# News this Week

Science  25 Jun 1999:
Vol. 284, Issue 5423, pp. 2062
1. SCIENTIFIC PUBLISHING

# Varmus Defends E-biomed Proposal, Prepares to Push Ahead

1. Eliot Marshall

This summer, many biomedical editors and publishers are wondering how their journals will survive if the government goes ahead with a plan to distribute biomedical papers for free on the Internet. But such qualms do not trouble the plan's author, Harold Varmus, director of the National Institutes of Health (NIH). He's charging forward with “E-biomed,” as he calls it. His idea is to create “free, fast, and full access to the entire biomedical research literature” for anyone with a computer and an Internet connection. E-biomed would distribute unedited preprints, as well as articles that have been through the traditional mill of peer review. Varmus's reason for doing this: Taxpayers have paid for the research already, he says, so NIH should make the results widely available.

Varmus released a written description of E-biomed in April (Science, 30 April, p. 718), and since then he has defended the proposal in public and private venues. His most recent defense came in an “addendum” posted on NIH's Web site on 21 June (www.nih.gov/welcome/director/ebiomed/ebiomed.htm). This six-page memo suggests that the planning is moving toward a dramatic climax. Indeed, in an e-mail response to questions from Science, Varmus said “we are in the process of assembling” cost estimates to submit to Congress.

The addendum also indicates that the venture has picked up key support from Europe. Both the European Molecular Biology Organization and the European Molecular Biology Laboratory have expressed their support, Varmus writes. “We are discussing a potential partnership with them,” he says, which would allow joint development of technology and “encourage other organizations to collaborate.” Varmus told Science that “we agree on the basic principles” and that the Europeans “are presenting the issues to their boards” right now. He hopes to create an international governance structure. Varmus is also sounding out private backers: In mid-June, he met with Vitek Tracz, chair of Britain's Current Science Group, which publishes Internet-distributed journals, but they apparently did not reach any agreement.

Despite Varmus's forceful advocacy, E-biomed has been taking flak in the past month. On 2 June, Varmus met privately with leaders of journals published by the Federation of American Societies for Experimental Biology (FASEB), who worry that the project could undermine not just their societies' revenues but cherished traditions of scientific publishing. Those who attended say Varmus seems determined to launch E-biomed in some form in the next 9 months and describe the session as “tense.” “We were not reassured,” says FASEB publications committee chair Ed Rekas. Individual researchers appear to be more receptive, but in hundreds of responses to the original proposal posted on Varmus's Web page, many worry about the threat to peer review, the need for editorial independence, and how E-biomed would be financed.

The New England Journal of Medicine has weighed in heavily, firing a blast at E-biomed in its 10 June issue. In an editorial, former editor Arnold Relman wrote: “A system that allowed immediate electronic publication of new clinical studies without the usual careful process of peer review and revision would be risky at best and might well fill the clinical databases with misleading and inadequately evaluated information.” He suggested that E-biomed might undermine the clinical journals “enough to threaten their survival.” The American Physiological Society, the Journal of Immunology, and other society-based publications have expressed similar concerns. But the American Association of Pharmaceutical Scientists is “enthusiastic” about E-biomed, its president Larry Augsburger wrote, as are some other clinical groups and even basic science organizations within FASEB, like the American Society for Cell Biology (ASCB). Elizabeth Marincola, ASCB's executive director, says, “My feeling is that a society like ASCB has more to gain than to lose” from E-biomed, as “we are not making any money on our journal.”

Varmus released the addendum to his proposal to try to allay the worries and to review some of the many unresolved practical issues. He writes that E-biomed “most emphatically” would not eliminate peer review or create “vast quantities” of useless data. Existing journals, he hopes, will “establish peer-reviewed electronic journals operating within E-biomed.” And although the system will permit the posting of unreviewed material, Varmus argues that “few scientists would knowingly” put sloppy reports in the public domain, “because it would soon diminish their reputations.”

But tough questions remain unanswered, including: Who will run the operation, and who will pay for it? It is “an unfortunate misreading” of the proposal, Varmus writes, to think that the government will be in charge. “It would not be owned by the NIH or any other component of the U.S. government.” NIH would only provide technical and financial support. But Varmus leaves many details to be filled in by a proposed E-biomed governing board, whose authority and composition remain undefined. Critics are annoyed by what one editor calls this “foggy” aspect of the plan.

As for financing, Varmus notes that E-biomed could charge authors a fee—perhaps a low one to handle a simple submission and a higher one for peer-reviewed publication. How high? That will depend on several factors, Varmus says, “but will likely be in the range of $100 to$1000 per article” to cover the participating journals' costs. However, publishers of some FASEB journals report that they already have costs in the range of $1000 to$4000 per page, and that converting from subscriptions to a page-charge method of financing would drive authors away. FASEB members were not reassured when Varmus suggested in a meeting—as in the addendum—that societies should find other ways of raising money, such as raising meeting fees. As one observer said, “People didn't appreciate being told they should go out and sell Girl Scout cookies.”

In general, editors who liked the original E-biomed idea are enthusiastic about the addendum; those who didn't are as cool as ever. But many society chiefs are reluctant to sound off in public, says Michele Hogan, executive editor of the Journal of Immunology. The E-biomed plan casts them as defenders of the status quo, even though many journals have led the way to e-publishing, she says: “We're a little afraid of how the scientific societies are going to look.” Some will have a chance to air their views at a “summit meeting” of electronic publishers being held at the National Academy of Sciences in Washington, D.C. this week. But no matter what attendees think of E-biomed, says Marincola, it appears that Varmus considers this “possibly one of the most important things he's done as NIH director.”

# Library-Society Alliance Puts Bio Journals Online

1. David Malakoff

Research libraries are taking their fight to hold down the cost of journals into cyberspace. This week the American Institute of Biological Sciences (AIBS) announced it will work with major U.S. research libraries and a private printer to distribute electronically dozens of nonmedical biological journals published by its 55 member societies. The arrangement is aimed at allowing smaller societies—some of whose journals may be threatened by the advent of preprint servers (see previous story)—to stay in the publishing business, while giving libraries a say on subscription prices, which are rising much faster than inflation, and access to archival material. But some observers predict the collaboration will face many of the same challenges confronting other scientific publishers that have gone online.

View this table:

The joint venture will create a collection of Web material, called BioONE, that will debut in 2001. BioONE will initially offer about 50 journals published by AIBS member societies, which range from the 6000-member Ecological Society of America to the American Fern Society, with fewer than 1000 members. Most of the journals, including the AIBS's flagship BioScience, do not now offer full text online. Without financial help, many of the member societies might be forced to lease or sell their relatively low-cost journals to for-profit publishers, says Rick Johnson of the Scholarly Publishing and Academic Resources Coalition (SPARC), an organization of research libraries. “What's motivating us is the plight of the small society,” he says. “If their journals can't make the jump to electronic dissemination, [the society] may get squeezed out of publishing.”

Research librarians have become increasingly alarmed in recent years about rising journal prices. Since 1986, median prices for journals issued by both commercial and nonprofit publishers have increased by more than three times the rate of inflation, according to the Association of Research Libraries (ARL) in Washington, D.C. which represents more than 120 collections in the United States and Canada. As a result, library budgets are being stretched to the breaking point to cover key commercial titles that can cost up to $15,000 annually. In 1997 ARL officials formed SPARC, which has helped to sponsor several new journals that compete head to head with pricier existing titles (Science, 30 October 1998, p. 853). SPARC and its allies want to promote competition by helping small, cash-strapped scientific societies jump into online publishing. Over the next 2 years, it will work with the Washington, D.C.-based AIBS, the University of Kansas and Allen Press, both in Lawrence, Kansas, and the Big 12 Plus Libraries Consortium, a group of 23 midwestern research collections, to create a Web-based service that could eventually provide access to 200 journals in the biological, ecological, and environmental sciences. Although the framework for BioONE is still under discussion, its organizers hope to raise$750,000 in start-up funds and in-kind donations from foundations, libraries, and other university departments, says Johnson. The University of Kansas, for instance, is planning to donate the technical expertise and computers needed to store and operate the database. Campus staff will also work with nearby Allen Press—which already prints more than 100 AIBS journals—to prepare papers for Web publication. The company will be paid for its work.

In return for their support, libraries will have a significant voice in setting BioONE's subscription prices and access policies. One major issue involves electronic back issues. Although libraries routinely store past issues of a printed journal, they are sometimes denied access to archives of electronic publications once a subscription lapses. At the same time, societies are seeking assurances that Web publishing won't shrink revenues by reducing the number of library subscriptions to their print journals.

BioONE isn't the only Web journal publisher looking for a winning formula. In recent years more than a dozen groups—both commercial and nonprofit—have become online “aggregators” of scientific journals, creating Web pages that allow subscribers to retrieve papers from a large number of related titles. The American Chemical Society, for example, has created ChemPort, which provides access to dozens of chemistry journals, while HighWire Press of Stanford University Press in California has created a site with some 130 crosslinked online journals, including Science Online. Commercial giant Elsevier Science also has backed several sites, including BioMedNet and ChemWeb, stocked with dozens of its titles.

But many aggregators continue to hunt for the right combination of pricing, advertising, and access policies. And some predict BioONE may have trouble satisfying everyone. Librarians, for instance, may be loath to sign up for the whole collection when they now have the freedom to select individual journals. “It should be a very interesting experiment—they will be wrestling with the same economic issues we do,” says Don Muccino, executive vice president of the Online Computer Library Center in Dublin, Ohio, a library-backed nonprofit aggregator that puts more than 1600 journals online.

BioONE backers, however, are confident they can devise a workable solution that others may want to emulate. Says the University of Kansas's Beth Forrest Warner: “We're real excited about the possibility of breaking some new ground here.”

3. SCIENCE IN SOCIETY

# Panel Discounts Implant Disease Risk

1. Jocelyn Kaiser

A blue-ribbon panel has concluded that silicone breast implants do not increase the risk of diseases such as lupus or cancer, rejecting a theory invoked in countless claims against implant manufacturers. But the report, released earlier this week, is unlikely to be the last chapter in the lawsuit-weary saga: The Institute of Medicine (IOM) panel cites evidence that silicone implants can leak and cause infections or painful scarring around the implants.

Anecdotal reports blaming implants for serious health problems first arose in the late 1980s and led to billions of dollars worth of legal claims against manufacturers. Most of those claims are now being resolved, as Dow Corning, Bristol-Myers Squibb, and other manufacturers have agreed to create settlement funds totaling about $7.2 billion. In the meantime, studies on implants and chronic disease risk have been coming up empty-handed. The IOM panel “is simply saying over again what we already knew—that the case for autoimmune disease was extremely weak,” says Yale University immunobiologist Charles Janeway. But he and others say the imprimatur of the nation's top medical advisory body gives that conclusion more weight, shifting the scientific focus and legal battleground from systemic disease to local problems caused by ruptured implants. The IOM stepped into the thorny arena of implant science in late 1997, at the request of the Department of Health and Human Services. The 13 panelists, led by Stuart Bondurant, a professor of medicine at the University of North Carolina, Chapel Hill, examined some 2000 peer-reviewed studies and 1200 other data sets and reports, searching for links between implants and lupus, rheumatoid arthritis or other connective tissue diseases, cancer, or neurological diseases. The committee also heard testimony from sick women and was “moved by their suffering,” it said. But the touching personal stories failed to sway the panel's views on the data. It concluded that the 1.5 million to 1.8 million U.S. women with implants are no more susceptible to serious diseases than are women without implants, according to available evidence. This conclusion, the panelists noted, is consistent with reports in the past year from U.K. experts and a panel appointed by a U.S. judge overseeing breast cancer litigation (Science, 11 December 1998, p. 1963). But the panel did not give implants a clean bill of health. Essentially plastic bags filled with silicone gel, implants can rupture in an unknown percentage of women—studies have cited rates as low as 0.3% or as high as 77%. The pain of breast tissue contracting around implants, as well as infections and other health risks from surgery to replace implants, are “the primary safety issue[s] with silicone breast implants,” the report found. The panel recommends more research to track women with implants to get a better handle on problems such as rupturing and second surgeries, and improved tests to gauge silicone concentrations in body fluids and tissues. Some observers who are not ready to dismiss the disease threat say they are waiting for the results of a major epidemiological study on women with implants by the National Cancer Institute (NCI), due out later this year. That study, however, may be predetermined to find problems, says IOM president Kenneth Shine: Materials for recruiting participants may have “encouraged women with symptoms and problems to enroll,” he says, rather than gathering a sample that would include healthy women with implants who could serve as controls. NCI study leader Louise Brinton responds that “there has been some misrepresentation of our study,” but she declined to address its design until the findings are published. The IOM report may carry weight, but it will not be the last word on the contentious issue of silicone implants and health. 4. GENE THERAPY # RAC Urges Changes to Retinoblastoma Plan 1. Ken Garber* 1. Ken Garber is a writer in Ann Arbor, Michigan. Since losing its approval authority over gene therapy protocols 2 years ago, the Recombinant DNA Advisory Committee (RAC) has been out of the news—and out of the minds of many people working in the field. 1997 appointee Jon Wolff says that some colleagues have been baffled by his RAC activities. “People's first reaction was, ‘Is it still in existence?’” he says. But the RAC may be regaining its clout. One sign of life came last week when the committee turned thumbs down on a gene therapy protocol for treating retinoblastoma, a rare childhood cancer of the eye—and the lead investigator listened. Also last week, RAC got a strong endorsement from National Institutes of Health (NIH) director Harold Varmus, who 3 years ago proposed doing away with the committee but then, in response to much protest, agreed to keep it to advise on policy matters. Speaking to gene therapists assembled in Washington, D.C. for their annual meeting, Varmus delivered a barely veiled threat. Gene therapy protocols now need only the approval of the U.S. Food and Drug Administration (FDA). But in an otherwise optimistic talk, he warned that the “double approval” process that ended 2 years ago might be restored if researchers don't submit their proposals simultaneously to RAC and the FDA. “Departure from standards of gene therapy must be publicly discussed,” Varmus said, emphasizing that the RAC is the proper forum for hashing out the scientific, ethical, and societal issues surrounding new forms of gene therapy. NIH isn't likely to restore protocol approval power to the RAC, but the director did capture his audience's attention. “Varmus's talk really put us back on the map, in terms of clarifying our role,” says Wolff, a gene therapist at the University of Wisconsin, Madison. Such clarification may be needed. NIH records show that since 1997 about 10% of gene therapy protocols haven't gone to the RAC. Others, including the one for the retinoblastoma trial, came in late. The trial's lead investigator, Richard Hurwitz of Baylor College of Medicine in Houston submitted his protocol to the RAC only after the FDA had approved it, he says, because of a “completely inadvertent” oversight. When the RAC finally reviewed the protocol, safety was the big sticking point. The strategy is not new. It involves injecting the eye tumors with an adenovirus vector that carries the herpes simplex virus thymidine kinase gene into cells, making them susceptible to killing by the antiviral drug ganciclovir. The goal, Hurwitz says, is to reduce the size of the tumors at least to the point where they can be removed by freezing or by laser surgery. Standard therapy, which almost always cures the disease, is to remove the eye. Gene therapy holds out the promise of saving the eye and some vision. But the trial also raises a policy issue—or so RAC members believe. In the words of RAC consultant Pedro Lowenstein of the University of Manchester, U.K. this may be “the first gene therapy protocol focused on improving quality of life” as opposed to just curing disease or disability. Although removal of an eye seems draconian, retinoblastoma expert Thaddeus Dryja of Harvard Medical School in Boston describes its impact as “gratifyingly tolerable.” Thus, the Baylor trial involves treating babies, whose average age is 18 months, for a condition that already can be cured. RAC members cited several risks: The needle could lead to the cancer's spread through the blood vessels; the adenovirus vector might trigger inflammation that would damage the diseased eye and perhaps the normal eye as well; and ganciclovir could damage normal tissue. “I question whether we are introducing a very dangerous protocol … or one with unknown risk, in something that usually, with standard care, can be 100% cured,” said RAC member Louise Chow of the University of Alabama, Birmingham. Consequently, the RAC voted unanimously (with four of 13 members abstaining) to urge Hurwitz to treat only patients with tumors in both eyes, because these children face blindness anyway, making the therapy's risks easier to justify. Hurwitz argued that retinoblastomas do not usually metastasize through blood vessels, and that Baylor's eye surgeons can largely avoid them, anyway. As for the risk of inflammation, he notes that the aim of this first trial is precisely to see whether the gene therapy produces such toxic effects. Hurwitz also pointed out that evaluating adverse effects would be harder in patients with bilateral retinoblastoma because they, unlike the patients he proposes to treat, have already had potentially toxic therapies. FDA reviewers deemed the protocol safe enough to proceed, although they would have preferred that Hurwitz treat bilateral retinoblastoma. “A close call,” remarks the FDA's Philip Noguchi. But Hurwitz will comply with the RAC's decision—“at least [with] our first few patients,” he says. “We want to proceed very carefully.” After that, he's keeping his options open. With gene therapy trials expected to increase by up to 25% this year, the RAC may see more controversial protocols. “I think there are some real cowboys out there,” says Wolff. But will Varmus have to make good on his threat to restore teeth to the RAC? “We don't want to go there,” says NIH science policy director Lana Skirboll, speaking on behalf of Varmus, who was unavailable for comment. Much depends, she says, on how the gene therapy community responds. 5. IMAGING # 3D Camera Has No Lens, Great Depth of Field 1. Daniel Radov* 1. Daniel Radov is a free-lance writer in Brookline, Massachusetts. The traditional camera is a threatened species. Digital cameras, which replace photographic film with electronic light detectors, are on sale at your local photo shop. Lensless cameras, in which a computer does the job of the lens and digitally processes light to make an image, are taking shape in the lab. And in this issue of Science, the camera takes another step away from its roots. Since the days of Louis Daguerre, cameras have captured reality in two dimensions. But the lensless camera that a team of electrical engineers at the University of Illinois, Urbana-Champaign, describes on page 2164 makes the jump to three. By bathing an object in ordinary light, rotating it on a stage, and recording the interference of thousands of pairs of light rays reflected from or transmitted through the object, the system builds up a 3D representation that captures far more information than a hologram or stereo images. The “lens” responsible for this feat is a pair of mathematical algorithms, one borrowed from radio astronomy and the other from x-ray imaging. Other researchers are impressed, saying the technique could capture cells and tissues in three dimensions and give depth to machine vision. Kelvin Wagner, an electrical engineer at the University of Colorado, Boulder, who is familiar with the work, calls the group's method “an amazingly elegant way of turning the problem from something very messy into something far simpler.” The technique grew out of a mathematical insight that joined two traditionally separate imaging tools. One, widely practiced in radio astronomy, is interferometry, in which radio waves collected by separate dishes from the same point in the sky are allowed to interfere. The waves' interference can be translated into the position and intensity of their source, and combining interference data from many different points yields precise 2D maps of quasars, supernovae, or galaxies. The other mathematical tool is tomography—the T in x-ray CT scans, which pinpoint the body's internal structures by analyzing x-rays sent through the body along many different paths. The mathematical match was made when Daniel Marks, an Illinois graduate student, noticed that applying a mathematical tool called a Fourier transform to interference measurements would yield a data set ready-made for a particular type of tomographic analysis. By dovetailing the tomographic and interferometric algorithms and applying both of them to visible light, the group came up with its 3D lensless camera. The imaging begins as the object—in this case a small plastic dinosaur—is rotated in front of an interferometer. For each viewing angle, the interferometer collects light that follows many different paths from each point on the object, filters it, and allows it all to interfere. The result is a pattern of light and dark spots, captured by an array of detectors. Then the algorithms kick in. First comes the analysis of the interference data, which transforms it into a two-dimensional projection— something like a shadow of the object. Next is the tomographic algorithm. It analyzes the two-dimensional projections, each one analogous to the x-rays collected along a single viewing direction in CT scanning, to build up a set of image slices representing the three-dimensional object. The tomographic algorithm was designed for x-rays, which pass virtually unhindered through most tissues, but the Urbana-Champaign team has found that it also works surprisingly well for light reflected from opaque objects, allowing them to map the surface of their dinosaur with a resolution better than 1 millimeter. And the result is far richer than a hologram, which is made by recording the interference patterns of laser light, says David Brady, an electrical engineer with the Illinois group. Holography, which generally does not scan all viewing angles, “is not really a 3D imaging technique; it's a 3D perspective preserver,” explains Brady. Thomas Cathey, an electrical engineer and colleague of Wagner's at Boulder, cautions that the technique may be too slow for use in real-time applications such as robotic vision or automated quality control in manufacturing. But George Barbastathis, an electrical engineer at the Massachusetts Institute of Technology, thinks that for imaging biological samples, the new system could ultimately surpass techniques such as confocal microscopy, which builds up 3D images by illuminating and imaging samples point by point. “Confocal systems acquire intensity data one point at a time,” notes Barbastathis. “With Brady's method you scan in parallel. If they can manage to make the resolution comparable to confocal microscopy—and I believe that with their method it's actually possible to make the resolution better—then in that case it wins hands down.” 6. BIOPATENTS # Legal Fight Over Patents on Life 1. Eliot Marshall Biologist Stuart Newman of the New York Medical College in Valhalla is trying to get a patent on a “humanzee”—a chimeric animal made from human and chimpanzee embryos. Not because he really wants to create one, but because he wants to prevent other people from making one, and to challenge the rules for patenting life. Together with Jeremy Rifkin, president of the Foundation on Economic Trends in Washington, D.C. Newman is embroiled in a strange legal contest with the government that entered a new phase last week as the duo announced that—to their delight—the Patent and Trademark Office (PTO) had turned down their patent application. Newman and Rifkin have never seen a humanzee, much less created one. But on 16 June, they put out a press release saying that they had applied for U.S. patents on many types of chimeras. The government actually rejected this application 3 months ago, says Rifkin, but he kept the information quiet until now because he wanted to avoid publicity while drafting an appeal. That appeal was completed last week, and Rifkin's attorney submitted it to the PTO. “If we win,” Rifkin claims, “we'll hold the patent in trust for 20 years” to prevent others from commercializing human-animal combinations. But he seems equally enthusiastic about losing: “We will appeal all the way,” he says, even to the Supreme Court if possible. Rifkin wants to provoke a debate about what it means to be “human” and to undermine the legal basis for patenting organisms—particularly those containing human genes. The first round of the contest shows that the PTO has begun the debate right where the duo wants it, says Rifkin: on the question of whether it is acceptable to patent human tissue. In its rejection letter, the PTO says that Newman's claimed invention—which relies on the use of human embryos—“includes within its scope a human being, and as such falls outside the scope” of what the PTO regards as legally patentable. Some patent attorneys suspect, however, that government lawyers may sidestep the big issues in future proceedings. For example, says Paul Clark of the Clark and Elbing law firm in Boston, lead attorney on Harvard's “oncomouse” patent, the courts could simply dismiss Newman's claims because he has never created the exotic chimeras he aims to patent. 7. CLIMATOLOGY # The Little Ice Age--Only the Latest Big Chill 1. Richard A. Kerr BOSTON—George Washington's winter at Valley Forge in 1777–78, when temperatures fell as low as −15°C, was relatively mild for those days; some years, New York Harbor froze solid. Indeed, so bitter were the centuries from about 1400 until 1900 that they have been dubbed “The Little Ice Age.” But new evidence appears to confirm that the long cold snap was nothing exceptional. Instead, it was only the most recent swing in a climate oscillation that has been alternately warming and cooling the North Atlantic region, if not the globe, for ages upon ages. At the spring meeting of the American Geophysical Union here earlier this month, paleoceanographers William Showers of North Carolina State University in Raleigh and Gerard Bond of Columbia University's Lamont-Doherty Earth Observatory in Palisades, New York, reported that they had found tiny bits of ice-borne rock in North Atlantic sediments laid down during the Little Ice Age. Identical rock fragments show up in older sediments every millennium or two, beginning at least 130,000 years ago. The finding implies that Earth has experienced a long string of Little Ice Ages, perhaps driven by variations in the sun or by changes in ocean currents. It also suggests that the world will be warming naturally, as part of this roughly 1500-year climate cycle, on top of any human- induced greenhouse effect. “It's intriguing,” says paleoclimatologist Jonathan Overpeck of the National Geophysical Data Center in Boulder, Colorado. “It's not a plausible explanation of the rapid warming we're having today, [but] it's very important that we study these century- and millennial-scale phenomena” to predict accurately any future warming. Historical records testify to the chill of the Little Ice Age—essential Dutch canals froze up all too regularly—and clues from the sea floor, lake beds, and glacial ice confirm that after 1400 temperatures in the Northern Hemisphere dropped 0.5° to 1°C below 20th century levels. Taking a longer view, Bond and his colleagues have spent the past few years assembling a 140,000-year record of climate cycles in the northern North Atlantic by counting microscopic bits of rock deposited during cold periods. The debris, picked up by ice on or around North America, Greenland, and Iceland, floated into the North Atlantic in icebergs and sank when they melted. During the ice age, Bond found, the debris jumped in abundance every 1500 years (give or take half a millennium) as the great ice sheets surged toward the sea. The oscillations continued after the ice age ended 10,000 years ago, although at greatly reduced levels. The trace of the most recent cold pulse in this series had eluded Bond until now, however. The conventional coring method of dropping a long pipe end first into the bottom mud flushes away the soupy upper centimeters of sediment accumulated in the past thousand years or so. To capture the topmost part of a core, Bond gently pressed a set of short pipes into the sediment, then sealed them before bringing them to the surface. Using both methods, Bond took cores near Newfoundland. After tallying the ice-rafted debris particles, he could see that the 1000- to 2000-year oscillation runs “through the Holocene and right into the Little Ice Age,” says Bond. “The Little Ice Age was not an isolated event.” Bond has “a reasonable argument” that the ice-rafted debris corresponds to the Little Ice Age, agrees paleoceanographer Jerry F. McManus of the Woods Hole Oceanographic Institution (WHOI) in Massachusetts. “There are others who think they have something like this marching through the record,” he adds. Lloyd Keigwin of WHOI found signs of a couple of cycles, including the Little Ice Age, off Bermuda. And earlier this year, Giancarlo Bianchi and Nicholas McCave of the University of Cambridge reported that the varying size of sediment grains down a core taken from south of Iceland reveals a rough 1500-year cycle in the speed of bottom currents during the past 7000 years, slower during the Little Ice Age and other cold phases. Linking the Little Ice Age to a long-running climate cycle lets Bond and Showers eliminate possible mechanisms behind the cycle. Even though it coincides with periodic surges of the Northern Hemisphere ice sheets, ice can't be the driving force, because the oscillation continues in a milder form during interglacial periods, when those ice sheets disappear. And the sun-shielding debris of volcanic eruptions, often proposed as the cause of the Little Ice Age cooling, seems to be ruled out, says Bond; there is no known 1500-year cycle in volcanic activity or any good reason for one. That leaves variations in solar activity—there were almost no sun spots during much of the Little Ice Age, implying that solar activity was at a low ebb—or some sort of oscillation in the ocean's operation, as hinted by Bianchi and McCave's deep-water flow record. Whatever the cause, says Bond, “the impact of this millennial-scale oscillation will have to be taken into account” as researchers plot the climatic future in hopes the world will fare better in the warming than Washington's army did in the chill at Valley Forge. 8. PRIMATE ABILITIES # Chimps in the Wild Show Stirrings of Culture 1. Gretchen Vogel TAÏ NATIONAL FOREST, CÔTE D'IVOIRE, WEST AFRICA—At the foot of a buttress tree, in the dappled sunlight of the rainforest floor, a young chimpanzee named Lefkas is working hard for his lunch. He holds a rock with both hands and a foot and slams it down with a sharp crack on a round coula nut, a bit smaller than a golf ball, which is balanced on a flat rock on the ground. After a few tries the nut cracks. The chimp pops the meat in his mouth and scampers off. The ground where Lefkas was sitting is strewn with coula nut shells, the leavings of other chimpanzees' meals. Indeed, from December through February, coula nut cracking is one of these chimps' main pastimes; primatologist Christophe Boesch, who has studied Lefkas's group at Taï for 20 years, says he watched another young chimp crack nuts nonstop for 5 hours. But chimps from just a few hundred kilometers away would probably stroll right past Lefkas's dining site. In a survey of chimps throughout Côte d'Ivoire, Boesch found no evidence for nutcracking anywhere east of a river called the Sassandra-N'Zo, even though both nuts and rocks are readily available throughout the forest. To Boesch, who is director of the Max Planck Institute for Evolutionary Anthropology in Leipzig, Germany, such differences in customs are akin to the use of chopsticks in Japan and forks in France: signs of distinct cultures, in which groups develop their own sets of behaviors based on social ties and shared history. Most people think of culture as encompassing such uniquely human skills as language, music, art, and clothing styles. But some biologists have a simpler definition: any behaviors common to a population that are learned from fellow group members rather than inherited through genes. By this generous definition, bird song dialects and the calls of whales might qualify as animal “culture” (Science, 27 November 1998, p. 1616). Most anthropologists stick to a narrower definition, requiring culture to include language and whole systems of behavior. But in the past decade, a growing number of primatologists and psychologists have sought to approach the question more rigorously, defining specific elements of culture that could potentially be observed in animals, then seeking these behaviors in the wild and in labs. They are turning up increasing evidence that nonhuman primates, in particular chimpanzees, may have a primitive type of culture that bridges the gap between the two definitions. Their argument rests on two main kinds of evidence: examples in which one chimp learns from another, and the results of such learning—the seemingly arbitrary differences in habits between chimpanzee groups at different sites. Although most examples of “culture” among animals involve just one or two behaviors, chimpanzees have dozens of learned behaviors involving tool use, social customs, and calls, says Andrew Whiten of the University of St. Andrews in Fife, Scotland. Of course, no primate society can build a mud hut or do any number of other tasks that are relatively easy for humans to master. Some researchers argue that that is because our primate cousins do not learn as we do, by imitation and instruction. And most agree that primates don't seem to be able to build on previous inventions, an ability that “is the hallmark of human culture” and that allows us to develop complex technologies and rituals, notes psychologist Bennett Galef of McMaster University in Hamilton, Ontario. Even so, Boesch and others argue that the nascent cultural stirrings of our primate cousins may help uncover the roots of human culture, showing that, for example, gregariousness—hunting and foraging together rather than alone —may have spurred cultural development. To see the beginnings of culture in other species, says Boesch, “helps us to see what is unique about humans.” ## Multicultural chimps Some of the best evidence for primate culture has come from field studies comparing the repertoire of chimpanzee skills and behaviors in groups around Africa. For example, in 1974 William McGrew of Miami University in Oxford, Ohio, detailed how chimps at Jane Goodall's Gombe site in Tanzania used sticks to fish driver ants out of their nests. A decade later at Taï, Boesch and his colleagues noticed a slightly different technique. At Gombe, chimps use 60-centimeter-long sticks to probe an ant nest. They wait for the insects to swarm halfway up the stick, then withdraw the tool and sweep ants off with their free hand, gathering a crunchy mouthful of hundreds of ants. At Taï, chimps use sticks about half as long, wait only a few seconds, then use their lips to sweep about a dozen ants directly into their mouth. The Taï method, analogous to eating soup with a tiny sugar spoon, collects only one-fourth as many ants per minute, but in 2 decades of observation, no animals at Taï have ever eaten ants Gombe-style, presumably because no chimp there ever discovered it. “A Gombe chimp would laugh at [the Taï chimps]” for their “primitive” method of ant fishing, says McGrew. Social interactions vary among groups, too. For example, McGrew, primatologist Linda Marchant, and their colleagues have recently documented a new behavior they call “social scratch,” in which one chimp rakes its hand up and down another's back after grooming. The behavior is common at Mahale in Tanzania but never seen elsewhere. Like some human fads and fashions, the behavior isn't utilitarian, but a part of social etiquette that apparently caught on simply because it feels good. “It's unlikely to be related to functional significance of grooming,” McGrew says, but rather helps to reinforce the social hierarchy. In preliminary studies, higher ranking chimpanzees received more social scratches per grooming session. Such examples add up to an impressive list. In last week's issue of Nature, researchers from the seven longest established chimpanzee field studies combined observations and listed 39 behaviors, from tool design to grooming to mating displays, that are distinct to particular groups and not readily explained by ecological differences. “We now have, in a sense, an ethnographic record” of chimp populations, McGrew says. “We have enough data in enough populations that we can start doing the sorts of comparisons that cultural anthropologists do across human populations.” Such geographical differences suggest that a chimpanzee's specific behavior and skills are shaped by where it is raised. That idea “is the most exciting finding” in chimpanzee field research this decade, says primatologist Tetsuro Matsuzawa of the Primate Research Institute at Kyoto University in Japan. Yet simply noting these geographical differences begs the question of how they develop and how they are maintained. ## Do apes ape? A chimpanzee pant-hoot sounds like nothing else in the forest: who-ho-who-ho-who AH AH AH AH. Another voice usually responds, and soon the din drowns out even the copulation cries of monkeys and the screech of the hyrax. “Chimps are the loudest animals in the forest, except for humans,” Boesch notes, when the din dies down. Researchers are now analyzing these hair-raising hoots for another proof of culture, one that helps explain the origin of geographical customs: that chimps learn from one another. In 1992, primatologist John Mitani of the University of Michigan, Ann Arbor, reported that different chimpanzee groups had distinct pant-hoot patterns and pitch, suggesting the possibility of learned chimpanzee “dialects.” But earlier this year he noted that those differences correlate with factors such as average body size and so might be genetic rather than “cultural” in origin. To find out, anthropologist Richard Wrangham of Harvard University and his colleagues studied calls in two captive groups where chimps from a mix of wild populations live together. In spite of the mixture of genetic backgrounds, each colony had a characteristic style of pant-hoot. “This is some of the best evidence for learning” of vocalizations, says Wrangham. “It's very difficult to think of an alternative hypothesis here.” Evidence that chimp behaviors can spread from one group to another would also strengthen the case that they are learned. Successful human practices tend to spread when people travel, and Matsuzawa has shown that in at least one case, a chimp skill spread the same way. He studies a community near the village of Bossou, Guinea, where the chimps are skilled tool users and frequently use rock hammers and anvils to crack the hard shells of oil palm nuts to get at the fatty meat inside; coula nuts do not grow here, although they are found on nearby Mount Nimba. In a 1996 experiment, Matsuzawa and his colleagues left rocks, oil palm nuts, and the unfamiliar coula nuts in a clearing, then hid behind a grass screen and videotaped the chimps. Several chimps picked up the unfamiliar nuts, but only an adult female named Yo cracked and ate them. Although other adults ignored Yo's nutcracking, a few young chimps watched her intently and later picked up and cracked nuts themselves. Matsuzawa suspects that Yo, who joined the group as an adolescent and may have been raised in the coula-rich Mount Nimba area, remembered the skill from her childhood. The fact that she passed it on to other young chimps shows, he says, that chimpanzee behaviors can spread from one group to another throughout a region, just as human cultural behaviors do. But the field is divided over whether monkeys and apes learn from one another the same way humans do, and researchers interpret the same experimental results in very different ways. For example, in the very first evidence of possible primate culture, reported in 1958, primatologists Shunzo Kawamura and Masao Kawai of Kyoto University observed as a young female macaque living on the small island of Koshima discovered how to wash sandy sweet potatoes (provided by the researchers) in a nearby stream. Eventually most of her group was doing it too. Kawamura suggested that this was a “precultural” behavior, and the observations were touted in textbooks for decades as evidence for culture among animals. But in the early 1990s McMaster's Galef and other animal behaviorists pointed out that the skill took several years to spread through the group and suggested that troop members, once they paid attention to the potatoes, discovered on their own how to wash them—essentially reinventing the wheel. In contrast, humans learning a new skill tend to carefully mimic the exact movements they see in an expert and are often deliberately taught by another person. Although reinvention might work for learning to crack nuts or fish for ants, says psychologist Celia Heyes of University College London, it wouldn't work for passing on more sophisticated cultural behaviors such as chipping arrowheads or weaving baskets. Such critiques sparked a flurry of new work in both the field and lab to discern whether great apes do in fact imitate. St. Andrews's Whiten and his colleagues developed “artificial fruits,” which required several steps to open, and found in 1996 that chimpanzees tended to complete the steps in the same order as the demonstrator. Primatologists such as McGrew say that the experiments have “nailed down” the point: “In the right sorts of circumstances, chimps imitate.” The observations of the different ant-dipping methods offer an example in the wild, adds Whiten. “It's difficult to see how such consistent behaviors could come about with anything but imitation,” he says. But the animal behaviorists aren't so sure. Following the order of simple actions is not the same as humans' imitation of fine motor movements such as dance steps, says Heyes. And Matsuzawa cautions that chimp imitation is rare in the wild. “Imitation is much more difficult than we expected,” he says. “Yes, there is imitation, but it is very, very difficult for the chimpanzee.” He and others also note that active, deliberate teaching, which some claim is a prerequisite for culture, is also rare among chimpanzees. Boesch has described two instances of mothers helping their offspring with the fine details of nutcracking, but as Galef points out, only two clear examples in 20 years of observation suggests that teaching is very rare. “The primatologists are pushing very hard for a rich interpretation of the data that are available,” he says. “Given that imitation is rare in nonhuman primates and teaching is essentially nonexistent, it's hard to see how you're going to get the cumulative culture which is the hallmark of our culture.” ## The benefits of tolerance Whether you call primate behaviors “culture” or not, researchers say that primate traditions may offer insight into the origins of human culture. Take orangutans, which love to eat the high-fat seeds of the neesia fruit. “It's like chocolate; they eat it for hours,” says Duke University biological anthropologist Carel van Schaik. Most orangs won't touch the fruit after it ripens, however, because the seeds are then surrounded by stinging hairs. But one population, in Sumatra, uses sticks to scrape out the hairs and get at the seeds. “The whole population knows the trick,” van Schaik says. “It's very similar to what we see in some chimp populations.” And it's the only case in which orangs—skilled tool users in captivity —have been spotted using tools in the wild. Orangs that avoid ripe neesia have the same sticks available for tools, so lack of materials can't explain why their behavior differs, van Schaik says. The key difference, he and his colleagues found, is that whereas most orangs are solitary, the Sumatran tool-using animals travel and feed close together, perhaps because there is plenty of food to go around. In most environments, food is thinly distributed and the animals “can't afford” to forage together, says van Schaik. The extra interaction in Sumatra allows an invention by one animal to spread when its compatriots observe it, he adds. The pattern also holds for chimpanzees, as van Schaik and his colleagues report in this month's issue of the Journal of Human Evolution. In a survey of the behaviors reported at the five longest running chimp field studies, the researchers found that those with higher “social tolerance” (measured by the amount of meat sharing, female-female grooming, and similar indicators) have more varied tool use. The theory could help to explain why captive primates are better at using tools than wild ones, as animals in captivity have more chances to observe one another and have plenty of food, van Schaik says. The correlation might help explain the rise of human tool use as well. The earliest tool-using hominids “didn't have a much bigger brain yet, so we shouldn't look for major cognitive advances,” van Schaik says. “I hypothesize that there was a social change that made them tolerate each other,” which led to increased opportunities to learn and build on each other's inventions. The fossil record might support such a theory, says anthropologist John Fleagle of the State University of New York, Stony Brook. Ancient humans have small canine teeth and lots of tools compared to other apes, he notes, and “when you look at the fossil record, you see reduction of canines early and tools later.” He thinks smaller teeth might be a sign of increased tolerance, as canines are often used in fighting among group members. “And once you have tolerance, you have bigger tool kits.” But the researchers attempting to learn the roots of culture by studying wild primates worry that they are running out of time. Habitat loss and increased hunting are pushing many great ape populations to the brink of extinction. Illegal loggers are threatening the Sumatran orangutans that van Schaik studies. And on a recent market day at the village of Taï, just outside the park where Boesch works, three chimpanzee heads were stashed in the game warden's office, confiscated from poachers. If Boesch and his colleagues are correct, says Whiten, such sights mean “we're not just losing chimpanzees; we're losing lots of different chimpanzee cultures.” That, he says, would be a major loss for humans. “If we want to understand how humans came to have the minds we have and the cultures we have, then we're only going to learn about that by looking for similar characteristics in our close relatives”—close relatives who are fast disappearing. 9. PRIMATE ABILITIES # Life--and Death--in the Forest 1. Gretchen Vogel TAÏ NATIONAL FOREST, CÔTE D'IVOIRE, WEST AFRICA—A day spent watching chimpanzees here begins before sunrise, with a headlong crash through tangled forest to the trees where researchers watched the 32-member group nest the night before. Later there may be another dash through the jungle, trying to keep up with a hunting party as they race through the treetops chasing their favorite prey, red colobus monkeys. There are quiet moments as well, of patient watching and waiting as the animals nap, notes veteran chimp watcher Christophe Boesch of the Max Planck Institute for Evolutionary Anthropology in Leipzig, Germany. But often the forest seems to be crawling with chimps—three or four juveniles swing in trees, adult females sit with young ones eating fruit, and an adolescent male doggedly follows his latest crush, a female currently in heat. Keeping track of who is doing what with whom sometimes seems like trying to keep track of a kindergarten class on a field trip. It's an exhausting way to gather data. But for researchers seeking to test theories about chimpanzee behavior, including the idea that the animals have a sort of rudimentary culture (see main text), watching the animals in the wild is the only way. So scientists rearrange their work and lives to accommodate the chimpanzee rhythms, spending several months a year in remote jungles, largely out of contact with the rest of the world. The first lesson for wannabe chimp observers is patience. Chimpanzees are wary creatures and flee the moment they spot an intruder. It can take as long as 5 years to accustom a group to the presence of note-taking humans, a process called habituation. When Boesch began to work here as part of his graduate studies back in 1979, he spent endless hours just chasing dark shadows in the forest. “In the first 2 months, we never saw a chimp. We only heard them running away,” recalls Boesch, who worked with his wife, Hedwidge Boesch-Achermann, to habituate the chimps. “In the first 2 years, we spent full days in the forest and saw chimps only 1% of the time. In the third year, we had a dramatic increase—to 5%,” he says. It was 5 years before a chimp first looked at a researcher without running away or otherwise changing its behavior. In order to get his thesis done before his adviser and funders lost patience, Boesch studied nut-cracking behavior, in part because the sharp, rhythmic ring of rock on nut can be heard throughout a forest even when no chimps are in sight. By the time his thesis was done, the chimpanzees were nearly habituated, and he could move on to more firsthand studies. After 20 years, the data-gathering is highly systematic. Boesch still takes notes by hand in a spiral notebook kept in a plastic bag in his breast pocket, but several of his students and assistants type three-letter codes for behaviors into hand-held computers, then download the data directly to a laptop back at camp. Often a researcher will follow one animal for the whole day, noting its behavior and interactions and photographing and videotaping it. Still, even after a group is habituated and has been studied for years, the entire research program is vulnerable to everything that threatens the chimps themselves, from poaching to disease. The group here has suffered two epidemics of Ebola, and last month, a suspected measles outbreak killed another eight chimpanzees, leaving the group's survival in question. Boesch and his colleagues have successfully habituated another group in the southern part of the forest and are working on a third, but studies of family relationships and social structures in the first group have been crippled. “It is the way of nature,” Boesch says sadly, “but that does not make it easier.” 10. PRIMATE ABILITIES # Are Our Primate Cousins 'Conscious'? 1. Elizabeth Pennisi With animals brandishing both tools and symbols, consciousness seems the last stronghold of human uniqueness. But might primates also have some elements of self-awareness? A new generation of researchers seeks to find out When Marc Hauser sat down to write his soon-to-be-published book, Wild Minds, he knew he was in for a wild ride. The Harvard University cognitive neuroscientist was about to ask questions that philosophers have struggled with for millennia—and he was asking them about animals, not people. How do they think? Are they self-aware? Might they even be conscious beings—and if so, how could we tell? Hauser admits that even approaching such questions can be maddening. It's almost impossible to know what another person is experiencing unless they tell you, so how can scientists ever know what nonverbal animals are thinking? And there's no consensus on exactly what consciousness is, much less how to test for it. All the same, Hauser and increasing numbers of neuroscientists, psychologists, and ethologists hope to yank such questions out of the realm of philosophy and into empirical science. They seek to create a scientific foundation for understanding just what it is that makes the human mind so different from those of our hairier cousins. Researchers are designing clever new ways to test primates for some of the concrete abilities long considered to be prerequisites for consciousness, such as overcoming instinctive behavior; being aware of oneself and of others (and knowing the difference); and, most sophisticated, understanding that others also have mental states and thoughts. By borrowing from studies of infants and comparing results among primates and children of various ages, these scientists are beginning to understand where on the continuum of intelligent beings chimps and monkeys fall. Less advanced primates are turning out to be capable of sophisticated activities such as tool use (see sidebar on p. 2075), while other primates appear to be closer to humans than has often been assumed. For example, some monkeys can overcome instinctive behavior to solve a problem more easily than can 2-year-old children. Other experiments seem to show that chimpanzees can attribute thoughts and intentions to each other. Species “have conscious behavior attuned to their ecological niches and show different levels of conscious behavior depending on the situation,” says ethologist Irene Pepperberg of the University of Arizona, Tucson. But other scientists find these experiments unconvincing, no matter how cleverly they are designed. Given that animals can't talk, says Celia Heyes, a psychologist at University College London, “I'm just mystified how anybody thinks you can find out about consciousness in other creatures.” Even some animal-cognition researchers caution that interpreting states of mind from an animal's behavior is always problematic. They heed a warning by the 19th century psychologist Lloyd Morgan, who argued that one should always look for a simple, mechanistic explanation for even the most complex animal behaviors, because complex behaviors don't always require complex thought. Children use correct grammar, for example, long before they understand what nouns and verbs are. “When we engage in certain behaviors, we're convinced that it's thought that prompted the behavior,” says Daniel Povinelli, a cognitive scientist at the University of Southwestern Louisiana in Lafayette. “But the exact same behaviors can be generated by other means.” Such skeptics argue that although animals may be smart, in the sense of having excellent information-processing capabilities, they lack the subjective experiences that are the essence of human consciousness. Thus, opinions on how wide a chasm separates us from other primates diverge wildly. Yet researchers on all sides agree that finding just what abilities lie in the gap will help us learn more about both primates and ourselves—and perhaps our ancient hominid ancestors to boot. What we learn about chimps, our closest living relatives, “will help us reconstruct the evolution of the human mind,” predicts Andrew Whiten, an evolutionary psychologist at the University of St. Andrews in Fife, Scotland. ## From tools to empathy Back in the 1950s, anthropologists drew the line between human and ape at the use of tools; thus any ancient hominid associated with stone tools was automatically assigned to our genus, Homo. But then in the 1960s primatologists found that chimpanzees can use tools, and now researchers know that many other primates can too. Next it was language that was held to be the truly unique human skill—but then in the 1970s, primates were found to have symbolic representations for objects, although they do not fully master syntax. Now the distinction chiefly rests on what is called consciousness, and in psychological circles the term has come to include an ever-expanding range of cognitive abilities, says evolutionary psychologist Richard Byrne of the University of St. Andrews. On the simplest level, consciousness is being aware of oneself and others; some researchers also say that it is correlated with creativity, language, and some form of empathy—putting oneself into another's shoes. Clearly, other primates lack this full package of abilities. But in the past 5 years, researchers have devised new tests that dissect consciousness more finely. For example, Hauser has decided to tackle what many think is an important step on the road to consciousness: the higher order cognitive function that enables individuals to override instinct and solve a problem in a new way. People do this all the time, for example, every time they see a cookie in a bakery window and walk away from the window and up to the counter to buy one, rather than succumbing to the impulse to reach through the glass and grab it. Traditionally, behaviorists had assumed that animals behaved instinctively and could not restrain a particular response to a problem even when it failed repeatedly. The tendency to do what's routine is called perseveration, and neurobiological studies have located the ability to overcome it in the prefrontal cortex, a part of the brain that is much enlarged in humans (Science, 15 August 1997, pp. 900, 953). Human adults are able to judge immediately when they need to do something differently. But new results blur the animal-human distinction by showing that human infants can't always vary their instinctive response, whereas other primates sometimes can. For example, in experiments Hauser described in January in Denver at the annual meeting of the Society for Integrative and Comparative Biology, he and Bruce Hood of the University of Bristol in the U.K. and their colleagues tested whether primates and children could overcome their instinctive anticipation of where a ball or a food pellet would drop. They used a frame with three chimneys at the top and three boxes lined up under the chimneys. Instead of falling straight to the ground, however, the ball was sometimes shunted over to a different box through either an opaque or transparent shunt; the subject's task was to predict where a ball dropped down a chimney would land. Children over age 3 could always predict the ball's course based on how the shunt was positioned. But children under 3 and monkeys had similar difficulties with this task. Young children predicted the right landing spot only if the ball dropped straight down or they could see the ball moving through a transparent shunt. Cottontop tamarins—primitive primates weighing only 500 grams—reacted much the same way: They were able to find a food pellet if it fell straight down but not if it was shunted to another landing spot in an opaque shunt. But when the researchers put the apparatus on its side so that the objects were moving horizontally—thereby avoiding any gravity-related instincts—the monkeys did much better than the younger children in anticipating where the ball or pellet would emerge, says Hauser. No one is suggesting that cottontop tamarins are conscious, but the work shows a continuum of abilities in primates and humans, rather than a single cutoff. And because 2-year-olds make many of the same mistakes as the monkeys and on some tasks do worse, even though they have language, the experiments also suggest that overcoming perseveration has little to do with language. For example, language can't help children predict where the ball will fall. “We think of ourselves as thinking in language, and thus it's easy to conclude that language is doing all the work,” Hauser explains, but it's not. William Kimler, a historian of science at North Carolina State University in Raleigh, agrees: “It's not about language; it's about planning.” ## Beyond mirrors Almost everyone agrees that self-awareness, or being cognizant of one's body and thoughts, is another crucial element of consciousness, and many researchers think that chimps possess it. Alone among primates, chimps can recognize themselves in a mirror. But Robert Seyfarth and Dorothy Cheney at the University of Pennsylvania, Philadelphia, who have studied vervet monkeys and baboons in the field in Africa for 20 years, wondered if these primates might have self-awareness, too. Seyfarth and Cheney contend that the mirror test isn't relevant for species that in the wild would never have the opportunity to look in one. Because primates live together in tight hierarchies, they argued that a better test would involve “social self-awareness” —whether individuals understand themselves and their relations to other group members. “If you divide self-awareness into its components, then here's an aspect where we may be able to make progress,” Seyfarth says. For example, in one recent study, he and Cheney studied how pairs of female baboons reacted to recorded sounds of other adults in a fight. If the adults were not their own kin, the females didn't react. If the cries came from one female's relatives, the other female would look at her, and if both the females were related to the rabble-rousers, then the two females looked at each other and, eventually, the dominant one came and sat down in the place of the subordinate one, reasserting her place in the hierarchy. “This suggests they know the individuals [and their calls] and also the family relationships of each individual,” Seyfarth reported at this year's integrative biology meeting. The work “nails down the fact that these animals show an awareness of their own position in society and their position with respect to others,” says Carel van Schaik, a biological anthropologist at Duke University in Durham, North Carolina. That sophisticated knowledge of their social selves, says Seyfarth, “raises the possibility that they have a sophisticated sense of self.” ## The mind problem If self-awareness is part of consciousness, perhaps the next significant step is the ability to attribute mental states to others. In 1978, a hand-raised chimp named Sarah seemed to be able to understand what a human tester in a video should do to solve problems such as reaching food on a high shelf. To some, this suggested that chimps had a “theory of mind”—that they understood that other individuals had thoughts and mental states, too. Critics argued that the experiment was very contrived, however, and it took researchers years to come up with better tests. Now several groups are doing such experiments on primates. And the primate work fired the imaginations of child development researchers, so that there are now hundreds of papers on children's development of theory of mind. That work shows that children become sensitive to what others are thinking at an early age, but are unable to attribute false beliefs to others until around age 5. For chimps, however, the results are conflicting. For example, as part of a major program tracking cognitive development of both chimps and children for the past 8 years, Southwestern Louisiana's Povinelli had his subjects gesture to one of two people —one with a blindfold and another with a gag over the mouth—in order to ask for a treat. Children of age 2 understood that the blindfolded person could not see their gesture and asked the gagged person, but chimps were just as likely to gesture to either person. “They are not reasoning about seeing,” says Povinelli. He concludes: “Humans have a whole system that we call theory of mind that chimps don't have.” Work by Josef Call, a psychologist at the University of Liverpool in the United Kingdom, agrees in part with this conclusion. In the March-April issue of Child Development, Call's team reports that they could find no evidence that five chimps and two orangutans could figure out where a tester should find a hidden piece of food whose position has been switched without the tester's knowledge, although the animals themselves observed the switch. They were not sophisticated enough to realize that the tester had the wrong knowledge of the food's location, presumably because they couldn't fathom that the tester had knowledge different from their own. Still, this was a test of one of the most sophisticated aspects of the theory of mind, says Call. “The theory of mind is not just one skill; it's a series of skills,” he says, and he thinks that primates might still understand something of others' thoughts. Indeed, some positive results are now appearing. For example, Harvard graduate student Brian Hare, who works with Call in the lab of developmental psychologist Michael Tomasello, now at the Max Planck Institute for Evolutionary Anthropology in Leipzig, Germany, sought to design a scenario more relevant to chimps' lives than previous tests were. Because chimps forage in groups and have an elaborate set of rules about who gets to eat food first, Hare devised a test that looked at whether one chimp could tell what another chimp, rather than a human tester, was seeing—and presumably, thinking. Their setup involved three opaque cages in a row, with a chimp in the first and third cages and two pieces of food in the middle cage. The doors from the outer cages to the middle one were first opened just enough that each chimp could peek at the food and see that the other chimp was eyeing it too. When the door was opened fully, only the dominant chimp of the pair retrieved the food, as would have been the case in the wild. The researchers then placed a barrier in such a way that the dominant chimp could see only one piece of food, but the subordinate could see both and could also see that only one piece of food was in view of the dominant chimp. This time, the subordinate took the piece of food that the dominant couldn't see, suggesting that it knew the dominant was unaware of this food's existence. And when the dominant chimp was replaced with a chimp even lower on the hierarchy, the newly dominant chimp first went after the food both chimps could see—grabbing the potentially more contested item first—and then retrieved the second piece. Thus the chimp's response varied depending on its fellow's identity and what it could see, suggesting an understanding of another's visual perspective. Similar hints that chimps know what is going on in each other's heads come from Tetsuro Matsuzawa and his team at Kyoto University in Japan. They also looked at food retrieval, this time by pairs of chimps—a “witness” chimp who had seen where food was hidden and a “bystander” who hadn't. In a variety of experimental protocols, the “bystander” tended to follow the “witness” around and so appeared to understand the witness's knowledge, says Matsuzawa. In addition, the witness sometimes misled the bystander by leading it to an empty box, the team reported in June 1998 at the Napoli Social Learning Conference in Italy. In both sets of experiments, chimps are behaving as if they have a rudimentary awareness of their fellows' desire to find food—the first stirrings of mind-reading, says Matsuzawa. A few other experiments show similar abilities. One provocative study by psychologists Charles Menzel, Sue Savage Rumbaugh, and Duane Rumbaugh of Georgia State University in Atlanta and their colleagues involved a chimp named Panzee, who learned to communicate with a special computer keyboard outfitted with symbols. When seeking objects hidden outside her habitat, Panzee apparently understood that certain human keepers did not know what was hidden and where, and she told them both what the object was and what they needed to know to help her find it, according to a paper in press in the Journal of Comparative Psychology. These data are too new to be definitive and are bound to elicit tough scrutiny by both supporters and skeptics. Researchers like Tomasello, for example, don't think that chimps have a full theory of mind. “Seeing and perceiving is not the same thing as knowing and believing,” he says. Tomasello adds that the conflicting evidence about chimp consciousness may reflect the difference between animals raised in the wild and in captivity. He suggests that chimps raised by humans, such as Panzee, may be more likely to develop a sense of self and possibly an awareness of others. Human babies, he speculates, learn to recognize how others react to them and become self-aware because of the attention they get from adults. Thus, a human raised in isolation might not have the same “consciousness” as the rest of us, whereas chimps raised by people do uncharacteristically well in theory-of-mind experiments. Researchers admit that they expect to puzzle over the theory of mind for a while. But as they design cleverer and cleverer experiments, they are optimistic about beginning to chart the still-unknown territory that divides the human and animal minds. “These are exciting times,” says Harvard's Hauser. “I predict we will make immense progress.” 11. PRIMATE ABILITIES # In Labs, Organ Grinders Take Up Tools 1. Elizabeth Pennisi Consciousness isn't the only ability once thought to be uniquely human that may be slipping down the evolutionary scale (see main text). Decades ago, it was tool use that was said to mark the divide between humans and other primates—but then chimpanzees and other apes were spotted using tools. Now even the capuchin monkey, the organ grinder's accompanist, turns out to be adept at turning sticks into digging implements and leaves into containers. These 4-kilogram, nimble-fingered primates split off from the more intelligent apes a full 30 million years ago, and human observers hadn't seen many feats of intellectual prowess when watching them in their native rainforests of South America. But laboratory experiments tell a different story, says Greg Westergaard, who now directs the research division of a private primate facility, LABS of Virginia Inc. in Yemassee, South Carolina, and is an adjunct researcher with the National Institutes of Health. To prove an animal is smart, says Westergaard, the key is to watch it learn in new situations—and not to teach it. In the past, good training strongly reinforced with food rewards was often confused with intelligent behavior. So as a staff fellow at the Laboratory of Comparative Ethology at the National Institute of Child Health and Human Development, Westergaard and NICHD psychologist Stephen Suomi rarely included training with food rewards as part of the plan. Rather, they exposed the monkeys to everything from cups to clay and designed experiments based on spontaneous behaviors observed as the groups play. In one experiment, for example, they buried peanuts, provided branches of various sizes, and observed how the monkeys retrieved these subterranean treats. Many tried to dig the peanuts out with their hands; some tried poking at the ground with various branches; and a few even broke the branches and removed the bark and leaves, making more effective digging tools. That shows sophisticated tool manufacture and creative problem-solving, says Westergaard. Westergaard attributes his discoveries about the capuchins to working in the lab, where he can keep the monkeys happy and alert and observe them intensively, noting unexpectedly sophisticated behavior. Capuchins revel in shaping and painting clay, for example, and do it for hours with no food reward, he says. And in what appears to be rudimentary symbol use, they can learn to associate a particular color chip with a particular tool and can “ask” for a certain tool by giving an experimenter the right color. Now the lab-based discoveries about tool use have been verified in the wild. Last year, Kimberly Phillips of Hiram College in Hiram, Ohio, reported in the American Journal of Primatology that capuchins in Trinidad use leaves as sponges and rudimentary water containers; there's even one report of capuchin monkeys clubbing a snake with a stick. Among primates, it seems, tool use is a popular trick. 12. INTERNATIONAL COOPERATION # Finding a New Home for BESSY in the Middle East 1. Oliver Morton* 1. Oliver Morton is a writer in London. Germany wants to make a donation of a used synchrotron. But how will this scientific foster child fit into the Middle East's dysfunctional family? PARIS—Germans are well known for their environmentally responsible attitude toward reusing and recycling, and now they are extending that attitude to large research facilities. Faced with the need to decommission BESSY 1, a successful synchrotron x-ray source in Berlin, German physicists and their colleagues around the world decided it would be a shame to just sell it off as scrap. Instead, why not give it to some part of the world that would like such a machine but couldn't afford to build one? Somewhere like the Middle East. This somewhat quixotic idea is now on a fast track to reality. At a meeting held here last week under the auspices of UNESCO, the United Nations Educational, Scientific and Cultural Organization, representatives of five governments from the Middle East region—including Palestine—expressed official interest in hosting an upgraded BESSY 1a. These countries and others, together with well-wishing members of the synchrotron community around the world, will now look at ways to set up and fund an international center to house the machine—a center open to Arab and Israeli, Turk and Cypriot alike. As Federico Mayor, the director-general of UNESCO, put it when opening the meeting: “Such a center would encourage regional and international cooperation in science [and offer] an impressive practical illustration of ‘science for peace.’” When you whirl electrons around in circles vigorously enough, they give off energy in the form of peculiarly pure x-rays. This is something of an irritation for particle physicists, but a boon for their solid-state colleagues, as well as for structural biologists, surface scientists, environmental chemists, and a growing number of other specialists in all sorts of fields. When the usefulness of these very bright x-rays became clear in the 1970s, governments started to purpose-build storage rings to produce x-rays—dedicated synchrotron light sources. The more light sources that have been built, the more uses researchers have found for their light, and so yet more light sources have been commissioned. According to Herman Winick of the Stanford Synchrotron Radiation Laboratory, the person who first suggested giving BESSY 1 to the Middle East, about 45 synchrotron sources are in use around the world, with 11 more under construction and 16 more being designed. Unlike most big-science installations, these are inherently multidisciplinary, which is one of the things that makes them so attractive. The field's fast and continuing growth means that new machines are coming online before older ones have outlived their usefulness. Hence, a few older machines have found a new life in retirement. A Japanese synchrotron built for a fixed-term industrial research program has recently been shipped to Thailand. And a Dutch accelerator and storage ring used for nuclear physics is being moved to Dubna, outside Moscow, to add to Russia's synchrotron capability. BESSY 1, which is being replaced by a bigger machine, BESSY 2, still has plenty of life left in it. Synchrotron researchers say that although these machines may be secondhand, they need not be second rate. According to Gustaf-Adolf Voss of DESY, Germany's national accelerator center in Hamburg, the upgraded BESSY 1a that would be sent to the Middle East would be a “world-class machine.” It would have a new control system and vacuum system and room for more “insertion devices”—arrays of magnets that kink the beam in order to produce x-rays of particular brilliance. Although in early discussions of the project some scientists from the Middle East were a little leery about a cast-off machine, the proposed upgrades appear to have convinced everyone that the region could do very nicely with BESSY 1a. If, that is, the interested parties can find a place to put it, money to pay for it, and scientists to use it. For any big science project, site selection is all-important and usually deeply contentious. Big machines are prizes that bring opportunities, prestige, and money—a lot of the spending in such projects is local. Such factors no doubt motivated the five territories whose representatives in Paris expressed an interest in providing BESSY's new home—Iran, Egypt, Palestine, Turkey, and Cyprus. The Palestinian delegation was particularly strident—“This is the least the world community can do [for us],” said Hanna Hallak of Bethlehem University in the West Bank. After Hamid Mohamed Roushdy El-Kady of Egypt's National Centre for Radiation Research had gone through his country's many successes in the running of physical science institutes, Said Assaf of the Arafat National Science Centre for Applied Research came back with a grin: “You have many institutes—you need one more?” By the middle of July, Voss and his colleagues on a committee looking at the technical aspects of the BESSY 1a program will have drafted requirements for the site, such as the nature of the bedrock, local sources of vibration, and stable electrical power. (The synchrotron will need perhaps 3 megawatts.) Interested nations will have until the end of November to put together a bid, and the decision might then be made by the end of the year. If this seems terribly fast, that's because it is. The German government wants the synchrotron gone within a year of its closing down late in 1999, not least because the Max Planck Society wants to move a new center for the history of science into the vacated building. Whatever final requirements the site has to meet, one of them is clearly not negotiable: It must be accessible to all nations of the region, including Israel. Israel made it clear at the Paris meeting that it is not offering to host the facility—there would undoubtedly be resentment if it did—but it is vital to the project's success for two reasons. One is that without Israel's involvement it would hardly look like science for peace. The other is that in one small country the Israelis have more expertise in synchrotrons and their use than the rest of the region put together. There are more than 20 Israeli teams working on synchrotrons around the world, and the country is an associate member of the European Synchrotron Radiation Facility, whose machine in Grenoble is one of the world's biggest and best. Israel thus has many academics interested in using such a facility, and it also has industries that might conceivably wish to participate. Being able to attract such fee-paying users will undoubtedly decide the project's success. Many in Paris felt that the German estimates of$10 million for the upgrade, $10 million for infrastructure at the new site, and about$4 million for running costs was at best optimistic. Various possible donors were discussed, including the European Union's Mediterranean development budget and U.S. aid toward the Middle East peace process, which runs to billions of dollars. Closer to home, there are the oil-rich Arab states. None of these were represented in Paris, but if they could be persuaded to join the project, they could be a valuable source of cash.

But as James Vary of the International Institute for Theoretical and Applied Physics at Iowa State University in Ames points out, the big sum up front is not the most serious worry. “After the money for science for peace, you still need money for science.” And the Middle East is not renowned for its generous research budgets. Khaled Elshuraydeh of Jordan's Higher Council for Science and Technology estimates that Arab governments spend on average 0.2% of their mostly rather modest national incomes on R&D. A state-of-the-art synchrotron, together with the beamlines needed to channel its x-rays and the experimental setups required to use them, would be a very big fish in a small pool—possibly, given what else might be done with the money, an inappropriately big fish.

The researchers' solution to this would be to increase the size of the pool. As Miguel Virasoro, director of the Abdus Salam International Centre for Theoretical Physics in Trieste, Italy, points out, there is no conservation law keeping Middle Eastern research budgets at their current low level. The need to make accommodations for the synchrotron could be a way of focusing attention on research. “This means that governments [have to] raise it to a higher level on their agenda,” agrees Vary. But for that to work, there have to be researchers and they need to be nurtured and trained now, even if the synchrotron will not start work for several years. The Abdus Salam center already runs courses in synchrotron radiation applications, which have been put to good use by the Thais in setting up their center based around the Japanese synchrotron and also by the Brazilians, who built their own light source from scratch. According to Stanford's Winick, the U.S. Department of Energy might be willing to provide training at its facilities, although it would not cover all the costs.

By the end of the Paris meeting, the disparate group of participants had organized themselves into an interim council for the project with various committees looking at different aspects, such as training and funding. Arabs and Israelis nominated each other to the committees with a clear concern both to get the right people and the right balance. In this genuinely good-natured and open tone, the Paris meeting proved that the builders and users of synchrotrons are a community in more than name. “It's amazing how open people are here,” said a watching particle physicist. “If only we could transmit this spirit to the people back home,” said Voss. Even if it can't be transmitted directly, they'll do their best to put it into a storage ring.

13. PHYSICS

# Will the Higgs Particle Make an Early Entrance?

1. James Glanz

A raison d'être for the potent accelerator now being built at CERN in Europe, this long-sought particle may be within reach of an existing U.S. machine

BATAVIA, ILLINOIS—When physicists at the Fermi National Accelerator Laboratory discovered the top quark 4 years ago, the choice of music to play over the lab's public address system was obvious: Cole Porter's “You're the Top.” Now they—and their colleagues from another major particle physics laboratory, CERN, in Geneva, Switzerland—are speculating that a new theme could soon be in order: “You Turned the Tables on Me.”

The music would celebrate the Higgs boson, an eagerly sought particle believed to account for the origin of all mass, including that of the top quark. Most estimates had placed the mass of the Higgs itself too high for Fermilab's Tevatron accelerator to create it—leaving the search for the Higgs to CERN and its Large Hadron Collider (LHC), a far more powerful accelerator that won't be completed until at least 2005. But now it looks as if Fermilab might just be able to steal a march on CERN.

As the Tevatron and other accelerators have measured particle masses more precisely and theorists have refined their calculations, the best estimates for the Higgs mass have narrowed. Those calculations, the subject of intense discussion last week at a conference* here at Fermilab, near Chicago, “tell us the Higgs has to be rather light,” says Marcela Carena, a CERN theorist on leave at Fermilab. “This is where the Tevatron becomes interesting,” she adds. The calculations imply that the Tevatron, newly upgraded at a cost of $260 million, just might spot the Higgs before the LHC does. But the Fermilab-first scenario could depend on whether the upgraded Tevatron is allowed to run for longer than the 2 years now planned—and on whether the Higgs's low mass has opened the way to a surprise winner, CERN's current accelerator (see sidebar). The Higgs is crucial for modern particle theory, which is based on an interplay of symmetries but must also explain why particles in nature have wildly different masses. A “Higgs field,” which is envisioned as permeating all of space like a sort of unchanging voltage, spoils the perfect symmetry and results in the array of different masses. Quantum-mechanical excitations of the field should yield the Higgs particle itself, just as photons, or particles of light, emerge from a smooth electromagnetic field. The Higgs particle should be shaken loose by sufficiently energetic collisions in accelerators. “One of the raisons d'être of the LHC was to discover the Higgs,” says Georges Azuelos of the University of Montreal and the LHC's ATLAS collaboration. As Fermilab theorist Joe Lykken explains, “The thinking was that [the Higgs mass] could be as high as 1000 GeV”—or 1000 billion electron volts, about 1000 times the mass of a proton. Although the upgraded Tevatron hurls particles together with 2000 GeV of energy, not all that energy goes into creating new particles. As a result, making heavy Higgses in detectable amounts seemed likely to require the 14,000-GeV,$5 billion LHC.

Gradually, the mass estimates changed. Because quantum mechanics lets any one kind of particle temporarily exist as another particle, their measured masses are all closely related. So a refined top-quark mass, say, leads to a better estimate of the Higgs mass. Those refinements alone have pushed the estimated Higgs mass to below 230 GeV. But it also turns out that the Standard Model of particle physics—the accepted theory of particles and forces—goes haywire at very high energies unless the Higgs mass is less than 180 GeV. And a more comprehensive theory called supersymmetry, still hypothetical but popular among theorists, would be “strongly disfavored” unless the mass is below roughly 130 GeV, says Carena. Many of the new estimates emerged from a multi-institutional Higgs Working Group convened at Fermilab last year, she says.

The clearest signature of a low-mass Higgs would be its decay into a bottom quark and its antimatter counterpart, an anti-bottom quark, which would be manifested as “jets” of particles. But because of LHC's great energy and particle-beam intensity, it would generate massive numbers of jets by other processes and would have to rely on much-less-common decay patterns to detect a low-mass Higgs. Calculations Azuelos presented at the meeting suggest that it might take the LHC several years of running to discover a light Higgs.

As a result, says Azuelos, “there are chances that it will be seen at the Tevatron.” Speaking of physicists working on the LHC, Azuelos says “of course they are concerned” about that possibility, but “they are not worried about it”—as the Higgs will have to be studied in detail no matter where it is found, and the LHC is well suited to doing so by making the particles at a brisk pace.

The chance that the Tevatron will stake the first claim on the Higgs will rise, says Gordon Kane, a particle theorist at the University of Michigan, Ann Arbor, if Fermilab's new director, Michael Witherell, decides to push for more Tevatron run time. Witherell says, noncommittally, that he likes the idea. “All of the indirect evidence we have tends to push the Higgs mass down to the low end of the range,” he notes—“which is our end.” At the same time, he cautions that the mass estimates are not airtight and that any time-consuming upgrades or repairs to the Tevatron could erode its lead over the LHC. But if the Tevatron hits the right note in time, it could have particle physicists singing “Goin' to Chicago.”

• *7th International Conference on Supersymmetries in Physics (SUSY99).

14. PHYSICS

# A Tentative Nondiscovery of the Higgs

1. James Glanz

BATAVIA, ILLINOIS—Don't think of elephants. Now, are you thinking of elephants?

This classic psychological ploy captures the dilemma facing one scientific collaboration at the Large Electron-Positron collider (LEP), a particle accelerator at CERN in Geneva, Switzerland. The subject of the “don't think of” scenario is the Higgs boson, the hypothetical particle that is thought to explain how everything else in the universe—including all the particles in elephants—acquired its mass. The collaboration, called OPAL, is trying to dampen speculation that a handful of unexplained events in its data point to a Higgs discovery. The result is to fuel the rumors.

“It's not an effect—but may be interesting,” said Eilam Gross of OPAL and the Weizmann Institute of Science in Rehovot, Israel, during the SUSY99 conference here last week. Gross's disclaimer, given as he presented a viewgraph on the data, drew knowing titters from the audience. He explained that the gap between the data points and the green peak of expected “background” counts could either represent a statistical fluctuation or the first hints of a Higgs with a mass of about 91 billion electron volts (GeV), or 97 times the mass of a proton. “It's exactly the right sort of Higgs mass,” said Gordon Kane, a theorist at the University of Michigan, Ann Arbor, who adds that the signal should soon be either “golden or gone” as more LEP data stream in.

LEP has been smashing together electrons and their antimatter counterparts, positrons, at gradually increasing energies, and Gross said that the intriguing data came from runs since 1997 at collision energies from 183 to 189 GeV. If a Higgs materialized briefly in the debris of those collisions, it would be expected to decay most often into a bottom quark and an anti-bottom quark, each of which would be manifested as a “jet” of protons, neutrons, and pions in the OPAL detector. Because other particles created in the collisions—especially the Z boson, which mediates the weak nuclear force—would produce similar events, their expected contribution is a “background” that must be carefully subtracted from the overall count. Still, said Gross, the subtraction leaves an excess of remaining counts that has only a 4% chance of being a statistical fluke.

That level of certainty is not considered strong enough for a claim of discovery in particle physics—particularly because, according to Gross, LEP's other three detectors see no clear evidence for a peak there. The fact that the suggested value for the mass is almost precisely the same as that of the Z boson has raised eyebrows. “It's probably just a misunderstanding of their background,” says Georges Azuelos of the University of Montreal and the ATLAS collaboration of the Large Hadron Collider, a much larger accelerator to be built in the LEP tunnel.

Nevertheless, Gross and others pointed out that within the next year or so, when LEP reaches its ultimate energy of 200 GeV, it should be capable of a firm detection of the Higgs if its mass is anywhere below 109 GeV. (Much of the collision energy goes into creating other particles.) After that, Fermilab's Tevatron has a shot at the discovery if the mass is below about 180 GeV (see main text). “If God is on our side,” said Gross, playing on the almost theological significance the discovery would have for particle physics, “let Higgs appear this year. Amen.”

15. AMERICAN GENETIC ASSOCIATION MEETING

# Genomes Reveal Kin Connections for Whales and Pumas

1. Elizabeth Pennisi

STATE COLLEGE, PENNSYLVANIA—At this year's meeting of the American Genetic Association, held here on 12 and 13 June, researchers discussed genomic data that shed light on both ancient evolution and the relationships among modern species.

## A Puma Is a Cougar Is a Panther

Pumas are known by many names—panther and cougar among them. Indeed, experts on the animals thought they were so genetically diverse as to constitute a menagerie of 32 subspecies. Now an extensive genetic analysis has turned up just six puma subspecies. The finding sheds light on the evolution of the 60-kilogram cats and suggests that keeping some of these supposed “subspecies,” such as the Florida panther, from becoming extinct may be easier than previously thought.

As part of a DNA study of the world's cats, Stephen O'Brien and his team at the National Cancer Institute (NCI) in Frederick, Maryland, collected blood and tissue samples from 209 pumas in zoos, museums, and the wild across North and Central America, and from 106 of the animals in South America. They then looked for sequence differences in three mitochondrial genes and 10 microsatellites, short bits of repetitive DNA sequence that lengthen and mutate through time and thus indicate the relatedness of organisms.

The researchers found no differences in the mitochondrial DNA from North American pumas, and their microsatellites were “virtually indistinguishable,” NCI's Melanie Culver reported at the meeting. This suggests that only one kind of puma inhabits North America, rather than the 15 subspecies previously identified on the basis of where they live and differences in appearance. The DNA analyses also showed that only one subspecies lives in Central America and that just four others prowl South America. The NCI team found the most genetically diverse pumas in Paraguay and Brazil south of the Amazon River. This indicates that these populations are the oldest, dating back some 250,000 years, and that northward migrations gave rise to the others over time, Culver adds.

The work is “a tour de force,” says Oliver Ryder, a geneticist at the Zoological Society of San Diego. Moreover, with North American pumas so closely related, zoos should be able to breed endangered Florida panthers with others from the continent without fear of contaminating the genome of that “subspecies,” notes geneticist James Womack of Texas A&M University in College Station. On the other hand, if the researchers hope to introduce more diversity into North American pumas, they will have to travel far afield for appropriate mates.

## Whales and Hippos: Kissing Cousins?

Smooth, blubbery, and aquatic, whales and hippos look like plausible relatives. Now their DNA agrees. Based primarily on fossil comparisons, paleontologists had thought whales arose tens of millions of years ago from a hyenalike ancestor called a mesonychian. Over the past several years, however, comparisons of the genetic material of whales and other living mammals suggested that they belong instead among the even-toed ungulates, which include cows, deer, hippos, pigs, and camels. At the meeting two groups presented new molecular evidence that pointed to hippos, not cows or deer, as the closest cousins of sea-going mammals such as whales, porpoises, and dolphins.

Most molecular biologists try to sort out kinship between species by determining the degrees of difference in the same gene across species, then calculating the most plausible tree to fit them. But because whales, ruminants, and other close relatives, such as camels, have split apart so recently, their individual genes have few differences, making statistically significant results hard to get. So geneticist Norihiro Okada of the Tokyo Institute of Technology in Yokohama, Japan, and his colleagues used a different strategy.

They ferreted out “short interspersed repetitive elements,” or SINEs—bits of chromosomal DNA that at some point in history were transcribed into RNA and then, after being copied back into DNA, accidentally reincorporated in a new location in an organism's genome. Because these events are rare and SINEs, which can be recognized by their distinct sequences, stay put once they get back into the genome, two species sharing a particular SINE at the same site must have a common ancestor.

In 1997, Okada showed that whales and ungulates, such as hippos, cows, and giraffes, share three SINEs, indicating that the species are related. Now, with almost twice as many SINE insertions in hand, the group has identified SINEs common to whales and hippos but not present in the cud-chewing ruminant branch of the ungulates, which includes cows and giraffes. Thus, they conclude that whales and hippos evolved from a hippolike ancestor that had split from the ruminants some 55 million years ago. (The results are in press in the Proceedings of the National Academy of Sciences.)

A more traditional comparison, but one that uses 8200 nucleotide bases from eight genes, not just a single gene, in two whale and 24 ungulate species, supports that conclusion. “Until now, no one had sequenced multiple genes to see what they can tell us,” says geneticist Conrad Matthee of Texas A&M University in College Station, who presented the work.

To geneticist Masatoshi Nei of Pennsylvania State University in State College, these data mean that the question of whale evolution has “finally been decided.” Not everyone is so sure, however. Anatomist and paleontologist Hans Thewissen at Northeastern Ohio Universities College of Medicine in Rootstown says that although the SINE data in particular make a “compelling argument” that whales and hippos are cousins, he is still “on the fence.” Even though Thewissen himself reported a fossil last year that weakened the ties of whales to the hyenalike mesonychian, he says more fossil evidence is needed to convince paleontologists that the molecules are telling the truth.

16. EVOLUTION

# Ecology Returns to Speciation Studies

1. Virginia Morell

Evolutionary biologists rediscover their roots, as field studies highlight the importance of ecology in the formation of species

Charles Darwin got his ideas about how species arose by poring over his voluminous notes taken aboard the Beagle and from hours of observation in his native England. He watched to see what pollinators visited particular orchids, grew 233 cabbages of different varieties near each other to see how many offspring were true to their kind, and counted the number of Scotch fir seedlings that filled the gaps between adult trees.

For Darwin, such loving attention to the details of nature was the underpinning of his theory, but such ecological minutiae have had little place in this century's more sophisticated evolutionary science. Among evolutionists trying to understand how species arise, “ecology has been out of favor” for over 2 decades, says James Patton, an evolutionary biologist at the University of California (UC), Berkeley. Ecology was neglected, he and others say, as researchers buoyed by the power of new genetic techniques were “swept up in creating molecular family trees,” focusing on the relationships among species, not what drove them apart in the first place.

Now, however, Darwin's obsessive attention to ecological detail is making a comeback. Researchers probing the mechanics of how one species splits into two are once again taking copious notes about such things as the number of predators lizards face in different forests and the angle at which bottom-dwelling fish feed. It's a back-to-basics approach that has led to some sophisticated and surprising science—and revived an old idea: the ecological speciation model.

In the dominant picture of speciation, put forth in 1942 by Harvard University's Ernst Mayr, a geographic barrier develops between two populations, interrupting gene flow between them. Even if the populations live in identical environments, gradually they diverge through random mutations, so that if they ever encounter each other later, they will be unable to mate—a condition called reproductive isolation, the sine qua non of speciation. But the ecological speciation model offers another possibility: The barriers that spawn species can be ecological rather than geographic, and selection may be paramount. Different ecological pressures will favor changes in body shape and function that eventually make populations unable or unwilling to mate with each other, even if they have never been physically separated.

Thus, researchers studying speciation find themselves paying attention to environmental factors as well as genes. Some field studies demonstrate how a particular selective factor can push two populations down separate evolutionary paths. Others try to probe the factors that keep incipient species from mating, testing whether genetic or ecological differences make the best chaperones. And a few studies are trying to put it all together to document ecological speciation. The new work “shows the importance of ecology in speciation, which has been almost entirely neglected,” says John Endler, a longtime proponent of ecological speciation at UC Santa Barbara.

This view of speciation “is not new,” adds David Wake, an evolutionary biologist at UC Berkeley. “It traces its roots to Darwin. But what's new and nice is the sharp focus on testing hypotheses [via] natural systems.”

The shift in emphasis is allowing the field to move beyond the debate about whether speciation happens mostly when populations are geographically separated or when they are next door (Science, 13 September 1996, p. 1496). The new view implies that both distance and habitat differences can split a species. Also, because ecological speciation is spurred by strong selection and rapid adaptation, this model fits well with field data showing that evolution can be rapid and that a few mutations of large effect can support key adaptations (see sidebar). “We're trying to find what causes [speciation],” says Patton, “and we're finding that geographic isolation by itself doesn't always provide the best answer. Something else is driving it—and we think that ‘something else’ is often the ecology.”

## How fish choose their mates

Everyone agrees that geographically isolated populations do drift apart in the wild. But populations not isolated by geography have to be actively pushed down separate paths by natural selection. So one step in documenting the ecological speciation model is simply to show that natural selection does indeed push populations to diverge, and a number of studies have documented this with everything from Darwin's own Galápagos finches (Science, 26 February, p. 1255), to Trinidadian guppies (Science, 28 March 1997, pp. 1880 and 1934).

In some cases, selection can cause morphological change surprisingly quickly. For example, in an unpublished study of Cameroon rainforest birds called little greenbuls, Thomas Smith of San Francisco State University found that after 20 years in a secondary, more open forest, the birds evolved longer wings at the “huge rate” of 120,000 darwins. A darwin is a unit of proportional change per unit time, and artificial selection experiments on mice show rates of up to 200,000 darwins.

Such speed is important to help explain bursts of speciation, such as that seen in the cichlid fish in Africa's Lake Victoria, which have evolved into hundreds of species in only 12,000 years. But did minute ecological differences actually trigger the formation of that rich diversity? No one has actually witnessed the birth of a species in the wild, so researchers must come up with clever experiments to see whether differences in ecology, and the adaptations they spur, can isolate species reproductively. Dolph Schluter, an evolutionary biologist at the University of British Columbia (UBC) in Vancouver, and his colleagues are addressing this question using marine and stream-dwelling stickleback fish from British Columbia and their ecologically similar but genetically distant counterparts in Japan; the freshwater fish may have speciated from marine ancestors only 13,000 years ago.

The freshwater descendants on both sides of the Pacific look nearly identical to one another—small and husky, with deep jaws and snouts that point down so they can suction food off the bottom of the stream bed. In both cases, the dumpy shape lets the fish swim while feeding from the bottom. The ancestral marine forms of all these fish are also remarkably similar, with streamlined, torpedo-shaped bodies. “It's a great example of parallel evolution; I guarantee you can't easily tell them apart,” says Schluter.

Indeed, even the fish can be fooled. To see whether ecology was a strong force in causing reproductive isolation regardless of geography and shared history, Schluter teamed up with Jeffrey McKinnon of the University of Wisconsin, Whitewater, and Seiichi Mori of the University of Gifu-keizai in Ogaki, Japan, gathering Japanese and Canadian sticklebacks and putting them in his tanks at UBC. The team released female Japanese marine sticklebacks, heavy with eggs, one at a time into a tank containing a single male—a Canadian marine or freshwater form, or a Japanese freshwater form. If the female found the male acceptable as a mate, she entered his nest.

Geographic speciation would predict that fish from each continent—most similar genetically—would mate with each other and be reproductively isolated from those from the other continent, says McKinnon. But in this case, ecology won out. Japanese marine females spurned their closely related freshwater cousins but mated with their distantly related Canadian marine counterparts. Canadian freshwater females also accepted Japanese freshwater mates, and vice versa; these crosses produced viable hybrids, says Schluter. Thus, both the marine and freshwater species preferred to mate with others from their own environment rather than with more closely related fish from a different habitat.

Further analysis of the data showed that the females were choosing their partners based primarily on size. Thus, in terms of reproductive isolation in the sticklebacks, “genetic history doesn't matter,” says Schluter. “It's how they look that counts.” Fish from different environments were most likely to be reproductively isolated, even if they had close genetic and geographic ties.

Other researchers praise the work. “They've shown that common environmental differences can produce common patterns of speciation,” says UC Santa Barbara's Endler, right down “to the same isolating mechanisms. It's the first time we have definite evidence of this rather than speculation.”

## Separating skinks

Halfway around the world in Queensland, Australia, Christopher Schneider, an evolutionary biologist at Boston University, is studying similar questions in the leaf-litter skink, Carlia rubrigulais, a small, reddish lizard that lives in both wet rainforest and drier open forest. The setup is perfect to test whether geography or ecology drives speciation: A well-known biogeographic barrier, the Black Mountain Corridor, physically splits the skink's range into two large populations, but on each side of the mountains the lizards inhabit both closed rainforests and more open forest. Based on the differences between the two populations' mitochondrial DNA, Schneider estimates that the single ancestral population split apart several million years ago.

Lizards living cheek-by-jowl—in some cases only 500 meters apart—in the two different forest types have similar mitochondrial DNA, suggesting recent or current gene flow between them. Yet Schneider found that the neighboring lizards vary more in size and shape than do those inhabiting the same environment on the other side of the barrier. “Morphologically, the ancient isolates are very similar,” he says, “but there are whopping great differences” in size and shape between lizards separated by “very short distances.” Open forest lizards are smaller, with shorter limbs and bigger heads, and they become sexually mature at a smaller size than those in the rainforest.

Schneider and his colleagues believe they have found an ecological force responsible for these differences: predation. Earlier reproduction and smaller size are often found in species under high predation, as individuals that manage to reproduce before being picked off are favored. More species of lizard-eating birds hunt in the open forest, Schneider notes, and by placing clay lizard models in both environments, his team gathered evidence that lizards there are more likely to be attacked.

Of course this is only one case, but as genetic data on various organisms roll in, this pattern—of geographically separated populations being similar in size and shape, while neighboring populations in slightly different habitats vary—turns out to be quite common, says Berkeley's Patton. He cites similar findings for snails and bats across the Black Mountain Corridor in Australia and rodents in the Amazon River Basin. “We find these widespread species that have deeply divergent molecular histories yet haven't changed morphologically, apparently because they continue to inhabit the same environment. Time and isolation alone don't necessarily result in new morphologies—whereas a new environment does,” he says.

And because new morphologies may lead to new species—perhaps even in the face of gene flow—the vagaries of ecology may be a driving force in more cases of speciation than researchers have imagined, Patton says. In the case of the skinks, for example, if size and shape are important in mate choice, then the ecologically distinct lizards may have taken the first step down the road to speciation, says Schneider; the critical test will be whether the geographically or ecologically separated skinks have more reproductive isolation.

Schneider's and other studies are not yet complete, and no one is ready to toss out the notion of geographic speciation. Indeed, ecology and geography may work together, says Schneider. He expects that the next round of skink studies will find the greatest reproductive isolation between populations that have been separated for a long time and also occupy different habitats. The bottom line, says Patton, is that geography alone may not be sufficient for speciation. In many cases an environmental nudge may give populations a bigger shove down the path to speciation. “That's the way to generate diversity,” he says—an observation worthy of Darwin himself.

17. EVOLUTION

# Size Matters: The Genes Behind Adaptation

1. Virginia Morell

To a fruit fly, nothing is more important than the kind of fruit it chooses to live on. Here it will hatch, dine, mate, and leave its young, and the peculiarities of a particular fruit affect almost every aspect of a fly's brief life. That's why graduate student Corbin Jones of the University of Rochester in New York expected to find a big genetic gulf between two related fruit flies of the Seychelles archipelago in the Indian Ocean: Drosophila simulans, which lives on a variety of succulent fruits, and D. sechellia, which lives only on the prune-sized Morinda fruit, a knobby, foul-smelling fruit poisonous to most insects. While D. simulans and other fruit flies struggle to evade even the scent of the Morinda, D. sechellia happily settles in to live and lay eggs.

But to Jones's surprise, it appears that this dramatic switch stems from only a few genetic differences. Genetic mapping last year showed that only a handful of genes confer resistance to the Morinda fruit's poison. And Jones's latest mapping work shows that only a few genes may account for D. sechellia's attraction to the Morinda scent. “It looks like this adaptation requires only a few genes, but with big effects,” he says.

Such findings are a surprise to many researchers, because big, beneficial mutations were thought to come along so rarely that many models simply assumed that they play no part in adaptation. But as evolutionists begin to probe the genetic basis behind important adaptations, they are uncovering examples of such large mutations, dramatically revising how biologists think about evolutionary change. “Evolution is all about adaptation, and for the first time, we're actually getting a look at the genetics of adaptation,” says H. Allen Orr, Jones's adviser and an evolutionary geneticist at Rochester. “And it seems to go against all the old models—it's faster and uses bigger genes.”

Charles Darwin, of course, knew nothing about mutations—he wrote 100 years before DNA was discovered. Even so, he thought that natural selection acted on “successive slight variations,” and for much of this century researchers agreed that evolution was the sum of many mutations of small effect. New species were thought to emerge from the slow accumulation of mutations (see main text). And the population genetics model put forward by R. A. Fisher in 1930, which Orr says is still the leading explanation of adaptation, argues that the accumulation of very small mutations is the essence of evolutionary change.

But as researchers begin to uncover the specific mutations that separate species, their first findings show the opposite pattern: Big mutations lead the way in adaptive events. In addition to the fly study, for example, evolutionary biologist Doug Schemske and geneticist Toby Bradshaw of the University of Washington, Seattle, have found through genetic mapping that bee-pollinated and hummingbird-pollinated monkey flowers in California's Yosemite National Park differ from each other in only a few sets of genes. But these few genes have large effects, changing flower color, petal shape, and the amount of nectar (Science, 13 September 1996, p. 1499)—all crucial variables for luring the two pollinators and keeping the two plant species reproductively isolated.

Such studies “show how important large beneficial mutations are in the first stages of an adaptation,” says Schemske. A new adaptation must be acquired fairly quickly, or else organisms will be poorly adapted to both the new and the old conditions and will not survive. So it makes sense that the first genetic changes have large effects, he explains; later, smaller mutations fine-tune the adaptation.

But these results also pose a problem, because “they contradict theory,” as big mutations were thought to be mostly rare and mostly disadvantageous when they did happen, says Orr. “We're in a funny situation—we're about to have a wave of data crash down on us and no theory to hang it on.” Orr has made a first stab at filling this void, presenting a mathematical model in Evolution last year showing how the big-adaptations-first pattern would work. “This is what we need,” says Schemske: “a theoretical framework for the genetics of adaptation, something we can test.”

18. EVOLUTION

# Test Tube Evolution Catches Time in a Bottle

1. Tim Appenzeller

By running experiments on microbes for thousands of generations, researchers are exploring the roles of chance and history in evolution

For most living things, 24,000 generations is a daunting span of time. Go back that many human generations, or about 500,000 years, and Homo sapiens had not yet evolved. Even for the fruit flies beloved of geneticists, 24,000 generations equals about 1500 years. But in Richard Lenski's laboratory at Michigan State University in East Lansing, 24,000 generations ago is a recent memory. The year was 1988, when he and his students first introduced 12 genetically identical populations of the bacterium Escherichia colito their new homes: 50-milliliter flasks filled with sugary broth.

Since then, those bacteria have been clocking up the generations at a rate of about one every 3.5 hours, mutating and adapting right in front of Lenski's eyes. Lenski is a founding member of a subculture of evolutionary biologists—many of them his former students and colleagues—who are watching evolution unfold in laboratory cultures of microbes, where a single experiment can span enough generations for major evolutionary change. These laboratory microcosms, whether of bacteria, viruses, or yeast, can turn evolution into an experimental science, says Michael Travisano of the University of Houston. “You have the luxury of making a prediction, and then you can test it. It's almost like physics.”

Researchers can subject populations to the same environmental stresses again and again—a procedure that Paul Sniegowski of the University of Pennsylvania calls “analogous to being able to revive the fossils and rerun the evolutionary events.” They can thaw out ancestral forms, stored in laboratory freezers in what Lenski calls a “frozen fossil record,” and compare them to their descendants. And they can monitor the microbes' genomes as they evolve, tracking the ultimate roots of those changes in DNA or RNA. “It's some of the most exciting stuff in evolution,” says Stephen Jay Gould of Harvard University.

These laboratory microcosms are allowing researchers to address some of the field's biggest questions, such as how often the twists and turns of evolution are the result of chance rather than adaptation. Researchers can study how evolutionary baggage from one round of selection affects how an organism fares in the next, and how adaptive radiations can arise from a single organism. And they can address a question that has preoccupied evolutionary thinkers like Gould: How reproducible is evolution? If the history of life could be replayed from the same starting point, how differently would it unfold? So far they are finding that identical populations facing similar conditions can follow parallel courses, although the underlying genetic changes often differ. But over time, in new environments, the effects of those differences can grow, steering evolution into radically different courses and giving chance and history ever larger roles in a population's fate.

With the enormous complexity of nature reduced to test tube systems, researchers have to approach such questions with humility, Gould notes: “Of course, you're looking at a very different world at a different time scale.” Nor can researchers even be sure that what they see in one evolutionary microcosm will apply to any other, adds Holly Wichman of the University of Idaho, Moscow, who studies evolution in viruses. “One of the questions is how well [test tube findings] are going to generalize. … Is every case going to be a special story?”

Still, the granddaddy of these experiments—the 11-year, 24,000-generation E. coli cultures in Lenski's laboratory—is telling stories about predictability, chance, and history that other experiments have echoed. All 12 of Lenski's cultures experience the same stresses: a daily boom-and-bust cycle, in which the bacteria are transferred to fresh glucose medium every 24 hours, then undergo 6 hours or so of plenty followed by 18 hours of starvation. All 12 lines have adapted to this regimen; when the researchers do a head-to-head comparison between the evolved bacteria and the ancestral strain, plucked from the freezer and revived, the descendants now grow about 60% faster in their standard glucose-containing medium. All 12 populations show other parallel changes, too—for example, a still-unexplained, twofold increase in cell size.

Yet underneath these consistent responses to selective pressure, says Lenski, “you see all this hidden variation.” The fitness increases were almost identical in all of the populations, but not quite; the cell size expanded in all 12 lineages, but by different amounts. And when Lenski and his colleagues, including Michel Blot of the University of Grenoble in France and Werner Arber of the University of Basel in Switzerland, analyzed the genomes of their adapted bacteria, the similarities vanished. By chopping up the bacteria's DNA with enzymes and applying probes that home in on known sequences, they found that after thousands of generations, the populations' genomes were riddled with changes. The changes were different in each population and had accumulated at very different rates, the group reported in the March Proceedings of the National Academy of Sciences, even though the fitness increases were similar. That indicates what the authors called “conspicuous and significant discrepancies” between genomic evolution and its visible effects.

Lenski and graduate student Mark Stanek are now trying to pinpoint the particular beneficial mutations that boosted the bacteria's fitness. They've found one so far—and it is present in just one lineage, strengthening the idea that the others have found different paths to higher fitness. When it comes to organisms' adaptive performance, says Lenski, “evolution is remarkably reproducible. But as you move away from performance, to cell size or genes, things are less and less reproducible.” Because all 12 populations started out genetically identical and have experienced the same selective pressures, the differences underscore the role of chance in setting evolution's course.

## Evolutionary baggage

The role of chance becomes even more obvious over time, as those genetic differences become part of the baggage that organisms carry to their next evolutionary challenge—baggage that can dramatically affect how they fare, as Travisano and Lenski have shown. They took samples of the 12 E. coli populations after the bacteria had been growing in glucose for 2000 generations. By that point, all 12 populations had improved their ability to grow on glucose by about the same amount. But when they were put in a different sugar, maltose, some populations thrived while others languished. For each population adapting to limited glucose, says Travisano, “it seems likely that glucose uptake was tweaked in subtly different ways. And those subtly different tweaks had big effects in a different environment.”

He and Lenski then allowed all 12 lineages of bacteria to evolve for another 1000 generations on their new staple, maltose. Evolution did its work, and after months of mutation and selection, all 12 could grow well on maltose. But the fitness improvement was not as consistent as it had been on glucose, where the starting genotype had been identical. Evolution was no longer as reproducible as before, because of chance variations in how the populations had adapted to their earlier environment. “Once we had diversity, we could prune it back tremendously with adaptation. But not completely. Once you are different, that difference tends to persist,” says Travisano.

To Travisano, the results are a lesson in the importance of prior history in shaping the way organisms respond to an adaptive challenge. They “tell you that variation arises very easily … and it doesn't arise in ways that are easily predicted.”

Other researchers are weighing the roles of predictability and chance in adaptive radiations, in which one form gives rise to many. Paul Rainey at the University of Oxford in England seeds vials of sugar water with cells of the common plant bacterium Pseudomonas fluorescens. He avoids shaking the containers, allowing the environment to stratify into regions that are chemically and physically different, with oxygen-rich layers near the surface and oxygen- depleted but nutrient-rich layers beneath. The result is a diverse array of ecological niches for the bacteria to fill—what an animal species newly arrived on an empty continent might find. He then follows their evolution for 10 days.

In his original work, done with Travisano and published in Nature last year (also see Science, 17 October 1997, p. 390), Rainey found that in virtually every one of these microcosms, the bacteria evolve into three major forms. He named them for the appearance of their colonies when he grows them on culture plates: wrinkly spreader, fuzzy spreader, and smooth morph, which is the unchanged ancestral form. Each has a taste for a particular niche, with the wrinkly spreader congregating at the surface of the broth, the smooth morph spreading through the liquid, and the fuzzy spreader hugging the bottom.

Rainey is now trying to account for these tastes. So far, he and his students have learned that wrinkly spreader overproduces a cellulose-based polymer, which helps glue the cells together into a mat. The mat supports them at the surface, where the wrinkly spreader cells benefit from the abundant air supply.

These miniature adaptive radiations unfold in the same way every time, governed by the available environmental niches. And Julian Adams, at the University of Michigan, Ann Arbor, saw some of the same repeatability in his experiments, where diversity arises seemingly out of nothing. Adams, with Frank Rosenzweig, now at the University of Idaho, Moscow, and their colleagues, grew genetically identical E. coli populations in a device called a chemostat, which kept conditions for the bacteria blissfully constant, except for a steady shortage of glucose. But in spite of this uniformity, two or more E. coli variants—an ecosystem in miniature—regularly made their appearance after around 200 generations, or about a month, says Adams.

The group first got a clue that one strain had turned into several when they extracted samples from their cultures, grew them on plates, and saw colonies of different sizes, rather than the uniform colony size expected of genetically uniform bacteria. “The differences were so dramatic that we thought we had contamination” and shut down the system, Adams recalls. “I don't want to tell you how many times we did that before we cottoned on to what was happening.” He and his colleagues went on to show that at least two strains had evolved in their chemostats.

Originally, Adams explains, natural selection favored mutants that had a souped-up appetite for glucose and so could outgrow its neighbors. But bacteria can metabolize only so much glucose; as their biochemistry got clogged with the sugar, the glucose-hogging mutants shunted the excess from aerobic metabolism to the less efficient anaerobic pathway, which generates a waste product, acetate. As Rosenzweig, Adams, and their colleagues described in the August 1994 issue of Genetics, the acetate buildup created a new ecological opportunity, and eventually a mutant emerged that could fill it: a new acetate-scavenging strain. Adams and his colleagues reported last summer in Molecular Biology and Evolution that the acetate scavengers appeared in six out of 12 populations they studied, and each time a mutation in the regulatory region of a gene that influences acetate uptake was responsible.

“It's the first stage in speciation,” says Adams. “Diversity can exist even if you don't seed it with something that can drive diversification.” And like other studies, this one shows that diversification is not only inevitable but also follows a predictable course.

But even if the general outline of such experiments is predictable, in many cases the genetic pathway they take depends on chance, as Travisano saw when he transferred glucose-adapted bacteria to maltose. That seems to be the case for Rainey's wrinkly spreader strains, too. When his group took 24 wrinkly spreader strains that had evolved independently and then forced them to evolve back into a smooth form by shaking their vials to keep the culture medium from becoming stratified, Rainey says, “some go back easily; some sort of struggle,” implying differences in their genetic makeup. Thus, Rainey concludes that “you can become wrinkly spreader by a variety of different paths.”

The influence of chance and history on how organisms diversify is still more vivid in Rainey and his students' new experiments, in which they introduce an additional evolutionary force: a predator. After allowing the microcosms to diversify, they infect them with a bacteriophage, a virus that kills bacteria. The population crashes, then rebounds as a resistant strain takes over. The resurgent strain diversifies again—but it does so differently within each microcosm, spawning odd new variants including a strain that secretes a mucoid slime.

“What it comes down to is just a chance thing,” Rainey says. “The phage puts the population through a bottleneck, which increases the role of chance. The reproducibility goes out the door.” Only individuals that happen to be resistant to the phage pass through the bottleneck, and the array of genes they carry varies from microcosm to microcosm. As a result, each miniature ecosystem rediversifies from a different starting point and reaches strange new adaptive peaks.

## Carbon-copy evolution

In some experiments, however, evolution seems truly reproducible down to the level of genes—for example, Adams's work in which genetically similar acetate mutants appeared six times out of 12. Now researchers are trying to work out why. Travisano, for example, has reversed the experiment in which he switched glucose-adapted bacteria to a diet of maltose and saw a wide variety of responses. In work published in the June 1997 issue of Genetics, he adapted 12 identical populations of E. coli to a restricted diet of maltose. After 1000 generations, he switched them to glucose. But this time, every population responded to the diet switch in the same way, continuing to thrive. Apparently all 12 populations had evolved in the same way—perhaps, Travisano suggests, because bacterial physiology offers just one way to do better in maltose, forcing all of the populations down the same evolutionary path.

Similarly, Wichman, James Bull of the University of Texas, Austin, and their colleagues have found that the mutations underlying high-temperature adaptations in a particular bacteriophage are surprisingly reproducible, right down to the specific changes in the DNA sequence. Now Wichman, Bull, and their students are trying to identify the factors that favor this kind of predictability. “It's really too early to tell what the rules are,” she says, but she is enjoying her privileged view of evolution. “It's amazing to watch changes sweep through a population in a way we knew happened but had never seen before.”

19. EVOLUTION

# Early Life Thrived Despite Earthly Travails

1. Richard A. Kerr

Life learned resilience in a school of hard knocks, where rocks rained down from the sky, the climate control was precarious, and oxygen was either too scarce or too abundant—and that was just the first 2 billion years

Earth's youth is its most mysterious epoch, only dimly registered in a geologic record largely erased by billions of years of plate tectonics and erosion. Yet the first 2 billion years of our planet's history saw the first stirrings of life, when systems of molecules began reproducing themselves and deriving energy from chemicals and from sunlight. Because fossils from those early eons are rare and frustratingly cryptic, many researchers trying to understand the origins of life are turning to geology, hoping to learn the state of the planet's surface and atmosphere when it first hosted living things.

This approach to understanding life's early days joins clever techniques to wrest more direct information about life from long-dead rocks (see sidebar) and laboratory efforts to explore the likely chemistry of early life or infer its genetic nature from the genes of living creatures. But the geologic approach is adding unique plot elements to the story of early life: a series of hair-raising escapes.

For its first half-billion years, Earth endured a punishing rain of impacts, which vaporized the oceans and scorched the globe so fiercely that some researchers now propose that life could have first evolved on a more hospitable world, then later hitchhiked to Earth on a meteorite. After the heavy bombardment stopped came a new trial: The young sun was faint and relatively cold, leaving Earth perennially teetering on the edge of a planet-enveloping ice age. And as that threat began to lift, oxygen—for eons little more than a nuisance waste product—suddenly flooded the atmosphere, threatening the anoxic life of the time but also perhaps sending the planet into the threatened deep freeze.

Obviously, life survived all these trials—and emerged with resilience as its prime characteristic, says paleontologist J. William Schopf of the University of California, Los Angeles. “Environment and life go hand in hand,” he says. The imprint of those cataclysms remains in the hardy survivors from early times, such as blue-green algae, and indeed in the genes of all life, as the ancestor of every organism on Earth passed through these trials. “Life is persistent,” concludes paleontologist Andrew Knoll of Harvard University. “It can absorb a range of shocks from the environment.”

## Heavy bombardment

In the beginning, from 4.5 billion to 3.8 billion years ago, rock bodies left over from solar system formation, called planetesimals, were still battering Earth and the other planets. At least one of these impactors was gigantic; the size of Mars or larger, it struck Earth within about a hundred million years after the planet began forming and dislodged a mass of material that became the moon. Soon thereafter, water from Earth's own rock and impacting comets collected to form a planetwide ocean, providing a cradle for life.

But the pummeling wasn't over, as the 1000-kilometer-wide impact basins still visible on the moon as its dark “seas” testify. Earth, being a bigger target with a stronger gravitational pull, would have suffered blows from hundreds of objects of that size between 4.0 billion and 3.8 billion years ago, geophysicist Norman Sleep of Stanford University and planetary physicist Kevin Zahnle of NASA's Ames Research Center in Mountain View, California, noted in the Journal of Geophysical Research last year. A few of these impactors were probably 500 kilometers in diameter—big enough to create a superheated atmosphere of vaporized rock that would in turn have vaporized the oceans for 2700 years and sterilized even the subsurface, say Sleep and Zahnle.

Yet despite this brutal environment, the very first, simple organisms emerged at most only a few hundred million years after the bombardment stopped. The first recognizable fossils appear at 3.5 billion years ago, and there are controversial isotopic traces of life as far back as 3.7 billion or 3.8 billion years ago (see timeline). Indeed, life might have arisen even earlier, during the millions of years between sterilizing impacts—only to be wiped out by the next big one, say planetary physicists Kevin Maher and David Stevenson of the California Institute of Technology in Pasadena, who call this the “impact frustration” of life.

Even after the large, sterilizing impacts ended, smaller impacts continued until about 3.8 billion years ago, say Sleep and Zahnle, so any survivors were probably adapted to living deep in the crust where temperatures are high. That may explain why the genes of living microbes suggest that the ancestor of all life had much in common with today's hyperthermophiles, microbes that thrive in hot springs at temperatures of 113°C or more.

To Sleep and Zahnle, the severity of the bombardment coupled with the early appearance of microbes suggests another possibility: that Earth was seeded with life from elsewhere, namely Mars. Calculations have shown that rocks blasted off Mars by large impacts might have fallen to Earth quickly enough to deliver martian microbes, perhaps as rock-encased spores, before they succumbed to the rigors of space.

And Sleep and Zahnle say that in those days, Mars had the more benign environment. It is a smaller target than Earth, and its weaker gravity would pull in fewer planetesimals, so big, potentially sterilizing impacts were less frequent there. Mars probably had no oceans, or only small and shallow ones, so the steam from a 500-kilometer impactor would have condensed out within a mere decade, giving subsurface life a fighting chance of survival. And Mars's interior was cooler, say geophysicists, allowing microbes to penetrate further into the planet, away from the searing surface. “It's possible Mars would have been more suitable for life,” agrees Stevenson.

## The snowball threat

Even as blasts of heat were weeding out cold-adapted microbes, early life faced another threat: a long-term deep freeze. When Earth formed 4.5 billion years ago, astrophysicists believe, the young sun provided 30% less warmth than it does today. All else being equal, Earth and all its oceans should have been frozen over from pole to pole until the sun had brightened significantly, about 2 billion years ago, says planetary climatologist James Kasting of Pennsylvania State University (PSU) in University Park.

Yet the record shows that this chilly fate didn't befall Earth—or at least not for very long. Not only are there clear signs of life at 3.5 billion years ago, there are signs of running water and erosion, too. And traces of photosynthesis—a telltale pattern of isotopes—found in marine rocks from about 2.7 billion years ago make it seem unlikely that the oceans were constantly frozen over, says Knoll. The first fossil of a eukaryote—the term for organisms, from yeast to humans, that have a cell nucleus—appears in the form of the alga Grypania at about 2.2 billion years ago, when the ice would have been breaking up. So what kept the world from plunging into a deep freeze and allowed life to thrive and diversify?

For years, researchers thought that the solution to this “faint young sun” paradox was an atmosphere with 300 to 1000 times as much carbon dioxide as today's, creating a powerful greenhouse effect. But such high levels of the gas would have combined with iron in the soil to leave iron carbonate in ancient rock, which has not been found. The late Carl Sagan and cosmochemist Christopher Chyba of the SETI Institute in Mountain View suggested that high levels of atmospheric ammonia could have shored up the early greenhouse (Science, 23 May 1997, p. 1217). But that idea has problems too, because the sunlight-sensitive ammonia would have required a methane haze for protection. Yet that haze would ultimately have reflected light back to space and cooled Earth as much as the ammonia greenhouse could warm it, as cosmochemist Christopher McKay of Ames reported in Icarus early this year.

Now Kasting is offering another solution: a methane greenhouse spawned by life itself. Methane is a powerful greenhouse gas, and at July's International Conference on the Origin of Life in San Diego, Kasting will argue that methane produced by ancient methanogenic bacteria—the ancestors of organisms now responsible for the methane oozing from swamps, river bottoms, and landfills—could have reached levels 1000 times higher than today's.

The key to such high methane levels is a lack of free atmospheric oxygen, says Kasting. A molecule of methane wafting into today's atmosphere of 20% oxygen lasts about 12 years, on average, before being oxidized to carbon dioxide. But Kasting believes oxygen was nearly absent 2.8 billion years ago, so that a methane molecule could have survived 20,000 years, according to calculations he did with Lisa Brown of PSU. Combine that with a bit of carbon dioxide, which would prevent a cooling methane haze from forming, and the paradox could be resolved: By exuding a methane blanket, life itself could have warmed the potentially frigid world to within a few degrees of its current temperature.

McKay thinks the idea is plausible, although he says more lab work on the atmospheric chemistry of methane is needed. Studies of minerals and isotope ratios from before 2.2 billion years ago do suggest that microbes were turning out methane and free oxygen was scarce. Most researchers agree that although marine photosynthesizers were cranking out oxygen by 2.7 billion years ago, almost all of it was used up in oxidizing things like the copious iron dissolved in the ocean and emissions from volcanoes, and little made it to the atmosphere.

Still, the existing geologic data put the upper limit on atmospheric oxygen at around 0.1%—and Kasting's scenario requires virtually no oxygen in the atmosphere, notes geologist Roger Buick of the University of Sydney in Australia. Kasting recognizes the problem. “The story I'm telling is self- consistent,” he says. “It's just hard to prove.”

If methanogens did warm the world, other microbes may have inadvertently cooled it, by pumping out oxygen. At about 2.2 billion years ago, just as the sun was strengthening and eukaryotes appeared, oxygen levels apparently jumped in what Buick calls “one of the biggest environmental changes of Earth history.” Isotopic and mineralogical data have suggested that atmospheric oxygen rose to perhaps 10% to 15% of the present-day level, and the most recent data suggest that the change was abrupt, from 1.8 billion to 2.2 billion years ago (Science, 5 March, p. 1519). Somehow, the oxygen being turned out by photosynthesizers finally gained the upper hand, perhaps because volcanoes had slowed their emission of compounds that react with oxygen. “It's looking suspiciously like [the oxygen spike] might be real,” says Buick.

Such an oxygen jump would have driven some major changes, forcing anoxic organisms to adapt, perish, or hide out in the oxygen-poor mud on the sea floor and lake beds, where they remain today. Even worse, if Kasting's scenario is right, oxygen would have destroyed the methane and collapsed the greenhouse. “You can imagine that would trigger [Earth becoming] a big snowball,” says Kasting.

A snowball Earth is a startling idea, but that's just what paleomagnetist David Evans of the University of Western Australia in Perth and his colleagues suggested in 1997. They found 2.2-billion-year-old glacial sediments in South Africa that apparently formed when the site lay near the equator and close to sea level, and they concluded that glaciers penetrated deep into the tropics just when oxygen appears to have shot up. If ice were in the tropics, they reasoned, it probably covered the rest of the planet, including a thick layer on all the oceans.

Most researchers aren't quite ready to accept this scenario. But evidence reported last year strengthened the case for another global deep-freeze, 600 million years ago, when oxygen was rising again and multicellular animals may have been appearing (Science, 28 August 1998, p. 1259). The more recent cold snap, perhaps triggered by waning atmospheric carbon dioxide levels, lends a touch more credibility to the older one.

Such a global freeze would mean hard times for Earth's microbes—but not extermination. Photosynthesizers, for example, could have gotten by with the low levels of light trickling down through thin spots or cracks in the ice. However it managed, life not only survived but thrived after the snowballs, just as it had dodged impacts, suffocation, and oxygen poisoning. From its beginnings, it seems, life has been honed by crisis.

20. EVOLUTION

# Going Beyond Appearances to Find Life's History

1. Gretchen Vogel

You wouldn't think of watching a 3D movie without the proper goggles, but that's the situation of paleontologists trying to view the early history of life on Earth from the small, flattened blobs of rock that are all that's left of the earliest organisms. Researchers have been missing an entire dimension of information hidden in these fossils: the chemical traces of life, left behind billions of years ago as organisms took in nutrients and built cellular structures. Now, with an array of high-tech instruments and an influx of funding from NASA—which seeks new ways to track possible life-forms on other worlds—researchers are starting to read these faint chemical signatures.

Isotopic and chemical methods are uncovering billion-year-old traces of cellular and even molecular structure, even in rocks where no visible fossils remain. “This is a fundamentally different way of looking at the evolution of life,” says paleobiologist J. William Schopf of the University of California, Los Angeles. “There are a number of us who have been looking for tools for 20 years to sort out the biochemistry and physiology and metabolism of ancient life. Now the technology has finally come along that allows us to get at these questions.”

The earliest fossils themselves are few and far between. The most ancient are 3.5 billion years old, but the cells are arrayed in long filaments, mats, and other complex clusters—implying a long history of even earlier evolution that is missing from the record. And determining what kind of organism is preserved in a microfossil by looks alone is difficult at best. Inorganic activities can leave little round blebs behind too—hence many researchers' skepticism about putative martian fossils (Science, 20 November 1998, p. 1398). “All [researchers] have is the morphology to look at,” says geochemist Clark Johnson of the University of Wisconsin, Madison. “Everyone thinks [microfossils] are prokaryotes [when] they're little. When they get bigger they think, ‘Well, they must be eukaryotes.’”

So researchers are striving to put the science of tracking life on a more sophisticated footing. Whereas some scientists try to infer life's history by reconstructing the environment on early Earth (see main text), others are trying to glean more detail from the fossils themselves. NASA's \$24 million astrobiology program, which funds this and other work, began only last year, but it is building on some promising results from earlier studies.

To find the signature of life itself, some methods exploit the fact that living organisms tend to preferentially take up lighter isotopes of various elements. So, for example, former cells leave traces of carbon that have unusually high amounts of carbon-12 compared to the heavier isotope carbon-13. Researchers have found such isotopic anomalies in some of the earliest known sedimentary rocks—so-called banded iron formations from Greenland, which are older than 3.7 billion years (Science, 29 January, p. 674). But this kind of evidence doesn't convince everyone. “We would like to believe that those low [carbon isotope] ratios are indicative of life,” says Johnson. “But there are no fossils to prove it,” and heating, high pressure, and weathering can produce similar isotope signatures.

Isotopes of other elements might be more persuasive. Organisms also preferentially take up light isotopes of iron, for example, and this element is heavy enough that its isotope ratio is not thought to be affected by most geologic processes. Johnson and his colleagues Brian Beard and Kenneth Nealson of NASA's Jet Propulsion Laboratory in Pasadena, California, have found that iron in the purported bacterial layers of banded iron formations has relatively less of the heavy isotope iron-56 than other rocks (Science, 4 December 1998, p. 1807).

Other techniques promise to reveal what kind of life-form created a particular microfossil. For example, geochemist Roger Summons of the Australian Geological Survey Organization in Canberra noted that bacteria and eukaryotes—complex cells with nuclei, like those in our bodies—differ in much more than size. Bacterial cell membranes include hopanoids, large organic molecules with a five-ring carbon backbone, whereas eukaryotes produce sterols—similar molecules with a four-ring backbone. Billions of years of heat and pressure transform the cell's remains into insoluble organic matter—but, amazingly, the telltale backbones of the rings can remain intact and can be detected.

Summons and colleagues analyzed the hydrocarbon content of 2.5-billion-year-old black shale from Western Australia. At a recent meeting on ancient fossils,* he reported finding molecules characteristic of both bacteria and eukaryotes—300 million years before the first suspected eukaryotic fossils. “It's very, very nice work, and very difficult work,” says Schopf. “Now we've got to find [biomarkers] of intermediate age to show that this is part of an evolutionary continuum.”

Studies using a variety of other clever methods are in the pipeline. Schopf is using ion probes to measure the isotope ratios in individual microfossils and learn about their metabolism; other researchers are tracking the evolution of biological polymers with nuclear magnetic resonance spectroscopy, which can identify telltale carbon-nitrogen and carbon-carbon bonds. And Nealson and colleagues are blasting synchrotron radiation at rocks between 2.5 billion and 3.2 billion years old. The brilliant x-rays could reveal the remains of bacteria inside iron or manganese oxide deposits; and because they cause different molecules to fluoresce in predictable ways, they could also reveal chemical composition.

The new techniques will have their pitfalls and may take some years to perfect, cautions Schopf. “We're crawling along now,” he says. “But at least we're moving forward.”

• *Bridging Two Worlds: From the Archean to the Proterozoic, University of California, Los Angeles, 18 to 20 February.