# News this Week

Science  11 Oct 2002:
Vol. 298, Issue 5592, pp. 338
1. SCIENTIFIC EVIDENCE

# Judge Rejects Cancer Data in Maryland Cell Phone Suit

1. Mark Parascandola*
1. Mark Parascandola is a writer in Washington, D.C.

A Baltimore federal judge has effectively scuttled the arguments behind an $800 million lawsuit brought by a man who claims he got a brain tumor from using a cell phone. Last week's ruling by U.S. District Judge Catherine Blake is seen as a major setback for those claiming that cell phones can damage health. The 23-page ruling, the most substantial court review of the cell phone issue to date, provides a textbook example of how federal courts are now taking a closer look at the quality of scientific evidence. It concluded that lawyers for Christopher Newman, the Baltimore neurologist who brought the lawsuit, had provided “no sufficiently reliable and relevant scientific evidence” to support the claim that using an analog mobile phone for 6 years caused his tumor. The judge served notice that she intends to dismiss Newman's claims if she receives no objections before 30 October. Newman's suit became a leading test of how the federal courts would respond to scientific arguments that underlie pending legal actions against cell phone makers. Newman's legal team, led by the Baltimore firm of Peter Angelos, hoped to establish that Motorola Inc., a group of local wireless service providers, and the Cellular Telecommunications and Internet Association had sold defective and dangerous products without informing consumers about the risks. The lawyers planned to call epidemiologists and experts on the biological effects of radio frequency emissions to support their allegations (Science, 16 November 2001, p. 1440). But a year ago, several industry defendants in the case filed a motion to block Newman's scientific witnesses from testifying. They cited the landmark 1993 U.S. Supreme Court decision, Daubert v. Merrell Dow Pharmaceuticals Inc., which directs judges to determine whether experts' reasoning is both relevant and scientifically valid before allowing them to appear in a trial. Blake probed the scientific experts during an intense 5-day hearing in February. But it was Newman's experts who were in the hot seat because the burden of proof rests with the party bringing the lawsuit. Newman's argument relied heavily on the testimony of Swedish oncologist Lennart Hardell of Örebro Medical Center. Hardell is the only scientist to date to report in an epidemiologic study a positive association between cell phone use and brain tumor risk. In his study, analog (but not digital) cell phone users were 26% more likely to have a brain tumor than nonusers. In contrast, five other epidemiologic studies have failed to find any increased risk. Industry witnesses said Hardell's conclusions were based on a distortion of the data. Meir Stampfer, chair of the department of epidemiology at the Harvard School of Public Health in Boston, pointed out that the difference in risk appeared only when acoustic neurinomas, a benign type of tumor, were included in the analysis. No increase in risk was found for cancerous tumors. The possibility remains, defense experts admitted, that a larger study over a longer time period might detect a health risk that has so far eluded researchers. But judges cannot wait for all the evidence before reaching a verdict—and Blake did not find Newman's experts persuasive. Daubert also instructs judges to consider whether the scientific conclusions at issue have been subjected to peer review. In this instance, the judge actually reviewed the confidential notes of peer reviewers after Hardell admitted in a deposition that The Lancet had rejected a paper from his study. Defense attorney Thomas Watson, who previously represented biologists Robert Gallo and Thereza Imanishi-Kari against charges of scientific misconduct, demanded to see the anonymous reviewers' reports, which the court ordered Hardell to turn over. Although the reviewers' comments themselves remain sealed by the court, the judge quoted from the editors' correspondence with Hardell in her opinion. A Lancet editor told Hardell that reviewers had said his conclusions were written “much too forcefully.” The paper was eventually accepted by the European Journal of Cancer Prevention and published in June 2002. Another paper based on the same study is slated to appear in The International Journal of Radiation Biology, although the editor cautioned Hardell in his letter of acceptance that “never before have we accepted a paper in the face of such low scores by referees.” Hardell could not be reached for comment. Newman's attorneys plan to challenge Blake's ruling but had no comment on her decision. Meanwhile, several similar cases introduced around the country have been transferred to her courtroom in recent months. Norman Sandler, a spokesperson for Motorola, says that the company and its lawyers will file a motion to have these dismissed on the same grounds as the Newman case. “This particular decision sends a pretty strong message,” he says. Although Blake's ruling on the scientific facts in the Newman case does not bind other courts, other judges will likely be influenced by her thorough review of the evidence. Michael Green, an expert on product liability law and a professor at Wake Forest University School of Law in Winston-Salem, North Carolina, explains that when different judges are considering the same body of scientific information, “decisions like this do get paid attention to.” 2. BACTERIAL MENINGITIS # Appeal to Thwart Deadly Outbreak 1. Gretchen Vogel Fearing a repeat appearance of a rare strain of bacterial meningitis that killed 1500 people last year in Africa, the World Health Organization (WHO) last week issued an urgent plea to drugmakers to make an effective vaccine available at reduced cost. The most common forms of meningitis can be prevented with inexpensive, readily available vaccines, but the only vaccine that protects against strain W135 costs$4 to $40 a dose—far more than most African countries can afford. At last week's emergency session in Burkina Faso, officials were pessimistic about staving off another epidemic of W135 without help from the world's largest drug companies. If a major outbreak of W135 were to strike again, WHO says, there would not be enough vaccine doses available at any price. Outbreaks of bacterial meningitis plague Africa's so-called Meningitis Belt every winter as dry winds blow south from the Sahara, spawning dust storms that irritate the lungs and leave people vulnerable to the airborne bacteria. Last year, W135 hit with a vengeance, killing 1500 people in Burkina Faso. Most of the victims were young children. The death toll tells only part of the story: One in 10 survivors suffers brain damage, deafness, or limb amputation. It's impossible to predict whether the rare W135 strain will return this winter, says WHO spokesperson Iain Simpson, but if it does, the organization doesn't want to be caught empty-handed. The only vaccine available against W135 is a tetravalent preparation that also staves off three other strains. But at a cost of at least$4 per dose, the vaccine—which is in routine use for military recruits and travelers from wealthy countries—is out of reach of vulnerable countries in Africa, which can afford to pay at most about $1 per dose, according to WHO. Even if the price were slashed, there would not be enough of the tetravalent vaccine to respond to a large epidemic. In a severe outbreak, some 3 million doses would be needed; at the meeting, drug companies estimated that only 2 million doses are available, says Simpson. “If there are one or two small outbreaks, that might be enough. If there is one large outbreak, we wouldn't even be close.” On 4 October, following negotiations with WHO and other officials, the drug giant GlaxoSmithKline offered to produce a new triple-action vaccine against W135 and two common strains, A and C, for$1 per dose. But there might not be enough time to get the new vaccine ready for winter. It would have to receive regulatory approval in Belgium, where it would be manufactured, and in countries where it would be used. That could take months, says Daniel Tarantola, WHO's director of vaccine development. And financing is far from certain. “We do not want to overstate our hope,” he says.

WHO began talks with vaccine manufacturers in late spring, Tarantola says, but received little positive response until last month. Some observers question the strategy of vaccinating after an outbreak starts. Mass vaccination campaigns before an epidemic could save thousands of lives a year, asserts John Robbins, a vaccine expert at the National Institute of Child Health and Human Development in Bethesda, Maryland.

He and several colleagues have been pushing WHO to launch mass campaigns immediately to protect more people and drive down the tetravalent vaccine's cost. “There is a primary rule about vaccines: If you want to make it cheap, use it,” Robbins says. “No manufacturer will be able to store enough vaccine and deliver it within the time frame required to stop an epidemic.”

Tarantola and others at WHO say that such a strategy would work only with a conjugate vaccine that elicits a stronger and longer-lasting immune response against a range of strains. Such a vaccine is under development but won't be ready for several years, Tarantola says. In the meantime, health officials in Burkina Faso and the rest of the Meningitis Belt can only hope that they are luckier this year than last.

3. HERPETOLOGY

# 100 Frogs A-Leaping for Biodiversity

1. Elizabeth Pennisi

While many herpetologists have been scrambling to understand why frogs, toads, and other amphibians are declining worldwide, one research team has been cataloging more than 100 new species, all from one postage stamp of a rain forest in Sri Lanka. On page 379, Christopher Schneider, a systematist at Boston University, and his colleagues describe their discovery of this biodiversity hot spot, increasing the number of known frog species on the island of Sri Lanka fivefold. “It's quite amazing [that] they have found this incredible diversity,” comments David Wake, a herpetologist at the University of California, Berkeley.

In 1993, Rohan Pethiyagoda used his own money to found the Wildlife Heritage Trust in Colombo, Sri Lanka. He and his colleagues began a census of Sri Lanka's disappearing wildlife, systematically searching the 750 square kilometers of remaining rain forest, which once covered 15,000 square kilometers. To his surprise, he kept finding frogs he couldn't identify.

Pethiyagoda first showed the animals to frog systematists. They estimated there might be 200 new species, based on the morphology and other characteristics of 1000 specimens. Subsequent genetic studies reduced that number to 120 or so. Claims of a vast number of new species are often greeted with skepticism, but the new data are compelling, say the researchers' colleagues. This work, which combines traditional and molecular approaches, “is right at the forefront of what work in biodiversity should entail,” says Roy McDiarmid, a systematist at the U.S. Geological Survey who is based at the Smithsonian National Museum of Natural History in Washington, D.C. Adds Wake, “The molecular data gives a certain validation to the assertion that these things are really different species.”

The new species fall into two groups. One consists of just five species, all of which lay eggs in foam nests on leaves, rocks, or branches suspended above water. Once big enough to be out of danger from many aquatic predators, the hatched tadpoles slide off into the stream or pond below. But most of the newfound frogs are “direct developers” whose young never get their feet wet. These eggs incubate individually in leaf litter instead of foam nests, and they hatch as miniature adults, skipping the tadpole stage altogether. This water-free lifestyle “gives species a lot more latitude,” McDiarmid explains, and “lends itself to geographic isolation and speciation.”

Schneider thinks that these frogs have escaped the fate of other amphibians because disease, ultraviolet light, and other potentially deadly influences appear to be most dangerous to water-based young. “By skipping the aquatic [stage], they may bypass a life stage when they are most vulnerable,” he suggests.

But there seem to be some dangers even these direct developers have not escaped. Co-author Kelum Manamendra-Arachchi of the Wildlife Heritage Trust traveled to museums containing specimens similar to the newly discovered ones, looking to confirm his species designations. He found many frogs—perhaps 100 species—that had been collected from Sri Lanka more than 100 years ago that were not among their current finds. “It means a huge number of species must have gone extinct already,” says Wake, most likely because so much of the island's rain forest has been lost.

Therein lies the challenge for biodiversity's champions, says Paul Speck, president of the Arlington, Virginia-based Amphibian Conservation Alliance. “We're at a very precarious moment,” he says; “there are a lot of species still out there [and] there is the opportunity to save many of them, but we need to act quickly.”

4. TOXICOLOGY

# Protecting Liver From Painkiller's Lethal Dose

1. Jean Marx

Last week a committee of the U.S. Food and Drug Administration recommended that medicines containing acetaminophen, a commonly used painkiller sold under the brand name Tylenol and in many over-the-counter cold and flu remedies, carry stronger warnings about its dangerous side effects. Every year in the United States overdoses of the painkiller cause acute liver failure in as many as 800 people, one-third of whom die as a result. New results help explain just how acetaminophen harms the liver. They may also provide a target for treating liver failure due to overdoses of acetaminophen and perhaps of other drugs as well.

On page 422, a team led by David Moore of Baylor College of Medicine in Houston, Texas, reports that the so-called constitutive androstane receptor (CAR) plays a central role in bringing about the biochemical changes that underlie acetaminophen toxicity. Moore and his colleagues “have shown clearly that the receptor is important,” says Steven Kliewer, a liver toxicity expert at the University of Texas Southwestern Medical Center in Dallas.

The current work is an outgrowth of previous research by Moore's team and others showing that CAR helps the liver eliminate foreign chemicals. Ordinarily, Moore says, this system is protective, but in some cases it has the opposite effect, creating products that are more toxic than the originals.

A clue that CAR might be involved in acetaminophen toxicity came a few years ago. The drug phenobarbital increases susceptibility to acetaminophen damage because it fosters production of two enzymes in the CYP family that convert acetaminophen to a highly toxic compound called NAPQI. The cell normally eliminates NAPQI by tying it up with a detoxifying molecule called glutathione. But if NAPQI production outstrips the glutathione supply, cell damage occurs. “Everything is fine until you run out of glutathione,” Moore says. CAR comes into the story because Moore and other researchers found that phenobarbital works through that receptor to enhance CYP enzyme production.

Following up on that observation, Moore and his colleagues now report that CAR is involved even more directly in acetaminophen toxicity. In normal mice, high doses of acetaminophen increased production of the CYP enzymes and also an enzyme called GSTPi that attaches glutathione to NAPQI and other molecules. That could be a double whammy for liver cells, increasing NAPQI production at the same time glutathione supplies are depleted. Indeed, the animals showed signs of severe liver damage, such as patches of dying liver cells. In animals in which the CAR gene had been inactivated, the same doses of acetaminophen did not increase production of CYP or GSTPi enzymes. Although those animals suffered some liver damage, it was much less than that of the normal mice.

Inactivating the human counterpart of the CAR gene before an overdose isn't a likely therapeutic strategy. But Moore's team found that androstanol, a compound that inhibits CAR activity, protected mice against acetaminophen-induced liver damage even after exposure to the painkiller. It provided 100% protection 1 hour later and 50% protection 3 hours later.

Different species can vary significantly in their reactions to foreign chemicals. To determine whether the human CAR receptor responds to acetaminophen the same way the mouse version does, the Baylor group created a line of mice whose liver cells contain only human CAR. When these “humanized mice” were treated with phenobarbital or acetaminophen, their livers showed damage similar to that in normal mice, Moore says.

More work will be needed to show whether CAR inhibitors might be useful treatments for human liver toxicities, however. One problem is that androstanol does not inhibit the human CAR receptor, so researchers would have to develop new inhibitors that do. And any new treatment would also have to outperform the one currently available, a compound that replenishes the liver's glutathione stores; it works well if given within several hours of the toxic painkiller dose. Moore is looking at whether other liver toxins also work through CAR and might be candidates for treatment.

Kliewer, for one, is hopeful that work on CAR might lead to new therapies: “The more we understand about the mechanisms [of liver toxicity], the more opportunities we will have for treatment.”

5. PHYSICS

# Quantum Experiment Asks 'How Big Is Big?'

1. Charles Seife

Watch out, world: Erwin Schrödinger's infamous cat is straining at its leash. The cat—a seemingly ridiculous example of a familiar object ruled by quantum-mechanical laws—symbolizes the gulf between our world and the world of the very small. Now quantum physicists propose to bridge that gap by creating by far the biggest quantum- mechanical object ever constructed. If they succeed, the experiment might reveal whether the bizarre quantum antics of small things such as atoms and photons can crop up in large things such as cats and bricks—and if not, why not. “Basically, if this works, you've extended the validity of quantum mechanics by nine orders of magnitude,” says Max Tegmark, a theorist at the University of Pennsylvania in Philadelphia.

The work explores the concept of superposition, a quantum object's ability to be two opposite things at the same time, like a switch that is at once both on and off. Physicists have long wondered why superposition works with small objects such as photons but not with large ones such as cats. To find out, they have been trying to put increasingly large objects in superposition or teach them other quantum tricks. Anton Zeilinger of the University of Vienna, Austria, for instance, has shown that 70-atom buckyballs, enormous by quantum standards, can still act like quantum objects (Science, 25 May 2001, p. 1471).

Now Roger Penrose of Oxford University, U.K., Dik Bouwmeester of the University of California, Santa Barbara, and colleagues have designed a Schrödinger's cat billions of times larger than Zeilinger's record-holding molecules. “It was not clear whether it was feasible at all,” says Bouwmeester. “But the more calculations we did, the more it seemed feasible.”

The proposed experiment, described in a paper submitted to Physical Review Letters, starts with an interferometer. The device is normally used to split a light beam and then, using mirrors, to reroute the two halves so that their waves amplify or cancel one another. Physicists have found that, perversely, a single photon fired into an interferometer behaves the same way: On reaching the beam splitter, it shoots off in two directions at once—a clear case of superposition.

In the Penrose team's device (see graphic), the photon's two-pronged path leads to a pair of mirrored cavities, where the photon bounces around for a while before escaping whence it came. One cavity sports a mirror 10 micrometers wide—about the breadth of a red blood cell—perched at the tip of a cantilever tuned so that the mirror moves if a photon strikes it. That mirror is the “cat.” If subatomic particles played by common-sense classical rules, the photon would follow just one of the two paths, either striking the mirror and making it move, or hitting the other cavity and leaving the mirror stationary. But because the photon is in superposition, it both makes the mirror move and leaves it stationary at the same time.

En route back from the cavities, the photon is shunted into detectors that reveal whether the mirror did or did not move. By sending photon after photon through the device, the scientists can learn whether the mirror is in the “did move” state, the “did not move” state, or a superposition of the two, and they can measure how long the superposition lasts.

If the experiment actually puts a mirror into superposition, it will suggest that there is nothing fundamental about large things that makes them behave like classical objects rather than quantum ones. “If quantum mechanics hasn't gone wrong at the size of a cell,” Tegmark says, “it probably won't go wrong with something the size of a mouse or a human.” Bouwmeester's team is already testing components that will go into the lab setup, which requires very high vacuums, very cold temperatures, and very precise measuring equipment. Should the team overcome those formidable hurdles, Schrödinger's hypothetical tabby might become more than a pet notion.

6. SWITZERLAND

# Compromise Allows Transgenic Trials

1. Min Ku*
1. Min Ku is a writer in Bern, Switzerland.

BERN—Swiss biologists believe they have won a decisive victory in a 9-year battle with campaigners opposed to genetically modified (GM) organisms. After 11 hours of rancorous debate, the National Council, the lower house of the Swiss parliament, voted last week to accept watered-down legislation governing gene technology. To the relief of researchers, the council deleted measures that would restrict research and impose a 5-year moratorium on commercial release of GM organisms. The new version is “a good sign for young scientists to stay in the field,” says Heidi Diggelmann, president of the Swiss National Science Foundation's Research Council.

Anti-GM groups first aired the Gene- Protection Initiative in 1993 to ban all research on GM animals and patenting of any GM organism. Researchers went on the offensive, attempting to convey to the public the importance of their research. In 1996 a bill was put forward as a compromise, plugging gaps in existing law without aggressively curbing GM research. The initiative was decisively defeated in a national referendum (Science, 12 June 1998, p. 1685).

However, under pressure from anti-GM groups, the bill was modified to include provisions such as limiting releases of GM organisms to biosafety experiments that could be guaranteed to be 100% risk free and could not be performed with conventional means. These were poison pills to many scientists, who argued that the terms would amount to a de facto ban on basic transgenic research. The proposed legislation would also have banned antibiotic resistance genes in deliberately released GM organisms and would have held producers of GM products liable for damages for up to 30 years, even for nondefective, grossly misused products.

Some members of parliament rallied against these measures, and last-minute negotiations succeeded in defanging the bill. Under the approved measures, there will be no commercial moratorium, and restrictions on the release of GM organisms for research will be eased. In addition, antibiotic resistance genes in released GM organisms will be allowed until 2008, and in general, producers of GM products will be liable only for defective products. The legislation still has to go to the upper chamber of parliament for final revisions, but a moratorium looks unlikely.

Although many scientists are satisfied with the outcome, they are not dancing in the streets. Gottfried Schatz, president of the Swiss Science and Technology Council, a government advisory group, says that the approved measures were “the best that could be hoped for given the current climate of mistrust.” Daniel Ammann, director of the Swiss Gene-Technology Working Group, an anti-GMumbrella organization, confirms that the group plans to continue its fight by seeking a ban on any commercial release of GM organisms.

The modifications dispel a cloud of uncertainty that had been hanging over Switzerland's first-ever field trial of GM plants. On 13 September, Moritz Leuenberger, head of Switzerland's Department of Environment, Transport, Energy, and Communications, approved the trial of wheat engineered to make a protein toxic to a crop pest called stinking smut fungus. Regulatory authorities had rejected the proposed experiment last year on the grounds that the transgenic plants contain an antibiotic resistance gene (Science, 7 December 2001, p. 2067). Leuenberger approved the trial “solely for legal reasons,” as current laws do not require such an experiment to exclude all possible safety risks. Final approval is expected in a few months.

7. ASTRONOMY

# New Results Reawaken Quasar Distance Dispute

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

Two groups of researchers last week sought to deliver separate killer blows to a controversial theory about quasars. Instead, they seem to have simply stirred up the same old hornets' nest. “It's hard for me to believe that people on either side of the debate will ever admit they're wrong,” says Jack Sulentic of the University of Alabama, Tuscaloosa.

Conventional wisdom says that quasars are the extremely luminous cores of very distant galaxies. A quasar's distance is derived from the redshift in its spectrum, which is thought to be caused by the expansion of the universe. But a small group of prominent astronomers, including the late Fred Hoyle and Margaret Burbidge of the University of California, San Diego (UCSD), believes that quasars are bits of star stuff ejected by quite nearby active galaxies and that at least part of their redshift has an intrinsic cause.

The evidence? Quasars are much more numerous around these nearby active galaxies, and quite often they seem to be connected to them by luminous bridges and filaments. Moreover, studies carried out in the 1990s indicate that the redshifts of these quasars cluster around certain periodic values. All this is impossible to explain in the conventional view of big bang cosmology.

But now, astronomers at the University of Nottingham in the U.K. report in the current Monthly Notices of the Royal Astronomical Society (MNRAS) that they can find no evidence for redshift periodicities in a much larger sample of purported quasar-galaxy pairs taken from the recently completed Two Degree Field (2dF) survey. “Our plot shows there's nothing there,” says team member Edward Hawkins.

End of controversy? Hardly. Geoffrey Burbidge, Margaret Burbidge's husband and fellow UCSD astronomer, says the article “is a real piece of dishonesty.” According to him, his collaborator Bill Napier of the Armagh Observatory in Northern Ireland had already pointed out a serious statistical flaw in the analysis before it was submitted for publication. Napier, who is “quite open-minded” about the interpretation, believes the periodicity is there. “Further analysis … would seem to be called for,” he and Burbidge write in a rebuttal paper submitted to MNRAS. However, Hawkins counters that there was no need to change anything in the paper. “We got a very favorable referee report,” he says.

Meanwhile, the Space Telescope Science Institute (STScI) in Baltimore, Maryland, issued a Hubble photo last week of the galaxy-quasar pair NGC 4319 and Mrk 205, which have very different redshifts. Almost 20 years ago, Halton Arp of the Max Planck Institute for Astrophysics in Garching, Germany, working with his then-grad student Sulentic, showed that these two objects appear to be linked by a luminous “bridge,” suggesting a physical connection. The new Hubble picture shows no obvious bridge, and the STScI Web site describes its existence as “debatable.”

Arp accuses STScI of “deliberately misleading the public.” And this week the University of Alabama released an enhanced version of this same Hubble image, produced by Sulentic, that clearly shows a bridge. STScI astronomer Keith Noll, who helped create the Hubble image and the Web site, says there might be a “bridge” between the galaxy and the quasar, but there's no evidence that it is a real physical connection. “We are giving [Arp's] ideas more credence than they really deserve,” says Noll.

Arp still insists that sensitive spectroscopic observations with the new generation of 8-meter telescopes will be able to resolve the issue. But he says it's very hard to get access to a large observatory to collect the necessary data. Geoffrey Burbidge calls it a “sociological problem associated with the need to believe” that redshifts are related to distances. Clearly, neither side has spoken its last words on the subject.

8. YOUNG INVESTIGATORS

# European Program to Fund the Best

1. Erica Goldman

European research officials are hatching a bold scheme to attract young scientists—and preparing to put big bucks behind it. The talent competition, part of a growing move to bolster European science across national boundaries, will offer grants totaling roughly $1.5 million over 5 years. The contest will be open to researchers in any discipline—provided that they agree to work in one of the participating countries. Only the best and the brightest from around the world need apply for the new European Young Investigators Awards. “The award [program] will be the first of its kind, guided by scientific excellence, originality, and potential,” says Finland's Lea Ryynänen-Karjalainen, who serves as scientific secretary to the European Union's Heads of Research Councils, the organization introducing the program. Research councils in the 15 European Union countries (and three associate members) will meet later this month in Athens, Greece, to hash out the details. To launch the program in January, officials say they need a minimum of five participating countries, representing at least 100 million people. “I think it will go forward,” says Heidi Diggelmann, president of the Swiss National Science Foundation's Research Council in Bern. “Even if we can't launch it on a full scale, there are enough countries on board to get started.” The competition will be open to scientists from anywhere in the world, but the winners must go to work in a participating country. Scientists from noncontributing countries will apply through a member organization. Finalists at the national level will be evaluated by disciplinary panels managed by the European Science Foundation in Strasbourg, France. Applicants must be under the age of 35 or have completed a postdoctoral appointment within the last 2 to 5 years, with consideration given for career breaks such as maternity leave, explains Ryynänen-Karjalainen. The awards will cover salary, overhead, and personnel. Participating countries hope to fund between 30 to 50 awards in the first round, with the number rising as more countries join in. But there are no guarantees that a country that contributes to the pot will have even one winning applicant working at any of its institutions. “There will be no quota of ‘just return,’” explains Christoph Mühlberg of the German Research Council (DFG) in Bonn. 9. MICROBIOLOGY # The Science of Pfiesteria: Elusive, Subtle, and Toxic 1. Jocelyn Kaiser Ten years after ecologists in North Carolina found evidence that a toxic microbe caused mass fish die-offs, the toxin remains unidentified and the research is being challenged ## The Science of Pfiesteria: Elusive, Subtle, and Toxic Ever since it was blamed for massive fish kills in estuaries in the eastern United States a decade ago, a tiny one-celled organism known as Pfiesteria piscicida has fascinated and horrified the public. This dinoflagellate has been accused of slaying more than a billion fish and, in a superaggressive mode, releasing a potent neurotoxin that has sickened fishers and lab workers. It is also at the center of a raging scientific debate. Pfiesteria has adorned journal covers, been discussed in Congress, and even inspired a popular scientific thriller about its co-discoverer JoAnn Burkholder, an aquatic ecologist at North Carolina State University (NCSU) in Raleigh. It was Burkholder and her assistant Howard Glasgow who first described the microbe's bizarre “phantomlike” behavior—how it lies dormant in sediments, then changes form and “ambushes” live fish with toxin. But almost every aspect of this research is now disputed. Criticism of Burkholder's work reached a crescendo this summer. The Pfiesteria toxin has yet to be identified. And in three papers, scientists attempting to replicate her results found no evidence that Pfiesteria has a complex, 24-stage life cycle—or that it produces a deadly toxin. They agree that Pfiesteria kills fish, but their strain killed merely by feeding on them. Burkholder says the recent skeptics have shown only that they don't know how to follow her published protocols for culturing toxic Pfiesteria. “[Critics are] saying it's a hoax based on these three papers, which all come from a few nontoxic strains,” she says. One reason the debate has grown contentious is that Burkholder has been reluctant to share her cultures, citing costs and other concerns. “We have had so many problems with discrediting efforts and industry threats,” she says. Burkholder has churned the waters, too, by challenging grant awards and firing off long critiques of rivals when criticized at meetings and in the press. One group that sympathizes with Burkholder gave the controversy a new spin in August. An environmental group filed Freedom of Information Act (FOIA) requests for all the records of Pfiesteria researchers who have received more than$16 million in federal support since 1997, arguing that because some groups promised but failed to produce a toxin, they misused the funds. Donald Anderson of the Woods Hole Oceanographic Institution in Massachusetts laments the “National Enquirer” tone of the debate: “It's been very damaging to our field.”

Anderson is one of several prominent scientists who defend Burkholder and say some of her critics are drawing conclusions in the media that go beyond their data. Another is Donald Boesch of the University of Maryland (UMD) Center for Environmental Science, who says: “She's very controversial. She annoys some people. But her hypothesis really hasn't been significantly challenged by these papers.” Oceanographer Ted Smayda of the University of Rhode Island, Narragansett, blames the “anti-Burkholder camp” for upping the ante.

Burkholder and her collaborators are confident that she will be vindicated. They suggest that key results will be presented at a harmful algae meeting later this month—including, perhaps, proof of the elusive toxin. But convincing the scientific community that this toxin is real and comes from Pfiesteria will require that Burkholder share her materials more widely. Given the bitter history of Pfiesteria research, that possibility seems remote.

## Something in the water

The Pfiesteria saga began 1988, when scientists found that tilapia added to an aquarium at the NCSU veterinary school kept dying. Fish biologist Edward Noga and Burkholder, then a new professor, identified the cause as a new dinoflagellate from North Carolina estuarine water that could live on either algae or fish. They later learned that Pfiesteria blooms had appeared in sync with some fish kills in North Carolina. And under the microscope, Burkholder found, Pfiesteria was a fascinating shape-shifter, assuming many guises, including large amoeba-like blobs and various cysts, as well as dinospores.

The NCSU group published its findings in a 30 July 1992 letter to Nature, “New ‘phantom’ dinoflagellate is the causative agent of major estuarine fish kills.” At the time, “we were discovering a lot of new and unique dinoflagellates and algal blooms,” says Wayne Carmichael of Wright State University in Ohio, an algal toxin expert not involved in these studies. Pfiesteria, he adds, “was an interesting and intriguing possibility.”

It was also controversial. Burkholder soon found herself battling state officials and scientists who questioned whether the microbe had any role in fish kills and whether, as Burkholder also claimed, manure runoff from a booming hog industry stimulated Pfiesteria blooms. Then in 1993, a dramatic incident gave new weight to Burkholder's studies. Glasgow, her research assistant, reported mood swings, difficulty in reading, and severe memory loss after working for months in the presence of aerosols from Pfiesteria fish tanks. Several other workers, including Burkholder, experienced problems they attributed to Pfiesteria. In response, NCSU and federal officials mandated that Pfiesteria could be studied only using biosafety level 3 (BSL-3) precautions—the same as for HIV—and the state closed a river where Pfiesteria was found at the site of fish kills. These events are described in a 1997 book, And the Waters Turned to Blood by Rodney Barker, which Burkholder considers generally accurate.

## Pfiesteria fever

Pfiesteria hit the national radar screen in 1997. Maryland state scientists linked Pfiesteria to several fish kills in the Chesapeake Bay estuary that, because people stopped buying seafood, cost Maryland's seafood industry $50 million. Thirteen watermen working near the site of the kills reported memory loss, sores, and other symptoms. Also that year, at a meeting at the National Institute of Environmental Health Sciences (NIEHS) in Research Triangle Park, North Carolina, Burkholder announced that her team had “isolated and purified” a water-soluble toxin, according to an NCSU press release. (Burkholder now says the term “purified” was “not correct.”) At the same event, biochemist Daniel Baden of the University of Miami in Florida described a substance he had purified from Noga's fish tanks, a fat-soluble toxin. Congress subsequently appropriated new money for Pfiesteria research partly through ECOHAB, an interagency grants program, and NIEHS kicked in more. And Burkholder received the 1997 Scientific Freedom and Responsibility Award from the American Association for the Advancement of Science (publisher of Science). As public concern escalated, the governor of North Carolina convened a 14-member panel of North American experts on toxic algae that identified gaps in Pfiesteria science. High on the to-do list, they said, was distributing Pfiesteria cultures through the Bigelow Laboratory, a repository in West Boothbay Harbor, Maine. The lab received$275,000 from ECOHAB but never got cultures from Burkholder's group.

The ECOHAB program funded only part of a large proposal submitted by Burkholder. But she received $1.5 million from the North Carolina legislature and a foundation to build a new lab and a large biosafe facility, where she began generating large quantities of Pfiesteria. A collaborating group led by chemists John Ramsdell and Peter Moeller of the National Oceanic and Atmospheric Administration's (NOAA's) lab in Charleston, South Carolina, was supposed to isolate the toxin from her cultures. Other labs set out to pursue various studies; because Burkholder's cultures were not available, they attempted to replicate her experiments following her protocols. Researchers took heart from one technical advance at this time: UMD School of Medicine researcher David Oldach and Parke Rublee of the University of North Carolina (UNC), Greensboro, developed a molecular probe for Pfiesteria, which previously could be identified only by microscope. But in other ways the science got murkier. Scientists realized that Pfiesteria-laden waters contained other, similar-looking species. One of these, dubbed Pfiesteria shumwayae, was also sometimes toxic, Burkholder reported. Fish biologists suggested that low oxygen could explain many fish kills and that the deep sores typically found on affected fish could have many causes, such as a fungus that is usually found in the sores. Efforts to isolate the toxin dragged on. Lacking any clear biomarkers for human exposure, the U.S. Centers for Disease Control and Prevention (CDC) gave Pfiesteria-suspected illness the nebulous title “possible estuary-associated syndrome” and launched a cohort study. At a CDC workshop in 2000, an expert panel found “firm evidence …that Pfiesteria is toxic to fish” but concluded that the “extent and nature of the hazard to human health … remain unknown.” ## Toxin-free? This summer, several of Burkholder's critics launched a blitz: Accompanied by a flurry of press releases, they published three papers at odds with Burkholder's central claims about Pfiesteria. In the Journal of Phycology, a separate NOAA team using molecular techniques found that P. piscicida had a normal dinoflagellate life cycle (see sidebar on p. 348). Another team, led by fish pathologist Wolfgang Vogelbein of the Virginia Institute of Marine Science (VIMS) in Gloucester Point, reported in the 5 August Nature that P. shumwayae can attack and kill fish apparently without making a toxin. Attempting to follow Burkholder's procedures, they kept dinos and 25 to 40 tilapia in a fish tank for several weeks until fish began dying, then added fish daily to replace dead ones. Then, to study fish death in a simpler system, they tested how quickly water from their fish-killing tanks could kill larval minnows in small dishes. The team reported that the fish larvae died at rates comparable to those in Burkholder's toxic cultures—in 24 to 48 hours. But water with dino cells removed didn't kill fish. In addition, when fish were separated from Pfiesteria cells with a membrane, they didn't die, suggesting that they succumbed only when Pfiesteria made physical contact. The VIMS group also produced videos of the dinos aggressively feeding on the tilapia. Vogelbein thinks that P. shumwayae may be lethal to fish only in the lab. In a river, he says, the fish would presumably swim away. Says Vogelbein: “I don't know if anything like this occurs in the field.” In another paper in the 5 August online Proceedings of the National Academy of Sciences (PNAS), an overlapping team led by chemist Robert Gawley of the University of Miami suggested that Pfiesteria lacks the genes to make the polyketide toxins that are typically produced by fish-killing dinoflagellates. This paper suggests that other factors caused the fish kills, such as a different toxic alga. Burkholder says the VIMS researchers are grasping at straws because they failed to culture toxic Pfiesteria. The time to fish death is the givewaway, she says: Vogelbein misread her papers, which say that in her studies, it takes “minutes to hours” rather than a day or two. (She also says, however, that only in the early 1990s did she have strains “hot” enough to kill tilapia in minutes.) Burkholder says the VIMS team ignored her protocols specifying limits for ammonia and pH levels in their fish tanks, and they didn't expose dino cultures to fish long enough to set up a toxic attack. She says her group has reported that Pfiesteria attaches to and feeds on fish. And in her published studies, she adds, fish separated from toxic cells by a membrane still die. Burkholder is pleased that another group at Old Dominion University in Norfolk, Virginia, has recently confirmed that Pfiesteria produces a toxin. In the March issue of the new journal Harmful Algae, a team led by Andrew Gordon and Harold Marshall—who received cultures from Burkholder—report that extracts of water from tanks of P. piscicida and P. shumwayae do indeed kill fish, although more slowly than in direct contact. Gordon says it's important to follow Burkholder's protocol closely. It took his group a year to learn how to rapidly restock tanks with fish, making them “hotter” and more toxic. At first, “we were ready to give up and say there was no soluble toxin,” he says. Vogelbein says his group can get Pfiesteria to kill fish in 45 minutes if there are enough dinoflagellates in the tanks. What Burkholder calls “hot” tanks, he suggests, could be “entirely a cell density effect.” But he adds that “It's entirely possible that there are strains that produce toxins.” The only way to know for sure, he says, is for Burkholder to share her cultures. ## An unsharing culture But that's a touchy subject. Burkholder feels strongly that she cannot just give cultures away, mainly because of the cost. She insists on training to make sure other labs grow cultures the right way, safely, and don't wrongly discredit her. Generating culture plus “proper training” costs$40,000 per lab, she says. Nevertheless, using largely her own funding, she says she has given 41 researchers in 33 labs toxic cultures and training.

However, most of these groups received the dinos growing on algae or as buffered cells. Only one of these labs—Gordon and Marshall's group at Old Dominion—are replicating Burkholder's experiments by actively culturing “toxic” Pfiesteria in their own fish tanks. Like other Burkholder collaborators, they have agreed that they won't share cultures with other groups. “It puts me in a position that I'm kind of uncomfortable with,” says Gordon, who notes that sharing is expected in his field, microbiology. But, he says, “the ethics seem to be different in the world of [toxic] marine plankton.”

Plans to distribute cultures through the Bigelow Lab in Maine fell apart for other reasons. Burkholder said she never sent cultures because Bigelow wasn't prepared to feed fish to the cultures to keep them toxic, and it didn't have a BSL-3 facility. Robert Andersen of Bigelow, however, says that at the time, “there was no consensus in the community” that the toxin existed. Burkholder cites another reason for her reluctance: She learned that she was on a “top 10 list of the worst scientists in the country according to some conservative group” and that chicken farm barons Frank and Jim Perdue were “protesting” her research. “I really pulled back because I feared people working on these cultures, coming out saying they're not toxic, [were] not culturing them correctly; I felt I was in a major Catch-22.”

Burkholder disputes claims in a 6 August New York Times article—cited in a Science editorial—that she has refused to give cultures to various groups (Science, 23 August, p. 1237). Some of these requests Burkholder disqualified because they were verbal. She says NCSU considers only written requests. The University of Maryland Biotechnology Institute (UMBI) once offered several thousand dollars for cultures, but “it wasn't enough.” An e-mail that microbiologist Pat Gillevet of George Mason University sent in 2001 requesting amoebas was never received, according to university logs, Burkholder says.

Her supporters say she's right to be cautious. “Some people have not dealt with her in a professional way,” says marine biochemist Craig Cary of the University of Delaware, Newark and Lewes. “JoAnn often takes it personally, and rightfully so.” Cary says Burkholder was initially reluctant to send his group toxic cultures. “We had to really gain her respect and her confidence. It was very, very frustrating. But I never held it against her,” Cary says.

Burkholder doesn't hesitate to stand up for herself. Twice, grant award decisions have been reviewed because she felt she had not been treated fairly. And she is known for scathing, single-spaced, multipage letters she sends her critics. Once, she acknowledges, after a scientist made remarks at a Gordon Conference that she says she found “very denigrating” and “personally directed,” she sent a letter to his supervisor, John Wells, director of the Institute of Marine Sciences at UNC Chapel Hill. Wells says he investigated, but “no action was taken,” and the scientist “is in excellent standing at this institution.”

In August, a group sympathetic to Burkholder launched a strike on her critics. The Waterkeeper Alliance, an environmental group led by Robert F. Kennedy Jr., filed FOIA requests demanding the records of every researcher who has received federal funds since 1997 for Pfiesteria research (Burkholder included). Some labs are negotiating with the group or sending copies of old grant proposals as a first response. But if they have to turn over e-mails and lab notebooks, it could mean hundreds of hours of work, Vogelbein says. “To me, it looks like an attack on the way science is done in this country.” Burkholder says she's not directly involved in the filing but agrees that teams that couldn't grow toxic Pfiesteria should not have received funding.

One of the group's targets is Baden, who once collaborated with Burkholder but is now co-author on the PNAS paper that challenges her research. Baden, now at UNC Wilmington, received funding to study the toxin he unveiled in 1997, but has never published on it. Baden says by e-mail that, at the time, “I believed we had isolated a toxin from Pfiesteria,” but now thinks “there is no way of telling” what organism produced the toxic effects he observed.

## Closing in?

After the latest round of critical papers, some scientists are privately calling the hunt for Pfiesteria the “cold fusion” of biology—a wild goose chase whipped up by media coverage. “It's beyond science. It's a sickness,” says protistologist Tom Nerad of the American Type Culture Collection in Manassas, Virginia. But Burkholder thinks that many of her critics are simply impatient. She points out that identifying algal toxins can be difficult: It took scientists 25 years for brevetoxin, 7 years for maitotoxin.

Even some of Burkholder's fiercest critics think there was a harmful substance of some kind in her Pfiesteria tanks. The strongest evidence, they say, is that lab workers got sick. Theories abound, however, on what this substance was.

The problem all along, says Allen Place of UMBI, is that “it was never just Pfiesteria in the fish tanks.” Burkholder's gold standard for determining toxicity is to add Pfiesteria to fish tanks and see if the fish die. But this process adds an assemblage of organisms. UMBI researchers and Gillevet have identified more than 60 organisms—bacteria, eukaryotes, fungi—living in the fish tanks. “It's a milky mess,” says Place, and any of these components could have produced a toxin.

Gawley, the Miami chemist, suggests that because the VIMS group took steps to remove contaminants, such as dosing the broth with antibiotics, they may have unwittingly removed the source of a toxin. Burkholder agrees that a bacterial “cofactor” might be involved. Still another theory being investigated by a group at the UMD School of Medicine and the VIMS team is that Pfiesteria-rich water contains a bioactive agent that affects the brain of mammals, but not fish.

Burkholder, meanwhile, insists she and her team are close to identifying the toxin, and Ramsdell says they've made a recent “breakthrough.” After years of working with fractions separated from a gemisch of materials, keeping those that trigger a cell assay for toxicity, they have found a way to grow toxin-producing cells on algae free of contaminants. A compound isolated from these relatively clean cells yields the same nuclear magnetic resonance signature as the mixture originally found in fish tanks, Ramsdell says. “We have been tracking the same molecule the whole time,” he says.

Ramsdell has a video that shows this compound killing a larval fish in 5 minutes, and NOAA's Moeller says they will describe partial structural data at the harmful algae meeting in St. Petersburg, Florida, later this month. But even if they have isolated a lethal molecule, “we're not making the claim that it's the Pfiesteria toxin,” Ramsdell says; “we'll still need to prove it's what's killing fish.”

## Flu dynamics

Success in protecting against another notorious bug, the influenza virus, has prompted leading researchers to argue that public health recommendations should encourage vaccination of infants. John Modlin, head of pediatrics at Dartmouth Medical School in Hanover, New Hampshire, used data from several studies over the past decade to show that more young children than adults develop serious complications from influenza and end up hospitalized. Current CDC recommendations, however, call for routine immunizations only for adults older than 50 and children at high risk of developing serious disease, such as those with asthma or cardiovascular problems.

Modlin cited studies showing that healthy children under 2 who become infected by influenza virus end up hospitalized more than people in any other age bracket—including those who are 65 and older. He also pointed out that many children at high risk of developing serious complications from influenza go unvaccinated each year and end up hospitalized. Two recently published studies of children with asthma dramatically made this point: In the mid-1990s, 75% or more did not receive the flu vaccine. In contrast, 66% of adults 65 or older received the vaccine in 1999.

Official CDC vaccine recommendations come from a group that Modlin chairs, the Advisory Committee on Immunization Practices (ACIP). For the upcoming flu season, they changed the vaccination recommendation for children ages 6 months to 23 months to “encouraged when feasible.”

After Modlin's talk, pediatrician Ram Yogev of Children's Memorial Hospital in Chicago asked why ACIP had not strongly recommended universal vaccination of children. Modlin explained that it had been “slow to dawn on all of us” that influenza causes as much hospitalization of children as it does, and ACIP had “backed away” from recommending universal vaccination now because of two main obstacles: It remains unclear who would pay for the vaccine, and wide-scale use further complicates an already complex childhood vaccination schedule.

Both Yogev and Modlin emphasized that vaccinating all children against influenza likely would have many ancillary benefits. In particular, more high-risk children would receive the vaccine, fewer children would transmit the virus to their elderly relatives, and physicians would cut down on prescriptions of antibiotics, which don't fight the flu but are often mistakenly prescribed.

## Generation gap

Vaccinating children does not always benefit adults. On the final day of the meeting, John Edmunds, an epidemiologist at the Communicable Disease Surveillance Centre in London, presented a model suggesting that the success of the chickenpox vaccine might exact a steep tariff. The recently introduced vaccine works so well in children that, paradoxically, it might lead to more cases of shingles, a related form of the disease sometimes called herpes zoster, in adults.

Varicella, the virus that causes chickenpox, stays in a person's body for life. In some elderly people, it reactivates to cause nerve problems and skin blisters. Occasionally the condition is extremely painful and debilitating.

Edmunds explained that recent evidence from studies by his team and others gives “considerable weight” to the longstanding hypothesis that repeated exposure to varicella boosts the immune systems of adults, helping them keep the virus suppressed. His model predicts that as more and more children receive the chickenpox vaccine—which was introduced in the United States in 1995 but is not widely used elsewhere—adults will have less external exposure to the virus and rates of shingles will skyrocket. “This is a serious problem for countries outside of the U.S. considering varicella vaccination,” he said.

Edmunds said he has received mixed reactions to his work, and the meeting proved no different. “It was a very interesting presentation and very provocative,” said Michael Oxman, a virologist and clinician at the University of California, San Diego, and the Veterans Administration (VA). But epidemiologist Jane Seward, who heads a CDC division that does surveillance of chickenpox and shingles, stressed that they have yet to see any increase in cases of shingles and questioned some key parameters of Edmunds's model. “It's very early yet,” added Seward. “The data from next year will be more important to look at.” She added that adults might receive immunological boosts from their own varicella periodically reactivating but not causing disease.

Many researchers suspect that the chickenpox vaccine, if given to adults, might help thwart shingles. The VA has a study under way, headed by Oxman, to evaluate this possibility. The placebo-controlled study will involve 38,000 people, and Oxman said they expect to have the first data in 2004.

Edmunds did emphasize a silver lining. Children who receive the vaccine are less likely to develop shingles, he predicts, because the strain of varicella used in the vaccine is much weaker than the natural virus. Widespread use of the chickenpox vaccine thus should ultimately lead to a decrease in shingles. “It will come down quite dramatically,” he said, “after 50 to 60 years.”

• * 42nd Interscience Conference on Antimicrobial Agents and Chemotherapy, sponsored by the American Society for Microbiology, 27–30 September.

13. PROFILE: DENNY MORE

# Learning to Speak the Amazon's Languages

1. Bill Hinchberger*
1. São Paulo-based writer Bill Hinchberger is editor and publisher of BrazilMax (http://www.brazilmax.com/).

Denny Moore is bringing new linguistic approaches to old, dying languages; the work might elucidate the history of native Amazon Basin peoples

THE AMAZON RIVER BASIN—When Claude Levi-Strauss contacted the Mondé tribe in the Amazon in 1938, he found exactly the kinds of subjects he had been seeking for his research: an isolated tribe not then found in ethnographic literature. But the legendary French anthropologist was so weakened by his journey through the Brazilian outback that he had to give up on properly studying the group he called “my savages.” Levi-Strauss did, however, take note of their language, describing it as pleasant. He noted that many words ended with accented syllables: zip, zep, pep, zet, tap, kat. The Mondé, he recounted, peppered their speech with sounds that evoked the clash of cymbals.

More than 60 years later, linguist Denny Moore sits in a ramshackle house in a poor section of Porto Velho. A two-time Amazonian boomtown (first rubber, then mining), the city of 350,000 on the Madeira River has devolved into a sleepy state capital and reputed way station for cocaine. Across a small table, Maria Salomãy sits patiently. Moore, the director of the Amazonian linguistics center at the Goeldi Museum, a leading Brazilian research institute in Belém, tapes Salomãy's halting efforts to recall and articulate words in Mondé. Salomãy forgets most verbs but manages to come up with something for sun, moon, forest, sundry animals, body parts, and household and hunting implements.

One of the last three known semispeakers of Mondé, Salomãy is older than 60 and hasn't exercised her native tongue in decades. Following a familiar script, the Mondé were plagued with disease, death, and diaspora after coming into contact with Western civilization. Neither of Salomãy's two adult sons speaks a word of the language. “The only existing Mondé tape in the world,” announces Moore, pointing to his tape recorder.

The sessions with Salomãy could do more than preserve vestiges of a dying language. By comparing Mondé vocabulary and grammar with that of the languages of other tribes, Moore and his team of researchers hope to reconstruct “Proto-Tupi,” the extinct language from which modern Tupian languages like Mondé evolved. Called the Goeldi Comparative Tupi Project, this ambitious endeavor aims to unlock some of the mysteries surrounding the origins of the little-studied native Amazon languages. Because language reveals information about human activities, Moore and his team also hope to dig up clues about the prehistory of a region weak in archaeological evidence.

Moore's research might ultimately define the current epoch in linguistic history, predicts anthropologist Michael Silverstein of the University of Chicago. Moore, who grew up in Michigan and earned a Ph.D. at the City University of New York, helps indigenous people preserve their own heritages, unlike many scholars who dash in, gather data, and leave without providing any aid to their informants. In this way, Moore's work “is different not just in terms of consciousness of the languages but also in terms of the self-consciousness of the people who speak them,” Silverstein says.

A boxer in college, Moore takes a no-holds-barred approach to protecting the languages and cultures of Amazonian Indians. For 3 decades he's sparred with all comers. He's survived everything from tropical sprue, a rare malady of the small intestine that interferes with the absorption of food, to a murder attempt by a couple of men from a tribe at odds with one he was studying.

For his efforts, Moore won a $365,000 “genius grant” from the MacArthur Foundation in 1999. Although suddenly flush with cash, Moore sticks to a regimen worthy of a graduate student. The 57-year-old scholar recently spent 2 weeks crisscrossing the Amazonian state of Rondônia by bus to seek out informants like Salomãy. One 17-hour leg was marked by the partial collapse of a bridge underneath the bus; later the vehicle ran aground in a gully and nearly capsized. “If I keep my costs down, other researchers at the museum have to watch themselves, too,” Moore explains. “They'll be embarrassed if they outspend me.” ## Back through time Snuggled up against landlocked Bolivia, Rondônia state is home to half of the 10 branches of the Tupi language trunk, and most scholars suspect it is the original site of the Tupian people. Moore and a team of Brazilian linguists hope to reconstruct Proto-Tupi using an approach known as diachronic research, which identifies phonological, syntactic, and other features peculiar to a specific time period. Proto-Tupi probably dates back at least 2000 years, estimates Moore. In reconstructing languages, Moore and his colleagues also hope to piece together Amazonian prehistory, which is complicated by a hot and humid climate that accelerates decay and destroys much archaeological evidence. Using linguistic techniques to reconstruct ancient language forms, scientists can compensate for this problem and identify many ancient human behaviors. “The vocabulary of a language is an inventory of the culture of its speakers,” wrote Moore and Luciana Storto in a paper published this year. For instance, they reported that Proto-Tupi has the words “acaí” (a tropical fruit) and “ax,” indicating that its speakers used axes and inhabited a region with acaí trees. The researchers are racing against time. Mondé already appears on many lists of extinct languages. At least two other of the nine target languages in the Goeldi Comparative Tupi Project are similarly endangered. And Tupi is not alone. Of the estimated 1200 languages spoken in the Amazon Basin 500 years ago, at least three-quarters have disappeared. About 160 indigenous languages are still holding on within Brazil's national borders, but one-quarter of them are spoken by 50 or fewer individuals. A decent linguistic description—which would consist of an analysis of phonology, syntax, and morphology (the different forms words take, such as in declensions and conjugations)—exists for only a small fraction of these languages. Even fewer can boast a dictionary and a collection of texts. “It is a parallel to biology” and its accelerating loss of species, says linguist Norvin Richards of the Massachusetts Institute of Technology. “There is a lot to learn, and you have to be quick.” To counteract the perhaps inevitable loss, Moore and his Goeldi team hope to videotape 10 hours of standardized linguistic data on every native language in Brazil. Like everything else collected in the Goeldi project, these tapes will be made available to the Indians as well as to linguists. “The depth of this project will set it apart when seen 25 years from now,” says Silverstein. A language called Gavião will figure prominently in the project. Moore has been working on the language since 1975, when he entered a forest village of the Gavião Indians where he would conduct his dissertation research over the next 3 years. Contrary to the consensus opinion at the time among linguists, who were predominately missionaries, Moore discovered early in his Gavião research that it is a tone language like Chinese. Tone refers to the relative pitch of a syllable or word; a difference in the tone of an otherwise identical word results in a difference in meaning. Moore was terrified by his discovery—he's tone-deaf. To overcome this handicap, Moore persuaded his Indian informants to help him by whistling the tones, a technique they sometimes use for unobtrusive communication while hunting. Gavião now counts among the seven languages, all previously unwritten, for which the Goeldi Museum has implemented literacy projects. As Moore was touring Rondônia, his colleague and former student Vilacy Galucio, a native of the Amazon, was holed up in Costa Marques, a backwater town across the Guaporé River from Bolivia. She was trying to collect a word list in Puruborá, but she was having trouble. Her informant, Paulo Aporeti Filho, a Puruborá semispeaker, had recently suffered a stroke, and his memory was failing. To jog it, Aporeti suggested that it might help if they brought in a friend, another semispeaker with whom Moore had worked briefly several years ago. José Evangelista agreed to come. Galucio got him on an airplane. This group approach is unorthodox among linguists, says Moore. Yet the banter between the two informants seemed to prod them into remembering words and phrases. “We thought that it would be nice to do in human terms, at least,” says Moore. “And now it seems to have contributed to linguistic methodology.” One preliminary result of Galucio's research: Puruborá seems to have seven vowels instead of the five or six found in most Tupian languages. This provides some important insights into vowel correspondences in the Tupi language stock, says Moore. Specifically, it provides evidence suggesting that the Puruborá family is closer to a related family, called Ramarama, than previously thought. “Depending on what happens with Vilacy's test case, I think we may try to bring together as many of these old people as possible,” says Moore. Galucio is one of the many Brazilian linguists Moore has trained. A fellowship to his Goeldi institute now pretty much represents a ticket to a good graduate school; no other program in South America has sent so many students to study abroad. As Moore never tires of repeating, Galucio earned her Ph.D. with distinction from the University of Chicago. And anthropologist William Balée of Tulane University in New Orleans notes that Moore “built a whole linguistics program, perhaps now the leading program in Brazil.” After checking in with Galucio in Costa Marques, Moore takes the bus to the town of Cacoal, where he needs to consult with a tribal leader. Coincidentally, this person had accompanied a friend who tried to clobber Moore from behind with a wrench some 25 years ago. (He backed down when Moore whirled around.) The two apparently were making a point about a disagreement their tribe had had with the Gavião, with whom Moore was living. Both Moore and the Cacoal leader are willing to forget past grievances. They work out an agreement for collecting data and perhaps developing a literacy program for the tribe's language. It's nothing strange given the culture of constantly shifting alliances among tribes, Moore says: “You have to look at it in the Indian way.” 14. WOMEN IN SCIENCE # Can Equality in Sports Be Repeated in the Lab? 1. Jeffrey Mervis A 30-year-old federal education law caused participation in sports by women to skyrocket. Can it do the same thing for science? Physical chemist Debra Rolison has spent the past 2 years promoting a controversial idea for improving the status of women scientists in academia. But last week, when the U.S. Senate staged its first-ever hearing on the topic, Rolison was 1800 kilometers away, watching her niece play high school volleyball and basketball. Ironically, the reason Rolison couldn't attend was related to the theme of the hearing—whether a 30-year-old federal law that has increased opportunities for women athletes in schools and colleges could do the same thing for women scientists. Title IX of the Educational Amendments of 1972 prohibits sex discrimination by any educational institution that receives federal funds. Although the law applies to all activities, it has had its greatest impact in sports. By threatening schools with the loss of federal funds if they do not offer equal athletic opportunities to both sexes, Title IX has generated explosive growth in the number of women participating in competitive sports, including an 847% rise at the high school level. Rolison, who heads the advanced electrochemical materials program at the Naval Research Laboratory in Washington, D.C., would like to see the law help women pursue their intellectual dreams, too. So would Senator Ron Wyden (D-OR), chair of the science panel of the Senate Committee on Commerce, Science, and Transportation. “If Title IX can do that on the playing field, it should certainly do so in the classroom,” he declared last week at a hearing on the topic. But Wyden also knows that he's playing with fire. Title IX has sparked a political backlash among segments of the athletic community, who say that it's forced many financially strapped schools to eliminate boys' teams in a misguided effort to achieve gender parity. Title IX is currently being reviewed by a blue-ribbon panel appointed by Department of Education Secretary Rodney Paige; its report is due in January. Even supporters of the law are troubled by the specter of the federal government withholding research funding from universities that aren't doing enough to attract, retain, and promote women in science and engineering. “Title IX is a very big hammer,” Rolison admits. But for the past 2 years she has argued before numerous academic gatherings that this tool is needed to spur university administrators into action. Wyden was intrigued by the idea when it came up at a July hearing he chaired on the barriers facing women in science. So last week he gathered together five activists, including Birch Bayh, the retired senator from Indiana who co- authored the original law, as well as an official from the Department of Education, which enforces the law. Predictably, the activists agreed with Wyden that gender inequity within science faculties is a serious problem. The latest evidence arrived last month in a campus survey by the University of Michigan, Ann Arbor, which found a chillier environment for women faculty members on measures such as sexual stereotyping and harassment as well as in contract negotiations, mentoring, and leadership opportunities (see graphic). These distinctions exist not just between men and women but even between women in the hard sciences and those in the social sciences, says Abigail Stewart, a psychologist who led the study. “People have wondered if women are just more likely to complain,” she says. “But that's not true. The climate for women [natural] scientists is worse than for those in the social sciences.” Only the Education Department official disputed Wyden's bleak assessment of the current situation. “I am here to deliver good news,” announced C. Todd Jones, deputy assistant director of the department's Office for Civil Rights. Jones confessed that he was not yet in school when Title IX was passed. But he argued that the gains made by women in science over the past 30 years are due in part to Title IX, citing girls' recent dominance in winning prestigious high school science prizes and a 30-year rise in the proportion of undergraduate degrees earned by women in several technical fields. But despite the near unanimity that more must be done, many academics aren't sure that Title IX is the right vehicle. Nobody at the hearing supported any type of quotas, and few share Rolison's view that universities can be “bullied” into doing better. “It was a necessary tool in athletics,” says Stewart. “But it will be viewed by some as punitive if it's applied to academic life.” A better approach, says Geraldine Richmond, a chemical physics professor at the University of Oregon in Eugene who testified last week, might be to stiffen the backbone of federal agencies, some of which are already trying to address the problem. She applauds a new policy at the National Science Foundation to return any grant application that does not explicitly state how the proposed project addresses “broader implications” such as diversity and scientific literacy. In contrast, says Richmond, who is also chair of a scientific advisory panel to the Department of Energy, most federal research agencies “don't pay as much attention to these issues as they should.” Still, as the clock ran down on last week's hearing, one witness made it clear that she doesn't expect universities to improve without a strong push from the federal government. “Digit” Murphy, coach of the women's hockey team at Brown University in Providence, Rhode Island, told Wyden that the recent gains in women's athletics would never have occurred without the threat of fiscal punishment. “Absolutely not,” she said. “The only thing that schools pay attention to is when the NCAA comes in and does an audit. Then the university officials ask, ‘Are we in compliance [with Title IX]?’” Wyden has asked the National Academy of Sciences to review how U.S. universities apply Title IX to their science and engineering faculties. But Rolison sees Murphy's response as proof that a stick works better than a carrot. “Carrots are for vegetarians,” Rolison says. “And we are dealing with carnivores.” 15. ORNITHOLOGY # High-Flying Science Seeks to Reduce Toll at Towers 1. David Malakoff After years of debate about how best to reduce massive bird kills, researchers are beginning to receive funding to investigate solutions Next spring, if all goes as planned, Bill Evans will spend his nights chasing migrating birds under cloudy Midwestern skies. Using bright lights that normally warn pilots away from tall communication towers, the independent scientist hopes to understand why flocks are drawn to deadly encounters with the spires and discover lighting schemes with less allure. It's the first significant research project to be spawned by a recent debate over the threat to birds from a growing thicket of towers, and—if successful—it could help reduce a massive avian death toll. Ornithologists estimate that at least 4 million birds—mostly night-migrating warblers, vireos, and other songbirds—die annually at about 140,000 existing U.S. towers (Science, 16 March 2001, p. 2081). Conservationists worry that the toll will rise under plans to add thousands more towers to the landscape. But their pleas for studies that might identify solutions have gone largely unanswered. “It's been a Catch-22,” says Ellen Paul, executive director of the Ornithological Council, a Washington, D.C.-based group that represents 10 major bird science societies. “Industry and [tower regulators] say they need better science to justify taking action, but then they won't fund the necessary studies.” Industry officials say it's partly a matter of finding money in a depressed economy. The Bush Administration also hasn't made it a priority. And some question whether tower kills are really a major problem: A host of other factors, from cats to habitat loss, are believed to kill far more birds. But bird advocates say that's no reason for inaction. Some creative financing is now helping tower research get off the ground. The Evans study, for instance, will receive$50,000 from the settlement of an environmental lawsuit in Florida. And conservation groups are lobbying Congress for more money. Environmental attorneys, meanwhile, are preparing a legal challenge that could force regulators to cough up more cash. The legal strategy has already prompted one state, Michigan, to mull funding research in order to settle one case.

Researchers say that there's no shortage of questions. Members of the Communications Towers Working Group, set up by the U.S. Fish and Wildlife Service (FWS) to bring government, industry, and academia together, have called for a nationwide survey to refine tower kill estimates, which range from 4 million to 40 million birds a year. They'd also like to examine the conventional wisdom that the kills are limited, occurring primarily in eastern North America during the fall migration, and that the worst episodes—up to 12,000 birds in one night at a single tower—occur during overcast weather at towers higher than 75 meters, often marked by red, blinking lights

But answering such questions could cost millions of dollars and take years. So Evans decided to focus on a narrower issue: the impact of the red-and-white blinking lights that the Federal Aviation Administration requires on all towers over 65 meters. On misty nights, the warning lights attract birds, which often become confused and smash into tower girders and cables.

Evans, who runs Old Bird, a nonprofit research organization in Mecklenburg, New York, is working with a tower-lighting firm to build a rack of multicolored bulbs that blink at various rates and intensities. Instead of putting the lights on a tower and waiting for next fall's migration, Evans decided to “take the lights to the birds and the weather.” He'll drive the light box into the northern Midwest, where he expects to find plenty of spring migrants and overcast nights. Then he will monitor how the birds react to different lighting schemes, hoping to find one that could reduce collisions—but also meet federal visibility requirements and be inexpensive to install.

Other researchers are probing the underlying neurological reasons why some birds become confused when exposed to colored light. Ornithologists believe that night migrants typically navigate using the stars, but rely on Earth's magnetic field on overcast nights. But colored light can cause birds to become disoriented, suggest laboratory experiments by Roswitha and Wolfgang Wiltschko at Goethe University in Frankfurt am Main, Germany, and others. Red wavelengths, in particular, may interfere with vision-related pigments that also play a role in magnetic navigation, says Bob Beason of the University of Louisiana, Monroe, who has conducted similar work. The findings may explain “why towers with red lights seem to be a bigger problem,” he says.

Further research might depend on some help from Congress. Representatives John Dingell (D-MI) and Edward Markey (D-MA) have proposed giving FWS \$350,000 for peer-reviewed studies already identified by the agency's working group. But budget squabbles in Congress may block that earmark.

Other groups are looking to regulators and the courts for help. They want FWS to prosecute the owners of especially deadly towers under migratory bird laws, and they are also pressuring the Federal Communications Commission (FCC), which licenses towers. Over the last few years, the American Bird Conservancy, Friends of the Earth, the National Wildlife Federation (NWF), and other groups have challenged FCC permits for thousands of towers. They want the agency to conduct better environmental studies and consider the cumulative impact of all towers on birds. FCC has so far rejected most challenges on technical grounds, setting the stage for a federal court battle. In the meantime, NWF is negotiating with Michigan officials over remedies for apparent legal violations in building 181 towers for emergency communications, including several sited squarely in migratory pathways.

Evans and other scientists would prefer to work voluntarily with tower owners and users to find acceptable solutions. But legal wrangling may be unavoidable, he adds, noting that it took the threat of court sanctions to convince the wind-power and power-line industries to pour millions of dollars into understanding how to make their facilities safer for birds.