# News this Week

Science  01 Dec 2006:
Vol. 314, Issue 5804, pp. 1366
1. SCIENTIFIC CONDUCT

# China's Fraud Buster Hit by Libel Judgments; Defenders Rally Round

1. Jia Hepeng,
2. Hao Xin*
1. Jia Hepeng is a writer in Beijing.

BEIJING—China's self-appointed science cop, Fang Shi-min, was dealt a pair of setbacks last week in his high-profile crusade against academic misconduct. Two Chinese courts handed down libel judgments against Fang, known by his nom de guerre Fang Zhouzi, and the newspapers and Internet sites that have featured his writings on pseudoscience and fraud. Fang's revelations have cost several scientists their jobs and reputations.

With Fang now on the defensive, his backers are setting up two funds to help foot the costs of litigation. “If you strike false science, false science [makers] will strike you,” says Guo Zhengyi, a science writer and a co-organizer of one foundation. Guo and others say they hope that, by drawing attention to what they call “absurd” court rulings, they may force the government to crack down on corruption.

Fang received a Ph.D. in biochemistry and did a postdoc in the United States before becoming a science essayist. He got fired up about fraud in 2001, after reading dubious claims in the Chinese media about “nucleotide supplements.” Fang then started using his Web site, Xin Yu Si (“New Threads”), to debunk pseudoscience and expose alleged misconduct, from résumé padding to data fabrication (Science, 10 August 2001, p. 1039).

By Fang's tally, New Threads has aired allegations against more than 500 individuals. Fang uncovered some cases himself, but most were e-mailed to him by others. Few exposures have led to official investigations, and fewer still have resulted in punishment—the most notable being the dismissals earlier this year of an assistant dean of Qinghua University's medical school in Beijing and a dean at Tongji University in Shanghai, both for having falsified their résumés and exaggerated achievements.

The anonymous allegations published on New Threads trouble some people, who liken them to dazibao, or posters, used during the Cultural Revolution to denounce “class enemies.” Fang and his supporters contend there's a big difference: The Web postings are individual actions not directed by the state. The Chinese government takes an ambiguous stance: It blocks access in China to New Threads' U.S.-based site, www.xys.org, but allows access to mirror sites.

Fang's recent setbacks came on consecutive days. On 21 November, a Beijing intermediate court ruled that an article Fang wrote in 2005 defamed the late Liu Zihua, a Sichuan provincial government employee. In a dissertation written in France in the 1930s, Liu presented calculations based on the eight trigrams of an ancient divination text, I Ching (Book of Changes), predicting the existence of a 10th major planet in the solar system. Liu's prognostication was resurrected after last year's announced discovery of 2003UB313 (now officially a dwarf planet named Eris). A Sichuan newspaper ran a story extolling Liu's prophecy.

In an essay, Fang labeled Liu's prediction “pseudoscience” and noted that a Chinese astronomer discredited it in the 1940s. Liu's widow and son sued Fang and several newspapers and Internet content providers for libel. The court judged Fang's words “insulting” to Liu and ordered him to apologize publicly and pay Liu's family $2500 plus legal fees. The family did not respond to an interview request. Then on 22 November, a court in Xi'an slapped another libel judgment on Fang, ordering him and Beijing Keji Bao (Beijing Sci-Tech Report) to pay Xi'an Fanyi University$18,750 and its president Ding Zuyi $1250 in damages plus legal fees. In 2004, Chinese newspapers ran stories citing a “report” in the Los Angeles Times lauding Ding as one of China's most respected university presidents and his private college for training translators as the 10th-ranked university in China. In a 2005 article in Beijing Sci-Tech Report, Fang quoted an education ministry spokesperson, who stated that investigations showed the report to be “a self-paid advertisement.” Ding sued Fang for libel. Ding could not be reached for comment. Fang is appealing another libel verdict by a Wuhan court last July. In this case, Xiao Chuan-guo, a urology professor at Huazhong University of Science and Technology in Wuhan and a clinical associate professor at New York University School of Medicine, sued after Fang accused him in an essay last year of counting conference abstracts as publications in international journals to inflate his achievements. Fang also challenged Xiao's claim that a surgical procedure he invented is recognized internationally and has won neurourology's “highest award.” The presiding judge ruled that Fang's criticisms “seriously lacked facts” and ordered him to apologize publicly and pay Xiao$3750 in compensation. A final ruling is expected in early December.

Xiao told Science that the accusations are groundless and that Fang “intentionally confused” Xiao's urology awards. Xiao says he supported Fang until 2002, after which he concluded that Fang had begun to “misguide the public” with less-than-solid accusations.

In response to the Wuhan ruling, Zhang Feng, a Florida-based financial analyst and college classmate of Fang's, along with eight other expatriates, last month established the Organization for Scientific and Academic Integrity in China to raise money for Fang and other anticorruption campaigners. So far, the nonprofit has received more than 10,000 in donations. And in China, Guo and others are creating a separate science fraud-fighting fund. Fang's lawyer, Peng Jian, hopes the foundations will raise money to “implement systematic investigations into some individual cases or organize seminars to discuss legal punishments against proved misconduct makers.” Fang vows to continue “using sharp-tongued criticism1” to expose misconduct and folly. But he doubts that his freelance fraud busting can play a “decisive role” in cleaning up Chinese academia. To be more effective, he says, he intends to report future allegations, when appropriate, to a new disciplinary office at China's Ministry of Science and Technology and wait for a response before posting them online. 2. DEVELOPMENTAL BIOLOGY # Fraud Investigation Clouds Paper on Early Cell Fate 1. Gretchen Vogel A surprising report in a contentious area of developmental biology has sparked a scientific misconduct investigation at the University of Missouri, Columbia. Until that inquiry is complete, the results of the implicated paper, published in Science earlier this year, remain in limbo. Contrary to prevailing dogma, the report claimed that mouse embryo cells have distinct fates from the time of the very first cell division. If true, those findings would dramatically change the current understanding of mammalian embryo development—and could also play a role in ongoing political and ethical debates over cloning and stem cells. But the senior author of the paper now says the results are not trustworthy and predicts that the paper will be retracted as soon as the university completes its investigation. The paper, which was accepted as Science editors were embroiled in the scandal surrounding Woo Suk Hwang's human cloning papers in early 2006 (see Editorial on p. 1353), again raises questions about the limitations of the peer-review process in detecting fraud. Published in the 17 February issue of Science (p. 992), the paper caused an immediate stir. It is well known that embryonic cells of insects and amphibians have distinct fates from the first cell divisions, but the picture for mammalian embryos has been far murkier (Science, 6 May 2005, p. 782). Experiments in which mouse embryos are teased apart and cells transplanted from one embryo to another have suggested that until about a week after fertilization, mammalian embryo cells are quite interchangeable. There is an ongoing debate, however, over whether very early embryo cells—when the embryo is at the four- or eight-cell stage—might have a tendency toward one fate or another, although they are not yet committed. The results published in February indicated a much earlier differentiation than anyone else in the field had suggested. The corresponding author, R. Michael Roberts, is an expert in bovine embryology and until this year had not been involved in the debate. In the paper, Roberts, with post-doctoral fellows Kaushik Deb and Hwan Yul Yong and microscope technician Mayandi Sivaguru, claimed that in most mouse embryos there was a distinct difference between cells from the first cell division on. One cell had strong expression of a gene called Cdx2, the paper claimed, and eventually went on to give rise to the placental tissues. The other, which had less Cdx2 expression, went on to form the eventual fetus. The scientists argued that their observations might help explain why cloning in mammals is so inefficient. If Cdx2 expression is disrupted by cloning, they speculated, then embryos might have a hard time developing further. Proponents of so-called alternative nuclear transfer were also excited by the results. This technique has gained some support among people otherwise opposed to stem cell research and human nuclear transfer because it hypothetically offers a way to derive stem cells without destroying an embryo. If scientists use egg cells or somatic cells lacking Cdx2 for their nuclear transfer, the reasoning goes, the resulting cells would be unable to form a true embryo. The Roberts results seemed to help the supporters' case: If the gene is so crucial from the very beginning, then that would strengthen the argument that cells lacking it could not be called an embryo. The results, however, “were so drastically different from any of the results obtained by any other group” that most people viewed them skeptically from the start, says Magdalena Zernicka-Goetz of the University of Cambridge, U.K. However, it wasn't immediately clear why the results were so different, she says. A different strain of mice or a different labeling technique might have been the cause, she says. Others in the field were less willing to suspend their disbelief. Within weeks of the paper's publication, Roberts says, several scientists wrote to Science, to Roberts, and to the University of Missouri, pointing out problems with the data. Some of the images seemed suspiciously similar to each other, they said. In others, the staining didn't seem to line up exactly with the cells. By late April, Roberts says, the university had started an investigation. It was soon clear that there was reason to worry about the data's veracity, Roberts says. “In my view, there are a number of questionable images,” he says. But until the university investigation is complete, he says, the team will not be able to explain the details of what is wrong or retract the paper. Roberts says the university is being very cautious about assigning any blame before the investigation is complete. All the co-authors have since left the university. Two have found other jobs, and a third has apparently dropped out of contact. In the meantime, Science issued an “Editorial Expression of Concern” to alert the community that it should not trust the published results (Science, 27 October, p. 592). Some critics question why the paper was published in the first place or why image-analysis techniques—which Science editors said they put in place at the beginning of the year—didn't spot the apparent problems. Davor Solter of the Max Planck Institute for Immunobiology in Freiburg, Germany, one of the scientists who wrote to Science, contends that the review process was flawed. Science editors declined to discuss the specifics of the review process, which is confidential, but Katrina Kelner, deputy editor for biology, says, “Science published the paper based on feedback we got from the field.” Solter speculates that leading scientists in the field did not review the paper, noting that if they had, they likely would have caught the problems. But Richard Behringer of M. D. Anderson Cancer Center in Houston, Texas, is not sure the problems could have been spotted ahead of time. Although Behringer now says he can see the evidence of duplicated images, he says at first reading the paper seemed solid, if surprising. “I can understand why referees would say OK,” he says. Kelner and Science Editor-in-Chief Donald Kennedy add that even if the data are found to have been manipulated, the new image-analysis techniques would not have picked it up. Those techniques can flag unmatched pixels that are signs of deletions or cut-and-paste manipulations. But duplicated images—like those in the Hwang paper—are harder to spot, Kelner says. In retrospect, Roberts says he wishes he had been more cautious with the results his lab members presented to him. “I didn't go into this with preconceived ideas. I got into it by happenstance,” he says. The research was aimed at determining whether Cdx2 was involved in turning on another gene in bovine embryos, he explains, and the mouse embryos were used as controls to analyze Cdx2 expression. “But the results looked so beautiful, you couldn't come to any other conclusion.” Since questions about the paper were raised, he says, “I've obviously questioned myself and my judgment. I haven't had a good night's sleep since February.” The University of Missouri is expected to finish its investigation later this month. 3. MEDICINE # Squelching Progesterone's Signal May Prevent Breast Cancer 1. Jean Marx A woman who carries a mutated BRCA1 gene faces a daunting decision: She can opt for constant monitoring hoping to catch any cancer early, while it's still curable, or she can elect to have her breasts or ovaries removed to prevent cancer from developing in the first place. Results described on page 1467 now suggest that one day there may be a third option: using drugs rather than surgery to prevent BRCA1-mediated breast cancers. BRCA1 is a so-called tumor suppressor, a gene that in its normal form protects against cancer. One way it does this is by helping cells repair DNA damage that might otherwise result in cancer-causing mutations. The new work, which comes from Eva Lee and her colleagues at the University of California, Irvine, points to another cancer-preventing role for BRCA1. By aiding in the degradation of the receptor through which progesterone exerts its effects, the gene's protein product apparently checks the hormone's growth-promoting action on breast tissue. Lee's team also showed that mifepristone, a drug that induces abortions by inhibiting the progesterone receptor, blocks the development of mammary tumors in mice that have had the rodent version of BRCA1 inactivated in their mammary glands. “The paper has a mechanism [of BRCA1 activity] and has clinical implications. It's potentially important,” says Eliot Rosen of Georgetown University School of Medicine in Washington, D.C., who is also studying the interaction between BRCA1 and progesterone. Previous work had raised suspicions that progesterone fosters breast cancer development. For example, women taking both estrogen and progesterone to treat menopausal symptoms have a higher risk of developing breast cancer than women who take estrogen only. And working with human breast cancer cells in lab cultures, Rosen's team found that normal BRCA1 inhibits the action of the progesterone receptor, although how has been unclear. In the current work, Lee and her colleagues created mice that lacked functioning copies of the rodent versions of both BRCA1 and p53, another tumor suppressor that is frequently mutated in breast cancers. Although the female mice had never been mated, their mammary tissue showed increased cell proliferation—much as the breasts of pregnant woman do when high progesterone levels prepare the mammary glands for lactation. What's more, all the rodents developed mammary cancers by the age of 8 months. Mice treated with mifepristone, however, were still tumor-free at 12 months of age. Lee and her colleagues then took a closer look at the epithelial cells that give rise to breast cancer. “A lot more cells” from the double-mutant mice had progesterone receptors, she says, than did cells from normal animals or from animals in which only the p53 gene had been knocked out. Further work on cultured mouse and human cells revealed that the progesterone receptor is broken down less readily when BRCA1 activity is missing. As a result, “the [hormone's] signal goes on much longer,” Lee says. The excessive cell growth this produces provides extra chances for cancer-promoting DNA mutations to occur, especially because BRCA1 loss also handicaps the cell's DNA repair machinery. The participation of the progesterone receptor in BRCA1-mediated breast cancer could help explain why tumors occur specifically in the breast and ovaries even though the gene is mutated in cells throughout the body. Those other cells don't carry progesterone receptors. Lee points out that mifepristone itself may not be suitable for long-term use in cancer prevention because it acts on steroid receptors besides the one for progesterone. It might therefore cause unacceptable side effects such as immune suppression. Other more specific progesterone blockers are under development, Lee notes. There is uncertainty about how accurately the new mouse model reflects human breast cancer. Lee cites findings by Jeff Boyd's team at Memorial Sloan-Kettering Cancer Center in New York City that tissue adjacent to human breast tumors with BRCA1 mutations shows elevated progesterone expression compared to t issue from normal breast. However, Christine Clarke and her colleagues at the University of Sydney at Westmead Millennium Institute in Westmead, Australia, actually saw a decrease in progesterone receptors in tissue removed by mastectomy from BRCA1 carriers. The two situations aren't quite comparable. “The status of tissue around tumors is different from that of tissue taken from normal breast,” Clarke says. But that issue, and likely many others, needs to be resolved before cancer prevention trials of progesterone inhibitors can begin. 4. MOLECULAR BIOLOGY # Three Methods Add Up to One New Way to Genetically Engineer Fruit Flies 1. Jennifer Couzin When Koen Venken began a Ph.D. project on fruit fly genetics at Baylor College of Medicine in Houston, Texas, 4 years ago, he quickly became frustrated by limitations of the standard techniques for genetically engineering the insects. So he turned his attention to developing a novel procedure. The result, described in a paper published online by Science this week (www.sciencemag.org/cgi/content/abstract/1134426), appears to be a powerful new way of making transgenic flies, one that will likely make it easier to study fruit fly genes that were previously too large to work with and to compare the behavior of similar genes belonging to different fly species. The work eases two roadblocks that have long troubled the fly community: inserting genes longer than about 20,000 DNA bases—which make up more than 5% of the insect's genes—and controlling where in the genome those genes land, which impacts how they get expressed. “The real advantage here is that it's a way of putting in really big bits of DNA” into the fly, says Michael Ashburner, a fly geneticist at the University of Cambridge, United Kingdom. Traditionally, geneticists create transgenic flies with help from a piece of fly DNA called the P element. Roughly 2 decades ago Gerald Rubin, now director of the Howard Hughes Medical Institute's (HHMI's) Janelia Farm in Loudoun County, Virginia, and his colleagues spliced a stretch of DNA into a portion of a P element and found that the new DNA was easily incorporated into the fly's genome. The P element, however, can't integrate long stretches of DNA like the ones Venken wanted to work with when he joined the Baylor lab of geneticist and HHMI investigator Hugo Bellen. So Venken took components of the P element, including ones that allow it to integrate DNA into the fly genome, and added them to loops of bacterial DNA called plasmids. These bacterial artificial chromosomes (BACs) can more stably retain larger amounts of foreign DNA than a P element alone. To add into those BACs the DNA he wanted to insert into the flies, Venken next turned to a technique called recombineering, which was developed about 8 years ago. Recombineering involves allowing a BAC to recombine with an intended transgene within bacteria, isolating that BAC, and then using other bacteria to make multiple copies of it. Finally, Venken used a third existing technique to control where in the fly genome the BAC-ferried gene would land. When he injected the BACs into fruit fly embryos, Venken also injected messenger RNA that encodes an enyzme made by a bacterial virus called a phage. This enzyme normally inserts a phage's DNA into specific sites on a bacterial genome, but in these circumstances, it integrates the BAC-carried gene at similar DNA sequences engineered into the fly genome. Using this combination of methods, Venken, Bellen, and their colleagues have inserted DNA stretches as long as 133,000 bases into the fly genome. “It's now not clear what the upper limit is,” says Daniel Barbash, a geneticist at Cornell University. Bellen's lab is now assembling a library of fly DNA in the novel BACs for interested researchers. The new approach “is a significant technical advance,” says Rubin. It “allows us to do certain things we couldn't do before,” such as study the effects of whole gene complexes, like the homeotic genes that affect early development. And several scientists note that this blended approach might work to genetically modify other organisms. “The potential to export this system to … other animals is quite high,” says Barbash. 5. DRUG RESEARCH # WHO Panel Weighs Radical Ideas 1. Martin Enserink Lifesaving antiretroviral drugs have been available for a decade in wealthy countries, yet millions of HIV-infected people south of the equator still can't get them. The medicine cupboard is equally bare for people afflicted by tropical illnesses such as visceral leishmaniasis, sleeping sickness, and Chagas disease, for which there are no truly good therapies. Western medical science has not done well by the world's poor, and some critics blame this on its reverence for intellectual property (IP). Is it time to overhaul the IP protection system? A new working group hosted by the World Health Organization (WHO) will consider that question in a series of meetings beginning next week in Geneva, Switzerland. Critics of the current IP protection system hope that WHO's Intergovernmental Working Group (IGWG) on Public Health, Innovation and Intellectual Property will reform—or even hack down—the still-expanding worldwide patent system. They say it puts lifesaving new drugs beyond the reach of poor patients and hampers development of new medicines for tropical diseases. But others, including the pharmaceutical industry, argue that the IP protection system isn't the real problem and that the talks in Geneva risk distracting people from practical solutions. The IGWG—whose members will include representatives of governments as well as nongovernmental organizations—appears “motivated by anticapitalism rather than logical thinking about how to get drugs to patients,” says Trevor Jones, a former director of research and development at the Wellcome Foundation. Patents are designed to spur the invention of new products. But they also allow companies to charge high prices, putting people without purchasing power at a disadvantage. Many critics say it is not enough to help the poor get access to drugs; the system's incentives must be changed. To produce new drugs for neglected diseases, they say, the world needs a new R&D system that rewards not market sales but the potential to save lives and improve health. One such framework, which the IGWG may consider, is a hotly debated proposal for an international treaty to open up drug discovery, championed since 2002 by James Love, director of the Consumer Project on Technology in Washington, D.C. Under Love's “R&D Treaty,” countries would agree to spend a minimum percentage of gross domestic product on medical research, including a portion for neglected diseases. In addition, the treaty would promote open access to research findings and possibly add R&D incentives. For instance, governments could award big monetary prizes for those who invent important new medicines. Manufacturers would then be free to produce and market them cheaply. The treaty, recommended in a letter to the World Health Assembly by 162 scientists, health experts, and others last year, “is widely seen as the end of the pharmaceutical industry as we know it,” says Anne-Laure Ropars, a researcher at the George Institute for International Health in London. No wonder the industry is vehemently opposed. The treaty would create an “extremely complicated international bureaucracy,” says Eric Noehrenberg of the International Federation of Pharmaceutical Manufacturers and Associations in Geneva, adding that the award system would never work. Instead, Noehrenberg offers a different idea: The world should create markets where they currently don't exist. For instance, companies could be enticed with research grants from a “Global Tropical Disease Fund” or the promise of guaranteed sales should they develop an effective new drug. The industry also contributes through a model called the public-private partnership (PPP). Over the past 10 years, more than two dozen PPPs have sprung up to tackle diseases of the poor. Enlisting industry, academia, governments, and foundations, these partnerships, such as the TB Alliance and the Medicines for Malaria Venture (MMV), have produced many new candidate drugs (Science, 13 January, p. 167). And the IP protection regime has not been an obstacle, says MMV president Chris Hentschel: “If people spent less time thinking about IP and more about other things, we would make more progress.” But others point out that health PPPs have a narrow base: 60% of their funding comes from a single source, the Bill and Melinda Gates Foundation; governments contribute very little. Moreover, industry tends to help PPPs that work on diseases that affect both the poor and people from rich countries, such as malaria and TB, says Els Torreele, project manager at the Drugs for Neglected Diseases Initiative. Given the scope of the problem, something more radical is needed, she says. Whether the IGWG can deliver a solution remains to be seen. The group's predecessor at WHO, the Commission on Public Health, Innovation, and Intellectual Property Rights, issued a raft of recommendations in April—such as increasing contributions to PPPs and building clinical trial capacity—but could not agree on some key patent issues. Some predict that when the IGWG issues its final report to the World Health Assembly in May 2008, it may propose ways to implement the less controversial parts from the April review rather than a radical reform. But Love thinks the world may be ready for a change. He notes that, although the U.S. government has generally aligned itself with the pharmaceutical industry, it strongly supported increased access to HIV drugs in Africa. It also unexpectedly voted for the resolution introduced by Kenya and Brazil that called the IGWG into existence. (The drug companies and the European Commission opposed the plan.) Love is hoping for another surprise. 6. SCIENCE EDUCATION # Doing More With Less 1. Jeffrey Mervis Many U.S. educators think that a streamlined science curriculum with fewer topics per grade is a necessary first step toward boosting student achievement What do schoolchildren need to know to be scientifically literate? Scientists and educators keep coming up with new answers to that deceptively simple question. As states gear up for two nationwide assessments of student achievement in science, many educators think that the time is ripe to take another hard look at what children should be taught. But others worry that reviving debate on that contentious topic may divert attention and resources from the bigger challenge of actually improving student performance in science. Everybody agrees that current practices aren't good enough. “It is the height of national folly to think that America can maintain any competitive edge in science the way we are now teaching and testing it,” asserts Michael Casserly, executive director of the Council of the Great City Schools in Washington, D.C., after urban schools last month reported low performance in science. There's also consensus that the curriculum is a big part of the problem. A September report by a panel of experts assembled by the National Academies' National Research Council (NRC) deplores curricula that “contain too many disconnected topics that are given equal priority, with too little attention to how … [knowledge] is enhanced from grade to grade.” The result, says the panel in Taking Science to School (nap.edu), is that students receive a “fragile foundation” in science. That fragile foundation is exposed in both national assessments of what students know and in international comparisons with their peers. Those poor performances are fueling a campaign by the National Science Teachers Association (NSTA) to develop a national consensus around what NSTA Executive Director Gerald Wheeler calls “science anchors”: a small number of concepts that educators agree are essential for students to understand at any particular grade level. “There are way too many things in the standards,” Wheeler says, “and too much divergence in what's being taught across the country.” He sees the anchors as a de facto core curriculum drawn from topics that most schools are already teaching, “like Newton's law of gravity, or evolution and natural selection.” Wheeler hopes to influence two testing regimens that dominate U.S. elementary and secondary school education. The first, the 2001 federal No Child Left Behind Act (NCLB), requires states to test students in grades 3 through 8 each year in reading and mathematics. Its importance derives from the sanctions facing schools whose students do not make sufficient progress each year. Next year, science will be added to that lineup, although the law doesn't hold schools accountable for student achievement in that subject. The second is the National Assessment of Educational Progress (NAEP), a nonbinding, quadrennial federal assessment of student achievement in grades 4, 8, and 12 across several subjects. Although called the nation's report card, its results are not broken out by schools and districts, and there are no penalties for poor performance. One big sticking point is that, thanks to the country's 200-year history of local control over education, there isn't a national curriculum. Casserly and many educators would like to see voluntary national standards that would reduce variations among the 50 states and 15,000 local school districts. Two Senate bills introduced earlier this year would move the country in that direction by asking expert panels to identify common ground among state curricula and standards. One bill (S. 3790), from Senator Hillary Clinton (D-NY), would even develop a model math and science curriculum and sample assessment questions. The other (S. 2357), by Senator Edward “Ted” Kennedy (D-MA), would help states align their curricula and standards to national benchmarks. Neither bill attracted much attention this year, but that's likely to change next year, when Kennedy takes over as chair of the Senate panel with jurisdiction over federal education efforts. ## Do it again Standards-based instruction is not a new idea. And this is not the first time the concept is being invoked to help raise student achievement in science. In the early 1990s, scientists and educators rallied around the idea of describing the important concepts in biology, chemistry, physics, and the earth sciences that all U.S. elementary and secondary school students need to master, as well as the nature of scientific thought. The movement crested with the appearance of two acclaimed documents: the 1993 Science Benchmarks for All Americans from AAAS (which publishes Science), and the National Academies' 1996 National Science Education Standards. Educators hoped the standards would ensure not only that teachers covered the most important topics but also that there would be a seamless transition from one grade to the next—and, in a highly mobile society, that children wouldn't be shortchanged if they moved from one district to another. So far, so good. “Before the standards, teachers pretty much taught whatever they wanted to,” says Megan Lewis, who teaches physical sciences, chemistry, and physics to high school students in the rural Glen Lake, Michigan, school district. But because many officials see standards as a threat to local control over education, they are no more than voluntary yardsticks that states are free to adopt, modify, or ignore. Over the past decade, state and local education authorities have used those documents as a starting point for compiling their own standards. Unfortunately, the results have been less than ideal. Take Lewis's home state. Michigan was one of the pioneers in the standards movement, adopting science guidelines in 1991. In 2000, the document was revamped and renamed the Michigan Curriculum Framework. Since then, it's undergone another metamorphosis, emerging this summer as the Michigan Merit Curriculum. The current version, which describes what students should know at each grade level, is linked to tougher statewide graduation requirements that, for the first time, mandate 3 years of high school science. Lewis is very supportive of the state's attempt to upgrade science instruction. But she wonders why her state has just adopted its third set of science standards since she began teaching 14 years ago. “Why are we doing this again?” she asks. “Science is science.” ## Hydra-headed science If only it were that simple. For one thing, most experts agree that the nationwide standards that came out in the 1990s weren't really a bare-bones version of what students needed to master. “We pared down by 40% the amount of material that was being taught, we estimated,” says George “Pinkie” Nelson, former director of AAAS's Project 2061, which developed Benchmarks and another AAAS product, called the Atlas of Science Literacy, that presents the concepts in Benchmarks as a cluster of interlocking maps to help teachers prepare lessons on any particular topic. “But there was still way too much material.” Sally Goetz Shuler, executive director of the National Science Resources Center, a joint effort of the National Academies and the Smithsonian Institution in Washington, D.C., calls them “a good first effort. … They were a lot better than the mile-wide, inch-deep” curriculum most states were offering at the time. “But they were still way too complicated,” she adds, “especially beyond the fifth grade.” One problem in developing science standards is the multiple fields that must be included. “Remember, it's the sciences, not science,” says Janice Earle, a senior program director for elementary and secondary education at the National Science Foundation (NSF), which funded the recent NRC report and supported the creation of both 1990s standards documents. One consequence is what Shuler calls “the science wars,” in which experts lobby to make sure their specialty is adequately represented in any standards document. That effect is magnified as each state (Iowa is the lone exception) develops its own standards, says Nelson, who runs a science, math, and technology education program at Western Washington University in Bellingham. “It's easier to put things in than to take them out,” he notes. Expanding the standards, in turn, has led to ever-larger textbooks, as publishers scramble to make sure their materials cover all the topics state and local school districts had crammed into their standards. But despite their heft, those textbooks often fail to capture the idea that science is, in the words of the recent NRC report, “not only a body of knowledge, but also a way of knowing” about the world. That approach includes formulating and testing hypotheses, adjusting one's understanding to fit the data, and then blending that new information with what the student already knows to come up with a better understanding of any particular phenomenon. It's a process that doesn't fit neatly into a lecture, or even an experiment, the report points out. And it's something that few students have a chance to experience at any level. “My students have a hard time figuring out how things really work,” says Thomas Lord, a plant biologist and science educator at Indiana University of Pennsylvania in Indiana, Pennsylvania. “When I teach the water cycle, for example, I ask them about the role that plants play in the process. Not one kid mentions photosynthesis. And these are science majors.” Lord worries that any standards, especially something concise such as science anchors, could simplify the curriculum to the point of squeezing out the real science students need to learn. ## From zero to 10 To be sure, the standards are just one element in reforming science education, an effort that includes improved teacher training and professional development and stronger ties between school districts and university science faculty. “Gerry's idea of putting together some big ideas is an important one,” says Nelson. “And new standards are fine. But I think that it's zero on a 10-point scale of improving U.S. science education.” Wheeler agrees that science anchors won't suddenly make students smarter or give teachers a better understanding of fundamental scientific concepts that they never learned adequately before entering the classroom. But he thinks that the anchors might be attractive to states preparing for two upcoming major assessments. “It's a way of identifying the low-hanging fruit. It's a marketing technique,” he admits. “Once people buy into the concept, then maybe we can get them to develop better assessment items, and professional development, around them. If that happens, then I think we will be moving in the right direction.” The science component of NCLB begins in 2007-'08, for students in one grade at each of three levels: elementary, middle, and high school. But the results won't be counted as part of the law's requirement that students show adequate yearly progress (AYP). “Not being part of AYP means that science may remain on the back burner,” Wheeler fears. A more promising target may be NAEP, which will be given next in 2009. Earlier this year, an expert panel (Wheeler chaired its steering committee) sketched out a new “framework” of what the test should cover, as well as new ways to measure that knowledge. The draft NAEP framework (nagb.org) drew explicitly from the two 1990s documents, says Senta Raizen of the National Center for Improving Science Education run by WestEd, a California-based nonprofit with a contract from NAEP's oversight body to revise the assessment. “Our hypothesis was, if it's in both documents that it's in,” says Raizen, who co-chaired the project's planning committee. “If only one, then we'll think about it. So there's nothing fundamentally new about the content.” Raizen emphasizes that NAEP isn't trying to tell states what to teach—“we don't have a national curriculum in this country”—and that NAEP provides only “a snapshot” of what students have learned. But she agrees that it can “serve as a model” for the upcoming NCLB assessments that states must devise. ## A matter of time NSF's Earle thinks there are valid reasons to be optimistic about the latest efforts to clarify what students need to know in science. “I'm sensing that maybe it is time to think about taking the next step,” Earle opines. “The standards have been out there for a decade or so, and it takes people a while to digest them.” But don't expect anything to happen quickly, she counsels: “U.S. education is not efficient, by definition. We have a very decentralized system.” Michigan's Lewis doesn't have the luxury of time. As a classroom teacher, she's responsible for making sure her students understand the subject matter and can pass the highstakes tests. As a result, she suggests, only partly in jest, that some of the money being used to rework science standards might be better spent on her students. “With10,000, I could buy enough [PASCO] probes for every kid in my class,” she says, referring to equipment that allows experimental data on temperature, acceleration, and other features of the physical world to be collected and analyzed.

Wheeler agrees that the interaction of student and teacher is paramount to improving how schoolchildren learn science. “A clear set of standards aligned to the state assessment is a key first step,” he reiterates. “But unfortunately, we have to take about four first steps,” he adds, ticking off the need for better materials, improved professional development, and higher teacher retention rates. “Otherwise, there's going to be a lot of finger-pointing at the fourth grade teacher whose students didn't do well enough on the science assessment. And that doesn't help anybody.”

7. ASTROPHYSICS

# Burst-Hunter's Rich Data Harvest Yields a Cosmic Enigma

1. Tom Siegfried1
1. 1Tom Siegfried is a writer in Los Angeles, California.

The 2-year-old Swift gamma ray satellite has delighted astrophysicists with its versatility—and surprised them with observations that don't fit the models

SAN FRANCISCO, CALIFORNIA—In the quest for the secrets of cosmic explosions known as gamma ray bursts (GRBs), no satellite has been more successful than Swift.

Launched 2 years ago in November, Swift has outperformed expectations, providing NASA with a steady stream of news to report. Swift not only has been recording its predicted budget of bursts (about 100 a year) but also has gathered abundant data on other astrophysical phenomena, from nearby black holes in active galaxies to the most energetic magnetic flare ever detected on a star. “Swift has been a scientific bonanza,” says high-energy astrophysicist Ilana Harrus of NASA's Goddard Space Flight Center in Greenbelt, Maryland. “It's the satellite that keeps on giving.”

Now, Swift has given astrophysicists a major surprise. Its observation of a GRB earlier this year has challenged the standard classification system for such bursts.

GRBs are intense pulses of extremely high-frequency radiation emanating from distant space. Such bursts were first detected almost 4 decades ago by satellites designed to seek signs of nuclear weapons tests. When gamma radiation arrived from space instead of the ground, baffled astrophysicists groped for explanations. Among the more speculative suggestions was that the bursts signaled the demise of faraway civilizations that had annihilated themselves in nuclear wars.

In the 1990s, however, observations from the orbiting Compton Gamma Ray Observatory provided enough information to pin down key details, eventually establishing that some bursts were associated with supernova explosions. Others seemed to result from cosmic collisions, perhaps between neutron stars, the small, dense spheres left behind by supernovae.

Bursts believed to be associated with neutron stars were typically short—lasting less than 2 seconds. “Long” bursts lasted from seconds to minutes and were generally “softer”—meaning lower in energy—than the short, “hard” higher-energy bursts. Long bursts have been clearly linked to supernovae, but the short bursts' link to neutron stars is more speculative. “Short bursts lack a smoking gun,” says Joshua Bloom, a GRB investigator at the University of California, Berkeley.

In June, Swift further blurred the line between short and long by finding a long, soft burst with no apparent connection to a supernova, several astrophysicists reported at a recent meeting here.* That new burst and other Swift observations challenge the standard two-category scheme and raise questions about the nature of GRB progenitors.

“I see a growing crisis of classification,” said Bloom. “We don't just have long bursts and short bursts anymore that map directly to these progenitors. We actually have counterexamples in both cases that are really throwing a monkey [wrench] in the works.”

The prime culprit behind the category crisis is a burst recorded on 14 June that lasted 103 seconds, far into the range generally regarded as long. Observers eagerly awaited the appearance of stellar brightening signaling the supernova explosion responsible for the burst. But the supernova never showed.

Further observations suggested that the 14 June event was not a typical long burst in other ways. The initial pulse was high in energy but was then followed by a softer afterglow.

“That was kind of reminiscent of other short bursts we've seen,” Swift principal investigator Neil Gehrels of NASA Goddard said at the astrophysics meeting, suggesting that it belonged in the “short” category. “Short isn't the right word,” Gehrels said, but in many respects “it appears to group with the short bursts, and that could explain the lack of a supernova.”

In a paper posted online, Gehrels and collaborators argue that the 14 June burst requires a new categorization system. “This combination of a long-duration event without an accompanying supernova … opens the door on a new gamma ray burst classification scheme that straddles both long and short bursts,” the Swift scientists wrote.

In another online paper, Johan Fynbo of the Niels Bohr Institute in Copenhagen, Denmark, and an international team of collaborators including Berkeley's Bloom suggest that the 14 June burst implies a new type of explosive star death, producing a GRB but no supernova. If so, the burst represents the first of a whole new category of GRBs. Gehrels's and Fynbo's papers have both been accepted for publication in Nature.

Other astrophysicists, however, say it's too soon to junk the two-category system or invent new stellar death processes. “These guys are going off and making claims that you have a whole new class of [GRB] population never before seen,” says Bradley Schaefer of Louisiana State University (LSU), Baton Rouge. For such an extraordinary claim, he says, “you ought to have at least good evidence.”

In fact, Schaefer says, the mystery of the missing supernova could have a simple solution: The burst may have occurred far enough away to make the stellar explosion too dim to notice. Lack of a supernova seems mysterious only because the burst was estimated to be at a redshift of 0.125, relatively nearby in cosmic terms. But in a paper posted on the online astrophysics preprint archive (astro-ph/0608441), Schaefer and LSU collaborator Limin Xiao point out that that distance estimate is based solely on the measured redshift of the nearest galaxy along the line of sight to the burst. Perhaps that galaxy was not actually the burst's host and the burst was much farther away, too far for the supernova associated with it to be visible.

Data from Swift and other instruments can be used to estimate the intrinsic brightness of the burst, Schaefer and Xiao point out. Comparing the intrinsic brightness with the observed brightness gives a good measure of distance. Various indicators all suggest a high brightness for the burst, leading Schaefer and Xiao to assign it a redshift of about 2, far enough away to explain the lack of a supernova sighting. They calculate the odds of such a lineup of a galaxy with a more distant burst to be 1 in 125; because Swift has recorded more than 190 bursts, finding one such alignment is not surprising. “It's fully consistent with chance coincidence,” Schaefer says.

A similar conclusion appeared in a paper published 10 November in Astrophysical Journal Letters. B. E. Cobb and colleagues at Yale University determined the likelihood of lineup coincidence to be from about 1 in 50 to 1 in 200. Consequently, from one to four such coincidences would be expected in the bursts observed by Swift so far. “The conclusion that [the 14 June burst] requires a ‘new paradigm’ for gamma ray burst formation should be approached with caution,” Cobb and colleagues wrote.

Actually, a second such possible coincidence had already been recorded before the 14 June event. A burst detected in May also was technically “long”—at 4 seconds in duration—with no sign of a supernova. But that event was fainter and poorly observed, Schaefer and Xiao noted, and the burst might also have originated far behind the presumed host galaxy.

In any case, Swift's findings have surely complicated the older views of GRBs, providing much precise data that astrophysicists will have to digest to get a clearer understanding of the sources and properties of those cosmic flashes.

“Our real goal here is to attempt to uncover the progenitors of gamma ray bursts, whether they be long-duration gamma ray bursts or short-duration gamma ray bursts,” says Bloom. “We're trying to understand the diversity of the phenomenon. And because of Swift and other satellites, we're now in the position to really ask these questions in detail.”

• *High Energy Astrophysics Division of the American Astronomical Society, 4–7 October.

8. GLOBAL HEALTH

# South Africa Bolsters HIV/AIDS Plan, but Obstacles Remain

1. Robert Koenig

Ridicule at the Toronto AIDS Conference spurred South Africa's Cabinet to order a new plan to battle the epidemic

PRETORIA AND SOWETO, SOUTH AFRICA—When small baskets of garlic, lemons, and beets highlighted the South African exhibit at the XVI International AIDS Conference in Toronto last summer, many delegates were outraged. They viewed the display—intended to show the importance of nutrition in bolstering immune systems—as trivializing the response to the epidemic that now infects 5.5 million South Africans and kills an estimated 800 of them a day. As the meeting ended, Stephen Lewis, the United Nations special envoy for HIV/AIDS in Africa, lashed out at aspects of South Africa's AIDS policies as “wrong, immoral, [and] indefensible.”

The ridicule in Toronto was followed by a sharply critical letter to President Thabo Mbeki from 82 prominent international scientists, including Nobelist David Baltimore, virologist Robert Gallo, and 11 South African researchers. Arguing that garlic and lemons “are not alternatives to effective medications,” the researchers warned that “many people are … dying unnecessarily” in South Africa because they do not have access to antiretroviral (ARV) drugs to slow the progression of the disease. Although the country has the world's largest ARV program, it now reaches only about a quarter of the South Africans who are estimated to need the drugs.

Reflecting the outcry, some of South Africa's leading newspapers called in September for the resignation of the garlic-promoting health minister, Manto Tshabalala-Msimang. At about the same time, South Africa's ruling Cabinet, unhappy to again be the focus of international scientific scorn, decided to revive the near-moribund South African National AIDS Council. It named Deputy President Phumzile Mlambo-Ngcuka—rather than the controversial health minister—as the nation's point person for developing a more effective HIV/AIDS strategic plan for the next 5 years.

The deputy president planned to outline the framework of that new plan on 1 December, World AIDS Day. A draft of the wide-ranging plan, obtained by Science, features commitments to bolster prevention programs to sharply reduce the number of people being infected with HIV; better coordinate the government's often-fragmented response to the epidemic; support AIDS vaccine and antimicrobial research; and significantly expand ARV treatment—although the exact ARV target numbers were still being developed.

Many South African scientists, clinicians, and activists welcomed the long-overdue initiative to revamp HIV/AIDS programs. “For years, we had been confronted with obfuscation and confusion and a lack of leadership on HIV/AIDS,” says Francois Venter, who heads the Southern African HIV Clinicians Society. But he and others cautioned that the devil is in the details, some of which were not available as Science went to press. And no one was expecting that the announcement of a new action plan would end the debate on South Africa's HIV/AIDS policies.

The need for more effective government programs is clear. Although a draft of the plan cited evidence that “HIV incidence has started to decrease,” a November report by the Joint United Nations Programme on HIV/AIDS and the World Health Organization says that HIV prevalence—at nearly 19% of South Africa's adult population in 2005—“has not yet reached a plateau.” The nation's 5.5 million infected people include a quarter of a million children under age 15, the report said. It also warned of “a continuing, rising trend in HIV infection levels” among pregnant women using prenatal clinics.

## A history of controversy

International dissatisfaction with the country's HIV/AIDS policy is rooted in a series of government controversies and miscues over the past decade. In 1997, an attempt to fast-track clinical trials of a drug called Virodene ended in disgrace when a review panel found that the substance was toxic and had been prematurely tested on humans. Three years later, in early 2000, Mbeki sent a letter to the White House and to the U.N. Secretary-General suggesting that factors other than HIV could cause AIDS and asserting that it would be a “criminal betrayal” to “mimic foreign approaches to treating HIV/AIDS.”

Later that year, delegates to the International AIDS Conference in Durban were stunned that Mbeki and his health minister continued to question the connection between HIV and AIDS and failed to support ARV therapy. “The government took a strange position opposed by well-established science,” recalls the chair of the Durban meeting, pediatric AIDS researcher Hoosen Coovadia of the University of KwaZulu-Natal.

Meanwhile, as the epidemic worsened, the government came under increasing pressure to take decisive action. In 2002, Mbeki began to distance himself from the denialists and endorsed the concept of making ARVs available to pregnant women and rape survivors. Late in 2003, a panel developed an ARV rollout plan, which went into effect the following spring and now covers about 214,000 persons. Noting that the South African ARV program reaches more people than that of any other country, Medical Research Council (MRC) President Anthony MBewu contends that South Africa's recent initiatives on HIV/AIDS treatment and prevention have begun to blunt the epidemic.

Trying to rehabilitate the government's international image on HIV/AIDS policy, Cabinet officials are avoiding any public expression of AIDS denialism. Government spokesperson Themba J. Maseko told Science that “the position of the government is based on the understanding that HIV causes AIDS.” Even South Africa's most outspoken AIDS activist, Zackie Achmat, credits the ruling African National Congress party with exerting pressure to suppress AIDS denialism within its ranks. Although Mbeki has not made a definitive statement of his own position, Maseko says the president fully supports the Cabinet's recent HIV/AIDS decisions on developing a new action plan.

## A targeted approach

After the Toronto AIDS meeting, with the health minister hospitalized with a respiratory ailment, the Cabinet asked Deputy President Mlambo-Ngcuka to try to mend fences with interest groups and develop a stronger HIV/AIDS plan. Declaring that the nation's AIDS policy debate was at “a critical point,” she held meetings this fall with leading clinicians, scientists, and activists to try to resolve “difficulties and misunderstandings” and forge a consensus.

The most difficult single issue in reaching such an agreement has been setting targets for ARV treatments. (The government bears most of the costs of ARV drugs at public clinics, but international organizations and donors pay for ARV costs at many private or religious facilities, and medical insurance covers other individuals.) The government originally planned to announce specific ARV targets on World AIDS Day as part of the new plan. Indeed, one draft listed a goal of tripling the current ARV numbers, to 650,000 adults and 100,000 children, by 2011. But as soon as the plan's early drafts began circulating, AIDS activists and clinician groups began lobbying for much higher numbers, and they persuaded the deputy president to delay announcing specific targets until a compromise could be worked out, probably early next year.

Four influential AIDS groups that sought the delay—the Southern African HIV Clinicians Society, the activist Treatment Action Campaign, the AIDS Law Project, and the University of the Witwatersrand's Reproductive Health and HIV Research Unit—delivered a 28-page critique in November of an early draft of the action plan. That response, obtained by Science, argued that the proposed ARV targets “represent approximately 20 percent of those requiring treatment, and should be revised upwards. All epidemiologic data suggests that there are approximately 800,000 people who need ARVs at the moment … and that an additional 500,000 people will require treatment annually going forward.”

The critique also argues that the nation's research facilities are “largely uncoordinated when it comes to research on HIV” and recommends that a national health supervisory council find ways to improve coordination.

The draft HIV/AIDS plan did not address the coordination issue, but it confirmed that basic and clinical research into the epidemic were national priorities. Although shy on details, the draft specifically called for boosting research into microbicides and AIDS vaccines, a research strength of the country. MRC President MBewu told Science that “HIV/AIDS is the nation's top research priority.”

## Uphill battle

Although generally heartened by plans to boost HIV/AIDS prevention and treatment programs, South African scientists and activists caution that obstacles remain. For some, the chief problem is the recuperating health minister, Tshabalala-Msimang, who remains responsible for implementing the new HIV/AIDS plan. In November, she issued a statement lashing out at her critics and reaffirming her commitment to nutrition and traditional medicine in HIV/AIDS treatment.

University of Cape Town economist Nicoli Nattrass, an expert on the impact of the epidemic on South Africa, believes that HIV/AIDS activists “have won an important ideological battle,” but—with the health minister still involved in implementation—” the counterinsurgency remains strong.”

Clinicians and researchers are eager to see evidence that the government will back up its new HIV/AIDS commitments with more funding and improvements in health facilities. “Setting ambitious targets is good, but you have to have the resources and plans to meet those targets,” says researcher Coovadia.

Complicating the challenge, tuberculosis is rife among South Africans infected with HIV, and new drug-resistant strains are threatening to spread rapidly. “We've got two epidemics clashing in a dangerous way. We can't carry on with business as usual,” says immunologist Linda-Gail Bekker, co-director of the Desmond Tutu HIV Center in Cape Town.

Clinician Venter agrees that “a lot more needs to be done to get control of this epidemic.” Still, he says, “it helps to have support at the top.”

9. WILDLIFE CONSERVATION

# The Saola's Last Stand

1. Richard Stone

Wildlife experts say the rare Southeast Asian ungulate may soon disappear; a Vietnamese lab is undertaking a controversial attempt to clone it

PU MAT NATURE RESERVE, VIETNAM—Do Tuoc climbs a steep riverbank, entering the realm of the elusive saola. The creature, a cousin of cows, goats, and antelopes, is so rare that even Tuoc, who discovered the species in 1992, has never spied one in the wild. The forest ecologist finds safe footing on the slick slope and grabs a handful of broad, dark-green Araceae leaves. “Saola like to eat these,” Tuoc says. “At least, we have seen bite marks.”

A decade ago, the saola made headlines as the first large mammal new to science in more than half a century. Recent sleuthing suggests that the exotic ungulate is sliding toward extinction. At most, 250 saola are thought to roam the Annamites (called the Truong Son Mountains in Vietnam) of central Vietnam and Laos.

Now scientists are embarking on a lastditch effort to save the critically endangered species. Vietnam's National Saola Conservation Action Plan, expected to be approved by the government later this month, prescribes measures, including a hunting ban, that are deemed essential for the saola's survival. Meanwhile, a Vietnamese team is pursuing a conservation option of last resort: an attempt to clone the saola. But somatic cloning is supremely difficult even in the best-studied mammals—and “we know almost nothing about the saola,” says zoologist Nguyen Xuan Dang of the Institute of Ecology and Biological Resources in Hanoi.

More is at stake than one obscure relict species. The ecosystem that shelters the saola is home to an array of creatures, including at least two kinds of muntjac deer found nowhere else in the world. Saving this unique menagerie “would be a success story for other countries to follow,” says Barney Long, a conservation biologist with the World Wide Fund for Nature (WWF) who is working with local scientists and officials to protect the saola in central Vietnam.

The odds are against Long and company. “Foundations can easily raise funds for primates, tigers, elephants, rhinos,” says Dang. “For the saola, we can't even get money to educate the public, to tell people to stop hunting it.” As Vietnam's action plan notes starkly, “resources and attention afforded to the saola are currently insufficient to protect it from extinction in the immediate future.”

## Trophy hunting

Whereas biologists are captivated by the saola's unicornlike mystique, villagers in Truong Huong, on the edge of the Pu Mat Nature Reserve, are blasé about the beast. Few in this ethnic Thai community have seen a saola, and when they do, the outcome for the demure herbivore is almost invariably bad.

In a wooden house built on stilts, Lo Van Tinh, a farmer, sits cross-legged with four generations of family huddled around him and describes how, one day 10 years ago, he was hunting turtles in a mountain river. His dog spotted a saola mother and calf upstream and gave chase. The mother escaped, but her calf was cornered and assumed a defensive posture. Although a saola in captivity betrays no fear of humans, at the sight of a dog it snorts and hunkers head down, brandishing its long, straight horns, says saola expert William Robichaud, a zoologist with the Nakai-Nam Theun National Protected Area in Laos. That renders the saola easy to shoot, and for a juvenile, easy to grab. “I caught it with my hands,” says Tinh. The saola did not survive the 2-day hike back to Truong Huong, so Tinh and his family ate it. It was like beef, although not as tasty, he says.

In a home in nearby Truong Chinh village, a pair of saola horns hangs in a place of honor next to a poster of a smiling Vietnamese model. The dark-brown horns, about 40 centimeters in length, are more than twice as long as the head, which has short, coarse, chestnut-brown hair. To local people, the slightly diverging horns resemble the parallel wooden posts that support a spinning wheel (hence the name: Sao means “post,” and la means “spinning wheel”). Streaks of white hair above the eyes look like garish mascara.

It was in a home just like this that Tuoc discovered the saola. In May 1992, he was part of a team dispatched by the Ministry of Forestry, with WWF support, to Vu Quang forest, roughly 100 kilometers southeast of Pu Mat, to survey biodiversity in advance of Vu Quang's designation as a nature reserve.

Tuoc, schmoozing with the local villagers, wangled an invitation to a young hunter's home, where the team was shown a peculiar skull and horns. “I immediately thought it was a new species of antelope,” Tuoc says. But it was puzzling, as antelope prefer dry areas, and much of the Truong Son range is soaked by seasonal monsoons. Excited by the find, he asked local hunters to look for other specimens. Two more pairs of horns soon materialized, convincing the scientists that they had indeed found a new species, which they anointed the “Vu Quang ox.”

WWF funded a follow-up survey that November that turned up about two dozen pairs of horns and an intact saola skin. DNA analysis of the mitochondrial cytochrome b gene revealed a new bovid genus, and a paper in Nature in 1993 unveiled Pseudoryx nghetinhensis. (Subsequent DNA analyses suggest that cattle are its closest cousins.) The animal was confirmed in Laos through villager sightings and trophy horns in 1993. The species name is an amalgamation of the two Vietnamese provinces where specimens were first uncovered. The common name Vu Quang ox soon gave way to saola, a less parochial designation and one with historical roots. The first known written reference to the species is in an early 20th century Lao-French dictionary, which defines saola as an “antelope of the rocks,” says Robichaud.

The saola was the first large mammal discovery since the kouprey, a wild ox in Southeast Asia, in 1937. As an encore, Tuoc and colleagues first described the large-antlered (formerly giant) muntjac in 1994 and the diminutive Truong Son muntjac in 1997. (Both were discovered simultaneously in Laos.) With three mammal species under his belt, Tuoc has become a legend in cryptozoology, the study of previously unknown, presumed, or mythical creatures. “I've been very lucky,” he says.

## Zoological riddle

Ever since the saola's appearance, its biology, like the animal itself, has remained an enigma. In June 1993, Tuoc and colleagues at the Forest Inventory and Planning Institute in Hanoi took custody of two young saola that had been captured in Vu Quang. The animals ate several dozen kinds of plants and put on weight fast, Tuoc says. But after 2 months, they succumbed to infections. In all, 20-odd saola have been captured in Vietnam and Laos. All but two that were released into the wild died quickly in captivity.

The saola's fragility is no big surprise. “Certain animals in captivity, especially ungulates, are highly sensitive to stress,” says David Wildt, head of the Center for Species Survival at the Smithsonian's Conservation and Research Center near Front Royal, Virginia. Or the problem could be as simple as an “inappropriate” diet, says Wildt, whose team has pioneered techniques for breeding delicate creatures such as the Elds deer and the black-footed ferret. “A careful examination of why these animals die after capture is really needed,” he says.

What little is known about the saola has been gleaned primarily from the short-lived captives. In the mid-1990s, Cheng Syavong, a Lao general, offered a reward for the capture of a saola for his Lak Xao Zoo. In January 1996, Cheng procured an adult female. “I had the good fortune to observe her daily,” says Robichaud. The saola, he says, marked territory by flaring open a fleshy flap covering her maxillary glands on either side of the snout and stroking the underside across rocks, depositing a pungent, musky paste. The massive scent glands are thought to be the largest of any living mammal.

“Her most striking and endearing aspect,” Robichaud says, “was her utter calmness in the presence of humans.” Soon after arriving at Lak Xao, the saola allowed people to stroke her and fed from their hands. “She was tamer and more approachable than any domestic livestock I'd ever been around,” he says. But after a mere 18 days in captivity, the saola died suddenly, and no autopsy was performed—although she was found to be bearing a male fetus.

Saola are so rarely seen in the wild that it wasn't until 1998 that one was first caught on film in its habitat, by a camera trap near a mineral-rich spring in Pu Mat. Robichaud and Robert Timmons, an independent conservation biologist in Southeast Asia, have suggested that the survivors are descendents of a Pleistocene bottleneck, when their wet evergreen forests receded during cool, dry ice ages. “The current distribution of saola may reflect where these ice age refugia were,” says Robichaud.

Humans now have the saola on the ropes. In 1992, scientists pegged the population at 500 to 1000 in Vietnam, says Long. The estimate in Vietnam's action plan—“probably” fewer than 200—could be a large under- or overestimate, he says. But Long says a decline is evident “from the amount of hunted trophies that we see” and the lack of sightings in areas where the saola once roamed. Saola are also killed in snares set for more lucrative game such as bears, which fetch a high price for their gall bladders. Vietnam's action plan would ban snares in saola territory.

Habitat fragmentation further endangers the species. The action plan notes that the nearly completed Ho Chi Minh Highway, which will link northern and southern Vietnam, “must be viewed as the single largest threat to the connectivity of Saola populations and their habitat.” With support from the World Bank, the Dutch Development Organization, and the U.S. Agency for International Development, WWF is working with Vietnamese authorities to protect forests in two provinces, Thua Thien Hue and Quang Nam, where the largest saola subpopulation, approximately 50 individuals, is found. This “Saola Conservation Landscape” abuts forests in Laos, providing contiguous habitat for some of the few dozen saola thought to live across the border.

As an additional safeguard, Vietnam's national action plan would forbid keeping saola in captivity until 2010, unless one is confiscated from a hunter or liberated from a snare and is too injured to be released into the wild. To Wildt, this is a risky strategy. “I don't go along with the philosophy of leave them only in the wild and hope for the best,” he says. He suggests that saola experts convene a workshop that would take a hard look at captive breeding. “It's not like this has never been done before,” he says.

## A genetic “Hail Mary”

Long and others argue that without a robust effort to shield the saola from hunters and preserve its habitat, the animal is doomed. For all they know, the species may already have passed the point of no return.

That possibility is the main justification for a controversial, high-tech bid to keep the species on life support. On the tree-lined grounds of the Vietnamese Academy of Science and Technology in central Hanoi, a team at the Institute of Biotechnology led by Bui Xuan Nguyen is trying to clone the saola.

Nguyen knows the project is a long shot. But his lab has a chance at succeeding: He and his staff have been collaborating with top reproductive biologists in France, Japan, and elsewhere for 30 years and have racked up achievements in embryo transfer and in vitro fertilization in animals such as cows and rabbits. Nguyen is also credited with having developed a technique for rapidly freezing eggs and sperm that is particularly handy for preserving samples in the field. Building on this work, Nguyen is spearheading an effort to set up a lab network in Southeast Asia next year to “cryobank” frozen germ cells of rare species.

Soon after the cloning of the sheep Dolly in 1997, Nguyen says, he thought the revolutionary technique might be applicable to endangered species conservation. By then, the saola had become an icon in Vietnam. Nguyen struck up a collaboration with Tuoc's forest institute. “When someone finds a saola, the institute calls us and we immediately go take tissue samples,” Nguyen says. They have samples from one male and two females, including 30 immature eggs from one of the females that died.

They've held on to most of the eggs in the event that, someday, they might be able to attempt in vitro fertilization. But Nguyen has decided that “we cannot wait for a live female.” Working with Patrick Chesné from the lab of Jean-Paul Renard of the National Institute for Agricultural Research in Paris, Nguyen has used nuclear transfer to inject saola DNA into cow, goat, and swamp buffalo eggs. They have obtained early embryos—blastocysts—but these fail to develop. “We don't have any idea how to get past this stage,” Nguyen says. A fundamental hurdle is the dearth of knowledge about saola biology. “We have no information on the reproductive cycle, no idea how long pregnancy lasts,” he says.

Nguyen and his collaborators have filled in some gaps. For instance, they've established that the saola has 50 chromosomes. (Cows have 60, buffalo 84.) Nguyen now hopes to unravel how saola nuclei are reprogrammed. During reprogramming, an egg turns back the clock on an adult nucleus by removing chemical signatures of development, which returns it to an embryonic state—an essential step in somatic cloning. “We're interested in early molecular events in saola and closely related species,” says Renard.

When all the problems of interspecies cloning—such as different chromosome numbers and different mitochondrial DNA—are solved, then “cloning the saola will be possible,” predicts Takashi Nagai, a reproductive biologist at the National Institute of Livestock and Grassland Science in Tsukuba, Japan, who is working with Nguyen to conserve the genetic line of Vietnamese miniature pigs. Nguyen says he will persevere: “I'm a patient man.”

Some biologists, however, deem the effort hasty—or misguided. “Cloning is a tool for last-ditch heroics,” says Wildt. “It's too premature to consider it” for the saola, he says. “I don't see any conservation benefit from cloning the saola,” adds Long. “The money … would be much better spent trying to protect the species in the wild.” (Nguyen says his funding is “modest.”) To Long, the battle must be fought in the Truong Son Mountains. “If we lose the saola,” he says, “it will be a symbol of our failure to protect this unique ecosystem.”

That could jeopardize unknown species. “In Vietnam, there is still a lot of terrain not yet surveyed,” says Dang. Only in 2005, the kha-nyou, a bizarre, smallish rodent, was described from a specimen found in a Lao market; an expedition brought back the first live specimen last May. “There are small and medium-sized animals waiting to be discovered,” Dang says.

Optimists about the saola's fate are about as rare as the animal itself—but Tuoc is one of them. Natural enemies like the dhole are becoming scarcer, he says. Provided that snares are removed and vital habitat is preserved, the saola should be able to rebound, Tuoc says. “Maybe I'll never see one in the wild,” admits the cryptozoologist extraordinaire. “But I think—no, I hope—it will survive.” For the saola, survival will mean vanishing back into the misty sanctuary that hid it so well until humans came along.