News this Week

Science  20 Feb 2009:
Vol. 323, Issue 5917, pp. 992
1. U.S. BUDGET

Science Wins $21 Billion Boost as Stimulus Package Becomes Law 1. Eli Kintisch* 1. With reporting by Dan Charles, Lila Guterman, Andrew Lawler, and Jeffrey Mervis. The scene: a jittery Washington, D.C., amid the worst global economic crisis in decades. The main character: a young president who has promised to “restore science to its rightful place.” The supporting cast: a Congress dominated by Democrats heeding President Barack Obama's call not only to “create new jobs but to lay a new foundation for growth.” This week, those elements culminated in a historic$787 billion economic stimulus package that provides more than $21 billion for research and scientific infrastructure. Having a role in the biggest one-time federal outlay since the New Deal has been a surprise for scientists, on whom the Obama plan lavished a variety of cash investments. These include$10 billion for the National Institutes of Health (NIH), $3 billion for the National Science Foundation (NSF), and$1.6 billion for the Department of Energy's (DOE's) Office of Science.

Scientists are euphoric. “It's incredible; more than we've seen in a long time,” said Francis DiSalvo, a materials chemist at Cornell University. The infusion comes after a series of annual appropriations bills that doled out relatively flat budgets for many federal science agencies. Some research policy experts are wary of the audacious lump-sum approach. “It's an incredibly rapid infusion of an unprecedented amount of money for scientific research to be spent in an extremely short amount of time,” said Daniel Sarewitz of Arizona State University, Tempe.

U.S. science lobbyists first realized that they might be in for a piece of the stimulus pie soon after Obama was elected. As talk spread of combating the deteriorating economy with aggressive spending and tax breaks, the initial focus was on scientific construction projects, says Michael Lubell of the American Physical Society. But Obama's transition team and congressional staffers soon broadened their requests to include proposed investment in research, arguing that putting people to work on innovative ideas could generate even more well-paying jobs down the road (Science, 16 January, p. 318).

House Speaker Nancy Pelosi (D-CA) led the way by declaring that “science, science, science” was central to the Democratic agenda. In turn, the House version of the stimulus bill included the large boosts for DOE, NSF, and NASA that ended up in the final bill. Those increases mirrored levels Congress proposed as part of the 2007 America COMPETES Act, which authorizes spending that has never materialized.

Senate Majority Leader Harry Reid (D-NV) had a tougher political job. Two votes shy of a filibuster-proof majority, he had to woo a few Republicans, who generally wanted less spending and more tax cuts. Maine's two moderate Republicans, senators Susan Collins and Olympia Snowe, teamed up with Senator Ben Nelson (D-NE) to craft a compromise bill that lopped off $100 billion from what was then before the Senate. The science community was aghast when the first version of the compromise proposed dropping NSF from the package entirely. They enlisted heavyweights from academia and industry to make the case for the agency's role as an economic driver. At the same time Senator Arlen Specter (R-PA), a longtime supporter of biomedical research, successfully added$6.5 billion to the $3.5 billion that NIH had been allocated and made it clear that his support for the overall package was contingent on retaining a boost of that magnitude. “The cost-of-living adjustments have not been made,” Specter said during floor debate on the bill. “There has been an actual decline of some$5.2 billion of NIH funding in the last 7 years.”

The compromise Senate bill still provided generally lower levels for most science agencies, however, and lobbyists credit Pelosi as well as the White House for restoring the House levels in a conference between the two bodies. “She deserves great credit, as does the Obama Administration,” says Toby Smith of the Association of American Universities in Washington, D.C.

In the end, NIH received the biggest bounty. The $30-billion-a-year agency receives an additional$8.2 billion for research, spread out over the current and next fiscal years, as well as $500 million for construction and renovation of NIH's intramural labs and$1.3 billion for research facilities and instrumentation elsewhere. NASA will get $1 billion on top of its yearly$17 billion budget, with lawmakers targeting Democratic favorites such as supercomputing and earth science ($400 million), aeronautics ($150 million), and a new launcher ($400 million). DOE's$4 billion Office of Science receives an extra $1.6 billion, with no strings attached. The need is great, say scientists. “Most buildings here are very old,” says physicist Michael Norman of DOE's Argonne National Laboratory in Illinois, which is hoping for a new interdisciplinary center for energy research. Congress also added another$400 million for a new entity called the Advanced Research Projects Agency for Energy. The idea is to emulate the long-standing Defense Advanced Research Projects Agency (DARPA), credited with giving birth to the Internet and other important commercial technologies. Jane “Xan” Alexander, former deputy director of DARPA, says the new agency will help technologies avoid the socalled Valley of Death that separates basic and applied research at DOE.

But it's arguably NSF that will see the biggest impact from the stimulus bill. Steven Beering, chair of its oversight body, the National Science Board, says the $3 billion spending boost for the$6-billion-a-year agency “is phenomenal.… It'll enfranchise many people who wouldn't otherwise have been funded and allow NSF to fund more high-risk, transformational research.” Some $2 billion will be spent on research grants and$100 million for education programs. It also contains $900 million for various infrastructure projects, including$200 million to revive an academic facilities program. The board will take up how the influx affects NSF's priorities at its meeting next week, Beering adds.

Although Sarewitz says that funding science may be a worthy cause, he sees potential harm from its inclusion in the stimulus package. “If this spending is for [economic] stimulus, then it's not at all clear that R&D is a good way to get money into the economy quickly,” he says. “If, on the other hand, it's long-term investment, then there's no reason to ram it into the system so quickly.”

But such doubts are hard to find among a science establishment grown accustomed to flat budgets. Richard Marchase, president of the Federation of American Societies for Experimental Biology, admits to some worries of a poststimulus letdown. But for right now, he says, “we're very, very appreciative.”

2. PARTICLE PHYSICS

LHC Delays Give Tevatron a Shot at Higgs Boson

One lab's setback can be another lab's opportunity. Last week, officials at the European particle physics laboratory, CERN, near Geneva, Switzerland, announced that it will take longer than previously estimated to fix the world's highest energy particle smasher, the Large Hadron Collider (LHC), which suffered a catastrophic malfunction last year before ever taking data (Science, 26 September 2008, p. 1753). The LHC will not start up again before late September, 2 months later than previously announced. Meanwhile, the older Tevatron collider at Fermi National Accelerator Laboratory (Fermilab) in Batavia, Illinois, may be gaining the edge in the race to spot the Higgs boson, the last missing piece in the standard model of fundamental particles.

“Three years ago, nobody would have bet a lot that the Tevatron would be competitive [with the LHC] in the Higgs search. Now I think the tables are almost turned,” says Tommaso Dorigo, a physicist from the University of Padua in Italy who works with the CDF particle detector fed by the Tevatron and the CMS particle detector fed by the LHC.

Given the continuing delays with the LHC, Fermilab physicists are hoping for the chance to continue running their 26-year-old collider through 2011. Officials at the Department of Energy, which owns Fermilab, declined to comment because they are still formulating their budget for the next fiscal year, which starts on 1 October. But they would likely have to make that decision before the LHC comes back on.

Fermilab scientists say it's only prudent to keep the Tevatron running. “We don't think that a running accelerator complex should be shut off until it's displaced by another running accelerator complex that's producing physics,” says Fermilab's Rob Roser, co-spokesperson for the 602-member CDF team. But some say the call should not be made on the assumption that the LHC will have more problems. “You should look at your own program and assume the other guy will succeed,” says Sheldon Stone of Syracuse University in New York state.

The numbers show that the Tevatron is performing superbly. Physicists measure the amount of data produced in units of “inverse femtobarns.” From the beginning of the current run in 2001 until the end of September, the Tevatron produced 5 inverse femtobarns. It is now on pace to nearly double that amount by the end of 2010, and it could rack up 12 inverse femtobarns by the end of 2011.

That should be enough data to give researchers with CDF and the neighboring D0 particle detector a chance to see evidence of the Higgs boson—if it's there. “Evidence” is jargon for a signal that's stronger than a certain level but not strong enough to claim a definite discovery. The probability of seeing such a signal depends on what the Higgs might weigh. Researchers with CDF estimate that if the Higgs weighs between about 120 and 195 times as much as a proton—that's 114 and 182 giga-electron volts (GeV) in the units physicists prefer—then CDF and D0 working together should have at least a 25% chance of seeing it if the Tevatron runs through 2011 (see figure).

Scientists at the Tevatron seem to like their chances. More people are sticking around than a 2004 survey had projected, says Fermilab's Dmitri Denisov, co-spokesperson for the 530-member D0 collaboration. “We expected that we would lose between 10% and 15% of our people per year, but between a year ago and now, we're down about 5%,” Denisov says. CDF's losses are also lighter than expected, Roser says.

Ultimately, whether the Tevatron runs in 2010 or 2011 depends on Fermilab's budget. In preparing for life after the Tevatron, Fermilab is planning to build a proton accelerator to power neutrino beams and other experiments (Science, 31 August 2007, p. 1155). If there isn't enough money, says Fermilab Director Pier Oddone, “then we will stop the Tevatron to get the resources to develop the future of the lab.”

Across the Atlantic, CERN officials have laid out a timeline for repairing the LHC and making modifications to prevent a repeat of the 19 September failure. The problem began when a splice in a superconducting electrical line melted, triggering a leak of boiling liquid helium and a concomitant pressure wave that damaged 53 of the machine's more than 1600 superconducting magnets. Workers need an additional 6 weeks to make those fixes, says CERN spokesperson James Gillies. To make up for lost time, Gillies says, CERN will run the LHC through next winter despite the higher cost of electricity, which normally dictates a halt in operations.

Still, CERN officials can't guarantee that the LHC won't run into some new sort of problem that will cause further delays. Meanwhile, Fermilab researchers see a more predictable path to their shot at glory.

No News Is Good News for Holdren, Lubchenco at Confirmation Hearing

1. Jeffrey Mervis

Senate confirmation hearings generate news only if a presidential nominee unintentionally gets ahead of the boss in announcing some new policy or falls into a trap set by a legislator from the opposing party. But sometimes, if the committee is genuinely interested in the business of governing, the hearing can give consensus nominees a chance to discuss issues at the agencies they've been asked to lead even without making news.

That's what happened last week when John Holdren and Jane Lubchenco appeared before the Senate Committee on Commerce, Science, and Transportation. President Barack Obama has nominated Holdren, a physicist, to lead the White House Office of Science and Technology Policy and Lubchenco, a marine ecologist, to head the National Oceanic and Atmospheric Administration (NOAA).

The committee was more than gracious: “My hope is to move your nominations [through the Senate] as quickly as possible,” chair Senator Jay Rockefeller (D-WV) explained at the start and end of the 2-hour hearing. Senator David Vitter (R-LA) was the only committee member who challenged Holdren, grilling him over a series of papers the Harvard physicist had written or co-written, dating back to 1971, that discussed the possible dire consequences of several disturbing political, environmental, and cultural trends. Holdren, unruffled, parried the attack by saying that he was simply calling attention to mounting ecological worries or that accumulating scientific evidence now points in a different direction.

Not surprisingly, environmental issues dominated the discussion. But the tone was conversational rather than confrontational. Asked by Rockefeller how she deals with disagreement among scientists on climate change, Lubchenco explained that “science doesn't tell us what to do. It's one of many factors.” Holdren's answer was more assertive: “It's real, it's accelerating, it's caused in large part by human activity, it's dangerous, and it's getting worse.” But Holdren also used the question to offer advice on how policymakers might approach any science-based issue. “Policymakers should look at the range of scientific opinion, the center of gravity, and what most expert bodies have said [about an issue],” Holdren said. “If I were betting the public's money, I'd go with the mainstream.”

Holdren painted an expansive picture of his office, telling Rockefeller that “it is my responsibility to look at any place where science and technology is not being put to its best use.” Few of his predecessors have been given such a broad portfolio, and it remains to be seen whether Holdren will fare any better in the rough-and-tumble of White House politics. But he's going to try. Noting that an interagency coordinating body called the National Science and Technology Council had “languished” under the Bush Administration, Holdren promised it would have a voice on everything from Arctic exploration policy to water allocations during droughts. He also reiterated Obama's promise to create a National Space Council to coordinate federal policy.

Lubchenco endorsed a proposal from her predecessor, former Navy Vice Admiral Conrad Lautenbacher, to create a National Climate Service within NOAA that would “do the same thing for climate as the National Weather Service does for weather…coordinating a wealth of data from other agencies to model how the climate system works.” At the same time, she criticized NOAA's partnership with two other federal agencies in developing a multibillion-dollar system of Earth-monitoring satellites, calling the current collaboration an “embarrassment that needs to be fixed.”

4. RETROVIRUS MEETING

HIV/AIDS Researchers Reach for High-Hanging Fruit

1. Jon Cohen
1. Julia Higgins is professor of Polymer Science at Imperial College, London, and foreign secretary and vice president of the Royal Society.

MONTREAL, CANADA—As Ringo Starr once sang, “It don't come easy.” That was the unofficial refrain at the HIV/AIDS meeting held here last week, the largest annual gathering for the field in North America.

The 4200 researchers who attended the 16th Conference on Retroviruses and Opportunistic Infections heard about steady progress on several fronts, but unlike in years past, there was hardly a peep about new anti-HIV drugs and no major surprises surfaced about existing treatment or prevention strategies. “There's nothing that knocks my socks off,” said Mario Stevenson, a virologist at the University of Massachusetts, Worcester, who helped organize the meeting. “We're in an era of just hammering away at AIDS. We're not going to have breakthroughs and eureka moments at every meeting.” Yet the presentations here did offer many unexpected findings on a wide range of topics, including microbicides, the search for a cure, “elite controllers,” and even chimpanzees.

The most talked about prevention study starkly illustrated that “success” now often comes with a long list of provisos. Trials of microbicide gels to protect women against HIV infection have a perfect record: Not one has worked, and some were even harmful. Now, however, a large international study of a vaginal microbicide called PRO 2000 may have finally ended that curse, although the gel's benefits appear to be modest and require confirmation.

Epidemiologist Salim Abdool Karim of the Centre for the AIDS Programme of Research in Durban, South Africa, reported that PRO 2000 reduced the risk of HIV infection by 30% during the 3-year, $90 million study, which involved 3000 sexually active women in four sub-Saharan African countries and the United States. Before intercourse, the women used PRO 2000, another experimental microbicide called BufferGel, an inert placebo gel, or nothing at all. The 750 women in the PRO 2000 arm of the study had 36 HIV infections, whereas the other groups had between 51 and 54 infections each. Although this was a low level of protection and the finding did not reach statistical significance, Karim emphasized that women in much of sub-Saharan Africa often do not have the option of using condoms. “In that population, 30% protection to me is a big difference,” said Karim. “Finally, there's been a signal in the microbicide field, and that's a thrilling event,” said epidemiologist Sten Vermund of Vanderbilt University in Nashville, Tennessee. The study lends support to an all-but-rejected prevention approach, a microbicide gel with a so-called nonspecific mechanism. PRO 2000 has a negative charge that theoretically can bind to positively charged HIV surface proteins, blocking the virus's ability to infect cells. Several failures of such nonspecific approaches have led many investigators to place their bets on products that incorporate anti-HIV drugs. A larger study of PRO 2000 is under way and should determine whether it works later this year, but even the cautious are hopeful. “It's exciting to find a positive trend,” said a skeptic of nonspecific approaches, virologist Robert Grant of the University of California, San Francisco (UCSF). For the HIV-infected, the most pressing question is whether treatments can eliminate the virus altogether—a cure. Anti-HIV drugs can powerfully suppress the virus, but ultrasensitive assays have shown that no one has eliminated it. How the residual virus persists in blood and tissues is not clear. One camp holds that drugs do not stop all replication, allowing the virus to constantly infect new cells and copy itself at low levels. This, in turn, replenishes a reservoir of infected, long-lived cells that otherwise would die off in a few years. If so, then intensifying treatment with more potent drug cocktails could knock out the virus entirely. Virologist Robert Siliciano of Johns Hopkins University in Baltimore, Maryland, and colleagues tested that hypothesis by adding powerful anti-HIV drugs to the cocktails people with undetectable levels of virus were already taking. “What happens is that nothing happens,” said Siliciano. “Intensification has no effect on residual viremia.” This finding, he says, lends credence to another theory for HIV persistence: that the infected reservoir is made up of cells that provide immunologic memory for decades and thus is stable and does not need to be replenished. If so, drugs that target viral replication can never eradicate the reservoirs, he said, and a cure will require a different strategy. Researchers have long hoped that “elite controllers,” the small percentage of infected people in whom the virus remains undetectable for many years on the standard test without taking anti-HIV drugs, may have effective immune responses that hold clues for vaccine makers. Steven Deeks of UCSF, however, had dispiriting news about these lucky few. Deeks's studies suggest that in the majority of these controllers, the immune system works overtime to thwart the virus—with deleterious effects. These elite controllers have high levels of “activation,” an inflammatory state caused by the immune system laboring to control HIV. That activation depletes critical CD4 white blood cells, the hallmark of AIDS; it also causes systemic inflammation that contributes to atherosclerosis and other complications. Indeed, four of 58 elite controllers he studied progressed to AIDS despite having only residual viremia. “If I were an elite controller, I'd seriously think about going on treatment,” he said. In 40% of elite controllers, another immune mechanism seems to control HIV, but Deeks says he has no idea what that is. These controllers do not have high levels of activation, nor do they have much T-cell immunity— which many believe is a key immune response to battle HIV. Both of these findings throw a curve ball to vaccine makers. Researchers assume that even if an AIDS vaccine does not stop an infection, it will bolster immune systems so people who do become infected effectively become elite controllers. But if most elite controllers suffer from immune activation, this becomes a far-from-ideal outcome. The second group of controllers who do not have an activation problem further confound the vaccine search, as T-cell immunity is a cornerstone of most products under development. Is there another critical immune response that vaccine makers need to target? One of the biggest surprises at the meeting came from studies of wild chimpanzees. Researchers have long assumed that SIVcpz, the chimpanzee virus that infected humans and triggered the AIDS epidemic, caused no harm to apes. But new data reveal that wild chimps infected with SIVcpz are more likely to die than are uninfected chimps. The animals also show AIDS-like damage to their immune systems. The finding raises the possibility that some chimp populations are suffering from AIDS epidemics. Evidence suggests that SIVcpz, discovered in 1989, is most closely related to HIV-1 and predates it. Captive chimps experimentally infected with HIV-1 typically suffer no harm, which led several researchers to propose that chimps had lived with SIVcpz for centuries and that their immune systems had evolved to coexist with the virus. But few SIVcpz-infected chimps in the wild were identified until about a decade ago, when researchers led by Beatrice Hahn of the University of Alabama, Birmingham, developed a way to routinely test fecal samples for evidence of the virus. Although SIVcpz has not been found in several chimp communities studied, some have a prevalence as high as 35%. Rebecca Rudicell, a graduate student in Hahn's lab, reported that she and her colleagues analyzed 1099 fecal samples collected between 2000 and 2008 from chimpanzees living in Gombe Stream National Park in Tanzania. They found evidence of SIVcpz infection in 18 chimps. Seven of the 18 infected chimps died during the study period, compared with 10 of 76 uninfected animals, said Rudicell. When they corrected for age and other variables, the scientists found that the SIVcpz-infected chimps had a 15-fold higher risk of death than did virus-free apes, meaning that SIVcpz poses nearly as great a risk as HIV-1 does to humans. Studies of lymph nodes from two of the infected chimps that died also showed the type of immunologic destruction seen in HIV-infected humans. And these chimps had low levels of CD4 cells, the lymphocytes that are the main targets of SIVcpz and HIV-1. “We were shocked at the initial discovery of SIVcpz in the Gombe chimps and even more dismayed when we established that it seems to be pathogenic,” said behavioral ecologist Anne Pusey, who runs the Jane Goodall Institute's Center for Primate Studies at the University of Minnesota, Twin Cities, and collaborates with Hahn. “It must be the case that some of the [chimp] mortality over the last decades has been due to SIV.” During their study period, the prevalence ranged from 9% to 18%, which mirrors the devastating levels of infections seen in human populations in the hardest hit countries in sub-Saharan Africa. The finding raises provocative questions about the relationship between HIV-1 and SIVcpz. For instance, why does SIVcpz harm chimp immune systems when HIV-1 doesn't? The work might offer clues to vaccine makers, too, about which immune responses to target. Also unknown is whether SIVcpz has contributed to the alarming chimp decline seen elsewhere. But once again, the answers to those questions surely won't come easy. 5. SCIENCEINSIDER From the Science Policy Blog Lifestyles of the rich and famous. That's one way to describe the world of science this week. Here are some highlights from our science policy blog, ScienceInsider. First the rich. Science agencies continued to rack up billions, as Congress finally completed work last week on a$787 billion economic stimulus package. The biggest winner is the National Institutes of Health, which would receive $10 billion for research and facilities. Also seeing green is the Advanced Research Projects Agency for Energy, established by Congress 2 years ago to inspire risky energy and climate-related research. The concept hadn't gotten a dime in the regular appropriations, but lawmakers threw$400 million at it as part of the stimulus.

Now for the famous, and it doesn't get any more famous than Tom Hanks. This week, the movie star stopped by CERN near Geneva, Switzerland, to chat about Illuminati and antimatter. So what does he think of the world's largest particle physics laboratory? “I love seeing science fiction become science fact,” he told one of our reporters on the scene. Speaking of CERN, its Large Hadron Collider should be back in business by late September.

In Washington, John Holdren and Jane Lubchenco sailed through a joint Senate confirmation hearing, while the House of Representatives passed legislation to overhaul environmental and safety research related to nanotechnology. If the Senate follows suit, the new law will set up a nano czar in the White House and require a research plan.

Finally, what do you get when you cross a vacuum cleaner with a tree? A potential geoengineering solution called air capture, which aims to lower atmospheric CO2 levels by literally sucking the greenhouse gas out of the sky. We smell a sci-fi movie. Someone get Tom Hanks.

6. AAAS ANNUAL MEETING

Al Gore to Scientists: 'We Need You'

1. David Grimm

AAAS ANNUAL MEETING, 12–16 FEBRUARY 2009, CHICAGO, ILLINOIS

At a meeting full of scientific celebrities, a former politician proved to be the greatest draw. Of course, Al Gore is no ordinary politician, especially to any scientist interested in climate change. One might say he's a politician turned scientist, and as an invited speaker on the second night of the meeting, the former vice president sought to reverse the equation: He asked all the scientists in the audience to get involved in politics. Gore began by saying that the economic crisis is intertwined with the climate crisis. Both, he said, have their roots in our dependence on carbon-based fuels. He even went as far as to compare climate change with the mortgage meltdown. “We now have $7 trillion worth of subprime carbon assets whose value is based on the assumption that it is perfectly all right to put 70 million tons of global warming pollution into the thin shell of atmosphere surrounding our planet,” he said. Then it was on to An Inconvenient Truth territory, with Gore updating the doomsday scenarios he laid out in his book and 2006 movie: The Arctic ice is melting faster than we anticipated, the Maldives is trying to buy itself a country that won't be under water, and no one seems to notice anymore when 1 million people are evacuated from New Orleans. “Is this the new normal?” Gore asked. So what can scientists do? Educate the public, for one. These days, Gore's number-one enemy is “clean coal.” He says the coal industry is spending half a billion dollars to mislead society about the dangers of fossil fuels—much like the tobacco companies whose ads touted the health benefits of cigarettes in the mid-1900s. “When they spend$500 million putting their version of this story in the minds of the American people, it increases the importance of you being willing to speak out,” Gore said.

But Gore wants scientists to do more than talk. He wants them to get involved. Science and politics have been separated for too long, he said. “Now that the survival of our civilization is at risk, and now that the solution to this crisis depends on the rapid spread of understanding from the world of science into the world of policy,… scientists can no longer in good conscience accept this division between the work you do and the civilization in which you live.”

Some scientists have already joined the fray, Gore noted, referring to John Holdren, Jane Lubchenco, and other researchers who will advise President Barack Obama. “The policymakers are of you,” Gore said. “Keep your connections to them. Become a part of this struggle. We need you.”

Judging from the standing ovation, Gore might have won a few recruits.

7. AAAS ANNUAL MEETING

Will Many Endangered Species Recover?

AAAS ANNUAL MEETING, 12–16 FEBRUARY 2009, CHICAGO, ILLINOIS

In the past 35 years, only the peregrine falcon and a handful of other species have recovered enough to be taken off the U.S. government's list of threatened and endangered species. Others, such as the California condor, require constant help from humans to survive the threats they face. In fact, the vast majority of these species are “conservation reliant,” said John Wiens of PRBO Conservation Science in Petaluma, California. And they may never be taken off the list.

Wiens and Michael Scott of the U.S. Geological Survey in Idaho wanted to know how many of the 1300 listed species will require constant conservation to endure. They examined recovery plans issued for 1100 species and checked whether they will need continual help. More than 80% are conservation reliant and will need to remain on the list, they found. “We thought it would be lower,” Wiens said. “We were quite astounded.”

The situation is likely to get worse. Because habitat destruction and other threats are increasing, more species will probably need to be listed—adding to costs of keeping species on life support. That means more thinking will be needed about how to prioritize funds spent on endangered species.

8. AAAS ANNUAL MEETING

Time Traveling at AAAS

AAAS ANNUAL MEETING, 12–16 FEBRUARY 2009, CHICAGO, ILLINOIS

Here are some of the highlights of other stories filed by Science reporters at the AAAS meeting. For more complete coverage of the meeting, go to Findings.

Considering the theme of this year's meeting, “Our Planet and Its Life: Origins and Futures,” it's no surprise that many of the talks traveled through time. First, it was off to the future with a sober assessment of what lies ahead for our planet. Perhaps the most frightening news is that the worst case scenarios analyzed by the Intergovernmental Panel on Climate Change (IPCC) may not have been dire enough. Humans have recently been pumping out climate-warming gases faster than the IPCC anticipated. These changes spell bad news for humans, as coastal cities flood and diseases like malaria spread to new areas.

Another series of talks focused on the “Origins” theme of the meeting by peering into humanity's past. Particle accelerators are allowing scientists to view our history like never before, as beams a billion times brighter than a hospital x-ray bring ancient manuscripts and statues to life. Scientists are also learning more about our distant relatives. Neandertals at one site rarely knew their grandparents, for example, as most died before the age of 30. And what of those Indonesian “hobbits” that have confounded scientists since their discovery in 2004? Researchers presented additional evidence at the meeting that they are not deformed Homo sapiens but rather represent a small species of human.

Speaking of small, scientists reported a genetic mutation that explains why little dogs, such as dachshunds and Scottish terriers, have such stumpy legs. Other researchers, perhaps seizing on the fact that Valentine's Day fell during the meeting, probed the evolutionary importance of kissing. And if you ever wanted to fold a piece of paper into a three-dimensional rabbit, scientists are working on a computer program called an origamizer to help you out.

9. AAAS ANNUAL MEETING

1. Ann Gibbons

AAAS ANNUAL MEETING, 12–16 FEBRUARY 2009, CHICAGO, ILLINOIS

Ever since researchers found fossils of Homo erectus beneath a medieval castle in Dmanisi, Georgia, they have been chipping away at the image of this venerable human ancestor. At 1.8 million years old, the fossils are the earliest members of the human family known outside Africa. Now, it turns out that they managed to trek all the way across Africa and the Middle East with the most primitive kind of stone tools known rather than with more sophisticated stone hand axes that were thought to be essential for intercontinental travel.

The textbook vision of the first world traveler has changed, says paleoanthropologist David Lordkipanidze of the Georgian National Museum in Tbilisi. This is the third time Lordkipanidze's team has revised the textbook view of early H. erectus, suggesting that it was more primitive than expected. First, his group published the brain size of the fossils at Dmanisi, which had a volume of just 650 cubic centimeters—not much larger than an australopithecine's brain volume of 450 cc. Then, the team found leg bones and announced that the Dmanisi people were short. Now, they have found Oldowan, Mode 1 stone tools at Dmanisi (see picture), not the retouched Acheulean hand axes that were a kind of Swiss Army knife for H. erectus in Africa.

“I'm not at all surprised,” says paleoanthropologist Robert Blumenschine of Rutgers University in New Brunswick, New Jersey. He says it's not uncommon to find evidence of both types of technologies in the same fossil locality. “The Oldowan tools were still good tools—they used them for different things,” says Blumenschine. There was no such thing as technological obsolescence—yet.

10. ECOLOGY

Tree Rings Tell of Angkor's Dying Days

1. Richard Stone

AAAS ANNUAL MEETING, 12–16 FEBRUARY 2009, CHICAGO, ILLINOIS

Archaeologists have long puzzled over the collapse of the mighty medieval Khmer kingdom in Southeast Asia best known for its resplendent capital, Angkor. New findings suggest that a decades-long drought at about the time the kingdom began fading away in the 14th century may have been a major culprit.

Evidence for a megadrought comes from centuries-old conifers that survived the Angkor era. At a conference* earlier this week in Dalat, Vietnam, tree-ring scientist Brendan Buckley of Columbia University's Lamont-Doherty Earth Observatory in Palisades, New York, described how the annual growth rings of conifers in Vietnam reveal a sharp weakening of Asia's summer monsoon from 1362 to 1392 C.E. and again from 1415 to 1440 C.E., just as the Little Ice Age was setting in and right when the Khmer kingdom was reeling. The tree-ring chronologies “represent a major breakthrough in tropical dendrochronology,” says David Stahle, a tree-ring expert at the University of Arkansas in Fayetteville.

Finding this new trove of data wasn't easy. Many tree species in Southeast Asia lack growth rings or have ones that cannot be used to reveal annual patterns, says Buckley. But in the past 5 years, he and others have validated several species in the region with annual growth rings. The evidence presented in Dalat comes from a rare conifer, the po mu (Fokienia hodginsii), spanning 7 centuries.

Its drought-revealing tree rings corroborate similar climate data from coral reefs and, most recently, stalagmites and stalactites that peg monsoon changes to the fall of other Asian societies (Science, 7 November 2008, p. 837). “The evidence for pronounced weakening of the monsoon is indisputable,” says Daniel Penny, co-director of the Greater Angkor Project (GAP) at the University of Sydney in Australia.

The Khmer kingdom, which encompassed much of modern-day Cambodia, central Thailand, and southern Vietnam, would not be the first civilization to literally bite the dust. A series of droughts devastated the Maya city-states of the Yucatán Peninsula between 800 and 900 C.E., around the time Angkor was rising. It's still unclear whether prolonged drought several centuries later brought a vibrant Khmer kingdom to its knees or was a coup de grâce to a staggering society. Nevertheless, the medieval tree rings may offer a lesson for the modern world: Harsh weather events, predicted to grow more frequent with global warming, may imperil communities on the knife edge of sustainability.

Angkor's rulers, on the backs of a huge corvée labor force, built hundreds of temple complexes—including Angkor Wat, humanity's largest religious monument—and carved hundreds of kilometers of canals and massive reservoirs that appear to have been used both for irrigation and for religious ceremonies. Then, in one of archaeology's enduring mysteries, Angkor was largely abandoned by the 16th century.

Theories abound for how the wealthy kingdom fell, with one of the latest ideas being that Angkor's vaunted waterworks grew too complex to maintain (Science, 10 March 2006, p. 1364). Recent archaeological and pollen findings from GAP indicate that Angkor's great reservoirs and storage ponds began operating at sharply reduced capacity several decades before the back-to-back droughts. That evidence suggests Angkor was in trouble long before drought set in, says GAP co-director and Sydney archaeologist Roland Fletcher. “Climate instability would have been a severe problem for a massive and inflexible water network to manage,” he says.

“I'm very concerned that the story will end with, there was a drought and Angkor collapsed,” adds Penny. “It's not as simple as that—and [it's] far more interesting.”

• *Climate Variability in the Great Mekong River Basin, 16–18 February.

11. MATERIALS SCIENCE

Is Silicon's Reign Nearing Its End?

1. Robert F. Service

Silicon is almost synonymous with computer chips. But as the semiconductor struggles at the minute scales of today's devices, chipmakers are being forced to consider other materials.

Silicon is almost synonymous with computer chips. But as the semiconductor struggles at the minute scales of today's devices, chipmakers are being forced to consider other materials

If you want to see real nanotechnology in action, check out Intel's Penryn computer chip. It contains some 820 million transistors, each with features just a few tens of nanometers across. These transistors are so small that more than 2 million can fit on the period at the end of this sentence. A device inside each one flips an electrical switch on and off as many as 300 billion times a second. In the time it takes for one such flip, light travels less than half a centimeter.

Amazing—but not good enough. Late this year or early next, Intel plans to introduce a new line of chips that shrinks the components even smaller. For the past half-century, this perpetual contraction has been at the heart of the industry's favorite trend: Moore's Law, which holds that the number of transistors on chips will double about every 2 years. For most of that history, this downsizing, known as scaling, came about as engineers developed ever sharper chip-patterning techniques. But the past few years have brought a quiet revolution to chipmaking. Because conventional materials have started misbehaving at such small scales, engineers have had to call in reinforcements. Since the 1990s, chipmakers have gone from making their devices out of about 15 chemical elements to making them out of more than 50, in hope that the new additions will help them keep shrinking the devices.

Few doubt that Moore's Law will hold for a couple more generations of chips, and perhaps even longer. But continuing this trend will not be a straightforward enterprise. Researchers are looking at redesigning the way they make transistors, incorporating new insulators, and even replacing silicon as the semiconductor through which electrical charges flow in their circuitry. “There are very serious challenges ahead,” says electrical engineer Jesús del Alamo of the Massachusetts Institute of Technology (MIT) in Cambridge. Adds electrical engineer Mark Rodwell of the University of California, Santa Barbara, “Regardless of which [technology] wins, there's a real sense that there aren't too many years left to play this game.”

Matchmaker, matchmaker

At the center of this game are transistors, or more precisely, metal-oxide-semiconductor field-effect transistors (MOSFETs). They work by sending a pulse of electricity to a central electrode called the gate (see figure, below). The charge on this gate causes a spike in the electrical conductivity of a semiconducting channel that sits below it. This opens an electrical doorway in the channel, allowing an electric current to flow between two other electrodes—known as the source and the drain—that sit at opposite ends of the channel. The high electrical output at the drain is a “1” in the on/off digital world. If you turn off the charge on the gate, the channel door closes and the current flow to the drain shuts off, giving you a digital “0.”

Silicon MOSFETs have worked wonders in recent decades, thanks in large part to a unique marriage in the material world. In a MOSFET, the gate electrode and the conducting channel beneath it must be insulated from each other, so that when the gate is closed, any excess charges in the gate aren't able to open the channel and let current flow between the source and the drain. In traditional logic chips that are the brains of desktop computers and servers, that insulating layer has been provided by silicon dioxide (SiO2).

Silicon and silicon dioxide's storybook marriage has thrived for decades because where they meet, they form a near-perfect union. At these interfaces, each silicon atom binds readily to four oxygen atoms. That's good, because it ties up unfilled bonds at the silicon channel's edge that can trap electrical charges in place and disrupt the transistor's ability to switch when prodded. Whenever defects occur, engineers can easily neutralize them by piping in hydrogen to latch onto any free bonds. The result has been that although other semiconductors are faster and stronger, the near-perfect union between silicon and silicon dioxide has made it possible to continually improve transistors.

Recently, however, the stress of relentless scaling has proved too great for SiO2, which is only a moderately effective insulator. As transistors continued to shrink, engineers were forced to make their SiO2 layer as thin as 1 nanometer, or about three atomic layers thick. That was so thin that it started to leak. “We were just running out of atoms,” says Suman Datta, an electrical engineer and Intel veteran, who now works at Pennsylvania State University, University Park.

In the mid- to late 1990s, researchers at Intel, IBM, and elsewhere realized the days for silicon dioxide were just about up and that they needed to ditch silicon's devoted partner for a new trophy wife. Their goal was to find a material with a higher insulating value, known as its dielectric constant and denoted by the Greek letter kappa (κ). Researchers tested dozens of “high-k” alternatives, finally settling on a new bride called hafnium dioxide (HfO2).

“It started off innocently, saying we'll just swap out the SiO2 with a high dielectric constant [material] and be on our way,” says Supratik Guha, a materials scientist and senior manager at IBM in Yorktown Heights, New York. Unfortunately, the change wasn't so simple. Among many other problems, the new insulator didn't form a clean interface with the gate above it, which was made from polycrystalline silicon. So engineers were forced to replace the polysilicon gate with titanium-based alloys that performed better. That change raised new problems, because the metal alloys couldn't handle the high temperatures used in manufacturing some of the other components. Ultimately, Intel worked out a manufacturing strategy, and last year the company began shipping chips made with the new metal gates and HfO2 insulator. At the time, Gordon Moore, co-founder of Intel and eponymous drafter of the law, called the switch “the biggest change in transistor technology since … the late 1960s.”

What is sobering is that this relatively straightforward change took the industry a good 10 years to accomplish, and the challenges ahead appear to be far greater. “The complexity of the problems is only going to grow,” Datta says. For starters, as Intel and other chip companies prepare to contract from the 45 nanometer scale to 32 nanometers (the numbers refer to half the distance between adjacent lines of memory cells), it appears that silicon dioxide is set to cause a new round of trouble. Even though SiO2 was supposedly replaced as the gate insulator, a little bit has remained behind because it is nearly impossible to eliminate. When silicon is placed next to hafnium dioxide, Datta explains, some oxygen atoms at the interface invariably break their bonds to hafnium and hook up with silicon. This isn't all bad, he says, because the clean interface with that ultrathin layer of SiO2 seems to improve the conduction of electric charges through the semiconductor channel. But with all the device components set to shrink again, the SiO2 layer must also get thinner—and that could disrupt the flow of charges in the silicon. “It's a very tough problem,” Datta says.

At the International Electron Devices Meeting (IEDM) in San Francisco in December 2008, Intel researchers reported that they had solved this problem and many others. The company announced that it had completed the development phase of its 32-nanometer-manufacturing process and would begin turning out the new chips by the end of 2009. Intel researchers haven't revealed their latest tricks, but Datta calls the performance data they have presented “excellent.” Datta says he expects Intel to stick with HfO2 for now, because it was so expensive to make the switch in the first place.

A new look

So if the next step in shrinking silicon electronics is on track, what's next? After 32 nanometers, the next step down is 22 nanometers. “This is where things get very interesting,” Datta says. At this dimension, it's likely that HfO2 will also begin to fail to contain current within its walls. If that happens, one option is to change to a material with an even higher dielectric constant, says Darrell Schlom, a high-k expert at Cornell University. Schlom's group has tested more than a dozen. Among the best, he says, is crystalline lanthanum lutetium oxide, which has a dielectric constant of 40, more than 10 times that of SiO2 and nearly double that of HfO2. A big challenge, however, is that the more elements make up the insulating material, the harder it is to keep the perfect order of the material at the interfaces.

Another option at this scale is to redesign the architecture of transistors altogether. One alternative, considered most likely for 22-nanometer scale devices, is to move away from layered, sandwichlike devices and stand the silicon channel on its side. Such devices, known as FinFETs (because the vertical silicon channel looks a bit like a fish fin), in principle would allow engineers to surround the channel on three sides with dielectrics and gate materials. Then using several gates in concert to trigger the flow of current in the channel could make it easier for engineers to control when the devices flip on and off and how much current they put out when they do.

Academic researchers have been turning out FinFETs and other exotic-shaped devices for years. Now, however, even the big chip players are looking at exotic designs. At IEDM, for example, a consortium of researchers from Toshiba, IBM, and Advanced Micro Devices reported making the world's smallest FinFET transistors with high-k dielectrics and metal gates. The new transistors were only half the size of previous FinFET record holders, and studies revealed that their geometry made them more reliable than planar versions, according to company press materials. But, at least for now, the new devices suffer from poor electrical contacts between the source and drain electrodes, Datta says.

Beyond silicon

Even that switch, however, looks simple compared with the changes further ahead. After the 22-nanometer scale comes 15 nanometers, and it appears—at least for now—that improvements in transistor speed and performance will have to come from new materials rather than from scaling silicon further. “Now that you've gotten rid of the SiO2, why not have the high-mobility channel and high-k materials to put on top of it?” Schlom asks. In other words, why not replace the silicon with a better semiconductor?

Top candidates for this switch are semiconductor alloys know as III-Vs, due to the position of their elements in the periodic table. Examples include gallium arsenide, indium gallium arsenide (InGaAs), and indium antimonide (InSb), all of which can ferry electrical charges much faster than silicon can (see table, above).

But there are lots of problems in marrying III-Vs with modern logic-chip technology. For starters, it is not currently possible to make large-scale III-V wafers on which to construct the devices from the ground up. That means engineers need to integrate III-V materials on the silicon wafers that are the industry standard. That's no easy task, because the atomic spacing in III-Vs is significantly different from silicon. Try to grow InGaAs, for example, directly on silicon and you get an interface riddled with free atomic bonds that drastically slow the flow of charges in the device. If the interface between a III-V semiconductor and the silicon wafer beneath it is a problem, equally difficult is the interface to a high-k dielectric above it. And so far, efforts to neutralize defects have proven underwhelming. Unlike silicon and SiO2's perfect marriage, “there's no wonderful interface with the IIIVs,” Schlom says.

Another challenge is making transistors that conduct positive charges, called holes. That's important for chip designers, as it allows them to alternate transistors that carry negative charges (n-type) and positive charges (p-type), which, when working together, give clearer electronic readouts and use less power. Silicon ferries holes nearly as well as it does electrons. But most III-Vs are far better at conducting electrons than holes. “There are a lot of problems that may seem insurmountable,” Schlom says, “but I'm cautiously optimistic that they can be solved.”

Some steps toward those solutions are already evident. In 2007, for example, researchers at Intel led by Robert Chau reported growing InGaAs transistors atop a thin buffer layer sitting on a silicon wafer. The buffer allowed the two dissimilar materials to sit next to each other with relatively few problems. Researchers elsewhere have also had success with InGaAs. In one example, researchers led by Peide “Peter” Ye of Purdue University in West Lafayette, Indiana, reported in the April 2008 issue of IEEE Electron Device Letters that they had made high-speed InGaAs n-type transistors that had a large output current when switched on, an important hurdle for the field.

Still, Ye and others acknowledge that III-Vs aren't there yet. Ye's InGaAs devices, for example, work splendidly when switched on, says materials scientist Paul McIntyre of Stanford University in Palo Alto, California; but try to switch one off and current still leaks through, like water overtopping a levee. In addition, neither Ye's group nor any other has had much luck in making high-quality p-type transistors from InGaAs. At IEDM, Chau and colleagues at Intel did report novel InSb p-type transistors that are the fastest to date. But, for now, they too remain leaky when switched off.

There might be a workaround. Researchers at IBM and elsewhere have also had some success with making high-speed p-type transistors using germanium (Ge). And though Ge isn't itself a III-V compound, it seems to integrate fairly well with them. One hope, Ye and others say, is that researchers will be able to make integrated circuitry using InGaAs n-type transistors next to Ge or InSb p-type transistors. Even if they succeed, however, the circuits could be prohibitively expensive to manufacture, as they require patterning very different materials in the same layer of the chips. Still, despite these challenges, Ye remains optimistic. “I think the future is still bright because there is no other choice,” he says.

That's not entirely true. If III-Vs don't pan out, or perhaps even if they do and last only a generation or two, there are plenty of far-out ideas for reinventing microelectronics. Among them: transistors made with single-layer carbon sheets called graphene, carbon nanotubes, or III-V nanowires. But for now, these upstarts still need plenty of work to have a shot at dethroning more-conventional approaches.

In any case, it's clear that this work needs to happen soon, or the steady progress of Moore's Law will begin to slow if not stop. “My own view is that we will not see another decade of scaling,” McIntyre says. Perhaps not. But MIT's del Alamo and others point out that the $260-billion-a-year chip industry has been jumping over roadblocks for years. “The potential payoff is gigantic,” del Alamo says. Adds Datta: “One thing I've learned is, don't predict the end of Moore's Law.” 12. APPLIED PHYSICS New Facility Propels Korea to the Fusion Forefront 1. Dennis Normile Using innovative magnets that should confine plasmas for minutes rather than seconds, KSTAR is poised to become a premier testbed for fusion research. Using innovative magnets that should confine plasmas for minutes rather than seconds, KSTAR is poised to become a premier testbed for fusion research DAEJEON, SOUTH KOREA—At first glance, the fist-size bundle of wires on a conference table at the National Fusion Research Institute (NFRI) here looks like scrap metal doubling as a paperweight. But then NFRI President Gyung-Su Lee points to it as evidence of the engineering prowess that has thrust Korea to the fusion frontier: The wires are a sample of superconductive cable that's wound into magnets at the heart of the world's most advanced fusion research facility, the Korean Superconducting Tokamak Reactor (KSTAR). The ability to fashion high-performance superconducting magnets from a niobium-tin alloy was a key technology that NFRI's industrial partners mastered while building a machine that fired up its first plasma a few months ago. Engineers are now installing control and diagnostic equipment that will allow research to begin in earnest this fall. KSTAR's explorations will reverberate all the way to Cadarache, France, where a consortium is assembling the most important fusion experiment ever, the International Thermonuclear Experimental Reactor (ITER), expected to come online in 2016. “We definitely [wanted to] make KSTAR a really useful device for ITER preparation,” says Lee. Fusion physicists say the Koreans have succeeded. “KSTAR has a very important role to play [in providing data] that will be used to design operating scenarios for ITER,” says Hutch Nielson, a plasma physicist at the Princeton Plasma Physics Laboratory. ITER will tackle a decades-old question: whether the fusion process that powers stars can be harnessed to produce electricity (Science, 13 October 2006, p. 238). It will use powerful magnets to confine a plasma within a doughnut-shaped high-vacuum vessel called a tokamak. At about 1 million degrees Celsius, the plasma's charged particles fuse, releasing energy. ITER will be several times larger and, at$6 billion, far more expensive than any existing tokamak.

Because ITER could cost more than $1 million a day to run, “you're going to have to operate as efficiently as possible,” says David Campbell, ITER assistant deputy director general for fusion science and technology. Before ITER comes online, he says, researchers will be learning to control plasmas using the world's existing half-dozen major tokamaks. KSTAR,1/25 of ITER's size, is the youngest in this cohort and will cost about$800 million after upgrades planned over the next several years. KSTAR's biggest advantage is its superconducting coils, which will enable it to confine plasmas for up to 300 seconds, compared with the 20 or so seconds of older tokamaks.

Catching up

Soviet Premier Mikhail Gorbachev proposed building what became known as ITER to U.S. President Ronald Reagan in 1985. The original four partners represented all those nations with a serious investment in fusion research—the Soviet Union, the United States, the European Union, and Japan. Twenty years later, as ITER was moving from design to construction, three Asian nations wanted in. China and Korea joined the project in 2003, and India joined in 2005.

The newcomers were intent on showing that they could bring expertise as well as cash to the table. India embarked on its Steady State Superconducting Tokamak in 1994; technical problems have delayed commissioning. Korea initiated KSTAR in 1995, although it was put on hold for 2 years because of the 1997 Asian financial crisis. China completed its Experimental Advanced Superconducting Tokamak in 2006 (Science, 19 May 2006, p. 992). The three countries “are much more significant players in fusion than they were a decade ago,” says Nielson.

Of the new facilities, experts say KSTAR is particularly outstanding given Korea's limited experience in fusion research. Before KSTAR, Korea had a handful of fusion researchers working with pint-sized tokamaks. Building the most advanced tokamak to date “was courageous and visionary,” says Nielson, who consulted on KSTAR's design. He and others credit Korea's success to Lee. The physicist, like his colleagues in China and India, drummed up support for fusion research and for joining ITER by raising the alarm over future energy supplies. Korea is totally dependent on imported energy. “Because of the energy crisis and global warming, someday [fusion energy] had to get going, but Korea was not prepared,” Lee explains.

Lee first won backing from Korea's industrial titans by convincing them to get in on the ground floor or risk having to license fusion reactor technology from others. He then got the government to put up “a huge magnitude of money” by persuading bureaucrats that investing in a future energy source was like buying insurance.

Paving the way

As planned for ITER, KSTAR uses superconducting magnets for both the toroidal field, which vertically rings the vacuum chamber, and the poloidal field, which follows the curve of the torus horizontally. Only one other tokamak—China's—features fully superconducting coils and can confine plasmas for 300 seconds or more. Older tokamaks use copper magnets that operate in pulses of up to only 20 seconds or so before overheating. “The significance of KSTAR and [China's tokamak] is that they can run long pulses and explore how to operate for a long duration and also what kind of physics you're going to encounter in that long duration,” says David Humphreys, a plasma physicist at General Atomics in San Diego, California. ITER is expected to initially operate in 300- to 500-second pulses before ramping up to 3000 seconds.

KSTAR will also manipulate the plasma in ways particularly relevant to ITER, says Campbell. Some older tokamaks form plasma into a circular cross section. But ITER and KSTAR will aim for a sharply angled “D,” which is more effective for confining the plasma and reducing instabilities that can leak energy and damage the vessel. Scientists hope to refine the “D” in Daejeon. “The KSTAR magnets were designed to allow strong plasma shaping control research,” says Yeong-Kook Oh, head of experiments and operations for KSTAR.

Also paving the way for ITER, KSTAR uses the same systems to heat plasma and boost plasma current. These include injecting neutral particles and zapping the plasma with radio waves. KSTAR will also test exotic methods of taming instabilities, such as firing pellets of frozen deuterium into the plasma to release pockets of pent-up energy that otherwise cause turbulence.

The KSTAR tokamak will initially be lined with carbon-based tiles, but it might be modified later to test tungsten-based materials favored for ITER, says Oh. Already, KSTAR has proven the feasibility of working with the finicky niobium-tin alloy that ITER intends for its magnets.

No existing tokamak, KSTAR included, can achieve a burning plasma, in which at least half the energy necessary for fusion is generated internally. ITER is designed to produce more energy than it consumes. It will achieve that goal in part by relying on a fuel mix of deuterium and tritium, which fuses at a lower temperature than other gases, including deuterium alone, which is what fuels KSTAR and most other tokamaks.

KSTAR can't prove fusion is the energy of the future. But until ITER is fired up, this Asian upstart will be the hottest testbed in the world for fusion research.

13. RICHARD RICHARDS PROFILE

Making Every Drop Count in the Buildup to a Blue Revolution

1. Elizabeth Finkel*
1. Elizabeth Finkel is a writer in Melbourne, Australia.

Richard Richards, a geneticist at CSIRO Plant Industry, is breeding wheat varieties that can tough out prolonged droughts--and keep people fed.

Richard Richards is breeding wheat varieties that can tough out prolonged droughts—and keep people fed

LEETON, AUSTRALIA—Kneeling in verdant young wheat at Leeton Field Station, Richard Richards uproots seedlings of two varieties and splays them out for inspection. Looking on are several stern farmers and scientists from the Grains Research and Development Corp., which funds his work. They're a tough bunch to impress—but Richards has them spellbound. One seedling, “Vigour X-25,” has a coleoptile, or seed sprout, that's twice as long as the other's. To this audience, the meaning is clear. In drought, when the top few centimeters of soil dry out, seeds that grow longer coleoptiles can be sown deeper, in moist soil needed for germination.

In the 1950s, wheat breeder Norman Borlaug launched the Green Revolution by dwarfing wheat varieties, which diverts plant energy from stalk to grain and boosts yields several-fold. Today, a golden dwarf sea stretches from the U.S. Great Plains to southern Australia's Wimmera Plains. But dwarfs have a major shortcoming: short coleoptiles. That's a serious problem in many regions, where water availability is a key constraint for crop yields. With a rallying cry of “more crop per drop,” Borlaug exhorted fellow breeders to foment a Blue Revolution.

After 3 decades of dogged effort, Richards, a soft-spoken geneticist at CSIRO Plant Industry in Canberra, has nudged wheat to the brink of a Blue Revolution. “Richards uses a surgeon's scalpel” to tweak just the right physiological processes, says Brett Carver, a wheat breeder at Oklahoma State University, Stillwater. As a result, although most breeders these days content themselves with wheat yield gains of about 0.5%, Vigour X-25 and another Richards creation, Drysdale, are yielding 10% to 20% gains in arid conditions.

Hailing his achievements, the American Society of Agronomy last October awarded Richards its Martin and Ruth Massengale Medal for “significant contributions to new and innovative research in crop physiology and metabolism.” He's not resting on his laurels. Using DNA tags that track genes conferring more efficient transpiration and long coleoptiles, Richards is combining the best of Drysdale and Vigour X-25 in a variety that should be ready for commercialization next year.

The long and short of it

Australia's southern wheat belt has a Mediterranean-like climate, in which much of the annual rainfall comes during a mild winter. In the early 1980s, Richards's CSIRO team pondered how it could help farmers make the most of that precipitation pattern.

One day in 1985, a farmer sowed an idea in Richards's mind. “Wheat's slow off the mark, not like barley,” he recalls the farmer saying. Richards knew that barley is prized for its ability to thrive during drought. He speculated that barley's vigorous leafy growth acts as a shade to prevent evaporation from the soil. When wheat and barley are planted side by side, one obvious difference is that barley seedlings roll out wide leaves whereas wheat foliage looks spindly. A second difference lay nestled in the seed: Barley's embryo is nearly twice as big as a wheat embryo, giving it a substantial head start.

Guided by these clues, Richards scoured seed banks for wheat with larger embryos and wider leaves. Jinghong, a Chinese variety, has embryos that are 50% bigger, on average, than those of other wheats. And Kharchia from India has wide leaves. After crossing, the hybrids were even more vigorous than barley and cut evaporation from a field by 30%.

To Richards's surprise, the hybrids also had long coleoptiles. Those of Green Revolution varieties are 5 to 7 centimeters—short enough to lead to crop failure in parched conditions. “It's the baggage of the Green Revolution,” Richards says. The new varieties had another advantage: more highly branched roots, which double nitrogen uptake from soil. To add the benefits of dwarfing, Richards began crossing the hybrids with varieties of dwarf wheat with long coleoptiles. His newly developed Vigour X-25 is the result of a cross between an Italian dwarf and the Chinese-Indian hybrids.

U.S. farmers could benefit from this advance, says Carver. Great Plains wheat is sown in late summer to take advantage of sporadic summer rain, but because local varieties have short coleoptiles, farmers can lose up to one in 10 fields in dry summers and must resow. “We really need to incorporate [long coleoptiles] in the Great Plains,” says Carver, who has begun breeding Richards's varieties with local wheat.

Breeding for drought tolerance is notoriously difficult. Some breeders have selected varieties based on an ability to survive drought, but the result tends to be cactuslike plants that grow slowly and produce little grain. Breeders have also zeroed in on traits such as leaf curling to reduce evaporation, but these often vanish when bred into other varieties or when environmental conditions are altered.

Buoyed by the success of transgenic crops resistant to insects and to viruses, efforts are under way to design drought-tolerant crops (Science, 11 April 2008, p. 171). It's a big challenge, as many genes are involved. “To think gene transfer replaces conventional breeding for drought is unrealistic,” says Matthew Reynolds of CIMMYT, the International Maize and Wheat Improvement Center in El Batan, Mexico. “We don't yet understand the gene interactions well enough.”

Although many colleagues have embraced transgenic techniques, Richards is old school, patiently crossing varieties. He credits high school friend Graham Farquhar, a biophysicist at Australian National University in Canberra, for the inspiration that seeded Drysdale, his most successful cultivar. In the early 1980s, Farquhar, like Richards, was exploring how to alter plant physiology to resist drought. He targeted RuBisCO, an enzyme complex that captures CO2 and converts it into sugars. RuBisCO prefers CO2 with the carbon-12 isotope rather than heavier carbon-13. (About 1.1% of carbon in the atmosphere is carbon-13.)

Farquhar expected that the isotope ratio in plant tissue might shed light on plant breathing habits. Plants take in CO2 through stomata, which open and close like mouths. Farquhar imagined that some plants are like sponge divers, taking big gulps of air, then holding their breath for minutes. Other plants are like swimmers taking frequent sips of breath. With fresh CO2 every few seconds, swimmers have the luxury of picking out lighter carbon. Divers consume most CO2 in each gulp, including carbon-13. Compared with swimmers, the isotope ratio in divers should be closer to that of air. Crucially, because divers keep their stomata closed longer, they should lose less water.

Testing Farquhar's hypothesis, Richards confirmed that the most water-efficient wheat varieties had the highest carbon-13 to carbon-12 ratios. “That was a eureka moment,” says T. J. Higgins, deputy chief of CSIRO Plant Industry. At last, Richards had a solid physiological trait to select for drought tolerance. With that tool, he bred Drysdale, a variety that's more like a gulper than a sipper. Richards's isotope-discrimination technique, “Delta,” has revolutionized breeding for drought tolerance, says CIMMYT's José Luis Araus, who is applying it to breed drought-tolerant maize in sub-Saharan Africa.

As Australia's summers get hotter and drier, a new challenge is to breed wheat that matures faster and can be harvested before the grain withers. Again, Richards is eyeing barley: “It's got this magical property of producing a large amount of grain in a short period of time. If we can do that with wheat, it will be a massive breakthrough.” With his dedication to the Blue Revolution, that should only be a matter of time.

14. GENETICS

Rooting Around the Truffle Genome

1. John Bohannon

A favorite of gourmets, truffles are revealing their delicious secrets to the biologists studying the mysterious fungi. (You can join Science on a truffle hunt and an unusual taste test in the latest episode of the Gonzo Scientist.)

A favorite of gourmets, truffles are revealing their delicious secrets to the biologists studying the mysterious fungi

ALBA, ITALY—Upon first glance, it's hard to believe that this is one of the most prized and expensive foods in the world. The muddy clods laid out in the display case bear a striking resemblance to animal droppings. But their true identity is revealed the moment the seller, Stelvio Casetta, lifts the glass lid. The aroma—a potent, earthy cocktail of sulfurous chemicals—is unmistakable. Then you see the price tags, ranging from $100 to$400 for a lump of fungus smaller than Brussels sprouts. These are tartufi or, as the many English-speaking tourists here call them, truffles.

Casetta plucks out a $300 creamy, golden nugget with the same enormous hand with which he unearthed it just days ago at his secret location in a nearby forest. (You can join Science on a truffle hunt and an unusual taste test at http://www.gonzoscientist.org/.) “You can't know a truffle just using your eyes,” he tells potential buyers before handing it to a nearby woman, who takes it to her nose, inhales deeply, and smiles at the recognizable odor. Paola Bonfante, a microbiologist at the nearby University of Turin, knows her truffles, in some ways better than Casetta does. She is part of a Franco-Italian team exploring in intimate detail the prized fungi that make up the truffle genus. At a meeting in November,* the group gave a preview of the first full genome sequence of the black truffle (Tuber melanosporum), just the second symbiotic soil fungus to be so deciphered, and the genome sequence of the white truffle (T. magnatum) is expected by summer. Already, the European investigators have dug up several surprises among the fungal DNA sequence, including one that may help stem the truffle black market and another that rewrites the sex life of these subterranean organisms. The truffle is the latest in a series of gourmet genome projects pursued by European researchers. French and Italian scientists clinked glasses last year after sequencing the genome of the grape used for wine production (Science, 25 April 2008, p. 475). French scientists are now sequencing the genomes of the microbes responsible for Camembert and Roquefort cheeses. “It is natural that these should be French and Italian projects,… food is so important to our culture” says Francis Martin, a plant and fungal physiologist at the French National Institute for Agricultural Research (INRA) in Nancy who led the black truffle genome project. Truffle trouble For commodities such as gold and silver, the market value is the same regardless of where they are mined. But the price of a truffle is strongly determined by its geographic origin. The most highly regarded black truffle is la truffe noire du Périgord, harvested for centuries beneath oak trees in southwestern France. French researchers discovered how to reliably cultivate black truffles in orchards in the early 19th century and continue to do so to this day. But the white truffle has resisted domestication and thus commands far more money. The fungi can be found—with the help of pigs or trained dogs—only in a narrow swath of forests between the Istrian peninsula of Croatia and central Italy. And here at the Alba market, tartufo bianco d'Alba—the locally grown white truffle—is the undisputed king, routinely selling for$4000 per kilogram.

But a major problem for truffle buyers is “counterfeiting.” Black truffles bearing the prestige and price of the name Périgord sometimes originate from less famous regions. White truffles harvested in Croatia are brought to Italy and sold with “d'Alba” labels. Sometimes similar looking species, such as the plentiful but less aromatic Chinese truffles (T. indicum), which resemble black truffles, “are sold with a small amount of black truffle included to provide the right smell,” says Francesco Paolocci, a fungus researcher at the Institute of Plant Genetics (IGV) in Perugia, Italy.

So the truffle industry is turning to molecular biology for help. Researchers have “assumed for decades that truffles are almost clonal,” says Martin, with hardly any genetic differences between the fungi growing in different regions. Any distinct flavors, aromas, or appearances are chalked up to variations in the environment. Sampling the genomes of black truffles from around Europe has turned this view on its head, however.

The newly completed black truffle genome revealed areas with highly variable amounts of repetitive DNA. A team of Italian and French geneticists led by Paolocci, Andrea Rubini, and Sergio Arcioni at IGV and Claude Murat at INRA used these markers to take DNA fingerprints of more than 200 black truffles from 13 regions across southern Europe. Far from being a monoculture, black truffles form local varieties that are genetically distinct, the researchers reported at the November meeting. Paolocci's group has also fingerprinted more than 300 white truffles from 26 different areas across its range, and early results indicate that it may also form distinct regional populations. “Now that we have the global picture,” says Martin, “you see that the Tuber genome is like a mosaic,” with “islands” of stable genes separated by “an ocean of repeating DNA” that change rapidly over the centuries.

DNA analysis has already caught imposter species, such as the Chinese truffle, says Bonfante. The next step will be to refine the genetic fingerprinting to determine if truffles of the same species but different regions can be distinguished. “This will become important,” she says, if governments adopt the wine industry's “controlled geographical origin” system. The European Union will then require authentication of a truffle's birthplace.

Beyond fighting counterfeit fungi, the genetic data is filling out the truffle's evolutionary story. Using the new map of black and white truffle diversity, Martin and Murat recently modeled the spread of truffles going back 12,000 years to the last Ice Age. By the time that Europe was thawing, only two small populations of black truffles existed, in southern Italy and Spain, and white truffles were restricted to central Italy. Black truffles then spread over the Alps and across Europe, but white truffles never did. What held the white truffles back is a mystery. Considering changing climates, the question is of more than academic interest. “It is expected that the black truffle will be able to adapt to global warming by moving northward,” says Martin, but the white truffle, blocked by the Alps, could become extinct.

Sex in the soil

Everyone in Alba seems willing to testify that truffles are an aphrodisiac. The origin of this legend may be that, among the hundreds of volatile compounds that truffles emit, there is indeed a close mimic of androstenol, a mating pheromone secreted in boar saliva. This may explain the frenzied enthusiasm of sows when they locate truffles beneath leaf litter, says Bonfante. Evidence of any such behavioral effect on humans is lacking, but the question of the truffle's own sexual behavior has now been answered, thanks to the new genome data.

The only part of the fungus that ends up on dinner tables is the fruiting body, a temporary reproductive organ created at specific times of the year—between November and December for white truffles, December and February for black. The rest of the organism exists year-round as a fine web of hairlike cells, the hyphae, that sheath the roots of a tree, providing minerals in exchange for food. The spores packed into the fruiting body each contain two copies of the truffle's chromosome complement, whereas the hyphae cells only contain a single set. The elusive question has been what sexual acts hyphae get up to when producing spores.

According to the prevailing model, truffles do not have sex with strangers. Instead, they self-mate, with two hyphae of a single fungus fusing. The resulting cell, which would have two identical sets of the genome, then divides rapidly to form the spores of the fruiting body, surrounded by a matrix of cells, called the gleba, each containing single-copy genomes.

With the highly variable markers from the fully-sequenced truffle genome in hand, the IGV group has now tested whether the fungi are really so chaste. The investigators compared the DNA of gleba cells and spores in dissected black and white truffles. Like catching an adulterer in the act, they found different genetic fingerprints in the two cell types of both truffles. The gleba cells contained one set of DNA markers, and the spores carried those same markers plus a foreign set, the team reported at the meeting. In one stroke, says Paolocci, this work showed not only that the fruiting bodies were the result of a sexual encounter between two different fungi but also that gleba and spores have different cellular origins. Like a mother's womb, all of the fruiting body's gleba cells have just the “maternal” genotype, whereas the spores carry both, like a fertilized egg.

Confirming this, the researchers inoculated the roots of potted tree seedlings in the laboratory with black truffle spores. In each case, the hyphae cells that grew contained a Mendelian assortment of genotypes, exactly as predicted for sexual reproduction.

The discovery has “major implications” for the truffle industry, says Charles Lefevre, a mycologist based in Eugene, Oregon, who is considered one of the world's experts on the delicacy. Companies that sell tree seedlings colonized by black truffle fungi “have benefited from low or nonexistent seedling quality standards,” he says. Now sellers may have to prove that their trees are colonized by multiple truffle mating types that can produce the edible fruiting bodies.

What's that smell?

The sex life of truffles is only one of their secrets, says Martin. “We really don't understand truffle ecology.” For one thing, the nature of the chemical crosstalk between the hyphae and roots is still “a black box,” he says. And then there are the countless interactions with all the other residents of the soil. One clue to this ecosystem may literally be under our noses. “All those volatile chemicals that make truffles smell delicious to us may serve other purposes,” says Bonfante. When truffle fungi colonize a tree, a denuded zone, or brulé, that looks as though the ground has been scorched often develops around the trunk. Bonfante's group has shown that the microbial community structure in brulé soils is dramatically different when compared with soil around neighboring trees. “Truffles seem to trigger these changes with chemical signals,” possibly with the help of the tree, says Bonfante, “but we don't know how it works.”

Martin and his Italian colleagues plan to take a crack at the problem by creating a “volatile map” of truffle aroma in the field. “The idea is to capture the gases released by truffles at various stages of maturation and freeze samples of the tissue immediately,” Martin explains. Back in the lab, gene-expression patterns in the fruiting body and other tissues will be correlated with changes in the cocktail of chemicals released. The ultimate aim is to find the genes responsible for fruiting body development and for the symbiosis between tree and fungus.

Of course, learning the delicious secrets of truffle biochemistry could make it possible to genetically engineer truffle aroma into more easily cultivated organisms, such as mushrooms. But would Europeans accept genetically modified portabellos with the whiff of tartufo bianco d'Alba? “Absolutely not!” says Martin.

• *3rd International Truffle Congress, Spoleto, Italy, 24–28 November 2008