# News this Week

Science  06 Mar 2009:
Vol. 323, Issue 5919, pp. 1274
1. U.S. BUDGET

# Amid the Gloom, Researchers Prepare for a Boom in Funding

1. Jeffrey Mervis*
1. With reporting by Eliot Marshall.

Your retirement account may be taking a beating, but if you have a grant application pending at a U.S. science agency, there's an upside to the global financial meltdown: Your chances of being funded have never been better. And if your application isn't already in the pipeline, don't despair. The competition for funds should be eased significantly next year.

This improved research outlook comes partly from the $787 billion stimulus package signed by President Barack Obama last month. It provides an additional$22.5 billion across several research agencies, including $10.4 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. Agency officials are under orders from White House budget czar Peter Orszag to spend the money “quickly and wisely,” and they are now working out procedures for getting it spent. The stimulus money is in addition to agencies' regular budgets, which were expected to be finalized this week as part of a$410 billion spending bill covering the rest of the 2009 fiscal year. The version passed 25 February by the House of Representatives contains a 20% boost for DOE's Office of Science, a major supporter of basic research across the physical sciences, and a 6.5% increase for NSF. NIH would receive a $937 million bump. (The Senate was due to act after Science went to press.) And that's not all. On 26 February, Obama delivered a 2010 budget request to Congress that would start to make good on his campaign promise to double the federal investment in basic research over the next 10 years. If Congress goes along, the$3.6 trillion proposal would boost NIH spending on cancer research by 21% during the fiscal year that begins on 1 October, raise NSF's budget by another 8.5%, to $7.04 billion, and give “similarly large increases” for DOE science and for research at the National Institute of Standards and Technology (NIST). A detailed breakdown for each agency won't be available until April. That trifecta of spending will mean tremendous opportunities for the U.S. research community. It puts NSF, DOE science, and NIST back on the doubling track that President George W. Bush proposed in 2006 and that is enshrined in a 2007 law filled with good intentions but no cash. “I'm very pleased that Congress and the White House have provided us with an unprecedented level of resources for this year,” gushed NSF Director Arden Bement at a briefing last week for the National Science Board, NSF's oversight body. Adding the stimulus package to NSF's regular appropriation, Bement explained, gives NSF a total budget of$9.5 billion in FY 2009. That's nearly 60% more than it received in 2008. And Obama's first budget request, Bement adds, “contains a very good number for NSF.”

It's also welcome news for biomedical researchers, who have felt the squeeze from an NIH budget that has remained essentially flat since 2004 after a 5-year doubling. “We are very happy that we are getting out of the doldrums,” says cell biologist Richard Marchase, president of the Federation of American Societies for Experimental Biology. Marchase expects that many high-scoring grant proposals sidelined in 2008 will now get approved, adding that the 2010 budget framework holds out hope for steady and sustained annual growth of roughly 7%.

For research agencies, the immediate challenge is how to spend their unexpected wealth from the stimulus package. The Obama Administration has assured Congress and the public that these funds will go out the door as quickly as possible without lowering standards, and that their impact on the economy—in particular, on the number of jobs created—will be monitored closely. Although each agency is setting out its own guidelines for how to spend the money and what information grantees will need to provide, the rules drawn up by NSF and NIH appear typical.

Both agencies plan to dip into the existing pool of applicants for the bulk of the new awards. For NIH, that includes proposals that didn't make the merit review cutoff in 2008; for NSF, it means proposals submitted since last fall that seek funding in the current fiscal year. These peer-reviewed projects could be called shovel-ready science. To avoid commitments beyond 2010, most NSF awards will be in the form of standard grants, in which the full 3-year total is funded up front.

NIH will give top priority to its bread-and-butter R01 grants, but most other categories are also eligible for stimulus money. Unlike at NSF, NIH is also preparing a solicitation for 2-year challenge grants, using up to $200 million from an$800 million pot given to the NIH director. The agency hopes to fund challenge grants in 14 areas, with an application deadline of 27 April.

Agency officials promise that speed won't compromise the quality of the reviews. NSF's Division of Mathematical Sciences, for example, is in the midst of running some 50 panels that are sifting through roughly 2500 proposals submitted last fall. “Our instructions to the panels haven't changed,” says division director Peter Marsh. “We're still looking for the best science.” But Marsh, whose 2008 budget was $212 million, expects that the additional funds (which haven't yet been allocated among NSF's six research directorates) will result in a healthy rise over last year's 31% success rate. That could be good news for mathematician Andrew Belmonte of Pennsylvania State University, University Park. Belmonte is awaiting word on a grant proposal he submitted in November to NSF's applied math program after taking an unsuccessful shot the previous year. His work on the transformation of materials during fluid flows is expensive—costing 10 times the normal math project, he estimates—because it requires a well-equipped wet lab. And if funded, he'll need to hire a postdoc and several undergraduates, a priority for the stimulus package. Belmonte says he's ready for any additional paperwork: “It's the government's money, and if they want more reporting, I'm happy to do it.” The emphasis on funding what's already in the pipeline obviously favors those lucky enough to have applied in the current funding cycle. But Marsh says that because NSF will be funding so many more grants this year, there should be fewer applications competing for funds next year. At the same time, the recipients of this year's spending spree are likely to show up 3 years hence in the competition for renewals. “So the stimulus will have both a ripple and an echo effect,” he says. 2. U.S. BUDGET # Budgets in Brief 1. News Staff ## News Staff Some highlights from the stimulus package, the 2009 budget before Congress, and President Barack Obama's 2010 request: ENERGY: DOE's Office of Science gets$4.77 billion in 2009, up 19%. The ITER project in France gets $124 million,$90 million less than requested. DOE's national labs and other facilities divide the $1.6 billion stimulus money, and the department gets$400 million to create the Advanced Research Projects Agency-Energy. No 2010 research numbers are available.

NSF: The $394 million boost for 2009, to$6.49 billion, is on top of $3 billion in stimulus money for research, infrastructure, and education. The agency's six research directorates would grow to$5.18 billion, and education to $845 million. The$11 million Robert Noyce Scholarship program for prospective teachers gets its second straight $40 million bump, and a program to help researchers in states that struggle to win NSF grants would grow by$20 million, to $133 million. NIH: The agency's 3.2% increase this year, to$30.3 billion, is supplemented by $10.4 billion from the stimulus package. The 2010 budget mentions only “over$6 billion within the National Institutes of Health to support cancer research.” The National Cancer Institute's budget this year is $5 billion. NASA: The space agency will receive$17.8 billion in 2009, $380 million more than in 2008. A$200 million drop in the agency's science programs, to $4.5 billion, is remedied in its$1 billion stimulus package. The $18.7 billion request for 2010 bolsters earth sciences and robotic probes to visit other planets. The new launcher to send humans to the moon wins funding, but the White House may review the target of a 2020 lunar landing. NIST: A$63 million jump in 2009, to $819 million, maintains the$65 million Technology Innovation Program. Obama would raise that to $70 million in 2010. The stimulus funds add$220 million to a $470 million research budget,$180 million to a $172 million lab-construction program, and provide$180 million for a facilities grants competition.

3. EARTH MONITORING

# Loss of Carbon Observatory Highlights Gaps in Data

1. Eli Kintisch

With rising temperatures altering a variety of ecological and weather systems on Earth, the current patchwork of sensors can't answer all the questions that scientists are asking. Land-based sensors have provided a conclusive picture of rising CO2 levels worldwide, for example, but researchers don't fully understand where all the carbon that humans and natural sources are pouring into the atmosphere ends up. How much is being absorbed and where?

NASA's Orbiting Carbon Observatory (OCO) was supposed to provide some answers about the nature of carbon sinks on land and in the oceans. But on 24 February, the rocket carrying the $278 million satellite crashed shortly after takeoff, the victim of a failure related to the nose cone. “Bang—it's gone. It was absolutely terrible,” says Pieter Tans, an atmospheric scientist at the National Oceanic and Atmospheric Administration's laboratory in Boulder, Colorado. Tans was hoping to improve his four-dimensional maps of CO2 flows—useful to close the global carbon budget—by using OCO's unique ability to detect tiny carbon fluxes. The loss of OCO is part of a bigger problem of data gaps related to climate. Some gaps, like OCO, rest upon relatively new measurements. In other cases, climate scientists worry that satellites well beyond their operational lifetimes will fail before their replacements are in orbit. Sometimes, as with the monitoring of forest biomass, polar ice, and sea levels, the records go back for decades. “The need for a systematic and comprehensive approach to collecting climate observations has taken on new urgency,” concluded a panel from the National Academies in a report issued last week. No single agency or person has the overall authority for the multibillion-dollar challenge of observing Earth's climate, it noted. Researchers are hopeful that a$400 million boost for space-based Earth sensors included in the stimulus package (see p. 1274) will make a difference. But a 2007 academy panel estimated it would cost $6 billion through 2020 to fix a system it said was “at risk of collapse” in 2005. This backlog exists at a time when better observations are urgently needed, especially at Earth's poles, say researchers. The 2007 report from the Intergovernmental Panel on Climate Change identified basic observations to better understand ice sheet physics as a key requirement for more reliable predictions of sea-level rise. POLENET, an effort to install seismic monitors all over the Antarctic continent, will provide measurements of the temperature of the bedrock on which ice sheets sit, a major factor that affects the speed at which ice slides toward the coast. But bad weather, compounded by rising fuel costs that forced the National Science Foundation to trim spending on polar research, allowed scientists to place only one new station in the field out of 16 originally planned; most operational stations are clustered near McMurdo Station, the main U.S. base on the continent. The speed at which ice sheets are declining is also governed by the interaction of glaciers with water at the ocean shore. A robotic submarine has recently collected data on the Pine Island Glacier, the fastest moving glacier in Antarctica. Scientists would like a long-term picture of shifting dynamics at the edge of the ice sheet as well as this snapshot. But a project to set up stations on the glacier's edge to monitor the region below the ice was delayed by logistical hurdles: The ice was too rough to allow planes to land on skis, and arranging helicopter facilities wouldn't be possible until the 2011–12 season. “Logistics in Antarctica proceed slowly, [and] we only get to make progress for about 3 months in any year,” says NASA's Robert Bindschadler. Modelers say more data could also clarify the role of atmospheric aerosols in global warming. Aerosols help form clouds, which can both warm and cool the atmosphere. The main sensor on the$14 billion National Polar-orbiting Operational Environmental Satellite System (NPOESS), scheduled for a 2010 launch, will provide aerosol data that will be inferior to data gathered by a sensor on an existing NASA craft called Terra, already 2 years beyond its operational life. In 2007, the National Research Council's (NRC's) decadal study for the field suggested that NASA launch a replacement between 2013 and 2016, and Michael Freilich, NASA's earth sciences chief, says his staff has begun to scope out the mission's requirements. “We're committed to the decadal,” he says.

Even Earth-observing mainstays like Landsat are at risk. Landsat 5 and Landsat 7 are both well beyond their nominal design life, regularly missing swaths of images during malfunctions. Scientists who rely on the crafts to study forests' contributions to the carbon cycle say a gap in land imagery is virtually assured because the next satellite in the series, the Landsat Data Continuity Mission, won't be launched before 2012. “We may lose one or both of these satellites before then,” says Compton Tucker of the White House Climate Change Science Program (CCSP).

Another problem area is ocean color, a variable that scientists use to measure photosynthesis in the ocean—a giant and poorly understood contributor to the global carbon cycle. NASA's orbiting SeaWiFS sensor has been steadily recording data since 1997. But the instrument is 7 years beyond its design life, and the sensor meant to measure ocean color on the first NPOESS mission will likely be unable to provide sufficient data because of technical problems (Science, 15 February 2008, p. 886). That has spurred some managers to push for a “gap filler” mission sooner, says Tucker. But Freilich is hesitant to commit.

CCSP is leading an interagency effort to finalize a plan that would for the first time provide a set of government priorities for Earth observations, says Tucker. Scientists say that's the first step needed to fill the gaps.

In some cases, international partners might help, he says. Japan's recently launched Greenhouse Gases Observing Satellite, for example, detects CO2 starting at the ground level, like OCO. Although its resolution is lower, it well complements an existing NASA satellite that only picks up CO2 readings at an altitude of 5 km. “We're going to see what we can continue to do with our Japanese partners to fill the gap,” says NASA's Stacey Boland.

The effort to better coordinate climate-related monitoring will no doubt be strengthened by the NRC report released last week and presented on 26 February to presidential science adviser John Holdren. Holdren has extolled the virtues of “observation and scientific study of the condition of our home planet's land, vegetation, oceans, and atmosphere.” Climate researchers are hoping those words, delivered at his confirmation hearing to be director of the White House Office of Science and Technology Policy, will soon translate into support for filling the holes in the nation's current Earth-monitoring system.

4. EDUCATION

# Study Questions Value of School Software for Students

1. Jeffrey Mervis

U.S. students using educational software do no better learning primary school math and first-year algebra than their counterparts who follow a traditional curriculum. That's the conclusion of a new federally funded study that is loaded with caveats about what it means for students, educators, and the companies that make the software.

The $14.5 million study, funded by the U.S. Department of Education and conducted by Mathematica Policy Research Inc. in Princeton, New Jersey, was designed to find out whether students are benefiting from the growing use of educational software. Preliminary results released in April 2007 suggested that the answer, for first- and fourth-grade students in reading and for sixth-and ninth-grade students in math, was no. But those results were widely criticized by educators and software makers for lumping together the outcomes from many different products and for testing their impact on student achievement too early, in the first year the teacher had used the material. The second-year results, posted 17 February, address those concerns by reporting results for individual software packages and by testing a new cohort of students whose teachers already had a year of using the software under their belts. What it found is that none of the four math products tested produced a statistically significant difference in achievement between control and treatment groups over the 2-year study, which began in the fall of 2004. (One of the six reading packages registered a meaningful jump in test scores by fourth graders.) Students of teachers using Cognitive Tutor, a computer-based curriculum for algebra I students developed by Carnegie Mellon University researchers (Science, 2 January, p. 64), for a second year showed a meaningful improvement in test scores. But the software had no overall impact on student achievement when data from both student cohorts were combined. Education research is rarely definitive, however, and this study is no exception. For one thing, the students and teachers were not randomly assigned to the products, explains Mathematica's Mark Dynarski. That rules out comparing one software package with another. It also means that the results shouldn't be generalized to different student populations. “Everyone began at different starting points. So we can only say how this product works at a particular school,” he says. Another confounding factor is that researchers didn't have enough money to continue observing teachers in their classrooms or surveying them about how they used the product, something that was done in the first cohort. That makes it difficult, for example, to interpret large fluctuations from one year to the next in the amount of time spent using particular software. Does an increase for algebra I students, for example, mean that teachers found the product more helpful once they became comfortable with it? That seems a logical explanation to Cognitive Tutor's Steve Ritter, who points to his company's emphasis on supporting teachers. “For us, usage went up, and second-year teachers achieved statistically significant improvements,” Ritter says. “That's worth talking about.” In contrast, does a decline in usage mean that teachers found that it wasn't right for their students, who are disproportionately from large, poor, and urban schools? Or was it simply due to a dearth of computer facilities, inadequate tech support, or scheduling problems? And what effect does an extremely mobile teacher corps—only one in four taught the same grade at the same school for both years—have on student outcomes? Dynarski cautions against overinterpretation of the data. He notes that Congress wanted to know whether education software improves student learning, not whether individual products can help certain populations taking certain subjects. “There might be a lot of positive things going on in the classroom, such as greater fluency with computers,” he says. “But it's not leading to the kind of results that people want, which are better test scores.” 5. EPIDEMIOLOGY # Author of Iraqi Deaths Study Sanctioned 1. John Bohannon The lead author of a controversial study that concluded that the U.S.-led invasion of Iraq in 2003 caused a high number of violent deaths—more than 600,000 Iraqis—has been sanctioned for a lapse in ethics. The Bloomberg School of Public Health at Johns Hopkins University in Baltimore, Maryland, found last week that epidemiologist Gilbert Burnham had committed “violations of the Bloomberg School's policies regarding human subjects research” by failing to fully protect the confidentiality of interviewees. It ordered a 5-year suspension of “Burnham's privileges to serve as a principal investigator on projects involving human subjects research,” the school announced in a press release. “The [Bloomberg] School has asked me to make no comments at the present,” Burnham said to Science. A university investigation discovered that the full names of Iraqi people interviewed appeared on some of the 1800 data-collection forms filled out during the house-to-house survey designed by Burnham. The university gave Burnham approval for the study with the understanding that no unique identifiers would be recorded. The survey was carried out by a team of Iraqi researchers who met with Burnham in Jordan to receive instructions and hand over data (Science, 20 October 2006, p. 396). According to a Hopkins press release, the investigation found “no evidence that the violations caused harm to any individuals involved in the study” because the data sheets were “never out of the possession of the research team.” As principal investigator, Burnham was held responsible. “The punishment is very severe, and making it public is unusual,” says Gary King, a statistician at Harvard University. Les Roberts, a co-author of the 2006 Lancet study who is now at Columbia University, acknowledged in an e-mail exchange with Science that a “serious” breach of confidentiality occurred in the survey forms but said that Burnham was not initially aware of it. “When the forms arrived in Jordan all filled out, [Burnham] asked and was assured that these words at the top of the forms were not complete names or unique identifiers,” Roberts wrote, and “most are written in Arabic, which [Burnham] does not read.” The debate is likely to continue. The sanction “has nothing to do with the scientific merits or defects of the study or its results,” notes Debarati Guha-Sapir, an epidemiologist at the World Health Organization's Collaborating Centre for Research on the Epidemiology of Disasters in Brussels. Statistician Seppo Laaksonen of the University of Helsinki notes that Hopkins didn't evaluate the study's sampling methodology or statistical approach, which he cares about more than the punishment. But Roberts dismissed criticism of the study. In an e-mail, he wrote: “There is a way to verify [our] findings: replicate. The posh academics who spend their lives in offices are offended by that, but in this arena, rigor is more about ground activities than software manipulations.” According to the Hopkins press release, “an erratum will be submitted to The Lancet” concerning the unauthorized recording of names. 6. SCIENCE AND BUSINESS # India Allows Government Scientists to Own Companies 1. Pallava Bagla NEW DELHI—In 2001, Swami Manohar and three colleagues at the Indian Institute of Science (IISc) in Bangalore invented the Simputer, a simple and cheap hand-held computer. But as civil servants, the computer scientists by law could not commercialize their invention. “I had no choice but to resign,” says Manohar, who is now chief of intellectual property and strategy at Geodesic Limited, a telecom firm that bought the company Manohar and his colleagues founded after leaving IISc in 2001. Indian scientists will no longer be forced to make such a stark choice. On 24 February, the Department of Scientific and Industrial Research issued regulations that permit researchers at government-funded institutions to hold equity stakes in scientific enterprises and spinoff companies. The “historic decision,” says Science Minister Kapil Sibal, will “unleash the latent entrepreneurial potential of Indian scientists.” The policy shift is expected to have a profound impact in India: Some 400,000 scientists, about three-quarters of the scientific work force, are employed at public institutions. By bringing India in line with the United States and other Western nations, the new rules should create an attractive environment for talented expatriate scholars to return to India, says Samir Brahmachari, director general of the Council of Scientific and Industrial Research in New Delhi who helped shepherd the regulation through 18 ministries over 18 months. The new rules also permit research institutes to hold equity stakes in commercial enterprises. To facilitate this process, the government will encourage the lateral mobility of researchers between institutes and industry. “Cross-fertilization between the academics and industry is very much necessary,” says Sibal. Although the regulations came too late for Manohar to keep his post at IISc, he applauds what he sees as a long-overdue change. “Scientists need the freedom to flourish, and now they have gotten it,” he says. 7. UNION OF CONCERNED SCIENTISTS # From Protest to Power: An Advocacy Group Turns 40 1. Dan Charles On 4 March 1969, some of the most prominent scientists at the Massachusetts Institute of Technology (MIT) issued a declaration of political dissent and scientific self-criticism. Stirred into action by student protests against the war in Vietnam, the professors convened a campuswide meeting and declared that the “misuse of scientific and technical knowledge presents a major threat to the existence of mankind.” The statement bore the name of a previously unknown organization: the Union of Concerned Scientists (UCS). Forty years later, UCS has 80,000 members and a staff of 130 working in four cities. The organization campaigns to shrink nuclear arsenals, fight global warming, and reduce agriculture's harm to the environment. For 8 years, it was a thorn in the side of the Bush Administration, criticizing White House policies and zinging its appointees for “politicizing science.” That aggressive approach raised its visibility and helped triple its budget this decade, to almost$20 million.

But just as some MIT faculty members of yesteryear ignored the teach-ins and went ahead with their normal duties, some scientists believe that UCS cannot claim to be the conscience of the scientific community. “Many of its statements and conclusions … were perverse oversimplifications of complex issues,” says John Marburger III, science adviser to President George W. Bush and a frequent target of UCS attacks. “I think it's hard, maybe impossible, for an advocacy organization to be entirely science-based.”

Although its launch attracted wide attention, UCS almost succumbed to the status quo soon afterward. “Many of us were quite disturbed that we'd spent several months playing hooky,” recalls physicist Kurt Gottfried, one of the primary organizers of the MIT event. As students and faculty members returned to their normal routines, Gottfried says, “this hotbed of activity sort of faded away.”

That it did not is due largely to MIT physicist Henry Kendall, a respected scientist—he won the Nobel Prize in physics in 1990—and heir to an industrial fortune. Kendall, who died in 1999, once said he learned a valuable lesson about advocacy during the 1960s as a member of the JASON group that provided confidential technical advice to the Department of Defense. “Being inside [the government] was not an effective way to change national policy,” he told the American Institute of Physics in 1986, because the government “would not take advice it didn't like.”

“For the first couple of years, Henry Kendall pulled some money out of his own pocket” to cover expenses, recalls Daniel Ford, a young economist who became UCS's first executive director. Kendall and Ford teamed up on studies that were critical of reactor safety, bringing UCS to national prominence. Perhaps their biggest media coup took place in 1976, when Ford arranged for Robert Pollard, a safety official at the U.S. Nuclear Regulatory Commission (NRC), to announce on 60 Minutes that he was resigning in protest and going to work for UCS. “In the 1970s, the organization was 90% focused on nuclear energy,” says Ford.

Yet nuclear power was not a simple issue for the organization. “A lot of physicists were strong supporters of nuclear power; they'd been involved in its invention,” says Gottfried, who is stepping down this year as chair of UCS. He will be replaced by James McCarthy, past president of AAAS (which publishes Science). By calling for tighter safety regulations while insisting that it was not “antinuclear,” UCS walked a fine line. It tried to strike a similar balance when criticizing other technologies, such as missile defense or genetically engineered crops, that some scientists strongly support.

Its opponents accuse it of giving mere lip service to scientific objectivity. George Keyworth, science adviser to President Ronald Reagan, says that “their political agenda has been manifest for their entire career. … I just never take them very seriously.” But John Ahearne, a former NRC chair who now directs Sigma Xi's ethics programs, says UCS has been “a good counterweight to industry.”

Along with successfully pushing for tighter regulation of nuclear power in the 1970s and 1980s, UCS played a prominent role in opposing deployment of antisatellite weapons and missile defenses during the 1980s. In recent years, its attention has shifted toward finding ways to reduce greenhouse gas emissions.

As it marks its 40th birthday, UCS has never been closer to the center of political power. Issues on which UCS has campaigned—energy efficiency, clean cars, and big cuts in greenhouse gas emissions—have been embraced by the Obama Administration. Yet political prosperity may also leave it weakened. UCS was a child of the 1960s antiwar movement, and the organization has prospered during periods when it strongly opposed government policies. In contrast, membership and donations fell during the 1990s when President Bill Clinton was in office.

Kevin Knobloch, a former journalist and congressional aide who has been president of UCS since 2003, sees no signs of waning support now that another Democrat occupies the White House. “We have this tremendous window of opportunity to win adoption of farsighted policies,” he says. “This is our moment.”

8. SCIENTIFIC MISCONDUCT

# Retractions Put Spotlight on China's Part-Time Professor System

1. Hao Xin

BEIJING—Traditional Chinese medicine (TCM) is a slippery subject: In herbal concoctions, it's often difficult to discern active compounds from fillers and contaminants. Sorting wheat from chaff is all the harder when data themselves are concocted. In the latest scandal to grip Chinese academia, Zhejiang University (Zheda) in Hangzhou last November fired an associate professor in a recently established TCM research lab after finding him guilty of scientific misconduct, the university said in a statement. Three journal articles have been retracted so far out of eight that the university has charged contain plagiarized or fabricated data. Since then, more questionable papers have come to light.

The case, which came to national attention in February, could have widespread fallout. Some observers criticize Zheda for firing the junior researcher, He Haibo, before the investigation is complete. “I don't think Zhejiang University handled the case properly,” says Zhang Haixia, a professor at Peking (Beijing) University who led a misconduct inquiry a few years back. “What about the responsibility of the corresponding author?” asks Zhang. The handling of the case has sparked an outcry in Chinese media and on the Internet focused on the senior scientist who established the TCM lab and, more broadly, on a system that critics say gives power but little accountability to academicians.

In 2004, Zheda recruited Li Lianda, who was elected the previous year as an academician of the Chinese Academy of Engineering, to be part-time dean of its College of Pharmaceutical Sciences. Li, an expert on TCM modernization, also set up a lab at Zheda and has supported it with a 5-year, $430,000 grant from the Ministry of Science and Technology to study TCM drugs for cardiovascular disease. But Li spends most of his time in his main lab in Beijing, where official meetings require his presence, as he has explained to Chinese media. He Haibo joined the Hangzhou lab as its first postdoc in July 2006 after receiving a Ph.D. from China Pharmaceutical University in Nanjing 1 month earlier. Judging from his publication record, He was a dynamo. Between April and November 2007, he submitted eight manuscripts on which he was first author to international journals; all appeared in print in 2007 or 2008. According to a Chinese reporter who tracked down a former grad student in the lab, He was well-liked because he helped students fulfill the college's requirement that Ph.D. candidates publish one first-authored and at least a second paper in a journal in the Science Citation Index database. (Since 2006, Zheda has published more indexed papers than has any other Chinese university.) As a reward for He's remarkable productivity, Zheda hired him as an associate professor last July. Things quickly unraveled for He. In October, an editor at the International Journal of Cardiology came across a paper in Phytotherapy Research with figures and tables similar to those in a manuscript IJC was about to publish. The manuscript reported a study on two molecules in a signaling pathway involved in irregular heart palpitations caused by inflamed cardio muscles in rats, whereas the published paper is about the same pathway during a heart attack. On 11 October, the editor contacted the manuscript's senior author, pharmacologist Dai De-Zai of China Pharmaceutical University, for an explanation, according to Dai's blog. Dai—who was He's Ph.D. adviser—wrote in his blog that he was surprised to see that his former student had, as Dai alleged, plagiarized graphs, tables, and texts from his lab. In his blog, Dai says he got in touch with the executive vice dean of Zheda's pharmacology school and contacted the journal that published He's paper. Contacted by e-mail, Dai declined to provide further comment. According to a Zheda statement received by Science on 13 February, the college confronted He on 16 October; administrators asked him to retract the fraudulent paper, reflect on his mistakes, and apologize to his former adviser. Ten days later, according to the statement, He submitted a self-criticism essay in which he cleared Li of any knowledge of his wrongdoing and then left the college. To protect privacy, the university says, it is not making He's essay public. Zheda President Yang Wei told Chinese media that He, in his self-criticism, said he committed misconduct because he was desperate for a faculty position, which would enable him to remain at Zheda and earn a living. After ascertaining that He had e-mailed the requisite retraction to Phytotherapy Research, Zheda says, the university fired him on 13 November. He Haibo could not be reached by telephone and did not respond to e-mail messages. The story did not end there. Li and his lab came under fire on the Internet. On 23 October, an anonymous posting on New Threads (http://www.xysforum.org/) alleged that Li's Zheda lab published duplicate articles in Pharmacological Research and Naunyn-Schmiedeberg's Archives of Pharmacology. Apparently, before he disappeared, He e-mailed both journals to retract the papers, giving “duplicate publication” as the reason, according to e-mails provided to Science. The journal editors alerted Zheda to the retractions. Three days later, another anonymous posting on New Threads alleged that the duplicate papers, along with a third in the Journal of Ethnopharmacology, contain fabricated data. The postings caught the eye of Zhu Guo-Guang, a TCM practitioner in Oulu, Finland. Zhu downloaded the three papers for a closer look and says he found fabricated data “written in black and white.” To Zhu, the alleged misconduct is a big blow to TCM research. “Fake research worsens the reputation of traditional Chinese medicine,” he says. As a vice chair of the Pan European Federation of TCM Societies, Zhu says he felt a responsibility to report his findings to the three journals and to Zheda. Zhu assembled the evidence and mailed hard copies to the journal editors and to Yang Wei in mid-November. As Chinese netizens turned up more questionable journal articles from Li's Hangzhou lab, Zhu set up a blog to post his analysis and provide updates. Zhu's allegations grabbed national attention after they were reported in a 3 February article in the Chinese newspaper 21st Century Business Herald. Zheda issued a statement the next day, laying the blame on He and stating that Li's name had been given as a coauthor of the fraudulent papers without Li's authorization. Yang Wei says he is now personally sending out retraction requests. Meanwhile, more anonymous postings appeared on New Threads; Zheda now says that 19 publications allegedly have problems, including papers from Li's lab at the China Academy of Traditional Chinese Medicine in Beijing and by other members of the Hangzhou lab that did not include He as an author. A team led by Harold Garner of University of Texas Southwestern Medical Center in Dallas has analyzed several papers and added them to the Déjà vu database of extremely similar publications (spore.swmed.edu/dejavu/duplicate/74404/); also see entries 75181 through 75184. “We try to provide the most unbiased data possible which an appropriate body, such as an editorial board or a university ethics board, can use to make their own evaluations,” says Garner. The scandal has prompted fresh debate about moonlighting by academicians. Many, like Li, hold two or more positions in addition to their main job, often in different cities as part-time deans or lab chiefs. Universities eagerly recruit academicians because these elite scientists advise various levels of government on research and funding priorities, review grants, and evaluate achievements. At least two-thirds of Zheda's 15 science and engineering deans are such part-timers. Li acknowledges that he visits Hangzhou only a half-dozen times a year, spends 3 to 5 days in the college, and has only 1 day for his five or six grad students in the TCM lab. Wu Limao, who received his Ph.D. from Li's Beijing lab in 2004, ran the lab's day-to-day operations until August 2008 when Wu left to spend 2 years as a visiting researcher at Yale University. Li says that, as a victim of He's fraud, he has become a subject of investigation. In the meantime, Zheda has expanded its probe and has asked Wu to return to Hangzhou to answer some questions. Wu is corresponding author on eight papers that Zheda has found to be fraudulent and first or corresponding author of other papers called into question. According to a preliminary investigation report received by Science on 25 February, the university faults Wu for not reviewing the papers before publication. Wu declined to comment for this article. The saga is inflicting collateral damage on researchers who have collaborated with Li's labs. Phytotherapy Research Editor-in-Chief Elizabeth Williamson says the journal is preparing to retract more articles from Li's labs and has held up all manuscripts with authors whose names appear on suspicious papers, pending Zheda's further investigation. “I don't want to penalize honest people, but it's getting difficult to know who to trust these days,” she says. 9. # ScienceInsider: From the Science Policy Blog From his “State of the Union”-like speech to Congress last week, to his new budget, to his seemingly never-ending selection of advisers and Cabinet members, the U.S. president commanded the science policy spotlight over the past week. Here are some highlights from ScienceInsider: The big news has been Obama's 2010 budget. Mirroring the enthusiasm Congress has shown for science, the president asked for big boosts for the National Science Foundation, NASA, and the Environmental Protection Agency. The EPA budget includes a$19 million increase for the agency to create a greenhouse gas emission inventory—an important step in preparing the United States for a cap-and-trade system. Also seeing green is autism research. Obama requested \$211 million as part of the Department of Health and Human Services budget for research into causes and new treatments. But the president isn't showing everyone the money. Fulfilling a campaign promise to oppose plans to build a storage vault for nuclear waste at Yucca Mountain in Nevada, Obama's budget scales back funding for the project.

To many scientists, the president's budget gets a lot right. But in Obama's speech to Congress a couple of days earlier, he got at least one thing wrong. The president promised to ensure that “by 2020, America will once again have the highest proportion of college graduates in the world.” The United States is already there, however. The country leads the world, with roughly 30% of its adult population holding 4-year college degrees.

In non-U.S. news, U.K. Prime Minister Gordon Brown assured scientists that U.K. research will not be “a victim of the recession.” Brown's address highlighted his vision for science as a driver of the United Kingdom's economic future, a desire that has triggered much debate within the scientific community. Meanwhile in Japan, the incoming president of the University of Tokyo dismissed the importance of school rankings. “University research should be not to satisfy individual egos but to enrich the entire society and humanity,” he said.

For the full postings and more, go to blogs.sciencemag.org/scienceinsider.

10. BIOSECURITY

# The Danger Within

1. Yudhijit Bhattacharjee

The suicide of Army researcher Bruce Ivins will inevitably mean changes in biosecurity policy, but government and academic scientists are divided on how stringent the new procedures should be.

The suicide of Army researcher Bruce Ivins will inevitably mean changes in biosecurity policy, but government and academic scientists are divided on how stringent the new procedures should be

On the morning of 2 August 2008, 3 days after Army researcher Bruce Ivins committed suicide, the Biodefense Policy Coordination Committee assembled in a conference room of the Eisenhower Executive Office Building, a stone's throw from the White House. The mood around the table was somber when Robert Kadlec, committee chair and member of the Homeland Security Council, spelled out the meeting's agenda, according to one of the participants. If Ivins had indeed carried out the 2001 anthrax letter attacks, as the Federal Bureau of Investigation (FBI) claimed, what could be done to prevent another bioterrorist act by an insider at a government or academic lab?

The insider threat has never loomed so large in policy discussions on bioterrorism. The massive expansion of biodefense research after the anthrax mailings in 2001, which killed five people and sickened 17, was largely predicated on the concern that a terrorist group could launch an attack using biological weapons it had developed or stolen from a military facility. The FBI's implication of Ivins in the mailings—although still questioned and untested in a court of law—sparked the unsettling realization among researchers and biosecurity experts that “the insider threat is a lot bigger problem than we ever thought,” says a former scientist at the U.S. Army Medical Research Institute of Infectious Diseases in Frederick, Maryland, where Ivins worked.

In recent months, that concern has echoed through the halls of the U.S. government. Soon after the 2 August meeting, the White House asked the National Science Advisory Board for Biosecurity (NSABB)—a panel set up in 2005 to address the potential security risks from life sciences research—to consider how best to ensure that academic and industry researchers working with select agents are trustworthy. It is expected to come up with recommendations next month. Independently, an interagency working group, co-chaired by the secretaries of Defense and Health and Human Services, is looking at ways to strengthen lab biosecurity.

Some change in biosecurity procedures is inevitable post-Ivins, experts agree. But how stringent any new steps should be is being hotly debated. Since NSABB got its charge from the White House, board members have been nervously discussing the costs and benefits of measures such as tougher background investigations, psychological screenings, drug tests, and credit checks. Other possibilities include video surveillance of labs and a two-person rule prohibiting researchers from working alone.

Some of those measures are standard practice at the Department of Defense (DOD), the Department of Energy (DOE), and other agencies where scientists work with nuclear, chemical, or biological agents. But some academics worry that extending such measures to academia would undermine the open culture of life sciences research and impede biodefense research. A better approach, some say, would be to trust lab managers to keep a watchful eye on employees for signs of troubling behavior. But even advocates of self-policing concede that that approach may not allay fears of the insider threat.

## Trying to foil an insider attack

The U.S. government already vets researchers working with select agents through rules that were put in place a year after the anthrax attacks. Under those rules, institutions and individuals who wish to work with any of 80 biological entities, including the Ebola virus and botulinum toxin, must go through a Security Risk Assessment (SRA).

The SRA process involves a basic FBI background check. Disqualifying factors include a criminal history, a substance-abuse or mental health problem, a dishonorable discharge from the military, or being a citizen of a country that the United States deems a sponsor of terrorism. The FBI runs an applicant's fingerprints through criminal and terrorist databases but generally doesn't investigate whether an applicant has honestly answered questions about substance abuse and mental illness. An SRA can take up to 45 days and is valid for 5 years.

Although many scientists complain that the SRA is onerous and impedes research, experts point out it is a fairly limited form of vetting. “The SRA catches the people who are dirty already,” FBI's Kristine Beardsley said at a 12 December 2008 NSABB meeting during which the board discussed personnel reliability. By contrast, a clearance looks at an individual's entire background—not just law enforcement records.

Personnel reliability procedures at federal agencies are already much stricter than those prescribed by the select agent rules. At DOD, employees whose work involves access to select agents and toxins have to undergo a secret-level Personnel Security Investigation that involves, among other things, credit checks and interviews with friends and acquaintances. (Like other defense researchers, Ivins had this clearance.) Once admitted into the lab, employees are monitored through drug tests and periodic medical evaluations. Although there is no formal psychological testing, those who display mentally unstable behavior—such as threatening colleagues—can be removed from the lab. The monitoring measures have been formally in place since 2004.

Since Ivins was implicated, DOD has considered making the system even more stringent. An agency task force has recommended background checks equivalent to those required for a “top secret” clearance. The prospect worries even those used to DOD's security-conscious culture. “They'll hunt down your ex-wife, your landlord, anybody who knows you,” Kenneth Cole, a senior Pentagon official, said at the 12 December meeting.

DOE's safeguards could also be a model for NSABB. At DOE's Lawrence Livermore National Laboratory (LLNL), where a security clearance is mandatory for most employees, the dozen or so researchers who work with select agents are further vetted through extensive conversations with past managers. Eric Gard, select agent manager at LLNL, says he has turned down individuals “who have trouble managing their temper, for instance,” assigning them instead to work with less hazardous pathogens in a biosafety level 1 lab. Once they are cleared to work with select agents, employees are given a psychological evaluation at least once a year. Occasional credit checks are done “to determine whether somebody might be tempted by their poor financial circumstances to do something desperate,” Gard explains.

These measures come at a cost: Academic researchers often suffer a rude shock when they come to work at Livermore, and the intense scrutiny there makes it difficult to attract and retain talent, Gard says.

## Peering into the mind

At the 12 December meeting, some NSABB members urged caution before embracing measures such as psychological screenings, which they worry would drive academic scientists away from biodefense projects, slowing vaccine development and other biomedical advances.

Richard Ebright, a chemist at Rutgers University, New Brunswick, doesn't buy the argument. A proponent of stronger regulations, Ebright points out that for all the grumbling, researchers entered the biodefense field in droves after the select-agent rules came into effect.

At the meeting, some board members wondered aloud whether psychological monitoring would yield too many false alarms without adding much security. Many scientists take “pride in their eccentricity,” said Stanley Lemon, a microbiologist at the University of Texas Medical Branch at Galveston.

Gerald Epstein, a biosecurity expert at the Center for Strategic and International Studies in Washington, D.C., agrees. He also questions whether the art of “peering inside people's psyches” to predict their future actions has been perfected yet. “If Ivins was the anthrax mailer, what could we have seen ahead of time in his behavior to stop him?” he asks.

In the summer of 2007, the Office of the Director of National Intelligence (ODNI) commissioned an exercise by scientists and security experts to answer a question along those lines. Group members were divided into red teams and blue teams, says ODNI's Lawrence Kerr, who coordinated the study. The red teams came up with three scenarios of bioterrorist attacks, while the blue teams identified points in the scenarios at which the offender could be captured and the attack foiled. “It was like reading a Michael Crichton book, except that the scenarios contained actual recipes for attacks,” says Kerr, a former academic microbiologist. (The study is classified.)

In one of the scenarios, a trained biologist plotted an attack from within a sophisticated lab. “The key judgment we arrived at was that if such an individual was truly intent on carrying out an attack, they would succeed,” Kerr says. “It scared the crap out of us.”

The exercise led Kerr and his colleagues to devise three categories of indicators that could help detect an insider threat: human, technological, and material. In all three, the study found, the lab's principal investigator, other researchers, and administrators were in the best position to spot red flags such as odd behavior or anomalous purchases of equipment and materials. The take-home message was that researchers need to be sensitized to security concerns. At the same time, institutions need to develop procedures for reporting potentially suspicious activities to higher-ups without fear of embarrassment or recrimination. Kerr is working with scientific societies including the National Academies and AAAS, the publisher of Science, to educate life scientists about indicators to watch for.

Even the idea of keeping an eye on colleagues makes some academics nervous. “The default position should be of mindful trust, not of distrust,” NSABB chair Dennis Kasper said at the meeting.

NSABB is likely to emphasize “local judgment” in its recommendations, says the board's Susan Ehrlich. But mere self-governance is likely to be a hard sell, board members admit. “We are in a classic double bind,” says Kasper. “If we don't implement a program that is very serious, we are going to get criticized. If we do and research shuts down, we'll get criticized.”

Kerr says once all the reviews are complete, the government is likely to prescribe a combination of grassroots vigilance and increased top-down oversight. But no matter what measures are taken, it may be impossible to guarantee that a insider attack won't happen. The reason, Kasper says, is “inherent imperfection in people.”

11. NIGEL FRANKS PROFILE

# Watching as Ants Go Marching--and Deciding--One by One

1. Virginia Morell

The "Ant Idea Man" challenges established views by using paint and radio tags to see each ant as an individual.

The “Ant Idea Man” challenges established views by using paint and radio tags to see each ant as an individual

Over the past decade, Franks's mix of experimentation and ant-watching has produced a steady stream of research, both controversial and groundbreaking, about how T. albipennis ants make decisions. Decision-making in groups, whether ant or human, is increasingly a hot research topic in everything from computer design to election committees to armies. In a special issue on the topic last month in the Philosophical Transactions of the Royal Society B, Franks and his team described how ants implement quick decisions in the best possible way even in a time of crisis, such as when their home has just been destroyed.

“Most people think that ants are stupid, but they're not,” says Franks. “They have very sophisticated behaviors, although they don't have language or theory of mind [the ability to understand what another individual is thinking]. And that's what intrigues me: this behavior that looks like thinking, yet we can reduce it to an algorithm.”

Franks is probably best known—some would say infamous—for showing what he called teaching in ants, as foragers led naïve nestmates on searches for food. It was the first time this talent was shown in a species other than humans. Franks's assertion outraged many, however, and prompted changes in the criteria for teaching as set out by animal cognition researchers. Franks's experimental work was “outstanding,” says bee cognition expert Lars Chittka of Queen Mary, University of London. “But what the ants are doing is not ‘teaching' in the human sense.”

Critics and supporters alike agree that Franks has a unique talent for asking questions and designing ingenious experiments. “You can tell which students have worked in Nigel's lab,” says his former mentor Bert Hölldobler, who was nevertheless dismissive of the teaching finding. “You can see his handwriting in just the way they ask questions.” Franks “always has a new angle,” adds Thomas Seeley, an insect sociobiologist at Cornell University. “In ant biology, he's known as the ‘Idea Man.’”

## Tracking ant opinion polls

Franks has been watching ants since he was a child in Yorkshire, when he kept a colony on top of his wardrobe closet. “Even then, I really wanted to understand how the individuals cooperated to make a functioning whole.” His interest took him to the famous ant lab of E. O. Wilson at Harvard University, where he worked with Hölldobler, now at Arizona State University in Tempe. Hölldobler and Michael Möglich had recently discovered a new ant communication signal in a Temnothorax species (Science, 13 December 1974, p. 1046), which they labeled “tandem calling.” When scouting for a new nest, individual ants led other scouts to the site using a mix of chemical signals and physical touching.

In the 1980s, on a research trip to Bulgaria, Franks encountered another Temnothorax species, and the proverbial light bulb went off: He'd found a good experimental species. Temnothorax colonies often live in cracks in rocks, so their habitat could be easily mimicked in a lab, and they tend to have between 40 and 400 individuals. “I could hold an entire colony in the palm of my hand,” Franks says.

He and his students collect colonies from rocks along the English Channel each year for the Ant Lab. The ants set up their colonies inside a cardboard layer sandwiched between two microscope slides, the whole colony about 3 to 4 millimeters high and set into a petri dish the size and shape of a CD jewel case. The lab's counters are stacked with the dishes, each one a world unto itself with a separate colony. In each dish, “I have everyone in their society,” Franks explains. That means, he argues, that he has a better understanding of their social behavior than do colleagues studying apes or birds in labs.

Most of Franks's experiments are based on the ants' readiness to move from one nest to another. He and his students have discovered things such as how the ants evaluate a new nest: They assess its interior size and shape by keeping track of how often they cross their own trails. The ants agree on which nest to move to via “opinion polls.” When enough scouts agree on a new nest, they stop leading other ants to it and switch to the faster method of simply carrying the remaining members of the colony there. “They basically vote with their feet,” says Franks.

In a current paper, Franks and his team investigate a problem faced by ants and many human decision-makers, too: how a speedy and so possibly poor decision can lead to an even greater disaster (think of Hurricane Katrina). In the ants' case, in a crisis, a handful of scouts might locate a poor but acceptable nest and begin to emigrate while some of their fellows continue the search. Without the usual number of scouts to lead the emigrants, that would slow the emigration and leave the colony dangerously exposed for a longer time. To speed things up, these scouts lead “some of their undecided scout colleagues” from the new nest to the old one. This technique apparently turns them into advocates for the new home, because they immediately switch their behavior from searching to emigrating, saving the colony.

## Ants as individuals

To understand how a colony makes decisions, Franks first separates it into its parts. He usually paints each ant by carefully wedging it headfirst into a piece of foam and touching its gaster with four hues of paint. It can take 10 hours—“a horrible job,” Franks says. But then the ants are easily identified as individuals. “An ant colony is an information-processing unit, a little brain,” says Seeley. “Nigel is dissecting it, then putting it back together again.” With his interdisciplinary team, Franks reduces the decisions of the ants to mathematical models, revealing how simple rules followed by individual ants produce the complexity of the colony.

Recently, Franks's team took another step in tracking ants, reporting online in this month's Behavioral Ecology and Sociobiology the first successful experiment with extremely small radio-frequency identification tags (RFID) glued to the backs of each ant. “It's a fabulous technical advance,” says Anna Dornhaus, a former postdoc in Franks's lab, now at the University of Arizona, Tucson. “It will make many things possible, such as monitoring the ants for 24 hours and finding out who is doing what.”

Franks and lead author and postdoc Elva Robinson used it to investigate whether social information or individual experience is most important in an ant's decision to leave the safety of its nest. T. albipennis ants forage after receiving information from a scout about a good food source or because they are hungry. To find out which trigger is stronger, Robinson prevented scouts from returning to the nest, then used the RFID tags to track when the remaining ants decided to go foraging and which ones did so.

Even ants without previous foraging experience ventured out to seek food if they were especially lean, the team showed. That finding adds to previous studies in Franks's lab showing that an individual ant's behavior has some flexibility. Such an assertion runs counter to previous studies by Wilson and others that correlate an ant's age with its given task, suggesting that tasks are hard-wired. “Our studies show the opposite, and they are frankly extremely controversial,” Franks acknowledges. “But we followed individual marked ants for 6 months, photographing each colony several times a day,” amassing data that he asserts no one else has gathered.

“When Wilson says, ‘It's this way,' many people won't investigate further,” says Dornhaus. “But Franks is different, which is why he has such original papers; he studies things other people think are already settled. That's why I asked to join his lab.” Still, Hölldobler warns that conclusions based on Franks's chosen ant species, with its small colony size, may not be a good model for ants in general.

Franks's creative defiance is apparent in his decision to test whether ants can teach. “There was a definition that everyone accepted,” he explained, “and so we looked at the ants' behavior to see if an ant ‘modifies its behavior in the presence of another, at a cost to itself, so another individual can learn more quickly.’” He and grad student Thomas Richardson watched hours of videotaped “tandem runs” of foraging ants leading naïve ants (all painted) from the nest to a new source of food. These tandem runs are carefully orchestrated, with the lead ant walking slowly and waiting for her follower to tap her antennae on the leader's legs and abdomen before continuing, and “fulfill the criteria” of teaching, the two wrote in Nature. Not everyone was impressed. “This might be a charming metaphor,” Hölldobler and Wilson wrote in their new book, The Superorganism, “but it adds little, if anything, to our understanding.” Franks insists that the behavior is teaching and says that the lead ant in effect evaluates her pupil, a behavior he and Richardson describe further in a 2007 Current Biology paper.

Back at Franklin's experiment, the scout ants were ferrying the last of the brood into the new nest. But these scouts were young juveniles, or “callows,” because Franklin had carefully brushed away the older ants. Were the callows up to the task? “There's one who has two ants following her,” Franks pointed out. “But she's moving too fast. … Oh, she's lost them now.” The lost ants milled around, looking … lost. What would they decide to do? Which elements of a tandem run are innate; which are learned? Franks fired off question after question. The Idea Man leaned in close, his eyes tracking the ants' every move: A new experiment was surely in the works.

12. ORIGINS

# On the Origin of Photosynthesis

1. Mitch Leslie

Where would we be without photosynthesis? In the third essay in Science's series in honor of the Year of Darwin, Mitch Leslie details researchers' efforts to piece together how and when organisms first began to harness light's energy.

Try to picture the world without photosynthesis. Obviously, you'd have to strip away the greenery—not just the redwoods and sunflowers, but also the humble algae and the light-capturing bacteria that nourish many of the world's ecosystems. Gone, too, would be everything that depends on photosynthetic organisms, directly or indirectly, for sustenance—from leaf-munching beetles to meat-eating lions. Even corals, which play host to algal partners, would lose their main food source.

Photosynthesis makes Earth congenial for life in other ways, too. Early photosynthesizers pumped up atmospheric oxygen concentrations, making way for complex multicellular life, including us. And water-dwellers were able to colonize the land only because the oxygen helped create the ozone layer that shields against the sun's ultraviolet radiation. Oxygen-producing, or oxygenic, photosynthesis “was the last of the great inventions of microbial metabolism, and it changed the planetary environment forever,” says geobiologist Paul Falkowski of Rutgers University in New Brunswick, New Jersey.

Given its importance in making and keeping Earth lush, photosynthesis ranks high on the top-10 list of evolutionary milestones. By delving into ancient rocks and poring over DNA sequences, researchers are now trying to piece together how and when organisms first began to harness light's energy. Although most modern photosynthesizers make oxygen from water, the earliest solar-powered bacteria relied on different ingredients, perhaps hydrogen sulfide. Over time, the photosynthetic machinery became more sophisticated, eventually leading to the green, well-oxygenated world that surrounds us today. In the lab, some biochemists are recapitulating the chemical steps that led to this increased complexity. Other researchers are locked in debates over just when this transition happened, 2.4 billion years ago or much earlier.

Looking so far into the past is difficult. The geological record for that time is skimpy and tricky to interpret. Eons of evolution have blurred the molecular vestiges of the early events that remain in living organisms. But “it's a terribly important problem,” says biochemist Carl Bauer of Indiana University, Bloomington, one well worth the travails.

## To catch a photon

Over more than 200 years, researchers have ironed out most of the molecular details of how organisms turn carbon dioxide and water into food. Chlorophyll pigment and about 100 other proteins team up to put light to work. Plants, some protists, and cyanobacteria embed their chlorophyll in two large protein clusters, photosystem I and photosystem II. And they need both systems to use water as an electron source. Light jump-starts an electrical circuit in which electrons flow from the photosystems through protein chains that make the energy-rich molecules ATP and NADPH. These molecules then power the synthesis of the sugars that organisms depend on to grow and multiply. Photosystem II—the strongest naturally occurring oxidant—regains its lost electrons by swiping them from water, generating oxygen as a waste product.

However, some bacteria don't rely on water as an electron source, using hydrogen sulfide or other alternatives. These nonconformists, which today live in habitats such as scalding hot springs, don't generate oxygen. Their photosynthetic proteins huddle in relatively simple “reaction centers” that may have been the predecessors of the two photosystems.

Envisioning the steps that led to this complex biochemistry is mind-boggling. Similarities between proteins in photosynthetic and nonphotosynthetic bacteria suggest that early microbes co-opted some photosynthesis genes from other metabolic pathways. But protophotosynthesizers might also have helped each other piece these pathways together by swapping genes. Biochemist Robert Blankenship of Washington University in St. Louis, Missouri, and colleagues say they've uncovered traces of these lateral gene transfers by comparing complete bacterial genomes. For example, their 2002 study of more than 60 photosynthetic and nonphotosynthetic bacteria (Science, 22 November 2002, p. 1616) suggested that bugs had passed around several photosynthesis genes, including some involved in synthesizing the bacterial version of chlorophyll.

Gene-sharing might also explain the puzzling distribution of the photosystems, Blankenship says. A cell needs both photosystems to carry out oxygenic photosynthesis. Yet modern nonoxygenic bacteria have the presumptive predecessor either of photosystem I or of photosystem II, never both. To explain how the two protein complexes wound up together, Blankenship favors “a large-scale lateral [gene] transfer” or even a fusion of organisms carrying each photosystem. However, other researchers remain skeptical, arguing that one photosystem evolved from the other, possibly through the duplication of genes, creating an ancient cell with both. No one knows for sure.

## The electron thief

Either way, it took some fancy fiddling to convert the primitive reaction centers to oxygen-generating photosystems. Oxygenic photosynthesis was a huge upgrade, leading to a land of plenty, says biochemist John Allen of Queen Mary, University of London. “Water is everywhere, so the organisms never ran out of electrons. They were unstoppable.”

But water clings to its electrons. With its oxidizing power, photosystem II can wrench them away, but the reaction centers in nonoxygenic photosynthesizers cannot. Biochemists James Allen (no relation to John Allen) and JoAnn Williams of Arizona State University, Tempe, and colleagues are working out how a bacterial reaction center could have evolved photosystem II's appetite for electrons.

Taking a hands-on approach, they have been tinkering with the reaction center of the purple bacterium Rhodobacter sphaeroides to determine if they can make it more like photosystem II. First they targeted bacteriochlorophyll, the bacterial version of chlorophyll that's at the core of the reaction center, and altered the number of hydrogen bonds. Adding hydrogen bonds hiked the molecule's greed for electrons, they found.

The water-cleaving portion of photosystem II sports four manganese atoms that become oxidized, or lose electrons. So the team equipped the bacterial reaction center with one atom of the metal. In this modified version, the added manganese also underwent oxidation, the researchers reported in 2005. James Allen says that their creations aren't powerful enough to split water. But eventually, they hope to engineer a reaction center that can oxidize less possessive molecules, such as hydrogen peroxide, that would have been present on the early Earth. Even if the researchers never replicate photosystem II, “if we define the intermediate stages, we've accomplished a lot,” he says.

## Something in the air

How the photosystems got their start is crucial for understanding the origin of photosynthesis. But the question that's drawn the most attention—and provoked the most wrangling—is when photosynthesis began. Most researchers accept that nonoxygenic photosynthesis arose first, probably shortly after life originated more than 3.8 billion years ago. “Life needs an energy source, and the sun is the only ubiquitous and reliable energy source,” says Blankenship.

The sharpest disputes revolve around when organisms shifted to oxygenic photosynthesis. At issue is how to interpret a watershed in the fossil record known as the great oxidation event (GOE). In rocks from about 2.4 billion years ago, geologists see the first unmistakable signs of significant, sustained levels of atmospheric oxygen. These signs include red beds, or layers tinged by oxidized iron, i.e., rust. Further support that the GOE marks an atmospheric revolution comes from a technique that detects skewed abundances of sulfur isotopes that occur if the air lacks oxygen. These imbalances persisted until the GOE, when they vanished.

Hard-liners construe these data to mean that oxygenic photosynthesis could not have emerged until shortly before the GOE. But other scientists disagree. “We are finding more and more hints that oxygenic photosynthesis goes deeper into the fossil record,” says astrobiologist Roger Buick of the University of Washington, Seattle. These hints could push the origin back 600 million years or more.

One line of evidence is oil biomarkers that researchers think are the remains of cyanobacteria. They've turned up in rocks that are up to 2.7 billion years old. And in western Australia, thick shale deposits that are 3.2 billion years old hold rich bacterial remains but no traces of sulfur or other possible electron sources, suggesting that the microbes were using water to make energy.

Geologist Euan Nisbet of Royal Holloway, University of London, and colleagues found additional support for an early origin when they went searching for traces of RuBisCO, a key photosynthetic enzyme. RuBisCO feeds carbon dioxide into the reactions that yield sugars. The enzyme version found in oxygenic photosynthesizers plays favorites: It prefers carbon dioxide that contains the carbon-12 isotope over the bulkier carbon-13. In 2007, Nisbet and his colleagues found disproportionately low carbon-13 values indicative of RuBisCO activity when they analyzed organic matter in rocks from three sites in Zimbabwe and Canada that are between 2.7 billion and 2.9 billion years old. Nisbet concludes that oxygen-making photosynthesis began at least 2.9 billion years ago.

The early-origin case isn't ironclad. For example, a 2008 paper that has some researchers fuming claims that the oil biomarkers are contaminants that seeped in from younger rocks. Advocates also have to explain why it took hundreds of millions of years for oxygen to build up in the air.

Although the last word on the origins of oxygen-making photosynthesis isn't in, researchers say they are making progress. One thing is for certain, however: Without this innovation, Earth would look a lot like Mars.