# News this Week

Science  31 Jan 2003:
Vol. 299, Issue 5607, pp. 636
1. CLIMATE CHANGE

# A Perfect Ocean for Four Years of Globe-Girdling Drought

1. Richard A. Kerr

It's a long way from Atlanta to Kabul, but parched crops in the U.S. Southeast and dry wells in Afghanistan have more than superficial similarities: It seems they share a cause. Researchers are now blaming both droughts on the distant tropical Pacific. Odd doings in that part of the ocean seem to be responsible for 4 years of dryness spread across North America, southern Europe, and central-southwest Asia from Iraq to Pakistan. The link is the grandest connection between ocean and recent climate since El Niño's reign was recognized and marks yet another instance in which the tropics have been tied to climate shifts both brief and prolonged (Science, 10 January, p. 183). Now that this drought is easing, as predicted in an encouragingly prescient forecast, researchers will be confronting the real possibility that greenhouse warming will make such droughts commonplace.

On page 691 of this issue of Science, climate dynamicists Martin Hoerling of the National Oceanic and Atmospheric Administration's (NOAA's) Climate Diagnostics Center in Boulder, Colorado, and Arun Kumar of NOAA's Climate Prediction Center in Camp Springs, Maryland, compare climate-model simulations with the pattern of Northern Hemisphere drought. The comparison shows how two parts of the tropical Pacific ganged up on the mid-latitudes to create persistent drought in widely separated, seemingly unrelated regions.

In a departure from usual practice, Hoerling and Kumar tackled an ongoing climate event—and a particularly widespread one at that—rather than something of smaller scale from deep in the data archives. And they threw an exceptional amount of computing power at the problem: They ran three models independently developed by three different institutions for a total of 51 runs. In each run, Hoerling and Kumar inserted the actual sea-surface temperatures observed from 1998 to 2002 and saw how the models' atmospheres responded.

Averaged together, the 51 simulations showed a striking similarity to the globe-spanning precipitation shortage that actually happened. “There was virtually not a single model run that had greater than normal precipitation in southwest Asia or the United States,” says Hoerling. “It was just remarkable.”

Hoerling and Kumar narrowed the ocean's influence to the tropical Pacific by running a fourth model, altering temperatures of only certain parts of the ocean from their long-term averages. Including just the hot spots—in those years the western Pacific and eastern Indian Ocean—produced some parts of the mid-latitude dryness. El Niño's opposite number, La Niña, provided the period's most abnormally cool area, in the central and eastern tropical Pacific. When the researchers included that chill in the model, it reinforced some parts of the warmth-induced dryness and produced other parts of the pattern. Only a combination of the two tropical Pacific regions of unusual cold in the east and warmth in the west fully reproduced the observed dearth of precipitation. The cool eastern Pacific and the warm Indo-western Pacific “are two superimposed, additive influences,” says Hoerling.

Other modelers find the NOAA work persuasive. “It's very convincing,” says modeler Siegfried Schubert of NASA's Goddard Space Flight Center in Greenbelt, Maryland. There's also “a very nice agreement between models and observational analysis,” says meteorologist Mathew Barlow of Atmospheric and Environmental Research Inc. in Lexington, Massachusetts.

Last year, Barlow and his colleagues traced the influence of the tropical Pacific of the late 1990s to the drought-stricken central- southwestern Asia region. In much the way an El Niño can reach across an “atmospheric bridge” to shift wintertime climate over North America, they suggested, the enhanced warmth in the western Pacific pumped up the region's towering rain clouds, thereby releasing additional heat to the atmosphere. That heat in turn altered winds of the jet stream that then steered away the rain-laden winter storms that normally march eastward across Iraq, Iran, and Pakistan, as well as Tajikistan, Uzbekistan, and Turkmenistan to the north.

This ocean-drought connection is receiving further support from the easing of the drought in recent months. Once La Niña gave way to El Niño last summer, heavy rains in the southeastern and eastern United States brought relief in the fall, and winter rains returned to central-southwestern Asia. In fact, last spring, a group at Columbia University's International Research Institute for Climate Prediction (IRI) in Palisades, New York, headed by climatologist Anthony Barnston, actually forecast the easing of the Asian drought, based on Barlow's ocean-drought link and the imminence of El Niño. Such long-range seasonal forecasting has been impossible outside of classic El Niño-La Niña conditions.

Despite the welcome rains, the trouble isn't over. The Indo-western Pacific remains warmer than normal. The IRI group expects that that lingering warmth could still pose a threat for the southern parts of the Iraq-to-Pakistan tier this winter. Indeed, “they may be in for a long-term problem,” says Barlow. The Indo-western Pacific has warmed since the 1970s and may continue to warm. “There's a strong suspicion that the Indo-western Pacific warming trend is related to the global warming trend,” says Barlow. If the tropical Pacific is responding to mounting greenhouse gases by warming in the west, drought in parts of central-southwest Asia could become commonplace, and the double whammy of western warmth plus La Niña cold could become more frequent in a greenhouse world. Then Atlanta and Kabul might be sharing their pain more often.

2. U.K. UNIVERSITIES

# Tuition Scheme Sparks Worries

1. Gretchen Vogel

A revolutionary proposal in British higher education—a system of graded tuition fees—may scare students away from science courses, researchers fear. The government plan, aired last week, would abandon the existing one-size-fits-all fee and allow universities to charge students up to $5000 per year. But it has come under fire from two directions: student groups and politicians arguing that the higher fees will dissuade poorer students from enrolling in expensive-to-teach subjects such as science, and university leaders contending that the plan would not bail them out of their current funding crisis. Reforms are necessary because public funding for education has not kept pace with a dramatic rise in university enrollment in the United Kingdom in recent decades. Indeed, a cash crunch is visibly taking a toll on campuses across the country. “We're £8 billion [$13 billion] in the hole in terms of infrastructure improvements that need to be made,” says Leslie Aiello, head of the graduate school at University College London. “Things are in crisis, and something needs to be done to inject money into the system.”

In a report last week from the Department for Education and Skills, the government argues that students who benefit from a university education should contribute directly toward its costs. Currently, students pay $1800 per year in fees, no matter where or what they study. Under the new plan, universities in England and Wales would be allowed to set their own fees up to the$5000 ceiling. (Scottish universities are governed by a separate system.) Students would not have to pay the fees until after graduation and after their annual income exceeds $24,500. Interest-free payments would be spread out at a rate commensurate with a graduate's income, and those who enter certain public-sector jobs or earn low salaries may have their fees paid by the government. Such a scheme might sound generous to North American students, who can expect to pay up to$30,000 a year in tuition up front at private universities. It has, however, caused an uproar in Britain. There, until 5 years ago, higher education was free for anyone who met the academic requirements for entrance.

In a statement last week, The Royal Society, the U.K.'s most prestigious scientific body, cautioned that higher fees for science courses, if implemented, might discourage students from enrolling. That's a well-founded concern, says Peter Cotgreave of the lobby group Save British Science. “If I'm a clever person from a relatively poor background, I would probably study law” rather than science, as fees might be lower and salaries higher, he says.

In the end, the new fees are unlikely to please anybody. The real cost of educating a student at Imperial College in London is closer to $17,000 a year, far more than the college will recoup in fees, says rector Richard Sykes. Imperial has said it will charge the maximum for all courses, as none costs less than$5000 per year. With students paying fees retroactively, Aiello notes that universities would see no extra cash under the plan for at least a few years.

The education department will accept comments on the report until the end of April. Legislation on the proposal is expected in Parliament later this year.

3. SPAIN

# Research Chief a Victim of the Oil Spill?

1. Xavier Bosch*
1. Xavier Bosch is a writer in Barcelona.

BARCELONA—Criticisms of the government's handling of the Prestige oil spill may have claimed the first victim in the top echelons of Spanish science policy. On 24 January, the government announced that it had accepted the resignation of Rolf Tarrach, president of the Higher Research Council (CSIC), Spain's main basic research agency. However, Tarrach suggests that he is being made a scapegoat for the government's stumbling response to last November's devastating spill.

Tarrach, a theoretical physicist at the University of Barcelona, had been appointed head of CSIC in September 2000. His appointment was largely welcomed by the scientific community, which hoped he would lure fresh talent into CSIC's network of 121 research centers. But when Tarrach assumed the reins of the agency and its $400 million annual budget, he found himself reporting to the science ministry, which had acquired CSIC from the education ministry in 2000. According to Tarrach, the situation deteriorated steadily. On 21 October, he sent a letter to Pedro Morenés, state secretary of scientific policy, in which he threatened to resign if the ministry did not find a solution to the salary gap between CSIC scientists and their university counterparts, who earned up to 10% more. The differential was causing a “brain drain of CSIC scientists” to the universities, Tarrach says. He found himself under increasing pressure from the ranks: In a 4 December letter, 10 directors and scientists of CSIC institutes urged him to deal with a series of concerns about his management of the agency. Tarrach's woes intensified after the oil spill. In a 24 January letter to Science (p. 511), 422 scientists accused the government of failing to adequately take into account the views of the scientific community. Although no government official was singled out in the letter, some researchers pin at least part of the blame on Tarrach. He has shown an “incapacity of leadership,” charges Juan Eugenio Iglesias of the Institute of Materials Science in Madrid. Tarrach, he says, has demonstrated that “he serves the government rather than the scientists.” If that's the case, the government has found an odd way to express its appreciation. On 24 January, the conservative newspaper ABC announced Tarrach's resignation. Tarrach himself says he learned of his resignation from the news article; he blames his fall on the Science letter, which generated widespread press coverage in Spain that day. In a written statement, the science ministry explained that Tarrach resigned because he “wished to return to the academic life.” A spokesperson declined to comment on suggestions that the Science letter precipitated Tarrach's departure. Ironically, the CSIC official most responsible for dealing with the oil spill—Emilio Lora-Tamayo, CSIC vice president and head of the agency's scientific commission on the spill—is being tapped as Tarrach's successor. 4. NEUROSCIENCE # Minor Variation in Growth-Factor Gene Impairs Human Memory 1. Jean Marx Buried deep within the brain, the sickle-shaped hippocampus helps determine what a person learns and remembers. Now, researchers have identified a tiny genetic variation that may influence just how effectively the hippocampus functions. The genetic twist may also affect a person's susceptibility to brain diseases such as Alzheimer's. People carrying a particular variation in the gene for a protein called brain-derived neurotrophic factor (BDNF) didn't perform as well on a memory test as people with the standard version of the gene did, according to a report in the 24 January issue of Cell by Michael Egan, Bai Lu, Daniel Weinberger, and colleagues at the National Institutes of Health (NIH) in Bethesda, Maryland. Brain-imaging and other studies point to abnormal functioning of hippocampal neurons in those with the variation, which changes just one amino acid in BDNF, replacing valine with methionine at position 66 of the protein. Neurobiologist Susan Patterson of Columbia University in New York City says that the work “provides a very nice demonstration that BDNF plays a role in some forms of human memory.” The NIH team had ample reason to suspect that BDNF might be involved in memory and learning. Although it was originally discovered as a general facilitator of neuron growth and maintenance, over the past few years, numerous groups, including Lu's, have linked it to the neuronal remodeling that underlies learning and memory. In particular, they have found that it enhances a phenomenon called long-term potentiation (LTP), in which synapses, the connections between neurons, are strengthened when their neurons are stimulated simultaneously. In the first phase of their work, Weinberger and his colleagues searched gene databases for variations in BDNF that might influence the protein's function. They also wanted to see if any variations could be linked to schizophrenia, which is associated with derangements in hippocampal function. The valine-to-methionine switch at site 66 looked promising, Weinberger says. It's a fairly common variant: Of 600 people the researchers examined, 32% had at least one copy of the oddball gene. And the variation is located in the so-called ZIP code of BDNF, a sequence that directs the protein to its correct destination in the cell. Thus, the amino acid change wouldn't alter BDNF function directly, but it could do so indirectly by causing the protein to end up at the wrong place. Subsequent testing revealed no link between the methionine variant and schizophrenia risk. But the researchers did find that both schizophrenic and healthy subjects who carried the variant fared worse on a test that measures episodic memory—the ability to remember past experiences—than did people who had two copies of the valine version. Results on the test, in which subjects are asked to recall the elements of a short story that they have read, have been linked to activity in the hippocampus. Brain-imaging studies confirmed that hippocampal function was abnormal in people with the methionine variant. In a different memory test, in which the hippocampus is normally deactivated, functional magnetic resonance imaging showed that activity in this brain region was actually turned up in people with the variant. In addition, Weinberger's group used another imaging method to look at levels of a chemical called N-acetyl aspartate (NAA), an indirect measure of the richness of synaptic connections among neurons. Hippocampal NAA levels were lower in people with the methionine variant. All in all, Weinberger concludes, “the form of the gene you have influences how well your hippocampus works.” Further studies by the NIH team point to a possible reason for the impaired hippocampal function in people who make the variant BDNF. Normally, neurons secrete BDNF when they are stimulated. This is thought to help strengthen synapses, as in LTP. But studies of cultured hippocampal neurons showed that the methionine variant wasn't transported to the nerve endings as it should be. As a result, BDNF wasn't secreted when the neurons were stimulated. Neuroscientist Mu-ming Poo of the University of California, Berkeley, describes these findings as “very intriguing … a very bold link of a molecular defect with a defect in cognitive function.” Even so, Egan and others note that the variant's effect on memory is small, and it has no apparent effect on one's IQ. But the findings raise the possibility that people with the variant may be more vulnerable to memory loss that may come with advanced age or brain injury. “Everybody may not have the same genetic toolbox to deal with additional insult,” as Weinberger puts it. 5. METABOLIC ENGINEERING # Researchers Create First Autonomous Synthetic Life Form 1. Robert F. Service Peter Schultz is dissatisfied with the number 20. That's the number of amino acids that virtually all organisms use to construct proteins, the molecules that carry out the lion's share of chemistry within cells. Two years ago, a team led by Schultz, a chemist at the Scripps Research Institute in La Jolla, California, engineered the genes of a live organism—an Escherichia coli bacterium—to incorporate a 21st amino acid into its proteins. That made the bug the first synthetic life form with a chemistry unlike anything found in nature. To make proteins with unnatural amino acids, however, the researchers had to put that new amino acid into the bacterium's growth medium. Now, in a paper scheduled to appear in this week's issue of the Journal of the American Chemical Society, Schultz's team reports going one step further by engineering an E. coli that not only incorporates a 21st amino acid into its makeup but also manufactures the compound by itself. “These results are very exciting,” in part because they provide researchers with a new way to explore evolution, says David Liu, a chemist and specialist in molecular evolution at Harvard University in Cambridge, Massachusetts. “It's tantalizing to ask how can this organism evolve, now that it is equipped with a method for not only using a nonnatural building block but also for creating that building block,” says Liu. Getting an organism to expand its standard amino acid repertoire took some doing. Schultz's team has had to reengineer the basic machinery that cells use to make proteins. That machinery starts with trios of nucleotide bases, called codons, that make up DNA. Each codon directs a particular amino acid to be added to a protein chain. To create its first synthetic bug, Schultz's team co-opted a little-used codon—known as the amber stop codon —and reengineered the cellular enzymes to add a new amino acid, O-methyl-L- tyrosine, whenever it saw that codon. In the latest work, Schultz—with students and colleagues at Scripps, the University of California, Berkeley, and the Genomics Institute of the Novartis Research Foundation in La Jolla—followed that same strategy with a new amino acid, p-aminophenylalanine. To enable the microbes to make the new amino acid on their own, they also spliced a trio of enzyme- producing genes from a strain of Streptomyces bacteria into the E. coli. Together with another enzyme already present in E. coli, these enzymes enabled the bugs to turn a common compound in E. coli called chorismate into p-aminophenylalanine, which protein-building enzymes then picked up and inserted into growing proteins. What difference does the altered chemistry make? To find out, Schultz says, his team is randomly inserting amber stop codon mutations into the E. coli genome. Next, they plan to put these E. coli and the 20-amino-acid variety under selective pressure by changing their food supply and other factors, to see whether the bacteria with a 21st amino acid fare better than natural ones do. If so, he says, it would suggest that although biology has made do with 20 amino acids for billions of years, evolution could make use of plenty more. 6. RUSSIA # Prestigious Plant Institute in Jeopardy 1. Paul Webster* 1. Paul Webster is a writer in Moscow. MOSCOW—A renowned plant sciences institute in St. Petersburg is facing eviction because of the Russian government's desire to beef up its presence in the former Imperial capital. The All-Russian Vavilov Institute of Plant Industry, home to the world's oldest and second-largest plant gene bank, has already beaten off one attempt by city authorities. But this time, its opponent is none other than Russian President Vladimir Putin, a St. Petersburg native. This is not the first time that Vavilov researchers have found themselves staring into the abyss. A decade ago, after the Soviet Union unraveled, the institute received$5.5 million from Western institutions for repairs and renovations to facilities housing its vaunted collection of 320,000 plant germ-plasm accessions.

The latest threat originated last year, when officials in the office of St. Petersburg vice governor Valeriy Nazarov, the federally appointed chair of the City Property Administration Committee, hatched a plan to acquire one of the Vavilov's buildings. The institute hit back with a civil suit, arguing that the land grab was illegal because the building's owner is the Russian Academy of Agricultural Sciences. The Vavilov has occupied the building in question since 1930, some 36 years after it was erected for the Russian Ministry of Agriculture. Last month, the court found for the institute; last week, Nazarov's office announced that it would appeal.

The appeal is backed by a federal decree, issued 17 December, that transfers control of both of the institute's buildings from the Agricultural Academy to the Putin administration. “The president wants to move bureaucrats from Moscow into our building,” says the institute's director, Viktor Dragavtsev. “If they take it, we will have nowhere to go.” He estimates that relocation into new facilities would cost $70 million, including roughly$30 million for a new germ-plasm storage facility. Nazarov declined to speak with Science.

Western experts hope that Russian officials will recognize the value of the Vavilov's germ plasm before inflicting irreparable damage. “It's a very important collection. Anything that threatens its ability to keep operating matters deeply, not just to Russia but to the entire world,” says Ruth Raymond, coordinator of the Global Conservation Trust at the International Plant Genetic Resources Institute in Rome, which is campaigning to raise $250 million to preserve international crop-diversity collections such as the Vavilov's. “The government must make provision for the collection,” she says; “otherwise it will be a tragedy.” 7. GLOBAL HEALTH # Bill Gates Plans a Hit List, With NIH's Help 1. Martin Enserink The richest man on Earth has a new plan to help the poorest—and maybe inspire others to follow him. Last Sunday, software tycoon Bill Gates announced that his philanthropic organization will spend$200 million on research into the diseases that plague developing countries, in a process that he hopes will entice other research organizations to take up the cause. The U.S. National Institutes of Health (NIH) and its director, Elias Zerhouni, are partners in the plan.

The program's key element is to create a list of “Grand Challenges,” about 10 concrete research projects that, if successful, could make a huge dent in human disease and suffering worldwide. The objectives will be drawn up by top scientists over the next half-year and published. The Bill and Melinda Gates Foundation will then give research groups up to $20 million each to address them over the next 3 or 4 years. The foundation hopes the global to-do list will stimulate complementary research and lure young scientists. Gates, who presented the plan at the World Economic Forum in Davos, Switzerland, says he was inspired by David Hilbert, a renowned German mathematician who formulated a set of unsolved math problems in 1900 that stimulated the field for decades. Noting at a press conference that some of the world's most deadly diseases are chronically underfunded, Gates said that he hoped a similar list of health problems would help attract resources for a focused assault. Among the potential challenges identified by the foundation are a drug that prevents latent tuberculosis from resurging and a malaria-resistant mosquito. “These will be risky things,” Gates said. “No guarantee of success.” The new program will be administered by the Foundation for the National Institutes of Health, a nonprofit corporation associated with NIH that allows the agency to accept money from nonfederal sources. NIH will help in a variety of ways, Zerhouni says—for instance, by offering advice or collaborating on grant proposals. Zerhouni, who will be a member of the panel that defines the grand challenges and picks the awardees, calls the list “an interesting model to watch. … Let's see if it works.” Other panel members are Anthony Fauci, head of the National Institute of Allergy and Infectious Diseases, director Julie Gerberding of the Centers for Disease Control and Prevention, and top scientists from developing countries. The chair is Nobel laureate Harold Varmus, a former NIH director who now heads the Memorial Sloan-Kettering Cancer Center in New York City. Others global health experts are enthusiastic. Gates has “a fascinating idea, money, and the right people,” says George Rutherford, interim director of the Institute for Global Health in San Francisco. “It's the perfect storm.” 8. ARCHAEOLOGY # Impending War Stokes Battle Over Fate of Iraqi Antiquities 1. Andrew Lawler While the world awaits a possible war in Iraq, a battle is already under way over how best to preserve the country's vast cultural heritage. At the center of the controversy is a group of wealthy and influential American antiquities collectors and curators with enough clout to wangle a meeting last week with U.S. Defense and State department officials. The collectors say that their goal is to save the country's myriad archaeological sites, museums, and invaluable data collections. But many scientists fear that the group is also eager to have a postwar government loosen Iraq's tight restrictions on the ownership and export of antiquities. Those changes would increase opportunities to obtain artifacts—a motive members of the group hotly deny. Although the 1991 Gulf War did negligible damage to Iraqi antiquities, an unsuccessful but bloody uprising against the government left museums looted and innumerable objects damaged or lost (Science, 6 July 2001, p. 32). A decade of poverty and economic sanctions also took its toll on ancient sites. That history leaves scholars and collectors alike more concerned with what happens after a war than with the immediate impact of bombs or troops. “Who knows what will happen in a post-Saddam Iraq?” says Ashton Hawkins, a former longtime counsel to New York City's Metropolitan Museum of Art and president of the New York-based American Council for Cultural Policy, which pressed for last week's meeting. “There is a real concern about the aftermath, and we would want to be helpful.” Iraqi officials say they welcome the focus on the country's rich cultural heritage. “We want people to be aware of what is here,” says Donny George, research director of Iraq's antiquities board in Baghdad. Although international teams have postponed plans to dig in Iraq because of the prospects of a war, the National Museum in Baghdad remains open, and domestic excavations continue. “If something happens, we will do our best to protect our objects,” says George. “We always prepare for the worst.” Seeking to avoid the worst, Hawkins and the council's vice president, Arthur Houghton, a former Getty Museum curator, met on 24 January with State and Defense department officials to discuss the impact of military operations on Iraq's cultural treasures. They also offered possible postwar technical and financial assistance, including conservation or computer-database support as well as access to private funding sources. But several archaeologists, while acknowledging the need to alert the U.S. government to the importance of Iraq's heritage, fear a hidden agenda in the council's moves. “There is a strong common interest here,” says art lawyer Patty Gerstenblith, a member of the Boston-based Archaeological Institute of America (AIA). “But one has the strong sense that this group is using this discussion as a pretext for their ultimate goal: to change Iraq's treatment of archaeological objects.” Hawkins strenuously denies this. “That's completely and utterly stupid and ridiculous, and I resent it,” he says. “The antiquity laws of Iraq are almost exactly where they should be.” Council members have spoken out strongly, however, against what they see as unreasonably strict controls in the United States on the import and sale of artifacts. Under Iraq's current system, which predates Saddam Hussein, it is illegal to export archaeological goods. Donny George says that those laws were tightened in November 2002 to impose severe criminal penalties on looters. William Pearlstein, the group's treasurer, describes the council as “a group of prominent collectors and curators who favor a rational and balanced approach to cultural heritage issues.” He characterizes Iraq's laws as “retentionist” and says he envisions a postwar government that would liberalize the issuance of foreign-dig permits. The council, he adds, supports “a sensible post-Saddam cultural administration,” with laws that should allow “some objects [to be] certified for export. If we can get the Defense Department to listen when it comes to targeting [bombs], and influence conservation of cultural heritage [with the State Department], then that's a pretty good start,” says Pearlstein, a lawyer who represents the National Association of Dealers in Ancient, Oriental, and Primitive Art. Talk of altering Iraqi laws is offensive to many archaeologists. “We have to be very careful,” says Gerstenblith. “This shouldn't be [preparation for] an American antiquities service in Iraq. It's not for the people in the United States or Europe to tell Iraq what laws they should pass.” AIA refused late last year to sign a statement drafted by the council calling for military forces to take into account the country's cultural heritage, she adds. A separate AIA statement calls for maintaining Iraq's current laws, including state ownership of sites and objects. University of Chicago archaeologist McGuire Gibson attended the Washington meetings after weighing the message his presence might send. “We're sort of caught: Either we get in and have a voice, or we will be totally ignored.” But he says that the community must remain vigilant lest “the carpetbaggers and scalawags go in to make money” from Iraq's ancient treasures. Others at the meetings included representatives of the Association of Art Museum Directors and the World Monuments Fund. State Department officials pledged at the meeting to include a cultural-heritage panel in their postwar planning, and Pentagon officials said they have coordinates for 150 major archaeological sites. Gibson and other scholars, meanwhile, are compiling a more detailed roster to help military planners who review potential bombing targets. It is a daunting task. “God knows how many sites there are—maybe 15,000 to 20,000 major ones,” Gibson says. Still, Gerstenblith says that scientists can only do so much. “If Saddam Hussein puts a command center next to a ziggurat,” she says, “it becomes a legitimate target.” 9. ECOLOGY # Conflict Over Cooperation 1. Ben Shouse* 1. Ben Shouse is a writer in Santa Cruz, California. A controversial push to focus on positive ecological interactions rather than competition and predation has ignited a debate among ecologists A brightly colored starfish with a voracious appetite for mussels occupies a key niche in ecological theory. In a series of seminal papers starting in 1966, Robert Paine of the University of Washington, Seattle, showed that by preying on mussels, Pisaster ochraceus creates an opening for other rock-dwelling shellfish, dramatically increasing the diversity of these sessile species. Take Pisaster out of the picture, and the riot of anemones, limpets, and barnacles on Washington's rocky shores would be replaced by a monotone of mussels. Paine called the starfish a “keystone” species because, like a keystone in architecture, it is crucial for maintaining structure. Since then, Paine's notion of keystone species has become a fundamental concept in ecology. But a group of renegade ecologists is now arguing that the concept is flawed. And they are taking shots at other long-standing bulwarks of ecology. In a provocative article in the March issue of Trends in Ecology and Evolution (TREE),* John F. Bruno of the University of North Carolina, Chapel Hill, John Stachowicz of the University of California (UC), Davis, and Mark Bertness of Brown University in Providence, Rhode Island, argue that much of modern ecological theory stems from a misleading fixation on the roles of competition, predation, and externally imposed stress in shaping natural communities. Missing from core concepts, they argue, is the growing realization that species can interact in positive ways—a process called facilitation—with major consequences for community structure. This more benign view of species interactions, the trio points out, can have profound implications for deciding which species to focus on in conservation efforts. “It is time to bring ecological theory up to date by including facilitation,” they write. And they warn: “This process will not be painless, as it will fundamentally change many basic predictions and will challenge some of our most cherished paradigms.” Take P. ochraceus. Bertness argues that the positive side—facilitation by mussels—is at least as important as the negative, predation by starfish. Mussel beds are home to hundreds of invertebrates that do poorly in the presence of the mussels' competitors. In studies that measured overall diversity, rather than the diversity only of sessile species, diversity was actually greater when Pisaster, the so-called keystone, was absent, says Bertness. “The classic keystone species example is wrong,” he says. This prompted the TREE authors to propose a theoretical alternative to the keystone species: the “foundation species,” which shapes communities by creating and enhancing habitat. Not surprisingly, this challenge to fundamental concepts is creating perturbations in the community of ecologists. Most of them agree that the subject has suffered undeserved neglect. “I have no problems with any of [Bertness's] work on positive effects,” says ecologist Bruce Menge of Oregon State University in Corvallis. But he and many other ecologists say that they have long recognized facilitation and have made up for earlier biases. Adds Paine, now a professor emeritus: “It depends [on] how [Bertness] casts it. If he says new sorts of facilitative mechanisms are going to turn ecology on its head, he's clearly wrong.” ## A different perspective That's exactly what Bertness and his colleagues are arguing. They note that competition and predation became the dominant forces in ecological theory beginning in the 1950s, when ecology began a transformation from a descriptive to an experimental science. Appreciation of the “largely unanticipated yet striking” influence of facilitation on the organization of terrestrial and aquatic communities came long after the core theories were well established, they write. And those theories have been slow to incorporate positive interactions, largely because their impacts are only now being put to the test. “We seem to finally be making headway, because we have unleashed experimental ecology to settle the debate,” says Bertness. Bertness and a parade of students built their case on experiments in marine systems, where harsh environments—salinity, pounding surf, alternating wet and dry conditions—are the norm. A study published online on 2 July 2002 by Oecologia, for example, shows that on the relatively saline south side of Cape Cod in Massachusetts, growth of transplanted salt marsh plants is more often enhanced by neighbors than on the milder north side, where growth suppression is more common. Facilitation is easier to find in such environments, which could bias their perspective. But some of the newest support for their views comes from plant ecology. Ragan Callaway of the University of Montana, Missoula, and his colleagues measured the effect of thinning out neighboring vegetation on the growth of small herbaceous plants on 11 mountain ranges from Alaska to Argentina to the Republic of Georgia. The standard competitive view would predict that the remaining plants should benefit from reduced competition for limited resources. That's what Callaway found for low-altitude plants: They grew 22% faster than controls did. But, as Callaway reported in the 20 June 2002 issue of Nature, high-altitude plants whose neighbors were removed grew 25% less, because neighbors improved microclimate, sheltered plants from wind, or stabilized the soil. “There's a ton of new evidence that has come out just within the last 10 years,” he says. For example, recent studies show that plants can change communities by providing shade or soil oxygen for neighbors. Despite such findings, and anecdotal evidence dating back as far as the 1960s, most ecologists still see facilitation as a collection of “cute little evolutionary stories,” says Bruno. “We're not saying people aren't aware of it; we're saying ecological theory doesn't account for it.” Most ecologists “definitely don't think of it as structuring communities in the way disturbance or competition or predation does,” he says. The new TREE article sets out to redress the balance. With its confrontational tone and its attacks on some of ecology's most sanctified ideas, the article is guaranteed to attract attention. First, it tackles the ecological-niche concept. According to this time-honored theory, the distribution of a species is restricted to a range of food sources and environmental conditions—a “fundamental niche”—that is whittled down by competition with other organisms to a “realized niche.” But the concept could be even more powerful if it included facilitation, the authors argue. Realized niches could turn out to be larger than fundamental niches, because some species create habitat or beneficial conditions for others. For example, the TREE authors note, by providing shade and moisture, intertidal seaweed canopies extend the distribution of many organisms to higher tidal heights than they would otherwise be capable of occupying. The authors also take on the intermediate-disturbance hypothesis. This pillar of ecology states that species diversity is highest when disturbance of an ecosystem is moderate: frequent enough to prevent the best competitors from dominating, but not so frequent that only fast-growing or resistant species survive. A classic 1979 study of algae living on intertidal boulders supports this idea. The boulders roll with the changing tides at a frequency determined by their size, crushing resident algae and providing a ready-made gradient of disturbance to study. It turns out that small boulders, which roll frequently, and large boulders, which rarely roll, both have low algal diversity. But medium-sized boulders have the highest diversity, because both dominant competitors and new colonizers have opportunities to survive. The TREE article points to little-noticed research on the same ecosystem in 1987, which took a broader view of diversity than just focusing on algae. The study, by Richard Dean and Joseph Connell, now at UC Santa Barbara, concluded that diversity of mobile invertebrates was highest at low levels of disturbance. The reason is facilitation: Less-disturbed algae provided more habitat for invertebrates. The challenge to the keystone-species concept is similarly based on the argument that ecologists should broaden their focus to include a wide variety of species and positive interactions. In the kelp forests off the West Coast of the United States, for example, sea urchins' appetite for kelp may increase the diversity of other seaweeds. But most ecologists recognize that kelp is a foundation species, and its removal greatly reduces the amount of habitat available for anemones, fish, and other dependent species. These arguments are more than academic. Bertness and company propose that facilitation may help devise strategies for dealing with invasive species, which cost billions of dollars a year in damages and can drive natives extinct. A large body of research suggests that diverse communities are less easy to invade because they are more competitive. But other research indicates that diverse communities can be easier to invade if they promote facilitative interactions (Science, 5 May 2000, p. 785). And if invaders facilitate one another, one invasion can open the door for others, precipitating what Daniel Simberloff of the University of Tennessee, Knoxville, calls an “invasional meltdown.” The TREE authors say conservation efforts that simply promote native diversity may therefore be doomed to failure. Managers might want to think twice, for example, before trying to restore grasslands by adding a diverse mixture of native species without knowing which are likely to facilitate invaders and which will enhance habitat for natives. ## Competing pressures Ecologists widely recognize and applaud the research by Bertness and his colleagues, but many question whether facilitation really is as important as competition or predation is. First in line with such questions is Paine himself. He praises Bertness's experimental work, but he questions his intense focus on habitat-forming species. “If you want to call that facilitation, fair enough, but it's boring,” Paine says. “His [Bertness's] current hobbyhorse … is much less studied and much less understood and much less experimentally tractable than the one that has made me famous—but not rich.” Paine argues that keystone interactions are actually the most important kind of positive interaction. Pisaster, by preying on the enemies of sessile invertebrates, facilitates those invertebrates. “When you add it, it's like hitting the system with a ball-peen hammer,” says Paine. In the kelp-forest example, Paine says that sea otters, not urchins, are the keystone. Otters eat urchins, which is what allows kelp to thrive and facilitate other species. Bertness's own research in fact underscores the importance of keystone species, Paine claims. In a study published online on 29 July 2002 by the Proceedings of the National Academy of Sciences, Bertness and student Brian Silliman attribute the decline of salt-marsh cordgrass in parts of the southeastern United States to herbivory by snails, which are plentiful because humans have overfished snail-eaters such as the blue crab. Paine calls this a clear example where a keystone species, the blue crab, is more important than the so-called foundation species, cordgrass. “I agree that our blue-crab work is a spectacular example of a keystone,” Bertness says. But “strong keystone species effects are almost always associated with predators controlling important foundation species.” To an outsider, the debate may seem like semantic wrangling, and some ecologists are inclined to agree. “I personally think the whole idea of positive versus negative interactions is not intellectually productive,” says Clive Jones of the Institute of Ecosystem Studies in Millbrook, New York. The struggle over which is more important “comes from a very strong desire: physics envy.” Ecologists would like to predict what happens in an ecosystem based on very simple data, he says, and Bertness and company may just be swapping the obsession with competition for an obsession with facilitation. Ecologists should focus on the conditions that foster positive and negative interactions, not on deciding which predominates, he says. Shahid Naeem of the University of Washington, Seattle, a veteran of a war of words over diversity in ecology (Science, 25 August 2000, p. 1282), says he is also bemused by the argument. One group focuses on the keystone species and the type of diversity it promotes, the other on foundation species and another type of diversity, he says: “But that's simply changing what you think of as diversity. … It serves us poorly to have people championing one cause over another.” In other words, strong words are no substitute for strong science. That may be one of the few points of agreement in this fractious discipline: Only creative, rigorous experiments can decide the outcome. “If you ask me if it's worth doing experiments [on facilitation], the answer almost certainly is ‘yes,’” Paine says. But ask him if he knows how they will turn out, and he answers, “I don't have the faintest idea.” 10. NEUROSCIENCE # Singing in the Brain 1. Greg Miller Researchers flocked here in December 2002 for the first international conference devoted to birdsong. New findings presented at the meeting shed light on the neural circuits that coordinate the intricate movements needed to create song NEW YORK CITY—Songbirds have long captivated certain humans. The English composer George Henschel, for instance, reportedly kept a highly trained bullfinch that sang “God Save the Queen.” Henschel was intrigued when an untrained canary kept in an adjoining room picked up the tune and finished it off properly whenever the bullfinch paused too long in midmelody. In recent decades, the fascination with songbirds has hatched a remarkably productive niche in neuroscience. By studying how male birds learn and produce their song (females generally listen and judge; see sidebar, p. 648), researchers have gleaned insights into the neural mechanisms of learning and motor control. Birdsong researchers were the first to discover that—contrary to decades-old dogma—new neurons can be born in the adult brain (Science, 3 January, p. 32). They've also revealed many mechanisms by which sex hormones set up differences between the brains of males and females during development. Despite all this interest, birdsong researchers had never come together for a conference of their own until last month, when 200-plus scientists from around the world gathered for a soggy few days at Hunter College in Manhattan. It felt something like a family reunion. The grand patriarchs of the field were there, including Peter Marler, whose work with sparrows in the 1950s pioneered the scientific study of birdsong; nearly all in attendance could trace their academic lineage to him. “It's like being at your wedding,” one researcher said. “Everyone you ever wanted to see in the whole world is there, but you only get to see them for 5 minutes.” Presentations covered everything from genetics to behavior to theories on song evolution. One area in particular, though, that has taken wing of late is research on the motor-control circuits in the songbird brain. New work has revised the view of how birdsong is produced and may yield clues about how the brain generates other types of sequenced behavior. Researchers also have found that a song-learning pathway in the bird brain has remarkable similarities to a crucial motor-control circuit in mammals, a finding that could lead to insights into brain evolution. ## Unraveling the song For years, researchers have suspected that a hierarchical chain of command exists in the songbird brain. According to one popular scenario, neurons in brain regions at the top of the chain serve as the conductor, issuing bursts of electrical activity like flicks of a baton to dictate the overall organization of the song. Neurons in midlevel areas, like the musicians, handle the details of which note gets played when—in this case, by sending commands to the bottom level, the brainstem regions that control the muscles that open and close the bird's vocal organ, the syrinx. In zebra finches, the lab rats of birdsong research, songs consist of roughly 20 syllables, each of which is a fixed sequence of several notes. Previous work by Albert Yu and Daniel Margoliash of the University of Chicago demonstrated that neurons in a forebrain area called the robust nucleus of the archistriatum, or RA, fire bursts of activity just before a particular note is sung (Science, 27 September 1996, p. 1871). Neurons in another forebrain area, known simply as HVC, fire rapidly throughout the song and modulate their firing rate according to which syllable is about to be sung, Yu and Margoliash found. This suggested that HVC is the maestro, RA the musician. But recent work from the lab of Michale Fee at Bell Laboratories in Murray Hill, New Jersey, presents a different view. Only a subset of HVC neurons sends signals to RA. Fee, reasoning that these so-called HVC(RA) neurons are likely the ones most directly involved in song production, improved his recording setup to zero in on these cells—something other researchers had been unable to do. Unlike the highly active HVC neurons described by Yu and Margoliash, the HVC(RA) neurons are fairly quiet, firing no more than a single burst during a song motif, a sequence of syllables that lasts about a second. The neurons' timing, however, is remarkably precise: Each one fires at a particular point in a motif each time it is sung. Simultaneous recordings from HVC and RA neurons confirmed that the message is getting through. Bursts from HVC(RA) neurons elicited bursts in RA, but RA stopped firing when a drug silenced HVC neurons. These findings, which Fee and colleagues Richard Hahnloser and Alex Kozhevnikov reported in fall 2002 in Nature, suggest that HVC(RA) neurons are indeed sending commands to RA—but they're not, as Yu and Margoliash concluded, ordering RA to play a particular syllable. Rather, the HVC(RA) neurons act something like the bouncing ball on the screen of a karaoke machine, keeping track of time and telling RA what to do from moment to moment. This may sound like a subtle difference, but Fee believes it has important implications for how birds learn to produce their song. This happens as a male compares his own song to a memorized version of his tutor's song. Because HVC represents time in a “sparse” way—each neuron is active only once per motif—if a bird needs to fix an error at a certain time in the song, it needs to tweak just the handful of HVC neurons active at that time, or about 1% of the total population, according to Fee's calculations. “It simplifies the learning process,” he says. Fee's team presented computer-modeling data that support this idea. The researchers found that the less active the HVC(RA) neurons are, the more quickly learning can occur. For example, with simulated HVC(RA) neurons that fire once per song motif, learning takes half as long as with HVC(RA) neurons that fire twice per motif. Sparse coding is an idea that's been floating around in the literature for some time, says Eric Vu, a neuroscientist at the Barrow Neurological Institute in Phoenix, Arizona, but the new findings are probably the best evidence yet that the brain actually uses such a system. And by suggesting that neurons can encode movement strictly in terms of time, the work adds an interesting twist to thinking on neural control of movement, which traditionally has focused on how neurons encode specific muscle contractions, Vu says. Margoliash is also impressed by the work. “Without a doubt, they've significantly changed and improved our understanding of the system.” But the case for sparse coding is not yet airtight, he says: “We don't yet know if all HVC(RA) neurons behave as Fee has described.” Moreover, he says, although Fee's model suggests that much of the learning takes place in HVC, previous work has shown that RA circuits change during learning. ## The brain's back roads The direct route from HVC to RA is the song-production freeway. It conveys the neural signals needed to generate song, and if it's disrupted at any time in a bird's life, song breaks down. Another more circuitous path from HVC to RA veers off into the anterior forebrain. This pathway is not essential for song production per se, but disruptions to it in juvenile birds cause serious deficits in song learning. Many researchers have been pecking away at this anterior forebrain pathway (AFP) in search of clues about the mechanisms of song learning. At the conference, several reported on their latest attempts to puzzle out its role in song learning. Although the picture is still somewhat fuzzy, one thing is coming into focus: The circuitry seen in songbirds is not unique. David Perkel of the University of Washington, Seattle, presented evidence that the AFP is wired up much like the mammalian basal ganglia. This is the network of brain nuclei that goes haywire in Parkinson's disease; it plays a key role in controlling movement and has been implicated in learning skilled movements. Anatomical experiments in the 1970s suggested a gross similarity between certain regions of the avian forebrain and the basal ganglia, but Perkel is the first person to describe the circuit in songbirds in cell-by-cell detail, says Harvey Karten, a neuroscientist at the University of California, San Diego. A main thrust of Perkel's recent work has been demystifying a neural black box in the AFP—a region aptly named Area X. He has found that Area × contains two classes of neurons that have remarkable similarities to components of the mammalian basal ganglia. One group of Area × neurons resembles neurons in the striatum; the other resembles those in the pallidum. Perkel has mapped out the connections of these neurons, described their electrical properties in detail, identified the neurotransmitters they release and respond to—and found that all of these properties are comparable to those of neurons in corresponding parts of the mammalian basal ganglia. In a series of recent papers, he described these findings and argued that birdsong researchers might be able to extract valuable lessons from the literature on mammals, where the role of the basal ganglia in learning has been well studied. More recent research presented at the conference suggests that even in birds, this network of neurons might not be dedicated exclusively to song learning. For example, Perkel's team reported that a similar circuit exists in chickens, for which no learned vocalizations have been documented. “I think it's reasonable to hypothesize that this is a generalized pathway for sensorimotor learning,” Perkel says. There may be an important lesson in these findings for students of brain evolution, says Karten: “If we could understand what the basal ganglia do in birdsong, that would be the first time we've understood the function of this ancient system in any nonmammalian vertebrate.” In mammals, the basal ganglia are thought to control movement through their connections with the outer layer of the brain, the cerebral cortex. But in birds, which lack an obvious motor cortex, the original function of the basal ganglia may be easier to discern. After all, says Karten, “Were the basal ganglia just sitting around for 400 million years waiting for the motor cortex to evolve? Not likely!” As the conference wound down, a roundtable discussion gave researchers a chance to voice their views on where the field was—and should be—headed. Many argued that songbird research needs to go genomic. Fledgling sequencing efforts at a half-dozen institutions have so far identified a total of about 75 zebra finch and canary genes, and the National Institutes of Health just awarded seven researchers$1.1 million over the next 3 years to help coordinate these efforts.

Others, including Marler, urged colleagues not to abandon the behavioral tradition in birdsong research in a rush to dissect its mechanisms with genetics and studies of neural firing patterns. Understanding behavior, he believes, will help researchers interpret the changes they see at the level of genes and circuits and help tie together different lines of investigation. “Let's not get so reductionist that we forget where it all began,” Marler said.

But at a basic level, the conference attendees clearly do appreciate the behavior of their subjects. Applause followed whenever a researcher played a snippet of song in the course of a presentation, and the occasional slide of a handsome zebra finch in midsong was greeted by “oohs” and “aahs.” Given the level of enthusiasm, it was hard to believe that this gathering was anything but the start of a tradition.

11. NEUROSCIENCE

# How to Please a Persnickety Female

1. Greg Miller

A great deal of research has been done on how the male songbird learns and produces his song, but scant attention has been paid to his feathered muse. Males sing to woo females, but researchers aren't sure what lady birds listen for in a song and why.

At the recent Hunter College birdsong conference in New York City, Stephen Nowicki of Duke University in Durham, North Carolina, presented recent work that may help explain one preference common among female songbirds: a soft spot for the boy next door. Like people, songbirds have local dialects. And just as a New Jersey accent doesn't always knock ‘em dead in, say, Mobile, Alabama, songs that don't adhere to the local dialect are a turnoff for female songbirds. Females, whether through inherited preference or a learned ability to recognize local songs, know what they're looking for.

The traditional explanation for this preference, Nowicki says, is that males with local-sounding songs are likely to come from a lineage that's been in the area for many generations. Thus they're likely to possess evolved traits that help them survive better in the local environment. A female looking for a mate would do well to secure some of the genes underlying these handy traits for her offspring, the thinking went.

But recent research by Nowicki and William Searcy of the University of Miami pokes a hole in that notion. Captured female song sparrows, for instance, discriminate only against the songs of males who live 30 kilometers or more away from the females' native area, Searcy, Nowicki, and colleagues reported last year in The American Naturalist. But in the wild, they found, the average song sparrow never wanders more than 5 kilometers from home. This means that most females never hear the songs of foreigners. In practice, then, it seems that the preference for local dialect wouldn't be useful for helping females reject potentially maladapted interlopers. So why do females prefer homegrown songs?

Nowicki thinks that what females are really looking for is a perceptive, intelligent male—or more precisely, one who has the wherewithal to make a faithful copy of the local song type. Males who do this will tend to sound local, true, but more importantly from the females' perspective, they may also be revealing a few things about their vigor.

For instance, Nowicki hypothesized, if a young male doesn't get enough food at critical times during development, the song system in his brain might not get wired up the right way. Males blessed with good genes (and genetically well-endowed parents able to provide adequately for their offspring) would be less vulnerable to this nutritional stress and sing faithful copies of songs they heard when young.

To test the idea, Nowicki recruited two groups of song sparrows shortly after hatching. He gave birds in one group as much as they wanted to eat and gave the other group 70% of that amount for 2 weeks. As adults, the food-deprived birds had atrophied song structures in the brain and hadn't copied their tutor's song as well, averaging only 16 syllables copied, compared to 20 in well-fed birds, he reported at the conference.

Wild female song sparrows notice the effects of past nutritional stress, Nowicki's team has found. The birds solicit copulation less frequently in response to poorly learned songs. The stressed males were smaller and had weaker immune systems, Nowicki told those at the conference, suggesting that females were wise to give them the cold shoulder.

The nutritional-stress hypothesis makes a lot of sense, says Clive Catchpole of the University of London. Because birdsong is such an intricate behavior, it may be a sensitive indicator of a male's fitness. Producing a song is a difficult task for the brain, and any additional challenge—lack of food, infections, or other types of stress—is likely to take a toll, Catchpole says.

Indeed, his group reported at the conference that adding the stress hormone cortisol to the food of zebra finches causes them to drop a few syllables from their song. “It doesn't sound like much, but the females don't like it,” Catchpole says.

12. SCIENCE POLICY IN JAPAN

# Researchers Feel Shut Out As Council Loses Its Clout

1. Dennis Normile

The once-vibrant Science Council of Japan has been eclipsed by a higher-ranking panel that some scientists complain speaks only for the elite

TOKYO—Rank-and-file Japanese scientists are losing their voice on the national stage. And a fight is under way over how—or even whether—they should regain it.

For the past 2 decades, scientists looking to influence government policy have turned to the Science Council of Japan (SCJ), whose members are elected by grassroots scientists voting through their professional societies. And the council has delivered: Its suggestions in the mid-1990s to create national laboratories that would focus on neuroscience and on advanced computing quickly became a reality, for example, and its 1994 report on the definition of brain death helped pave the way for public acceptance of organ transplants.

But the influence of the council, known as a “national assembly of scientists,” is now on the wane. Recent recommendations have been ignored, and a new, politically appointed Council for Science and Technology Policy has eclipsed SCJ in setting national science policy. “The impact of the SCJ has shrunk,” laments its president, Hiroyuki Yoshikawa, the former president of the University of Tokyo. In December, the new policy body delayed action on its proposal to reform the science council after several scientists complained that it would further dilute the council's clout.

What's at stake, say scientists, is the health of the country's basic research enterprise. Akiyoshi Wada, head of the RIKEN Genomic Sciences Center in Yokohama, says that the muffled voice of the science council has made it easier for the government to increase funding for research overly tilted “toward work expected to produce [economic] results quickly.” At the same time, says Yoshiki Hotta, director-general of the National Institute of Genetics in Mishima, “other mechanisms for the community to influence science policy are not well developed.”

SCJ was established in 1949, but until a 1983 restructuring, it was dominated by ivory-tower academics with little interest in public policy. A quasi-independent body that is nevertheless funded entirely by the government, SCJ is made up of 210 members elected every 3 years by a 2370-member liaison committee that represents 730,000 scientists from 1356 academic societies. Its working committees are made up of SCJ and liaison committee members. It puts out about 30 reports a year, with a budget of $11 million and a permanent staff of 62. The science council rose to prominence in the mid-1980s and remained influential for about a decade. But over the years, it became dominated by older and often retired researchers who, Yoshikawa admits, often put the interests of their society or field ahead of the scientific needs of the nation. Another unfortunate trend involves working committees that generate reports based on “discussions of their personal opinions” rather than on investigative studies, complains Kiyoshi Kurokawa, director of Tokai University's Institute of Medical Sciences and an SCJ vice president. Two years ago, legislators proposed abolishing SCJ as part of a wider reform. Although the council survived, its president no longer sits on the government's main science advisory body, which preceded the new policy council that reported to the Science and Technology Agency. And its overseer, which had been the prime minister's office, is now the Ministry of Public Management. To get back into the game, SCJ released a reform plan last April. A subcommittee of the policy council released its recommendations in November. Both reports stressed the importance of reviving SCJ. “We need the SCJ, which should provide unbiased scientific advice to the government,” says Shiro Ishii, professor emeritus of law at the University of Tokyo, who chaired the policy council's subcommittee. This role is expected to complement the more administrative activities of the 15-member policy council, which includes ministers and industry representatives in addition to a handful of scientists. But the two reports differ on how the science council should be organized. SCJ wants to convert itself into a body of 2500 leading scientists that would elect its own members. Yoshikawa says co-optation, as it's called, would encourage council members to represent the whole scientific community, not just their own disciplines. The council would then elect a 210-member managing committee, a self- sustaining body that might tap the larger pool for its expertise. The policy council's proposal would limit the science council to between 200 and 300 members and have them choose their successors. That approach looks like an attempt to further concentrate power at the top and make it “an organization of the elite,” says Wada. “Some people in the government don't like bottom-up input.” Ishii says the policy council is not trying to reduce the role of bench-level scientists. He thinks that an organization of 2500 scientists would be unwieldy and that a few hundred “is really the limit for rational discussions.” The policy council took the unusual step in December of deferring the recommendations of its subcommittee. Hiroyuki Hosoda, the minister for science and technology policy, is reportedly working behind the scenes to forge a compromise. In the meantime, Japan's scientific community will keep searching for a way to preserve its voice. 13. JOINT MATHEMATICS MEETINGS # Algorithmics = Has Trim Logic 1. Barry Cipra BALTIMORE, MARYLAND—Earlier this month, 5000 mathematicians converged here for the joint annual meetings of the American Mathematical Society and the Mathematical Association of America. Anagram enthusiasts love to take anything from a single word to a Shakespeare sonnet and rearrange its letters into something witty or weird. Dictionary-equipped computers capable of checking millions of possibilities in a flash added a powerful tool to their armamentarium. Now comes a novel twist: Noam Elkies, a mathematician at Harvard University, has found a way to use computers not to construct new anagrams but to discover ones that have been there all along. Elkies's algorithm allows him to take a large collection of words, phrases, or sentences and find two subsets that are anagrams of each other. Last year, he earned an “Anagrammy Award” at http://www.anagrammy.com/ and raised a hubbub in the alt.anagrams newsgroup by winnowing a compendium of clichés (including such classics as “a bird in the hand is worth two in the bush” and “every man has his price”) into two long lists so that each complete list anagrammed the other. Elkies has also found numerous examples in the periodic table, including one in which not only do the elements form anagrams, but the atomic numbers on each side also agree, both anagrammatically and arithmetically (see figure, above). In a talk at the joint math meetings, Elkies revealed how he does it. Given a collection of sentences, the algorithm starts by counting the number of A's, B's, C's, and so on in each item and recording the information in an ordered list (a “26-dimensional vector,” in mathspeak). For example, the typist's sentence “The quick red fox jumps over the lazy brown dog” becomes the vector (1,1,1,2,4,1,1,2,1,1,1,1,1,1,4,1,1,3,1,2,2,1,1,1,1,1), as it contains one A, one B, one C, two D's, four E's, etc. (The “periodic table” example uses 37-item vectors that record the letters in the element's name, the number of times each digit appears in the element's atomic number, and, finally, the atomic number itself.) After converting a big enough set of sentences or element names into vectors, the algorithm arranges the vectors into a regularly spaced array known as a lattice. The rest is pure mathematics. Through a technique called lattice reduction, for which fast algorithms were developed in the 1980s, the computer searches the lattice for two sets of vectors that, when added together component by component, yield the same sums in each place. That means that the sentences in each set have the same total number of A's, B's, and every other letter and symbol of interest. In other words, they are anagrams. There is no guarantee that such same-sum sets exist, but when the algorithm starts with a large collection of items, the odds are good, Elkies says. With an impressive string of “found anagrams” in the bag, Elkies is tackling a tougher challenge. Right now, his algorithm has to draw both sides of the anagram from the same collection of sentences. Elkies wants to expand that to two. By pitting, say, snippets of Emily Dickinson against passages from Stephen King, a playful user could find hidden anagrams “proving” that the two authors are one and the same, just as some literary sleuths have sought to do with Shakespeare and Christopher Marlowe or Francis Bacon. Adding this new wrinkle has been harder than expected, however. Thinking about high-dimensional vector spaces, after all, is not something people are naturally equipped to do. Or as Elkies, borrowing a bit of online abbreviation, puts it, “Homo sapiens = ape's son, IMHO.” 14. JOINT MATHEMATICS MEETINGS # How a Scribe Learned Math, ca. 1800 B.C. 1. Charles Seife BALTIMORE, MARYLAND—Earlier this month, 5000 mathematicians converged here for the joint annual meetings of the American Mathematical Society and the Mathematical Association of America. Suen-apil-Urim lost patience with the gods when he was forced to redo his homework. After writing out in tedious detail the results of multiplying 24 by various numbers, the aspiring Sumerian scribe ended his clay tablet with a flourish. “Praise Nisaba and Ea!” he added, in honor of two Sumerian gods. Alas, the scribe got some values wrong and had to repeat the exercise 4 days later. His second tablet mentions only Nisaba, leaving the water god Ea with no praise at all. Nearly 4 millennia later, Suen-apil-Urim's scribe-school exercises have given historians an unexpected new insight into the practice of ancient mathematics. The tablets, which scholars believe came from the Sumerian city of Larsa (near An Nasiriyah in the southeast of modern-day Iraq), show how long it took scribes to learn to multiply numbers. This is the first indication of how long scribes took to train. “It's a first for any scribal tradition,” says Eleanor Robson, an intellectual historian at the University of Oxford, U.K. “It's very important.” The clay tablets were among hundreds housed at the Ashmolean Museum in Oxford, but they were uncatalogued until last year. Sitting among more interesting-looking writings, such as incantations to prevent flatulence, Suen-apil-Urim's multiplication tables were overlooked, especially because Sumerologists have lots of similar tables. The students would create these tablets when learning multiplication tables “to show teachers that they've memorized the whole lot,” says Robson. Because of this written evidence, modern historians know a lot about how scribes learned mathematics in Sumer, says Stephen Tinney, a Sumerologist at the University of Pennsylvania in Philadelphia. “We have a good idea of the sequence. We know what they did and how they did it,” he says—down to the order in which they learned their lessons. But one critical piece of information was missing: “We had no idea how long it took them to go through the educational system.” According to Robson, who presented her findings at the Joint Mathematics Meetings in Baltimore earlier this month, that uncertainty may be over. She analyzed the two Ashmolean multiplication tables and realized that they are signed and dated. What's more, a third tablet, housed at Yale University, is also written by Suen-apil-Urim and dated. Instead of a 24-times table, though, like the Ashmolean tablets, the Yale tablet is a 4-times table, which comes somewhat later in the Sumerian math curriculum. (Scribes learned reciprocals, too, so it's a logical progression even though it seems backward to modern eyes.) The Yale tablet not only pinned down the year the tablets were created (1815 B.C.) but also showed Robson that scribes took 6 months to progress from learning the 24-times tables to the 4-times tables. Given the natural progression of the curriculum, Robson concludes that it took about a year for Sumerian scribes to learn multiplication. “This is a very interesting clue and very important if the extrapolation is correct,” says Tinney. Although he would like to examine the tablets himself to be sure, he says he believes Robson's conclusion. Suen-apil-Urim would no doubt be pleased that, nearly 40 centuries later, his hit-or-miss struggle to master arithmetic is helping scholars get their numbers straight. 15. JOINT MATHEMATICS MEETINGS # New Skating System Fails Virtual Replay 1. Charles Seife BALTIMORE, MARYLAND—Earlier this month, 5000 mathematicians converged here for the joint annual meetings of the American Mathematical Society and the Mathematical Association of America. Most mathematicians would prefer a double integral to a triple axel, but mathematical missteps could trip up new rules for awarding marks for competitive figure skating. According to a team of U.S. mathematicians, the International Skating Union's new judging methods are badly flawed. The International Skating Union proposed the changes in the wake of the scandal at the 2002 Winter Olympics, in which judges allegedly conspired to deny a Canadian figure skating team the gold medal. One key change is that the traditional panel of nine judges would be expanded to 14; five of those judges' votes would be randomly discarded. In theory, this would reduce the effectiveness of a corrupt judge or group of judges by raising the specter of their votes' not counting. But Elyn Rykken, a mathematician at Muhlenberg College in Allentown, Pennsylvania, and colleagues at two other colleges say the method has serious defects. “It's especially unfair and capricious for the competition,” she says. In computer simulations of real and fictitious skating events, Rykken and her colleagues showed that randomly tossing out five scores leads to dramatically unpredictable outcomes. In the 2002 Winter Olympics' ladies' freestyle event, for instance, American Sarah Hughes beat Russian Elena Slutskaya. But when the new rule is applied to the event (after systematically padding the judges' ranks up to 14), Rykken says, “Sarah Hughes comes in first about one-quarter of the time, while Elena Slutskaya comes in first three-quarters of the time.” An ideal judging method, she says, would yield an identical outcome for identical sets of judges' scores. “I understand that point of view,” says Roland Jack, the communications coordinator for the Lausanne, Switzerland-based International Skating Union. The new system might not be perfect, Jack acknowledged, but he noted that the same judges are eliminated throughout the whole skating program to make it as consistent as possible. 16. JOINT MATHEMATICS MEETINGS # Diagram Masters Cry 'Venn-i, Vidi, Vici' 1. Barry Cipra BALTIMORE, MARYLAND—Earlier this month, 5000 mathematicians converged here for the joint annual meetings of the American Mathematical Society and the Mathematical Association of America. Some things are so simple you'd think they'd have nothing new to offer. Take Venn diagrams. A staple of high school algebra, these diagrams use overlapping geometric shapes—usually circles—to represent the different ways two or three sets can intersect. What more is there to say? A lot, it turns out. Three mathematicians, including an undergraduate student, recently solved a 3-decade-old problem involving rotationally symmetric Venn diagrams. Imagine making such a Venn diagram with a rubber stamp, moving the stamp evenly around a circle N times. The result would look like a daisy with N petals overlapping at the center. The hard part is finding a petal shape that will result in all possible combinations of intersections. For which numbers of sets (or petals), mathematicians wondered, can such rotationally symmetric diagrams be drawn? Only prime numbers work, they quickly realized, but which ones? Now Carla Savage and Charles “Chip” Killian of North Carolina State University in Raleigh and Jerrold Griggs of the University of South Carolina, Columbia, have found a way to draw a diagram for any prime number of sets, no matter how large. Until 2 years ago, rotationally symmetric diagrams were known for only the first few primes. The familiar circular Venn diagrams with two and three sets fit the bill. There are many examples with five sets, including one made by rotating an ellipse (see figure), and many more with seven, although they were so hard to find that mathematicians initially doubted their existence. Two years ago, Peter Hamburger of Indiana University-Purdue University in Fort Wayne constructed an example for N = 11. It looked as though mathematicians might be in for an eternity solving the problem prime by prime. Fortunately, the new result takes care of everything at once. At a workshop* held in Baltimore a few days before the joint math meetings, Savage and colleagues described a systematic way of producing rotationally symmetric Venn diagrams of arbitrarily large (prime) size. Their proof, which produces snowflakelike patterns that Hamburger calls “doilies,” builds on a suggestion Hamburger made after constructing his N = 11 example. “We didn't have to do many new things,” Savage says. “When all the pieces were put together, it required only one new trick”: a clever way of ordering the intersections around the circle, which she credits to her then-student Killian (now a grad student at Duke University). “The solution is very elegant,” says Lenore Cowen of Tufts University in Medford, Massachusetts. Frank Ruskey of the University of Victoria, Canada, whose Web site survey of Venn diagrams (sue.csc.uvic.ca/∼cos/venn) has become a touchstone for researchers interested in the subject, agrees. “It's nice to have it finally resolved,” he says. The Carolina trio's result is not the last word on Venn diagrams, though. Their construction produces points where many curves come together. Partly for aesthetic reasons, but mostly for a new challenge, mathematicians now want to know if they can find rotationally symmetric diagrams with curves that meet only in pairs. Examples are known with two, three, five, and seven sets, but whether that continues for larger primes—even 11—remains to be seen. • * ALICE03 (Algorithms for Listing, Counting, and Enumeration), 11 January; sponsored by the Society for Industrial and Applied Mathematics. 17. JOSEP PIQUÉ INTERVIEW # Spain's Science Minister Sees Future in Telecom 1. Xavier Bosch* 1. Xavier Bosch is a science writer in Barcelona. Josep Piqué argues that his focus on innovation is essential to boosting his country's competitiveness, but it has left many fundamental researchers feeling piqued BARCELONA—When economist Josep Piqué was appointed Minister of Science and Technology last July during a Cabinet reshuffle, he knew he had a hard act to follow. His predecessor, Anna Birulés, had won respect for making Prime Minister José Maria Aznar's campaign promises in 2000 of more support for science a reality. Spain's science budget,$4 billion for 2003, has increased 28.7% in 3 years, and Birulés launched a clutch of innovative R&D efforts. Prominent among them is the lauded Ramón y Cajal program, which disbursed $300 million last year on tenure-track positions for some 2000 new postdocs at Spanish research centers. Birulés also had to whip into shape a superministry formed from three former ministries. Piqué is no stranger to Aznar's government, having spent nearly 3 years as foreign minister before the reshuffle. But this energetic 47-year-old Catalan is new to science and has already ruffled feathers over the government's response to the Prestige oil spill (Science, 24 January, p. 490) and over his desire to focus on telecommunications and other industrial areas of his portfolio. He has also spurned recent calls from academics who have urged him to try to persuade defense officials to relinquish a portion of Spain's hefty military R&D budget—estimated at$1.37 billion—for civilian research. In a wide-ranging conversation with Science, Piqué acknowledged that his top priorities are to boost innovation, telecommunications, and information technology. But he has pledged to fight to land the International Thermonuclear Experimental Reactor (ITER), a fusion-energy project that five countries are vying to host. An edited transcript, translated from Catalan, follows.

Q: Many people think that your ministry is preoccupied with telecommunications. Is it true that fundamental research takes a back seat to innovation?

A: We need a very potent telecommunications sector that helps us weave new technologies into the social fabric and make Spain an information society. It's logical that this receives the ministry's attention, particularly as it gets a lot of media attention. Conversely, basic science doesn't generate so much media attention.

Q: So science gets short shrift because it's not avidly consumed by the public?

A: Science only interests the public when it sparks a debate; take, for instance, the argument over embryonic stem cells versus adult stem cells. But this does not mean that basic science is not receiving enough attention from politicians. Last month, we hosted the seventh negotiating round of a great scientific installation, ITER. I have tried to give it the highest prominence I can. It would be good for our scientists if we could host it.

We are talking about a complex ministry that manages important resources, and so we have to pay attention to many things.

Q: Many European scientists think that Framework 6 [Europe's 5-year, \$17.5 billion flagship research program] is too bureaucratic and too industrially oriented. Do you favor the creation of a European Research Council [Science, 3 May 2002, p. 826] to boost basic science?

A: Make no mistake, research must be useful for society. That happens to a great extent, as basic science can feed into applied science, which in turn drives innovation and competitiveness. We have to strive to link the support of basic science to future applications. Logically, there is a problem of time: We cannot ask immediate results from basic science, and thus we have to find an equilibrium. As for Framework, I do share the criticisms over the excessive bureaucracy. Researchers should be doing research, not paperwork. But I do not think that a European Research Council is necessary.

Q: In 2001, the Catalan government launched Icrea, a program that, unlike Ramón y Cajal, aims to lure senior scientists working abroad back home. Have you considered a similar initiative for all of Spain?

A: We have had a conscious policy to recapture top-level scientists. This is perfectly compatible with Icrea, which is a fantastic initiative. But I would like to dispel the notion that the main problem of science in Spain is “brain drain.” This is not currently true. Now there are many more scientists from abroad working in Spain than there are Spanish scientists abroad. But I think that one challenge we have, in addition to assuring that prestigious researchers can develop their work in Spain, is how to lure more foreign scientists to our country.

Q: The government has come under fire for what is perceived as a disproportionately high level of spending on military R&D. What is your reaction?

A: It's a highly controversial issue and highly polarized. But in serious countries, this is not a matter of discussion. Nobody doubts that military research must be done. Whoever advocates to the contrary must know that he or she is telling Spain to resign its status as a first-class country. I know that what I am saying is politically incorrect. But I have long insisted that Europe as a whole and Spain in particular must make ourselves responsible for European security; we cannot count on the United States forever.

Q: So are you content with the current level of military R&D spending?

A: I want to see it sustained and, if we are able, to increase it in the future.

18. EVOLUTION

# Puzzling Over the Origin of Species in the Depths of the Oldest Lakes

1. Erica Goldman

The creatures in Baikal and other ancient lakes could help researchers unravel fundamental mysteries about why some life forms speciate and others do not

After 3 days of sorting through numbingly cold mud from the bottom of Lake Baikal in southern Siberia, Oleg Timoshkin was growing restless. Timoshkin, a biologist at the Limnological Institute in nearby Irkutsk, and his fellow scientists on this most venerable of lakes on a warm July afternoon had yet to find anything interesting.

Then at about 3 o'clock, the scientists aboard the Vereschagin, a vessel named after one of the first Russian biologists to study the lake, hit the jackpot. The next slug of mud winched to the surface was teeming with densely packed white mollusks, pink amphipod crustaceans, and writhing flatworms. Timoshkin, a flatworm expert, rushed some of the brownish goo to the ship's lab and placed it under a dissecting microscope. A few painstaking hours later, he tweezed from the muck a tiny, translucent worm with a yellowish-green stomach—a kind of specimen that he had never seen before in more than 2 decades of work. Timoshkin later confirmed that the mystery beast was a new deep-water subspecies, Geocentrophora wagini abyssalis, one of roughly 140 types of Baikal flatworms that exist nowhere else in the world.

Welcome to a freshwater horn of plenty, where scientists have racked up an average of 20 new species a year over the past decade while probing the oldest and deepest lake on Earth. The diversity of life forms in Baikal and other ancient lakes makes them “unique crucibles of evolution,” says Lisa Park, a paleolimnologist at the University of Akron, Ohio. The Speciation in Ancient Lakes group, which includes several dozen specialists from around the world, met last autumn in Irkutsk to mull over the forces that shape these biological hotspots. Researchers unveiled fresh findings on the genetics of ancient lake species that are helping to unravel why some creatures are prone to speciation and others are not.

## Forced to speciate

Baikal is one of a dozen present-day lakes that have persisted for at least 1 million years; most lakes clock out at about 18,000 years. And Baikal's biological diversity—more than 2500 identified faunal species, primarily invertebrates that are nearly all unique to this body of water—is unsurpassed, says Risto Väinölä of the Finnish Museum of Natural History in Helsinki. “We are seeing diversity that you would expect over a continental scale, taking place within a single lake,” he says. Illustrating that point at the meeting, Väinölä described new molecular data on amphipods, colorful, voracious scavengers that in Baikal can grow to a world-record length of 9 centimeters. Some of the lake's 350 amphipod species that look alike and live in the same sections of lakebed are in fact genetically divergent, Väinölä has found. The finding, he says, implies that there may be three times as many amphipod species in Baikal as are currently enumerated.

That only deepens the intrigue over the complex set of forces behind Baikal's startling diversity. One obvious factor is the lake's longevity: An estimated 20 million to 25 million years have passed since its primordial waters first began to pool in the rift formed when the central part of the Eurasian landmass began to separate from the smaller southern plates. In contrast, “most lakes are by nature suicidal,” mere water holes that accumulate sediments, turn into swamps, and become land, says evolutionary biologist Koen Martens of the Royal Belgian Institute of Natural Sciences in Brussels. Baikal has accrued a thick sediment layer that offers a window on ancient climates (see sidebar), but it is its depth—1637 meters at the deepest point—that is the lake's fountain of youth. That Baikal straddles a rift also helps prolong its life: Its lakebed is sinking by about 20 millimeters a year, roughly equivalent to its sediment buildup. The next two oldest and deepest lakes, Africa's Tanganyika at 9 million to 12 million years old and Malawi at roughly 3 million years old, are situated in rifts as well.

The peculiar geology of rifts—in particular, Baikal's patchwork of substrates—appears to influence speciation. For example, in the amphipod genus Plesiogammarus, six closely related species have staked out different turfs. P. zeinkowiczii, with its big eyes and long antennae, lives atop coarse, freshly laid sediment, whereas P. brevis, with its slitlike eyes and short antennae, burrows several centimeters into a siltier layer. Its cousins in between, not surprisingly, have eyes and antennae that are medium sized. As an animal adapts to a substrate type, it may lose the ability to breed with cousins on other substrates, even in close proximity, says evolutionary biologist Ellinor Michel of the Institute for Biodiversity and Ecosystem Dynamics in Amsterdam. Thus, Baikal may offer a clutch of rare examples of sympatric speciation, in which species can arise from an ecological specialization, rather than a physical barrier to gene flow.

## Driven to speciate

Also attracted to Baikal's biological bonanza are researchers probing the genetic underpinnings of speciation. In collaboration with Evgeny Sverdlov of the Institute of Molecular Genetics in Moscow, a team led by molecular biologist Dmitry Sherbakov of the Limnological Institute is using a novel approach to hunt for rapidly evolving DNA regions in Baikal species with relatively small genomes, including the omul fish and the amphipods. The researchers combine the DNA of two closely related species and isolate stretches that do not stick together. These leftover regions are the ones that have diverged. “By finding the genes that evolved fastest, we can identify those responsible for adaptation to environmental variables such as depth and cold,” explains Sherbakov, who described their findings at the meeting. For instance, some species of amphipods have an armor coat that appears to have evolved in parallel many times. The genes that encode armor may be among those that have evolved the fastest, he says.

Other ancient lakes may also contribute to this effort to tease out the genetic triggers of speciation. There are more than 1400 species of cichlid fish in three African lakes—Malawi, Tanganyika, and Victoria—representing the biggest radiation of any animal genus within a geographic region. Cichlids give scientists an “evolutionary playing field” on which to test hypotheses, says Axel Meyer of the University of Konstanz, Germany. Meyer, an evolutionary biologist, is homing in on quickly evolving DNA regions in the cichlids from Lake Tanganyika. He's looking for genes that encode differences in traits such as head shape and body color. Perceiving color is critical to cichlids: Females choose mates according to the color pattern of the males, which varies by species. So far, Meyer has found that a gene known to cause color loss in some mammals varies among cichlid species.

Color vision drives cichlid speciation, and a loss of ability to discern color patterns appears to be accelerating the demise of species in Africa's Lake Victoria, says Ole Seehausen, an evolutionary ecologist at the University of Hull, U.K. Today, there are roughly half of the 600 to 800 species that were present just 50 years ago. Many have fallen prey to Nile perch, which was introduced into the lake in the early 1950s. But for decades, Victoria's waters have grown steadily murkier, in part due to algal blooms linked to the loss of cichlids.

Because the cichlids have evolved so recently, species are still capable of interbreeding and producing viable offspring. In clearer waters in Victoria, species stick to their ecological niches and generally steer clear of one another, mating according to color pattern, Seehausen explains. But the murky waters have now made female cichlids blind to love. As a result, mating between species is now rampant, further reducing the lake's already crippled cichlid diversity (Science, 19 September 1997, p. 1808). Victoria and other ancient lakes “are poised to suffer a great loss” of biological diversity, says George Coulter, an African lakes specialist now retired in New Zealand. Conservation work, he says, must go hand in hand with research.

Further insights into the mechanisms of speciation could be gleaned from more comparative work across the ancient lakes, scientists say. For example, Park of the University of Akron and Elizabeth Gierlowski-Kordesch of Ohio University in Athens are collecting data on animal life from 35 present-day and now-dry fossil lakes with life spans exceeding 100,000 years. Their efforts so far suggest that the total number of species in ancient lakes may have been underestimated by as much as 50%, Park says. They hope to unveil their database before the next ancient lakes jamboree, set for 2006 in Berlin.

There should be plenty more findings to reveal then as well. Like its other long-lived brethren, Baikal and the secrets it may reveal about how species arise have scientists coming back, year after year, to its enchanted waters.

19. EVOLUTION

# A Window on Ancient Siberia

1. Erica Goldman

For decades, scientists have been mining the ocean floor for clues to past climates found in gases and fossils locked in the sediments. But it wasn't until 1989 that Douglas Williams, a geologist at the University of South Carolina (USC), Columbia, was able to mount an effort to extract such a record from a continent's interior. With Mikhail Kuzmin of the Institute of Geochemistry in Irkutsk, Russia, Williams set out to reconstruct Siberia's past from the muck at the bottom of Lake Baikal.

Perhaps the most provocative insights have come from studying the minuscule remains of diatoms whose silicon skeletons have accumulated in the sediment over the eons. Severe cooling can drive these single-celled algae to local extinction, whereas warming nurtures reappearance and diversification. These biological proxies for climate change have revealed that during some periods when ocean cores showed milder cooling over the past 8 million years, Baikal was in the throes of glaciation.

Past climate changes can also be inferred from the ratio of carbon isotopes in the sediments. Methane hydrates, found for the first time in fresh water in Baikal in 1997, are frozen, cagelike arrangements of methane and water. They contain a preponderance of carbon-12, which leaves a distinct signature in the composition of plankton and sediments. When the climate warms, methane is unlocked and released into the lake and eventually the atmo sphere. “It is somewhat like opening a Coke can,” says USC geologist Alexander Prokopenko. Thus a deficit of methane hydrates in the cores indicates regional warming.

Although the Baikal Drilling Project ended 4 years ago, its findings on everything from ancient temperature regimes to the rise and fall of diatom species are still cascading into the literature. The project has also provided the impetus for drilling at other ancient lakes: The International Continental Scientific Drilling Program, a nine-nation consortium, will sponsor an expedition next year to Lake Malawi to retrieve the first ancient sediments from the continental tropics that, researchers hope, will contain a fossil timeline of evolution in this African ancient lake.