# News this Week

Science  23 Feb 2007:
Vol. 315, Issue 5815, pp. 1062
1. U.S. 2007 BUDGET

# NSF Enjoys a Heartfelt Ending to a Difficult Budget Year

1. Jeffrey Mervis*
1. With reporting by Adrian Cho, Jennifer Couzin, and Eliot Marshall.

On Valentine's Day, the U.S. Congress sent the National Science Foundation (NSF) a long-delayed token of its warm feelings for the $6 billion basic science agency. It came at the end of a marathon budgetmaking process that stretched 4 ½ months past its intended deadline. The card was signed by the new House Speaker, Representative Nancy Pelosi (D-CA). The valentine was a$334 million increase in NSF's $4.4 billion research budget that matches its 2007 request. It was tucked into a$463 billion, yearlong spending plan that displays little budgetary love for most of the other civilian agencies in the package. The National Institutes of Health (NIH) and the Department of Energy's (DOE's) science programs were also modest winners, picking up $620 million and$200 million, respectively, while NASA's space science programs continue to be squeezed by plans to explore the moon and Mars. Most agencies get no increase over 2006 in the bill, which covers the rest of the 2007 fiscal year that began 1 October.

For many scientists, the year began on a high note some 54 weeks ago, when President George W. Bush proposed major increases for the physical sciences as part of his American Competitiveness Initiative (Science, 10 February 2006, p. 762). The increases, part of a proposed 10-year doubling for NSF, DOE science, and the core labs at the National Institute of Standards and Technology (NIST), were embraced last summer by House and Senate spending panels. But those efforts went by the boards after the outgoing Republican-led Congress failed to pass a spending bill and the new Democratic majority announced that it planned to hold 2007 spending to 2006 levels (Science, 5 January, p. 24).

Any boost for research in the so-called joint funding resolution is a tribute to the bipartisan support for science within Congress, say lobbyists. In particular, the bill tracks with an innovation agenda released by congressional Democrats, led by Pelosi, nearly 1 year before they were voted into power. And although science advocates pushed hard for increases that would match the president's levels, they were not optimistic about the outcome of negotiations that took place behind closed doors from mid-December to late January. So the result, which was approved without amendments by the House on 31 January (ScienceNOW, sciencenow.sciencemag.org/cgi/content/full/2007/130/1) and by the Senate 2 weeks later, was a pleasant surprise. NSF's research account received its full 2007 request, DOE's $3.6 billion Office of Science received 40% of its$506 million request, and NIST's core labs gained half of a proposed $104 million boost. “There are two words why any research money was added to the House bill: Nancy Pelosi,” says John Palafoutas, chief lobbyist for AeA (formerly the American Electronics Association). “The [congressional] staff was telling us not to get our hopes up and that research would be flat-lined along with everything else.” Despite strong letters from hightech CEOs and university and scientific society presidents urging support for basic research, he said, all indications were that adding money “was too big a lift.” An appropriations staffer who requested anonymity describes a free-for-all in which members' priorities were matched up with the money on the table. “We all submitted our lists [of exceptions to the 2006 spending levels], and some of them included offsets,” says the aide. “And I have to tell you, I was stunned by the NSF level.” The 2007 spending bill contains a special valentine for agencies whose budgets are riddled with congressional earmarks, projects inserted by individual legislators that were not requested by the agency. The Democrats' decision to erase all 2006 earmarks allows agencies to spend the “extra” money as they see f it, subject to the approval of their operating plan by appropriators. There was$128 million in such earmarks in DOE's 2006 science budget, for example, and $137 million in NIST's budget. Even so, some legislators are keeping close tabs on their favorite projects. For example, this year was to see the second of five$15 million awards, funneled through the Department of Veterans Affairs (VA), to the University of Texas Southwestern Medical Center in Dallas for research on a controversial theory linking neurotoxins to the mysterious symptoms that plagued veterans after the first Gulf War. Marc Short, a spokesperson for porkbarreling Senator Kay Bailey Hutchison (R-TX), says that the VA “signed a contract with UT Southwestern, … [and] we certainly would expect the VA to honor that commitment.”

The biggest wrinkle yet to be ironed out of the bill is whether agencies can use their money to start projects and programs. The previous spending measure that governed the first 4 ½ months of the 2007 fiscal year said no, but the final bill contains no such restrictions.

NSF Director Arden Bement hopes to begin construction on three major research facilities in his 2007 budget: the Ocean Observatories Initiative, the Arctic Research Vessel, and the National Ecological Observatory Network, although each one is being “rescoped” to reconcile their scientific objectives with rising costs. But DOE appears to be taking a different tack, according to lab officials. Brookhaven National Laboratory in Upton, New York, foresees a 20-week run of its Relativistic Heavy Ion Collider—which last year relied on a $13 million private donation—but no new design work for the proposed$775 million National Synchrotron Light Source II. A delay also seems likely for a suite of instruments known as SING-II at the Spallation Neutron Source at Oak Ridge National Laboratory in Tennessee, and a neutrino experiment called NOvA at Fermi National Accelerator Laboratory in Batavia, Illinois. At the same time, the Thomas Jefferson National Accelerator Facility in Newport News, Virginia, hopes to press ahead with an upgrade of its CEBAF particle accelerator even if it means running fewer experiments this year.

NIH hopes to make an additional 500 research grants, including $91 million for a new investigator fund,$40 million for short-term, high-risk “junior pioneer” awards, and $69 million for the National Children's Study (Science, 9 February, p. 751). The 2% overall boost may look small for a$28.3 billion agency, but it is a “tremendous victory,” says Jon Retzlaff, legislative director of the Federation of American Societies for Experimental Biology, compared to flat funding in the president's 2007 request and some 450 programs that have been cut from 2006 levels.

NASA is among that group, and the chair of the agency's spending panel, Senator Barbara Mikulski (D-MD), apologized for not doing better during the negotiations. “This joint funding resolution is not what anyone wanted,” she declared immediately after the vote. The only silver lining Mikulski could find in a $16.2 billion budget that falls$400 million below 2006 levels is that science accounts were not raided to provide some $460 million designated for new exploration vehicles. Despite the expectation that 2008 will be another tough budget year, Pelosi's spokesman, Drew Hammill, says that the 2007 budget “is the first step of good things to come for science funding.” Lobbyists sure hope he's right. “Science continues to win bipartisan support,” says Joel Widder of Lewis-Burke Associates in Washington, D.C. “But the arithmetic still stinks.” 2. PRIMATE BEHAVIOR # Spear-Wielding Chimps Seen Hunting Bush Babies 1. Ann Gibbons The right to bear arms has long been considered a distinctly human privilege. But apparently the Second Amendment to the U.S. Constitution applies to chimpanzees too, at least while they're out hunting small game. Researchers in Senegal recently spotted wild chimpanzees biting the tips of sticks, which they then used like spears to jab small primates called bush babies. Anthropologist Jill Pruetz of Iowa State University in Ames was astonished when her project manager saw a chimp thrust a sharpened stick into a hole in a tree and pull out a limp bush baby to eat, according to a report in the 6 March issue of Current Biology. “This is stunning,” says primatologist Craig Stanford of the University of Southern California in Los Angeles. It's the first time a nonhuman primate has been known to make a lethal weapon for hunting other animals, he says. “This is no anecdote, as they have 22 cases,” adds primatologist William McGrew of Cambridge University in the U.K. “Once again, chimpanzees exceed our imaginations.” Anthropologists have known for some time that chimpanzees are adept at making and using stick and stone tools, for example to probe termite mounds or crack nuts. And researchers have seen gangs of male chimps kill monkeys by beating and biting. But they thought only humans used tools to hunt. Pruetz's team, working at the Fongoli research site in the wooded savanna of Senegal, observed chimps breaking off green branches and in four cases using their incisors to sharpen the points. The chimps, which typically weigh 26 to 60 kilograms, were hunting nocturnal bush babies, 100- to 300-gram primates that hide by day in holes in trees. In all, Pruetz and Paco Bertolani, a graduate student at Cambridge University, documented 10 different chimps thrusting the tools into holes in 22 instances. “This is habitual,” says Pruetz, whose team logged 2500 hours of observations. Other researchers were impressed by the observations, although some noted that the researchers saw only one bush baby actually killed. “Could they have been rooting around for something else?” asks primate behavior ecologist John Mitani of the University of Michigan, Ann Arbor. Pruetz says the chimps' intent was clear: They jabbed the sticks in the holes with enough force to injure prey and far more vigorously than when probing for termites. And bush baby remains were common in chimp feces, indicating they were regular fare. In another surprising twist, most hunters were females. “It's a double whammy,” says Pruetz. “It doesn't fit the old paradigm of Man the Hunter.” Make that Chimp the Hunter. 3. SCIENTIFIC PUBLISHING # European Union Steps Back From Open-Access Leap 1. Martin Enserink BRUSSELS—Europe took center stage last week in the growing battle for free access to the results of publicly funded research. An online petition, signed by almost 14,000 researchers and 500 research organizations in the European Union (E.U.) and presented here at the start of a 2-day meeting, asked the European Commission to take bold action on so-called open access. Traditional scientific publishers launched a counteroffensive, arguing that the future of scientific communication—as well as their €3 billion European industry—is at stake. For the moment, the publishers' argument has carried the day: In a policy brief, the commission failed to enact a mandatory open-access policy for E.U.-funded scientists, to the disappointment of ardent supporters of the petition. “This doesn't reflect the spirit of what's happening in Europe,” says cognitive scientist Stevan Harnad of the University of Southampton in the United Kingdom. Open-access proponents argue that scientific papers should be available to everyone for free, instead of only from publishers at a cost. One way to achieve this goal is to ask researchers to make a copy of each paper freely available online, perhaps on their institute's Web site—a step called “self-archiving.” The U.S. National Institutes of Health asks researchers to do this on a voluntary basis; legislation to make it compulsory for most taxpayer-funded researchers has stalled in the U.S. Congress. Meanwhile, five research councils in the United Kingdom have made self-archiving within 6 months of publication mandatory, as have other research funding agencies. If the E.U. required the same from the scientists it funds through its €50 billion Seventh Framework Programme, many individual countries—within and outside the E.U.—might follow suit, contends Harnad. “It would be terrific if this big domino fell,” he says. Indeed, a commission-sponsored study of the publishing industry by Belgian and French academics recommended mandatory self-archiving in January 2006, as did a December report by the commission's European Research Advisory Board. The brand-new E.U.-funded European Research Council also supports the idea. But mandatory self-archiving has met stiff resistance from most scientific publishers. Making papers freely available after just 6 months may lead librarians to cancel subscriptions, threatening the entire publication system, they said at the 2-day, commission-sponsored meeting here. Publishers also questioned the economics of new-style online journals—such as the Public Library of Science (PLoS)—in which authors pay to publish and access is free. Such a business model is too young to know if it can work, they said. The commission agrees, for now. In a 14 February policy statement, it acknowledged that data from publicly funded research “should in principle be accessible to all” and offered steps to move in that direction, such as a promise to reimburse scientists publishing in journals such as PLoS. But it didn't endorse a mandate to self-archive, asking for more studies and debate instead. Robert Campbell, president of Blackwell Publishing, calls it a “sensible and encouraging” position. But Harnad says the commission's steps are “wishy-washy.” It appears to be protecting publishers' interests without realizing that open access would have much greater economic benefits overall, he says. Other supporters of open access take a more optimistic view. The commission is still new to the debate and may come around, notes Sijbolt Noorda, chair of the Association of Universities in the Netherlands. “Rome wasn't built in one day.” 4. MATHEMATICS # Quasi-Crystal Conundrum Opens a Tiling Can of Worms 1. John Bohannon The mosques and palaces of the medieval Islamic world are wonders of design. Because tradition forbids any pictorial decorations, they are covered with complex and intricate mosaics. These geometric patterns, called girih in Arabic, may be even more sophisticated than has been appreciated. On page 1106, physicists Peter Lu of Harvard University and Paul Steinhardt of Princeton University propose that architects made a conceptual breakthrough sometime between the 13th and 15th centuries. By first visualizing a surface as a tiling of polygons, these unknown scholars created girih patterns that conform almost exactly to a pattern called a quasi-crystal. If Lu and Steinhardt are right, then the Islamic world discovered a piece of mathematics 500 years before it was formally described in the West. But the paper has also sparked a rancorous dispute over who first made this insight, and whether it is true at all. Starting in the 1960s, mathematicians studying the geometry of tiling came up with the concept of the quasi-crystal. Tiling is crystalline if it is made up of an infinitely repeating pattern of some finite set of units. Quasi-crystals are also made up of a finite set of interlocking units, but their pattern never repeats even if tiled infinitely in all directions. Researchers also found that although pentagons and decagons don't fit easily into normal tiling, in a quasi-crystal such fivefold and 10-fold rotational symmetries are integral. The most famous quasi-crystal pattern is “Penrose tiling,” named after Oxford University mathematician and cosmologist Roger Penrose. In 2005, Lu, a doctoral student at Harvard, noticed a geometric pattern on the wall of an Islamic school in Uzbekistan with surprisingly complex decagonally symmetric motifs. “It got me thinking that maybe quasi-crystals had been discovered by Islamic architects long ago,” he says. Islamic architects began to explore motifs with fivefold and 10-fold rotational symmetry during a flourishing of geometric artistry between the 11th and 16th centuries. Back at Harvard, Lu began to study architectural scrolls from that period. On many scrolls, faintly sketched beneath the intricate lines of the girih design, was a polygonal tiling pattern. “I found the outlines of the same tile shapes appearing over and over,” he says. Lu realized that Islamic architects could have used a pattern of polygonal shapes—which he calls girih tiles—as the starting point for their designs, creating a wonderfully complex girih pattern by tracing lines from tile to tile following local rules. And if the right shape of girih tiles were laid together just so, the resulting pattern could be extended forever without repeating—a quasi-crystal. Lu examined “a few thousand” photos of real mosques and found that although decagonal girih patterns became increasingly common from 1200, nearly all are periodic and so are not quasi-crystals. But then he found a photo of the Darb-i Imam shrine in Isfahan, Iran, built in 1453. Its decagonally symmetric motifs on two different length scales are a telltale sign of a quasi-crystal. Working with Steinhardt, his former undergraduate adviser and a quasi-crystal expert, Lu found that the Darb-i Imam girih pattern can map onto a Penrose tiling. There were a few defects, but these are superficial, says Lu, and were likely introduced by workers during construction or repair. “We realized that by the 15th century, these architects had the makings of quasi-crystals,” says Lu. The paper has had a mixed reception. Crystal expert Emil Makovicky of the University of Copenhagen, Denmark, studied girih patterns for 2 decades. His analysis of the patterns on a tomb in Maragha, Iran, built in 1197, concluded that they map onto Penrose tiles and was published in a 1992 book about fivefold symmetry. Lu and Steinhardt cite his work, he says, but “without proper quoting and … in a way that [the ideas] look like their own.” Physicist Dov Levine of the Israel Institute of Technology in Haifa agrees that Makovicky deserves more credit than he is given in the paper. “His analysis of [the Maragha tomb] patterns anticipates some of the ideas in the Lu and Steinhardt paper,” he says. Joshua Socolar, a physicist at Duke University in Durham, North Carolina, agrees that Makovicky deserves credit for discovering “an interesting relation between the Maragha pattern and a Penrose tiling with a few defects.” Both Levine and Socolar doubt that the architects truly understood quasi-crystals but say Lu and Steinhardt have generated interesting and testable hypotheses. Lu and Steinhardt say they were aware of Makovicky's published work on the subject, but “we have found serious problems with both his technical reconstruction and general conclusions.” They say that they decided to limit their references to Makovicky “to avoid having to address the serious technical problems with his work.” Makovicky disagrees that his work is flawed. Beyond the question of credit, just how mathematically sophisticated these medieval architects really were remains open. “We haven't done an exhaustive search of Islamic architecture by any means,” says Lu. “There could be a perfect quasi-crystal pattern waiting to be found.” 5. ARCHAEOLOGY # Clovis Technology Flowered Briefly and Late, Dates Suggest 1. Charles C. Mann For almost 80 years, one of the most enduring puzzles in the archaeology of the Americas has been the “Clovis culture,” known for its elegant, distinctively shaped projectile points. Was Clovis the progenitor of all later Native American societies, as many researchers have long maintained, and, if so, how and when did it arrive in the Americas? On page 1122 of this week's issue, Michael R. Waters of Texas A&M University in College Station and Thomas W. Stafford Jr., proprietor of a private-sector laboratory in Lafayette, Colorado, use new radiocarbon data to argue that Clovis was a kind of brilliant flash in the pan—a movement that may have flourished across North America for as little as 2 centuries around 13,000 years ago. The new dates also put Clovis a bit later than thought, making it harder to accept that it was the first in the Americas. “What this paper does is reinforce how unusual was the phenomenon we call Clovis,” says Michael R. Bever of the University of Texas, Austin. “To have it rise and fall [throughout North America] in as little as 2 centuries” is a phenomenon with few equivalents in the archaeological record. Waters says that he and Stafford, an expert in the complex art of radiocarbon dating, set out “to nail down the most basic question: When was Clovis?” The heyday of the technology has typically been set between 11,500 and 10,900 radiocarbon years B.P. (The radiocarbon calibration is disputed for this period, but the widely used IntCal04 calibration puts the dates at 13,300 to 12,800 calendar years B.P.). In a controversial move, Waters and Stafford argue that no fewer than 11 of the 22 Clovis sites with radiocarbon dates are “problematic” and should be disregarded—including the type site in Clovis, New Mexico. They argue that the datable samples could have been contaminated by earlier material. Of the remaining 11 sites, Waters and Stafford found that five had been recently dated by higher-precision techniques. The pair decided to redate the others, succeeding in all but one case. The results, Waters says, “were a real surprise.” All of the new dates—as well as all of the previous acceptable dates—occurred within, at most, a 450-year band. Indeed, they say, Clovis probably existed for as little as 200 years, between 11,050 and 10,800 radiocarbon years B.P.—a cultural flowering both somewhat later and considerably shorter than thought. The later, more precise dates support the emerging view that Clovis was not the progenitor culture, because it overlaps or occurred after other cultures, including one in Monte Verde, Chile, dated to 1000 years before Clovis. The real surprise of the paper, according to David Meltzer of Southern Methodist University in Dallas, Texas, “is the compressed time frame for Clovis writ large.” So fast was its apparent spread that Stafford suggests that Clovis may have been a set of technologies that were picked up by a mosaic of different cultures across North America rather than a single, fast-moving society. “These tight dates, if they hold up, may help us resolve that long-standing debate,” says Meltzer, who questions the decision to discard the 11 sites. Meltzer stresses that the dates used are from a minority of North American sites, most in the west, whereas most Clovis points have been found in the east. Until more data are compiled, he says, researchers “can't know whether this is a real effect or simply a consequence of sampling.” In a sense, Stafford agrees. “We need to get more people out in the field,” he says. “We hope these dates motivate that.” 6. AAAS ANNUAL MEETING # Wedging Sustainability Into Public Consciousness 1. Robert Coontz* 1. With reporting by David Grimm, Eli Kintisch, Greg Miller, and Erik Stokstad. SAN FRANCISCO, CALIFORNIA—In a darkened ballroom in the Hilton San Francisco, 413 people tap numbers onto slate-gray keypads, each the size of a thick paperback book. Around them, almost 600 others watch as two screens at the front of the room reveal the results of their manipulations: a selection of strategies for taking wedge-shaped bites out of a graph of projected levels of atmospheric carbon over the next 50 years. Their mission: to whittle future CO2 levels down to a plateau in time to avert intolerable greenhouse warming. The “Wedge Game,” based on “stabilization wedges”—a concept developed by Robert Socolow and Stephen Pacala of Princeton University (Science, 13 August 2004, p. 968)—was part of a town hall-like session for teachers and students at the AAAS Annual Meeting, held here from 15 to 19 February. The game, designed to convey the scale of the effort needed to stabilize carbon emissions and the pros and cons of possible options, was just one of some 200 sessions, ranging from “Addiction and the Brain” to “Education, Learning, and Public Diplomacy in Virtual Worlds.” (For coverage of selected sessions, visit http://www.sciencenow.org/.) But one theme dominated the meeting: “Science and Technology for Sustainable Well-Being.” AAAS President John Holdren of Harvard University and the Woods Hole Research Center in Massachusetts set the stage with an opening address in which he warned of the dangers of complacently expecting technological fixes such as nuclear fusion to solve our problems. “I'm a great believer in science and technology, but the notion that science and technology will ride to the rescue is a pernicious one,” Holdren told reporters at a morning briefing before the talk. “Believing in technological miracles is usually a mistake.” Instead, he said, a huge effort on many fronts will be needed. Holdren urged scientists to “tithe” 10% of their time to working on four key challenges: global poverty, the competition for resources, the “energy-economy-environment dilemma,” and the threat from nuclear weapons. For its 4000 participants and 3000 visitors, including some 1000 reporters, the meeting offered a crash course in those challenges and how scientists are tackling them, from “big picture” strategies to technical nuts and bolts. Researchers monitoring the state of the planet reported warning signs from several quarters. Glaciologist Lonnie Thompson of Ohio State University in Columbus said ice cores from the Quelccaya Ice Cap in Peru—the largest body of ice in the world's tropics, 5670 meters above sea level—show that the ice is now melting faster than precipitation can replenish it. “All things being equal, those glaciers should be growing,” he said. Thompson, who has been studying such glaciers for decades, estimates that the 5000-year-old glacier could be gone within 5 years. Because temperatures at high altitudes are more stable than those below, he says, melting tropical mountain glaciers could be a “canary in the coal mine” for global climate change. Their loss could devastate the millions of people who depend on them for water. Meanwhile, in the Pacific Ocean, a research cruise from Tahiti to Alaska has shown that the upper 700 meters of the northeastern Pacific have increased their acid content by about 5% within the past 15 years. The change matches what computer models predicted would happen as more carbon dioxide from the atmosphere dissolves in seawater, said Richard Feely, an oceanographer with the U. S. National Oceanic and Atmospheric Administration in Seattle, Washington. Largely as a result, Feely calculates that the zone within which marine creatures can grow calcium carbonate shells is growing shallower by 1 to 5 meters per year. Even some of the supposed good news about climate change is looking less rosy. “You tell farmers in high latitudes they're going to get warming temperatures and longer growing seasons—end of story, they're happy,” plant ecologist David Wolfe of Cornell University said at one session. But recent outdoor field studies with carbon dioxide suggest that “yield benefits are about half what we thought they were,” he said. Also, hotter weather could damage milk production and crop yields. New work suggests that high levels of atmospheric CO2 emboldens weeds more than crops and could require farmers to double the amount of herbicide they use. Problems dominated news reports from the meeting, but more than three times as many sessions focused on the quest for solutions: economically competitive biofuels, better-managed water resources, and more efficient fish farms, fisheries, and livestock grazing. The tone ranged from matter-of-fact to unabashedly techno-optimistic. In a fast-paced pep talk in the run-up to the Wedge Game, for example, long-time alternative-energy advocate Amory Lovins of the Rocky Mountain Institute in Snowmass, Colorado, hymned the virtues of greener living through engineering. Ultralight low-drag cars, better-insulated houses, and decentralized low-carbon “micropower” energy sources, he predicted, would stabilize Earth's climate while reaping huge profits for businesses that seize the opportunities they present. “The low-hanging fruit is mushing up around our ankles,” Lovins said. Perhaps influenced by Lovins, the Wedge Gamers voted for a deep-green mix of two parts increased efficiency and one part each solar electricity, wind power, driving less, switching from petroleum to natural gas, and “biostorage” (planting forests to absorb CO2). It's far from current U.S. energy policy, but it reflects much of the thinking on display at many other sessions at this meeting. 7. ENVIRONMENTAL REGULATION # U.S. Courts Say Transgenic Crops Need Tighter Scrutiny 1. Dan Charles* 1. Dan Charles is a Washington, D.C.-based science writer. Citing a broad range of risks, U.S. federal judges in three separate cases have asked the U.S. Department of Agriculture (USDA) to examine genetically engineered crops more closely. The courts said the department had violated the National Environmental Protection Act (NEPA) in approving commercial sales of transgenic alfalfa and field trials of turf grass and plants engineered to produce pharmaceuticals. Critics of genetically engineered crops say the decisions, two issued this month and one last August, will compel tighter regulation of transgenic crops. Will Rostov, an attorney for the Center for Food Safety in Washington, D.C., which filed all three lawsuits, called the alfalfa decision, rendered 12 February by U.S. District Judge Charles Breyer in San Francisco, California, “another nail in the coffin for USDA's hands-off approach to regulation.” But Stanley Abramson, a lawyer who represents several biotech companies, pointed out that the courts raised questions about USDA's procedures, not its substantive decisions. He predicted that USDA's final judgments would hold up in court. The alfalfa verdict could have the most significant impact. In 2005, USDA approved the sale of Roundup Ready alfalfa, jointly developed by Monsanto and Forage Genetics International, which can withstand the popular herbicide glyphosate. But last week, Breyer said that the department should have first prepared an environmental impact statement (EIS) as required under NEPA. Joseph Mendelson of the Center for Food Safety said that his group may demand an end to sales of genetically engineered alfalfa or even a ban on planting transgenic seed already in farmers' hands. USDA officials declined to discuss the government's position or whether it plans to appeal. A spokesperson for Monsanto, which sells genetically engineered alfalfa but was not a party to the lawsuit, said he did not expect sales to be halted. Breyer gave both sides until next week to propose regulatory fixes. The second verdict, handed down 5 February by a Washington, D.C., district judge, found that USDA should have carried out an EIS or a more modest environmental assessment before it allowed a 162-hectare field trial of transgenic turf grass near Madras, Oregon, in 2003. And last August, a federal court in Hawaii faulted USDA for approving field trials in Hawaii of corn and sugar cane engineered to produce experimental pharmaceuticals without considering the state's numerous endangered species. In two of the cases, the judges expressed concerns about potential risks that USDA has dismissed as insignificant or outside its mandate. Breyer, for instance, complained that USDA ignored the cumulative impact of glyphosate-tolerant alfalfa, corn, and soybeans. Greater use of glyphosate increases the odds that weeds will develop resistance to it. Breyer also said USDA erred when it dismissed as not “significant” the concerns of organic farmers who don't want Roundup Ready pollen or seeds spreading to their alfalfa fields. The possible replacement of traditional varieties is itself significant, he noted. “An action which eliminates or … greatly reduces the availability of a particular plant—here, nonengineered alfalfa—has a significant effect on the human environment,” he wrote. USDA argued that cross-pollination wasn't a serious problem in alfalfa, because farmers typically harvest their fields before the plants have a chance to flower, much less produce seeds. Producers of commercial alfalfa seed, however, would have to make sure their conventional and transgenic fields were widely separated. Alfalfa is pollinated by bees, which can carry pollen at least 3 kilometers. In the turf grass case, Judge Henry Kennedy found that transgenic bentgrass from a large field trial in Oregon threatened a nearby area's “aesthetic and recreational” value. Pollen from the bentgrass spread up to 20 kilometers into the nearby Crooked River National Grassland. Many scientists, including some critics of genetically engineered crops, say the bentgrass poses no real ecological threat in that area because it isn't well adapted to the region's arid climate. But the spread of this “confined” field trial proved embarrassing to the Scotts Co., which hopes eventually to sell bentgrass seed to golf courses. 8. BIODIVERSITY # Reconstructing Brazil's Atlantic Rainforest 1. Bernice Wuethrich* 1. Bernice Wuethrich usually writes from Maryland. Marcelo Rideg provided translations in São Paulo. Some reforestation projects results in little more than tree plantations. An ambitious project in Brazil's São Paulo state is trying to go further and create a real working ecosystem CUNHA, BRAZIL—Benedito de Carvalho Filho strides across his yard, through an empty cow pen and uphill. He clambers over rocks and searches, pushing aside shreds of barbed wire and vegetation, as his 12-year-old son shadows him. Finally, the farmer stops and points to the eye of a spring that barely percolates up from the earth. His water—or what's left of it. Throughout this pastoral region that was once the heart of Brazil's Atlantic rainforest, extensive deforestation has not only changed land cover, it has also altered the hydrologic cycle. The forest once stretched over 1 million square kilometers along Brazil's coastal region with extensions inland. Today, only 7% of that original extent persists. And now, de Carvalho and other farmers are grappling with a growing realization: The fresh and abundant ground water they have relied on for decades is disappearing. For his family and his crops to have water, de Carvalho needs the rainforest back. He may be in luck. With little fanfare, this past December, biologists and farmers planted the first seedlings of an ambitious project that ultimately aims to reforest 1 million hectares (2.47 million acres) of riparian rainforest across São Paulo, Brazil's most populous state. The goal is not just to recreate this globally unique ecosystem but also to reclaim the so-called services the rainforest once provided, from the maintenance of natural springs and soil fertility to the sequestration of carbon. It's a tall order. Funded with seed money from the state and the Global Environment Facility (GEF), the Riparian Forest Restoration Project (RRP) will require an estimated 2 billion seedlings of hundreds of species of trees and take decades to complete. The total bill could top$2 billion—about $2000 per hectare—just a fraction of which has been raised. “It is hard to think of a reforestation project undertaken anywhere that is quite this ambitious,” says Thomas Lovejoy, formerly the World Bank's chief biodiversity advisor. Even if the RRP achieves only some of its ambitious goals, say conservation scientists, it appears set to test the limits of nature's resilience, the science of ecological restoration, and societal commitment. ## Dimensions of diversity The RRP will focus on restoring forest along the denuded margins of rivers and streams that establish migratory corridors for animals and plants and protect the ecological health of waterways. The eventual goal is to recover a significant portion of the staggering diversity of the rainforest's flora and fauna—its bromeliads, lianas, shrubs, grasses, birds, bats, butterflies, insects, microbes, mammals, and amphibians. How exactly to do that is being tested in pilot projects in five watersheds across São Paulo state, with$19 million in start-up funds provided primarily by the state, GEF, and the World Bank. Each of the five watersheds is in turn divided into three “microwatersheds” that represent a range of ecological as well as social challenges. In Cunha, for instance, located within the larger Paraíba do Sul watershed, subsistence farmers work within a mosaic of forest fragments. It is a sharp contrast to the large northwest watershed, Aguapeí, where sugar cane is king and any remnant forest is hard to find.

In these pilot projects, the state Department of the Environment and its partners are experimenting with a variety of restoration methods. Some emphasize replanting trees alone; others aim to return a variety of plants and animals simultaneously. “We want to see how the forest can best develop on its own after a starting push,” says project director Helena Carrascosa von Glehn.

At each site, the selection of trees and methods will depend on local conditions and the specific goals for the area, whether that be soil stabilization, water maintenance, or production of fruit and nuts from rainforest trees. “Our emphasis is on bringing people to this for 20 to 30 years and involving them in restoration,” Carrascosa says.

## Search for Mother Trees

At the heart of the effort is the Mother Tree project: a detailed plan for identifying the starting stock for the forest's regeneration. Conceived by Ricardo Rodrigues at the Laboratorio de Ecologia e Restauração Florestal in Piracicaba, Brazil, the Mother Tree project aims to find and mark the location of 15,000 trees of 800 different species to support seed collection programs and ensure adequate genetic diversity in the replanted forests. There is now a master list of 780 species appropriate for reforestation in designated watersheds.

To find the seeds, biologists are scouring remnants of the Atlantic forest across the state. So far, says Rodrigues, 20 biologists have recorded 10,200 mother trees.

One of those biologists is forest engineer Renato Lorza, who conducts his search for mother trees in the mountainous terrain of Cunha on foot, aided by a long-handled scissor. On a morning last December, a fine mist laces the air as he scoops up a handful of gleaming seeds from the forest floor. Most plentiful are the soft, marble-sized seeds of the Jussara palm (Euterpe edulis). Severely overharvested to extract heart of palm, the tree grows up to 25 meters to reach the light through the forest's dense canopy. The palm is but one of some 1300 tree species in the Serra do Mar State Park in Cunha that Lorza has come to know well during 2 years of collecting.

University researchers analyze and sometimes genetically sequence the material he collects. After the species is definitively identified, Lorza returns to nail a small metal label into the trunk of each mother tree. He has tagged 750 individual trees, representing 250 species in this remnant forest. Each one must be in a reproductive stage of life, but some species flower only every 20 years.

One giant, Number 2496, is Sapium glandulatum, a bird-dispersed pioneer tree. Number 2497 is a climax species, Aspidosperma parvifolium. Slow growing, the tree is a rare find, as it has been overharvested for use in construction. Both pioneer and climax species are essential to the reforestation effort. Fast-growing pioneer species are the first to establish in gaps that form when trees fall in the forest. They germinate and grow in light. The secondary species that follow germinate in light but grow in shadow; slow-growing and long-lived climax species germinate, establish, and grow in the shadows.

A state law—the most far-reaching of its kind in Brazil—mandates that each reforested hectare include a minimum of 80 tree species, with each represented by at least 12 mother trees from distinct populations to ensure good genetic variability. The law also mandates that the 80 trees reflect the different successional stages of the forest, which means a mix of both pioneer and nonpioneer species, in a ratio that is close to 1:1.

This basic approach grows out of the work of Paulo Kageyama, director of biodiversity conservation at the Brazilian Ministry of the Environment, who spent decades planting experimental forests for the hydroelectric industry to restore biodiversity around dams. Beginning in 1988, he began growing diverse plantations that included species representative of each major successional stage.

One such plantation, the state's Paraibuna forest, has stood as an emblem of success. Its 100 different tree species have thrived, and now that “other kinds of plants are arriving by wind or birds, the biodiversity is increasing,” Kageyama says.

It took well over a decade, however, for the Paraibuna forest to cross that threshold and begin recruiting the other species needed to recycle nutrients, disperse seeds, and pollinate plants. It is now a functioning forest rather than a plantation that must be maintained, says Kageyama.

But many other reforested areas in Brazil have failed to make that transition. Of some 98 publicly funded reforested areas evaluated in 2000, only two did well, according to Luiz Mauro Barbosa, director of the São Paulo Botanical Institute and a co-director of the RRP. In most areas, after initially flourishing, the trees died, and weeds took over.

Barbosa says that diversity is a key ingredient of success: Forests with 30 tree species generally did better than those with only three or four. The premise is that if a sufficient and balanced diversity of trees is planted with species appropriate to the local conditions, the flourishing forest will recruit the other flora and fauna—essentially, build it and they will come.

## Trees in context

Yet species diversity in and of itself has not always been sufficient. For instance, one reforestation effort in western São Paulo state began with 42 tree species, but after 10 years, just four pioneer species dominated the upper canopy, says Daniel Piotto, a forest engineer who has worked on industry-financed reforestation projects in São Paulo and is now a Ph.D. candidate at Yale University. Piotto thinks that the difficulty may lie in the limited availability of factors such as water, mineral nutrients, and suitable microclimate that put slow-growing climax species at a competitive disadvantage relative to faster growing pioneer tree species and grasses.

In addition, some biologists believe that the rainforest restoration will require more than a narrow focus on trees. Ademir Reis, now at the University of Santa Catarina, and Fernando Bechara, now at the Casa da Floresta Assessoria Ambiental in Piracicaba, have developed an experimental technique known as “nucleation.”

It aims to jump-start the establishment of ecological relationships essential to plant and animal life through creating a wide variety of niches. In test plots, the biologists have combined a number of techniques: They erected perches to attract birds and bats, built shelters for small mammals, and planted herbs and shrubs (shunned as competitors to trees in traditional reforestation) to attract butterflies. They also transposed squares of topsoil from intact forest, delivering soil microbes, earthworms, and fungi. After just 1 year, Bechara has noted the return of 35 species of birds to one experimental area.

“Forest restoration is extremely expensive and labor-intensive. Anything that can mimic natural succession in depleted areas is worthwhile,” says Gustavo Fonseca, chief conservation and science officer with Conservation International.

## Building support

Equally important to testing restoration methods is ensuring the support of the local community: “We can do nothing without the involvement and support of the people,” Lorza says. The RRP aims to build this support by convincing people that devoting some portion of their land to rainforest restoration would benefit them in the long run, even if it means giving up some hectares used for farming or pasture.

In Cunha, that means educating the community about the role of the rainforest in maintaining their natural springs, which have slowly disappeared over the past 40 years. Honario Eliane, a farmer who has lived in Cunha all of his 71 years, is convinced. “About two-thirds of the natural springs that once irrigated my land are gone. The river I used to swim in when I was a kid now barely covers my feet. We need this,” he says. Honario's daughter is one of 19 teenage “environmental monitors” who work with a local nonprofit group known as Living in the Atlantic Forest, one of many partners in the state's reforestation effort.

Today, after some initial opposition, about 60 farmers in Cunha want to reforest some of their land, says Leila Pires, who directs the RRP in Cunha. Due to limited resources, reforestation has begun on just six farms.

To speed the project along and ensure its financial sustainability, São Paulo state is exploring the establishment of a fund for ecosystem services, says Carrascosa. The state has already implemented new environmental legislation in two watersheds, charging industry, large-scale farmers, and the government for water use. Other possibilities include GEF payment for carbon sequestration achieved through reforestation and the establishment of a private market for rainforest seedlings, Carrascosa says. She suggests that the RRP could prove to be a model for both the Amazon and the nation.

9. U.S. DEPARTMENT OF AGRICULTURE

# Farm Bill May Contain Seeds for More Robust Research

The Bush Administration wants to trim farm subsidies, restructure USDA, and do more research on biofuels, fruits, and vegetables

For the past decade, advocates for U.S. agricultural research have watched with envy as the federal government has pumped money into other disciplines, notably medical research. Despite daunting environmental problems and growing international competition for producing cheap food, the research budget of the U.S. Department of Agriculture (USDA) has been stuck at roughly $2 billion. But advocates say that a new proposal by the Bush Administration to revamp agricultural subsidies provides an ideal cue for pitching research to a Democratic-controlled Congress that intends to pass some sort of legislation this year. Congress periodically adjusts the nation's agricultural policies in a massive farm bill, and the Administration's latest proposal, submitted 31 January, does more than tinker: It would cut$10 billion over 10 years from subsidies paid to farmers. The president's plan also proposes a restructuring of USDA by merging two research agencies and creating an Office of Science. And it would add $100 million for research on specialty crops—fruits, vegetables, tree nuts, and nursery plants, for example—and$50 million for biofuels.

Ag lobbyists say the move is a good start, although it falls short of the doubling of basic research that a major panel recommended several years ago. “USDA should be in a much stronger leadership role” if the bill is approved, says Jeffrey Armstrong, dean of the College of Agriculture and Natural Resources at Michigan State University in East Lansing. “To integrate and create an Office of Science—it's huge.” And although the science reorganization is relatively uncontroversial, the bid for additional spending could run afoul of a push by growers to protect their subsidies as appropriators try to tighten belts.

USDA's research is divided into three main parts: the $1.3 billion intramural Agricultural Research Service (ARS); the Cooperative State Research, Education, and Extension Service (CSREES), which gives out$180 million in competitive grants through the National Research Initiative and $180 million in so-called formula research grants to agricultural universities; and the$340 million research effort at the U.S. Forest Service. Each program has its own staff, and there's considerable duplication in areas such as plant genetics, soil science, and air quality. “An integrated approach would be more effective,” says Fred Cholick of Kansas State University in Manhattan.

Change has been in the air for several years. In 2004, a blue-ribbon panel called for an independent research agency (Science, 10 December 2004, p. 1879). But USDA disliked the idea, and a bill to implement the findings died. Last year, the National Association of State Universities and Land-Grant Colleges (NASULGC) proposed a suite of reforms, called CREATE-21, that included unifying USDA research efforts.

The Administration's bill, expected out in the next few weeks, would merge ARS and CSREES and give a chief scientist control over both intramural and extramural programs. “We'll be better able to plan and coordinate comprehensive programs like food safety,” says Lowell Randel of USDA's research policy shop, as well as pursue pressing topics such as obesity. Forest Service research would remain separate, Randel says, because it focuses on forest issues. (It's also overseen by a different congressional committee, he notes.)

The farm bill would nearly double research on specialty crops and biofuels. USDA spent $44 million on biofuels research in fiscal year 2006 and somewhere between$120 million and $150 million on specialty crops. The new initiatives, totaling$150 million a year for 10 years, would be financed by the government's Commodity Credit Corporation, which places them outside the annual appropriations cycle.

Even so, appropriators could divert some of the money, as they did repeatedly with the \$200 million Initiative for Future Agriculture and Food Systems (IFAFS), a competitive grants program set up in 1998 (Science, 21 January 2000, p. 402). IFAFS also garnered little support outside the academic community, as growers which didn't see its value for specific crops or agricultural problems. Randel says that focusing on specialty crops and biofuels is a better way to “show that our science is world class and [that it is] solving problems that matter to all Americans.”

Research advocates hope Randel is right. But they would like the bill to go further. “I don't have any quarrel with the targeted areas, but there are other areas that are equally critical,” says Ian Maw, vice president for food, agriculture, and natural resources at NASULGC, which is asking for the USDA research budget to double over 7 years. The group also wants competitive funding, now 10% of the department's total research budget, to grow to 58%.

Research advocates admit that farmers' cries for maintaining subsidies could drown out their call for a bigger research pot. But they argue that international pressure to eliminate farm subsidies may work in their favor. In 2005, the World Trade Organization (WTO) ruled that USDA cotton subsidies were illegal because they undermined fair trade. But getting Congress to listen is another matter. The chairs of the two committees that will write the bill have said that they don't feel bound by WTO rules. Their goal is to complete work on it by September.

10. NICOLA CLAYTON PROFILE

# Nicky and the Jays

1. Virginia Morell*
1. Virginia Morell is a writer in Ashland, Oregon.

Experimental psychologist Nicola Clayton and her colleagues are proving that birds are more intelligent than most of us give them credit for

“I'm very birdlike, don't you think?” chirps Nicola Clayton. In fact, the petite Clayton does indeed remind one of a bird: She stands just 1.62 meters tall even in her stilettos and has a slightly beaked nose, a plume of blond hair, flashy clothes, and a hummingbird's quick, darting moves. So apparent is the similarity that her students at the University of Cambridge, U.K., have dubbed this 44-year-old experimental psychologist a new bird species, Claytonia professorii.

But Clayton's “birdiness” is not just a matter of style. She's a keen student of bird behavior, drawing on natural history to investigate experimentally a trait that most scientists did not believe existed in birds: cognition. Over the past 9 years, she and her colleagues have pumped out 75 papers on the surprisingly humanlike mental skills of food-storing birds, particularly the western scrub jay (Aphelocoma coerulescens). In this week's issue of Nature, Clayton's lab chips away at another “human-only” cognitive ability by showing that scrub jays can plan for the future.

“They've devised a clever experiment for these jays which does fulfill the [psychological] criteria for future planning—which makes this a first,” says Sara Shettleworth, a comparative psychologist at the University of Toronto in Canada. Other papers have argued that monkeys, bonobos, and orangutans plan ahead (Science, 19 May 2006, pp. 1006, 1038), but because the animals were given numerous trials, they might have been “trained” to act as though they foresaw the future, Shettleworth points out.

Although it might seem that some animals prepare for the future, for instance by moving to a cave for hibernating, researchers consider these behaviors to be merely mechanistic responses to a seasonal cue. In contrast, Clayton's graduate students Caroline Raby and Dean Alexis placed the scrub jays in a novel situation. The birds first spent the night in a suite with two adjoining rooms. In the morning, the researchers moved the jays to one of the end rooms. One end room was like a bed and breakfast, and the jays received breakfast pine nuts; the other room was more like a cheap motel, and the jays went hungry.

After a few days, the jays were allowed access to either end room and provided with pine nuts in the main compartment. Spontaneously, the jays stashed the nuts in the breakfastless room, hiding them in a sand-filled ice cube tray (the jay's equivalent of a refrigerator). They seemed to recall going hungry during their past stay in the motel room and, based on that experience, saved the nuts for a future breakfastless morning.

“Who would have thought you could ask a bird what it understands about the future?” asks Joan Silk, a primatologist at the University of California (UC), Los Angeles. “It's incredibly difficult to design an experiment that would demonstrate this. But they have. It adds to a growing body of work that shows that jays, crows, and rooks can do anything primates can do, especially things that we once thought were important links to understanding the emergence of cognition in humans.”

## For the birds

Clayton's talent for designing such ingenious experiments comes, she says, from understanding what scrub jays do in the wild. During the early 1990s, in John Krebs's University of Oxford lab, Clayton worked on memory development in food-caching Eurasian jays and tits, species known to have excellent memories. Her work there showed that the hippocampus region of these birds' brains responded like a muscle to their food-hiding behaviors, growing markedly in size as they stored seeds in numerous locations.

When she moved to UC Davis in 1995, she found that the western scrub jays—gray-and-blue birds with a raucous squawk—were common, making it easy for her to observe them in their natural setting. “One thing I noticed almost immediately was that the scrub jays would steal bits and pieces from people's lunches and hide these,” she says. “Later on, they'd return to these caches and move them. But why?”

She decided to use this food-stashing behavior as a starting point for her experiments, using what she knew about the jays' wild activities to reveal their psychology. For these tests, Clayton provides them with sand-filled ice-cube trays. Previous studies have shown that in the wild, food-stashing birds use nearby landmarks to remember their stashes. To give the lab jays similar cues, Clayton attaches a colorful stack of Lego pieces to each cube of the tray. She offers the birds nuts and fresh wax worms and films the birds' behaviors.

She had observed that wild scrub jays stole one another's caches, if they could, and also stored worms and bugs, which decay relatively quickly, as well as longer-lasting seeds and nuts. “They not only had to remember where they buried things but when they did,” she points out. That observation led to a controversial Nature paper in 1998.

With her co-author, University of Cambridge comparative psychologist Anthony Dickinson, she asserted that scrub jays have “episodic-like memory,” the ability to remember specific things they did in the past, a skill akin to what humans do. “People seemed to think that if a monkey had done it, that would be one thing,” Clayton says, “but a bird?”

Dickinson says he would never have considered making such a claim prior to meeting Clayton. Indeed, Dickinson, a learning theorist, confesses that when they met at a 1996 conference, he was initially dismissive of Clayton's research, saying there was absolutely no reason an animal needed to have personal, episodic memories. “It was an outrageous statement,” recalls Clayton, who countered that the birds had to remember when they buried something, particularly fresh food, otherwise they'd be wasting time and energy flying back and forth constantly to make sure the food was still edible. “That wouldn't be at all adaptive,” says Clayton. Intrigued, Dickinson soon found himself collaborating with Clayton on experiments designed to investigate whether the birds were making cognitive decisions or behaving mechanistically.

“I wouldn't have thought of using a natural behavior—such as the food-caching—to ask a psychological question, prior to meeting Nicky,” says Dickinson. “But that's her strength: She's a biologist who's capable of thinking about psychological processes.”

Clayton met her other main collaborator (and husband), Nathan Emery, at about the same time. He was then studying primates at UC Davis and delighted in boasting to Clayton about his monkeys' many talents. “He made so many ‘ape-ist’ remarks, about how they could do things that no other animal could do, and I'd say, ‘But that's not true!’ and tell him all about my jays,” Clayton recalls. To his credit, she adds, Emery immediately grew interested in the jays, and soon he and Clayton were also collaborating on papers, adding his primatologist's credentials to their claims that jays and their relatives are as smart as apes. They moved to England in 2000, bringing the jays with them.

## Pilferers and physicists

In her lab at Cambridge, Clayton leads the way into a room with large wire enclosures lining the walls. Inside, the scrub jays squawk and leap from the sides of the cages to their perches and back, watching her every move. “I find it so curious, the way they look you right in the eye,” says Clayton, who hand-raised most of her charges. “I think it's because they recognize individuals—that's not a scientific finding, just a gut feel. But I think it's another factor in their pilfering; they know each other as individuals.”

She, Emery, and postdoc Joanna Dally have shown that the jays are well aware of each other's thieving natures and, once they themselves have stolen the cache of a fellow jay, learn to watch out for thieves. “They might know something about what the other bird is thinking; they seem to take his point of view,” says Clayton. “They keep track of who was watching when” (Science, 16 June 2006, p. 1662).

Many researchers think that such an ability, called “theory of mind,” is something that requires language and therefore can only be done by humans. But Clayton disagrees. “I think [thievery] is one reason they've become so cognitive: When they stash their nuts, they have a lot to worry about—they have to think about it from their own perspective, as well as the thief's.”

Other comparative psychologists are not persuaded, however. “It's a giant controversy,” says Shettleworth, explaining that a jay stashing a nut may simply be mentally recording the presence of another bird near its cache; the jay returns to hide its food elsewhere when the other bird is gone. “The hoarder doesn't need to know what the other guy is thinking to act this way.”

If jays are thieves, then it seems that rooks (Corvus frugilegus)—raven-sized birds with dark plumage—are physicists. Or, at the very least, the birds' ability to manipulate things suggests they have some basic understanding of the properties of objects and gravity, even though rooks don't use tools—and therefore don't need this knowledge—in the wild.

Clayton, Emery, and graduate student Amanda Seed reported in the April 2006 issue of Current Biology that rooks easily mastered the simple physics required to retrieve a piece of meat from inside a tube, using a twig to push or pull as necessary until the food fell to the ground. “You can see them assessing the tube and the stick and the bit of food” to arrive at a solution, says Clayton. This behavior again suggests that these birds are quite cognitive creatures.

Clayton, Emery, and University of Cambridge postdoc Lucie Salwiczek plan to expand their lab's investigation of bird intelligence this year by testing young jays newly arrived from California for their understanding of object permanence: that is, their ability to know that something is still there even if they can't see it. The adult birds must have this ability, because after burying a nut, they return to retrieve it. But Clayton and her colleagues want to know at what age they develop this sense.

Despite her intense curiosity about these birds' brains, Clayton says she will stay away from neuroanatomical studies, as she is loath to sacrifice any jays. “I'm too attached,” she says. That may limit her progress, as that's the logical next step, says neurobiologist Erich Jarvis of Duke University in Durham, North Carolina, who compares bird and mammalian brains. Yet even without taking this step, Clayton is changing the way neurobiologists think, he notes. “There's a general belief that the further away a group is from humans, the less intelligent it will be. So people are surprised—even scientists are surprised—that some birds can do many of these things.” For Claytonia professorii, that's something to crow about.