# News this Week

Science  23 May 2008:
Vol. 320, Issue 5879, pp. 996
1. SICHUAN DISASTER

# Landslides, Flooding Pose Threats as Experts Survey Quake's Impact

1. Richard Stone*
1. With reporting by Chen Xi and Hao Xin.

CHENGDU, CHINA—Wei Fangqiang knows what it's like when a mountain crumbles: The Longmenshan, or Dragon's Gate Mountains, are prone to landslides. But when the physical geographer and seven colleagues with the Chinese Academy of Sciences' Institute of Mountain Hazards and Environment (IMHE) in Chengdu trekked into the area devastated by the Sichuan earthquake, they were stunned. It looked as though the hills had been blown apart. Landslides had flattened several-story buildings in the town of Beichuan and annihilated villages that clung to the steep slopes. In Wenchuan, Wei and his comrades picked their way across a 70-meter-high, 300-meter-wide rubble pile that had crushed a hydropower station and blocked the Chaping River. If an aftershock had struck, it could have spawned a new landslide where they were walking. “It was very, very dangerous,” Wei says.

Landslides unleashed by the rupture of a more than 200-kilometer section of the Longmenshan fault, followed by powerful aftershocks, dammed parts of nine rivers, creating 24 new lakes. The biggest and most threatening is 3.5 kilometers upstream of Beichuan. If the debris dam were to break, the resulting flood would threaten relief workers and researchers in Beichuan. “We're worried about another catastrophe,” says Wei. As Science went to press, experts with the Ministry of Water Conservation were weighing options for how to relieve pressure building up behind the dam. They had at most a week to act, said Cheng Genwei, IMHE's vice director.

Down the road from IMHE, researchers with the Chengdu Institute of Biology (CIB) were in mourning. Three senior staff members died when the wall of a hostel in the mountains collapsed as they were dashing out of the door for safety. (IMHE lost one staffer in Beichuan.) After a 20 May memorial service, CIB scientists were hoping to return to work with an ambitious research agenda, including an examination of habitat fragmentation and ecological succession in landslide areas. “The earthquake will be a big driver for research,” says CIB ecologist Bao Weikai. He and colleagues will also be alert to a grave threat to Sichuan's famed giant pandas: the possibility of a massive die-off of bamboo, the panda's staple, like one recorded in a quake 30 years ago.

At 2:28 p.m. local time on 12 May, the Sichuan earthquake struck with a magnitude of 7.9. It “was not a total surprise to geophysicists,” says Mian Liu, a geophysicist at the University of Missouri, Columbia. It occurred on a well-known, active fault system, he notes, which in 1933 produced a magnitude-7.5 quake that killed about 9000 people.

But the death toll of the Sichuan earthquake is horrific. As of 20 May, more than 40,000 people are known to have perished, including thousands of children. Experts are asking whether better construction, especially at schools, could have prevented many deaths. “Earthquakes themselves do not kill people,” says Liu. The biggest killer, he says, is structural collapse—“a point so sadly illustrated by this earthquake.” It appears that many wrecked buildings were not reinforced. “One hardly sees steel beams extruding from the collapsed buildings,” Liu says. “When they are seen, they are so thin that they bent with the debris like overcooked noodles.”

Under a makeshift canopy next to a swimming pool at a community center in the hard-hit historic town of Dujiangyan, west of the epicenter, geophysicist Miao Chong-Gang points to a map on his laptop overlain with seven circles in a line on the Longmenshan fault. It's the latest data from China's seismic monitoring network showing that the Sichuan earthquake was composed of seven powerful sequential ruptures unleashed when the fault ruptured southwest to northeast. “Several years ago, we could not do an analysis like this,” says Miao. But with more than 1000 seismometers now in a digital network, China can now parse data like this in a few hours.

Within 30 minutes after the quake hit, the China Earthquake Administration (CEA) in Beijing had crunched the numbers and issued a preliminary forecast of at least 7000 deaths. Their assessment would prove to be an underestimate, but it was alarming enough to prompt CEA to mount a full-scale response. Miao, vice-director of CEA's Earthquake Emergency Management Department Response Command Center, led a 230-person team to Dujiangyan late in the evening on 12 May. His group, one of 187 rescue teams in the disaster area, has saved 48 people; in the morning of 19 May, they were elated to have saved a 61-year-old woman who had survived 163 hours in the rubble.

Miao's team was about to switch from rescue to recovery. Among their tasks over the next 2 months, Miao says, is to ground-truth the computer-generated data. That will mean conducting seismic, strong-motion, and geologic surveys and running tests on everything from geomagnetism to water chemistry. Such research must wait until the aftershocks have subsided. Several CEA volunteers who were ferrying food and water on foot into the disaster zone were among more than 150 relief workers known to have died in aftershock-induced landslides. The slides also claimed the lives of two Sichuan Earthquake Administration researchers who were measuring crust deformation. “We have almost no experience in responding to an earthquake in a mountainous area,” says Miao.

Back in Chengdu, CIB scientists are itching to get out into the field. A week after the quake, 10 of their colleagues were alive but stranded at CIB's Maoxian Mountain Ecosystem Research Station in a pine forest 220 kilometers northwest of Chengdu. The institute had a couple of dozen long-term projects in the disaster area, a biodiversity hot spot that encompasses 22 nature reserves. They'll have to write a new research plan. “The earthquake has dramatically changed the landscape,” says CIB ecologist Luo Peng.

One urgent task is to monitor bamboo. The plant flowers once every 70 years or so. Shortly after a powerful earthquake in the 1970s, large swaths of bamboo suddenly flowered and died, says CIB ecologist Pan Kai-Wen. How a quake might trigger flowering is a mystery, but a large-scale die-off, he says, could pose a big threat to China's endangered giant pandas.

To map the landslides, Wei and his IMHE colleagues ventured into the danger zone on 15 May. They had to abandon their car where a landslide had blocked the highway and head toward Beichuan on foot. Traveling in the other direction was a ragged stream of refugees. When the researchers reached Beichuan the next day, they found that although many buildings had collapsed from the shaking, many others were demolished by massive boulders. “In some places, the landslides did more damage than the earthquake,” Wei says. “We know the rock is very loose here. But still I was surprised that the landslides were so severe.” In a nearby village, a woman was on top of a pancaked building. “She was calling her son's name, trying to wake him up.” There was no one else around.

Wei and his colleagues could not get past a blocked mountain pass leading to the biggest landslide, a 2-kilometer-long debris flow that had clogged the Qingjiang River. To ward off a catastrophic breach, Cheng says, the preferred option is to dig a canal that drains the lake gradually. If that's impossible, he says, they'll have to blast the dam and allow a more chaotic release. Sichuan's rainy season starts in late June; if the rains start early, before the problem is dealt with, the situation could be very dangerous, says Wei.

The IMHE researchers plan to head into the field as early as next week to sample landslide material and draw topographic maps. A future task is to advise authorities on a safe place to rebuild Beichuan city. The original site will almost surely be abandoned. “It should be a memorial to the earthquake victims—and a reserve for seismic research,” says Miao. CIB scientists hope to turn the disaster into an opportunity to advise Longmenshan residents about more sustainable livelihoods in the fragile mountain ecosystem. One practice they want to see ended is farming on the steep slopes. Better forest cover could reduce the landslide risk, says Luo: “We need a new strategy of mountain development.”

Others say the Sichuan disaster should stimulate China to rethink its entire approach to earthquake research. “In recent decades, geophysicists have spent too much energy and funding on research on deep-earth structure or tectonics,” says Zhou Shiyong, a geophysicist at Peking University. He argues that more attention should be devoted to earthquake prediction. “We could find some precursors,” he says, such as abnormal patterns in seismic stress or underwater variation before a huge quake occurs. Miao counters that any precursors of the Sichuan quake were minimal. “They could not have given us any warning,” he says.

One thing that will surely come under scrutiny is China's construction standards. “More effort should be devoted to earthquake hazards analysis and management, including developing and enforcing proper building codes, especially for schools, hospitals, and other public buildings,” Liu says. For thousands of victims in Sichuan, that lesson came too late.

2. U.S. AGRICULTURE

# Farm Bill Gives Agriculture Research a Higher Profile in the Department

1. Constance Holden

Spending on basic agricultural research in the United States could grow significantly thanks to a massive farm bill that Congress approved overwhelmingly last week. The bill also calls for a larger competitive grants program within the U.S. Department of Agriculture (USDA). “We view this as a real win,” says Ian Maw, vice president of the National Association of State Universities and Land-Grant Colleges (NASULGC) in Washington, D.C.

The changes are part of the Food, Conservation, and Energy Act of 2008, a 5-year, $307 billion measure that preserves massive subsidies for farmers. It renames the department's major extramural arm, the Cooperative State Research, Education, and Extension Service, as the National Institute for Food and Agriculture (NIFA). The institute is to be headed by a “distinguished scientist” appointed by the president to a 6-year term. The competitive grants portion of the new institute, to be called the Agriculture and Food Research Initiative, will replace the National Research Initiative (NRI). Its budget will be authorized at$700 million a year, $200 million more than the level for NRI, which actually receives only$180 million a year.

Supporters of agricultural research hope these changes will be more than cosmetic. They have long pressed for an entity within USDA analogous to the National Institutes of Health (NIH) or the National Science Foundation. Although Congress rejected a proposal from President George W. Bush to combine the department's intramural and extramural research programs into an office of science (Science, 23 February 2007, p. 1073), the legislation seeks to better coordinate USDA's $2 billion research portfolio by requiring an annual “roadmap.” It assigns the job to the undersecretary for research, education, and economics, currently Gale Buchanan, former agriculture dean at the University of Georgia. The arrangement combines the recommendations of a group headed by William Danforth, former chancellor of Washington University in St. Louis, Missouri, and a proposal from NASULGC. Danforth calls the measure “a great breakthrough” but adds that “what will really be necessary will be to build competitive funding.” Almost all of NRI's current budget is spent on “formula-driven” research, says Maw, whereas the new bill designates that 60% must go to basic research. It keeps separate the department's intramural arm, the Agricultural Research Service. The NIFA chief would also oversee$308 million over 5 years for competitive grants in two new areas: organic crops and “specialty crops,” otherwise known as fruits and vegetables. This funding becomes an actual spending level unless Congress explicitly decides otherwise.

Bush says that the bill is too generous on agricultural subsidies. But both houses passed it with veto-proof majorities, and it was expected to become law as early as the end of the week. It would take effect in October 2009.

3. RESEARCH FUNDING

# Australia's New Science Budget Gets a Mixed Review

1. Cheryl Jones*
1. Cheryl Jones is a writer in Canberra, Australia.

CANBERRA, AUSTRALIA—Two of Australia's science agencies are shedding jobs and trimming programs to comply with a new national budget that's both praised and criticized by research leaders. The spending plan announced by the Labor government last week—its first since coming to power in 2007—provides more money for education initiatives, including a $10.5 billion trust fund for higher education infrastructure, but less for two key players, the nation's premier science agency, the Commonwealth Scientific and Industrial Research Organisation (CSIRO), and the Australian Nuclear Science and Technology Organisation (ANSTO). The cuts are troubling, some say, because the government expects to reap a$20.7 billion surplus over Australia's next annual budget cycle, which starts 1 July.

The reduction at CSIRO is “a disappointment,” says Chief Executive Geoff Garrett. Combined with a cut announced previously, it will shrink the agency's appropriation through the 2008–09 budget cycle by roughly $15 million, or just over 2%, to$660 million. “Our aim will obviously be to preserve core capability and the science and research activities that we're doing, … but the arithmetic is such that … there will be some staff losses.” He says CSIRO will benefit from funding for energy technology, water management, and climate change adaptation. But it will probably have to cut its 6350-strong workforce by about 100. ANSTO, meanwhile, is to lose about 80 of its 1009 staff as it deals with rising costs and a cut of about 2.6% from its $144 million appropriation in the year ending on 30 June 2009. Australian Academy of Science President Kurt Lambeck welcomed the education investment fund, saying it will “put us on a path to a world-class higher education and research sector.” He is also enthusiastic about a promise to create new scholarships and 1000 fellowships for midcareer researchers: “It creates opportunities for retaining people in Australia and attracting overseas researchers at a stage when they are most productive.” But climate change research is a different story: An outlay of$2.2 billion over 5 years for global warming R&D, including clean coal and renewable energy projects, “does not reflect the urgency of the problem,” he says.

Lambeck worries that a series of reviews into the national innovation system and universities could be “used as an excuse for inaction.” Minister for Innovation, Industry, Science and Research Kim Carr could not be reached for comment.

4. GLOBAL WARMING

# Hurricanes Won't Go Wild, According to Climate Models

1. Richard A. Kerr

If you put much faith in the world's most sophisticated climate models, there's good news about how hurricanes will react to global warming. Two new model studies project a modest increase or even a decrease in the frequency and intensity of Atlantic tropical cyclones. “The Atlantic isn't going to be swallowed by repeats of the [disastrous] 2005 hurricane season,” concludes hurricane researcher Hugh Willoughby of Florida International University in Miami, who did not take part in the work.

But even some of those involved in the studies urge caution in interpreting the results. “I'm much less sanguine about models solving the problem,” says Kerry Emanuel, lead author of one of the papers and a hurricane researcher at the Massachusetts Institute of Technology (MIT) in Cambridge. There's still too much messiness beneath the surface of all such studies, he says.

Using two different approaches, both model studies tweak the big global climate models to simulate tropical cyclones. Global climate models can't form tropical cyclones because their picture of the atmosphere is too fuzzy. So, as they report in this week's issue of Nature Geoscience, climate modeler Thomas Knutson and colleagues at the National Oceanic and Atmospheric Administration's Geophysical Fluid Dynamics Laboratory (GFDL) in Princeton, New Jersey, put extra computing power into simulating the tropical Atlantic in enough detail to form storms while embedded in a fuzzy global model. And Emanuel and his MIT colleagues randomly “seeded” seven different global models with incipient storms that grew or died depending on whether conditions favored them, as they reported in the March Bulletin of the American Meteorological Society.

On average, the two approaches yielded much the same results for the Atlantic, where actual hurricane numbers have doubled in the past 25 years. The GFDL model produced a modest 18% decrease in the frequency of Atlantic hurricanes by the end of this century under global warming and a few percent increase in the intensity of storms. The MIT group reported just a couple of percent increase in frequency and a 7.5% increase in intensity.

Broader interpretations, however, differ. Pointing to their model's striking ability to reproduce variations in hurricane frequency during the past 25 years, Knutson concludes that his group's work “does not support the notion that increasing greenhouse gases will support large increases in hurricane or tropical storm frequency.” In contrast, Emanuel finds his “results to be very different when you [run] different models and very different in different ocean basins.” For example, using the MIT approach, the GFDL model produces a 23% increase in storm frequency rather than a decrease of 8%, he says. The models do predict a smaller increase in Atlantic hurricane activity than has been seen in the past few decades, Emanuel concludes. That implies that global warming was not the prime driver behind the recent burst of activity.

On the other hand, Emanuel adds, the models may not be properly handling global warming and its effects on tropical cyclones. Other studies have statistically linked the tropical Atlantic warming both to the greenhouse and to the jump in storm activity, many researchers note. What's more, the MIT group's seven different models yield a disturbing variety of predictions—from a 23% increase in frequency to a 29% decrease. And most researchers are concerned that the GFDL group generated input for its Atlantic storm model by averaging together a large range of predicted conditions from global climate models. As might be expected, researchers say bigger and better models are needed to make the message clear.

5. CLIMATE CHANGE

# Polar Bear Listing Opens Door to New Lawsuits

1. Dan Charles*
1. Dan Charles is a freelance writer in Washington, D.C.

The Bush Administration's decision last week to list the polar bear as a threatened species is about to spark a new round of litigation over greenhouse gas emissions. After analyzing climate models that predict the bear's sea ice habitat would continue to shrink due to global warming, the U.S. Department of Interior ruled that the animal deserves some protection under the Endangered Species Act (ESA). Several environmental groups are preparing to use the ruling to argue that cuts in greenhouse gases are now legally required to protect the polar bear, whereas conservative legal groups are planning to challenge the ruling itself.

When he announced the polar bear's new status, Interior Secretary Dirk Kempthorne tried to preempt litigation to force cuts in greenhouse gases. No specific source of these gases, Kempthorne asserted, will kill any individual polar bear, so the ESA doesn't require power plants, refineries, or even the nation's fleet of automobiles to reduce their emissions. But attorneys on both sides of the long-running legal war over endangered species predict that some courts will reject that argument. “The secretary can't dictate to the courts how they interpret the law,” says M. Reed Hopper, a principal attorney for the Pacific Legal Foundation (PLF) in Sacramento, California, a conservative critic of environmental regulation. “I think the environmentalists will find sympathetic judges who will rule that there is a causal connection and give them standing to bring their suits.”

Kassie Siegel, an attorney for the Center for Biological Diversity (CBD) in Joshua Tree, California, is leading the environmentalists' strategy. The “attempt to exempt greenhouse gas emissions is illegal and won't stand up,” she says. CBD, Greenpeace, and the Natural Resources Defense Council jointly filed their initial legal challenge on 16 May. Siegel, who also filed lawsuits that forced the government to list the polar bear, plans to argue that the ESA requires every government agency to consult with polar bear experts at the U.S. Fish and Wildlife Service before taking any step that could increase emissions of carbon dioxide. Such steps include authorization of oil and gas drilling, issuing permits for coal-fired power plants, or writing new fuel-economy standards for sport utility vehicles and trucks. “It's high time that federal agencies rolled up their sleeves and did what they're supposed to do on greenhouse emissions,” she says.

Some environmentalists doubt that such lawsuits ultimately will reduce greenhouse gas emissions. “I think it's highly unlikely that any court will say, ‘This source of emissions has to be halted because it's adding to the burden of carbon dioxide, which is melting ice in Alaska,’ “says Michael Bean, a specialist on wildlife conservation at the Environmental Defense Fund in Washington, D.C. Such legal actions could, however, capture public attention, says Holly Doremus, a professor of environmental law at the University of California, Davis. “It can really help people agree that, ‘Okay, we've got to act. And we've got to act now.’”

Meanwhile, PLF has announced that it will challenge the Administration's polar bear decision in court. Hopper says it makes no sense for the government to declare the polar bear threatened while insisting that it can do nothing to change the situation. “A listing that cannot address the alleged problem … should not have occurred,” says Hopper. “The listing can't affect the melting, but it does open the floodgates to litigation.”

Even some environmentalists say the ESA isn't well-designed for dealing with the broad impact of climate change. Doremus points out that global warming may create new dilemmas that the law didn't foresee. “Suppose we decide we can't save all species. Which ones should we concentrate on? The law doesn't allow us to give up easily,” she says. “But in situations where we may have to give up on some, we may need better mechanisms for triage.” And Bean finds it “worrisome” that “you have a species that is at risk of extinction, but the law that was designed to protect endangered species lacks the tools to deal with the threat.”

More animals may also be called as witnesses in the fight against U.S. climate change policy, as the polar bear is far from the only animal threatened by the shrinking field of arctic ice. Ice-dwelling mammals such as the Pacific walrus and several species of seals “are in even worse shape,” says G. Carleton Ray, an environmental scientist at the University of Virginia, Charlottesville. CBD has already filed petitions demanding that the ribbon seal and Pacific walrus also be listed as threatened.

6. CLIMATE CHANGE

# The Threat to the World's Plants

1. Dan Charles

A day after polar bears made headlines last week, the world's leading botanical gardens issued a call to remember threatened plants, too. Their new report, Plants and Climate Change: Which Future? makes the case for protecting the botanical foundations of terrestrial life. “If you read any report about the impact of climate change, it's almost always about polar bears or tigers,” said Suzanne Sharrock, director of Global Programmes for Botanic Gardens Conservation International (BGCI) in London and a co-author of the report.

But BGCI, a network of 2000 organizations involved in plant conservation, says climate change could kill off half of Earth's plant species. Plants that grow on islands or on mountainsides are at greatest risk because they have “nowhere to go” as the climate shifts around them.

BGCI also announced its own global effort to catalog and preserve threatened plants. It will update a 10-year-old survey of the world's trees, identifying species that need additional protection in their native habitat and collecting others for preservation in botanic gardens and arboreta. BGCI plans to reintroduce some threatened plants into their former habitats.

Thomas Lovejoy, president of the H. John Heinz Center for Science, Economics and the Environment in Washington, D.C., welcomed the new initiative. “At the outset, plants were scarcely mentioned in the Endangered Species Act. Now, it's an integral part,” he notes.

7. MICROBIOLOGY

# Bacteria Are Picky About Their Homes on Human Skin

1. Elizabeth Pennisi

Julie Segre is touring the microbial landscape of our body's biggest organ, the skin. In anticipation of a $115 million, 5-year effort by the U.S. National Institutes of Health (NIH), she's traveling from head to toe, conducting a census of some of the trillions of bacteria that live within and upon human skin. Although their project is just getting off the ground, Segre, a geneticist at the National Human Genome Research Institute (NHGRI) in Bethesda, Maryland, and her colleagues have already uncovered a surprising diversity and distribution among skin bacteria. And a few oddities have emerged, too: Microbes known mostly from soils like healthy human skin, living in harmony with us; and the space between our toes is a bacterial desert compared to the nose and belly button. Segre's work on what bacteria live where “is cool stuff,” says Steven Salzberg, a bio-informaticist at the University of Maryland, College Park. “We need to increase our own and the public's awareness of the diversity and quantity of bacterial species on our own skin. The more people are aware, the more we can do to control infection.” Bacteria and other microbes that colonize our skin and other tissues outnumber the human body's cells 10 to 1, forming dynamic communities that influence our ability to develop, fight infection, and digest nutrients. “We're an amalgamation of the human and microbial genomes,” says Segre. Recognizing this, NIH last year designated the Human Microbiome Project as one of its two Roadmap initiatives (Science, 2 June 2006, p. 1355). Researchers will sequence the genomes of about 600 bacteria identified as human inhabitants and get a handle on the 99% of bacteria that defy culturing but thrive in the skin, nose, gut, mouth, or vagina. “You have to understand what is the normal flora in the healthy skin to understand the impact of flora on disease,” says Kevin Cooper, a dermatologist at Case Western Reserve University in Cleveland, Ohio. As a first step, Segre, NHGRI postdoctoral fellow Elizabeth Grice, and their colleagues have studied five healthy volunteers, swabbing the insides of their right and left elbows. The site chosen isn't as unusual as it sounds; people with eczema often develop symptoms there. To survey the full thickness of skin, the researchers also used a scalpel to scrape off the top cel ls. And to reach even deeper, they took small “punches” of skin, a procedure akin to removing a mole. From all the samples, Grice, Segre, and colleagues pulled out 5300 16S ribosomal RNA genes, which vary from microbe to microbe. After lumping together the most similar 16S genes, they came up with 113 kinds of bacteria and identified these dermal residents by matching the 16S genes to those of known bacteria. (Segre described the results at a recent meeting at Cold Spring Harbor Laboratory, and they are being published online 23 May in Genome Research.) “That's a lot of diversity, a lot of different organisms,” says Martin Blaser, a microbiologist at New York University, who has done a similar survey of microbes living on the forearm, also finding a lot of diversity. Yet just 10 bacteria accounted for more than 90% of the sequences. Almost 60% of the 16S genes came from Pseudomonas, Gram-negative bacteria that flourish in soil, water, and decomposing organic debris. The next most common one, accounting for 20%, was another Gram-negative soil and water bug, Janthinobacterium. Neither had been considered skin microbes before this census. Although there were some differences among the volunteers in the microbes present, their elbows did share a common core set of microbes, the group reports. The three sampling methods yielded slightly different results, with “punches” revealing a surprising number of bacteria under the skin—1 million bacteria per square centimeter compared with 10,000 from the scrapes. “I would have thought under the skin there would be fewer,” says Salzberg. Segre and her team have also begun sampling 20 other skin sites, including behind the ear and the armpit, from the bodies of volunteers. Skin varies in acidity, temperature, moisture, oil accumulation, and “different environments select for different microbes,” says Blaser. Bacteriawise, reports Segre, “no subsite is identical.” Some researchers suspect that shifts in the makeup of skin microbial communities activate the immune system to cause diseases such as eczema. “If you know what the [healthy] flora is, then one strategy is to recolonize the area with the right flora,” says Cooper. 8. ENGINEERING # A New Great Lake--or Dead Sea? 1. Richard Stone Turkmenistan intends to create a huge lake in the desert by filling a natural depression with drainage water. Critics say it's a bad idea that could even spark a war. Turkmenistan intends to create a huge lake in the desert by filling a natural depression with drainage water. Critics say it's a bad idea that could even spark a war ASHGABAT, TURKMENISTAN—Bone-dry and as forbidding as California's Death Valley, the windswept, 120-kilometer-long Karashor Depression—a natural bowl speckled with the ash-gray, mica-laden sand that gives the Karakum, or “Black Sand,” Desert its name—might seem the last place in the world to put a lake. But on a fine day in October 2000, some 450 kilometers south of Karashor, President Saparmurat Niyazov leaned against a spade and breached a few-meters-wide earthen dam. Laborers took over, and soon water was gushing into the initial segment of a canal intended to fill Karashor to its rim. Golden Age Lake, the late president said, would become “the symbol of revival of the Turkmen land,” covering 3500 square kilometers—nearly the area of Utah's Great Salt Lake. With that gesture, Niyazov—known as Turkmenbashi, or “Father of the Turkmen People”—launched one of the most grandiose water projects ever undertaken. According to the plan, two canals that bisect the country will funnel runoff from heavily irrigated cotton fields into Karashor. The$6 billion project is designed to drain swamps and combat the buildup of salt and other minerals that have degraded three-quarters of Turkmenistan's arable land and eroded renowned archaeological monuments. “The lake will solve many problems,” says Paltamed Esenov, director of the National Institute for Deserts, Flora, and Fauna in Ashgabat. Turkmen officials predict that the project will reclaim 450,000 hectares of water-logged agricultural fields and create a habitat for migratory birds and an inland fishery.

Next month, Turkmen engineers say they will complete the mammoth effort's first phase: excavation of the two “collector” canals, each hundreds of kilometers long. Water apparently has already begun trickling into Karashor. “We are carrying out a unique, pioneering project,” says a senior engineer at the Turkmen State Water Research, Production, and Design Institute in Ashgabat, which leads construction of Golden Age Lake. “Everything we are doing is aimed at increasing agricultural productivity,” says the engineer, who requested anonymity after agreeing to be interviewed without permission from Turkmenistan's Ministry of Foreign Affairs.

But Golden Age Lake has unleashed a torrent of criticism as well. “There's no sense in this,” says Timur Berkeliev, a geochemist who coordinates the Worldwide Fund for Nature's Econet project in Turkmenistan. He and others are skeptical of plans to purify the runoff, laden with pesticides and fertilizers, and contend that the lake will become an artificial Dead Sea. “Trying to find value in this lake may be like trying to put lipstick on a pig,” says Michael Glantz, director of the U.S. National Center for Atmospheric Research's Center for Capacity Building in Boulder, Colorado. “A bad idea, even for the best of intentions, is still a bad idea.” Some experts believe that runoff will be insufficient to fill the lake, as the drainage water will evaporate or seep into the desert through unlined feeder canals.

That prospect raises fears that the lake could trigger a water war. Some observers worry that to prevent Golden Age Lake from running dry and to dilute tainted water, Turkmenistan might top it off with fresh water from the Amu Darya, a river on the border with Uzbekistan to the north. Uzbeks rely on the river for irrigation, and their leaders have said they would not tolerate a reduced share of the Amu Darya. “The lake project has incredible geopolitical implications,” says Johan Gely, who works on water issues in central Asia for the Swiss Agency for Development and Cooperation. The senior water engineer insists such fears are unfounded: “Every drop of the Amu Darya is valuable, and nobody is planning to use this water for Golden Age Lake,” he says.

Some see a window of opportunity to coax Turkmenistan to reconsider. Niyazov died in December 2006, and his successor, Gurbanguly Berdimuhamedov, has not yet spoken publicly about the project. Foreign leaders have remained mum as well, perhaps in deference to Turkmenistan's growing clout as owner of the world's fifth largest natural gas reserves. In the meantime, Berdimuhamedov has promoted a gradual opening of the isolated country. “The leadership is now sensitive to world opinion,” says Berkeliev. There might be one last chance, he says, to persuade authorities to convene an international scientific review before irreversible steps are taken to fill the lake. “This is the right time to do something,” he says.

## Back in the USSR

Centuries ago, central Asians learned how to make the most of the region's scarce water with networks of underground canals that conserved water for irrigation and drinking. “The tragic irony is that this region was home to one of the largest and most efficient irrigation systems in history, until the Mongol invasion destroyed much of the network,” says Peter Sinnott, director of the Caspian Project at Columbia University.

Josef Stalin managed to outdo the Mongols. During the Cold War, when central Asia was part of the Soviet Union, Stalin's water managers cooked up a notorious fiasco. In the 1950s, they began to divert massive amounts of water from the Syr Darya into a network of canals to irrigate cotton fields in Uzbekistan. The Syr Darya is one of two main sources of water for the landlocked Aral Sea; the river's reduced flow resulted in the Aral's shrinkage to less than a quarter of its original surface area.

Soviet planners were pushing cotton in Turkmenistan as well, and in 1954, work commenced on the Karakum Canal, which would feed water from the Amu Darya—the other big Aral Sea source—into the Turkmen heartland. At 1375 kilometers in length, the Karakum waterway, completed in 1988, is the world's longest irrigation canal. It has been a boon for agriculture—it tripled the arable land in its vicinity—and provides water to the capital, Ashgabat.

But it has a dark side: A sizable fraction of the water that enters the canal (15% to 50%, depending on whom you ask) seeps through its unlined bed into the surrounding soil. The hemorrhaging created a patchwork of ponds and swamps and has exacerbated salinization. As the ground became water-logged, the water table rose, bringing salts—primarily sodium sulfate—to the surface by capillary action. With evaporation, the brine crystallizes into mirabilite, a corrosive mineral that ruins oases and poisons fields. “Several kilometers to the left and right of the canal is a death zone,” says a Turkmen government scientist who asked to remain anonymous to keep his job. “If you step in the extremely salty water, your shoes are destroyed within a week,” adds a Western technician in Ashgabat who has visited the construction site of Golden Age Lake.

The Karakum Canal is not the only villain in the salinization saga. In the mid-1970s, Soviet engineers constructed drainage canals to discharge runoff into the desert. Dumping, coupled with overirrigation of farm fields, has saturated the ground and brought salt to the surface across the watershed. The water table is so high in the Dashoguz region, researchers say, that dozens of saline lakes have formed from water burbling up from the ground. “About 80% of arable land is damaged to different degrees,” says Berkeliev. Many Turkmen farmers soak fallow fields in winter, wrongly believing that as fresh water seeps into the soil, it takes salt with it. “But this has the opposite effect,” concentrating mirabilite, Berkeliev says: “This is a very complex problem, and the level of study is not adequate.”

That hasn't stopped Turkmen authorities from forging ahead with a solution: the resurrection of a 1970s idea to divert Turkmenistan's irrigation runoff into Karashor, near the border with Uzbekistan. Niyazov dusted off a Soviet rough blueprint for an artificial lake, Glantz and others assert, as a strongman's way of showing dominion over nature. “Only a powerful state can build such a gigantic thing,” Niyazov said in 2003. Turkmenistan's leader from the country's independence in 1991 until his death, Niyazov was anointed by parliament as Saparmurat Turkmenbashi the Great and, in 1999, made president for life. Golden Age Lake was not put to public consultation or debate. “It was almost impossible to object before,” says Berkeliev. In 2004, after merely asking whether the project included ecological expertise, the country's sole homegrown environmental group, the Katena Ecological Club, was shut down.

One potential beneficiary of the lake project is the region's archaeological treasures. “Water and salt are the main enemies of archaeological sites,” says the government scientist, who says that farmland and runoff have begun to encroach on what might be Turkmenistan's most famous site, the Bronze Age ruins of Gonur Depe (Science, 3 August 2007, p. 586). Salinization has already taken a heavy toll at one ancient monument, Little Kyz Kala in the medieval city of Merv, which has deteriorated especially rapidly in recent decades. The water table rose, soaking the foundations of the 1400-year-old brick fortress with salt and weakening them (see photos, below). With archaeologist Tim Williams and colleagues at University College London, SébastienMoriset's team at the International Centre for Earth Construction of the Grenoble School of Architecture in France has helped Turkmen conservators improve drainage and apply sacrificial soil layers at monuments that will bear the brunt of erosion rather than the original walls.

Draining the runoff water from the landscape should, in theory, ameliorate salt-induced erosion of the monuments, says the government scientist. “How it will work in practice,” he says, “we don't know.”

## Salvation or damnation?

To turn a dusty depression into a lake requires a whole lot of moisture. So the first and perhaps most formidable task was to excavate the two cross-country collector canals. Specialists plotted out routes that would make best use of natural topography. “In some places we had to dig as deep as 50 meters,” says the senior water engineer. In other areas they built platforms or added boulders as obstacles to suppress the flow rate. When they encountered giant stone slabs, they invented equipment that could be inserted in cracks between layers to lift the rock out. Blasting was considered too expensive, and “we don't have reliable professionals for that purpose,” says the senior water engineer.

The crew dug the northern canal in the Dashoguz region wider and deeper to allow for a larger water flow. For about half its length, the 432-kilometer Dashoguz Collector follows the bed of the ancient Uzboy River. The 720-kilometer Great Turkmen Collector starts in the Lebap region in the east and links up with the Dashoguz Collector 75 kilometers upstream of Karashor. About 45 kilometers from the depression, engineers built a 30-meter-tall, 600-meter-long dam to steer the water; otherwise it would have followed the lower-elevation Uzboy riverbed to the Caspian Sea. The senior water engineer says his engineers have also done some “sculpting” of Karashor's contours.

Water is now moving the length of the Dashoguz Collector and beginning to flow in the Great Turkmen Collector, the senior water engineer says. Satellite images confirm this. “It looks like canals, even unlined, can convey the drainage flow,” says Leah Orlovsky, a water researcher at Ben-Gurion University of the Negev in Israel who works in Turkmenistan. On a flight from Tashkent to Tel Aviv last October, Orlovsky noticed that an area of roughly 20 to 25 square kilometers at the southern end of Karashor was flooded.

Filling the lake should take several decades, says Esenov of the desert research institute. Water must first flow into the capillaries—a 1000-kilometer network of small feeder canals linking at one end to agricultural drainage ditches and at the other to small reservoirs or to the vast collector canals. Pumping stations regulate the flow into the collectors. Eventually, the senior water engineer says, the groundwater table should drop by a couple of meters, allowing for the gradual desalinization and reclamation of farm fields.

Although Golden Age Lake could save some iconic monuments, lesser known archaeological sites were damaged during construction. “They just bulldozed some small monuments and sites that hadn't been excavated yet,” says the government scientist. The project's design called for an archaeology rescue program, he says, but it had no funds. Living heritage is being lost as well. The collectors have raised the water table along their length, spoiling drinking water wells in some desert settlements. “Villages with ancient roots are being moved,” says the independent scientist. “It's a degradation of the cultural landscape.”

Future plans call for widening and deepening both collectors, says the senior water engineer. But there are no plans to line them. He referred questions about their dimensions and anticipated flow rates to institute colleagues, who were not available for interviews. One told Science privately that the lake's depth should reach 130 meters and its anticipated volume is 135 to 145 cubic kilometers.

“Data from Turkmenistan are hard to come by … and not so reliable,” says Glantz. But even rough approximations suggest that the project is doomed, says Berkeliev. The quality of the lake will depend on what goes into it, and Turkmen authorities in the past have predicted a water inflow of 10.5 cubic kilometers a year. About two-thirds will come from Dashoguz, including cross-border runoff from the Khorezm region of Uzbekistan; the Great Turkmen Collector will supply the other third of the water. However, Uzbekistan plans to build a drainage canal from Khorezm to the Aral Sea, so the amount feeding Golden Age Lake would eventually taper off, says Kai Wegerich, a central Asia water expert at Wageningen University in the Netherlands. “If the Uzbek drainage canal is built, it might not make sense anymore to construct the lake,” he says.

Khorezm canal or no, Berkeliev says his calculations are damning. Based on the high evaporation rate in Karakum, he asserts, “there will never be a water body there.” Others say Golden Age Lake may well come into being but is fated to become an environmental nightmare: a salty broth of organic pesticides and fertilizers.

Not so, says Vyacheslav Zharkov. He and his colleagues at the desert research institute in Ashgabat are devising filter media that absorb heavy metals and organic contaminants from runoff. These can be installed at treatment plants at points where water enters the collector canals—if the Turkmen government finds money to build such treatment plants. “After treating water with our sorbents, it is suitable for agriculture and for drinking,” Zharkov says. He claims that salt will be drawn out as water moves along the canals. “We have asked how the salt will be removed. They say the water will clean itself. Nobody is able to explain to me how this works,” says the Western technician. Berkeliev too says he is mystified.

## Looming shortage

The overarching question is whether Turkmenistan might tap the Amu Darya to improve the new lake. Under the Soviet-era water-sharing agreement, Turkmenistan and Uzbekistan each can use up to 22 cubic kilometers of water flowing out of Afghanistan and along their shared border—despite a huge difference in population size. (Turkmenistan has 5 million people; Uzbekistan has 28 million.) “The Uzbeks will not tolerate any ‘vanity diversions’ to the new lake,” says Glantz. In a tense situation, “new diversions will lead to a real war.”

Even if water isn't diverted to the lake, Afghanistan's plans to rev up irrigation are likely to curtail the Amu Darya's flow. Currently, it uses only a few cubic kilometers each year. “They are planning a massive expansion of irrigation,” says Wegerich. Several major projects launched in the last 2 years aim to irrigate more than 1 million hectares, with completion dates staggered over the next 5 to 15 years. Adds Glantz, “The Uzbeks think it is decades away. Wrong.” A complicating factor is the retreat of glaciers in the Pamir Mountains—the source of much of central Asia's fresh water. “Eventually, there will be no Amu Darya, no Syr Darya,” Mamadsho Ilolov, president of Tajikistan's Academy of Sciences, told Science. Golden Age Lake, he says, “will be very dangerous for neighboring countries.”

The best solution to Turkmenistan's water problems, Berkeliev and others argue, is conservation. Currently, Turkmenistan uses 5000 cubic meters of water per capita per year. That's twice the rate of Uzbekistan and more than 10 times that of Israel. “We are the champions of water waste,” says Berkeliev. It's high time, he and others say, that the country revises its Soviet-era agricultural system and switches to water-saving technologies, like drip or subsoil irrigation, and converts a significant portion of farmland to less water-intensive crops like wheat, corn, grapes, and olives.

Turkmenistan must also solve another problem arising from its poorly maintained infrastructure: water hoarding. Public supplies are sporadic, and when the spigot is on, Turkmen farmers funnel off as much as they can. Upgrading the irrigation system would be a much better investment than the lake, says Aral Sea expert Philip Micklin, a geographer at Western Michigan University in Kalamazoo. In his view, Golden Age Lake is “a big waste of money.” When the plan was being put together in the late 1990s, it had a conservation component—“but that disappeared,” says the independent scientist. “If they spent half the budget of the lake on water conservation,” he says, “they would not have had to build the lake.”

The senior water engineer says he is not bothered by the criticism and that it will not derail the lake project. “We faced the same opposition when we built the Karakum Canal,” he says. “Any such great project will have negative effects. But these are out-weighed by the benefits.”

Berkeliev says it's refreshing to be able to have this debate; it could never have happened under Niyazov. “But to change the minds of decision-makers, we need strong support from the outside,” he says. “We must have an international review of this project while there's still time,” adds geographer Igor Zonn of the Engineering Research Center on Water Management, Land Reclamation, and Environment in Moscow. That might be possible, as Turkmenistan continues a cautious opening up to the world. “We are trying to increase international cooperation on environmental issues,” Ogulsona Karyeva of the Ministry of Nature Protection told a Fulbright conference in Ashgabat last month.

“We would be very happy to work with foreign scientists,” says Esenov. “It's a complex problem.” That's something everyone can agree on.

9. ENGINEERING

# The End of an Intellectual Dark Age?

1. Richard Stone

This autumn, 80 top university graduates will begin postgraduate studies in Turkmenistan--the country's first crop of postgrads since 1997.

ASHGABAT, TURKMENISTAN—This autumn, 80 top university graduates in this central Asian nation will take part in a revived system of candidate (the Russian equivalent of a Ph.D.) and doctoral degrees in fields as diverse as art history and zoology. If that sounds modest, consider how many students last year began postgraduate studies in Turkmenistan: zero. This is the country's first crop of postgrads since 1997.

That year, the nation's authoritarian former leader, Saparmurat Niyazov, abolished advanced degrees. Other elements of his stultifying program included halving undergraduate education to 2 years and lopping a year off secondary school. Niyazov also closed the Academy of Sciences in 1997, citing “the lack of any practical scientific results.” Perhaps most insidious of all, his underlings enforced rote memorization of a book—the Rukhnama, a banal spiritual primer that Niyazov himself penned—as dogma.

Since Niyazov's death in December 2006, his successor, Gurbanguly Berdimuhamedov, has made education reform a top priority. He has upped university education to 5 years—six for aspiring physicians—and reinstated the lost year of secondary school. Science is back in fashion: “Science plays the leading role in the strong state, and therefore we should keep pace with its latest achievements,” Berdimuhamedov said recently, according to the state press. But recovery will not be easy. “After so many years of the forced degradation of the education system, it's really hard to revive it,” says one Turkmen government scientist. “The serious scientists didn't wait for changes within the country—they left.”

Turkmenistan is not the first modern nation to willfully erode its intellectual capacity: Afghanistan under the Taliban, for instance, suffered severely. But Turkmenistan's descent took place largely out of sight, as Niyazov isolated the country and placed sharp limits on international cooperation. In this twilight, in 2001, the Rukhnama appeared. The book is a mix of folksy guidance about how to lead a good life and a history of the Turkmen people that mangles the chronology of real events and fabricates others. “It did great damage for historians,” says the government scientist. Workplaces formed Rukhnama study circles, and TV programs showed children reciting passages while professing their love for Niyazov. Rukhnama knowledge was necessary to pass exams, including the driver's test.

The Rukhnama is still for sale in Ashgabat, and in some primary and secondary schools “it remains a strong part of the curriculum,” says Leon Yacher, a geographer at Southern Connecticut State University in New Haven, who lectured in Turkmenistan last month. When he visited a school in Turkmenbashi, a city on the Caspian coast, “every student had a copy of the book on their desk, and they were expected to read from it every day.” But to the relief of scholars, the Rukhnama is being phased out in universities and government offices.

Turkmen academics are trying to pick up the pieces. “A change of the curriculum is needed badly,” Yacher says. One problem is that there are few solid Turkmen textbooks, and no recent textbooks in Russian or in English, says the government scientist. That matches the general decrepitude of the faculty. Even in a field that was in favor under Niyazov—archaeology—the department was eliminated at Turkmen State University in 1999 and, says the government scientist, “the youngest archaeologist we have is a 60-year-old guy. When the last generation of archaeologists is gone, only foreigners will work here.”

Among signs of progress, construction has begun on a $35 million building for Turkmen State's physics and mathematics faculty, and a new campus is in the works for Turkmen State Medical Institute. The country is looking beyond its borders as well, with plans this fall to dispatch 1500 students to overseas universities, including Columbia University. “If [students] are off-the-charts good, we should do what we can to overcome any obstacles and get them here,” says Peter Lu, a physicist at Harvard University, who lectured in Turkmenistan in 2005. Foreign institutions can play a critical role in the intellectual revival, starting with the next generation. 10. COMMUNITY ECOLOGY # All That Makes Fungus Gardens Grow 1. Elsa Youngsteadt The discovery of a parasitic yeast draws attention to the ways that pathogens can stabilize ant agriculture and other symbiotic networks. The discovery of a parasitic yeast draws attention to the ways that pathogens can stabilize ant agriculture and other symbiotic networks Fifty million years ago, while the earliest primates were still scurrying from tree to tree, scrounging fruits and insects, attine ants were growing their own food. They were so adept at domesticating mushrooms that hundreds of species have descended from the original farmers, all of them cultivating fungi. Humans could learn a lot from the ants' success. Over the past 10 years, researchers have come to realize that the fungus gardens thrive because of an intricate web of bacteria and fungi that includes both pests, such as a newly discovered black yeast, and partners, including bacteria that keep pathogens in check. By studying these relationships, biologists hope they'll uncover lessons about the evolution of such interactions, knowledge that will help humans better manage microbes in medicine and agriculture. “It's a system that works,” says John Morrissey, a microbiologist at University College Cork in Ireland. “If you could develop a bacterial inoculant that was as successful in controlling a specific pathogen as the [beneficial bacteria] are for the ants, you'd be on to a real winner.” ## Complex network Ant agriculture runs the gamut. For leaf-cutter ants, farming is big business. They're the most notorious of the more than 230 described species of fungus gardeners, forming colonies of millions of workers that can defoliate a tree or crop in mere hours. The ants use the harvest to fertilize hundreds of separate fungus gardens in an elaborate subterranean compound. Most attine gardens, however, are small-scale operations: Their inconspicuous colonies are tended by as few as a dozen workers that scavenge bits of detritus to feed a spongy handful of fungus. But from the most primitive gardener to the dreaded leaf-cutter, all attines would starve if deprived of their fungal crops. When an ant queen leaves home to mate and found a new colony, she must take a little mouthful of the fungus with her to start a garden. Although naturalists have known since 1874 that the attine ants are fungus gardeners, more than a century passed before Ph.D. student Cameron Currie began to chip away the microbial complexity underlying the ant-fungus symbiosis. While at the University of Toronto in Canada, he discovered that ant gardens often contained a second fungus, Escovopsis. When he grew it on culture plates with different food sources, Currie determined that Escovopsis is a pathogen with a sweet tooth for only the ants' cultivar. What's more, the pathogen's evolutionary tree had the same basic shape as those of the ants and their crop, indicating that all three had coevolved since the beginning of ant agriculture, Currie and colleagues reported in 2003 (Science, 17 January 2003, pp. 325, 386). Although Currie isolated Escovopsis from up to 75% of the gardens of several attine species in Panama, this pathogen rarely seemed to do much damage. The reason, it turned out, was a fourth symbiont: Currie found that actinomycete bacteria, housed and nourished in pits on the ants' bodies, produce chemicals that keep Escovopsis in check. The four-part garden symbiosis of ant, cultivar, pathogen, and bacteria interrupted a scientific tradition of studying symbionts two at a time—think corals and algae, for example, or soybeans and nitrogen-fixing bacteria. The discovery accelerated a transition toward thinking of interacting organisms in trios or networks, not pairs. “When I got into this stuff, it was two symbionts,” says Ted Schultz, an entomologist at the Smithsonian National Museum of Natural History in Washington, D.C., who has studied attine evolution for nearly 30 years. “I was stunned” when Currie identified two more. Now, in a paper in this month's issue of Ecology, Currie and his colleagues introduce a fifth symbiont. “[It] just continues the trend of being repeatedly surprised by how complex this system is,” Schultz says. The first hint of the new player came when Currie, now at the University of Wisconsin (UW), Madison, cultured the actinomycete bacteria from an ant called Apterostigma. In addition to white bacterial spots, a black yeast often appeared on the same culture plates. Ainslie Little, now a postdoctoral fellow at UW Madison, took a closer look at the yeast. She treated worker ants in a vial coated with a selective antibiotic that would rub off on the ants and kill the yeast but not the other symbionts. At first, it looked like the experiment was a bust: Getting rid of the yeast had no effect on the ants or their crop. But when Little spritzed half the ants' gardens with a solution of Escovopsis spores, the yeast suddenly revealed its true colors. Over 3 days, ants with black yeast infections lost twice as much of their crop to Escovopsis as the yeast-free ants. When Little grew the actinomycetes in petri dishes with the yeast, the yeast ate the bacteria, demonstrating that they rob ants of an important defense against Escovopsis. DNA studies showed that the black yeasts are widespread on the attine ant family tree, thriving near the pits where the ants house the actinomycetes. Like the cultivar, Escovopsis, and actinomycetes, the yeast has been part of the attines' microbial balancing act since the ants first began to farm, Currie says. “It's really exciting,” Schultz says of the fifth symbiont. And Ulrich Mueller, an integrative biologist at the University of Texas, Austin, agrees: It's “interesting to what extent the presence of [another] symbiont can fundamentally change the interaction of two other symbionts.” ## No cheating allowed Currie thinks that understanding three-, four-, and five-way interactions like the ones in the fungus gardens may ultimately revolutionize the way we think about the evolution of mutualism. Why two parties should cooperate—whether it's two species over evolutionary time or two people over the course of a day—is one of science's big mysteries (Science, 1 July 2005, p. 93). In a two-player partnership, cheaters should be able to get ahead by reaping benefits without paying their dues, destabilizing the agreement. But the ants have lived stably with two mutualists for millennia. Moreover, the two antagonists rarely get too far out of line. Diseases and pests of humans and their crops, on the other hand, have evaded control measures in a matter of decades. So everyone, from crop scientists to basic evolutionary biologists, is itching to know the secrets to both cooperation and control in the ants' gardens. Morrissey, for example, would like to incorporate beneficial microbes into human agriculture to reduce chemical input. The attine system “shows it is possible to set up a structured [microbial] community … over a very long term” for biological control, he says. The “key question” for figuring out how to manage beneficial microbes, says R. Ford Denison, a crop ecologist at the University of Minnesota, St. Paul, is why the ants' actinomycetes never turn against their hosts. Antifungal compounds are expensive to make, and selfish bacteria that don't make the compounds should be able to reproduce faster and eventually overrun the helpers. But somehow the ant system is robust to cheaters. This is where the multipart interactions come in, Currie says. He thinks there is a reason that each partnership in the garden has a parasite: Escovopsis intrudes on the ant-fungus mutualism, and the yeast disrupts the antactinomycete mutualism. These parasites may be the very thing that keeps the mutualists cooperating, Currie adds. Currie and his colleagues tested this idea by forcing the fungus or the ants to cheat on each other. A selfish fungus would reproduce more and feed the ants less. To simulate this, Little removed specialized nutrient-laden structures from much of the cultivar. To shortchange the fungus, she reduced the proportion of fungus-tending workers in the colony. With no Escovopsis around, the negative effects of cheating were minimal, suggesting that it could become common over time. But add Escovopsis and cheating was a disaster, Currie reported last summer at the Gordon Research Conference on Microbial Population Biology in Andover, New Hampshire. When either the ants or the fungus were cheating, Escovopsis took over more of the garden, killing the fungus and leaving the ants with little to eat. Parasites, Currie says, may play a crucial and underappreciated role in keeping cooperators honest by raising the costs of cheating. But that doesn't explain what keeps the parasite itself from overrunning the fungus garden, the same way human pathogens have outpaced antibiotics. New results from Currie's lab, not yet published, indicate that Escovopsis does evolve resistance but never gets ahead because the actinomycetes themselves are so adept at evolving new antifungal compounds. And if they don't do it fast enough, the ants can acquire new actinomycete strains. But other organisms, yet to be described, may also play a part in stabilizing fungus garden ecology. “There are additional biofilms associated with [the fungus], stuff that grows on the fungus, or in the substrate, wherever anything else can move in,” says Mueller. When the ants transplant a garden, they do so by choosing a little piece of the existing garden, including any other microbes that are mixed in with it. “In a sense, they're selecting on an entire community that has desirable properties,” Mueller says. The idea still needs to be tested. Meanwhile, only the daring would place bets on how many symbionts have yet to turn up. Currie expects a few more; Mueller predicts hundreds. Evolutionary biologist Jacobus Boomsma of the University of Copenhagen in Denmark says that regardless of the number of symbionts, the fungus gardeners are poised to answer critical questions about cooperation, conflict, and microbial ecology. “This system,” he says, “will keep inspiring us for at least a decade more.” 11. ASTROPHYSICS # GLAST Mission Prepares to Explore the Extremes of Cosmic Violence 1. Yudhijit Bhattacharjee NASA's new gamma ray observatory will probe the most energetic radiation ever studied, the product of cataclysmic events deep in space. NASA's new gamma ray observatory will probe the most energetic radiation ever studied, the product of cataclysmic events deep in space In July 1967, U.S. surveillance satellites looking for signs of a Russian nuclear test in space recorded two flashes of gamma radiation. Scientists quickly determined that the high-energy bursts did not come from a nuclear explosion, which would have generated a more sustained stream of gamma rays and also produced lower energy radiation detectable by other satellite instruments. Only years later did they realize that the flashes—named gamma ray bursts (GRBs)—originated in violent events deep in space. In scanning the heavens for an enemy secret, they had stumbled upon a cosmic one. That serendipitous discovery opened a window on previously unknown phenomena whose signatures lay at the gamma end of the energy spectrum. Since then, astronomers have dispatched a number of gamma ray telescopes into space to glimpse the pyrotechnics unleashed by violent events such as collisions between neutron stars and the emission of particle jets by massive black holes. Now, researchers are opening the window wider with a new telescope designed to record gamma radiation several orders of magnitude higher in energy than current instruments can detect. NASA's Gamma-ray Large Area Satellite Telescope (GLAST), scheduled for launch next month, will also be the first instrument of its kind to survey the entire sky several times a day, increasing the chances of finding and following extreme astronomical phenomena anywhere in the universe. Researchers say GLAST's powerful combination of sensitivity and sweep will yield a rich harvest of data that could answer a host of astronomical questions such as how super-massive black holes behave and how cosmic rays originate. A tantalizing possibility is that observations from GLAST will help physicists discover the fundamental nature of dark matter, which makes up 10 times as much of the universe as the familiar matter of planets and stars. The mission represents a convergence of the quests to understand the large and the small, says Rene Ong, an astronomer at the University of California, Los Angeles. “We are just at the start of a very exciting time in astronomy and fundamental physics,” he says, noting that unlike data from most previous missions, GLAST's will be available in real time to scientists—and the public— anywhere in the world starting a year after the launch. ## Catching rays Built over a decade at a cost of$690 million, GLAST is a feat of engineering. Its main instrument is a 3-ton detection system called the Large Area Telescope (LAT), which consists of two devices to track the direction and energy of incident gamma rays—energetic photons of extremely high-frequency electromagnetic radiation. The direction detector is a four-by-four matrix of towers that are essentially layers of tungsten and silicon stacked one on top of another. When a gamma ray slams into a tungsten layer, there's a chance its energy will be transformed into an electron and a positron that are propelled along the path the ray would have taken. As the two particles travel through silicon layers in the stack, they generate currents that reveal their direction. When they emerge from the bottom of the stack, the particles enter a chamber of cesium iodide—the telescope's energy-detecting device—producing a flash of light whose intensity shows how fast they had been moving and thus the energy of the gamma ray.

The stacked design gives LAT a much larger collecting area than previous gamma ray telescopes such as the Energetic Gamma Ray Experiment Telescope (EGRET), which flew on NASA's Compton Gamma Ray Observatory from 1991 to 2000. More collecting area means increased chances of a collision. And that's exactly what's needed in order to detect higher energy gamma rays, explains Steven Ritz, GLAST project scientist at NASA's Goddard Space Flight Center in Greenbelt, Maryland, because they are so rare that a less sensitive instrument would miss them. As a result, LAT can detect gamma rays of up to 300 billion electron volts, 10 times EGRET's upper limit (see figure).

GLAST has a second instrument designed to detect lower energy gamma rays that LAT would not register. Called the GLAST Burst Monitor (GBM), it's a set of 12 sodium iodide disks and two bismuth germanate disks pointed in different directions, covering practically the entire sky. The disks produce light when struck by photons at the lower end of the gamma spectrum; scientists can trace the direction of the incident rays simply by noting which disk bears the brunt of the collision. GBM will detect rays between 10 KeV and 25 MeV, overlapping with LAT's lower limit of 20 MeV. “Together, the two instruments give us vast energy coverage,” says Ritz, adding that “if GLAST were a piano, it would have 23 octaves.”

One of the challenges in designing the system was to ensure that it would run on the small amount of power available from the satellite's solar panels. Robert Johnson, a physicist at the University of California, Santa Cruz, who led the engineering of the tower array, says simplifying the electronics was part of the solution. “We ended up at 160 watts,” he says. “That's a couple of light bulbs of power for over 900,000 channels.”

## Black hole, bright lights

Astronomers will be eagerly scanning GLAST data for clues to what goes on near monstrous black holes that sit at the centers of galaxies. Such objects can be as massive as hundreds of thousands or even billions of stars. As their enormous gravity sucks matter into a whirling disk around them, opposing jets of particles shoot away from their poles at nearly the speed of light. If a jet from such an active galactic nucleus (AGN) happens to be pointed at Earth, astronomers call it a blazar. The process generates radiation across the entire electromagnetic spectrum, including high-energy gamma rays.

The Compton Observatory identified 66 blazars during its time in orbit. Astronomers have since puzzled over how these beasts accelerate particles to such high speeds. GLAST is expected to find thousands of new blazars because of its sensitivity and periodic surveying of the sky, and the data it sends back should provide a sharper, more dynamic picture of these events than Compton did, says Ritz. He expects blazars and AGN to be a “bread and butter” topic for researchers analyzing GLAST data.

Alan Marscher, an astronomer at Boston University, agrees. Last month in Nature, Marscher and colleagues presented x-ray, radio, and visible light observations from a blazar, suggesting that the “accretion disk” spinning around the black hole had caused the magnetic field in the galactic center to coil into a spiral, leading to the emission of particulate jets from its core. He says GLAST will help him test the theory by observing how the brightness of gamma radiation from different blazars changes over time. “We expect time delays between the peaks in the flares at different gamma ray energies and relative to the flares in x-ray, visible-light, and radio emission,” Marscher says. That signature, he says, could offer a deeper look into the heart of a blazar.

## Outracing LHC?

Researchers involved with GLAST call the mission a unique marriage between particle physics and astronomy. Some are hoping to justify that description in grand fashion in the years to come by carving out a prominent role for GLAST in finding the elusive particle that constitutes dark matter.

As its name implies, astronomers cannot see dark matter; only its gravity gives it away. Many theorists think it consists of still-unknown “weakly interacting massive particles” (WIMPs). Detecting WIMPs is one of the goals of the Large Hadron Collider (LHC), the \$5.7 billion underground accelerator at CERN that is expected to come on line this summer. But if the hypothesized particles do turn up there, physicists will still need to confirm that they make up the dark matter out in space. That's where GLAST would come in.

According to theory, in the rare event when two WIMPs collide, they annihilate each other and give off gamma rays. Such collisions are most likely in galactic regions where dark matter is densely concentrated.

“We would look in those known directions to see if GLAST is picking up an excess of gamma rays,” says Johnson, who started his career as a particle physicist before being completely “consumed” by the GLAST project a decade ago. He's now the co-convener of the mission's dark matter science working group. “The dream scenario is that we see the signature of dark matter before LHC turns on,” he says with a chuckle. “But we'd be perfectly happy if they saw it first and determined the particle's mass, and then we went out and found it in space.”

By the same token, observations from GLAST could help LHC in its quest to identify WIMPs, says Dan Hooper, a theoretical physicist at Fermi National Accelerator Laboratory in Batavia, Illinois. He says LHC will generate such a huge volume of data that any hint from GLAST about the energy released by an annihilating WIMP pair would help LHC scientists to focus their search. “If you are doing this needle-in-a-haystack search, knowing how big the needle is could be key,” he says. However, Hooper cautions that gamma rays produced by WIMP collisions could turn out to be too faint for GLAST to see.

When the data start streaming in, researchers will be able to sink their teeth into them for insights into other fundamental problems. For example, says Neil Gehrels, deputy project scientist, GLAST might catch small, primordial black holes in a vanishing act, confirming a prediction by Stephen Hawking that such objects shrink by emitting radiation and eventually evaporate into a crematory flash of gamma rays.

“It could well be that the most interesting observations turn out to be something entirely new and unexpected,” says Ritz. If that were to happen, as witnesses to the discovery of GRBs can testify, it would not be the first time.

12. # The Inner Lives of Sponges

1. Gretchen Vogel

Symbiotic ties, bioactive compounds, and mysterious distributions of bacteria characterize these ancient invertebrates.

Symbiotic ties, bioactive compounds, and mysterious distributions of bacteria characterize these ancient invertebrates

A spongeful of bacteria is the last thing a dishwasher wants to think about. But for Jörn Piel, the more microbes he finds in a sponge, the better. Not a synthetic one, of course, but those that adorn tropical reefs and populate the ocean bottom.

One of evolution's more ancient animals, sponges at first glance seem quite simple—little more than loose consortiums of semiautonomous cells, stuck in one place filtering food from the water column. But a closer look reveals a surprising twist. “With many species, under the microscope you see almost exclusively bacteria” among the cells, says Piel, an organic chemist at the University of Bonn in Germany. Just as microbial ecologists are demonstrating the extent and importance of microbes in ecosystems as diverse as guts and glaciers (see p. 1046), Piel and others are slowly uncovering a hidden microbial world inside sponges.

It's a difficult job, as almost none of the sponges' inhabitants grow in the lab. But through genetic studies, researchers are revealing the rich diversity and unusual distribution of these microbes. Some investigators are pinning down the roles bacteria play in sponge biology and ecology. The microbes are teaching scientists about evolution, symbiosis, and the mind-boggling variety of life on our planet. “It never ceases to amaze me that a sponge, an organism that just sits there and pumps bucketfuls of water through its canals,” has such a rich and varied, yet highly specific, inner life, says marine microbiologist Michael Taylor of the University of Auckland in New Zealand. The research also has a practical side: Piel and others are betting that sponge-dwelling bacteria could be the source of potentially valuable compounds for treating cancer, malaria, and other human diseases.

## Sharper focus

The first hint of the sponges' pervasive inhabitants came in the 1960s and '70s, as new equipment allowed longer and deeper dives that gave researchers their first up-close look at the diversity of life on the ocean bottom. It quickly became clear that something else was living among the sponges' cells. Looking at the first electron microscope images of sponge tissue, marine biologist Jean Vacelet and his colleagues at the University of Marseille spotted what looked like a half-dozen different types of bacteria. Other researchers “took sponges and squeezed them out over culture plates to see what would grow,” recalls marine ecologist Robert Thacker of the University of Alabama, Birmingham, but it was difficult to follow up on the finds. At most, 5% of sponge-dwelling species have thrived in the lab, says microbial ecologist Ute Hentschel of the University of Würzburg in Germany. And the sponges themselves “are incredibly hard to keep alive,” Thacker says.

Therefore, Hentschel, Thacker, and others have been using indirect methods to piece together a picture of this reclusive community. Most of the evidence comes from studies of the gene for 16S ribosomal RNA (rRNA), a piece of the genome that scientists use to identify unknown microbes in the environment. Differences in this gene can serve as a useful measure of the kinship between two species.

These genetic studies uncovered a distinctive and extensive community, identifying more than 100 species of microbes that are found in sponges but not in the surrounding water. This distribution indicates that these bugs are long-term residents rather than passersby. An individual sponge might host dozens of different species, and overall, the molecular analyses have found an impressive variety: 14 bacterial phyla, two phyla of archaea, and several types of eukaryotic microbes.

Such diversity initially suggested multiple, independent acquisitions of microbial symbionts. But evidence is building that sponges of different types and in different oceans host strikingly similar microbial communities. Hentschel and her colleagues showed in 2002 that sponges from the coast of Japan, the Red Sea, the Mediterranean, and the Republic of Palau in the South Pacific contained microbes that are more closely related to each other than to the microbes in the seawater from which the sponges were harvested. “It's astounding,” says Susanne Schmitt, a postdoc in Hentschel's lab. The different sponges the scientists sampled diverged millions of years ago, she says, but they are home to very similar, and very complex, microbial communities. In 2007, Taylor and his colleagues found the same result when they analyzed the entire database of 16S rRNA sequences available from sponge-dwelling microbes collected from all over the world—nearly 2000 sequences in all.

But Taylor, Hentschel, and their colleagues are still trying to work out what the results mean. Microbes might have colonized a sponge early in the group's evolutionary history and acquired characteristics that enabled them to live in sponges full-time, Taylor proposes. Those sponge-loving microbes could have then spread to other sponges—and other oceans. And such a scenario could explain what may be a new phylum called Poribacteria, after Porifera, Latin for “sponge.” Poribacteria have been found throughout the world, albeit exclusively in sponges.

## Fruitful partnership

As with much of microbial ecology, the sponge specialists have been focused primarily on taking a census. “I go in just trying to figure out what's there—what people did collecting insects in the forest 100 years ago,” Taylor explains. But he and his colleagues are now starting to take the next step, because census data can't tell researchers what each side gets out of the relationship. Ecologists want to know if the microbes and their hosts are obligate symbionts, unable to survive without each other, or whether the microbes are tolerated but dispensable guests, says Michael Wagner, a microbial ecologist at the University of Vienna in Austria: “If we want to understand these communities, we have to know the function each member plays.”

Yet even after decades of study, scientists are still not exactly sure what sponges and their microbes are doing for each other. Living in nutrient-poor but sunlit waters in the lagoons of the Republic of Palau, sponges of the family Dysideidae are home to blue-green algae that probably provide their hosts with energy and carbon. The sheer mass of the microbes may help support the meter-high giant barrel sponge Xestospongia muta, in which bacteria can sometimes make up 40% of a sponge's volume. Microorganisms may even help defend their hosts against disease-causing bacteria.

But those are educated guesses rather than proven observations. “And there are a whole lot of other things that are going on that we just don't know about,” says molecular ecologist Russell Hill of the University of Maryland Biotechnology Institute in Baltimore.

To try to get a picture of the daily goings-on inside a sponge, Wagner and his colleagues are catching sponge microbes in the act of “eating.” The researchers have just started experiments on several species of sponges that host Poribacteria. No Poribacteria have ever been cultured in the lab, but the scientists are able to keep the host sponges alive in aquaria, at least for a short time. They use a technique that allows them to observe the metabolic activity of individual microbes and sponge cells. They expose the sponge to fluorescently labeled rRNA markers, which lets them know what species they are dealing with, and to radioactively labeled “food”—amino acids, bicarbonate, and other molecules. They then watch which cells take up the labeled morsels and follow how the morsels are processed, including whether the sponge consumes compounds excreted by the microbes. “We're asking not only ‘Who are you?’ but also ‘What are you eating?’ “he says.

Whatever their function, the microbes seem important enough for sponges to pass on to future generations. In the female sponge, nurse cells, which provide the “yolk” for developing eggs, also ferry blue-green algae from the sponge's outer layers to the developing oocytes located deeper in the sponge matrix. In 2005, Kayley Usher and her colleagues at the University of Western Australia in Perth even found blue-green algae in the sperm of the sponge Chondrilla australiensis. A year later, Julie Enticknap, a postdoctoral fellow in Hill's lab, was able to culture a sponge-dwelling alphaproteobacterium from the larvae of a sponge collected off the coast of Florida, another indication of possible parent-to-offspring transmission.

But that study highlights what may be the most baffling mystery in sponge microbiology. Usually when symbionts are passed from parent to offspring, the partners undergo what is called cospeciation, and the microbes develop a unique genetic signature and become confined to that particular host. “But that doesn't happen here,” says Hentschel. The bacteria in the larvae proved closely related to those cultured from unrelated sponges growing in Jamaica, Indonesia, and the Chesapeake Bay in the United States. The best explanation for the broad distribution of this bacterium—and for many other species found across the globe—she says, is that sponges acquire their resident bacteria both from their parents and from the environment.

To date, no sponge-specific microbe has turned up in seawater, but scientists have a distinct disadvantage when it comes to sampling. Although a 1-kilogram sponge can filter thousands of liters of seawater a day, Hills says that “if we are lucky, we filter 200 liters,” so the chances of finding an uncommon microbe, such as the larvae's alphaproteobacterium, are small.

If sponges are taking microbes in from the surrounding environment, they need to be able to tell friend from foe from food. And the microbes need a way to protect themselves against accidental or intentional rebuff by their hosts. Electron microscope images reveal that most sponge-dwelling bacteria have either thickened cell walls or slime capsules that might prevent the sponge cells from digesting them. Once established, these resident microbes, or the sponge itself, seem to produce chemicals that discourage interlopers. Several antibiotic compounds isolated from sponges efficiently kill bacteria found in the water column but do not affect sponge-dwelling organisms.

## Sea-based drugs

Those antibiotic compounds are driving at least some of the interest in sponges. For bio-prospectors looking for potential new drugs from the sea, “sponges are one of the best sources of bioactive compounds,” says Hill. The chemical from which the antiretroviral drug AZT was derived was first found in a Caribbean sponge. In the lab, other compounds from sponges kill cancer cells and malaria parasites.

AIDS, cancer, and malaria are not the sponge's concern, but a powerful chemical defense arsenal is, points out microbial ecologist Julie Olson of the University of Alabama, Tuscaloosa: “Sponges can't evade predators, and if something blocks their [water-filtering] passages, it's a death sentence.” To protect against unfriendly microbes, a successful sponge probably needs a range of chemical weapons, she says.

Initially, marine biochemists using the “grind and find” approach assumed that most of those chemicals came from the sponge itself. But as the diversity of the sponges' residents became clear, many began to suspect that at least some of the compounds might come from the lodgers rather than the hosts. The quest to come up with enough of a bioactive compound for clinical testing is proving that these suspicions are well-founded.

To date, few sea-based drugs have made it to the clinic. “Supply is the primary reason there is no blockbuster so far,” says Piel. It's been almost impossible to purify a compound in quantities large enough for animal and human testing, and the chemical structures are often too complex for large-scale chemical synthesis. Halichondrin B, for example, is a powerful antitumor compound in lab tests. But scientists calculated that clinical trials would require at least 10 grams of the substance. The best producer, a New Zealand sponge group called Lissodendoryx, yielded 300 milligrams per metric ton of sponges. Because the entire population of Lissodendoryx was estimated at 280 metric tons, it was clear that harvesting from the wild was not sustainable, Piel says.

Piel is trying to get around that problem. His lab is fishing for genes involved in making promising compounds. The goal is to find the set of genes that codes for the potential drug's synthesis, and if the original host won't grow in the lab, to transfer those genes to a microbe that is happy in an artificial environment. The designer microbe would then pump out enough of the drug for testing.

The technique is pointing to bacteria as the source of many of the compounds. “So far, every time we've found a gene cluster, we found typical bacterial genes” nearby, Piel says. In 2004, the group reported that they had pinned down the genes responsible for producing compounds called polyketides in a dark-red sponge called Theonella swinhoei that lives in coral reefs. (In the lab, polyketides can kill tumor cells, and several types are in clinical trials.)

Based on the genes' similarity to known genes, Piel and his colleagues concluded that the genes most likely come from a still-uncultured microbe. What's odd, however, is that these genes are quite similar to polyketide genes belonging to a bacterium that lives in the guts of beetles. The researchers are currently working to transfer the sponge microbe's genes to a lab-friendly host.

Hill and his group have focused on trying to harness the original bacteria producers. “Part of the problem is that people have in their heads that all symbionts are difficult to grow,” Hill says. But patience and hard work can pay off. “Sometimes we get new colonies after months of incubation.”

In recent work, Hill's lab has homed in on the source of a particularly promising compound called manzamine A. In lab tests, manzamines kill malaria parasites more efficiently than either chloroquine or artemisinin, two of the leading antimalarial drugs. The compound was first identified in a sponge collected off the coast of Okinawa, but related compounds have since turned up in dozens of unrelated sponge species all over the world—a strong hint, Hill says, that they are produced by a microbe shared by all these species.

In as-yet-unpublished work, his group has isolated the bacterium that produces manzamine A. The microbe should give scientists their first steady supply of the compound, allowing them to make and test new derivatives, Hill says. Such studies led to the eventual development of AZT, Hill points out. And he is hoping sponges—or at least their microbes—will again lend a hand in the fight against deadly disease.

13. # Confusing Kinships

1. John Bohannon

Understanding microbial evolution and ecology rests on a solid classification system, but coming up with one is difficult.

Understanding microbial evolution and ecology rests on a solid classification system, but coming up with one is difficult

Along the parched slopes of a canyon in Israel, in salty pools in Spain and Algeria, and on countless petri plates in laboratories around the world, a scientific debate is playing out: how to classify microbial organisms accurately. Not only are bacteria and their ilk amazingly diverse, but genes cross species lines so frequently that researchers argue whether microbial species exist at all. At stake is much more than the esoteric record-keeping of taxonomists, says R. Thane Papke, a microbiologist at the University of Connecticut, Storrs: “This is about our fundamental understanding of evolution.” Without rigorously categorizing diversity, “we're really stuck.”

The problem is that prokaryotes, the single-celled organisms without a nucleus, are promiscuous. Instead of one cell splitting into two genetically identical daughter cells, over and over, most take part in a global orgy of gene swapping, passing genes between different taxa. This spells trouble for traditional systematics, built as it is on the assumption that organisms' genes faithfully reveal their common ancestry. Whereas most genes in a particular microbe do come from its direct ancestor, many may not, making lines of descent difficult, if not impossible, to describe (Science, 1 May 1998, p. 672).

This is not the first time that microbial systematics has been disrupted. The tremors began 4 decades ago when DNA sequence became the gold standard for classifying organisms, revolutionizing our understanding of how microbes fit into life's grand scheme. Now, the availability of whole genomes and DNA sequence from complete microbial communities is shaking up the field anew. “An earthquake is coming for microbial systematics,” says Hans-Peter Klenk, a microbiologist at the German Collection of Microorganisms and Cell Cultures (DSMZ) in Braunschweig. And after the dust settles, what microbial “species” will look like is anyone's guess.

## What's in a name?

Classifying microbes has never been easy. Well into the 20th century, bacteria were considered members of the fungi, themselves erroneously classified as plants. At the first International Microbiological Congress in Paris in 1930, scientists decided that microbes needed their own scheme. At that time, members of the new Commission on Nomenclature and Taxonomy called microorganisms “in part plants, in part animals, and in part primitive.” They therefore concluded that these single-celled creatures belonged with neither.

From then on, microbiologists assigned names to the organisms they found wriggling and dividing under their microscopes based on the few characteristics that could be reliably observed. Whether microbes irreversibly soak up a dye called crystal violet designated them as Gram-negative—such as our common gut inhabitant, Bacteroides fragilis—or Gram-positive, such as the food-poisoning microbe Clostridium botulinum. The Bacillus species were distinguished by their need for oxygen and ability to form spores. Others, such as Streptococcus pneumoniae, earned their monikers based on the diseases they caused. “But everyone understood that these species definitions were rather arbitrary,” says Klenk, because there was no way to confirm that they reflected genetic relatedness. Nonetheless, “it was a practical system that could help microbiologists know what they were talking about.”

And help it did. Microbiology exploded during the second half of the 20th century, transforming every field it touched, including medicine, agriculture, ecology, and even geology. The enterprise was built on an ever-growing microbial family tree and “type cultures” of each microbial species, representative batches kept in laboratories around the world. Type cultures made it possible for researchers to replicate and build on previous experiments, says Klenk.

By the late 1970s, there were some 40,000 type cultures. “And that's when we had our first big shock,” says Klenk. Using the newly available tools of molecular biology, scientists compared DNA sequence from different microbial species and found that “we were completely wrong about evolutionary relationships.” The prokaryotes split, some staying in the familiar Bacteria and others shifting into Archaea, single-celled organisms that are superficially similar to bacteria but whose genetic architecture is more like our own.

Starting in 1980, the naming of microbial species went in for a complete overhaul by the International Committee on Systematics of Prokaryotes. The standards for type cultures were made far more rigorous, says Klenk, calling for DNA sequence data and more thoroughly documented isolation. The 40,000 type cultures were pared down to about 4000—growing since then to 6800.

But DNA studies have continued to muddy the microbial waters. “In some ways, the more genetic data arrives, the less clear things get,” says Christophe Fraser, an epidemiologist at Imperial College London.

One problem revealed by DNA is the vastness of microbial diversity, says Fred Cohan, a microbiologist at Wesleyan University in Middletown, Connecticut. “All of microbiology was based on what we could culture in the lab,” with heavy emphasis on pathogens. But lab-culturable bugs turn out to be “certainly less than 1%” of living microbes. As technologies improved, ever more DNA sequence has been harvested from environmental samples, representing all the microbial genetic material in a pinch of soil or milliliter of seawater. The realization that “a soil community contains tens of millions of bacteria” is “humbling,” says Cohan.

Sequencing of whole genomes has presented microbiologists with an even more daunting challenge: Microbes have an active “sex” life. Scientists have long known that genes can move between microbes; the spread of antibiotic-resistance genes since the 1950s is a case in point. “But what's surprising is how frequent and widespread it turns out to be,” says Fraser.

For traditional species to be well-defined, their genes need to flow vertically, from parents to offspring and nowhere else. But among microbes, genes can move along a bewildering variety of routes between genomes: sliding through bridges between cellular membranes, hitchhiking inside viruses, or even getting sucked up from the environment as naked fragments. That means that any given microbe can have a large number of “parents” from many different species. But “if the genes are moving freely, then how can you nonarbitrarily define the relations between different microbes?” says Papke. “Do you really have species at all?”

## The earthquake begins

Yet Fraser and others haven't given up on classifying microbes. They think there are other ways to define “species.”

Take recombination, for example. In plants and animals that have no choice but to reproduce sexually, recombination happens every generation: Matching strands of DNA on chromosomes line up and swap segments, producing offspring with a shuffled deck of genes from each parent. This can happen within microbes, too. When foreign DNA finds its way inside a microbe's cell membrane, it has a chance of lining up with a similar sequence and swapping segments with the genome. Such recombination events happen infrequently, but when they deliver a more useful version of a gene, the recombined genome can even sweep through a population to become the norm. DNA sequencing of many microbes from the same population has revealed that recombination is much more common than was ever thought.

But the “crucial insight” is that the frequency of recombination depends on the kinship between the donor and recipient microbes, says Fraser. The more closely related two microbes are—and hence, the greater the similarity between their genomes—the greater the chance that recombination will happen. When the microbes are too different, “the recombination rate drops off steeply,” he says, effectively blocking gene flow. “If we can define this threshold, then that could be a rigorous way to define microbial species.”

Last year, Fraser and colleagues used computer simulations to study microbial evolution (Science, 26 January 2007, p. 476). The results suggested that when the rate of recombination is high enough, genetically isolated groups can emerge that are “analogous” to traditional species, he says.

Recent evidence for Fraser's view has come from a study of real-world microbial recombination (Science, 11 April, p. 237). In this case, a species is being lost. A team led by Martin Maiden, a microbial geneticist at the University of Oxford, U.K., studied two species of Campylobacter bacteria. The genomes of C. coli and C. jejuni share only 86.5% of their most conserved DNA sequence, due to millions of years of adaptation to different wild host animals. But in the 10,000 years since the advent of agriculture, the two species have been living together cheek-by-jowl in farm animals, and there, recombination is on the rise, blurring what in other environments is a clear species line. Because recombination is occurring nearly 20 times faster in the genome of C. coli, that genome is becoming ever more like C. jejuni's, and Maiden predicts that the two will eventually become indistinguishable.

But other tests of recombination as a gold standard for microbial species have not yielded such clear results. Papke was part of a team that studied microbes in three hypersaline pools, two in Spain and a third 250 kilometers away in Algeria. The salt-loving microbes isolated in each pool, all members of the genus Halorubrum, should represent separate species. Yet among the 153 strains tested, there was a bewildering degree of recombination—even between “species” in pools separated by the Mediterranean, Papke and his colleagues reported last August in the Proceedings of the National Academy of Sciences (PNAS). Considering their blush-worthy promiscuity, “we'll just have to accept that microbes evolve in ways that don't allow them to be pigeonholed into species,” says Papke.

## The eco-challenge

But some researchers argue that similarity in lifestyle, not just genes, is the way to classify microbes. Even with DNA flowing willy-nilly, microbes pigeonhole themselves into coherent groups by adapting to different niches, says Cohan: “The key to understanding microbial diversity is ecology.” Cohan and others would like to do away with microbial “species” as the “fundamental unit” of diversity. Instead, microbes would be divided into “ecotypes,” based first on genetic relatedness and more finely on shared adaptations to a particular habitat.

To demonstrate the existence of ecotypes among real-world microbes, Cohan and a team led by David Ward, a microbial ecologist at Montana State University in Bozeman, have studied Bacillus bacteria from a group of arid canyons in Israel. The bacteria have adapted to the canyons' various microenvironments, says Cohan, from the harsh, dry northern slope to the relatively mild, lush southern slope, and a periodically flooded streambed between. The team isolated 218 bacteria from these locations, all members of the genus Bacillus, and sequenced the DNA of five genes from each. On the basis of their highly conserved 16S ribosomal RNA gene sequences, bacteria from one canyon divided neatly into two species, B. simplex and B. licheniformis. But stopping there “would not be informative,” says Cohan, because the sequences of other genes revealed substantial diversity between bacteria with identical 16S sequence. That diversity is driven by adaptation to different microhabitats, says Cohan.

So Ward and Cohan's team devised a method called “ecotype simulation” to categorize the canyon microbes based on lifestyle. First, they generated a phylogenetic tree based on variation in the four marker genes that were different in the various bacteria. Then, the simulation comes up with putative ecotypes: It divides the bacteria into clusters of genetically similar individuals, all the while checking to make sure that the similarities are due to adaptive changes and not to chance. Finally, the researchers use what is known about the actual environment of each microbe—solar exposure, moisture, soil type, plant resources—to test whether the predicted ecotypes correspond to various microhabitats.

The team identified as many as 30 distinct ecotypes across all canyons studied. Rather than species, these are the “fundamental units” of microbial diversity, the team concluded in a paper published in PNAS 19 February. Cohan says they are now “preparing to propose” some of these ecotypes, such as Bacillus simplex ecospecies Graminifolius, to the systematics community for formal recognition.

Ward and Cohan aren't the only researchers using ecology to make sense of microbial diversity. A study of marine microbes led by Martin Polz at the Massachusetts Institute of Technology in Cambridge found that groups of bacteria of the same species occupy different parts of the plankton community, and even during specific seasons (see p. 1081).

But an ecology-based classification faces an uphill battle for acceptance. “I just don't see how ecotypes can work,” says Papke. He thinks gene swapping is so frequent among microbes that recombination, not ecological adaptation, is the main cause of diversity. Ecotypes “still have a way to go,” agrees Ford Doolittle, a microbiologist at Dalhousie University in Halifax, Canada.

Doolittle is pessimistic that anything better than a “compromise solution” can ever be achieved for microbial systematics. When conditions are just right, microbes “may cluster into what we could all agree to call species,” says Doolittle, on the basis of either “ecotypes” or rates of recombination. “But there is no reason to suppose that conditions will often or even ever be right and thus no reason to suppose that there must be ‘fundamental units’ of bacterial diversity.” Traditional systematics requires such units, he says, “but needing something to be true does not make it so.”

And how will all this affect day-to-day microbiology? “I expect species names and type cultures will continue,” says Klenk. Papke agrees: “We need to be able to have a conversation.” But as for the “fundamental unit” of microbial diversity and what, if anything, is represented by microbial species, “we'll probably need philosophers to sort that out.”