# News this Week

Science  20 Jul 2012:
Vol. 337, Issue 6092, pp. 274
1. # Around the World

1 - Tanami Desert, Australia
Conservation in the Desert
2 - London
U.K. Moves Toward Open Access
3 - Washington, D.C.
Panel Pushes to Scale Down Agro-Defense Lab
4 - Paris
Scientists Protest Canada's Budget Cuts as Antiscience

## Tanami Desert, Australia

### Conservation in the Desert

The Australian government has designated 10 million hectares of land in the Tanami Desert of the Northern Territory as a conservation zone—the largest in the country. The Tanami Desert is home to many endangered species, including the bilby, a long-eared marsupial (shown), and the great desert skink, a burrowing lizard that builds elaborate tunnel systems. The animals have been threatened by feral cats and foxes, as well as brush fires and cattle grazing.

The region will be administered as an Indigenous Protected Area, under the supervision of indigenous communities. The government plans to spend $1.6 million over the next 2 years to fund the rangers; the Nature Conservancy will also contribute$500,000 to help manage the area.

## London

### U.K. Moves Toward Open Access

The U.K. government has accepted the recommendations of a report released last month that encouraged greater open access publishing of publicly funded research. Officials announced on 16 July that over the next year, they will work to require studies wholly or partly funded by tax-payers to be published in journals that are free to readers. The government's decision means that by 1 April 2013, all papers from government-funded research must be published in an open access journal; if not, the journal must make the paper open access after 6 months.

The original report, chaired by sociologist Janet Finch, recommended the open access publication requirement because it would maximize the economic benefit of government-funded research. But the move is getting a critical response from some researchers and universities who fear it could increase costs. The government's decision could also mean extra costs for U.K. universities, critics say, because in addition to paying article processing charges, research and university libraries will also have to continue paying for subscriptions until other countries adopt similar policies. http://scim.ag/Finchrep

## Washington, D.C.

Earlier this year, the Obama Administration suspended plans for a $1 billion, highly secure laboratory for studying dangerous animal diseases, called the National Bio and Agro-Defense Facility (NBAF). Now, a new study from the U.S. National Academies suggests rescuing the facility by scaling it down. NBAF was intended as a replacement for the aging Plum Island Animal Disease Center in New York. But amid controversies—including environmental groups' concerns for the site's planned location in Manhattan, Kansas, over the potential release of pathogens—the Department of Homeland Security (DHS) has struggled to secure funding from Congress, and the estimated cost of the facility grew from$450 million to more than $1 billion. DHS had asked the National Academies to study whether NBAF's mission could be achieved by continuing with the Plum Island center or relying on biocontainment labs in other countries. The study concludes that neither option can enable the United States to protect animal and public health in the long term; instead, it recommends reducing the price tag for NBAF by scaling down activities such as training, research on small animals, and vaccination research, and outsourcing them to existing animal pathogen labs. http://scim.ag/AgroLab ## Paris ### Researchers Receive Controversial UNESCO Awards A tumultuous and divisive episode at the United Nations Educational, Scientific and Cultural Organization (UNESCO) ended Tuesday when three researchers received a life sciences award sponsored by Teodoro Obiang Nguema Mbasogo, the dictator of Equatorial Guinea. A broad coalition of human rights advocates, scientists, and health experts had fought almost until the last minute to prevent the ceremony from happening. They say the award is an attempt by Obiang to buy credibility for his regime, which stands accused of human rights violations. The winners of the prize are Maged Al-Sherbiny from Egypt, for his research on vaccines and diagnostics against hepatitis C and schistosomiasis; plant scientist Felix Dapare Dakora from Tshwane University of Technology, Pretoria, in South Africa for his work on legumes and soil bacteria; and Rossana Arroyo of the Centre for Research and Advanced Studies of Mexico's National Polytechnic Institute, who studies trichomoniasis, a parasitic disease. All three were at the ceremony, which was boycotted by many Western countries. UNESCO Director-General Irina Bokova, who opposed the prize but was overruled in March by UNESCO's 58-member Executive Board, didn't attend either. http://scim.ag/Obiangaward ## Ottawa, Canada ### Scientists Protest Canada's Budget Cuts as Antiscience Led by the Grim Reaper, a mock funeral procession of more than 2000 scientists marched on Canada's Parliament last week to protest what organizers called a “systemic attack on science” by the country's Conservative government. The “Death of Evidence” rally was organized by students at the University of Ottawa to protest budget cuts that will close the Experimental Lakes Area in northwestern Ontario and the Polar Environment Atmospheric Research Laboratory, two key environmental research programs; eliminate funding for a national science adviser; and convert the mandatory long form of the Canadian national census into a voluntary one, among other changes (Science, 29 June, p. 1627). The rally was timed to coincide with a large, international meeting of evolutionary biologists, many of whom joined the march. “Recent actions by the federal government suggest that our state is frightened by evidence and is retreating into a fantasy world,” says Arne Mooers, a biologist at Simon Fraser University, Burnaby. The Canadian government says the cuts are necessary to balance the budget and fund other, more-applied areas of science. 2. # Random Sample ## They Said It “I can't find anyone in my peer group who believes in BioWatch. … The only times it goes off, it's wrong. I just think it's a colossal waste of money. It's a stupid program.” —Ned Calonge, former chief medical officer for the Colorado Department of Public Health and Environment, to The Los Angeles Times on 7 July about the Department of Homeland Security's efforts to equip cities with devices to detect a biological terrorist attack. ## Swan Song for the Cernettes A woman in a puffy pink prom dress croons plaintively; three backup singers with beehive hairstyles sway and sing. Behind them all, a familiar giant particle accelerator curves away into the distance. “You never spend your nights with me. … You only love your collider.” “Ooooh … wop wop.” For 22 years, Les Horribles Cernettes—CERN's favorite 1960s-style girl group—have been serenading Geneva-based physicists with songs such as “Strong Interaction,” “Antiworld,” “Mr. Higgs,” and the aforementioned “Collider,” all written by the band's creator, manager, composer, and keyboardist/bassist, Silvano de Gennaro. But on 21 July, the Cernettes' era will come to an end, with one last performance at the CERN Hadronic Festival. (The show will be available on YouTube, http://www.youtube.com/user/cernettes). It all started in 1990, when a CERN secretary—tired of feeling neglected by her always-on-shift physicist boyfriend—asked de Gennaro (then a CERN computer scientist and former musician) to turn her tale of woe into a song. De Gennaro says he opted for doo-wop “because the average age at CERN was mid-50s.” He assembled singers and a band to perform the song, and the Cernettes were born. Over the years, about a dozen different singers have been in the group—CERN secretaries and interns, and outside talent from Geneva's operatic society. “We never played for money,” de Gennaro says. “We played to get the physicists to smile and dance.” But now de Gennaro is retiring, so the act must disband. Ironically, the Cernettes went briefly viral last week, when news outlets reported that a 1992 still of the group was the first photo ever uploaded to the World Wide Web. De Gennaro demurs: When the photo was posted to CERN's Web site, “they told me that this was the first photo of a band that appeared on the Web,” he says. “Somehow, this got twisted.” ## By The Numbers 1.5 million — Number of different plankton taxa in the oceans according to a preliminary analysis of 27,000 samples collected by Tara Oceans, a scientific sailing expedition, from the Mediterranean Sea and the Red Sea to the Pacific Ocean and the Antarctic region. 91 — Percentage of the world's 103 lemur species that are listed as critically endangered, endangered, or vulnerable by the International Union for Conservation of Nature. ## ScienceLIVE Join us on Thursday, 26 July, at 3 p.m. EDT for a live chat with leading experts on a hot topic in science. http://scim.ag/science-live 3. Profile: Bert Vogelstein # Cancer Genetics With an Edge 1. Jocelyn Kaiser Their lab helped reveal how faulty genes cause cancer, but Bert Vogelstein and Kenneth Kinzler sometimes irk colleagues with their “reality check” comments on genomic medicine. In April, as scientists at a cancer meeting in Chicago were wowing audiences with their new DNA studies, geneticist Bert Vogelstein stood by with a bucket of cold water. At a briefing, he told a roomful of reporters that his own work shows how little value a key genetic approach—whole-genome sequencing—will have for preventing cancer. Using data from identical twins, his team estimated how well whole-genome tests could spot individuals at risk. Most people, the model showed, would get a negative result because inherited mutations very rarely predispose us to cancer. Yet about one-third of those who test negative will still develop cancer, triggered by environmental factors and bad luck. “It would be great if we could determine who will and will not get cancer from sequencing their DNA, but the reality is, we won't,” Vogelstein said later. Vogelstein's sobering message, picked up by The New York Times and other publications, irritated some scientists: They claimed the study had flaws and wasn't saying anything new. True, papers with similar findings had been published before, unnoticed by the press. But Vogelstein says that although the bottom line may have been familiar to some scientists, his study used a novel model to quantify the limits of personal genomes for public health. He says he also wanted to drive home the message that the only sure way to reduce cancer is through screening and a healthy lifestyle. Some colleagues say the paper was a typical dart thrown by Vogelstein, who is known as a contrarian. Vogelstein (a longtime member of Science's board of reviewing editors) and collaborator Kenneth Kinzler, who co-directs their lab at the Johns Hopkins University medical campus, helped lay the foundation of cancer genetics by revealing how mutations in key genes lead to a tumor. Their work helped inspire others to use genetics to predict cancer risks and develop personalized cancer treatments. Yet today, Vogelstein often offers what he calls a “reality check” on such efforts. Last month, for example, his team warned about a flaw in so-called targeted drugs that dramatically shrink tumors: They inevitably fail when resistant tumor cells take over. “He sees things in very interesting and provocative ways,” says cancer geneticist Stephen Chanock of the U.S. National Cancer Institute (NCI) in Bethesda, Maryland. Vogelstein “wants to make a quick difference and a quick hit. I would guess he thinks his job is to push the field along,” says cancer geneticist Joe Gray of Oregon Health & Science University in Portland. Gray adds: “But whenever Bert talks, I listen.” In the past few years, Vogelstein and Kinzler have shifted away from discovering new cancer genes to a less glamorous pursuit: using genetic tests to detect common tumors as early as possible, when they are easiest to cure. This is not mainstream work, they say: “Society is fixated on curing advanced cancer. It's considered a success even if you're just reducing patients' cancer burden for a few months, not prolonging their lives for a long time,” Vogelstein says. But it's a natural progression of earlier research, the two said in a recent interview at their upper-floor lab in downtown Baltimore, Maryland, two blocks from where Vogelstein was born. Vogelstein, 63, small and trim with a grizzled beard, irreverent and humorous, wore his usual jeans and a white dress shirt. Kinzler, a tall, boyish-looking 50, was in khakis. Their research was never driven by curiosity about biology, Vogelstein says, but by “an overwhelming desire to empty the cancer clinics across the street.” ## Cancer genes After studying mathematics at the University of Pennsylvania, Vogelstein went to medical school at Johns Hopkins, becoming a pediatric oncologist. His frustration at being unable to explain cancer—to tell parents why their 4-year-old daughter had leukemia, for example—inspired him to pursue research. In 1989, while working with colon tumor tissue, which they could obtain for different cancer stages, his group showed that the mutated TP53 gene was not a cancer driver as had been thought but a tumor suppressor. Mutations that cause the gene to lose its function occur often in many different cancer types. Vogelstein says his decision to work with “messy” human tumors, not cell lines or animals, might be the reason why it took him three tries to get a grant from NCI. In the late 1980s, Vogelstein also unveiled what's now a textbook model of how colorectal cancer develops. He and Kinzler, who joined the lab as a graduate student, later showed how the slow accumulation of mutations in specific genes leads to tumor growth. They went on to discover a string of major colon cancer genes. These studies all started by analyzing human tumors for mutations. Their guiding principle was that if a gene drives cancer, it will often be mutated. Trying to figure out a gene's importance by studying its function, they say, can be misleading. Just because the mutation alters a disease process doesn't mean it's the cause. It was “a completely different view” from what many other scientists believed, Vogelstein says. The two built a highly competitive lab whose members played as hard as they worked. They installed a Ping-Pong table near the postdocs' desks; the lab held basketball and tennis tournaments. They had a rock band, Wild Type, composed of cancer biologists, with Vogelstein on keyboard and Kinzler on drums. Lab members got free use of a beach house. A quirky tradition arose in interviews of prospective postdocs: Candidates were asked to wear a Burger King crown when they spoke about their graduate work. “If you were not willing to wear the crown, you were probably a little too high-strung for the lab,” says Devin Dressman, a former Vogelstein postdoc who is now at Life Technologies in Beverly, Massachusetts. Vogelstein and Kinzler rarely travel to speak at meetings. They prefer to stay involved in their lab's experiments and avoid being swayed by what's popular, or the “herd effect,” Vogelstein says. Instead, Vogelstein keeps up on the latest science by devouring the literature: “He reads more than any other human I know,” Dressman says. ## Cancer's own genome project After human DNA was fully sequenced in 2003, cancer geneticists around the world began to envision their own genome project. Instead of homing in only on suspected cancer genes, they would do a more sweeping search and systematically sequence the entire genome of patients' tumors and compare the results to DNA in normal cells to find all possible cancer genes. Vogelstein and Kinzler declined to join the pilot phase of a proposed$1 billion project funded by NCI called The Cancer Genome Atlas (TCGA). Instead, using private funding, they set out to sequence the first “cancer genome” in a handful of breast and colorectal tumor samples—enough, according to their theory, to find the most important genes. They narrowed the search to focus on protein-coding DNA, the so-called exome.

It was an ambitious goal, especially because high-throughput sequencing machines weren't available. They accomplished it by brute force, performing about 200,000 polymerase chain reaction experiments on each of 22 patient tumor sample to amplify 13,000 or so genes. They sent the DNA to a commercial lab for traditional Sanger sequencing. Then they waited for results.

The gamble paid off: In 2006, Vogelstein's group published the first rough exomes for breast and colorectal cancers in Science (13 October 2006, p. 268). Data revealed both known and new cancer genes in varying frequencies—the authors called them mountains and hills—as well as daunting variations in patterns from one tumor to the next. Some researchers were critical: In Technical Comments in Science, TCGA leader and genome researcher Eric Lander of the Broad Institute in Cambridge, Massachusetts, among others, argued that the Johns Hopkins team had used faulty statistical methods and tested too few tumors to yield meaningful results.

Vogelstein and Kinzler moved on to glioblastoma exomes in 2008 and found an important new oncogene, IDH1. This was a slight embarrassment for the NCI collaboration, whose glioma project sequenced only a set of candidate cancer genes in a large number of tumors and missed IDH1.

Reviewing work by his group and others in his talks, Vogelstein seems ambivalent about the value of tumor DNA scans. More than 1000 of these surveys have added only a few dozen new cancer-driver genes to the 80 or so that were previously known, Vogelstein says. This includes some intriguing new genes, however. He and Kinzler agree that the more comprehensive NCI-backed genome atlas work is “invaluable” because it will find rare mutations and fill in gaps in our knowledge about the pathways that lead to uncontrolled tissue growth. But their own lab is moving on to other things after sequencing exomes for 20 cancer types. “We're focusing on different questions now—applications as opposed to discovery,” Kinzler says.

## Twins and cancer

One question that drew the Johns Hopkins group was: What fraction of the population would benefit from whole-genome sequencing? They had already used exome sequencing in 2009 to look for genes linked to pancreatic cancer common in one family. (Researchers compared the genomes of affected and nonaffected family members to pinpoint a responsible gene, PALB2.) Like screening for the BRCA breast cancer genes, identifying individuals with rare but high-risk pancreatic cancer genes might help some avoid death from the disease. This work led the group to wonder what might be gained if everyone's entire genome were sequenced, Vogelstein says. If a major fraction of the population carried high-risk genes, then sequencing everyone might make sense.

Because homozygous twins have identical genomes, the researchers could learn about their inherited disease risks without DNA sequencing, by comparing health records. Data on nearly 54,000 twin pairs, most in Europe, allowed the Johns Hopkins group to model how risks were distributed across the population. They found that genome testing had potential value. With a positive result defined as an overall disease risk of 10%, the average person would likely test positive for at least one disease, such as heart disease or diabetes. But genome scans would not help much with identifying risks for common cancers, they reported in Science Translational Medicine and at the meeting of the American Association for Cancer Research (AACR) in Chicago in April. For most people, the risks associated with inherited DNA are small or nonexistent, they found—and trivial compared to risks from random mutations and factors such as smoking and obesity.

NCI's Chanock suggests that the twins data included too few cases of specific cancers to yield definitive results: “I would not take this as the final word on the value of the genome,” he says. But Lander, despite his criticism of the group's work on tumor exomes, says the twins study “makes an important point” even if it just “confirms what we have known all along: Genes are not everything.” Lander says his own view has always been that genetic risk studies are most useful for pointing to the biology underlying diseases, not for predicting personal risks.

## Antihype

Vogelstein seems to enjoy pricking balloons. Recently, he has focused on a new target: exuberance over targeted cancer drugs. He says he got interested after seeing a paper last year on melanoma therapy. It included photos of the torso of a man with melanoma who had received a new drug aimed at a mutated gene called BRAF. Before treatment, the patient's skin was riddled with metastatic tumors; soon after treatment, the tumors vanished, and the man looked perfectly healthy. Five months later, the tumors reappeared in exactly the same locations. The photos “blew my mind,” Vogelstein says. “Why do the tumors all return at roughly the same time? It's almost as miraculous as when they disappear.”

Targeted drugs for other cancers usually stop working after about the same number of months, presumably because rare resistant cells in the tumors continue to grow and ultimately proliferate. To investigate, Luis Diaz and others in the Vogelstein-Kinzler lab drew on a sensitive technique they had developed for detecting mutations in the very small amount of tumor DNA present in a cancer patient's blood. They collected a series of these “liquid biopsy” measurements from patients with advanced colorectal cancer whose tumors had become resistant to a targeted cancer drug. With Harvard University computational biologist Martin Nowak, they devised a model showing that even before the patient begins treatment, some tumor cells always carry genes with random mutations that can support resistance to targeted drugs. This form of resistance, they wrote last month in Nature, is therefore “a fait accompli.”

But the modeling study also suggested that this resistance can be delayed by combining two drugs that target different pathways. Indeed, Vogelstein and colleagues suggest that once a targeted drug has passed initial safety trials, it's so clear that single-drug therapy will fail that they consider it unethical to give patients just one such drug. “Why shouldn't you design a large, very expensive trial to incorporate more than one agent?” Vogelstein asks.

Memorial Sloan-Kettering Cancer Center cancer researcher Charles Sawyers, who helped develop Gleevec, the leukemia drug that paved way the way for other targeted therapies, says he agrees with Vogelstein's message: “We have to move to combinations as quickly as possible.”

The Vogelstein-Kinzler lab is attacking cancer from several new angles, including an unusual clinical trial that involves destroying tumors by injecting them with bacteria. But the project closest to Vogelstein's heart right now is an effort to develop the liquid biopsy method as a general cancer-screening test. The aim is to look for a set of mutations commonly found in tumors that can be isolated from circulating blood.

Vogelstein and Kinzler argue that DNA mutations offer “clarity” compared to other blood-sampling approaches that focus on more complex protein or gene-expression markers. At the AACR meeting, Vogelstein described his group's latest blood test results for 304 patients with several types of cancer. Nearly all with metastatic tumors had detectable mutations in their circulation; more than 50% of those whose cancer had not yet metastasized had them; and none of 82 controls. “I think we're getting very close with the technology,” says Vogelstein, who owns stock in a German company, Inostics, that is commercializing the test.

Although Vogelstein seems as passionate about research as ever, he has stepped back in recent years, others say. He now goes on vacations. Lately, he's been spending two afternoons a week with his two grandchildren, a time that Johns Hopkins neuroscientist Joshua Vogelstein (their uncle) thinks is probably “the best part of his week.” The lab has gotten “leaner and meaner,” Kinzler says: It's down from a peak of about 40 lab workers (including six faculty members) to perhaps 30—one reason is that sequencing now involves less lab work. Wild Type hasn't picked up their instruments since Kinzler's wife fell ill with leukemia 10 years ago and died 6 months later. “We had to stop and we never started up again,” Vogelstein says.

Vogelstein has no plans to retire: “I was born here and when I die they'll wheel me out on a cart,” he says. Before that, he says, he has one overriding goal: that a cancer-screening test based on mutations in tumors will become a routine part of everyone's annual physical exam. “I just want one of those tests to be the standard of care. That's the one thing.”

4. Ecology

# Taking the Measure of Madidi

1. Jean Friedman-Rudovsky*

Researchers journey into one of the world's most diverse areas to help predict the future of our planet's trees.

MADIDI NATIONAL PARK, BOLIVIA—It felt like an old-fashioned treasure hunt. For 5 days, two biologists, a botanist, and their students braved collapsing cliff-side roads, clambered across rushing rivers, and hacked their way through forest to reach this spot on their map. The researchers fanned out through the jungle, navigating thorny underbrush and armies of stinging ants, in search of small metal tags. A shout broke through over the rush of the nearby creek: “Number 191!” Moments later, calls from all directions: “Sixteen is over here!” “Seventy-two!” “I've got 12!” The excitement mounted as they located and read off dozens of numbered tags that had been nailed into trees 7 years ago.

Despite the giddiness, the endeavor was serious. This was day one of fieldwork in a 3-week expedition, part of the ongoing Madidi Project, to measure tree growth here in one of the planet's most unusual protected areas. For the students, locating silver strips obscured by moss or leaves in the middle of a forest the size of Vermont was like discovering buried treasure. For the Madidi Project's director, Peter Jørgensen of the Missouri Botanical Garden in St. Louis, who started tagging Madidi's trees almost a decade ago, it was “like finding old friends.”

Those friends, it turns out, have an increasingly important tale to tell. Botanists and biologists have long grounded their understanding of the natural world on long-term projects such as Madidi. Recurrent surveys of the same communities generate data that help researchers figure out how species interact in the face of environmental change or nearby human activity. As climate change descends upon us, it's becoming critical to identify factors influencing tree survival. “Local populations of trees have three options,” says Nathan Kraft, a biologist at the University of Maryland, College Park. “They can move, adapt, or die off.” By “move,” he means that trees have the ability to shift their range when threatened in their current habitat, say by long-term drought or a more competitive species. Through seed dispersal and other regeneration mechanisms, trees can move to a better area. Otherwise, to survive, they must adapt, evolving traits better suited to the new conditions. The choices will determine the biodiversity of future forests. “Only once we know how species will behave can we figure out how to save them,” not just in Madidi but elsewhere as well, Jørgensen says.

## It's not easy being a botanist

Madidi National Park is home to 11% of the world's bird species and an estimated 12,000 plant species. Located in northwestern Bolivia, it is one of the largest, most diverse protected areas on the planet, on par with Madagascar's national parks and Colombia's Chocó jungle. It is also remarkably untouched. Aside from the handful of communities living inside the park and limited ecotourism on its eastern edge, human activity is scarce in the area. “Madidi is exceptional for scientific study,” says Peter Raven, president emeritus of the Missouri Botanical Garden, the main sponsor of ongoing studies here, “because of this lack of disturbance.”

For biodiversity researchers, it has another attribute: the longest continuous elevation gradient on the planet. The park ranges from Andean peaks of just over 6000 meters above sea level to the Amazon basin, approximately 180 meters above sea level. “Madidi is like a laboratory for climate change,” says the Madidi Project's Bolivia coordinator, Alfredo Fuentes, a La Paz–based botanist at the Higher University of San Andrés in La Paz, Bolivia, Bolivia's largest university. Because temperatures drop as the elevation increases, species in Madidi can theoretically shift uphill instead of adapting to global warming in place. “Madidi offers an excellent opportunity to watch reactions to climatic change,” Fuentes says.

The Madidi Project, launched by the Missouri Botanical Garden in 2001, has begun to capitalize on an opportunity not only to look at the effects of climate change but also to test existing theories about what controls biodiversity. Researchers are looking in particular at factors that influence the composition of plant communities, a concept known as beta diversity (see sidebar). The project's entire data set includes more than 206,000 trunks and about 2400 tree species. “Nothing along these lines has been done before,” says Thomas Lovejoy, biodiversity chair at the Heinz Center in Washington, D.C.

From 2001 to 2010, the project established 442 small plots (0.1 hectare) and 50 large ones (1 hectare) throughout the park, each of which are now demarcated by small plastic tubes topped by a strand of orange fabric. Within each plot, researchers mapped, identified, tagged, and noted the characteristics of all trees with at least a 10-centimeter diameter, establishing a baseline.

Now, several expeditions a year here measure these trees, ideally returning to each plot every 5 years to take stock of the changes. During this visit in April, the group will remeasure more than 1600 trees from three plots in a semideciduous tract of Madidi ranging between 900 and 1100 meters above sea level. (One plot is on the banks of a river, one on a slope, and another on a crest.) Every tagged tree's location is double-checked against a map from the plot's establishment, and if a tagged tree has died, the team tries to establish the cause of death by examining the stump or remaining limbs for evidence of infection or other impacts such as a lightning strike.

Before their first trip, “everyone thinks it's like being on vacation,” says Denmark-born Jørgensen as he swats away swarming mariwees, gnatlike insects that sting. It's the first day working the plots, and his neck is already swollen from bee stings and ant bites in the campground the night before.

Around him, the number calls continue, but they are no longer random. Rather, at each tagged tree, the researchers methodically sound off with codes that correspond to every possible characteristic: width, height, smell, bark texture, closeness to canopy, leaf patterns, and more. “I've got to be very careful to enter everything right,” says Esther Mosqueria of the Higher University of San Andrés. Her undergraduate biology thesis will be on Madidi's tree population. She scribbles furiously to keep up with her colleagues' shouts.

One of the local guides accompanying the group is 3 meters off the ground, balancing on a thick branch leaned against a tree. He wraps a tape measure around the midsection of a large trunk. Many trees in this area have aboveground root structures, so team members must climb to measure the trunk diameter accurately. Tree height is estimated (with roughly 90% accuracy) because taking an exact measurement through triangulation would triple the fieldwork time.

Species identification is a key component of the project; the team double-checks trees identified in the past, taking samples back for confirmation if there's any doubt. “You can have a ton of data, but it doesn't mean anything if you can't consistently identify and standardize the species,” says Brian Enquist, an ecologist and evolutionary biologist at the University of Arizona in Tucson. While identifying tree species might seem the task of a third-grader, “in tropical areas, the extreme number of species means identification is exceptionally difficult,” says Iván Jiménez, a botanist at the Missouri Botanical Garden who analyzes the Madidi Project's data.

More than 6 hours later, everyone has worked up a sweat—better to be covered from head to toe than ravaged by insects and spiky trunks and bushes. Just before sundown, they head back to the small campsite, which is roughly one-quarter the size of a football field and took a day of machete-slashing to clear. Shared tents ring a common area with a large table and benches built out of branches and 8-meter-tall river grass stalks, held together with string. The wife of one of the guides prepares meals and boils stream water for drinking. But not even the burning fire can mask an omnipresent odor of nearby “wild-garlic” trees.

At night, wearing headlamps, team members press leaf samples into old newspapers to store them intact for the trip back. They listen to Bolivian rock on a battery-operated MP3 player and tell bad jokes until they collapse onto their foam sleeping pads, anticipating a dawn wake-up call to repeat the day's activity.

“It's not easy being a botanist,” says biology student Eber Renjito of the Higher University of San Andrés.

## How forests work

The scale of the Madidi Project makes the hardships worthwhile. Already a decade old, the project is slated to go on for at least one more year or longer if continued funding can be found. Other similar biodiversity studies are smaller and have a shorter life span. Relatively few are based in the tropics, even though tropical plant species are among the most susceptible to climate change. And very little is known about tree patterns in tropical mountainous zones such as the Andes.

What is known points to a move-rather-than-adapt strategy for coping with climate change. Miles Silman, a biologist at Wake Forest University in Winston-Salem, North Carolina, has studied an area of Peru with a similar elevation gradient to Madidi. His work has shown that many trees are in fact shifting their ranges to higher, cooler ground rather than trying to adapt in place to increasing temperatures.

That's what Sebastian Tello, a biologist with the Madidi Project based at the Missouri Botanical Garden, thinks is going on in Madidi. Adaptation requires genetic change that occurs over the course of generations. Woody plants have long lives, so the shifts in genetic composition are extremely slow. “In general, we hypothesize that it's easier for tree species to move than adapt,” Tello says.

Yet other environmental factors may impede geographical shifts by certain species, causing them to be left behind even as community compositions change. For example, the project's research has found a species-rich belt within Madidi, between 1000 and 1500 meters above sea level, that mixes trees from the lowlands and the mountains. “What will happen to this belt is uncertain,” Jørgensen says, “since species adapted to lowlands and flat areas are not particularly well adapted to growing on sloping terrain.” Trees with large buttresses, for instance, might be in danger because they may not be able to survive on a hill. “The high-diversity belt may shrink or disappear completely,” he says.

Having even a vague idea of whether a community or species is at exceptional risk of disappearing is a main goal. “This kind of research is fundamental if you want to have representative ecosystems going into the future,” Lovejoy says. Ideally, research like the Madidi Project gives policymakers an idea of which areas are most in jeopardy and in need of protection. “Since there is limited money for conservation,” Jørgensen says, “we want to make sure the areas with greatest range of species, as well as the species that may have less chance for survival, are protected.”

He warns, though, against drawing firm conclusions about climate change's effect on forests even after accumulating a mass of data over 10 years. A decade is not long enough. Sustaining the work over many decades may be the most important element to advancing this research. But Jørgensen admits that his own days of fieldwork might be nearing their end. “I'm getting a little too old for this,” says the 54-year-old, red in the face and tired already. He hopes his Bolivian colleagues will take over the expeditions while he focuses on taxonomy and data analysis back at his desk in St. Louis. “It may be decades before we can really understand how our forests will change,” he says. “But Madidi is a start.”

• * Jean Friedman-Rudovsky is a journalist based in La Paz, Bolivia.

5. Ecology

# Probing Diversity's Complexity

1. Jean Friedman-Rudovsky

By identifying all trees growing along an elevation gradient from above 6000 meters in the mountains to 180 meters above sea level, researchers hope to model the true complexity of our planet's biodiversity.

Rainforests may be the conservationist's poster child, but they fall short as models of the true complexity of our planet's biodiversity. So says Peter Jørgensen, a botanist at the Missouri Botanical Garden in St. Louis, whose long-term project in Bolivia promises to shake up our understanding of the distribution of tree species in tropical South America. The Madidi Project (see main text) charts the changes in tree communities growing along an elevation gradient that plunges from above 6000 meters in the mountains to 180 meters above sea level. By identifying all trees at least 10 centimeters in diameter within hundreds of research plots along this gradient, Jørgensen and his colleagues have built a database that allows them to examine spatial patterns of diversity in ways not possible from studies limited to rainforests, which tend to be comparatively homogeneous because they are confined to the lowlands.

The Madidi Project focuses on comparing the rate of turnover of species in a sample of plots—a concept known as beta diversity. “It's about trying to understand why a certain species lives in one area and not in another nearby,” says Brad Boyle, a biologist at the University of Arizona in Tucson. To gauge beta diversity, researchers count the number of species in various plots. Then they ask: How many species are common to the plots? How many are different? The higher the number of species found in only one plot, the greater an area's beta diversity.

Researchers want to know why beta diversity is higher in some areas than others. Why do some species thrive in two different areas while some drop out? “There are many different factors,” says Nathan Kraft, a biologist at the University of Maryland, College Park, including climatic tolerance, soil quality, rainfall, and the differences in the dispersal abilities of species. “The really important next step is determining the relative importance of these factors and how they interact.”

His work shows that the high beta diversity along elevation gradients arises because so many species are capable of living in these areas that by chance each plot has numerous species not found in another nearby (Science, 23 September 2011, p. 1755). But Jørgensen and his colleagues suspect that chance is not the most important factor. They think that high beta diversity in tropical mountain forests like those of Madidi is driven by a special mix of environmental factors. They hope an analysis under way will prove them right. “The value of [the Madidi Project] is in testing diversity theories of a landscape with real actual data, which is very rare,” says Thomas Lovejoy, biodiversity chair at the Heinz Center in Washington, D.C. “It speaks to the importance of matching theories and what happens in nature.”

Kraft is eager to see the results: “The strength of the Madidi data set is that it's intensively sampled and spans a broad elevation gradient. It will be a great step forward.”

6. European Association for South Asian Archaeology and Art Meeting

# The Ingredients for a 4000-Year-Old Proto-Curry

1. Andrew Lawler

Recent studies presented at the meeting found a surprisingly diverse Indus diet that incorporated spices, beans, and grains commonly eaten today, and even bananas.

When cooks in the ancient Indus River civilization prepared their meals 4000 years ago, the results may not have been much different from what you might order today in an Indian restaurant. Recent studies presented at the meeting found a surprisingly diverse Indus diet that incorporated spices such as ginger and turmeric, beans such as lentils and mung, grains such as rice and millet, and even bananas. An explosion in food-related studies, thanks to both new tools and new interest in rural villages, provides exciting clues to day-to-day life in the Indus, says Indus expert Jonathan Mark Kenoyer of the University of Wisconsin, Madison, who was not directly involved in the studies.

With Egypt and Mesopotamia, the Indus was among the first urban civilizations, centered on today's Pakistan and India. The Indus people built a half-dozen massive cities around 2500 B.C.E. that mostly fell into ruin after 1800 B.C.E. No Indus texts have been deciphered, however, and few images found, leaving scholars with fundamental questions about how the people lived, worked, and worshipped. But some of their traditions, including food preparation, may live on.

Archaeologists have long spotted burnt grains such as wheat, barley, and millet at Indus sites, but identifying vegetables, fruits, nuts, roots, and tubers has been more challenging. Researchers are increasingly using phytoliths—the mineral secretions left by plants—to identify specific plant remains, as well as starch grain analysis (Science, 2 July 2010, p. 28). Plants store starch granules as food, and the microscopic leftovers can be identified by researchers. For example, anthropologists Arunima Kashyap and Steve Weber of Washington State University, Vancouver, analyzed starch grains from human teeth from the ancient town of Farmana, west of Delhi, and found remains of cooked ginger and turmeric. They also found those ingredients inside a cooking pot. Dated to between 2500 and 2200 B.C.E., the finds are the first time either spice has been identified in the Indus. Cow teeth from Pakistan's Harappa—a major Indus city—yielded the same material. “It's like India today,” Weber says. “Cattle wander around eating trash,” including the remains of cooked meals. In some Indian regions such as the western province of Gujarat, some families still leave food remains outside the house as a ritual offering to cattle.

Whether or not these spices represent the earliest curry is not clear: Kashyap and Weber note that what makes curry curry is disputed even today. Black pepper and chili peppers, for example, are common in the dish today but were later imports to India.

Even bananas, not known to have been cultivated here until late medieval times, have turned up at three scattered Indus sites. A team led by Marco Madella, a Barcelona archaeologist with the Spanish National Research Council, found phytoliths of banana on grinding stones at Farmana. Phytoliths at the site of Loteshwar in Gujarat and at Kot Diji in the Indus heartland in Pakistan were also found. “I'm not confident in saying it was cultivated,” Madella says. “But clearly the Indus people were in direct contact with people to the east,” where the plant grew wild.

Indus farmers also grew a surprisingly wide array of grains and beans. Many archaeologists once thought that the society depended primarily on crops such as wheat and barley, which were planted in winter. But new data from rural villages challenge that idea. Examining two sites near today's Masudpur, west of Delhi, University of Cambridge archaeologist Jennifer Bates compared carbonized seed and phytolith density per liter of soil near hearths to determine the relative abundance of crops by period and site. Bates found that both villages practiced summer and winter cropping, and both ate wheat, barley, millet, and rice from early Indus times, as shown by nearby pottery; she also identified lentils and mung beans. Rice has long been assumed to be only a late addition in the Indus, yet one village apparently ate more rice than wheat or barley, although millet dominated.

Many of these crops have uses in addition to pleasing Indus palates, Kenoyer notes. Burned bananas produce salt, ginger can treat illness, and turmeric is used for both poultices and dyeing cloth. The data may also shed light on how specialized and exotic foods reflected class differences, he says.

7. European Association for South Asian Archaeology and Art Meeting

# Diving Into the Indian Ocean's Past

1. Andrew Lawler

Researchers studying what appears to be the oldest known shipwreck in the Indian Ocean say it promises to remake our understanding of the region and era.

Nearly 15 years ago, two fishers in the waters off the southern coast of Sri Lanka hauled up a stone slab etched with ancient Hindu symbols. During a brief 2008 dive, archaeologists retrieved pottery and glass ingots. Then, in December 2011, with funding from the U.S. National Endowment for the Humanities and other sources, researchers began the first systematic dives to examine what appears to be the oldest known shipwreck in the Indian Ocean, radiocarbon dated to between the 2nd and 1st centuries B.C.E. Because almost nothing was known about seafaring in this time and place, the wreck promises to remake our understanding of the region and era, says team member Osmund Bopearachchi, a historian at the University of Paris, Sorbonne.

Archaeologists suspected that traders crisscrossed the Indian Ocean at this time, but the evidence was slim. Just a century or so later, merchants from India and the Roman Empire plied the Indian Ocean, trading cotton, glass, and spices. But textual references and archaeological evidence of earlier maritime trade are extremely rare; the only known South Asian shipwrecks date from medieval times. Indians and Arabs dominated the medieval trade, which connected Europe and China, but its origins are murky.

The Sri Lankan wreck offers the first good look at such trade. The island, which lies off the southeastern coast of India, was an important Buddhist kingdom that eventually became a wealthy port of call. The ship remains lie near the estuary of Walawe Ganga, one of the nation's few navigable rivers. Just upstream lies Godavaya, a recently excavated monastery dating to the 2nd century B.C.E.; monastic settlements often played an important economic role at this time. Other nearby settlements go back to the 4th century B.C.E., says Bopearachchi, who works with underwater archaeologist Deborah Carlson of Texas A&M University, College Station, as well as the maritime unit of the Sri Lanka Department of Archaeology.

Only surface surveys have been done to date, examining artifacts spread across 40 meters of ocean floor that mark it as the site of a sunken ship. Massive glass ingots, used to make expensive vessels, were part of the ship's cargo and provide what Bopearachchi calls “the best physical evidence for the early exchange of raw glass in South Asia.” Samples from both the glass and nearby pieces of metal point to an Indian origin, and the ship may have been bound for the port, he says.

Its remains lie in 34 meters of water amid strong currents, but the water is crystal clear, and conditions are good for diving during the winter monsoon. Archaeologists hope that the ship itself may lie below the bottom, providing evidence of shipbuilding technology. The team plans to begin excavating at the end of this year. The results “might lead to a rewriting of the economic, social, religious, and cultural history of the area,” Bopearachchi says. Others agree that the find will shed needed light on a critical trade route. “This is the Silk Road of the sea,” says Hans-Joachim Weisshaar of the German Archaeological Institute in Bonn.

8. European Association for South Asian Archaeology and Art Meeting

# Persians Made the Afghan Desert Bloom

1. Andrew Lawler

A French team surveying a region of north-central Afghanistan has recently discovered large settlements organized around an impressive water system dating back 2500 years.

North-central Afghanistan is a harsh desert of dunes long thought to have been uninhabited save for the occasional hardy nomad; even today the region is sparsely settled. But a French team surveying the region north and west of Mazar-i-Sharif has recently discovered large settlements organized around an impressive water system dating back 2500 years. No other culture before or since has managed to support large settlements in this desolate region. The find could provide exciting new information on irrigation techniques that made the early central Asian desert bloom, say the small cadre of archaeologists who work here.

Roland Besenval, an archaeologist with the French national research agency in Paris; Eric Fouache, a geoarchaeologist at the University of Paris, Sorbonne; and others identified more than 30 large settlements, all dating to the period of the Achaemenid Persian Empire, which flourished from the shores of Greece to western India in the 5th and 4th centuries B.C.E. The empire was overwhelmed by Alexander the Great late in the 4th century B.C.E. The archipelago of northern Afghan settlements appears to have died out at that time or a little earlier, and has been largely uninhabited ever since.

Many of these towns or garrisons had massive walls that have been hidden in the sands. One site, Altin Dilyar Tepe, first noted by Russian researchers in the 1970s, sits on a hill that the French team determined was an artificial mound, showing that enormous manpower was used to build it, Besenval says.

Using a tiny, instrument-filled drone called a hexacopter, the team spotted a 2-meter-wide linear feature that led to a 60-by-80-meter basin near Altin Dilyar Tepe. The mud-brick canal stretches for nearly 10 kilometers, pierced periodically with at least two other large basins that coincide with settlements. Besenval says this apparent aqueduct likely carried water for farming from the Balkh River, which flows out of the mountains, across the harsh desert. Achaemenid pottery found nearby dates the structure as early as the 5th century B.C.E. By the 4th century B.C.E., a drying climate may have made this always-marginal land too difficult to farm even with irrigation, Besenval adds.

In Central Asia, where water remains a key and scarce resource, the discovery of a sophisticated water system this ancient excited archaeologists at the meeting. Besenval's interpretation of the feature “is very convincing,” says archaeologist Pierfrancesco Callieri of the University of Bologna in Italy, who has dug in Central Asia. Besenval also recently found signs of irrigation at a site in Tajikistan that dates back long before the Achaemenids, as early as the 4th millennium B.C.E., suggesting that irrigation has old roots in the area.

Because of security issues, the researchers couldn't stay overnight near the Afghan sites, so they have yet to establish details such as the aqueduct's gradient or whether it stretches another dozen kilometers to reach the Amu Darya river to the north. Besenval is eager to return to gather more data, but security concerns kept him out of the field entirely last season. “Our visits have been too short,” he says in frustration. “And we have so many questions to ask.”