# News this Week

Science  24 Aug 2001:
Vol. 293, Issue 5534, pp. 1410
1. THE GULF WAR'S AFTERMATH

# Kuwait Unveils Plan to Treat Festering Desert Wound

1. Ben Shouse*
1. Ben Shouse is an intern in Science's Cambridge, U.K., office.

LONDON— Ten years after the Gulf War ended, Kuwait's deserts are still drenched in crude oil, most of it spilled as Iraqi invaders beat a hasty retreat. Now the country is about to embark on a belated $1 billion effort to tackle the ecological calamity in one of the biggest environmental remediation projects ever attempted. “It's a living laboratory of a type mankind has never seen before,” says Paul Kostecki of the University of Massachusetts, Amherst. Despite its considerable wealth, Kuwait has made little headway in cleaning up its oil-contaminated deserts. An estimated 250 million gallons of oil—more than 20 times the amount spilled by the Exxon Valdez oil tanker off Alaska in 1989—despoiled one-third of the land. Kuwaiti scientists claim that wildlife took a heavy hit, particularly in the National Park of Kuwait, where the national flower, the arfaj (Rhanterum epapposum), was wiped out; it's now being replanted. Speaking here last week at the International Congress on Petroleum Contaminated Soils, Sediments and Water, Kuwait's oil minister, Adel Al-Sabeeh, asserted that his nation's oil industry has committed more than$630 million on projects related to health, safety, and the environment. However, Kuwaiti researchers counter that cleanup efforts have so far dealt only with contamination that impedes oil extraction. All told, they insist, only $13 million has been spent in the past decade to examine the true scope of the oil's harm. A delay in sopping up the crude was inevitable: Kuwait spent the first 6 months just putting out oil fires set by retreating Iraqi forces. Some also see the psychological factors of an aggrieved nation at play. “If somebody breaks into your car, you wait for them to pay,” says Andy Kwarteng of the Kuwait Institute for Scientific Research (KISR) in Kuwait City, which will oversee the environmental reparations budget. But the wait is over. In June, the United Nations Compensation Committee awarded Kuwait$108.9 million in reparations from U.N.-controlled Iraqi oil sales to be spent on addressing the environmental fallout from the Gulf War (Science, 29 June, p. 2411). First up is a 5-year project to catalog the environmental ills, followed by a remediation estimated to cost more than $1 billion. Any remediation efforts must be tailored to Kuwait's desert environment and the unprecedented extent of the contamination. Although the Valdez spill was a disaster in its own right, waves helped break up the slick to allow petroleum-eating bacteria to consume tons of oil, thus partly mitigating the harm to Alaska's coastal ecosystem. In soil, by contrast, adhesion and weathering make crude oil more stubborn, while a desert's dryness tends to deter natural degradation. Short on funds, the KISR so far has carried out only two pilot remediation projects. In one, Nader Al-Awadi's team from KISR, working with Japan's Petroleum Energy Center, showed how to remove 94% of hydrocarbons from soil underneath lakes of oil now covering 49 km2 of Kuwait. It is not a delicate process: The soil is excavated and washed with kerosene, piled up, and then pumped with air and water to nourish oil-eating microbes. If this process were used to treat all 70 million cubic meters of soil affected by oil lakes, it would cost$1.3 billion, says Al-Awadi. And that's leaving out contaminants such as soot and hardened tar mats, which cover a wider area but are deemed less serious ecological threats.

One novel project stems from the high concentration of petroleum in some of the spills. Researchers have proposed using the oily sand to pave roughly 5000 kilometers' worth of roads. In other words, when life gives you asphalt, make a highway.

Kuwait's bioremediation windfall “could provide an incredible amount of research,” says Kostecki, executive director of the U.S.-based Association for Environmental Health and Sciences, which sponsored the London conference. And although Kuwait has skimped so far, outside experts say the country's leadership has experienced a change of heart. “They don't really care about the cost,” insists Farouk El-Baz, director of the Center for Remote Sensing at Boston University. “If they can find a way, they will clean it up.”

2. COSMOLOGY

# Changing Constants Cause Controversy

1. Charles Seife

The times, they are a-changin', and so are the fundamental constants of physics, an international group of physicists reports. After analyzing light from distant quasars, the team has concluded that the fine-structure constant, which is related to the speed of light, has shifted over time. The claim is extremely controversial, but scientists are taking it seriously—if skeptically.

The fine-structure constant is an amalgamation of the speed of light, the charge of the electron, and the quantum-mechanical number known as Planck's constant. Combined, these values give a measure of the inherent strength of electromagnetic interactions, such as those that bind an electron to an atom. Like the speed of light, it is thought to be immutable: approximately 1/137. But in the 27 August edition of Physical Review Letters, a team of astronomers and physicists presents evidence that the fine-structure constant was different in the early universe. “One thing is clear. If it's correct, it's fantastically important,” says John Bahcall, an astrophysicist at the Institute for Advanced Study in Princeton, New Jersey.

The claim is based on observations of light from 72 distant quasars that has passed through light-absorbing clouds en route to Earth. Ions in those clouds, such as different valences of magnesium, iron, nickel, and zinc, each absorb certain narrow wavelengths of light, etching dark lines in the quasar's spectrum. Like a cosmic fingerprint, the pattern of the absorption bars tells scientists which ions reside in the clouds. And because an atom absorbs light due to the electromagnetic interaction between its nucleus and its electrons, the fine-structure constant affects where the bars appear. “The physics is pretty straightforward,” says team member Jason Prochaska, an astronomer at the Observatories of the Carnegie Institution of Washington in Pasadena, California.

When Prochaska and other physicists from Australia, the United States, and England collected data from the distant quasars and analyzed the patterns of bars, they noticed that the spacing of the bars wasn't quite right. The pattern seemed to indicate that the fine-structure constant was about 0.001% smaller when the light was absorbed billions of years ago than it is now. In other words, the fine-structure constant has been increasing over time.

But other physicists are skeptical. “There's more ways to go wrong than to go right,” says Bahcall. “This measurement is so sensitive to systematic uncertainties that I'm worried that one of them got them.”

Lennox Cowie of the University of Hawaii, Manoa, has an alternative explanation for the strange spacing of the absorption lines. “Generally, it's likely to be things like different ions having slightly different velocities, as they reside at different points in space,” he says. Because of the Doppler effect, the ions' different velocities shift the relative positions of the absorption lines. “In my own mind, that's the probable explanation,” says Cowie.

But the team says it has already accounted for that effect. “I will be very surprised if this is the explanation,” says team member John Webb, an astrophysicist at the University of New South Wales in Sydney, Australia.

Prochaska says he has unpublished data that strengthen the case for an inconstant constant, although he suspects they won't sway all critics. “Someone else needs to do it with a different telescope and a different instrument. That would be the proof of the pudding,” he says. Until then, he agrees that cosmic change remains in doubt: “I wouldn't bet my life on it right today.”

3. SCIENTIFIC MISCONDUCT

# Wellcome Rules Widen the Net

1. Robert Koenig

The U.K.'s biggest biomedical charity has filled a void by proposing its own guidelines and procedures for handling allegations of scientific misconduct. While generally winning high marks, the draft rules from the Wellcome Trust are likely to spark controversy by broadening the definition of misconduct beyond the U.S. government's standard and by offering relatively little protection to whistleblowers.

The draft guidelines, circulated late last month, would apply only to institutions receiving Wellcome funds. Even so, they could be a tonic for a scientific community that has been left to police itself without widely accepted definitions of both misconduct and good scientific practices. “Everyone in the U.K. seems to agree that something needs to be done, but no one seemed to be willing to take action,” says Wellcome's Robert Terry.

“This is an extraordinarily positive development,” says ethics expert C. Kristina Gunsalus, associate provost of the University of Illinois, Urbana-Champaign. “The most important thing is that someone in the U.K. has finally taken the initiative.”

The Wellcome document gives a fuller—and perhaps more contentious—definition of misconduct than parallel regulations governing U.S. federal funding developed by the Office of Science and Technology Policy (OSTP) (see table). While approving Wellcome's overall approach, some experts quibble with the details. Although both the trust's and OSTP's definitions include plagiarism, fabrication, and falsification of data, the Wellcome language “moves away from the clarity of the U.S. definition by reintroducing deviation from accepted practices as misconduct rather than as the basis for finding misconduct,” argues Fred Grinnell, director of the Ethics in Science and Medicine program at the University of Texas Southwestern Medical Center in Dallas. Adds ethics expert Howard K. Schachman, a biochemist at the University of California, Berkeley, “The definition of scientific misconduct presented in the Wellcome Trust document contains words and ideas that I think should be eliminated.”

According to Chris B. Pascal, director of the U.S. Office of Research Integrity, Wellcome's definition of what constitutes misconduct— including “deliberate, dangerous or negligent deviations from accepted practices”—“is considerably broader” than the OSTP's definition. “In theory, it would be easier to show misconduct under Wellcome's definition” than under the U.S. definition, says Pascal, who notes that the OSTP rules are likely to go into effect at several—but not all—U.S. departments by the end of this year.

One shortcoming in the Wellcome guidelines, contends microbial geneticist Herbert Arst of Imperial College in London, is their lack of strong provisions for protecting whistleblowers and ensuring that universities don't conduct cursory “whitewash inquiries” of misconduct allegations. Terry defends Wellcome's whistleblower section, explaining that its wording is limited partly by the U.K.'s strict libel laws, which make it easier for accused parties to win a defamation case.

Such concerns could be addressed in final guidelines set to go into effect in the fall of 2002. In addition to providing comments on the existing draft, due next month, the trust has asked organizations to describe how they deal with misconduct allegations. Oxford University's 2-year-old integrity code, for example, offers a relatively broad definition of misconduct and a set of procedures to pursue allegations. According to Wellcome, sanctions against a researcher found guilty of misconduct could range from a letter of reprimand to barring the individual from receiving trust funds “for a given period.”

The Wellcome guidelines could trigger a rush among U.K. research outfits to follow suit, predicts Drummond Rennie, deputy editor of The Journal of the American Medical Association. Nor are the guidelines intended solely for biomedical scientists: “We're hoping that this also can become the template for guidelines in other fields of science,” says Terry. In addition, Britain's research councils are weighing whether to require universities to adhere to “good practice” as outlined by the Medical Research Council. Although Wellcome does not have the power of a government agency, it does wield a sword of Damocles over universities and research institutes hesitant to enforce its planned rules: the threat of making them ineligible for Wellcome grants.

4. MATERIALS SCIENCE

# Silicon Lights the Way to Faster Data Flow

1. Robert F. Service

Computer engineers can design souped-up chips capable of performing billions of calculations per second. But their wizardry will be in vain unless they can also speed the flow of information between chips and other computer components. One way to do that is to replace today's sluggish metal wires with higher speed optical connections, using special semiconductors to convert electrical signals to a staccato of light pulses. Unfortunately, the best light-emitting semiconductors, such as gallium arsenide (GaAs), are hard to integrate with silicon, and the ideal material for the job—silicon itself—has been a poor light emitter.

Now Australian researchers have taken a big step toward making silicon shine. In this week's issue of Nature, physicist Martin Green and colleagues at the University of New South Wales in Sydney report a 100-fold boost in the efficiency of silicon-based light-emitting diodes (LEDs) using a trick for making solar cells. The devices still aren't as bright as ones made of GaAs. But there appears to be plenty of room for improvement. “If it can interact with transistors and memory, it would probably be really important,” says Daniel Radack, who oversees advanced computing issues for the Defense Advanced Research Projects Agency in Arlington, Virginia.

Silicon, it turns out, does only a mediocre job of both absorbing and emitting light. For solar cells, which absorb light and convert it to electricity, the result is that much of the light that hits a cell passes right through the material. In recent years, Green and his colleagues have found that texturing the top and bottom surfaces of the cells causes light to bounce around inside the cell so it can be absorbed. The best light-absorbing semiconductors are also the best light emitters, Green says, giving him the idea that texturing silicon could improve the efficiency of silicon LEDs as well.

An LED works like a solar cell in reverse. Negatively charged electrons and positively charged “holes” are injected into the device. When they collide, they give off photons—in this case of infrared light—with wavelengths similar to the ones used in optical communications.

Silicon is actually pretty good at getting these charges to combine: About 10% of the injected charges produce photons. The problem is that usually only about 0.01% to 0.1% of the photons ever get out; the rest just create unwanted heat. To improve matters, Green's team created an array of pyramid-shaped wells on the silicon's top surface and a mirrorlike flat bottom surface. The combination helped newly created photons bounce around inside the device until they could find an escape route, raising the efficiency to more than 1%.

The next step for the team is to convert the silicon LED's steady stream of light into a series of pulses that can encode information, by connecting the LED to a device called a modulator. Then both devices will be placed directly onto a silicon computer chip.

5. ECOLOGY

# African Elephant Species Splits in Two

1. Gretchen Vogel

As the largest land mammal, elephants should be hard to miss. But scientists have apparently overlooked an entire species. On page 1473, a team of geneticists and elephant experts describe new molecular evidence showing that forest-and savanna-dwelling elephants, currently lumped together in a single species called Loxodonta africana, each merits its own species name.

For more than 100 years, scientists have argued about the distinctiveness of forest elephants. The shy creatures are difficult to spot in their thick forest habitat, and only one is in captivity in the Paris Zoo. But those fortunate enough to have seen both them and their better known savanna-dwelling cousins note that the difference is striking. Forest elephants are not only smaller, but they also have straighter, longer tusks and round, as opposed to pointed, ears. “If you see a forest elephant for the first time, you think, ‘Wow, what is that?,’” says team member Nicholas Georgiadis, a biologist at the Mpala Research Center in Nanyuki, Kenya.

Despite the differences between savanna and forest elephants, most biologists had assumed that the two populations readily mix on the edges of forests. At best, forest elephants were designated as a subspecies, Loxodonta africana cyclotis. But when Georgiadis and his collaborators analyzed the DNA of the elephants, the results indicated that the two populations are as genetically distinct as lions are from tigers. Indeed, the researchers propose two separate species names: Loxodonta africana for the savanna elephants and Loxodonta cyclotis for the forest dwellers. “The morphological evidence has been very, very strong,” says conservation biologist Samuel Wasser of the University of Washington, Seattle. “When you see the genetic data, it seems almost a no-brainer.”

The team bases its claim on data from an extensive collection of tissue samples from 195 animals in 21 different elephant populations. Georgiadis spent 8 years collecting the samples, shooting needlelike darts into free-ranging elephants. The darts collected a plug of skin and then fell to the ground, enabling Georgiadis to retrieve them after the startled elephant ran away. The project was originally designed to collect genetic signatures so that ivory samples could be traced to their elephant populations—a goal other geneticists are pursuing. A preliminary analysis of the mitochondrial genes suggested significant differences between forest and savanna dwellers (Science, 7 March 1997, p. 1418), a finding that piqued Georgiadis's interest, but a more robust test with nuclear genes was needed to cinch the case, he says.

To pursue the question, Georgiadis teamed up with researchers at the National Cancer Institute (NCI) in Frederick, Maryland, to measure the genetic variation between the populations. Alfred Roca, a postdoc at NCI, with geneticists Stephen O'Brien and Jill Pecon-Slattery, sequenced portions of four nuclear genes, a total of 1732 nucleotides, from each of the samples. The researchers focused on the noncoding intron regions of the genes, which are not subject to natural selection; this makes them more reliable indicators of the random genetic changes that occur over time.

The team pegged the genetic distance between forest and savanna samples as more than half as large as the distance between Asian and African elephants—long recognized as distinct genera. Only one of the populations showed the type of genetic mixing that could come from interbreeding, and that apparently happened several generations ago. To O'Brien, that means that crossbreeding between the two populations “does occur once in a while, but not very often.”

The new genetic evidence has implications for conservation, says Georgiadis. Instead of assuming that 500,000 elephants exist in Africa, “there are many fewer than that of each kind, and they're both much more endangered than we presumed,” he says. Researchers estimate that up to one-third of African elephants are forest dwellers.

Ivory from forest elephants is especially prized for its hardness and sometimes pinkish hue. Wasser cautions that conservation organizations must be alert: The current international regulations list only Loxodonta africana as protected. If the law is not changed quickly to reflect a new species name, an inadvertent loophole might leave the vulnerable forest elephant even more at risk.

6. VIROLOGY

# Finally, a Handle on the Hantaviruses

1. Martin Enserink

A group of U.S. Army virologists has found by accident what researchers had been seeking for decades: an animal model to study hantaviruses, a fearsome group of rodent-borne pathogens that cause disease and death across the globe. In a paper accepted by the journal Virology, they report that Syrian hamsters get sick and die when injected with a hantavirus from South America—and that the animals' disease looks strikingly like hantavirus pulmonary syndrome (HPS), one lethal manifestation of the infection in humans.

The report, from a group at the U.S. Army Medical Research Institute of Infectious Diseases (USAMRIID) in Fort Detrick, Maryland, means that researchers at last have a way to study how one member of the family sickens and kills—and how the disease can be stopped. “I'm impressed,” says Heinz Feldmann, a virologist at the Canadian Science Centre for Human and Animal Health in Winnipeg. “This will definitely speed up vaccine and drug development.” Virologist Stuart Nichol of the U.S. Centers for Disease Control and Prevention in Atlanta hails the study as “a major breakthrough for the field.”

The hantavirus family grabbed world headlines in 1993 when an outbreak occurred in the Four Corners area of the southwestern United States. The culprit, now called Sin Nombre virus, is one of several hantaviruses that cause HPS throughout North and South America. No specific antiviral treatments exist, and usually between one-third and one-half of the victims die. In Europe and Asia, a quartet of different hantaviruses causes tens of thousands of cases yearly of a disease called hemorrhagic fever with renal syndrome (HFRS); mortality rates range from 0.1% to 15%. Hantaan virus, the first of that bunch to be discovered, caused more than 3000 cases of HFRS among United Nations troops during the Korean War between 1951 and 1954; the U.S. Army has had a keen interest in developing hantavirus vaccines ever since.

But Army researchers and others have been handicapped by the lack of animal models. No matter how many animals they injected with various hantaviruses, they could not produce anything resembling the ravages of either severe HFRS or HPS. For example, two teams have injected monkeys with a European hantavirus called Puumala, but it caused symptoms that would likely be too subtle for a vaccine trial, says USAMRIID team leader Jay Hooper.

Confronted with these obstacles, the team had to make do with another strategy. Instead of testing whether a vaccine protects against disease, they tested whether it could prevent infection altogether—a much stricter test because it flunks vaccines that let the virus enter the body and replicate but prevent illness.

The team recently started looking into vaccines for Sin Nombre and the Andes virus from South America, which together cause the great majority of HPS cases in the Americas. A key initial step was determining how much virus is needed to infect an unvaccinated animal. When Hooper injected hamsters with Sin Nombre, they became infected but stayed healthy. But after he injected adult Syrian hamsters with Andes virus, something unusual happened. One by one, the animals developed difficulty breathing, and most died within days—just as fast as human victims. Further studies of the hamsters by USAMRIID pathologist Tom Larsen revealed that after an incubation period of about 11 days, the microscopic blood vessels in the animals' lungs became permeable and their lungs and chest cavity rapidly filled with fluid, essentially causing them to drown. All these symptoms closely resemble HPS in humans, says Hooper.

Not only do researchers now have a better way to test vaccines and drugs, but they can also study the details of hantavirus pathogenesis. Several studies have suggested that hantaviruses aren't all that lethal to individual cells, says virologist Clarence Peters of the University of Texas Medical Branch in Galveston. Instead, the human immune system's response may wreak the real havoc. Researchers can now test that theory by blocking or stimulating the suspect immune messengers—studies that could provide new drug leads.

The model does have drawbacks, however. Because it works only with Andes virus (the team is still not sure why), it may not tell researchers much about the HFRS-causing hantaviruses. And the hamster, unlike the mouse, isn't a common lab animal, so researchers lack both an intimate knowledge of its biology and a wealth of reagents to study it. Even so, the dearth of models has been so frustrating that Peters predicts others will jump on the findings.

7. ANIMAL CARE

# Coulston Loses NIH Tie, Faces Hard Times

1. Josh Gewolb

A major U.S. primate facility has lost its permit to house and experiment on federally owned chimpanzees, raising questions about its viability.

In June the National Institutes of Health (NIH) ended its funding of the Coulston Foundation of Alamogordo, New Mexico, after finding a new caretaker, Charles River Laboratories, for 300 chimps housed there (Science, 10 September 1999, p. 1649). Later that month NIH officials let lapse a document, called an Animal Welfare Assurance, that allows Coulston to carry out federally funded research with animals.

The foundation was formed in 1993 when businessman Frederick Coulston united his several primate care ventures and created the nation's largest chimpanzee facility. At its peak, Coulston cared for more than 600 chimpanzees with a staff of 120.

Foundation spokesperson Don McKinney declined to comment on the number of chimps currently housed at the facility, the foundation's financial condition, or the size of its workforce. But available records suggest that the loss of federal funding will be a heavy blow. In the 1999–2000 fiscal year, 63% of the foundation's $4.1 million in annual revenues came from the government, according to tax returns obtained by animal-rights groups. And its ability to solicit contracts with private U.S. companies is restricted by a 1999 decision by the Food and Drug Administration that the center does not comply with good laboratory practice regulations, to which all advanced animal trials must adhere. McKinney says the foundation has active private contracts but that all details are proprietary. According to tax records, the foundation's private contracts declined by 35% from the 1998 to 2000 fiscal years. Ronald Couch, former president of the foundation, says that investigations into possible animal welfare violations hurt the foundation's ability to attract private clients during his brief tenure in 2000. Coulston still faces an investigation by the U.S. Department of Agriculture (USDA) over the deaths of two chimps in 1999 and 2000. If history is any guide, the foundation's future may depend on the size of Frederick Coulston's personal cash reserves. The 86-year-old Coulston has reported giving the foundation more than$7.5 million, according to NIH records and the foundation's tax returns.

# Scientists Want Tougher Endangered Species Law

1. Jay Withgott*
1. Jay Withgott writes from San Francisco.

Canadian biologists are trying to toughen proposed legislation designed to protect endangered species in Canada. Their stance puts them in the awkward position of resisting government efforts, almost a decade in the making, to pass the nation's first federal law on the issue.

After changes in government derailed two previous attempts (Science, 13 December 1996, p. 1827), the chances of passage of the proposed Species at Risk Act (SARA) this fall appear good. But many scientists believe that it doesn't do enough to protect species' habitat, and they want a scientific panel, not politicians, to have the final say in deciding which species are listed.

SARA differs from the equivalent U.S. Endangered Species Act in seeking first to work cooperatively with landowners and industry, offering incentives and financial compensation; enforcement of yet-to-be written regulations would be used only as a last resort. “We do not want to hamstring our own efforts to recover species with a confrontational and immediately prohibitive approach,” Environment Minister David Anderson told Science. “We want to promote voluntary action, individual responsibility, and cooperative, community-based solutions.” The goal, Anderson adds, is to produce “legislation that is effective on the ground, not just ‘strong’ on paper.”

But a number of scientists say that this particular carrot-and-stick approach is too much carrot and not enough stick. The bill provides no mandatory protection for species' habitats, they say, safeguarding “residences” such as dens or nest sites but leaving the designation of habitat and enforcement mechanisms open to influence from local and regional officials, landowners, and industry. “Anyone with Ecology 101 knows that without habitat, it is impossible for species to survive,” says ecologist David Schindler of the University of Alberta, one of the organizers of a letter being drafted to Prime Minister Jean Chretien (http://www.scientists4species.org/) that lays out their arguments.

Scientists also find fault with the proposed listing process. A panel of experts, the Committee on the Status of Endangered Wildlife in Canada (COSEWIC), has long maintained a list of species recommended for protection under the act (see graph), and scientists would like to see COSEWIC have the legal authority for listing species. But the bill leaves the decision on listing to Cabinet ministers.

Scientists also worry that the federal government will defer to provincial governments in enforcing the act. “Appeasing the provinces seems to be in vogue in Canada, so when the provincial bullies snarl, federal ministers turn and run,” Schindler says. He and others say that federal control is key for the 70% of threatened and endangered species, such as grizzly bears, wolves, and migratory birds, whose ranges extend into the United States. “It's really embarrassing that both Mexico and the United States have stronger laws,” says environmental lawyer Kate Smallwood of Sierra Legal Defense Fund, who is working with the scientists.

Not all of Canada's scientists oppose the bill, however, and many fear that too much criticism from biologists may derail it. “Anderson has gone an awful long way for this and is really doing a lot to make it happen,” says Fred Cooke, an ornithologist at Simon Fraser University in Burnaby, British Columbia. Canada's collaboration with industry, he adds, seems to be working better than the U.S. approach in helping such species as the marbled murrelet.

In contrast, a prominent legislator suggests that scientists might want to make an even bigger fuss. Charles Caccia, a longtime proponent of strong endangered species legislation and chair of Parliament's Standing Committee on Environment and Sustainable Development, says that some members of his committee “wish that scientists would be more forthcoming, more politically explicit” in explaining what changes are needed. Caccia's committee will consider amendments to the bill next month before forwarding it to the House of Commons.

9. NEUROSCIENCE

# New Brain Institute Struggles for Traction

1. Andrew Lawler

It has money, a Nobel laureate, and MIT's prestige, but a new neuroscience institute launched last year is off to a slow start

CAMBRIDGE, MASSACHUSETTS— Can the brain ever understand the brain, much less the mind? Researchers are betting they can, thanks to new tools, large sums of private and public money, and an influx of scientists into neuroscience. Universities around the world are busy launching brain-research programs, with an eye toward winning fame and fortune in the rapidly growing field.

One of the most highly touted efforts is at the Massachusetts Institute of Technology (MIT), which announced in early 2000 a determined initiative to seize the preeminent role in neuroscience. Armed with one of the largest academic gifts in history and a formidable faculty, the McGovern Institute for Brain Research aims to unravel the mysteries of behavior, communication, and perception. But a year and a half into the highly publicized start, the field's researchers say the effort is stalled. The new organization, led by Nobel laureate and MIT molecular biologist Phil Sharp, has yet to define its research direction, attract a senior outside neuroscientist, or find its own niche in a university notorious for its academic rivalries.

Scientists inside and outside the McGovern Institute acknowledge that it is too soon to make definitive judgments about the endeavor's success. But the troubles highlight the difficulties—despite money, smarts, and power—of creating a dynamic research enterprise at the cutting edge, particularly in this unruly interdisciplinary field where competition to hire the best and the brightest is fierce. Other universities, including Harvard just down the street (see sidebar, p. 1419), are aggressively organizing their own ambitious programs.

The competition stems in large part from new tools for molecular biology and medical imaging, which are enabling researchers to begin approaching the enormous complexities of the brain and behavior in a comprehensive way. The effort spans a wide spectrum. At one end are molecular and cellular biology, which examine single neurons. At the other end are cognitive studies that examine learning, memory, emotions, and other higher order brain functions. In between is systems neuroscience, which aims to understand the behavior of aggregates of neurons. But this interdisciplinary reality challenges a university system subdivided into disciplines and subdisciplines.

MIT hopes to pull together a variety of researchers in a single campus complex, with Sharp's institute at the center. But the bureaucratic, financial, and research challenges are formidable, and outsiders are watching closely. “There is no model on how to do this,” says William Newsome, a neuroscientist at Stanford University, “and that's why the McGovern Institute is so important.”

Powering the new institute is a $350 million gift received in March 2000—the largest in history at that time—from International Data Group founder Pat McGovern and his wife, entrepreneur Lore Harp McGovern. Pat McGovern, an MIT alumnus, says that the new institute's mission is to create “a world-class center for the study of how the brain affects human behavior, communications, and perception.” The gift will be dispensed gradually, at$5 million a year, with an endowment of approximately $250 million in the 20th year. MIT will roughly match those funds, primarily by constructing a massive 13,000-square-meter building on the northern edge of campus. The new complex will house not just the institute's 16 researchers (including 10 hired from outside MIT) and its total staff of 300 people, but the many other pieces of MIT's varied neuroscience and cognitive research effort. ## Inside job The leader of the McGovern Institute, a consummate MIT insider, is the former head of MIT's biology department and cancer center (see sidebar, p. 1420). But although Sharp has outstanding credentials in his field of molecular biology (including the 1993 Nobel Prize in physiology or medicine), he has no track record in neuroscience or cognitive studies. Thus neuroscientists greeted Sharp's appointment with surprise, to say the least, particularly because McGovern and MIT president Charles Vest emphasized that the new institute would study the higher end brain functions. “It was seen as an exceedingly odd choice,” says Charles Gallistel, a neuropsychologist at Rutgers University in Piscataway, New Jersey, and a member of an advisory panel for MIT's brain and cognitive sciences department. Adds Charles Stevens, a researcher at the Salk Institute for Biological Studies in La Jolla, California: “Most neurobiologists feel Phil isn't the optimal person to set the research direction for the new institute.” Sharp says he has no illusions “that I understand the deepest aspects of neuroscience.” He strongly courted Stanford's Newsome to serve as his right-hand man in navigating the complex terrain of the field. Newsome is a highly respected pioneer in using imaging technologies to investigate brain activity in awake monkeys—research that provides vital clues to the connections between brain activity and behavior. Hiring him might have muted the criticism. “Bill would have been a fantastic hire,” says Gallistel, “because he is one of the few bridges between neuroscience and cognition.” But Newsome ultimately declined the offer because of his wife's career in California, and Sharp, noting that there are “a lot of good people right here at MIT,” now says he has no immediate plans to hire a senior outside researcher. In recent months he has named six current faculty members—and only one outsider, a research associate at Houston's Baylor College of Medicine—to work at the McGovern Institute. That has raised eyebrows in the community. Only one of the six works solely on issues of human cognition; the others tend toward the molecular and cellular end of the neuroscience spectrum. Thus researchers interested in higher order functions are nervous. “The direction is toward systems and molecular neurosciences, not as much toward cognition— toward understanding emotion, language, and thinking,” says Steven Pinker, an MIT psychologist advising the new institute. “There should be more focus on humans in order to distinguish it from the hundreds of other research programs in neuroscience.” The basic problem with the McGovern Institute, Pinker says, is its play-it-safe approach. An emphasis on molecular neurosciences is most likely to yield grant support, corporate funding, and Nobel Prizes. Cognitive work, by contrast, involves controversial issues such as consciousness and provides few citations in major journals and few dollars from pharmaceutical companies, he adds. Others, including at least one McGovern Institute professor, share Pinker's concerns. Nancy Kanwisher, the institute researcher who focuses primarily on human cognition, says that “to realize the institute's stated mission of understanding the higher level functions of the human mind, it will be important to appoint more faculty who work on human cognition.” Members of an MIT visiting committee this spring shared that concern, recommending that the word “cognition” be carved into the neuroscience building's facade to ensure that this area is not overlooked. McGovern, who served on the visiting committee and is active in the new institute's affairs, dismisses the concerns about an overemphasis on biology. He says that 25% to 30% of outside hires will be cognitive scientists. And he adds that systems-neuroscience research will provide the basis for understanding higher functions. ## Building up Newsome advises that the new institute make its mark by taking a broad and inclusive view of systems neuroscience. “If you focus on the neurological basis of the higher brain functions, you could be a unique force intellectually,” he says. That entails recognizing the limits of biology. “A Palm Pilot and a Cray computer may have the same components, but they are not the same—it's all about organization and complexity.” A host of brain-and-behavior-related issues, such as attention, “don't reduce to the molecular level.” Other neuroscientists warn that molecular biologists like Sharp, heady with their decades of research triumphs and formidable arsenal of tools, may underestimate the complexity of this field. “Molecular biologists want to bring our cut-and-dried approach to neuroscience,” concedes MIT biologist Nancy Hopkins. Sharp must also contend with MIT's checkered history in hiring new neuroscience talent and determining a clear research direction. Several now-prominent researchers left MIT in the past after failing to receive tenure, and the administration's support for the field since the 1960s has waxed and waned. The university also has a reputation as a place of competing fiefdoms. “MIT is a notoriously fragmented place; everyone has their own empire,” says Anthony Movshon, a New York University researcher who specializes in visual systems. Sharp says his primary constraint in hiring outsiders and launching an extensive research program is space. Until at least the end of 2004, when the new building is expected to be ready, he has limited room for new faculty members. And the building will present its own scientific challenges: Straddling railroad tracks, it will require pylons sunk deep into the bedrock to minimize vibrations in the labs above. “It's one hell of an undertaking,” says Gallistel, who reviewed the plan while on the advisory committee. But while the building plans are being drawn up, other U.S. universities like Harvard, Stanford, and the California Institute of Technology are moving quickly to expand their own neuroscience efforts. At least one senior neuroscientist—David Tank of Lucent Technologies in Murray Hill, New Jersey—slipped from MIT's grasp after Harvard and Princeton each made him an offer. Tank will go to Princeton. “There are other programs out there snapping up people,” notes Max Cowan, retired chief scientist for the Howard Hughes Medical Institute in Chevy Chase, Maryland. “MIT risks losing momentum by waiting until a building is built,” adds another researcher familiar with the McGovern. Sharp and his team won't be alone in the new building. Bringing together the myriad pieces of MIT neuroscience in one complex, he says, is the university's central strategy for creating an interdisciplinary environment. But just how well—or if—those pieces will fit together remains uncertain. Just this week, MIT's 7-year-old Center for Learning and Memory, run by Nobel Prize-winner Susumu Tonegawa, won a$50 million gift from the Picower Foundation. The money will enable Tonegawa to add three or four more researchers to his current faculty of nine, increase the center's endowment, and pay for its portion of the new complex, which will be called the Picower Center for Learning and Memory. Tonegawa says he's actively recruiting: “We have our own vision and territory.”

The brain and cognitive sciences department, which will also be housed in the complex, intends to continue its work on a broad spectrum of research, says chair Mriganka Sur. Tonegawa's and Sharp's centers complement that, he adds, by providing insight into specific issues—such as molecular and cellular mechanisms for learning and memory in the case of the Picower, and possibly systems-level work on vision and movement systems in the case of the McGovern. Center researchers have joint appointments with the department. “We want multiple approaches,” Sur adds. The Martino Imaging Center, which develops imaging techniques, will also be part of the complex.

But whereas Tonegawa, Sur, and the Martino Center will build on their existing efforts and reputation, Sharp must build from the ground up. And although he wants the McGovern to do for brain and behavior what the Whitehead Institute has done for molecular biology in the last decade, he faces constraints that Whitehead founders, including David Baltimore, did not. “The Whitehead is off-campus and fantastically endowed,” acknowledges Sharp. By contrast, the McGovern must depend on MIT for its building allotment, and it will receive only a restricted annual sum from its donor for the first 2 decades. Sharp anticipates that the McGovern will begin to pull in more money and resources as it matures, providing additional autonomy.

Building up the institute's endowment, as well as its credibility, will take several years, Sharp and McGovern point out. And the current criticism is to be expected, says McGovern, as it reflects the usual cycle of raised expectations and disappointment that follows the start of any project of this magnitude: “This [institute] had a lot of publicity, and everyone thought it would boom immediately. But things are going pretty much according to the original plan.”

Many academics agree that it's too early to judge, given the inevitable inertia when large sums of money, a major university, and prominent researchers are involved. “There are concerns in the neuroscience community,” says Tonegawa, “but I want to be patient. Adds Newsome: “It's much more important that this be done well rather than quickly.”

Still, the community will be watching closely to see if MIT and the McGovern Institute can pull off a program that founders hope will be a model for 21st century neuroscience.

10. NEUROSCIENCE

# Nearby Rival Takes Quieter Course

1. Andrew Lawler

The Massachusetts Institute of Technology (MIT) announced its neuroscience initiative last year with a bang: a cascade of media events and press packets filled with glowing profiles of the donors. Longtime neighbor and rival Harvard University is taking a different approach. Harvard has quietly committed $50 million and created 15 new positions to boost its standing in neuroscience. The university also plans to construct a new campus building for the burgeoning field. “Fewer press releases, more recruitment,” is how one Harvard scientist sums up the strategy. Some researchers around the country think that Harvard will prove to be the tortoise to MIT's hare. “Harvard is already a molecular and biomedical juggernaut, and it has one of the best cognitive programs in the country,” an MIT professor acknowledges. Even now, the two universities are competing for the field's stars. Before MIT's McGovern Institute for Brain Research could arrange an interview with David Tank, a neuroscientist at Lucent Technologies in Murray Hill, New Jersey, Harvard had made him an offer to lead its initiative. Tank, however, turned that down a few weeks ago for a job at Princeton, and Harvard's search is back on. Researchers recall that in the 1960s, Steven Kuffler of Harvard Medical School in Boston created the first proper neuroscience program in the United States, providing a strong interdisciplinary vision in that nascent field. Since then, some of the luster has worn off; the university is now seen as lagging in neuroscience. “Somewhat belatedly, Harvard has realized this is a major area,” says Max Cowan, former chief scientist at Howard Hughes Medical Institute in Chevy Chase, Maryland. Harvard's new initiative will focus on systems neuroscience, with an emphasis on the functioning of the larger nervous system rather than on the molecular level, says Harvard neurobiologist Markus Meister. That requires marshaling a wide variety of researchers and giving them the chance to work closely together. “We don't have to cook up collaboration,” he adds, noting that physicists and engineers are already wandering into his lab. The new building—on the northern tier of its Harvard Square campus in Cambridge—should help heat that interdisciplinary interaction. “Harvard has, for the most part, had buildings divided along disciplinary lines,” Meister says. “So this is a somewhat revolutionary proposal.” As at MIT, the building is likely a long-term project, requiring up to 4 years for completion. “But it also could be under way very, very quickly,” says Harvard biologist John Dowling. “The genome center was built in 2 years.” And whereas MIT plans to wait until its building is complete to hire the bulk of outside researchers, Harvard plans to provide interim space for new faculty members. “We are limited just by who we can find and convince to come here,” says Meister. This approach gives Harvard more flexibility than MIT in luring new hires. Harvard also faces two hurdles that MIT does not. First, Harvard's medical school already has a well-established neuroscience effort. “It's hard to know where the competition is stronger—with MIT or with the med school,” says one outside neuroscientist. “We're not trying to compete with the medical school,” responds Dowling, merely complement it. “They have a big push in neurodegeneracy and other such areas” that focus more on the medical aspects of neuroscience. As evidence of this coordination, Meister notes that Carla Shatz, who chairs the neuroscience program at the medical school, is also on the steering committee for the new program. Second, conducting research on monkeys is a critical aspect of neuroscience—but a sensitive topic in Cambridge. Whereas MIT plans to do extensive monkey research, Harvard officials worry about community opposition. Although a small amount of work with monkeys is already being done at the university, the bulk of it is across the river in Boston at the medical school. Still, Dowling says, the steering committee has agreed that banning monkey research on its main campus would be a mistake. “We're trying to prepare the groundwork for this, and we are doing this with great care,” one Harvard official adds. In any case, outside researchers say that Harvard will surely give MIT and other institutions a run for their money. “We will be a major player,” promises Dowling. Down the street, Sharp says he isn't worried: “Healthy intellectual competition will be good for the field.” 11. NEUROSCIENCE # Sharp Edges Into New Terrain 1. Lawler Andrew Once you've been offered the presidency of a major university, started a roaringly successful company, and won the Nobel Prize, what else is there to do? For Phil Sharp, the answer is to jump into the thick of a young and burgeoning field. The 57-year-old molecular biologist at the Massachusetts Institute of Technology (MIT) is at the tiller of one of the world's most ambitious efforts to understand the brain and behavior. And he is being closely watched by neuroscientists—many of whom view him as an outsider—to see whether he can match his earlier successes. The son of poor Kentucky farmers, Sharp raised cattle and grew tobacco to pay his college tuition. He received his Ph.D. in 1969 from the University of Illinois, Urbana-Champaign, and after working at the California Institute of Technology joined MIT's faculty in 1974. Three years later, he discovered that genes in most cells are split rather than arranged in continuous strands—a finding that earned him a Nobel Prize in 1993. A year after that landmark discovery, Sharp and several U.S. and European colleagues founded Paris-based Biogen, a biotech company that posted$261 million in revenues last quarter. In 1989, he accepted the MIT presidency, only to stun the community by then declining the job, saying he preferred to stick with teaching and research. Through the 1990s, he chaired MIT's prestigious biology department.

Sharp has an almost shy demeanor. “He doesn't like to press the flesh,” says one MIT professor. And some colleagues note that along with his natural reserve comes a penchant for administrative control. Those traits, combined with his lack of neuroscience experience, could hamper his acceptance in the community and his ability to draw top-notch researchers, they suggest.

But Sharp's old friends don't buy that. He is a quick study, a precise and detail-oriented researcher, and an able administrator, says MIT biologist Nancy Hopkins: “He sees the cutting edge instantly, and young people flock to him.” Adds Robert Birgeneau, former MIT dean of science and now University of Toronto president: “It's easy to take shots at a Nobel Prize winner coming in from [another field]. But I'm confident in the long run that Phil will succeed.”

Science recently spoke with Sharp in his cramped office in MIT's biology department, where he was busy planning the building that will house the university's new neuroscience effort. An edited interview follows:

Q: Why did you get involved in neuroscience?

A: I came into science in 1966 as a chemist, and by 1969 I wanted to make a move into molecular biology because I wanted to understand its fundamental processes. I think the biological sciences are now ready for a wonderful expansion in neuroscience; it's a frontier that is exciting a lot of people. So when I was given this opportunity, I said yes. I thought it would be fun.

Q: What do you think will be at the forefront of the field in the next 5 to 10 years?

A: There's a great deal of excitement in bringing molecular and cellular approaches to the nervous system. There are advances in imaging, where clearly there is a lot more to be done. And advances are being made in systems neuroscience. The big challenge is how to integrate molecular and cellular systems with higher order systems.

Q: How important is it to advance brain imaging?

A: Increased resolution, both temporal and spatial, is critical to bridging the gaps between cognitive science and the more physiological aspects of neuroscience. A lot of people here at MIT know how to process such information, and it would be advantageous for the neuroscience community to find a way to interface with them in an effective way.

Q: Have you attended any neuroscience meetings in the last year?

A: I'm just beginning to do that. I've been mostly working with the neuroscience community here at MIT. I certainly will go to neuroscience meetings in the fall.

Q: How do you define consciousness?

A: [Laughs] Oh my goodness, don't ask me that! [Pause] I can't offer an answer.

Q: Many neuroscientists worry that you are not familiar enough with the field.

A: I'm aware of that. I will have to depend on my colleagues to help me work through the issues. But there is some advantage to having a little distance from the specific field: It helps get people to work together and ask how we can do something new. Having that distance, and a commitment to working with my colleagues, will enable us to put together something exciting, important, and at the forefront of the field.

12. DNA FORENSICS

# Laying Ghosts to Rest in Bosnia

1. Susan Ladika is a freelance writer in Vienna.

Scientists are embarking on a major effort to identify the remains of as many as 30,000 missing persons

TUZLA, BOSNIA-HERZEGOVINA— Exactly 4419 white nylon bags are stacked on bare metal shelves, row after row, in a chilly, dimly lit room. Inside the duffel-sized bags are the remains of anonymous victims of the worst massacre in Europe since World War II: Bosnian Muslims slaughtered when Serb forces overran Srebrenica in 1995.

This month, the biggest project ever to use DNA testing to identify human remains shifts into high gear here in Bosnia. Using techniques pioneered by the Armed Forces DNA Identification Laboratory in Rockville, Maryland, scientists plan to compare DNA from victims and survivors to try to bring closure to thousands of grieving relatives. “This is the first time this has been done on a scale this large,” says the project's leader, Ed Huffine, former head of the mitochondrial DNA (mtDNA) program at the Armed Forces lab.

The plan is to sequence nuclear and mtDNA to identify an estimated 30,000 people missing since the Dayton peace accord brought Bosnia's three-and-a-half-year war to an end in 1995. About 10% to 20% of the victims are expected to be Serbs and Croats. So far, the International Commission on Missing Persons in the former Yugoslavia (ICMP), created at the G-7 summit in France in 1996, has carried out limited DNA testing on war victims, shipping samples to labs in the United States and Poland where costs run up to $5000 per sample. Operating two Bosnian labs—joined by a third later this year at Banja Luka in the Republika Srpska, the Serb-controlled portion of Bosnia—will lower the cost to$300 to $350 per body. That savings, along with donated equipment, free rent, and lower salaries in the war-ravaged country, is expected to hold the overall price tag to no more than$25 million.

Despite the grisly nature of the work, Bosnia offers advantages over other war-torn regions. For example, in Rwanda, where more than half a million people were killed during 3 months of civil war in 1994, the warm, moist climate and widely scattered remains would make such an effort impossible. That doesn't mean the Bosnian project will be easy: Technicians must work meticulously to ensure that the bone samples, in particular, aren't contaminated with other DNA. “If I was asked to do it, I think it would be close to my vision of a nightmare,” says the father of DNA fingerprinting, Alec Jeffreys of the University of Leicester, U.K. Nonetheless, he says, “there are plenty of precedents in terms of this type of analysis for mass disasters. [The technique] does work.”

Traditional forensics can't begin to tackle the task of identifying the remains of the victims, many of them Muslim men and boys slain during the fall of Srebrenica, a United Nations “safe haven” that Western peacekeepers notoriously failed to protect. Many survivors of the rampage fled to Tuzla, the nearest city under Muslim control. And they've stayed here, loath to move back to Srebrenica, now part of the Republika Srpska. Thus it made sense to ship the victims' remains, unearthed from pits or collected from fields, to the Podrinje Identification Project's morgue here in Tuzla.

The ICMP project got going last year, when it began dispatching teams to collect blood from relatives of the missing persons. So far the ICMP has amassed more than 12,000 samples, with some relatives coming here from as far away as Australia. On average, it requires 2.5 donors to identify a body, says Huffine. The ICMP has 100,000 blood kits in hand, enough in principle to identify 40,000 bodies. “Once we have 100,000 samples, then we can expect that almost every body we find can be identified,” says Amor Masovic, director of the Bosnian Muslims' missing persons commission.

The blood samples are shipped to a newly outfitted lab at the Tuzla Clinical Center, where key DNA regions are sequenced. The ICMP will open a second DNA lab in Sarajevo to sequence DNA from the bone. The researchers use the polymerase chain reaction to churn out millions of copies of the target DNA regions. The DNA is tagged with fluorescent dyes; different colors bind to each of DNA's four repeating nucleotides. A laser reads the colors to sequence the DNA.

Nuclear DNA is the material of choice for samples in which it has been preserved. Because roughly half is inherited from each parent, it offers “a unique profile for every individual, just like fingerprints,” says molecular biologist Rijad Konjhodzic, ICMP's DNA lab coordinator. In Bosnia, where tens of thousands have been killed—many from large families—telling brothers apart or son from father would be virtually impossible without nuclear DNA. The lab also uses nuclear DNA to sequence telltale markers on the Y chromosome, which can reveal a male victim's paternal lineage.

A second procedure involves screening the mtDNA, which is “better preserved than nuclear DNA and will stay [intact] longer in a corpse,” Konjhodzic says. As mtDNA is inherited only from the mother, an individual will share virtually identical sequences with all maternal relatives. Because most of the survivors—and blood donors, therefore—are women, the odds of determining the families of at least the male victims are higher than they might have been through random screening of a gender-balanced population.

“There are a lot of maternal relatives [of the victims], so we will have a lot of matching DNA,” says Konjhodzic. Often, no single test will identify remains. But combining the results from the nuclear DNA and mtDNA tests, he says, “can give us an answer.”

If successful, the DNA work could soon expand to the rest of the former Yugoslavia. The ICMP recently signed an agreement to collect blood and bone samples in Kosovo, and it hopes that a new office in Belgrade will oversee the collection of blood from those who fled Bosnia and Croatia to avoid the fighting. There is also a preliminary agreement to incorporate DNA samples from a state-funded organization in Croatia into the ICMP database. Such international agreements, says Huffine, helped overcome “a certain degree of mistrust between parties” when surviving relatives live outside the country in which the victims' remains were recovered.

Before DNA testing from blood samples became automated about 5 years ago, Huffine says, the effort in Bosnia could not have been done. But for every victim that the latest DNA science helps the ICMP team to identify, that's one less ghost to haunt this troubled land.

13. ASTRONOMY

# India Seeks Partners for 'Himalayan Space Telescope'

1. Pallava Bagla

Ground-based astronomers lust after sites that are cold, dry, high, and dark. India's new Hanle observatory fits the description—and fills a geographic gap

HANLE, LADAKH, INDIA— It's a 9-hour Jeep ride from the nearest airport, over rugged terrain. And local officials recommend a 2-day layover before setting out, to get used to the Himalayan altitude. But the cloudless skies and cold, dry climate at Mount Saraswati make this remote, 4517-meter-high site ideal for optical, infrared, submillimeter-, and millimeter-wavelength astronomy. Next week Indian officials will dedicate a new 2-meter optical/infrared telescope here with the goal of turning the world's highest observatory into one of the world's most productive scientific locales.

Ground-based astronomers are forever trying to escape the limits that Earth places on their profession. That search has led them to higher and more remote regions of the planet. The Indian Astronomical Observatory at Hanle, operated by the Indian Institute of Astrophysics (IIA) in Bangalore, is perched some 200 meters higher than the Meyer-Womble Observatory operated by the University of Denver in the Rocky Mountains and at a similar altitude to the proposed Atacama Large Millimeter Array in the Andes Mountains of northern Chile. With several observatories already in operation, the Atacama desert region may be Hanle's chief competition for new astronomy facilities.

“It's a dream site,” says Yash Pal, an astrophysicist and former chair of the University Grants Commission, who has dubbed the facility the “Himalayan space telescope.” Other scientists add that Hanle's location also makes it an excellent site for research in several other disciplines, including geophysics, atmospheric science, climatology, and conservation biology.

The Indian government has already invested over $10 million to develop the infrastructure at Hanle, located along the eastern edge of the contentious Jammu and Kashmir region but well removed from the ongoing civilian violence and fighting between Indian and Pakistani troops. One of the site's major attractions is a dedicated satellite link that allows astronomers to operate their instruments remotely, from a branch of IIA at Hosakote near Bangalore. A 2-megahertz data link, via an Indian geosynchronous satellite launched in March 2000, transmits information much faster than the shared lines used by most isolated observatories. “That makes it rather special,” says Peter Wehinger, a staff astronomer at the Steward Observatory in Tucson, Arizona. Boasts IIA director Ramanath Cowsik, “no astronomer actually needs to visit Hanle to conduct his observations.” The 2-meter telescope being dedicated next week will be used primarily by Indian scientists. But some IIA astronomers are also involved in a joint project with Washington University in St. Louis, Missouri. The effort, called the Antipodal Transient Observatory, features two identical 0.5-meter photometry telescopes, one at Hanle and one half a world away in Arizona, that will continuously track highly variable objects, including a class of active galaxies, called blazars, with a supermassive black hole at their center. Government officials hope the 2-meter telescope, to be named after India-born astrophysicist and Nobelist S. Chandrasekhar, will be a test-bed and demonstration site for larger projects. “The telescope is a small steppingstone toward making this the best astronomical site in the Eastern hemisphere,” says Bhuwan Chandra Bhatt, who manages the Hanle facility. Valangiman Subramanian Ramamurthy, a nuclear physicist and secretary of the Department of Science and Technology, which has funded the Hanle observatory, says that the government is studying a$100 million proposal from IIA and the community for a binocular telescope, with two primary mirrors in the 6.5- to 8-meter range, that would have international partners. He says that India may include a down payment for the project in next year's budget.

Japanese and U.S. astronomers familiar with the site extol its suitability for a range of investigations. “The site appears to have great potential, especially for the submillimeter and midinfrared spectral regions,” says Wehinger. “In addition, it is extremely dark and has excellent seeing conditions.” Munetaka Ueno, an assistant professor of earth sciences and astronomy at the University of Tokyo, has visited the site twice and says that its low winter temperatures make it an ideal location for follow-up observations to ASTRO-F, an infrared imaging and surveyor satellite scheduled for launch in 2003. Japanese scientists from the Nobeyamo Radio Observatory have pronounced the geography suitable for a submillimeter array, and they are monitoring water vapor at the site.

Astronomy is not the only discipline expected to reap benefits from the Hanle observatory. Its location on the Asian tectonic plate fills an important gap in a network of geodetic research stations on the Indian plate. IIA scientists are collaborating with the Center for Mathematical Modeling and Computer Simulation in Bangalore to quantify the dynamic deformation field in Ladakh and have already confirmed predictions from models that India and Eurasia are colliding at the rate of 55 mm a year.

For atmospheric scientists, Hanle offers an opportunity to collect better data on factors that shape climate. Researchers from the National Physical Laboratory in New Delhi are already monitoring ozone and ultraviolet levels, while the Indian Institute of Tropical Metrology in Pune is gathering data on aerosol levels. A French group plans to install a permanent CO2 monitor at the site.

Despite its barren and desolate look, the region is home to highly endangered animals such as the elusive snow leopard, Tibetan gazelle, and the Tibetan wild ass. It is also the only known breeding ground for the highly threatened black-necked crane. Hanle is already providing scientists from the Wildlife Institute of India and World Wide Fund for Nature with a base camp for their expeditions.

To secure Hanle's future, Indian officials must convince the worldwide scientific community that it not only offers exquisite viewing conditions but also fills an important geographic gap in data collections. In the meantime, they plan to lavish resources on it. “No stone will be left unturned to make the site a world-class facility,” says Ramamurthy.

14. ECOLOGICAL SOCIETY OF AMERICA

# Forests: No Greenhouse Antidote?

1. Dan Ferber

MADISON, WISCONSIN—More than 3000 ecologists gathered here from 5 to 10 August for the 86th annual meeting of the Ecological Society of America (ESA). Hot topics included trees and global warming (see below), the risks of transgenic crops (see “Breeding a Hardier Weed”), and vanishing tropical mammals (see “Case of the Missing Mammals”).

Some experts claim that the world's forests can absorb enough carbon dioxide to reduce the impact of further global warming. But at least one type of hardwood forest may not be up to the job. Rather than storing extra carbon in long-lasting trunks and branches, an experimental sweetgum stand in Tennessee socks most of the CO2 in tiny roots that rapidly die and decompose. That process sends the gas right back into the atmosphere.

Researchers have long wrangled over the ability of forests to serve as carbon sinks for excess greenhouse gases. It's clear that saplings in open-top enclosures respond to high CO2 with growth spurts, stepping up photosynthesis and making more leaves and wood than would trees sucking unadulterated air. But what's true for a stand of saplings may not be true for a mature forest, says ecologist Rich Norby of Oak Ridge National Laboratory in Tennessee. That's because leaf coverage maxes out as a tree matures—putting limits on photosynthesis and, thus, on its capacity to soak up excess CO2.

To find out how much CO2 mature trees can absorb, Norby and colleagues built towers 4 years ago to pump CO2into the canopies of four stands of young sweetgums. As Norby reported at the meeting, during the first year most of the extra carbon went into wood, with the gassed-up sweetgums accumulating 35% more carbon than control trees grown in unsupplemented air. But 2 years later, that wood differential had narrowed to 7%. More than twice as much carbon as in the controls ended up in the fine roots—thin structures that fall off and die each year. Soil organisms quickly consume the detritus, releasing CO2 that diffuses out into the air.

Forest ecologist Adrien Finzi of Boston University calls the results “really interesting” but cautions that they may not hold true in other forests. The mechanism of carbon storage certainly differs in an experimental pine stand he studies in North Carolina. Although the loblolly pines there exposed to extra CO2 also store less extra carbon in wood after a few years as they run short on nutrients such as nitrogen (Science, 6 April, p. 36), the carbon ends up primarily in leaf litter, not the fine roots. That suggests to Finzi that researchers must check more than a couple of stands to understand how different forest types respond to high CO2levels.

The bottom line for sweetgum and loblolly pine, anyway, is that neither leaf litter nor fine roots offer long-term carbon storage. For that reason, says biogeochemist William Schlesinger of Duke University, planners shouldn't count on forests as CO2 saviors.

“These terrestrial sinks,” he contends, “are just not adding up to much.”

15. ECOLOGICAL SOCIETY OF AMERICA

# Breeding a Hardier Weed

1. Jocelyn Kaiser

MADISON, WISCONSIN—More than 3000 ecologists gathered here from 5 to 10 August for the 86th annual meeting of the Ecological Society of America (ESA). Hot topics included trees and global warming (see “Forests: No Greenhouse Antidote?”), the risks of transgenic crops (see below), and vanishing tropical mammals (see “Case of the Missing Mammals”).

In the vitriolic debate over the potential risks of transgenic crops, one big concern is that wild relatives may commandeer valuable traits and turn into “superweeds” that spread, unchecked, across the land. Two new studies add hard data to what has been mostly a theoretical discussion. One finds that genes from a crop can persist in a weed for many generations, while a second supports the idea that if genes that protect against viral infection slip into wild plants, there could be serious consequences.

Although neither finding pins down the risks, these and other studies have convinced some ecologists that genetically modified (GM) crops are being rolled out too hastily. “We really need a lot more data before we make assumptions” about safety, says Alison Power of Cornell University in Ithaca, New York, who presented her work on viruses.

Conventional wisdom says that crop traits are unlikely to persist in the wild in part because crossbreeding crops and weeds yields hybrids that tend to reproduce poorly. “In the crop-breeding and weed science world, there's always been a feeling that crop genes would not persist,” says Allison Snow of Ohio State University in Columbus, who described a 6-year experiment on half-wild, half-crop radishes planted next to wild radishes in Michigan. Snow's group found that crop genes had no trouble sneaking into the weeds—and staying there.

While the first cross between these relatives (the F1 generation) had low fertility—as few as 60% made seeds—several traits, including white flowers and variants of two enzymes, showed up in subsequent generations of wild radishes. And second-generation hybrids—crosses between F1 and wild plants—grew almost as well as the wild radish. Although it's not a big surprise that traits showed up in the wild radishes, “it's important to quantify persistence,” says plant scientist Neal Stewart of the University of North Carolina, Greensboro. Radish, he notes, “is a very nasty weed.”

But whiter flowers and a more croplike metabolism are hardly the makings of superweeds. What might help weeds outlast the competition, however, is if a jumping gene they acquired were able to help them fend off viral attack, says Power.

Her test case is crops modified to resist the barley yellow dwarf virus. To find out whether the trait could give a leg up to wild plants, Power first looked at whether the virus shows up much in nonagricultural ecosystems. The team tested for virus in wild grasses near Ithaca. Surprisingly, up to 60% of samples of 16 grasses, including yellow foxtail and wild oats, were infected. “To me, that's mind-boggling,” Power says, because it has been assumed that wild plants had evolved resistance to the virus.

Wild oats, not surprisingly, often grow in the same areas as crop oats, a variety of which has been equipped to withstand the virus. (Although companies have developed at least 18 virus-resistant crops, the only ones now being planted commercially are potato, papaya, and squash.) Hypothetically, the trait could slip into the wild oats. To see if this might make the weed hardier, the team grew cultivated oats and wild oats side by side in a greenhouse. Some plants were inoculated with the virus; others were disease-free. Compared to infected plants, healthy wild oats grew much better than healthy crop oats, producing about 25% more biomass; meanwhile, crop oat biomass actually dropped 13%. “If you remove the virus,” Power says, “it gives a huge advantage to wild oats because they're no longer suppressed.” The result could be a weed that aggressively invades croplands.

Although other meeting attendees found the results intriguing, they caution that only field tests of virus-free wild oats will determine the actual risks. Power expects to complete those studies this summer.

16. ECOLOGICAL SOCIETY OF AMERICA

# Case of the Missing Mammals

1. Jocelyn Kaiser

MADISON, WISCONSIN—More than 3000 ecologists gathered here from 5 to 10 August for the 86th annual meeting of the Ecological Society of America (ESA). Hot topics included trees and global warming (see “Forests: No Greenhouse Antidote?”), the risks of transgenic crops (see “Breeding a Hardier Weed”), and vanishing tropical mammals (see below).

As primates, tapirs, and other large mammals in the tropics get picked off by poachers, the harm could extend far beyond the devastation to the species themselves. Because these herbivores influence the makeup of plant communities by eating leaves and seeds, their loss could transform the overall structure of tropical forests, ecologists have argued. Experiments described here at the meeting suggest that this scenario is now playing out in Mexico.

Three decades ago, ecologists posited that herbivores and seed-eating animals might help maintain the remarkabg le diversity of tropical forests by thinning abundant seedlings to make room for less prolific species. Nobody tested this hypothesis on a large scale, however, until an alarming report on Mexico's fragile ecosystems appeared in 1991.

That's when ecologist Rodolfo Dirzo of the Universidad Nacional Autónoma de México (UNAM) noticed a striking contrast in the appearance of two forests in southern Mexico: Montes Azules in Chiapas, a largely intact forest with its original 40 or so species of nonflying mammals; and the smaller Los Tuxtlas, Veracruz, in which about 46% of the same suite of species— including jaguars, tapirs, deer, and monkeys —had in the last 3 decades been hunted down or captured for the pet trade. Los Tuxtlas's understory was carpeted with dense patches of seedlings from trees such as Nectandra ambigens and Brosiumum alicastrum—far fewer species, but far thicker growth, than the understory at Montes Azules.

Factors other than browsing, such as a heavier-than-normal seed production one year for certain tree species, might have explained the strangely uniform and thick understory in Los Tuxtlas. But the findings inspired Dirzo and other ecologists to examine what happens when mammals are absent. At the meeting, Dirzo and graduate student Eduardo Mendoza described recent experiments that firm up a role for mammals in fostering diversity in Los Tuxtlas and Montes Azules.

First, Dirzo's team set up fences to exclude all mammals from 2-square-meter plots located near five canopy trees at each of the two sites. After 2 years, the number of species of understory plants in fenced and control plots in Los Tuxtlas were, as expected, exactly the same. But in Montes Azules, which still has its mammals, the scientists counted a total of only 91 species in fenced plots compared to 112 in plots visited by animals.

The UNAM team is also exploring a possible mechanism for why large-seeded plants may be favored when large and medium-sized mammals are missing. Dirzo and Mendoza suspect that mice—which may be more abundant with their predators gone—are eating mainly small seeds in Los Tuxtlas. The two have found that caged mice seem to eat mostly small seeds in the lab, and they're now testing preferences in the field with cages that only small animals can enter.

Although these results appear to support a role for herbivores in promoting forest diversity, Dirzo cautions that he can't predict the long-term fate of Los Tuxtlas's species-poor understory or whether the findings apply to other forests. Indeed, researchers reported at the meeting that in much larger, 35-meter-by-40-meter exclosures at Barro Colorado Island in Panama, plant diversity is going up—not down—after 8 years. A similar 3-year study in Australia has gotten mixed results: more species, but certain species are flourishing and may be crowding out others. “It may take longer to see differences” at a larger scale, notes investigator Catherine Gehring of Northern Arizona University in Flagstaff.

No matter how their role in forests plays out, Dirzo says, the plight of tropical mammals demands action: “We need to view defaunation as another major environmental change.”