# News this Week

Science  14 Jun 2002:
Vol. 296, Issue 5575, pp. 1944
1. NATIONAL SECURITY

# Research Chiefs Hunt for Details in Proposal for New Department

1. Martin Enserink,
2. Andrew Lawler

President George W. Bush's proposed Cabinet-level department to combat terrorist threats would, on paper, include billions of dollars' worth of federally funded research and hundreds of government researchers. But the hurriedly assembled plan, released 6 June, is sorely lacking in details, leaving government research leaders scrambling to find out if, or how, their labs might be affected. The proposal must now work its way through Congress, where some lawmakers have already put their own ideas on the table.

The plan* was hatched by a small group of White House officials who met secretly for several weeks. Most Cabinet members, presidential science adviser John Marburger, and the institutes most affected by the change had no influence over the plan. “It was as much news to me as to the people in the Cabinet,” says Anthony Fauci, director of the National Institute of Allergy and Infectious Diseases (NIAID) in Bethesda, Maryland, who was informed the day of Bush's announcement by Health and Human Services (HHS) Secretary Tommy Thompson. NIAID would see $1.7 billion in bioterrorism funds in the president's 2003 budget request transferred to the new department. The same bewilderment was expressed by Bruce Tarter, outgoing director of Lawrence Livermore National Laboratory in California, a nuclear weapons facility listed as part of the new department. “Our lab has been cited in a number of media stories today relative to President Bush's proposal to create a new Homeland Security Cabinet-level agency,” Tarter said in a statement issued a few hours before the president's evening television address to the nation. “We have not yet received any official details on this proposal.” The plan would pull together in a single new department, the$37 billion Department of Homeland Security, hundreds of programs now scattered across dozens of federal agencies. A 29-page overview of the proposal pays due homage to science and technology, noting that it provides the United States with a “key advantage” over its more low-tech adversaries. And research against chemical, biological, radiological, and nuclear threats is highlighted as one of the four arms of the proposed department. But apparent mistakes and inconsistencies in the plan left agency managers and researchers wondering exactly what the White House wants to do. “They didn't do their homework,” says one government official.

For example, the report assumes that the new department would gobble up most of Livermore's $1.5 billion annual budget—but only 300 of its nearly 8000-strong workforce—to work on radiological and nuclear countermeasures. In fact, says Marburger, the majority of the lab's funding goes for work on the U.S. nuclear stockpile and won't be transferred to the new department. Still, in an interview on ABC News This Week, White House chief of staff Andrew Card suggested that some of the researchers not working on counterterrorism might be transferred to other weapons labs, such as Sandia National Laboratories, as Livermore becomes what one senior Administration official called “a center of excellence to help us deal with the development of technology” to combat terrorism. Bioterrorism researchers were also unclear about the plan's impact. A total of$2 billion in research to develop drugs, vaccines, and diagnostics—both at NIAID and the Centers for Disease Control and Prevention (CDC) in Atlanta—would be transferred from HHS to the new department. But only 150 researchers of the thousands working in this area would be transferred. An HHS spokesperson says the new department “wants to be able to drive the research agenda,” and the vast majority of scientists would work “on a contractual basis” without leaving their current institutes. But what role the new department would play in setting funding priorities for the National Institutes of Health and CDC remains unclear. “The details are currently being worked out,” says Fauci.

C. J. Peters, director of the Center for Biodefense at the University of Texas Medical Branch in Galveston, says he's opposed to the idea if it “separates bioterrorism from the best thinking in public health.” Last year's anthrax attacks demonstrated that doctors, nurses, and health officials are crucial in detecting and responding to an attack, he says.

A proposal for the new department to take charge of the Plum Island Animal Disease Center, a lab off New York's Long Island that is run by the U.S. Department of Agriculture (USDA), has so far provoked less opposition. The transfer of the entire annual budget of $25 million would make sense, says Plum Island director David Huxsoll, because the lab's focus on foreign diseases such as foot-and-mouth disease and African swine fever would “fit in well” with the new department's overall mission of safeguarding the borders. Huxsoll noted that one of the two USDA departments that uses Plum Island, the Animal and Plant Health Inspection Service, is also being transferred. Under the president's plan, the Department of Defense would not give up the nation's premier biodefense lab, the U.S. Army Medical Research Institute of Infectious Diseases in Fort Detrick, Maryland (see p. 1954). But it would apparently relinquish the proposed$420 million National Biowarfare Defense Analysis Center, requested in the current budget, to study the technology and tactics at bioterrorists' disposal.

Marburger says researchers shouldn't expect too many details at this stage. “This was done in a way to dramatize the scope of this change and generate support for a bold initiative,” he says. “It is still very much in the abstract and will be refined.” Marburger denies that the timing of the announcement—it came the same day an FBI whistleblower delivered damning testimony before Congress about U.S. intelligence gathering—was meant to deflect growing criticism of how the Administration responded to numerous bits of intelligence obtained before the 11 September attacks. “This has been planned for at least a month,” he insists.

Government researchers and managers, reluctant to criticize the White House, say they will wait and see what emerges from Congress, which by law must approve any plan of this magnitude. Several hearings are already in the works, some to explore proposals drafted before Bush unveiled his plan.

2. AIR POLLUTION RISKS

# Software Glitch Threw Off Mortality Estimates

1. Jocelyn Kaiser

The authors of a landmark air pollution study have found a problem with their software application that means they overestimated the risks of fine particles, or soot. The overall conclusions of the group at Johns Hopkins University in Baltimore linking soot and death haven't changed, but the discovery is providing fresh ammunition to industry groups that have criticized the science behind federal air pollution rules issued 5 years ago. The Environmental Protection Agency (EPA) says it will examine whether the rules need to be modified to reflect the new results.

The experience also serves as a cautionary tale to scientists who use off-the-shelf statistics software without questioning what's inside. The Hopkins group “is very good and very careful,” says Stanford University statistician Trevor Hastie, yet they used the program for 5 years before catching the problem.

The research, an ongoing project known as the National Morbidity, Mortality, and Air Pollution Study (NMMAPS), is led by Hopkins epidemiologist Jonathan Samet and biostatistician Scott Zeger and funded by the nonpartisan, nonprofit Health Effects Institute (HEI) in Cambridge, Massachusetts. Started in 1996, the project expands on earlier studies in several cities documenting that when daily levels of tiny soot particles rise, slightly more people die from heart and lung disease. These so-called “time series” studies helped persuade EPA to issue its first regulations limiting permissible levels of very fine particulate matter (PM), known as PM2.5, in 1997.

In NMMAPS, the Hopkins scientists sought to determine whether the case against fine particles held up across a much larger number of cities—90 in all. Such time-series studies are tricky because they seek to disentangle the role of particles from other factors that can also boost death rates, such as heat waves. The team used a model, the Generalized Additive Model (GAM), that is part of S-plus, a widely used statistical software package. The software searches for a pollution effect and smooth functions of the confounding variables in an iteration that continues until the results don't change much.

Since NMMAPS began, the Hopkins team has published more than a dozen papers linking fine particles and premature deaths (Science, 7 July 2000, p. 22). But about 10 weeks ago, says Zeger, “something struck me as funny about the way the software was working.” Eventually, his team figured out that the trouble was an S-plus GAM default setting. The software was set to stop calculating when a certain result differed from the previous one by 0.001. But the Hopkins researchers realized that because they were looking at a tiny rise in daily death rates, they needed to keep going. When they changed the default from 10−3 to 10−15, they got slightly different risks for most cities (see graph).

Their revised result for all 90 cities was a 0.27% rise in mortality per 10 micrograms per cubic meter (μg/m3) of PM10 (a class of particles that includes PM2.5) compared with 0.41% per 10 μg/m3 in the original study. The NMMAPS group informed HEI and is notifying the journals that published its papers.

Industry groups are crowing. Allen Schaeffer, executive director of the Diesel Technology Forum, says the error suggests that more work should be done before the current regulations are fully implemented. “If the risks have been exaggerated, we have to understand the real risks,” he says.

Industry complaints aside, both scientists and EPA officials say that the S-plus problem does not undermine the 1997 soot rule. “The underlying relationship is still solid,” says John Bachmann of EPA's air office. Indeed, another type of study looking at how death rates vary in polluted cities over many years makes an even stronger case against fine PM than the daily studies, notes HEI president Dan Greenbaum. However, Bachmann says, “a few” of the time-series studies that EPA drew on to set the daily limit on fine particles “used the same S-plus approach.” Agency scientists will take that into account in their latest review of PM2.5 science, which will delay the next version of the rule. Bachmann says the standard “could” change, but “it's too soon to tell.”

Scientists in other disciplines, from economics to genomics to ecology, use the S-plus GAM model. David Smith of Insightful Corp. in Seattle, which sells S-plus, says “it's really hard to say” whether many other researchers have had this problem, but his inquiries to some 2000 S-plus users on an e-mail list last week suggest not. Hastie, who co-wrote the S-plus GAM, says these pollution studies are “an unusual situation” because “they're doing very fine-scale modeling, and the effects are very small.”

Biostatistician Gerald van Belle of the University of Washington, Seattle, notes that a recent journal article pointed out that defaults can also gum up results with a popular stats package called JMP. Says van Belle: “99% of people are going to be working on problems for which the default settings are appropriate.” But when their problem is unusual, he says, they might need to take a look inside the box of their statistics package.

3. DOE WEAPONS LABS

# Livermore Keeps It All in the Family

1. Andrew Lawler

The appointment of an insider to head Lawrence Livermore National Laboratory ends a politically charged search that highlighted the sharp tensions between the lab's managers, the University of California (UC), and its boss, the Department of Energy (DOE). The new director, theoretical physicist Michael Anastasio, takes the job just as President George W. Bush has assigned the lab a more visible role in U.S. homeland defense (see p. 1944).

Anastasio, 53, was appointed 4 June to succeed Bruce Tarter, who is stepping down 30 June after 8 years as head of the $1.5 billion nuclear weapons lab. “He's the safe choice,” says one Livermore researcher about the 20-year Livermore veteran, who has led the division that designs plutonium triggers in nuclear weapons as well as the effort to ensure the safety and reliability of those weapons without testing them. In a press conference, Anastasio backed the Administration's policy not to test nuclear weapons and pledged good relations with the university and Los Alamos National Laboratory in New Mexico, Livermore's longtime rival. Those relationships require much mending. The university, which operates both Livermore and Los Alamos for DOE, came under withering fire this spring for attempting to appoint Ray Juzaitis, a senior administrator at Los Alamos, to head Livermore (Science, 3 May, p. 821). Juzaitis eventually withdrew from consideration, and last week a chastened UC president Richard Atkinson took responsibility for the episode, saying, “I failed to communicate with the key people. … It was my fault.” Both Livermore and Los Alamos have been criticized heavily in the past 2 years for cost overruns, breaches in national security, and alleged racial profiling and discrimination. DOE has pressured the university to tighten its managerial reins and reduce the traditional rivalry between the two labs. Choosing a Los Alamos employee to head Livermore was part of a strategy directed by John McTague, a former science adviser to President Ronald Reagan who now oversees the labs for the university. But when word leaked in April that Juzaitis was the favored candidate, Livermore's supporters went into high gear. They complained to the White House, DOE, and lawmakers that he was too junior—and that he had overseen the division that included Wen Ho Lee, a physicist accused of improperly copying classified material. On orders from DOE, Atkinson abruptly canceled a press conference at which he was to announce the new director. The following week, Juzaitis declined what he calls a firm job offer. In a 30 April letter to Atkinson, Juzaitis says he withdrew because of “negative reactions in Washington, within the university, and at Livermore.” He also decried the “unwarranted linking of my name to the Wen Ho Lee affair.” Government officials who decline to be identified complain that UC officials did not reveal Juzaitis's link with Lee in discussions with Administration managers and congressional lawmakers. “It shows a complete lack of political savvy,” says one. Representative Ellen Tauscher (D-CA), who represents the Livermore area, at the time criticized UC's failure “to be sensitive to national security, the culture, and the unique qualities of the laboratory as it selects a new director.” McTague, through a spokesperson, declined to comment on the selection process, although he praised Anastasio as “extremely impressive.” The appointment appears to have mollified Tauscher and other Livermore supporters. UC “made the right choice with Anastasio,” Tauscher said last week, adding that he “has always been the best choice for lab director.” Anastasio will have his work cut out for him. One key issue involves the allocation of time on the National Ignition Facility (NIF)—a$3.5 billion laser facility. NIF is being built to conduct tests to ensure the efficacy of existing nuclear weapons, but it also offers a platform for basic researchers. The new director must also deal with several suits charging the lab with racial discrimination in hiring and promotion. But the president's homeland security proposals no doubt will be on the front burner as Livermore charts an uncertain new course in protecting the United States from terrorists wielding chemical, biological, or nuclear weapons.

4. SPINTRONICS

# Magnetic Gate Opens New Computing Path

1. Adrian Cho is a freelance writer in Boone, North Carolina.

A tiny device that answers “no” when it's told “yes” and vice versa could mark the first step toward microchips that calculate magnetically, a team of physicists reports. The “NOT gate,” described on page 2003, uses a trick of geometry to manipulate magnetism as conventional devices do electric charge. “It's just very clever,” says Craig Lent, an electrical engineer at the University of Notre Dame in Indiana. “They're on the road to nanomagnetics.”

Electronic microchips crunch numbers by shuffling dollops of charge. But physicists and electrical engineers are striving to harness a more subtle property of electrons: the fact that the particles behave like spinning tops and are magnetized along their spin axes. Burgeoning “spintronics” technologies aim to use magnetic materials and magnetically polarized currents to store bits of information and perform calculations more efficiently (Science, 16 November 2001, p. 1488). So far researchers have developed devices that use layers of magnetic materials to read data from densely encoded disk drives or to store data in memory chips even when they're turned off (Science, 12 April, p. 246). They've also begun to manipulate magnetically polarized electrical currents flowing within nonmagnetic semiconductors, an approach that might lead to more efficient calculations and even to superfast “quantum computing.”

But researchers have yet to perform calculations with just changes in magnetization and no flow of electric charge. The new device, developed by Russell Cowburn and colleagues at the University of Durham, U.K., is a step in that direction.

The gate consists of a simple track of naturally magnetic nickel-iron wire, shaped like an upside-down Y. The magnetism of the alloy naturally runs parallel to the track, but it can be made to flip direction within a short length of the wire. In that case, the two opposing magnetizations meet at a region called a “domain wall.” There they either both point toward the domain wall (head to head) or both point away from it (toe to toe). Those two magnetic configurations can be used to encode 0 and 1 values for bits of information.

Cowburn and colleagues found a way to switch between the two arrangements by using a magnetic field to force the domain wall through a kink—the stem of the inverted Y. When a magnetic field points along a branch of the device, it pushes the domain wall along the track so that more of the wire is magnetized in the same direction as the field (see figure). If the track made a smooth curve, a rotating magnetic field would simply ease the domain wall around the bend. In the Y-shaped device, however, something else happens. As the domain wall moves up into the stem of the Y and down again, it flips from the head-to-head configuration to the tail-to-tail configuration or vice versa, something like a car backing into a driveway to turn around. By swapping domain-wall configurations, the device exchanges 0 for 1 and 1 for 0—the hallmark of a logical NOT gate.

Cowburn and colleagues have strung as many as 11 NOT gates together in a closed loop. The devices kept flipping bits faithfully while a domain wall went around as many as 100,000 times. Unlike some other budding technologies, the device also works at room temperature. Arrays of NOT gates can do little by themselves, the researchers acknowledge, but they hope to develop other devices, such as an “AND gate” that can compare two inputs, that will enable them to perform full-scale calculations. “We think we'll have a fully functioning logic [system] within a year,” Cowburn says.

Chips that manipulate magnetism should resist damage from radiation and retain information if they inadvertently lose power, so they might be useful in spacecraft and other harsh environments, says Russell Beech, an electrical engineer at NVE Corp. in Eden Prairie, Minnesota. However, Cowburn and colleagues must address some basic questions if they're to turn their promising idea into a useful technology, Beech says. For example, they must find ways to reliably feed domain walls into a circuit and to generate the rotating magnetic field from wires embedded in the chip itself.

But even if the new device does not blossom into a new technology, it could give researchers an important tool for probing the basic physics of magnetic materials, says David Awschalom, a physicist at the University of California, Santa Barbara: “It's a wonderful laboratory for studying domain wall motion.” Such studies should prove fruitful however the story of spintronics unwinds.

5. GENETICALLY MODIFIED FOOD

# TV Drama Sparks Scientific Backlash

1. Ben Shouse*
1. Ben Shouse is a writer in New York City. With reporting by Adam Bostanci in Cambridge, U.K.

Intending to discredit a biotech company, a farmer opposed to genetically modified (GM) foods slips a gene conferring resistance to the powerful antibiotic vancomycin into wheat. The protest goes horribly wrong, however, when the resistance gene moves from the wheat into the bacterium Staphylococcus aureus, provoking a deadly and uncontrollable outbreak of staph infections.

Sounds incredible? The plot of the BBC thriller “Fields of Gold,” which aired 8 and 9 June in the United Kingdom, is indeed far-fetched, many experts say. Some scientists, concerned that the alarming story line will erode already low public confidence over the safety of GM crops, mounted a high-profile attack last week. Leading the charge was Robert May, president of the Royal Society, the U.K.'s preeminent scientific body. “Fields of Gold,” he said in a statement, is “ludicrous” and “hysterically inaccurate” and is “propaganda.”

The vehement protests included calls for BBC to pull the program before it aired. BBC did not comply, although it posted on one of its Web science pages what could be construed as a disclaimer about the show's premise, noting that “most scientists think that the risk to our health from this is remote.”

The program was scripted by Ronan Bennett, known for his politically charged fiction about Northern Ireland, and Alan Rusbridger, an editor at The Guardian, a highly respected left-leaning U.K. newspaper. Last summer, the two asked University of Cambridge geneticist Mark Tester to vet their script. Tester provided suggestions to make the script more plausible, including a possible means for mixing GM wheat with the bacteria. But he harshly criticizes the accuracy of the final product, especially the ease with which the fictional farmer transfers the gene into the plant and the highly improbable transfer of the resistance gene from the wheat into bacteria. The program, he told Science, “raises concerns that have no scientific basis.”

Bennett says that he and Rusbridger rewrote some scenes according to Tester's advice. Moreover, he says it was Tester who brought the idea of horizontal gene transfer to their attention in the first place. When Tester reviewed the script, Bennett says, “he did not take objection to any of it, he just made suggestions.” Indeed, in an e-mail to the BBC in July 2001, Tester said he remained “open-minded” about the gene transfer scenario. As for the demand for the show's cancellation, Bennett says, “it has a whiff of book burning about it.” Rusbridger claims that many scientists gave positive feedback. He also points out that U.K. officials have expressed concern over horizontal transfer in government documents.

Despite the scant research on horizontal transfer of transgenes, several scientists who saw the program agree that it is implausible. “If genes moved with any frequency from plants to bacteria, we'd find them in bacterial genomes. We don't,” says Peter Lund, a molecular biologist at the University of Birmingham, U.K. An outbreak of vancomycin-resistant bugs, Lund and others say, is much more likely to come from existing resistant bacteria than from a plant.

But Lund adds that the show reflects real concerns among Britons: “We have no trust in politicians or big business, and very limited trust in or understanding of science.” Even a critic from the conservative Times of London sympathized with the urge to confront the dangers of GM crops: “The public have been taught fear and skepticism not by radical agitators but by businessmen and their political allies, who were prepared to take irresponsible risks with our health.”

The show's anti-GM message comes at a particularly inopportune time for the U.K. government, which announced last week that it would sponsor televised national debates on GM food safety this summer. Few, however, believe that television, whether in the form of public debate or a GM thriller, will reconcile such bitterly opposed viewpoints.

6. ECOLOGY

# A Coral by Any Other Name ...

1. Elizabeth Pennisi

Although the undersea landscape is peppered with corals of many shapes and sizes, there's no consensus about whether the different configurations denote different species. Some researchers think so and call each by a different name. But others argue that because many corals interbreed, they do not qualify as distinct taxonomic entities. Now on page 2023, Harvard University researchers say they have set the record straight, as least for three species found in the Caribbean. By performing more extensive genetic studies than had been tried before on corals, Harvard marine biologists Steven Vollmer and Stephen Palumbi have shown that appearances can be deceptive: Three “species” are really two.

The work is as noteworthy for its techniques as for its findings. Harilaos Lessios, an evolutionary biologist at the Smithsonian Tropical Research Institute in Panama, describes the study as “a first-rate application of molecular markers to solve an evolutionary problem that morphology and conventional crosses between gametes were unable to solve.” Although not everyone agrees, Lessios thinks the results will clarify when a particular coral is a separate species.

Coral experts had suspected that many coral species were promiscuous. In a maritime orgy, dozens of corals release their gametes on the same few nights once a year. Occasionally, sperm of one species pair with eggs of another and hybrids result—at least that's what lab tests indicate. But whether hybrids survive, or, more important, whether they can reproduce sexually, has been a matter of debate. If they can reproduce sexually with other hybrids, they have the potential to split off as new species, thereby contributing to the evolution of these organisms.

Vollmer and Palumbi evaluated DNA from three Acropora species”: staghorn, elkhorn, and Acropora prolifera, sometimes called fused staghorn. They focused on some noncoding sequences, or introns, from two genes and also some mitochondrial DNA. As expected in species, the mitochondrial DNA and the introns in both copies of each gene in the elkhorn were readily distinguishable from the staghorn's. But A. prolifera had one copy of each gene from each of the two species, indicating that A. prolifera is a first-generation descendent of elkhorn and staghorn corals.

The relatively recent origin of the hybrid was unexpected and suggests, at least to Vollmer and Palumbi, that A. prolifera doesn't warrant designation as a separate species. Had hybridization occurred long ago and the hybrid reproduced sexually, as some researchers suspected, then A. prolifera might have had time to evolve into its own species. Alternatively, others expected that the elkhorn and staghorn corals interbreed so much—with each other and with the hybrids—that their genomes would be too similar for biologists to call them separate species.

But neither hypothesis is quite correct, say Vollmer and Palumbi. Because this hybrid only very rarely reproduces sexually, it seems to be an evolutionary dead end even though it can live a long time and propagate asexually. Vollmer calls it an “immortal mule.”

But not everyone is convinced about this interpretation. “How common [these first-generation hybrids] are going to turn out to be is hard to know,” cautions Nancy Knowlton of the Scripps Institution of Oceanography in La Jolla, California. Bette Willis, a coral expert at James Cook University in Townsville, Australia, points out that the new data indicate that some interbreeding occurs between the hybrid and its parents, so the results actually support the idea that corals tend to be too intermingled to qualify as separate species. Contrary to what Vollmer and Palumbi say, “the paper adds to a growing body of evidence that [interbreeding] has played a role in the evolutionary history of the coral genus [with the most species], Acropora,” she suggests.

Figuring out what makes a coral a distinct species is not just academic, Vollmer, Palumbi, and others say. As Richard Aronson, a marine biologist and paleoecologist at the Dauphin Island Sea Lab in Alabama, points out, clarifying what is a species that can reproduce sexually can help conservationists decide which corals to protect.

7. ARCHAEOLOGY

# Millions Pledged for Afghan Restoration

1. Andrew Lawler

Afghan leaders are gathering this week in Kabul to map out the country's political future. Two weeks earlier, international donors met in the capital city to confront another daunting challenge—restoring Afghanistan's cultural and archaeological heritage after 23 years of strife. The meeting, blessed by the interim government and the United Nations Educational, Scientific, and Cultural Organization (UNESCO), produced millions of dollars in pledges to rebuild and restore structures and provide the human resources needed to maintain them. But participants agreed that an even more important ingredient is political stability.

The proposed restorations do not include rebuilding the famed Bamiyan Buddhas that the Taliban destroyed (Science, 9 March 2001, p. 1873). After much discussion, the participants agreed to leave that decision to the new government. But the plan would stabilize the fractured cliffs that sheltered them, build a new Kabul Museum to replace the one destroyed by war, and rejuvenate the country's archaeological institutes and its scattered community of researchers. Most Afghan researchers are “dead, wounded, or gone,” notes Robert Knox, Oriental antiquities chief at London's British Museum, who attended the meeting.

Afghan art and culture are a unique blend of civilizations—Greek, Persian, Indian, and Chinese—that have influenced the region. That rich amalgam has attracted support from European as well as Asian governments, institutes, and foundations. “We received lots of pledges; everyone is strongly motivated,” says Christian Manhart, UNESCO's Asian cultural heritage division chief. The notable exceptions are the British and U.S. governments, which instead are focusing on military and humanitarian aid.

A Japanese foundation has pledged $700,000 to begin a project this summer to shore up the Bamiyan cliffs, weakened by the explosives the Taliban used. After surveying the structure, engineers will insert steel cables and concrete. “There is great danger that the niches [surrounding the site] will collapse,” says Paul Bucherer-Dietschi, a Swiss architect who has visited Bamiyan and championed the rebuilding of the statues. There are also tentative plans for a museum and a sound-and-light show at the site. There were a few happy surprises for the scholars visiting Kabul, including invaluable painted terra cotta statues from Fondukistan that were saved by Afghan archaeologists. But the Kabul Museum, now an empty shell, lost 80% to 90% of its collections. Greece has promised$750,000 to reconstruct the current building, which is 8 kilometers from the city center. Others argue for a more central location, but the decision will be up to the new government.

Rebuilding a research community will require more than money and political will. The German Archaeological Institute of Berlin has offered $350,000 to jump-start Kabul's own archaeological institute, part of the country's academy of sciences, and the University of Kabul also has an archaeology department. “Both are completely wrecked” from the prolonged fighting, says Knox. “But the real problem is, who do you train?” A half-dozen Afghan archaeologists attended the Kabul meeting, but the rigors of life in modern-day Kabul might discourage prospective students and returning scholars. The German foreign ministry has promised$350,000 to conserve a number of archaeological sites. One of the most fragile is the Minaret of Jam, the second highest in the world. Built in the 12th century, the structure is leaning because its foundation is weakening. Italy will provide $800,000 for repairs and an additional$500,000 for other conservation projects.

Most of the pledges will be held in trust by UNESCO. But the largest gift—$5 million from the Aga Khan Trust for Culture, a Geneva-based organization—will go directly to the municipal government to recreate the Bagh-e-Babur gardens, planted by the famed first Moghul emperor and the site of his 16th century tomb. The funds also will help rebuild traditional Kabul residences and improve public sanitation. Interim President Hamid Karzai spoke at length to participants about the importance of rescuing and rebuilding the country's cultural heritage. But any long-term restoration requires an end to chaos and war, leaving scientists at the mercy of the political winds. “No one really knows what will happen,” says Bucherer-Dietschi. 8. EXOPLANETS # 'New Jupiter' Turns Up in Strange Company 1. Donald Goldsmith* 1. Donald Goldsmith, the author of Connecting With the Cosmos: Nine Ways to Experience the Wonder of the Universe, is a science writer in Berkeley, California. Planet hunters are expected to announce this week that they have discovered 13 new planets that orbit sunlike stars in our corner of the Milky Way. This new harvest, which brings the total number of known extrasolar planets close to 90, includes an exoplanet that orbits farther from its star than Jupiter orbits the sun—the greatest known star-planet distance of any exoplanet. “This [far-out planet] is the first good evidence that planets actually form where our basic theories say they ought to,” says Charles Beichman, a planetary expert at the Jet Propulsion Laboratory in Pasadena, California. The new planets were discovered by Geoffrey Marcy and Debra Fischer of the University of California, Berkeley; Paul Butler of the Carnegie Institution of Washington in Washington, D.C.; and Steven Vogt of the UC Observatories/Lick Observatory near San Jose, California. The astronomers used telescopes in California, Hawaii, and Australia to measure Doppler shifts in the spectra of light from the planets' stars. The Doppler shifts arise as the orbiting planets tug the stars in different directions. Like most other extrasolar planetary systems discovered so far, the newcomers include several enormous objects orbiting extremely close to their host stars. It's not surprising that star-hugging super-Jupiters unlike anything in our own solar system were the first to be discovered by the Doppler method, Marcy says. Not only do they exert the greatest gravitational pull on their stars, but their short orbital periods make their effects on starlight show up relatively quickly in astronomers' observations. But that doesn't mean that massive, close-in planets are the rule, Marcy says. View this table: The new far-out planet brings mixed evidence that solar system-like star systems are indeed out there. With a mass at least 4.05 times that of Jupiter (that is, 1290 Earth masses), the planet orbits the star 55 Cancri every 14.8 years at an average distance of 5.8 astronomical units (AU), slightly greater than Jupiter's distance of 5.2 AU. (An astronomical unit is the average distance of Earth's orbit from the sun, about 150 million kilometers.) So far, so familiar. The catch is that 55 Cancri also hosts two large, close-in planets, one just 0.12 AU from the star, the other 0.24 AU. Planets such as 55 Cancri's inner companions vex astrophysicists with two big questions: How did they get there? And what do they say about the likelihood that planets with Earth-like masses and orbits exist in the same system? Current theories imply that Jupiter-like “fluid giants” must form at least 3 to 4 AU from the star. To explain how they might move closer, Douglas Lin of the University of California, Santa Cruz, and others have proposed that newly formed giant planets follow a shrinking disk of preplanetary material that swirls into the star, drawing the planet inward by tidal interactions. Fred Rasio of Northwestern University in Evanston, Illinois, and others support a planetary pinball game scenario in which a close encounter between two massive planets shoots one into a much tighter orbit and the other into a somewhat larger orbit around the star. If the migration model is correct, Earth-like planets could potentially form after a giant planet had spiraled inward, if additional matter remained in a disk around the star. In the planetary-pinball model, however, the interactions between a close-in and a far-out giant would sweep up or kick out any Earth-like planets orbiting in between them. The next few years should bring a flood of new data to help resolve these possibilities, says Marcy's longtime collaborator Butler. With 1200 stars now under study and more to come, another half-dozen years of observations should bag several more Jupiter-like planets in Jupiter-like orbits, Butler says—objects that might help determine how hospitable, or hostile, the universe is to planets like our own. 9. ASTRONOMY # Lucky Catch Identifies Disintegrating Cluster 1. Govert Schilling* 1. Govert Schilling is an astronomy writer in Utrecht, Netherlands. ALBUQUERQUE, NEW MEXICO— The Milky Way galaxy is tearing apart its oldest inhabitants, and for the first time astronomers are witnessing the slaughter. A striking image presented here 3 June at a meeting of the American Astronomical Society* shows a globular cluster known as Palomar 5 being torn asunder by tidal forces of our home galaxy. Researchers say the observation and others like it could shed new light on the distribution of dark matter in the halo of the Milky Way. “This is a very exciting and beautiful result,” says theorist David Spergel of Princeton University. Globular clusters are large, spherical aggregations of old stars. They probably formed along with the galaxy, some 12 billion to 14 billion years ago. About 150 globulars are known to loop in wide orbits around the Milky Way's center. Some of them contain millions of stars; others are more sparse and diffuse. It's these loose systems that run the highest risk of being torn apart by tidal forces, which arise because the Milky Way's gravity is stronger on one side of the cluster than on the other. Astronomers have suspected that this fate might befall some clusters, but convincing direct evidence was missing. Now there's definitive proof in data from the Sloan Digital Sky Survey, a large international project to map one-quarter of the sky in exquisite detail (Science, 25 May 2001, p. 1472). A team led by Eva Grebel of the Max Planck Institute for Astronomy (MPIA) in Heidelberg, Germany, scrutinized stars near Palomar 5, a sparse globular cluster 75,000 light-years from Earth. On opposite sides of the cluster the astronomers found streams of stars stretching 13,000 light-years from end to end—20 times the apparent width of the full moon. These “tidal tails” form when stars are torn loose from the cluster and then slowly drift away. “It's a very amazing structure,” says team member Michael Odenkirchen, also at MPIA. “Nothing like this has ever been seen before.” Odenkirchen expects the cluster to disappear completely within 100 million years. To discover the tails, astronomers had to filter out the countless stars and background galaxies in the field of view that did not match the expected colors and brightnesses of globular cluster members. The finding suggests that many other sparse globulars have been torn apart completely in the past; detecting Palomar 5 in the process of being ripped to pieces was apparently just a lucky catch. Because the tails more or less delineate the orbit of the parent globular cluster around the Milky Way center—information that's impossible to come by otherwise—the data will help scientists map the distribution of dark matter in the Milky Way, Spergel says: “The tidal tail observations should enable astronomers to measure both the lumpiness of the dark matter and its central density.” That kind of information can help scientists rule out or refine models of what dark matter is and how it has shaped the evolution of the universe, he says. • *200th meeting, 2–6 June. 10. CLINICAL TRIALS # Agency Wants to Stop Shopping for Best Deal 1. David Malakoff Every child knows that if Mom says no, she can always ask Dad. Now, the U.S. Food and Drug Administration (FDA) is worried that clinical researchers might try the same trick: Find a new set of safety officials to approve a study involving human subjects that has been rejected by another panel. To deter such behavior, the government has proposed that researchers be required to tell safety panels about any prior reviews. But last week the biomedical community gave the plan* a mixed reception. The safety panels—known as Institutional Review Boards (IRBs)—must approve all research involving human subjects. Most academic researchers have no choice but to submit their plans to the IRB at their home institution. But a drug company sponsoring a multisite trial has more options, from submitting plans to multiple boards to employing a private “superboard” to which individual institutions have ceded authority. The FDA proposal is a response to a 1998 report by the department's inspector general that concluded new procedures were needed to prevent researchers from “IRB shopping” to find a more agreeable review panel. But critics aren't buying FDA's solution. They argue that shopping is rare and that disclosure rules would do little to improve patient protection. The debate “is part of a larger discussion about how to overhaul human subject protection” in the wake of the deaths of several study participants and a government crackdown on informed consent procedures at major research universities, notes Abbey Meyers, president of the National Organization for Rare Disorders in New Fairfield, Connecticut. Some respondents to FDA's request for comment doubt that shopping is a problem worthy of federal regulation. Officials at drug giant Merck & Co. in West Point, Pennsylvania, for instance, said they could identify only one instance in more than 1500 clinical trials the company has sponsored over the last 5 years where researchers even discussed approaching a second IRB when they were unhappy with a first ruling. Even that case, says company vice president David Blois, occurred only after the first panel imposed legal requirements that boosted study costs. He recommends that FDA delay regulation until more data are available. But some veteran reviewers say that big companies try to intimidate an IRB if it suggests changes in protocols or consent forms. Barbara Bigby, head of research subject protection at the Scripps Clinic in La Jolla, California, noted that research sponsors “frequently” tell her reviewers that they are the only ones expressing such concern. It's a tactic, she says, that's meant to get the panel to back down. “Yet when we communicate [with other IRBs that have reviewed the proposed study],” she wrote, “we find that our concerns are similar” to theirs. Commenters also disagreed about who should notify IRBs. Companies nominated investigators, saying they are closest to the studies. But the American Society of Gene Therapy spoke for many researchers and universities in arguing that sponsors should bear the burden, especially in clinical trials that might stretch across hundreds of domestic and foreign sites. There was more agreement on other issues. Both supporters and critics of disclosure warned that tracking multisite studies could cause paperwork-induced gridlock for researchers, institutions, and IRBs alike—although a Web-based filing system could help. They also worried about a “herd mentality” in which IRBs at large or prestigious research centers would set the pace. Some researchers fear that shoppers could avoid disclosure simply by tweaking their proposals to make them appear novel. There were also questions about how offenders might be punished. FDA officials are expected to spend several months chewing on the comments before deciding on their next move. In the meantime, the scientific community is already taking some steps in the direction that FDA might be headed. Biomedical researchers report that some IRBs have already begun to ask researchers about prior decisions. And the Association of American Medical Colleges (AAMC), which offered qualified support to disclosure, believes that a new voluntary accreditation program for institutions conducting human research will standardize reviewing practices. That step, says AAMC, would reduce the chances that researchers can get a better deal at the next IRB. 11. FORT DETRICK # On Biowarfare's Frontline 1. Martin Enserink Heightened fears of bioterrorism have shone the spotlight on the Army's biodefense lab— and pulled its researchers out of their isolation FORT DETRICK, MARYLAND— To outsiders, the research at this Army base an hour north of Washington, D.C., has always seemed slightly sinister. During World War II, scientists here embarked on a large, top-secret program to develop biological weapons that could kill thousands, even millions. Although that was ended in 1969 and scientists switched to purely defensive research, rumors about clandestine experiments in underground labs persist to this day. Even in the scientific world, the researchers have remained the odd ones out, studying exotic diseases that might cripple an army but have actually infected few people and that most other researchers cared little about. Now, everything is different. In the post-9/11 world, the expertise built up at the U.S. Army Medical Research Institute of Infectious Diseases (USAMRIID), the main research institute at Fort Detrick, has proven invaluable—and suddenly everyone is grateful. The lab is working closely with the Federal Bureau of Investigation (FBI) to help unmask whoever sent the anthrax letters last fall; the so-called Brokaw, Daschle, and Leahy letters were sent here to be investigated; and the lab serves as the repository for anthrax strains subpoenaed from other labs. Tens of thousands popped Cipro last fall because USAMRIID's Arthur Friedlander showed it could protect monkeys from anthrax. The institute is garnering some scientific respect, too. After two recent visits, Anthony Fauci, director of the National Institute of Allergy and Infectious Diseases (NIAID) in Bethesda, Maryland, says he was “quite surprised” by the lab's expertise and technical capabilities. Fauci hopes to collaborate closely with the lab now that President George W. Bush has requested a whopping$1.7 billion in new bioterrorism funds for NIAID. “This has pulled us out of the intellectual backwaters and into the mainstream,” says senior USAMRIID researcher Peter Jahrling, who studies smallpox and viral hemorrhagic fevers.

But at the same time that USAMRIID is helping in the bioterror investigation, the institute is under intense scrutiny as the potential source of the anthrax—and perhaps even the terrorist. The research strain called Ames, used in the letters, made its debut here after being isolated from a Texas cow, although it was subsequently sent on to more than a dozen other labs. The powder's sophistication suggests to some that the bioterrorist had connections to the earlier offensive biowarfare program, and the FBI has recently started giving employees lie detector tests. To complicate matters further, the lab has gone through several high-profile PR hiccups in recent months— including a finding of anthrax spores outside a high-containment facility that is still under investigation.

## Bugs and brains

At first, Fraser didn't intend to study microbes; 25 years ago she was set on a career in medicine. But when she was a premed student in her senior year at Rensselaer Polytechnic Institute in Troy, New York, she opted to do an independent research project on nearby Lake George. She was looking for bacteria that could feed on pesticides and fertilizers and thus clean up pollution, a field now known as bioremediation. The taste of research made her hungry for more, although she hadn't settled on a field.

Indeed, she chose her graduate school—the State University of New York (SUNY), Buffalo—not because it was the best academic fit but so she could be close to a boyfriend in Toronto. “I ended up in a graduate program in Buffalo probably for the wrong reason,” she concedes. Nonetheless, excited about the possibilities, she began checking out labs in her home department, pharmacology.

Fraser found a place in a lab Venter was running at SUNY—a move that would transform her life in unexpected ways. Venter was then studying how neurotransmitters interact with their receptors, and she joined in. Four years after working together, the two married in 1981, the same year she earned her doctorate. “Unlike so many men I had encountered, he was the first who was not the least bit threatened by my potential for success,” says Fraser.

Three years later, she and Venter moved to the National Institutes of Health (NIH), “a dream come true,” Fraser recalls. She continued studying brain receptors, first in Venter's group at the National Institute of Neurological Disorders and Stroke. By 1987 Venter had moved on to making gene sequencing faster and more efficient, and Fraser had taken on her own research group. But in 1988 she was denied tenure. “It was the first and probably only time that the old-boy attitudes of science really hit home,” she says. Too “reticent” to fight the decision, Fraser instead started a new lab in molecular neurobiology at the National Institute on Alcohol Abuse and Alcoholism, where her group looked for receptors important in addiction and alcohol tolerance. But when Venter resigned from NIH in 1992 and invited Fraser to join him at a new nonprofit research institute, she didn't hesitate.

Together they became leaders in a gene-discovery project that Venter had tried to convince higher-ups at NIH to invest in but ultimately launched with private money. It worked by capturing random fragments of DNA—called expressed sequence tags (ESTs)—and using them to reconstruct genes. After NIH turned down his request to scale up his operation, a private investor offered to set Venter up in a specially tailored arrangement, giving him a nonprofit research center that would be partnered with a biotech company called Human Genome Sciences. Venter agreed, and TIGR's doors opened in 1992.

Despite her excitement about the new EST project, Fraser was not convinced she was in the right place. “I was not fully into the genome thing,” she says. She was hoping to use ESTs to learn about a group of cell-signaling molecules called G protein-coupled receptors, important in a wide range of nerve cells, including those involved in taste. But TIGR's agreement with Human Genome Sciences imposed an 18-month delay on the publication of many of her discoveries. “I was increasingly frustrated,” she recalls.

By early 1995, Venter and his crew were midway through sequencing the first microbial genome, Haemophilus influenzae, and were eager to take on a second—Mycoplasma genitalium, which has the smallest genome among free-living bacteria—hoping to begin to define the minimum requirements for life. They were using a fast new approach called whole-genome shotgun sequencing, a method that some had called unworkable. Fraser agreed to shepherd the Mycoplasma genome to completion. Some 580,070 bases later, she was hooked. “I had never been so intellectually engaged as when the first gene list was in hand and we had to make sense of the information,” she recalls. She and her colleagues spent weeks perusing biochemistry books and journals, “trying to reconstruct the biochemistry of these bugs.”

## Hitting her stride

Although TIGR's agenda was growing, it faced an unexpected challenge in 1998: Venter left to create a sequencing company down the road, Celera Genomics. When Fraser took the reins, no one was quite sure what to expect. “People were cautious and curious” to see how well she would do, says molecular and cell biologist Rob Fleischmann, who was one of the first scientists TIGR lured away from NIH.

Any doubts are now gone. “TIGR has flourished since [Venter's] departure,” says Cold Spring Harbor's McCombie. When TIGR broke with Human Genome Sciences and lost some $38 million in support, Fraser helped her researchers get grants—a new strategy for the institute. Today, TIGR scientists are funded primarily by the Department of Energy, NIAID, and the National Science Foundation. There were some tradeoffs, however. “Some people miss Craig's approach of always wanting to stir the pot,” says Fraser. Impatient, he'd get an idea, call spur-of-the-moment meetings, and expect his ideas to be implemented by the end of the week. “He took up an awful lot of time,” she adds. Fraser has guided TIGR with a gentler hand. “When [Venter] was around, people scurried,” says Fleischmann. In contrast, “there's a calmness in [Fraser's] leadership.” TIGR molecular biologist Malcolm Gardner agrees: “She is perhaps a more steadying influence at the top.” Researchers seem to have more time to pursue their own interests, he adds. Fraser mentors young scientists, such as TIGR microbiologist Karen Nelson, and also goes to bat for them. When several funding agencies turned down Gardner's proposal to sequence a relative of the malaria parasite along with the parasite itself, thus enabling comparative studies, “[Fraser] was so taken by the science that she went to the board of trustees” and got the necessary funds from TIGR's private endowment. Under Fraser, TIGR now has a broader focus, expanding its analyses to include more comparative genomics and studies of gene function. Scientists at TIGR are using the institute's vast data collection to study evolution, probe for traits that create pathogenicity, and trace changes in gene expression over time. “They are getting more university-like and doing more science as well as genomics,” says W. Ford Doolittle, an evolutionary biologist at Genome Atlanta at Dalhousie University in Halifax, Nova Scotia. Sequencing remains a major thrust, however. TIGR has now completed a stunning 25 genomes, including that of anthrax, reported online by Science on 10 May and appearing on p. 2028 of this issue. About 40 more microbes, two fungi, eight parasites, and five plants are in the works, although not all will be sequenced in toto. The group has also contributed extensively to the Arabidopsis and rice genome efforts and has become a central repository for Arabidopsis data. As TIGR celebrates its 10th anniversary this year, Fraser and colleagues are charting an even broader course. Sequencing, TIGR's mainstay, will continue to be important, says Fraser. That means TIGR must stay ahead of rivals and on top of rapidly changing sequencing technology. The same is true for bioinformatics. Microbial and plant genetics will remain key, but the institute's scientists might also move on to mammals. They're also talking about joining the proteomics bandwagon. Fraser expects that there's “going to be some synergy” between TIGR and Venter's latest ventures: two new institutes, one dedicated to ethics and the other to clean energy and the environment (Science, 3 May, p. 824). But the lines of responsibility will be clearly delineated. Although these new ventures will share TIGR's$170 million endowment, they will be independent entities with different, but related, foci.

Whereas Venter's new energy institute will explore how to use microbes to solve energy needs, for instance, TIGR will continue its basic research on these and other microbes. In this way, Fraser and Venter will complement each other, as in the past. “I've been together with Craig for more than 20 years,” she explains. “[The relationship] works because we are so different. We represent parts of each other's personalities that each of us wishes we had more of.” That winning combination has ensured them a place as pioneers in this very dynamic field of science.

13. BRAZILIAN ECOSYSTEMS: HISTORY, MANAGEMENT, AND PRESERVATION

# South American Landscapes: Ancient and Modern

1. Anne Simon Moffat*
1. Anne Simon Moffat is a freelance writer in Chicago.

CHICAGO— Brazil's tropical rainforests, biological diversity, and freshwater resources are among the world's largest. On 10 and 11 May, more than 100 environmental scientists, anthropologists, archaeologists, geologists, and government officials gathered at Chicago's Field Museum to discuss the status of these exceptional resources and how to preserve them. Many talks focused on what can be learned from historical land use and what the priorities for modern conservation should be.

## Ancients Transformed Landscapes

Prehistoric human societies have long been considered too insignificant to have made much of an impact on Earth's landscapes. Now, a small but growing number of researchers is disputing that conventional wisdom. At the meeting, Anna Roosevelt of Chicago's Field Museum and the University of Illinois, Chicago, and anthropologist William Balée of Tulane University in New Orleans, Louisiana, described new findings indicating that people who lived in the Amazon basin, hundreds or even thousands of years ago, left a bigger mark on the ecology of the area than expected. “Paleoecological data don't fit preconceptions,” Roosevelt says.

Her perspective is a minority viewpoint. But, says ethnobotanist David Lentz of the New York Botanical Garden in The Bronx, “the more we look, the more we find evidence that supports her. Roosevelt has done us a service.” Whether pre-Columbian societies had an impact on the Amazon basin is no longer a matter of dispute, he adds; rather, the issue is how big that impact was.

Roosevelt bases her conclusions on about 20 years of work in the Brazilian Amazon, where she studies the interaction of humans with their environment by looking at the chemistry and taxonomy of animal and plant remains, including bones, seeds, and shells, that people used daily. This research shows, she says, that late prehistoric humans, who lived from A.D. 400 to 1550, left garbage dumps that changed soil quality over large areas, primarily along the mainstream Amazon.

From the same period, Roosevelt also found evidence of areas with huge earth mounds 1 to 10 meters high, sometimes occurring over many square kilometers, where people lived. This suggests that people created small, open clearings along the Amazon, which have since reverted to forest. “What is viewed today as the great, natural, original Amazon forest is not totally so,” Roosevelt says.

Balée offered other evidence, gathered from recent inventories of the very diverse plant life in two Amazon regions, that suggests how prehistoric societies transformed the Amazon landscape and allowed the invasion and establishment of new species. For example, the use of fire, both intentional and accidental, and the expansion of agriculture, involving the cultivation of crops including manioc and trees such as palms and Brazil nut, all changed the land on a broad scale. This, in turn, expanded the number of light gaps in the dense, tropical forest canopy, allowing the growth of novel, sun-loving species, such as corn. Evidence of earthwork complexes in eastern Bolivia, Marajó Island in the mouth of the Amazon, and elsewhere also indicates the intensification of food production and an altered landscape, Balée says. All this happened before the arrival of Europeans in the mid-16th century.

Balée also suggests that although many ecologists claim that natural environments are the richest in diversity, his research indicates otherwise. He claims that biologic diversity in forests occupied by prehistoric societies, particularly where pre-Hispanic agriculture existed, is “probably higher today because of [earlier] human occupation, use, and management of the land” than it would be if such settlements had not occurred. The enhanced biodiversity comes from the import of plants and seeds from other areas and small-scale transformations of the landscape from a moist, shady forest to a sunny, dry savanna that allowed the invasion of new species.

Roosevelt stresses that understanding the historical use of land is key to planning modern conservation programs. For example, the sustainability of agriculture in Amazonia is often questioned, but prehistoric evidence shows that some areas that are well forested today were used for thousands of years for a type of agriculture in which farmers shifted their plantings from area to area, allowing previously planted areas to lie fallow to recover. Lowland forests in Amazonia, she says, appear resilient, able to withstand modest deforestation by small-scale human societies. Over time, despite various trials and tribulations brought on by nature and indigenous humans, they “just go chugging along,” she says.

For a decade or more, the Amazon rainforest has been a poster child for the environmental movement. Over about 20 years, this large, exotic, and diverse region of South America has lost perhaps as much as 18% of its area, almost 400,000 square kilometers, to human activities, such as farming, logging, and gold mining. But as several presentations at the meeting made clear, the Amazon rainforest, some 40% of which is now protected, is not the only threatened ecosystem in South America. Three others are in even worse shape, a situation that is drawing calls for new conservation efforts in those areas.

The regions causing most concern are the Atlantic forest, a sliver of dense green stretching 23 degrees of latitude south of the equator along the Brazilian coast; the Caatinga, a semiarid region in east-central Brazil; and the Cerrado, a large, open savanna that occupies much of central South America in northeast Paraguay, eastern Bolivia, and western and central Brazil. “These areas have been ravaged, but they are just as important as the Amazon,” says ornithologist Joel Cracraft of the American Museum of Natural History in New York City.

When the first European settlers began to colonize the land that would later become Brazil, the Atlantic Forest, known locally as the Mâta Atlantica, stretched over 1306 million square kilometers. But as landscape ecologist Joäo Paulo Capobianco of the Instituto Socioambiental in Säo Paulo reported, it has since been nearly wiped out by human activity, including the sprawl of Säo Paulo, one of the world's largest cities. One recent study that charted land use change between 1990 and 1995 found that more than 500,000 hectares of forest were destroyed in states that encompass about 90% of what is left of the Brazilian Mâta Atlantica. “This is a destruction proportionally three times greater than that recorded for Amazonia within the same period,” says Capobianco. Today, only 6% of the original forest is left, and it's still under threat.

Even so, the remaining Mâta Atlantica is well worth saving. It has, Capobianco says, one of the richest collections of ecosystems on the planet. They contain numerous endemic species found nowhere else, including 73 species of mammals, of which 21 species and subspecies are primates, 160 species of birds, and 165 species of amphibians.

Equally endangered is the Cerrado, the world's largest tropical savanna. It covers about 1.86 million square kilometers, mostly in Brazil, although parts extend into Paraguay and Bolivia. As recently as 50 years ago, the Cerrado was almost pristine. But José Maria Cardoso da Silva of Conservation International do Brasil in Belém told the meeting participants that intense colonization over the 40 years since construction of Brasília, Brazil's new national capital, changed at least 70% of the region. Most of the changes were due to soybean and rice farming that “has not followed the most basic principles of conservation,” says Cardoso da Silva. He notes, for example, that habitats are often fragmented, resulting in loss of biologic diversity, and farmers often do not protect topsoil from erosion. Still, the Cerrado's biodiversity remains impressive with 10,000 plant species, 120 reptile species, 161 mammal species, 837 bird species, and 150 amphibian species.

The dryland Caatinga has also been so strongly disturbed, Cardoso da Silva says, mainly by fire, timber exploration, and cattle ranching, that it is almost impossible to describe what original vegetation thrived there 500 years ago. Current estimates suggest that at least 60% of Caatinga has already been converted to agriculture and other types of land use.

For the past 10 to 15 years, Bolivia and Paraguay have made notable efforts to conserve their bits of the Cerrado with the establishment of preserves, but Brazil's efforts to conserve the Cerrado, Caatinga, and Mâta Atlantica date only to the last 5 years or so. Cardoso da Silva and ornithologist John Bates of the Field Museum support the establishment of more Brazilian reserves and of conservation corridors to connect them. They also suggest increasing productivity on lands already under agricultural cultivation to reduce pressure on areas in these three biomes still covered by natural vegetation—a doable project, given Brazil's current commitment to conservation.

14. ASTRONOMY

# Hubble Gets New IR Eyes

1. Govert Schilling*
1. Govert Schilling is an astronomy writer in Utrecht, the Netherlands.

ALBUQUERQUE, NEW MEXICO— The Hubble Space Telescope's infrared vision has been restored, and it's better than ever. At an American Astronomical Society meeting here,* excited astronomers presented the first images taken by Hubble's Near-Infrared Camera and Multi-Object Spectrometer (NICMOS) after it was outfitted with a new cooling system in March. “It's absolutely fantastic that we now have infrared eyes back on Hubble,” says NICMOS principal investigator Rodger Thompson of the University of Arizona, Tucson.

Observing in the infrared enables Hubble to peer into dusty star-forming regions and to study the extremely distant universe. NICMOS was installed in February 1997, but the solid nitrogen used to cool the sensitive infrared detectors to 62 kelvin was depleted just 2 years later, much faster than anticipated. Now, with a new refrigerator-like cooling system, the camera should be back in business indefinitely. NICMOS is between 30% and 40% more sensitive than it used to be, says astronomer Daniela Calzetti of the Space Telescope Science Institute in Baltimore, Maryland, because its detectors work more efficiently at the new system's 77-kelvin operating temperature. “This will enable us to look deeper, or to do the same science in less time,” Calzetti says.

• *200th meeting, 2–6 June.

15. AMERICAN GEOPHYSICAL UNION MEETING

# Of Mars Water, Old Cold, and Deep Talk

1. Richard A. Kerr

WASHINGTON, D.C.— The 3500 attendees of last month's meeting of the American Geophysical Union couldn't complain about the variety offered among the 2200 presentations. It included yet more water on Mars, garrulous volcanoes off Japan, and the depths of the Little Ice Age.

## Dry Water on Mars

Planetary scientists “following the water” in pursuit of martian life hit a major lode of the icy stuff late last month in data returned by the Mars Odyssey spacecraft. But all the speculation about an ocean of water and media talk of quenching the thirst of astronauts overlooked an even more intriguing water find on the desert planet. In its search for water, Odyssey measured the abundance of hydrogen, the H in H2O, plotting areas of particular hydrogen abundance in shades of blue. Between the ocean-blue high-latitude caps denoting ice-rich dirt are two huge splotches of light blue. They too can denote water: water that is bound to rock, or at least rock that has reacted with water. By recording the interaction of rock and water over the eons, this water-altered rock might reveal far more about the history of water on Mars and the planet's suitability for life than the surface ice ever will. At first blush, however, the newly recognized distribution of watery minerals across the martian landscape is proving frustratingly inscrutable.

Researchers accomplished the first global mapping of hydrogen in the surface of Mars by keeping track of the subatomic debris produced by the high-energy cosmic rays battering the planet. On impact, the cosmic rays send off a splattering of neutrons that ricochet like pinballs within the top meter or so of rock, dust, and dirt before escaping. Some of them are then picked up by Mars Odyssey's neutron spectrometer, which is operated by planetary scientist William Feldman of the Los Alamos National Laboratory in New Mexico and Odyssey team members. The energy of these scattered neutrons reflects how often they hit hydrogen before escaping the surface. And at latitudes within about 40° of Mars's equator—where the combination of relative warmth and thin atmosphere rules out liquid or frozen water—the only hydrogen they're likely to encounter would be in water chemically or physically bound to minerals or in minerals that have combined with oxygen and hydrogen from water to form hydroxides.

Researchers expected to see bound hydrogen in at least a few areas from Odyssey's orbital perch. In 1976 the two Viking landers caught a whiff of water from midlatitude soil they scooped up and heated. And orbiters have seen suggestive spectral signatures of spots of mineral-bound water. But, as Feldman and colleagues reported at the meeting and in a paper published online by Science on 30 May (www.sciencemag.org/cgi/content/abstract/1073541), Odyssey discovered a huge, roughly circular area of enhanced hydrogen that spans nearly one-quarter of the planet's circumference, a surprisingly broad region of hydrogen. The circle is centered near the prime meridian and stretches across the tropics and subtropics (center of map). A second, more irregular region lies on the opposite side of Mars (right and left edges of map), mostly south of the equator. Taking the soil at the Viking 1 lander site to be 1% water, the soil in these two light-blue regions would contain 3% to 4% water (if the hydrogen is in the form of water). That's far less than Odyssey found as ice at high latitudes, but that ice might have simply “snowed out” on falling dust particles in the recent—and dry—geological past, whereas the low-latitude hydrogen might record ancient water from an earlier, wetter Mars more hospitable to life.

The recognition of great swaths of water-altered minerals on Mars is both enticing and frustrating. Both of the preferred landing sites for the next two U.S. Mars rovers happen to fall in these regions—the so-called hematite site, where the roots of a hot-spring system might be exposed, and Gusev crater, which holds a former lakebed (Science, 10 May, p. 1006). Finding water-altered minerals would seem to make them even more attractive targets for landers. But planetary geologists are having trouble making sense of the patterns of hydrogen-rich minerals in the context of what they already know about Mars. The hydrogen signature extends willy-nilly beyond the lander targets across any number of geologic terrains. The regions are “very hard to reconcile with what we know about geology or topography,” says planetary scientist James Bell of Cornell University. They don't fit the distribution of particular rock types, rock abundance, dust, or even atmospheric water vapor, notes planetary scientist Bruce Jakosky of the University of Colorado, Boulder.

The U.S. landers and a European lander might help sort out the nature of the water-altered rock, as could element-and mineral-identifying spectrometers on Odyssey. But the near-infrared mapping spectrometer aboard the 2005 Mars Reconnaissance Orbiter should provide “the definitive data set” identifying the minerals involved, says Bell, and thus help unravel the history of water on Mars.

## Chatty Volcanoes

In recent years, seismologists have begun eavesdropping on conversations among faults. Faults can tell each other when to fail, communicating across hundreds of kilometers through the stress changes earthquakes induce. If faults can talk, why not volcanoes? A group of geophysicists reported at the meeting that they seem to have overheard one side of a conversation between two Japanese volcanoes as one of them prepared to erupt. The discovery offers the prospect of understanding how groups of volcanoes work and, as with faults and their earthquakes, forecasting future behavior.

Geophysicists didn't mean to eavesdrop on volcanoes. The Japan Meteorological Agency (JMA), the organization responsible for earthquake monitoring in Japan, had deployed strainmeters in deep boreholes around the southeast portion of the main Japanese island of Honshu to watch for the great quake expected to strike offshore. But one of the strainmeters, which measure the subtle squeezing and stretching of the crust, was installed in 1981 on the small island of Izu-Oshima, the northernmost of a string of volcanic islands. In 1983, Miyake-jima, a volcano in the chain 70 kilometers south of Izu-Oshima, erupted.

The erupting volcano had no strainmeter, but the one on Izu-Oshima seemed to know what was going on, according to geophysicists Alan Linde and Selwyn Sacks of the Carnegie Institution of Washington's Department of Terrestrial Magnetism in Washington, D.C., and their colleague Osamu Kamigaichi of JMA in Otemachi. On three different time scales, the way the crust was deforming at the distant strainmeter changed at the time of the eruption, they reported. The rate at which the crust was being slowly squeezed had been constant for 2 years but dropped dramatically 2 days before the eruption. The extent to which the crust yields to the daily tidal kneading of the moon and sun had been increasing rapidly but slowed abruptly at the eruption. And the frequency of as-yet-mysterious squeezings lasting an hour or two jumped from one or two a year to three or four a month 6 months before the eruption and stopped entirely just before it.

The striking coincidence of deformation changes and distant eruption means that “there's probably some important Earth physics happening” between the two volcanoes, says Linde. Their best guess is that, perhaps 40 or 50 kilometers down, a deep-seated channel of magma mush connects the shallow magma reservoirs that feed the two volcanoes. Seismic probing has found such a deep channel beneath the similar volcanic chain of northern Honshu, the group notes. Pressure variations beneath one volcano could then travel to the other in a matter of days. Some of those communiqués, such as the mysterious abrupt squeezings, might herald a coming eruption, notes Linde. Deep connections might even explain the habit of some pairs of volcanoes to erupt one right after the other, says Sacks.

Volcano communication “is a good idea that needs more work,” says volcanologist Michael Ryan of the U.S. Geological Survey in Reston, Virginia. “I've done theoretical models of volcano interactions before, and you don't get much ‘news’ more than 10 or 15 kilometers down the line” from major deformations. Linde and his colleagues should show how 70-kilometer communication could possibly work, says Ryan, before he's convinced.

## Big Chill From Sun and Volcano

Anything as fiery as the sun or an erupting volcano would seem an unlikely agent of global cooling. But researchers are accumulating evidence that the sun and volcanoes ganged up on Earth's climate system 300 years ago to give Europeans, at least, a major chill. The sun became exceptionally quiet in the second half of the 1600s just as an unusual number of volcanoes popped off, a combination that a growing number of climate models suggests could have triggered intensely cold winters in Europe. The trick might have been the ability of a dimming sun to shift atmospheric circulation patterns over the North Atlantic.

The apparent progress in understanding at least a few decades of paleoclimate comes after climate researchers sharply narrowed their focus. Paleoclimatologists had long pondered the 3-century-long Little Ice Age, a time of frequent cold spells that ended around 1850 (Science, 25 June 1999, p. 2069), but its ultimate cause remained unclear. At the meeting, a morning session was devoted to a single cold spike, particularly well documented in Europe, that spanned from 1670 to 1710. That was just after sunspots nearly disappeared from the face of the sun for more than half a century. The event, called the Maunder sunspot minimum, is unique in the past millennium. By studying shorter term behavior of the sun and sunlike stars and indirect records of solar activity, scientists have estimated that during the Maunder minimum the sun grew a few tenths of a percent dimmer than it is today.

The coincidence of an extreme climate event and a possible solar dimming—a driver of climate change—proved an irresistible target for climate modelers. At the meeting, modeler Drew Shindell of NASA's Goddard Institute for Space Studies (GISS) in New York City and his colleagues reported that the GISS general circulation model, the sort of complex climate model used to predict greenhouse warming, cooled the globe as a whole a modest 0.3° to 0.4°C in response to the solar dimming of the Maunder minimum. The model's cooling agrees with the sketchy temperature record from historical accounts and paleoclimate proxies such as tree rings. But in the model, as in the real world, the extreme cold of the late Maunder minimum—1° to 2°C of cooling—was localized, primarily over Northern Hemisphere continents in winter.

The GISS results offer resolution of a paleoclimate quandary: how such a modest dimming of the sun could drive such a strong response. In the GISS model, the dimming doesn't cool the globe as much as redistribute heat around the globe. It weakens the westerly winds of the Arctic Oscillation (also called the North Atlantic Oscillation) that blow across the North Atlantic. The weaker westerlies pick up less heat from the Atlantic and thus carry less heat into Europe to moderate its winters. So a dimmer sun could mean more ice skating, at least on the canals of Europe. Rising greenhouse gases might be driving the Arctic Oscillation the other way and warming northern continents (Science, 9 April 1999, p. 241).

Shindell also tested the effect of the volcanic eruptions of the time as recorded in ice cores. The debris spewed into the stratosphere by volcanoes cools the surface by reflecting some sunlight back into space. “The volcanic signals help improve the agreement” of model and climate record, said Shindell. For example, in the model, solar dimming alone made Iceland warm, although some historical records say it experienced disastrous cold. Adding volcanoes to solar dimming chilled Iceland, as observed.

Other modelers presenting in the session reached similar conclusions. Both Ulrich Cubasch of the Max Planck Institute for Meteorology in Hamburg, Germany, running a different state-of-the-art greenhouse model, and Eva Bauer and her colleagues at the Potsdam Institute for Climate Impact Research in Germany, running a model of intermediate complexity, reported that both solar variations and volcanoes are needed to best simulate the climate late in the Maunder minimum. So a deep European chill might require a double whammy, at least.