# News this Week

Science  14 Jan 2005:
Vol. 307, Issue 5707, pp. 188
1. ATMOSPHERIC SCIENCES

# NOAA Loses Funding to Gather Long-Term Climate Data

1. Jeffrey Mervis

Congress has eliminated funding for a fledgling network of 110 observation stations intended to provide a definitive, long-term climate record for the United States.

The surprise assault on the Climate Reference Network (CRN) was buried in the 3000-page omnibus spending package for 2005 signed last month by President George W. Bush (Science, 3 December 2004, p. 1662). Legislators also took a bite out of a long-established atmospheric monitoring network that includes the historic time sequence of increasing carbon dioxide levels measured at Hawaii's Mauna Loa. Both networks are key pillars in a much-touted international “system of systems” for earth observation that the Bush Administration has called essential for resolving uncertainties in the connection between greenhouse gas emissions and climate change (Science, 20 August 2004, p. 1096). While federal officials say they plan to “limp along” this year and hope for better news in 2006, some scientists worry that the cuts signal a lack of political support for filling those gaps.

“[CRN] ties everything together,” says Richard Hallgren, former director of the National Weather Service and executive director emeritus of the American Meteorological Society. “Eliminating it would be an absolute disaster.”

The excision of CRN's $3 million budget is part of a$10.6 million cut in the $24.3 million climate observations and services program, which supports a far-flung monitoring system operated by the National Oceanic and Atmospheric Administration (NOAA). The reference network was part of the president's 2005 request for NOAA and was funded in separate bills that had moved through the House and Senate. But “it disappeared” after conferees completed work on the massive bill that bankrolled dozens of federal agencies, notes program head David Goodrich. CRN is meant to provide a 50-year climate record—including solar radiation, wind speed, and relative humidity—that is of much higher quality than existing temperature and precipitation records from weather stations. The weather stations are often staffed by volunteers, and the data are undermined by changing urban conditions, poor maintenance, and other variables. In contrast, CRN will rely on state-of-the-art equipment located in protected areas such as national parks and inspected regularly. “This network,” says Thomas Karl, its moving force as director of NOAA's National Climate Data Center in Asheville, North Carolina, “will eliminate the adjustments and corrections that we've had to make in the data” that have spawned so much debate about recent U.S. climate trends. But this year's budget squeeze, he says, raises questions about the viability of the network, begun in 2001 and with 56 stations now operating. For starters, the cuts will force 16 new stations scheduled to be commissioned this year into “hibernation.” It also means no money for some 20 technicians who crisscross the country to tend the equipment. Karl has siphoned off$1.5 million from other programs to keep on a skeletal maintenance crew. But he's worried that the hibernating stations could become degraded without proper maintenance and that further delays could trigger a clause in its site leases that requires NOAA to dismantle the entire system if the stations are not in use.

Also at risk are the five observatories operated by NOAA's Climate Monitoring and Diagnostics Laboratory (CMDL) in Boulder, Colorado. These sites, from Alaska to the South Pole, measure levels of carbon dioxide, carbon monoxide, methane, halogenated compounds, ozone, aerosols, and other atmospheric constituents. The data help researchers build better climate models.

A $2.5 million budget cut means that the observatories will be serviced less often, and several contractors will be given the boot, says CMDL Director David Hofmann. That will increase the burden on an aging system that, among other achievements, includes a Hawaiian project begun by Charles Keeling in 1958 that first alerted the world to a steady rise in C02 levels. “The road is barely passable now,” Hofmann says about the 180- kilometer roundtrip to the Mauna Loa summit. “At some point we won't even be able to make it up there.” Beyond the loss of data from individual monitoring stations, the cuts jeopardize the Bush Administration's Global Earth Observing System of Systems (GEOSS), a planned linking of existing networks to paint a comprehensive, real-time picture of what's happening to the planet. “It raises the question of whether the nation is willing to support a sustained, long-term effort to do the best possible job of monitoring our climate,” says Kenneth Kunkel of the Illinois State Water Survey, who chairs CRN's ad hoc science working group. To Kevin Trenberth, head of the climate analysis section at the National Center for Atmospheric Research in Boulder, the message from legislators is even bleaker. “It's almost as if some people don't want to know how the climate is changing,” he says. “Maybe they prefer uncertainty, so that they can avoid taking action.” 2. RAINFALL MONITORING # Report Bucks NASA's Plan to End Mission 1. Andrew Lawler The forecast for an aging NASA spacecraft that keeps tabs on tropical rainfall turned stormy last week. A National Academies' panel released an interim report urging the space agency to keep the satellite flying at least through the end of the year. But NASA officials insist they may have to shut it down as early as this summer, before the academy can finish its study. Both climate researchers and weather forecasters are eager to continue gathering data from the joint U.S.-Japanese Tropical Rainfall Monitoring Mission (TRMM) launched in 1997. They argue that the instruments could continue beaming back data for another 6 years. But NASA says that unless the National Oceanic and Atmospheric Administration (NOAA) agrees to take over operations, the constraints of time, money, and safety will force it to shut off instruments. NASA requested the study after scientists and members of Congress criticized agency plans to halt operations last summer (Science, 13 August 2004, p. 927). The academy panel, chaired by Eugene Rasmusson of the University of Maryland, College Park, “strongly recommends continued operation of TRMM,” at least through the end of 2005. The panel notes that TRMM's precipitation radar and microwave imager in particular provide a “powerful” set of data points for long-term understanding of rainfall patterns as well as near-term observation of hurricanes. It says TRMM also complements NOAA's polar weather satellites, which fly in a different orbit. “The instruments are in excellent shape,” says project scientist Robert Adler of NASA's Goddard Space Flight Center in Greenbelt, Maryland. But managers at NASA headquarters say they can't keep TRMM flying. “The real dilemma is physics, not money,” says one NASA official. The longer the satellite remains in orbit, the greater the risk that it cannot be sent into a controlled reentry above the Pacific Ocean and the more resources—personnel to monitor the satellite—will be needed. So while it would cost$4 million a year to continue operating TRMM, the reentry effort could take years and cost as much as $16 million. Meanwhile, NASA wants to spend every available penny to build a Global Precipitation Mission that would provide broader coverage starting later in the decade. NASA deputy science chief Ghassem Asrar said that, although TRMM has yielded “significant scientific data,” the agency must remain “vigilant” to ensure a controlled reentry. And that could mean shutting off the instruments as early as summer. “The sooner we prepare for deorbit, the better,” he adds. TRMM advocates say an uncontrolled reentry does not pose a significant risk, however, citing a 2002 finding by NASA's own safety directorate. “The community is going to have to speak out,” says Adler. But wanting the data isn't enough. Somebody—NOAA, Congress, the White House, or Japan—must also come up with the money and persuade reluctant NASA managers to keep TRMM on the job. 3. CLINICAL TRIALS # Facing Criticism, Industry Offers to Share Data 1. Jennifer Couzin Five trade groups representing pharmaceutical companies worldwide are urging members to release more information about clinical trials. However, some see the proposals as a way to stay ahead of legislation that could compel the release of such information. The companies have been under pressure since revelations that they kept trial data for antidepressants and other drugs secret. Congress failed to act last year on calls for a mandatory clinical trials registry, with penalties for noncompliance, but those bills are expected to reappear. The co-sponsor of one such bill, Representative Henry Waxman (D-CA), said last week that “nothing” in the industry's announcements “is going to dissuade me” from pursuing legislation. But the Pharmaceutical Research and Manufacturers of America (PhRMA), a Washington, D.C.-based trade group, says it would prefer for Congress to wait and “see if the voluntary efforts are going to work,” says spokesperson Jeff Trewitt. Voluntary registries in the past have included only a fraction of ongoing and completed trials. Seven of the nearly 100 members of the Association of the British Pharmaceutical Industry (ABPI) have participated in its registry, launched in May 2003. A 2003 study of U.S. cancer trials found that fewer than half of those sponsored by industry appeared on the government Web site (clinicaltrials.gov). The U.K.'s ABPI is pinning its hopes on the World Health Organization's efforts to establish a global trials database by July; it will recommend that members post trials and results there. The new PhRMA plan recommends adding trials for all ailments to clinicaltrials.gov. Other groups behind the effort include the European Federation of Pharmaceutical Industries and Associations, the International Federation of Pharmaceutical Manufacturers & Associations, and the Japan Pharmaceutical Manufacturers Association. They recommend the release of “all clinical trials to determine a medicine's therapeutic benefit,” says Richard Ley, an ABPI spokesperson. Critics such as Drummond Rennie, deputy editor of the Journal of the American Medical Association, aren't optimistic. “Marketing forces and self-interest … are going to win out every time over the ethics of doing the right thing,” he says. 4. INFECTIOUS DISEASE # Polio Eradication Effort Adds New Weapon to Its Armory 1. Leslie Roberts With success still frustratingly elusive, the leaders of the global program to eradicate poliovirus are reintroducing an old tool to fight the disease: an oral polio vaccine designed specifically to protect against the most pervasive strain of poliovirus, known as type 1. The only vaccine used in the 16-year eradication campaign targets three strains of the virus. The new monovalent oral polio vaccine (mOPV)—a version of which was used extensively before the adoption of trivalent OPV in the 1960s—offers “more wallop per punch,” says Bruce Aylward, who coordinates the program from the World Health Organization (WHO). It is not a silver bullet, caution officials at WHO and the U.S. Centers for Disease Control and Prevention. Program officials also stress that mOPV will augment, not replace, the well-honed strategy of immunizing every child under age 5 where polio remains a threat with several doses of trivalent OPV each year. But if mOPV works as hoped, “it may be what it takes to tip the scale,” says David Heymann, who heads WHO's eradication effort. The project, which began in November, is on an accelerated track. Sanofi Pasteur in Lyon, France, and Delhi-based Panacea Biotec have promised to deliver 200 million doses this spring. WHO officials say this could well be the fastest a vaccine has been produced and approved. The agency actually wanted the vaccine even sooner, says Francois Bompart, vice president of medical affairs for Sanofi, but the company simply could not retool production from trivalent OPV fast enough. Still, if the vaccine is ready by May, as planned, the partners should be able to deliver two rounds in Egypt and parts of India before the beginning of the high season in July to September, when viral transmission peaks. The mOPV plan, to be announced by the end of January, offers another key benefit: It will give officials a leg up on testing a key component of the vaccine stockpile needed to deal with emergency outbreaks once eradication is achieved. The use of mOPV is designed to root out the virus in areas where it is most entrenched—typically, overcrowded slums with abysmal sanitation and booming birthrates, like greater Cairo and parts of western Uttar Pradesh, Bihar, and Mumbai in India. Despite dramatic increases in the number of national immunization days last year and the percentage of children reached in each one, viral transmission still persists in these areas, notes Hamid Jafari, who directs the global immunization division at CDC. Meanwhile, the epidemiology of the disease has shifted, says Roland Sutter of CDC. The program has successfully cleared the world of type 2 poliovirus, he says, and type 3 is “just hanging on by its teeth.” In Mumbai and all of Egypt, for instance, type 3 has not been detected since October and December 2000. That opens the door for the reintroduction of mOPV against type 1 poliovirus. Since the early 1970s, polio experts have known that trivalent OPV simply isn't as effective in hot tropical climes, requiring perhaps five to eight doses to confer immunity instead of the standard three (Science, 26 March 2004, p. 1960). In Egypt and parts of India, especially, conditions are “very, very ripe for the virus,” says Jafari. Although Egypt recorded just one case of paralytic polio in 2004, environmental samples collected from open sewers show that the type 1 poliovirus is well established in the ecosystem. The same is true in parts of India; Mumbai, for instance, reported just one case of paralytic polio in 2004, but 84 environmental samples tested positive. “So the question is, do we keep pounding away, or do we get some sharper edge to our tool?” “very, very ripe for the virus,” says Jafari. Although Egypt recorded just one case of paralytic polio in 2004, environmental samples collected from open sewers show that the type 1 poliovirus is well established in the ecosystem. The same is true in parts of India; Mumbai, for instance, reported just one case of paralytic polio in 2004, but 84 environmental samples tested positive. “So the question is, do we keep pounding away, or do we get some sharper edge to our tool?” asks Jafari. He suspects that edge will come from a new version of mOPV. Past experience with mOPV has demonstrated that it is much more potent in prompting an immune response. Data from five tropical countries showed that just one dose of mOPV type 1 conferred immunity in 81% of those vaccinated, says Sutter. By contrast, the seroconversion rate for one dose of trivalent OPV in tropical countries is roughly 30% to 40%. The benefit occurs because the live attenuated vaccine virus, which replicates in the gut, doesn't have to compete with the other two virus types for cells susceptible to infection. MOPV also has a long safety record, notes Bompart. But because no company has produced it in years, and it is no longer licensed, the vaccine must be reviewed as a new product. Regulatory agencies in Egypt and India have agreed to expedite the review based on historical data, while also requiring new clinical trials and postmarketing studies. Sanofi is manufacturing 50 million doses for Egypt, and Panacea is ready to produce up to 150 million for India for introduction in May. Although all children under age 5 in the target areas will receive mOPV, the partners expect the biggest payoff to come from vaccinating very young children with low or little immunity, who are most likely to transmit the disease: “We really do need to get the youngest ones immunized as quickly as possible,” says Sutter. “MOPV will help us do it faster.” At this stage, cautions Aylward, the benefits are theoretical. And even if mOPV does boost immunity as expected, says Bompart, it is not clear that it will make a “real world” difference in terms of stopping transmission. One concern is that the promise of a more effective vaccine will divert attention from the need to reach every single child with multiple doses of trivalent OPV, which must continue, says Aylward. Even so, the idea is gaining steam. Polio expert Paul Fine of the London School of Hygiene and Tropical Medicine says the plan makes “good sense” scientifically and also shows that the program has an “open-mindedness” toward new tactics and vaccines, which may be needed to finish the job. 5. SOUTH ASIA TSUNAMI # U.S. Clamor Grows for Global Network of Ocean Sensors 1. Eli Kintisch An oft-ignored plea to the U.S. government to improve a federally funded tsunami warning system is falling on more receptive ears in the wake of the tragedy in South Asia. Scientists at the National Oceanic and Atmospheric Administration (NOAA), which runs a six-buoy network of pressure sensors in the Pacific Ocean, have seen previous efforts to expand the network rejected on fiscal grounds. But last month's earthquake and tsunami, which have claimed at least 150,000 lives, have changed the terms of the debate. “If there was a window of opportunity, this would be it,” says Jay Wilson, an earthquake and tsunami coordinator for Oregon's office of emergency management. Completed in 2001, the Deep Ocean Assessment and Reporting of Tsunamis (DART) network is made up of six sensors tethered to the ocean floor that can detect tsunamis as small as 1 centimeter, relaying data instantly via satellite from buoys to tsunami warning centers in Alaska, Washington state, and Hawaii. Two detectors currently sit off the coasts of Washington and Oregon, three operate near Alaska, and one sits about 1000 km south of the equator. NOAA scientists believe that about 20 detectors could provide adequate coverage for coastal warnings around the Pacific, and 50 would provide the basis of a global system. But NOAA's budget makes no provision for any expansion of the current network. Enlarging the DART system is “one of the things we're looking at,” says a spokesperson for the White House Office of Science and Technology Policy, which convened a meeting last week of several federal agencies that support related research. Last week, in separate teleconferences with Senate staff and House members and staff, Eddie Bernard, the director of NOAA's Pacific Marine Environmental Laboratory in Seattle, Washington, ended a presentation on previous tsunami studies with a proposal for a 53-detector global DART array (see map). “The grand scheme is a global approach,” says NOAA oceanographer Frank Gonzalez, who leads the agency's tsunami research program. The House Science Committee plans a hearing this winter on improved tsunami warning systems for U.S. and international shores, according to a staffer. Some legislators aren't waiting. For example, on 6 January, Senator Joe Lieberman (D-CT) proposed that the United States, along with “cooperating nations,” expand the DART network. However, even a global system would have limitations, notes U.S. Geological Survey (USGS) seismologist David Oppenheimer, pointing to a 1700 earthquake on the Cascadia subduction zone that sent giant tsunami waves crashing into the Pacific coast of North America in minutes. In such a situation, he says, “the buoys aren't going to save anybody; there's just so little time.” In the meantime, science agencies are already helping researchers eager to work at the affected sites. The National Science Foundation is funding several teams studying the tsunami's behavior along coastlines in Sri Lanka and India. The foundation has also described to White House officials how it could expand its portfolio in telemetry and sensing to improve the Global Seismographic Network, which it funds. And altimetry data from the joint U.S./French JASON-1 satellite have provided scientists a rare glimpse into the tsunami's birth. “The satellite just happened to be passing over as the tsunami was taking shape,” says NASA spokesperson Gretchen Cook-Anderson. 6. PALEONTOLOGY # New Fossils Show Dinosaurs Weren't the Only Raptors 1. Erik Stokstad The Mesozoic era is called the “Age of Dinosaurs” for good reason. For 185 million years, they diversified with ferocious gusto, evolving into a panoply of predators and prey that fill the record books for size and shapes. Mammals, meanwhile, were nocturnal, shrewlike nobodies that snatched insects and stole the occasional egg. Only after dinosaurs went extinct 65 million years ago could mammals escape from the shadows and begin to thrive. Or so the story goes. In this week's issue of Nature, Chinese paleontologists describe the largest Mesozoic mammal skeleton ever found, more than a meter long. And this furry Goliath wasn't content just to eat bugs: A smaller relative was discovered nearby with the bones of a baby dinosaur in its stomach. “This thing was probably hunting and eating relatively large-sized dinosaurs,” says Guillermo Rougier of the University of Louisville, Kentucky. “It forces us to think about [Mesozoic] mammals as a fully diversified group, not just in their typical role of insectivores.” The new fossils, each about 130 million years old, come from the famous fossil beds of Liaoning Province in northeastern China. Paleontologists had already discovered skulls of the smaller animal, called Repenomamus robustus (Science, 12 October 2001, p. 357), but could get only a vague estimate of its body size. Now the same team has found a fairly complete specimen of an adult. Squat, with powerful legs, it probably weighed about 4 to 6 kilograms. “We would say it looked something like a Tasmanian devil,” says team member Yaoming Hu, a graduate student at the City University of New York. Collaborators include his adviser Jin Meng of the American Museum of Natural History in New York City and colleagues at the Institute of Vertebrate Paleontology and Paleoanthropology in Beijing. While removing rock from the specimen, preparators made a rare discovery: teeth and bones strewn about inside the ribcage, in the likely position of the animal's stomach. The jumble included the remains of a herbivorous dinosaur hatchling, a 14-centimeter-long Psittacosaurus. One leg appears mostly intact, suggesting that the mammal dismembered and wolfed down its food. Given the large, sharp teeth and powerful lower jaw, the team suspects that Repenomamus was a predator, but Hu acknowledges it's hard to tell scavengers from hunters. R. robustus wasn't the only mammal that dinosaurs had to worry about. Another skeleton, better preserved, was even larger. Named Repenomamus giganticus, it was 1 meter long and weighed roughly 12 to 14 kg, as much as a modern coyote. “It was probably competing with carnivorous dinosaurs for food and territory,” Hu says. And that raises interesting questions, notes Anne Weil of Duke University in Durham, North Carolina. “What these finds really allow us to do—at least speculatively—is ask how mammals might have influenced dinosaur evolution,” she says. In other words, Mesozoic mammals may have cast a shadow of their own. 7. ITALY # Synchrotron Staff Protests Funding Cuts 1. Alexander Hellemans* 1. Alexander Hellemans is a writer in Naples, Italy. NAPLES, ITALY—The 250 employees of Sincrotrone Trieste, which operates Elettra, Italy's large synchrotron light source, put down their tools for a day this week to protest government funding cuts that triggered a financial crisis. After it lost half its income in 2002, the facility took out bank loans, which it assumed that the government would pay off. Staff and users now fear that if the government does not come to its rescue, the synchrotron may have to be mothballed. “The laboratory is suffering. If something breaks down, we cannot repair it,” says Silvia Di Fonzo, a physicist at Sincrotrone Trieste and a labor union representative who helped organize the strike. Like other synchrotrons, Elettra speeds electrons around a particle accelerator to produce x-rays that researchers use as probes in a wide variety of fields. Commissioned in 1993, Elettra hosts 840 users per year from across Europe and developing countries. But in 2002 the government drastically cut some research institution budgets, including one that supports Elettra. As a result, Elettra lost 50% of its$33 million yearly operating budget, although it retained the half that comes directly from government.

According to Alfonso Franciosi, CEO of Sincrotrone Trieste, the government encouraged the company to take out bank loans to cover the shortfall. “The lab operated for 3 years with loans from local banks, and the debts are now adding up to [$20 million],” says Franciosi. The government has repeatedly promised to restore Elettra's missing$18 million per year starting in 2005, he adds. But many were alarmed to see that Elettra is not included in the 2005 government budget, which was approved last month. Elettra officials are hoping that new funding will be included in a decree on national competitiveness that the government will issue at the end of January.

Guido Possa, Italy's deputy research minister, says the trouble is that Sincrotrone Trieste was set up as a private company, making it hard for the government to fund it directly. “The problem is when you have to manage public money, you have to follow certain rules.”

8. BIOMEDICINE

# As the Worm Ages: Epilepsy Drugs Lengthen Nematode Life Span

1. Ingrid Wickelgren

Although pharmacists have proven medications for ailments as varied as migraines and bacterial infections, they have little to offer in the fight against aging other than unproven remedies. But new evidence suggests that the right prescription for longevity may already be hidden behind the pharmacy counter.

Geneticist Kerry Kornfeld and his colleagues at Washington University in St. Louis, Missouri, report on page 258 of this issue that a class of antiseizure drugs markedly extends the life span of the roundworm Caenorhabditis elegans. The scientists screened 19 classes of medications prescribed for other uses for potential longevity effects. “These compounds are approved for human use, so they have [molecular] targets in humans,” says Kornfeld, although he cautions that there is no evidence yet that the anticonvulsants he tested slow aging in people.

Because these drugs act on the neuromuscular systems of both humans and worms, the finding also hints at a direct link between the neuromuscular system and the aging process, says geneticist Catherine Wolkow of the National Institute on Aging in Baltimore, Maryland. Furthermore, the data indicate that although the drugs' mechanisms of action partly involve molecular pathways already known to govern aging, those pathways tell less than the whole story. “The work opens up the possibility that there may be new targets not yet explored that affect aging and neuromuscular function,” says Wolkow. “That's a pretty important finding.”

With a life span of a few weeks in the lab, C. elegans is a favorite subject for longevity studies. Since the early 1990s, researchers have linked mutations in dozens of worm genes to extensions of the creature's lives. Given all the drugs on the market, Kornfeld speculated that at least one of them was likely to retard aging or promote longevity by affecting those gene targets.

So about 4 years ago, Kornfeld's graduate student Kimberley Evason began exposing separate groups of 50 worms to various drugs, from diuretics to steroids, at three different dosages. Most of the compounds the worms ate off their petri dishes had toxic effects. After 8 months of negative results, Evason tested the anticonvulsant ethosuximide (Zarontin). A moderate dose, she found, extended the worm's median life span from 16.7 days to 19.6 days, a 17% increase. Lower doses had a lesser effect, and higher doses were toxic.

Evason then discovered that two related anticonvulsants also lengthened worms' lives, one of them by as much as 47%. By contrast, a chemically related compound that does not have antiseizure activity had no similar effect. That is “nice evidence” that the compounds' ability to extend life span is related to their effectiveness as anticonvulsants, says geneticist Javier Apfeld of Elixir Pharmaceuticals in Cambridge, Massachusetts.

The drugs are thought to control seizures in people by acting on certain neuronal calcium channels. Exactly how the drugs extend life span in worms is unknown, although they seem to stimulate the nematode neuromuscular system. Kornfeld's team discovered that the drugs affect two types of neurons: those that govern egg laying, leading to earlier release of eggs, and those that control body movement, making the worms hyperactive.

Unlike many of the genetic mutations that affect worm longevity, the drugs don't act primarily through the worm's insulin-like signaling system, the St. Louis group revealed. For example, treatment with two of the anticonvulsants still lengthened the lives of worms with life-curbing mutations in an insulin-pathway gene. “We think the nervous system effects are more complicated than simply regulating insulin signaling,” Kornfeld says.

The next step is to test whether the drugs have any antiaging effects on higher organisms, such as flies and mice. “The nervous system might have a central function in coordinating the progress of an animal through its life stages, leading ultimately to degeneration,” Kornfeld speculates. Still, he adds, “it's very early days for understanding the connection between neural function and aging.”

# Plan for Chiropractic School Riles Florida Faculty

1. Yudhijit Bhattacharjee

Faculty members are questioning a plan to make Florida State University (FSU) in Tallahassee the first public U.S. university with a chiropractic medicine school. This week the faculty's graduate policy committee voted to examine the proposal amid concerns that implementing it would sully the university's reputation. But FSU administrators say such a graduate program, if ultimately adopted, would be a valuable addition to health care education and could benefit millions of Americans who suffer from back pain.

“There's a very good reason why no public university offers a degree in chiropractic medicine,” says Raymond Bellamy, director of orthopedic surgery at FSU's Pensacola campus and leader of the opposition campaign. “It's because having a chiropractic program would seriously undermine the scientific tradition of any institution.” Not so, says FSU provost Larry Abele, an invertebrate morphologist: “A graduate education and research program aimed at moving chiropractic medicine into a scientific and evidence-based realm is certainly worth exploring.” The flap is reminiscent of a dispute at York University in Toronto, Canada, when faculty members blocked a plan to offer an undergraduate degree program that would have been affiliated with the Canadian Memorial Chiropractic College (Science, 19 February 1999, p. 1099).

Last March, at the urging of a state senator who's also a chiropractor, the Florida legislature authorized $9 million per year to establish such a school. FSU administrators conducted a feasibility study and drew up a proposal for a College of Complementary and Integrative Health that would offer a 5-year Doctor of Chiropractic degree. That proposal, which cited studies that it claimed showed “why more than 15 million Americans use chiropractic care,” was to be presented this week to the university's board of trustees and 2 weeks later to the state Board of Governors. Abele says chiropractic medicine is a legitimate field of study that deserves a place in the academic mainstream. He also says the university will not implement the proposal unless it has the support of the faculty: “The legislation simply authorizes funds for setting up the school. It does not require that we do so.” Even so, FSU officials advertised in November for the position of dean of the proposed school. Richard Nahin, a senior adviser at the National Center for Complementary and Alternative Medicine at the National Institutes of Health, says the popularity of chiropractic care among Americans makes it important to understand whether “chiropractic works, what conditions it may work for, and how it may work. Having a state chiropractic school could be of benefit to the field,” he adds, “as that school would probably educate chiropractors using the same scientific, evidence-based approach used to train medical doctors.” None of those arguments is enough to convince neuroscientist Marc Freeman, one of 40 FSU professors—including Nobel Prize-winning chemist Harry Kroto and physicist J. Robert Schrieffer—who have signed a petition against the proposal. Apart from the lack of a scientific basis, he says, the chiropractic school is a threat to FSU's academic independence. “We cannot have the legislature forcing a program on a public university,” he says. 10. DEVELOPMENTAL BIOLOGY # Bird Wings Really Are Like Dinosaurs' Hands 1. Elizabeth Pennisi Molecular studies have smoothed a wrinkle in the assumption that modern birds had dinosaur ancestors. After tracing the expression of two genes important in the development of digits in wings and other limbs, researchers have concluded that the three digits in bird wings correspond to the three digits in dinosaurs' forelimbs. For years, most embryologists had considered them different. “This may settle a long-standing controversy and will strengthen the therapod [dinosaur]-bird link,” says Sankar Chatterjee, a paleontologist at the Museum of Texas Tech University in Lubbock. Over the past decade, new fossils and phylogenetic analyses have convinced most paleontologists that birds are dinosaurs. A few researchers have refused to accept this evolutionary pathway, and one tenet of their argument has to do with how to count fingers. Terrestrial vertebrates typically have five fingers, numbered 1 to 5. In both dinosaur fossils and birds, just three of these digits are fully developed, a trait that at first glance supports a dinosaur-bird connection. But dinosaur forelimbs have the first three digits, with stubs for the last two. In contrast, going by some embryological evidence, birds appear to have retained the middle three fingers. In 1997, for example, ornithologist Alan Feduccia, a noted critic of the bird-dinosaur link at the University of North Carolina, Chapel Hill, and a colleague tracked digit formation in turtles, alligators, ostriches, cormorants, and chickens. They concluded that the bird “fingers” were the middle three, whereas the reptiles' were the first three out of those five possibilities (Science, 24 October 1997, p. 666). That inference fueled arguments against a dinosaur-bird connection. In 1999, Yale University's Gunter Wagner and Jacques Gauthier, proposed a controversial compromise: that in avian ancestors, developmental signals transformed tissue in position to become digits 2, 3, and 4 into digits 1, 2, and 3. Determined to resolve the issue, Alexander Vargas, an evolutionary-developmental biologist at the University of Chile in Santiago, and John Fallon, a developmental biologist at the University of Wisconsin, Madison, compared the embryological development of digits of mice and chickens. Working in Fallon's Wisconsin lab, they traced the activity of two genes crucial for digit development, Hoxd13 and Hoxd12. Fallon and others had already shown that among other differences, the development of the first digit in mice relies on Hoxd13 but not Hoxd12, whereas the other digits need both. The first digit also forms differently. “There are several molecular and developmental reasons to consider that digit 1 is distinct from other digits,” says Vargas. When the researchers looked at the chick embryo, they found that the wing's initial digit—until now considered to be digit 2, especially by opponents of the bird-dinosaur theory—used Hoxd13 but not Hoxd12, indicating that it really is the first digit, developmentally speaking. Birds therefore have the same digits as dinosaurs, Vargas and Fallon conclude in the January issue of The Journal of Experimental Zoology Part B: Molecular and Developmental Evolution. In birds, the first digit is simply masquerading as the second one. “I think it's the best evidence yet that digits gain their identities from [their genetic milieu] and not from position,” says Richard Prum, an ornithologist at Yale University. Friesten Galis, a functional morphologist at Leiden University in the Netherlands, is not convinced. Studies of digit development in other animals do not show as clear a difference in Hoxd13 and Hoxd12 expression as Vargas and Fallon presume, he points out. Galis cites new evidence he's recently obtained by studying birds with abnormal digit patterns that continues to support the idea that the digits in bird wings are equivalent to digits 2, 3, and 4 in other animals. And Feduccia is even more skeptical about the study and its conclusion. Hand development is just not that malleable, he insists. The flap over bird wings continues. 11. STEM CELLS # California's Bold$3 Billion Initiative Hits the Ground Running

1. Constance Holden

Controversy over California's new stem cell initiative didn't end when the state's voters approved Proposition 71 in November by 59% to 41%. But now that the new California Institute for Regenerative Medicine (CIRM) is beginning to take shape, the debate has shifted from ethics and costs to how the enterprise will operate. Supporters are still brimming with confidence, however.

The new institute as yet has no staff, no home, and just a one-page Web site (http://www.cirm.ca.gov/). But at a press conference last week, Robert Klein, CIRM's newly elected chair of the board, repeated assurances that he expects grants to start flowing by May. “I admit that I am an optimist,” he added.

At its first full meeting, held on 6 January at the University of Southern California in Los Angeles, the 29-member board, called the Independent Citizen's Oversight Committee (ICOC), set up subcommittees to find outsiders for “working groups” that will establish policies on research funding, ethics, and facilities construction. They also launched the hunt for a president for CIRM—ideally a seasoned research administrator who will be in charge of recruiting scientific advisers, directing staff, and participating in the formation of policies from lab construction to intellectual property agreements. Klein will head the search.

At the meeting, ICOC also elected as Klein's vice chair Edward Penhoet, a chemist who has straddled many sectors as a Berkeley dean, co-founder of Chiron Corp. in Emeryville, California, and most recently as president of the Gordon and Betty Moore Foundation in San Francisco. As a scientist and public health expert, Penhoet has a “complementary set of skills” to Klein's, says ICOC member Edward Holmes, dean of the University of California, San Diego, Medical School. Penhoet is heading the search for space for the institute's administrative headquarters. Also on the front burner is securing a start-up loan of $3 million from the state. The critics have been busy as well. A primary concern, voiced by the Center for Genetics and Society in Oakland, among others, is that the initiative—which is immune from legislative tampering for the first 3 years—has been framed so that it may freely violate state and federal regulations on matters such as open meetings and conflicts of interest. Critics also worry that taxpayers won't get proper returns from patent and royalty fees, and some are troubled that Klein designed the entire initiative and slid into the top job without a hint of competition. But supporters seem to have limitless confidence in 59-year-old Klein, who put more than$3 million of his own money into the Proposition 71 campaign and helped raise more than $20 million. A graduate of Stanford law school and president of Klein Financial Corp. in Fresno, California, which finances the construction of low-cost housing, Klein was drawn into the stem cell issue because his 14-year-old son Jordan has juvenile diabetes. Committee members say they can negotiate the ethical minefield. “Whatever connections we might have anywhere” have to be a matter of public record, notes Holmes. Klein has pledged not to hold investments in biomedical or real estate enterprises “reasonably likely to benefit” from the stem cell program. He plans to step down after serving 3 years of his 6-year term. And he has resigned as head of the California Research and Cures Coalition (CRCC), which has been reconstituted as a nonprofit education and lobby group. CRCC hopes to build confidence with four community forums to be held around the state this month, at which citizens will discuss “practical and ethical issues” with scientists. For now, at least, supporters seem to outweigh critics. “I think [the organizers of the CIRM] are drawing in the best this country has to offer,” says Michael Manganiello of the Christopher Reeve Paralysis Foundation. Some scientists have expressed skepticism about the wisdom of funding research by means of popular vote and worry that the public has been oversold on the promises of the research. But it's hard to find a critic among stem cell researchers, who stand to benefit from the$3 billion and the new wave of attention that CIRM will bring to their field.

12. DRUG SAFETY

# Gaps in the Safety Net

1. Jennifer Couzin

After the discovery that several popular medicines may have harmed tens of thousands of people, experts are hunting for better ways to monitor drugs on the market

For those who trust government-approved drugs, 2004 was not a banner year. Merck, the maker of the anti-inflammatory medicine Vioxx, pulled the drug off the global market in September after a clinical trial linked it to heart attacks and strokes. In October, U.S. regulators concluded that a class of antidepressants can trigger suicidal thoughts in children and stepped up warnings of this danger. In December, studies of Celebrex, another arthritis medication, pointed to more cardiac risks. Just 5 days before Christmas, scientists running an Alzheimer's prevention study announced that Aleve, approved as a nonprescription painkiller in 1991, may also trigger heart problems.

These cases all involved drugs that had gone through extensive safety testing and had been on the market for years. And they raised disturbing questions: Should public authorities like the U.S. Food and Drug Administration (FDA) rethink what they consider acceptable risk? Should they move more aggressively to monitor approved drugs and restrict their use when problems surface among a fraction of patients?

The crises of 2004, some observers say, could trigger a shakeup in how drugs on the market are monitored. “I would like to believe that Vioxx could do for this decade what thalidomide did for the 1960s,” says Jerry Avorn, a pharmacoepidemiologist at Harvard Medical School in Boston and author of the book Powerful Medicine: The Benefits, Risks, and Costs of Prescription Drugs. In the 1950s and 1960s, women in 46 countries who took thalidomide for morning sickness gave birth to more than 8000 children with severe abnormalities. Governments worldwide passed legislation requiring meticulous safety tests before a drug could be approved.

Judging by the numbers, the Vioxx case should elicit at least as strong a response. David Graham, an FDA drug safety officer, says it may have caused 100,000 heart attacks and strokes, a third of them fatal. Regulators from France to New Zealand had nervously discussed “signals” hinting at harm caused by the drug before 2004 but were unable to nail down their suspicions. It took a company-sponsored clinical trial to accomplish that (Science, 15 October 2004, p. 384).

Since the Vioxx debacle, officials running postmarketing surveillance systems are considering how they might do better. The uncomfortable truth, some say, is that all such systems have gaps. Several nations and the European Union (E.U.) boast aggressive surveillance systems, but many are new and have not been rigorously tested. “Everybody's in bad shape here,” says Bert Leufkens, a pharmacoepidemiologist at the University of Utrecht in the Netherlands and an adviser to the Dutch and European Union drug agencies.

No public system is under greater pressure than FDA. Some members of Congress want to change it. Senator Charles Grassley (R-IA) plans to introduce legislation early this year to make FDA's existing Office of Drug Safety (ODS)—which is responsible for tracking the safety of drugs once they reach the market—independent of the drug approval mechanism in the Center for Drug Evaluation and Research (CDER), where ODS now resides. Academics and a few industry people say ODS needs a stronger legal mandate and more funds—but to make this happen, they must persuade a White House and Republican Congress that has traditionally recoiled from hands-on drug regulation.

Postmarketing surveillance systems, however, run on more than a legal mandate. Some of the strongest critics of the U.S. approach, like Avorn, say that FDA has all the police power it needs; it just needs to apply it creatively.

## Risk tolerance

Forty years ago, European countries seemed relatively relaxed about drug approvals in contrast to FDA, which had earned a reputation for caution. Europe released thalidomide onto the market in the late 1950s, for example, and left it there for years. But an FDA reviewer spotted potential problems; she declined to let thalidomide through, and it was not approved.

Today, the roles are often reversed: FDA is frequently the first to approve drugs. The FDA staff is paid in part by “user fees” from regulated companies. Industry and patient groups lobby for speedy decisions, and FDA now turns some applications around in 6 months.

FDA has allowed greater risks in recent years than some other regulatory agencies, according to observers such as Lucien Abenhaim, a pharmacoepidemiologist at the University of Paris and McGill University in Montreal, Canada. He recalls getting little attention when he flew to Washington, D.C., in 1995 to warn FDA about life-threatening heart and lung ailments associated with the diet drug duo fenfluramine and dexfenfluramine (fen-phen). A recent study Abenhaim led had suggested that they increased cardiopulmonary risks up to 23-fold; European governments responded by limiting access to them. But FDA approved dexfenfluramine “without proper warning,” says Abenhaim, only to see the drugs withdrawn in haste a year later after more than 100 people developed cardiopulmonary abnormalities.

Critics also fault FDA for its handling of the diabetes drug Rezulin. Two months after approving it in 1997, U.K. regulators pulled it off the British market because of concerns about liver failure. FDA read a different risk-benefit calculus in the data. “Most every country on Earth pulled the drug 2 full years before the FDA did,” says Avorn.

Graham, a career FDA employee, claims that pressure to move faster has made CDER a “factory” for approving new drugs. Graham recently made headlines when he asserted in a Senate hearing that consumers “are virtually defenseless” against a repeat of the Vioxx affair. He said in a later interview that “my experience with FDA has been that they don't have the will” to go after drugs with safety issues. Graham says ODS, where he works, is often shunted aside because its views on a particular drug may threaten the judgment of FDA officials who allowed that drug on the market.

In an e-mail, FDA's press office declined to make senior officials available to answer questions for this article.

## Shy gorilla?

Despite its woes, FDA remains a world leader in some areas—suggesting, perhaps, how tough it can be to police approved medications. “In many ways, the FDA is better able than we are at the moment to support independent research relating to pharmcovigilance,” says Panos Tsintis, head of pharmacovigilance, safety, and efficacy at the 25-member European Medicines Agency (EMEA), the E.U.'s London-based drug approval and surveillance agency formed in 1995. Abenhaim praises FDA for its expertise but thinks these talents are poorly applied to postmarketing surveillance. He attributes this to government policy that gives FDA little authority to aggressively track and test marketed drugs.

Like agencies in many industrialized countries, FDA has two methods of conducting postmarketing surveillance. One is to commission specific studies. The other is to gather spontaneous reports of adverse effects in a database called MedWatch. Britain's drug regulatory agency claims to have the “world's largest computerized database of anonymized patient records,” the General Practice Research Database (http://www.gprd.com/). It's a fantastic research tool, says professor of medicine policy Joe Collier of St. George's Hospital Medical School in London—if you have a specific question and can pay. Full access to GPRD costs $600,000 a year. No system is without flaws. One weakness of FDA's MedWatch, notes drug safety expert Alastair Wood, associate dean at Vanderbilt University in Nashville, Tennessee, is that it only skims the surface. He estimates that the 22,000 adverse events that are reported to the database each year represent only 3% to 10% of those experienced by patients. And the source could be biased: More than 90% of the reports come from companies, which are required to hand over reports given them by doctors, and fewer than 10% from doctors directly, FDA says. Furthermore, FDA's MedWatch is isolated from patient care. In parts of Europe, “pharmacovigilance” offices are housed in hospitals, and physicians can wander down the hall to report adverse events. “It's not … an office somewhere in [FDA] with 8000 people collecting data,” says Leufkens. Then there's New Zealand's Medsafe, which employs 10 people on a budget of under$1 million to oversee more than 10,000 drugs on the market. Seventy percent of adverse-event reports to Medsafe come from general practitioners, 20% from hospitals, and 10% from companies. Those who submit reports can expect to hear from a Medsafe employee who's hunting for additional details. According to the World Health Organization, New Zealand's reporting rate on drug adverse effects is among the top three worldwide, says Stewart Jessamine, a Medsafe spokesperson.

New Zealand's challenge is very different from FDA's: The country has just 5000 prescribers and 3.5 million people. That makes it both easier to staff an interactive surveillance network and tougher to detect signals from dangerous drugs because fewer people are ingesting them, says Jessamine. Medsafe was watching Vioxx, for example, but officials could only conclude that “there's something happening, but we don't know what it is,” he says.

This reflects the glaring limitation of even the best event-based reporting system: Doctors only report rare ailments that are easily linked to a drug. Vioxx and the heart attacks it induced are a different story altogether. “The doctor says … Mr. Blogg died from a heart attack, but he was 80, he did have angina and high blood pressure,” says Jessamine.

## Active surveillance

There are few ways to detect common but deadly hazards. One is through a clinical trial, like the one that brought down Vioxx. Another is by means of an epidemiology study that relies on massive databases, the kind maintained by HMOs such as Kaiser Permanente or government-funded health plans like Medicaid. Even though studies using these databases are cheap compared to clinical trials, running about half a million dollars, not many agencies fund them, says Brian Strom, a biostatistician and epidemiologist at the University of Pennsylvania in Philadelphia. Results from epidemiology studies sometimes carry less weight than those from clinical trials: Graham spent 3 years working with Kaiser in California on an epidemiology study of Vioxx and came to much the same conclusions as Merck eventually did, but his findings didn't prompt action against the drug.

FDA generally relies on companies to run postmarketing trials, called phase IV studies, often requesting them as a condition for a drug's approval. But follow-through is poor, a failing some blame on insufficient funds and others on a reluctance to confront drug companies. An FDA analysis released in 2003 found that more than 50% of phase IV studies don't even get started. FDA officials have said they need congressional authority to force companies to complete such studies.

Graham and Avorn think FDA has more muscle than its officials admit. If the FDA chief announced publicly that “there's a signal from Vioxx, the company's not responding,” says Avorn, “the mere threat would have been enough” “there's a signal from Vioxx, the company's not responding,” says Avorn, “the mere threat would have been enoughe” to force a clinical trial. The remedy, he and others say, is to give the drug safety office more clout.

Senator Grassley is proposing that the office remain within FDA but be distinct from CDER—a structure similar to that of the U.K.'s Medicines and Healthcare Products Regulatory Agency, in which safety regulators don't mingle with those who approve drugs.

Acting CDER chief Stephen Galson and other senior FDA officials declined to comment on FDA's postmarketing surveillance. But Jane Henney, FDA commissioner from 1998 until 2001 and now senior vice president and provost for health affairs at the University of Cincinnati, disagrees with Graham that FDA puts safety on the back burner, although she acknowledges that there will always be disagreement about how to handle drug risks. “As long as I was at the agency, the office of safety had a strong voice at the table,” she says. Henney attributes FDA hesitancy to a simple problem: lack of resources. “We made a number of requestse” to both Congress and the White House for increases in postmarketing surveillance funding, she says. Proposed changes included expanding FDA's access to large HMO databases to get a better grasp on adverse drug reactions and investing in research to more nimbly detect hints of drug problems. “Unfortunately, we just never got the money,” says Henney.

Today, FDA devotes 5% of CDER funds, about $24 million, to the center's drug safety office, a fraction on par with the United Kingdom but proportionally lower than some other countries (see table). Experts in both the United States and Europe believe that their countries should earmark far more money for postmarketing surveillance. But money works best when melded with creativity. Even if FDA's drug safety office is refurbished, pressing postmarketing studies into action could mean flexing muscles drug regulators aren't accustomed to exercising. Amid some controversy, France launched a new surveillance program several years ago that was spurred by the approval of Vioxx and Celebrex. EMEA had approved the drugs across Europe, but Abenhaim, then France's director general of health, wasn't convinced they worked as well as promised. He requested that a 2-year study of 40,000 people on Vioxx, Celebrex, or traditional nonsteroidal anti-inflammatory drugs begin before allowing France's national health care system to reimburse for the drugs. Abenhaim's position provoked an outcry, and he was asked to explain his position to the country's national ethics committee. In the end, the study was done. Since then, 50 more drug studies have been ordered. But, says Abenhaim, “there is still a lot of reluctance.” Nor is the system efficient: The Vioxx study, for example, has not yet been released. The Netherlands is eyeing a similar surveillance framework, says Leufkens. Meanwhile, EMEA, eager to harmonize drug approvals in Europe, will launch its own system in November 2005 to compel studies, using punishments such as financial penalties, says Tsintis. The greatest worry of those pressing hardest for change, particularly in the United States, is that even thousands of possible deaths due to Vioxx won't prompt an overhaul of postmarketing drug surveillance. “My fear,” says Avorn, “is that we will not be able to take advantage of this moment.” 13. RADIATION HAZARDS # Kyrgyzstan's Race to Stabilize Buried Ponds of Uranium Waste 1. Richard Stone With help from the West, local experts are devising ways to head off a potential landslide of Soviet-era mine tailings MAILUU-SUU, KYRGYZSTAN—Alexander Meleshko scrambles up a terraced hillside, skirting tons of gravel laid to buttress the slope. All seems quiet on a cool day in late autumn, but Meleshko, a geologist with Kyrgyzstan's Ministry of Ecology and Emergency Situations (MEES), knows that this tranquil setting in the southwestern corner of the country is a disaster waiting to happen. Looming above is a 250-meter-high sandstone ridge rippled with shades of brown, yellow, and ochre. In front, entombed in an artificial hill, are 115,000 cubic meters of slurry chock-full of radioactive metals—enough to fill a football stadium. The noxious cocktail includes isotopes of thorium, copper, arsenic, selenium, lead, nickel, zinc, radium, and uranium. Meleshko, decked out in Army fatigues, stamps a foot on the soil. “There's more than 10,000 microroentgens per hour of radioactivity under here,” he says—roughly 1000 times the local background rate. All that protects Meleshko and the surrounding region from the tailings in this impoundment (called T-3), a leftover of Soviet-era uranium mining, is a meter-thick layer of clay. Experts have identified T-3 as a far-reaching threat: In the scariest scenario, the ridge could dissolve in a landslide, sweeping the tailings into the nearby Mailuu-Suu River. That's a chilling possibility. The Mailuu-Suu is a tributary of the Syr Darya River, the main source of irrigation water for the 6 million residents of the densely populated Fergana Valley. “It's a huge potential danger,” says Vyacheslav Aparin, a senior scientist with the Complex Geological-Ecological Expedition in Tashkent, Uzbekistan. The valley, which extends southwest into neighboring Uzbekistan and Tajikistan, is a melting pot of peoples and beliefs, including enclaves of Islamic fundamentalists. A radioactive accident here could be traumatic to a region already simmering with tension. The risk of a catastrophe is rising. Heavy spring rains in recent years have made landslides a more frequent occurrence in mountainous Kyrgyzstan, and in this seismically active region, a tremor capable of unleashing a devastating landslide could strike at any time. “There's not much we can do if there's a strong earthquake,” says Isakbek Torgoev, director of the Geopribor engineering center in Kyrgyzstan's capital, Bishkek. Tajikistan and Uzbekistan are also grappling with the legacy of Soviet uranium mining. Anecdotal reports suggest that some sites in Tajikistan are in an even more precarious state than those in Kyrgyzstan. But Mailuu-Suu, poised like a match near Fergana's tinderbox, is deemed the top priority. After years of handwringing, Kyrgyz authorities are on the verge of doing something. In September, Kyrgyzstan received the first installment of a$6.9 million World Bank loan to deal with the most hazardous uranium sites, starting with T-3.

Work could begin as early as next summer—which would be none too soon. Authorities will be pacing anxiously when meltwater and rain renew their assault on the fragile land in the spring. “In our narrow valleys, gravity wins sooner or later,” says MEES's Nurlan Kenenbaev.

When the Soviet Union pushed its atomic bomb program to full throttle after World War II, Mailuu-Suu, nestled in the foothills of the Tian Shan mountains, was wiped off maps and became known simply as P.O. Box 200. Specialists arrived here in droves.

Officials in faraway Moscow pampered their uranium jocks with high salaries and ample food trucked in even during lean times. “The standard of living was much higher than it is today,” says longtime resident Ashir Abdulaev, an assistant mayor of Mailuu-Suu and local MEES representative. But many in Mailuu-Suu and other uranium towns in Cen tral Asia had no idea why they were so well off. Operated by the Ministry of Medium Machine Building, which ran the bomb program, the uranium facilities “were top secret,” says Alexander Kist, a radiochemist at the Institute of Nuclear Physics in Tashkent. According to Torgoev and others, the first Soviet bomb was made from uranium milled at Mailuu-Suu.

In those days, says Aparin, “there was no such science like ecology, so the idea was to just get the uranium out of the ground as fast as possible.” Nazi POWs and prisoners from Tatarstan, Ukraine, and elsewhere toiled in shafts laden with radon, a radioactive gas that wafts from the ore. “They didn't know what they were mining,” says Torgoev. Even the miners' housing was built from uranium-rich stone. (According to Kist, the skeletal remains of workers are radioactive.) Lavrenti Beria, one of Stalin's most feared henchmen and chief of the bomb project, would come to Mailuu-Suu to check on the mines. Today his former quarters, garishly decorated with yellow and blue plastic wall tiles, is part of a hotel.

Most people connected with the mines have left or died, but reminders of Mailuu-Suu's past linger. Tidy, two-story stone houses, built by German prisoners for the town's elite, line a street leading to a pair of former uranium mills. One mill was converted to a factory, Isolite, which makes insulation materials and glass wire. The other mill is a heap of rubble. The Soviets abandoned it in the 1960s after radioactive contamination of the machinery had grown intolerable even by the lax standards of the day, Meleshko says. Rather than dismantle the site, the Soviets blew it up. These days, locals have been seen scavenging tainted metal from it.

If anything, the shadows in Mailuu-Suu are deepening. Its population has dwindled from 36,000 to 23,000, in part due to an exodus after the uranium industry shut down. Local health officials assert that radioactive contamination is killing off many who stayed behind. “The cancer rate here is twice that of the rest of the republic,” claims Nemat Mambetov, chief of Mailuu-Suu's Sanitary and Epidemiological Station. Lung cancer is the biggest killer, he says, followed by stomach and digestive tract cancers—although he acknowledges that limited financing has resulted in poor record-keeping. Western experts are circumspect. “We've had trouble getting reliable epidemiological data,” says Peter Waggitt, an expert on uranium tailings with the International Atomic Energy Agency in Vienna. “You can't automatically just blame every cancer on the uranium.”

In 1958, flooding after a landslide ate into one of the impoundments at Mailuu-Suu (T-7), sweeping an estimated 300,000 cubic meters of tailings into the river, says Yuriy Aleshin, a geophysicist with Geopribor. The tailings, he says, are thought to have spread tens of kilometers downstream. The consequences of the accident may never be known: Soviet authorities hushed it up, and records of any follow-up studies have long since disappeared.

From qualitative analogies with Cold War-era tailings sites in the United States, Richard Knapp, a geoscientist with the Proliferation and Terrorism Prevention Program at Lawrence Livermore National Laboratory in California, has come up with a preliminary estimate of the potential risk posed by T-3: If it were to disgorge its contents today, the contamination would cause about 600 cancer deaths in the vicinity of Mailuu-Suu over 100 years, he estimates. In contrast, a 25-year cleanup at two dozen U.S. tailings sites has prevented about 1300 deaths combined, Knapp says.

## A uranium rust belt

Kyrgyzstan is not alone in its woes. Next door in Uzbekistan, the major headache is Charkesar, a fenced-off, decommissioned uranium mine that Aparin and others say may have sickened thousands of local residents. Uranium mining is still a big business there, unlike in Kyrgyzstan. These days, however, companies rely on a sulfuric acid process rather than miners to extract ore.

Tajikistan too was a major uranium producer in Soviet times. Processing took place at three sites: Adrasman, Chkalovsk, and Taboshar. According to a 2004 report from the state mining enterprise Vostokredmet, twice in recent years mudflows have destroyed impoundments at Taboshar. One Western expert who has visited the site describes having seen “mountains of tailings,” one 200 meters high, in the open air. Tajikistan will host a workshop in May, sponsored in part by the U.S. Department of Energy, to highlight the region's problems and attract international donors.

Nor is Mailuu-Suu the only worry for Kyrgyzstan. Another 12 hot spots are scattered across the country. After the Soviet breakup in 1991, says MEES Director Anarkul Aitaliev, “no maintenance was done on the tailings.” The U.S. State Department is funding a $500,000 effort, led by Lawrence Livermore with support from Russia, to deal with the Kadzhi-Say impoundment on the south shore of Lake Issyk-Kul. Kyrgyzstan has staked its development on tourism, and the lake is its biggest asset. “Anything that jeopardizes Issyk-Kul is a concern,” says Knapp. But the consensus of international agencies is that Mailuu-Suu poses the biggest risk. “Mailuu-Suu is critical because at the end of the road is another country,” says Waggitt. Exacerbating the situation is that the environment is literally falling to pieces. Meleshko has charted a steady rise in the incidence of landslides in Kyrgyzstan, from about 100 major slides per year in the 1970s to more than 200 last year. Last year, 45 people in Kyrgyzstan died as a result of landslides, including 33 in a single disaster last April not far from Mailuu-Suu. The higher frequency of landslides has followed, almost in lockstep, seasonal increases in precipitation. “The more rain and snow, the more chance of landslides,” Meleshko says. In May 2002, a slide just a kilometer upstream from T-3 engulfed several Isolite buildings. Today, an estimated 5 million cubic meters of soil at the site are at risk of sliding down. Although it wouldn't plow into T-3 directly, such a landslide could lead to a replay of the 1958 incident at T-7, this time disemboweling T-3. ## Move it or leave it? A fluke of Cold War political geography makes Mailuu-Suu—and T-3 in particular—more hazardous than other sites. From 1946 to 1967, more than 10,000 metric tons of uranium oxide were processed in Mailuu-Suu. Many more tons were shipped here for processing from Saxony, in eastern Germany, and elsewhere in the East Bloc. After some of the uranium was extracted, the leftover slurry was piped into the clay-lined impoundments. Tailings from the imported ore are hotter than those from local deposits, Torgoev says, accounting for a substantial fraction of the radioactivity sequestered in T-3. Last year, thanks to a grant from the European Union, gravel was laid to shore up the base of the 20-meter-deep T-3. Now Kyrgyzstan is about to embark on a broader$16.7 million effort to clean up Mailuu-Suu. An initial \$12 million from the World Bank, Japan, the Global Environment Facility, and the Kyrgyz government “will allow us to deal with the most dangerous parts of the problem,” says Meleshko. The first step is to remove soil from the ridge above T-3 that's deemed especially prone to sliding down. With funds in hand, Kyrgyz authorities are now selecting contractors; work could begin as early as next summer.

T-3's ultimate fate is unclear. “It's very difficult to come up with a solution; it's a huge volume,” says Meleshko. Complicating matters, the drainage system that prevented rain and groundwater from saturating the 50-year-old impoundment no longer works, says Knapp. Water percolating into T-3 explains why the tailings, which have the consistency of toothpaste or newly mixed cement, are unusually mushy—and unstable.

One option that Kyrgyz authorities are considering is to pump out the tailings from T-3 and store them at a more stable location nearby. Such a procedure has been carried out successfully in the United States. “About half of [the U.S. impoundments] were just picked up and moved somewhere else,” says Knapp. He advocates this solution for T-3, as it would be almost impossible to eliminate a landslide risk. Some experts in Kyrgyzstan, including Torgoev, also favor this strategy. But there are risks: Such an operation could expose workers to increased radiation levels, and if an accident were to occur, says Aparin, “you could contaminate the whole valley.” Also a big issue, says Waggitt, is where precisely to put the tailings. “If you look around the valley, there's an awful lot of instability in the landscape,” he says.

The other option is to leave the tailings in place and sculpt the ridges to avert a serious landslide threat. Although a massive job, it might be considerably cheaper than hauling out the tailings, says Meleshko. Experts in Uzbekistan are pressing for a third option: installing a pipe to divert any floodwaters generated by a landslide upriver around the T-3 impoundment. “I see this as giving a 100% guarantee of success,” says Vladimir Kupchenko, director of Uzbekistan's Complex Geological-Ecological Expedition.

It may take up to 2 years to make a decision and bring in new equipment and expertise, says Kenenbaev of MEES: “Everything we have is from the Soviet period.”

In the meantime researchers must play a waiting game. Making a brief stop on the long road back to Bishkek, Meleshko admires a landscape that could have been painted by El Greco. Dark-gray clouds cling to the mountains, their snowcapped peaks and glacial fields glowing eerily white in the twilight. The treeless land stretches like crumpled brown velvet as far as the eye can see. But Meleshko can't tear his thoughts from Mailuu-Suu. “We've waited 40 years to do something about it,” he says. “I hope nature will let us wait a few more months.”

14. SOUTH ASIA TSUNAMI

# Failure to Gauge the Quake Crippled the Warning Effort

1. Richard A. Kerr

Seismologists knew within minutes that the earthquake off Sumatra must have just unleashed a tsunami, but they had no idea how huge the quake—and therefore the tsunami—really was

When 1000 kilometers of subsea fault ruptured that Sunday morning west of Sumatra, seismologists knew a tsunami was on the loose, but they failed to grasp the true magnitude of the quake and therefore the hugeness of the tsunami it had spawned. Measuring earthquakes is no easy task, and only a single, unstaffed lab on the other side of the world had the proper tool.

“Everybody underestimated [the earthquake] in the beginning,” says Charles McCreery, director of the Pacific Tsunami Warning Center (PTWC) in Ewa Beach, Hawaii. That was because no seismologist was using the one, long-available technique that could nail down the magnitude of a truly great quake. Seismologists have long known that the commonly available methods underestimate any quake larger than about magnitude 8.5. The Sumatra-Andaman Islands quake turned out to be 9.0. That's 30 times stronger than initial estimates and was guaranteed to produce a deadly, far-ranging tsunami. A computer at Harvard University, using a mathematical technique called centroid moment tensors (CMT), automatically calculated a magnitude of 8.9 within 2 hours of the quake, but the results became available only when seismologists later checked its readout.

At PTWC, staffers calculating magnitudes from the seismic data circulating worldwide at first thought December's quake looked like a fairly run-of-the-mill magnitude 8.0. When the first informational PTWC bulletin went out 15 minutes after the quake, “there could have been a local [Sumatran] tsunami by then,” says McCreery, but at 8.0, nothing damaging would ever make the 2-hour trip across the 1600 kilometers of the Bay of Bengal to India or Sri Lanka. So that first bulletin, sent to participating Pacific Rim countries that PTWC is mandated to alert, reported the 8.0 magnitude and the absence of any threat around the Pacific.

As more seismic data arrived, the quake's perceived size grew. The magnitude 8.0 estimate had come from a technique dubbed Mwp, which was designed for speed and used some of the first seismic waves arriving at seismometers. But speed had a drawback. With Mwp, the rupture is assumed to be a one-dimensional point. That works pretty well up to magnitude 7.5 or 8. However, faults rupture along planes, not at points, and a bigger quake can rip hundreds of kilometers along the fault. The P waves used in Mwp zip through the earth much more directly than seismic surface waves do, but surface waves paint a clearer picture of the full, two-dimensional extent of a great earthquake's rupture. After gathering a full hour of data including late-arriving surface waves, McCreery and his colleagues were confident they had a magnitude 8.5.

So an hour after the quake—with the tsunami halfway across the Bay of Bengal— PWTC issued a second bulletin reporting the higher magnitude. Within minutes, the U.S. Geological Survey's National Earthquake Information Center (NEIC) in Denver, Colorado—the world's de facto seismic clearinghouse—sent out its own, independently calculated surface wave magnitude of 8.5 to its worldwide alert list. Any seismologist aware of the quake would now know it was underwater and sizable.

What that meant for the tsunami threat was unclear, even to McCreery and his colleagues. “Around 8.5 is when we start to feel there's some kind of reasonable threate” at greater distances from the quake, says McCreery, “but it's not consistent.” Lacking a system of sea-floor sensors to detect and gauge tsunamis in the Bay of Bengal, “we felt pretty frustrated,” he says. But “none of us was thinking it would be a 9,” he adds, so PTWC's second bulletin merely noted “the possibility of a tsunami near the epicenter.” Meanwhile, according to news reports, low-level scientists across Asia were passing word to superiors of a large, threatening underwater quake in the region, but their similarly vague warnings went unheeded.

Chances are that alarms would have traveled faster and farther if seismologists knew what a computer in Cambridge, Massachusetts, was learning. By the time the first waves hit India, it had automatically calculated a magnitude of a little over 8.9, according to Harvard seismologist Göran Ekström. That was 30 times more powerful than an 8.0 and easily large enough to produce waves that could damage India and Sri Lanka. The Harvard technique used not just the size of seismic waves but also their varying shapes, as recorded at varying distances and directions from the rupture. That extra information enabled the computer to gauge the true size of the fault rupture and thus the true magnitude of the quake, known as a CMT magnitude.

Ekström, then on vacation and away from his lab, logged in to the computer remotely after happening on an NEIC alert while checking his e-mail. Four-and-a-half hours after the quake, he and Harvard colleague Meredith Nettles e-mailed a recalculated magnitude to NEIC and PTWC. That was after India and Sri Lanka were hit but before the tsunami reached East Africa, where it killed more than 100 people.

If the Sumatran quake—which might recur once a millennium—had struck a year later, Ekström says, the world could have marked it as a killer more than an hour before it struck India. By then, under a USGS grant issued before the quake, NEIC will be receiving Harvard's automatic CMT analysis in real time 24/7. And a little fine-tuning can accelerate such real-time magnitude estimates to within three-quarters or even half an hour after a quake, says Ekström.

In the end, scientists did not have the fastest, most accurate warning tool at hand because no one had fully grasped the need. “We've known there was a probleme” off Sumatra, says Bilham, but “I'm surprised out of my wits about the magnitude of it.” It's clear now, he says, that “seismologists have to grapple with absolutely worst case scenarios.”

15. AMERICAN GEOPHYSICAL UNION MEETING

# A Lively Core Turns Mercury Into an Enormous Electromagnet

1. Richard A. Kerr

SAN FRANCISCO, CALIFORNIA—More than 11,500 earth scientists from around the world gathered 13 to 17 December at the fall American Geophysical Union meeting to discuss everything from Mercury's core to the rings of Saturn.

By all rights, the Mariner 10 spacecraft should have found a geophysically dead planet when it flew by Mercury in the mid-1970s. But to everyone's surprise, Mariner detected a weak magnetic field emanating from the sun's closest companion. A still-molten iron core churns out Earth's field, but Mercury's field seemed too weak to be generated that way. And besides, planetary scientists thought Mercury's big iron core must have frozen solid eons ago. Alternatively, if an early field-generating core had locked its field into Mercury's crust before freezing up, the field would be much stronger than Mariner's discovery. No spacecraft has revisited Mercury, but at the meeting, two groups of researchers built a strong case that Mercury generates its magnetic field in a lingering remnant of a molten core, much the way Earth's geodynamo operates.

The trick to diagnosing Mercury's interior without leaving Earth was measuring the planet's rotation rate to 1 part in 100,000. A combination of asymmetries links Mercury's interior to its rotation, as planetary scientist Jean-Luc Margot of Cornell University explained in his presentation. Mercury itself is slightly egg-shaped rather than spherical, so the sun's gravitational pull tends to align a bulge of the planet sunward. But Mercury's orbit is elliptical, not circular, so the planet's orbital motion tends to drag it out of its sun-induced alignment. The sun then tugs Mercury back toward alignment, ever so slightly slowing its rotation rate. Further along in the planet's orbit, the sun speeds up the rotation rate.

The amplitude of this rotational slowing and speeding up, or libration, depends on how much of the planet the sun must tug on. If even just the outer core is molten, that would disconnect the interior from the rocky outer shell, greatly reducing the mass that must be realigned and increasing the amplitude of Mercury's libration to at least double that of an entirely solid body.

Margot and his colleagues used a previously proposed ground-based radar technique to precisely measure variations in Mercury's rotation during the past 2 years. They repeatedly beamed a radar pulse at Mercury from the 70-meter antenna at Goldstone, California, and picked up the reflected signal at both Goldstone and the 100-meter antenna at Greenbank, West Virginia, 3200 kilometers to the east. Matching up the distinctively “speckled” pattern in the signal received at each station, they gauged the time lag of reception between stations and thus calculated the rotation rate precisely. It varied with Mercury's 88-day libration three times as much as it would if the planet were solid throughout.

Given such a definitive result, “it looks as if [a molten core] is the only explanation,” says planetary geophysicist David Smith of NASA's Goddard Space Flight Center in Greenbelt, Maryland. That still would leave the difficulty of why Mercury's magnetic field has only 1/100 the strength of Earth's geodynamo-generated field.

In a poster presentation at the meeting, planetary geophysicist Sabine Stanley of the Massachusetts Institute of Technology and her colleagues showed how the Mariner measurements could be misleading. They ran a computer model developed to simulate the geodynamo churning in the molten outer core of Earth, between a rocky mantle above and a solid-iron inner core within. On the assumption that Mercury's molten outer core had shrunk to a thin shell by now, they ran the model with progressively thinner outer cores. The model's dynamo continued to generate a relatively strong field within the core, but the field that it could project outside the core weakened to the point that a passing spacecraft would detect a very weak field even while a strong field dominated the core.

The Messenger spacecraft, launched last August, should be able to test the state of Mercury's core and the nature of its magnetic field after entering orbit in 2011.

16. AMERICAN GEOPHYSICAL UNION MEETING

# What's Going On in Saturn's E Ring?

1. Richard A. Kerr

SAN FRANCISCO, CALIFORNIA—More than 11,500 earth scientists from around the world gathered 13 to 17 December at the fall American Geophysical Union meeting to discuss everything from Mercury's core to the rings of Saturn.

Saturn's faint, broad E ring encircles the planet beyond the main rings with no visible means of support; no one ever has figured out what it's doing there. And no one can figure out what it was up to late last winter, either, when it apparently spewed out a cloud of water equal to its own mass. Whatever created the E ring in the first place—collisions of stealth moonlets or eruptions of icy volcanoes on the moon Enceladus, perhaps—may be responsible.

The E ring outburst came just as the Cassini spacecraft approached Saturn, carrying its Ultraviolet Imaging Spectrograph (UVIS), an instrument well suited to map out the glow of oxygen atoms near Saturn. At the meeting, UVIS principal investigator Larry W. Esposito of the University of Colorado, Boulder, and Donald Shemansky of the University of Southern California in Los Angeles described how 500,000 tons of oxygen atoms appeared during 2 months as an ultraviolet glow in the UVIS images. The oxygen formed a doughnut-shaped ring engulfing the E ring, then faded just as rapidly, leaving Saturn's magnetosphere depleted of ions.

That sequence of events suggests to Esposito and Shemansky that half a million tons of water ice crystals were suddenly added to the E ring, which already contained an equal mass of 1-micrometer ice particles. Colliding energetic ions would have knocked oxygen atoms free of the newly released ice. The resulting neutral oxygen atoms could then pick up charge from magnetospheric ions and eventually be ejected from the saturnian system, leaving the E ring much as it was.

What could have injected that much ice into the E ring so suddenly? Esposito favors a catastrophic collision of two unseen icy bodies orbiting in the E ring. Such embedded moonlets sustain the faint ring of Jupiter, but they do it through continual erosion by impacting micrometeoroids, not by collisions among themselves. It would take an improbable coincidence or a great many embedded moonlets to explain a major collision just as Cassini approached. Ring specialist Joseph Burns of Cornell University doubts that there are enough E ring moonlets. A Cassini camera search for such bodies larger than 1 to 2 kilometers in diameter is 95% complete, he says, but none has been found.

Alternatively, the water might have been blasted off the moon Enceladus in a volcanic eruption. But the two Voyager spacecraft found no signs of ongoing eruption there in the early 1980s, although they did find plains that might have been smoothed by geologically recent watery volcanism. “You've got several bad alternatives,” says Burns. Puzzled ring scientists hope that three Cassini close flybys of Enceladus this year, the first on 17 February, will improve their choices.

17. AMERICAN GEOPHYSICAL UNION MEETING

# Scary Arctic Ice Loss? Blame the Wind

1. Richard A. Kerr

SAN FRANCISCO, CALIFORNIA—More than 11,500 earth scientists from around the world gathered 13 to 17 December at the fall American Geophysical Union meeting to discuss everything from Mercury's core to the rings of Saturn.

The past three Septembers have seen the Arctic ice pack shrink dramatically to a record low amid signs that greenhouse warming could be melting the ice, threatening to clear the Arctic Ocean within decades. Researchers are still worried, but a study presented at the meeting offers some reassurance. A natural, temporary shift in the wind may have been largely to blame for the recent shrinkage.

Winds of the high northern latitudes are the domain of the Arctic Oscillation (AO), an erratic atmospheric pressure seesaw (Science, 9 April 1999, p. 241). Over weeks, years, or even decades, pressure can fall over the pole while rising around a circle near the latitude of Alaska. The resulting steeper pressure drop across high latitudes increases the generally westerly winds blowing there. When the pressure seesaws the other way, the winds drop to weaker than average.

Wondering how the AO had been influencing Arctic ice, meteorologists Ignatius Rigor and J. Michael Wallace of the University of Washington, Seattle, created a model that keeps track of ice as it forms and blows around the Arctic Ocean, thickening with time. In the 1980s, the AO was in its so-called low-index phase, with higher than average pressure over the pole and therefore weaker westerly winds. In the model, those winds tended to drive the ice around in circles off the Alaskan and Siberian coasts, giving it a chance to thicken for an average of 10 years or more. But in the 1990s, the AO swung into its strong-wind phase. In the model, the new circulation tended to blow old, thick ice out of the Arctic Ocean through the Fram Strait and into the North Atlantic. The remaining ice was thinner than under the opposite AO phase and thus easier to melt away. In fact, ice did surge through Fram Strait in the early 1990s, and the ice has thinned, culminating in the record low ice extents of recent years.

At least some of the recent ice loss is indeed “a hangover effect” of the early '90s swing in the AO, says meteorologist Mark Serreze of the University of Colorado, Boulder. The AO index fell back toward more normal levels in the late '90s, he notes, but the ice hasn't recovered yet. Because Arctic warming has been lengthening the period in the summer during which ice can melt, he says, Arctic ice may well continue to shrink, although probably not as rapidly as it did recently.

In the long term, Serreze adds, climate models predict that greenhouse warming should lead to increased melting over coming decades. Some models even have the intensifying greenhouse pushing the AO into a permanent positive phase, he says, which would favor still-greater ice losses.

18. AMERICAN GEOPHYSICAL UNION MEETING

# Snapshots From the Meeting

1. Richard A. Kerr

No vestige of a beginning. Seismologists got their most detailed look at an earthquake last fall when 30 kilometers of the San Andreas fault ruptured through the town of Parkfield, California, and its dense array of instruments, but they still missed something. “This is the best data we've got,” said geophysicist Malcolm Johnston of the U.S. Geological Survey in Menlo Park, California, but there is still no sign of the slow, hesitant onset of the fault rupture that some seismologists have been looking for (Science, 6 January 1995, p. 28). If earthquakes were to begin as slow slippage on a small patch of fault, well-placed instruments might detect it days or even weeks before the slippage took off and produced a quake. But the Parkfield data limit any such nucleation patch to a few tens of meters or less in size, says Johnston. So, even if nucleation occurs, detecting it looks improbable.

A nudge toward magnetic flip-flop. Two paleomagnetists found themselves presenting adjacent posters that argued for a previously unrecognized precursor to the most recent reversal of Earth's magnetic field. Researchers had thought that the field generated by the churning molten iron of the outer core had simply weakened and reorganized itself for a few thousand years as it got ready to flip about 775,000 years ago. Not so fast, say Laurie Brown of the University of Massachusetts, Amherst, and Bradley Singer of the University of Wisconsin, Madison. Brown, working on the paleomagnetic record frozen into lavas of central Chile, and Singer, studying lavas in Tahiti, found that the field had actually weakened and moved toward a reversal 18,000 years earlier. The prolonged precursory move toward reversal may have given the liquid outer core time to overcome the stabilizing influence of the solid inner core.