already warming faster than any other region on Earth is, a trend that climate models predict will continue. Although researchers are struggling to arrive at a bottom line, they suspect that thawing permafrost will drive global temperatures higher over the next century. No one has a clue, though, how much higher. Thawing permafrost “is a real wild card in the carbon cycle,” says Lawson Brigham of the U.S. Arctic Research Commission in Fairbanks.

## Cold storage

Any soil that stays frozen more than 2 years in a row counts as permafrost. The perpetual cold makes these soils an excellent carbon sink, because Arctic plants, such as sedges and mosses, decompose slowly after death. That results in living plants sucking from the atmosphere more carbon dioxide, the chief greenhouse gas, than is released from dead matter, building up thick organic soils. For tens of thousands of years, the far north has been accumulating carbon this way. In the upper reaches of Siberia, for example, peat deposits extend for thousands of kilometers and are hundreds of meters thick. Today, permafrost covers roughly a quarter of the land in the Northern Hemisphere.

Researchers first noticed a trend toward a widespread thaw after the U.S. Geological Survey started measuring rising temperatures in abandoned Arctic boreholes in the 1960s. It took another 2 decades to launch a concerted effort to probe how permafrost thawing—and the concomitant changes to overlying soils—might influence the amount of greenhouse gases released into the air.

One team, led by Walter Oechel, an ecologist at San Diego State University in California, started with the assumption that the Arctic was still a carbon sink. In the early 1980s, the researchers conducted experiments at Toolik Lake and Barrow, Alaska, where they measured gases wafting from tussocks and other typical High Arctic vegetation. To their amazement, they found that the Alaskan tundra was releasing more CO2 than it was absorbing. “This was contrary to all that was known about Arctic system functioning,” Oechel says.

The shift was primarily due to changes, brought on by warming, in how water moves through the soils. As Arctic soils warm, water seeps out of thawed permafrost and evaporates, often leaving the soils drier and more oxygenated. Under these conditions, microbes more readily break down dead plant matter, the bulk of organic material in Arctic soil. This decomposition releases CO2 into the atmosphere.

By 2000, however, the amount of CO2 escaping from the tundra had begun to taper off. Permafrost still appeared to be drying out, but woody shrubs and other plants that thrive in drier conditions were becoming more abundant—and absorbing CO2 from the air as they photosynthesized. “It was a total surprise that it was happening so quickly,” says Oechel. That observation has been backed up by aerial photos revealing the northern advance of shrubs over the past 3 decades.

Behind this trend are some complex phenomena. Thicker, colder deposits take longer to warm up. And vegetation can influence how much solar radiation reaches the soil: Trees reflect much less solar energy than do snow-covered sedges, for example. And denser vegetation is a better insulator of permafrost—leading to the counterintuitive situation of rising temperatures spurring plant growth and thus helping to preserve underlying permafrost.

Another key factor in determining the impact of thawing permafrost is topography. Where water can drain away easily, soils tend to release more CO2 and less methane. That's because plant roots and some microbes exposed to oxygen will produce CO2, whereas methane-making microbes need a wet, oxygen-free environment. “That makes it hard to find a common pattern of CO2 and methane impact across the complex Arctic landscape,” says Oechel. “The hydrology is key, but it's not well understood.”

In Manitoba, for example, the thawing of frozen wetlands has resulted in wetter soils and greater carbon uptake. Sphagnum mosses thrive in warm, wet bogs, and when they die, much more carbon is stored than in nearby permafrost-rich spruce forests, Camill reported in 2001. “The sheer act of just thawing out the permafrost is enough to double the peat accumulation,” he says, noting a similar pattern across western Canada. That implies a net CO2 uptake. It's unclear, though, how warming is affecting Siberia's prodigious peatlands, which have accumulated some 70 gigatons of carbon in the last 11,000 years (Science, 16 January, p. 353).

But CO2 is only part of the story. An ongoing study has documented that much of Sweden's northern tundra has grown wetter over the past 3 decades, and it is giving off increasing amounts of methane—an even more powerful greenhouse gas than CO2. As the permafrost thaws, the tundra is changing into marshland, with permafrost having disappeared entirely from some of the new bogs. Thawed marshes release methane because standing water leaves them oxygen-deprived, prime conditions for bacteria that convert plant detritus into methane. (As noted above, bacteria that liberate CO2 thrive in aerobic conditions.) “We're seeing dramatic changes,” says biogeochemist Torben Christensen of Lund University in Sweden. In one well-studied bog, he says, methane emissions appear to have risen by up to two-thirds since the early 1970s.

To chart the changes, Christensen and colleagues compared aerial photos of the Stordalen mire taken in 1970 and 2000. They tabulated four types of vegetation and checked the accuracy of the 2000 photo with a detailed field survey. The extent of drier plant communities, mainly mosses and shrubs, had declined from 9.2 to 5.9 hectares, the team reported in the 20 February issue of Geophysical Research Letters. Meanwhile, the abundance of sedges and other marshy plants increased by more than half. “The thing that surprised me is the rate of change,” says Christensen. “Almost from year to year, we can see the vegetation changing.”

His team calculated that the shift in plants is correlated with a rise of between 22% and 66% in methane production by soil bacteria and plants. Methane measurements from the early 1970s at Stordalen back the increase. “It's a pretty massive change,” notes William Reeburgh of the University of California, Irvine. Yet this may be a local phenomenon: Average regional temperatures in northern Sweden had hovered near freezing in recent decades, rendering the permafrost vulnerable to even a slight warming. Tundra in colder regions may not be so susceptible, Christensen notes.

All this uncertainty makes it difficult to predict the overall response of Arctic soils to global warming. Even basics are lacking, such as the precise extent of frozen peatlands. “So it's difficult to calculate the gigatons of carbon they would release if they all thawed,” says Camill. Further complicating the forecast is the fact that few spots in the circumpolar north are studied well or at all—a point emphasized in a report on permafrost released this spring by the U.S. Arctic Research Commission (www.arctic.gov/files/PermafrostForWeb.pdf). That uncertainty makes any conclusion about carbon flux risky, says Christensen: “I would be hesitant to come up with any major statements saying the Arctic is a source or sink.” Still, he bets that even if the most northern, drier Arctic soils revert to being a carbon sink, methane emissions from thawed permafrost will likely accelerate global warming.

Some experts, however, predict that the methane will be accompanied by CO2, and perhaps many gigatons could be released in the far north over the next century. “The Arctic is likely to be a huge positive feedback on global warming,” Oechel says, citing the fact that warmer, drier soil releases more CO2, warmer, wetter soil releases more methane, and loss of snow and ice cover mean more solar rays warming the ground. Increasing rates of forest fires could also unleash large amounts of carbon. The greening of the Arctic may eventually absorb much of the carbon released from cold storage, he says. But several generations of people will still have to cope with what is shaping up to be a fearful loss of permanence.

18. # The Secret Life of Fungi

1. Elizabeth Pennisi

Once considered pathogens, microscopic fungi that live in soil are shaping plant communities and aiding efforts in environmental restoration

It all started with truffles. In the late 1880s, German foresters eager to grow these fungal delicacies asked a colleague, A. B. Frank, to figure out how they propagate. He undertook the quest and unearthed a hidden world in which truffles and other fungi work with plants to create a mutual support network underground. It has taken more than a century for Frank's scientific descendants to make sense of the tangles of fungal threads and plant roots that he dug up, and what they've learned is changing the way people think about plant ecology.

The mycorrhizal species of fungi live hidden in the soil most of the time. They are probably best known for their fruiting structures—such as toadstools—that often pop up next to tree trunks. Less diverse but more pervasive are the arbuscular mycorrhizal fungi, which never see the light of day. Both types send out extensive networks of fine threads called mycelia, sometimes up to 20,000 km in 1 cubic meter of soil, which link with and extend the reach of plant roots. Because of their small size, one-60th as thin as roots, the threads can get into tight spaces and retrieve hard-to-get nutrients. The fungi efficiently deliver soil minerals, particularly phosphorus and nitrogen; the plants reward them with energizing sugars.

Ecologists and physiologists are finding that mycorrhizae—the union of roots and these soil fungi—do more than enhance a plant's nutritional status. By hindering water loss and erosion, they improve the soil. They also protect against pathogens and dampen harm from toxic wastes—talents that researchers exploit to reduce fertilizer use and remake damaged ecosystems. In addition, new studies are showing that these fungal threads link one plant to another, transferring nutrients not only among fungi but from plant to plant as well, shaping the biological makeup of whole communities. All in all, “I suspect that many plants would not be able to survive in nature without being aided by these small and unseen [partners],” says Marcel van der Heijden, an ecologist at the Free University Amsterdam, the Netherlands.

## The dirt on mycorrhizae

Frank was among the first to realize that—in contrast to a widely held view—soil microorganisms aren't always harmful. He demonstrated this by planting seeds in soil collected from pine forests, some of which had been sterilized, some left in its natural state. “Only in the natural soil did he get a big growth of plants,” says David Read, an ecologist at the University of Sheffield, U.K. About the same time, other researchers were discovering that mycorrhizae were widespread: Boreal, temperate, alpine, and tropical forests all have them, as do grasslands and tundras. The few exceptions were lava fields, strip mines, and other places robbed of topsoil—or farmland that had been heavily fertilized. And many plants seem to benefit from these plant-fungi partnerships.

All told, biologists have found, the roots of about 80% of plants are entwined with mycorrhizal fungi. “It would be hard to go outside anywhere and pick up any handful of soil and not have mycorrhizae,” says John Klironomos, a soil ecologist at the University of Guelph in Ontario, Canada. They help plants settle into damaged areas, such as those destroyed by fire. And their role appears to be ancient. Four hundred million years ago, “when the first plants colonized land, mycorrhizal fungi were there,” helping newcomers survive in a harsh, dry landscape, Klironomos explains. Since then, the support system has grown, says Katarzyna Turnau, an ecologist at Jagiellonian University in Krakow, Poland. “Mycorrhizae [gave] the plants the ability to use different areas and to explore new niches,” making possible the incredible diversity of modern flora.

The intimate details of this plant-fungi partnership remained elusive for decades, mainly because they were hard to observe. Few species grew well in the lab. Field studies were a challenge: “We're talking about small things in a cryptic environment,” says Klironomos. Now, molecular techniques are enabling mycorrhizal experts to break new ground, says Read.

Today's molecular ecologists don't need to isolate the organism; they can look for variation in the DNA coding for a large piece of the cell's ribosome, which makes proteins, and from those sequences distinguish one fungus from another. They also learn which fungi are present, whether they have begun to infiltrate a particular plant root, and how nutrients flow between plants and their fungal partners. The insights have been startling: “There's diversity out there that's never been seen [before],” says Peter Young, a molecular ecologist at the University of York, U.K.

Basic resources are expanding as well. Over the past few years, several groups have established repositories of mycorrhizal species. This year the U.S. Department of Energy (DOE) began deciphering the genomes of two widespread types: Glomus intraradices and Laccaria bicolor, a pink mushroom. DOE chose them because they partner with commercially important plants whose genomes are also being sequenced. Commercial foresters use Laccaria to spur poplar seedling growth, and Glomus interacts with crop plants including rice, maize, alfalfa, and wheat. “The genome information will allow us to find ways to better utilize fungi in combination with their host plants,” says Gopi Podila, a fungal and plant molecular biologist at the University of Alabama, Huntsville.

## Pioneers, providers, protectors

These technical advances have come just in time to help answer some fundamental questions. Intrigued by studies showing that not all mycorrhizal plant partnerships are equal, a growing number of ecologists are discovering the natural history of these fungi.

In the mid-1980s, Read began urging fellow mycorrhizal aficionados to broaden their traditional perspective and look beyond the two-partner model of symbiosis. “I wanted to look at how the symbiont might change the structure of the plant community,” Read recalls. It turns out that the fungi play an important role in plant succession, a pattern in which one set of plants gradually replaces another, and that fungi have a broad physical and ecological reach.

In addition, Read, Klironomos, and others were beginning to realize that fungi and plants are like people: Some pairings are more compatible than others. “You get just as many positive effects as negative effects,” says Klironomos.

In 1998, Read, van der Heijden, and their colleagues decided to look at compatibility. They built a series of boxes and filled them with soil of two types. Some boxes contained sterilized dirt; others included natural concentrations of fungi. The researchers recorded all the mycorrhizal species and the variety of seeds planted in each box. After the seeds had sprouted, they monitored the status of the plants for months, showing that the number and kind of plants depended on which fungi were present. Other work has shown that the more plentiful the fungal species, the more diverse the floral landscape above.

Klironomos took a closer look at these relationships. He matched each of 10 plant species with each of 10 mycorrhizal species and carefully monitored plant growth. “The effect was dependent on the specific combination and source [of the fungi],” he reported in the September 2003 issue of Ecology. For this work, he used plants from a meadow in Guelph where he and others had already characterized many of the mycorrhizae. He extended the original experiment by mixing and matching the Guelph specimens with exotic ones. The results showed that exotic species got along just fine with local partners. The outcome of plant and fungal matchups ran the gamut. At one extreme, both partners prosper. At the other, one partner, such as the fungus, takes complete advantage of the other. In some of these less-than-ideal matches, the fungus steals plant sugars without giving nutrients in return. This negative relationship was also common when Klironomos mixed local and exotic species.

In work yet to be published, York's Young has found that these differences have a biochemical basis. By tracing a carbon isotope as it made its way from plant to fungus, he has been able to see how much is taken up by various symbiotic species. “Some fungi are more active in taking carbon from a [particular] plant,” he says. If this carbon appetite translates into greater nutrient flow back to the plant, then that particular species has an advantage that could help it outcompete its neighbors, say by producing more seeds or by growing taller so it gets more sun, Young explains.

These results are solidifying the idea that the fungi are movers and shakers in the plant world. They set up some plant species to thrive and others to fail, depending on whether seeds settle in friendly fungal territory. “The fitness of [a particular species] depends largely on where it happens to land,” Klironomos says. And this variability intrigues researchers such as Thomas Bruns, a fungal expert at the University of California, Berkeley. He and others have sought out sites where they can watch mycorrhizae in action.

Bruns has been following changes in the fungal communities at California's Point Reyes National Seashore and the Sierra National Forest. In addition, he and his colleagues have looked at how soil depth, host, root disturbance, detritus enrichment, and prescribed burning affect the mix of fungal species. Their recent work suggests that some fungi build up a “bank” of stored spores that become active after fires and help rejuvenate the mycorrhizae network.

Van der Heijden and his colleagues are testing a radical study method in the dunes of the Netherlands that buffer against floods. He has bulldozed away all the topsoil in a research area and plans to add mycorrhizal species and seeds to some plots but just seeds to others—then monitor the role of those fungi in restoring the landscape. “In this experiment, we specifically test whether the inoculation of soil microorganisms could be used as a tool for nature conservation and restoration,” he points out.

Turnau is convinced that fungi can help. Indeed, many horticulturists already inject mycorrhizae in plant beds to reduce the damage done by fertilizers and to stimulate productivity. Turnau's graduate student Szymon Zubek has found that mycorrhizae may help save an endangered plant, Senecio umbrosus, which right now exists only in botanical gardens, according to Zubek. His work indicates that this flower will have a much better chance in the wild if planted with a mycorrhizal mix that he has worked out. The same is proving true for other rare species.

Heavily polluted sites such as strip mines may benefit from fungal treatment as well, Turnau and others believe. “Many plants are highly dependent on mycorrhizal fungi, especially when grown under stress conditions, such as an excess of heavy metals in soils,” he explains. Often the first plants to repopulate polluted or highly disturbed sites are weeds that don't form mycorrhizae. Only when mycorrhizal fungi move in, Turnau suggests, does diversity blossom. Adding these fungi and a variety of seeds to damaged soil can speed this process.

The mycorrhizae seem to create conditions in which these plants can thrive. For example, one study has shown that when fungi are present, they can modify a heavy metal, such as cadmium, making that toxin unavailable for uptake by the plant. Mycorrhizae also cause soil particles to clump, enabling them to hold on to nutrients. At waste sites, these clumps keep toxins from becoming airborne and dangerous to people, says Turnau.

## Subterranean liaisons

But mycorrhizae may do much more than sequester toxins and make soils more amenable to diverse growth. The fungi manipulate carbon's flow through the environment. Plants are continually shipping carbon in the form of plant sugars—sometimes as much as 20% of their photosynthetic production—down their roots and from there through the mycorrhizae. The more sugar provided, the more energetically the fungi work to provide the plant with other nutrients. But it turns out that the fungi may dole out the “food” they receive to other members of the plant community.

“What the mycorrhizae do is form a vast network, so [the species] in the community are all connected,” says Klironomos. The mycelia pipeline can extend well beyond the immediate partners; it also provides for two-way traffic of carbon resources through the tangled maze of plant roots and fungal threads. In this way, a large oak tree may be feeding not just its fungal partner but also other plants in its neighborhood.

When a seed germinates, it sometimes taps into this nutrient pipeline, Read explains. This is especially true of some orchids, whose seeds are like specks of living dust, carrying no food reserves. On its own, an orchid seed would have little chance. “But if the seed lands and attaches to the network, it has [instant] access to nutrients,” Klironomos points out. And other seedlings struggling to survive in dense forests or underbrush, where they don't get much sunlight, may also find the mycelia to be a lifeline. Carbon from a species different from the seedling's may sustain the nascent plant until it can make its own food. It's an idea that “violates” long-accepted views of how plants establish and maintain themselves in a community, he adds, but it is beginning to catch on.

Certain plants tap into this network as adults to survive. Marc-André Selosse of the National Museum of Natural History in Paris says that researchers now know of 17 such plant families with some members that don't bother with photosynthesis. “They fully rely on the fungus for their nutrition,” he says. Even if they can produce their own sugars, some plants don't bother. They just tap into the underground highway. For example, green orchids can get up to 90% of their carbon through their association with fungi, says Selosse. He is learning how such plants may dispense with their photosynthetic machinery.

At the International Symbiosis Society meeting in Halifax, Nova Scotia, last summer, Selosse explained that some orchids are not picky about which fungi they pair with, opening the way to new and possibly fruitful partnerships. Recently, Selosse examined the fungi that link up with a green orchid called Epipactis microphylla, some individuals of which remain dependent on fungi for sugars even though this orchid can photosynthesize carbon. In DNA studies just published online in Microbial Ecology, he found that these fungi are not the ones typically associated with orchids. (More than 78% of the orchid roots he studied are colonized by truffles.) Selosse suspects that the alternate partners he found enable some orchids to avoid photosynthesis. Over time, the whole species might switch over, if this unusual fungal-orchid relationship becomes dominant. And in this relationship, Selosse concludes, the truffles “appear to be evolution drivers.”

Truffle-hunter Frank would be proud. Truffles and their relatives have gained a lot of respect since he and the German foresters worked with them. Biologists are taking notice. Indeed, says Klironomos, “when it comes to ecology and evolutionary biology, the [mycorrhizal] field is cracking wide open.”