wvi Seeking Agriculture's Ancient Roots As they pinpoint when and where many crops were first domesticated, researchers are painting a new picture of how—and perhaps why— humans began to change their relationship to plants

JALES, FRANCE—In his lab in a 12th century from Nevali Cori," he says. So in the earliest fortress that now houses the Archeorient cultivated fields, wild and domesticated research center here, archaeobotanist grew in close proximity. George Willcox pops the top off a plastic The scarred spikelets under Willcox's capsule filled with tiny black particles, microscope represent one simple, physical spills them out into a petri dish, and puts the sign of a very complicated process: the dish under a binocular microscope. Magni- rise of agriculture. Farming was revolu- fied 50 times, the particles leap into focus. tionary in its implications for humanity, They are charred fragments of wheat providing the food surpluses that later spikelets from a 10,500-year-old archaeo- fueled full-blown civilization, with all of logical site in Turkey called Nevali Cori. its blessings and curses. Domestication— Wheat spikelets are attached to the central defined as the physical changes plants stalk of the wheat ear and carry the seeds, undergo as they adapt to human cultiva- or grain, that humans grind into flour. tion—was key to this transformation. It "Look at the scar at the lower end of the allowed former foragers to increasingly spikelet, where it has broken off," Willcox control when, where, and in what quanti- says. The scar is jagged—a hallmark of ties food plants were grown rather than domesticated wheat. It's a sign that the simply depending upon the vagaries of spikelet did not come off easily but nature. And unlike other aspects of early detached only when harvested, so the plant agriculture, such as whether a seed was probably needed human help to disperse its planted or simply gathered by human seeds. "This is the earliest evidence for hands, "domestication is visible" in the domesticated wheat in the world." archaeological record, says archaeologist Willcox spills the contents of a second Timothy Denham of Monash University in capsule into another dish. The scars are Clayton, Australia. round and smooth, showing that these Over the past decade, a string of high- spikelets easily detached and dispersed their profile papers has pinpointed the time and stores of grain. "This is wild wheat, also place of the first domestication of crops, ranging from wheat and maize A decade or so ago, most archaeologists to figs and chili peppers. Now saw the advent of agriculture as an abrupt researchers are beginning to fit break with the hunting-and-gathering all of these into a larger story of lifestyle on which hominids had relied for worldwide plant domestication. millions of years. Researchers thought that At Nevali Cori, where wild domesticated crops appeared very soon and domesticated plants grew after people began to cultivate fields, first in the same fields and perhaps in the Near East as early as 13,000 years even exchanged genes, Willcox ago, then somewhat later in a handful of and colleagues conclude that other regions. full domestication might have But the new data suggest that the road taken thousands of years rather from gathering wild plants to cultivating than the 200 years or fewer that them and finally domesticating them was some archaeobotanists had long and winding (see chart on p. 1835), predicted. "They could not unfolding over many millennia. "If the • i have gone from one kind of agricultural revolution is supposed to be ;iij,.'.'. economy to another in just a evidence for a punctuated change in human few generations," Willcox says cultural evolution, it seems to have taken of the early cultivators. "These quite a long time to get to the punctuation things happened gradually." point," says archaeobiologist Melinda

1830 29 JUNE 2007 VOL 316 SCIENCE www.sciencemag.org Published by AAAS NEWSFOCUS

Wheat's-eye view. Crop plants adapted slowly to human cultivation, evolving on a time scale of millennia rather than centuries.

Zeder of the Smithsonian Institution in has pushed the dates for the first domestica- Wild plants: The long goodbye Washington, DC. Douglas Kennett of the tion of squash and other crops back to about In his writings about evolution, Charles University of Oregon, Eugene, agrees. 10,000 years ago, making the roots of farm- Darwin argued that domestication was a "Agriculture was not a revolution," he says. ing in the New World almost as deep as clear example of selection in action. By cul- "People were messing about with plants for those in the Old World. tivating plants—growing them deliber- a very long time." Moreover, new archaeological work ately—humans intentionally or unintention- Clues to how this slow transition took shows that plants were domesticated ally select certain traits. Today, researchers place are accumulating rapidly. An alliance independently in many parts of the globe. define domestication as the genetically of archaeologists and geneticists armed There is now convincing evidence for at determined physical and physiological with new techniques for probing plant least 10 such "centers of origin," including changes a plant has undergone in response to genomes and analyzing microscopic plant Africa, southern India, and even New human behavior. "Domestication is the remains (see sidebar on p. 1834) has been Guinea (see map on p. 1833). "All around result of genetic changes that have evolved tracing the route to farming in much closer the world, people took this very new step because of cultivation," explains archaeolo- detail. In the Near East, for example, and started cultivating plants," which led to gist Dorian Fuller of the Institute of Archae- researchers are finding that domestication their domestication, says Smithsonian ology at University College London (UCL). itself happened a bit later than had been archaeobotanist Dolores Piperno. The rush These alterations make up what botanists thought, although humans apparently culti- of new data could help eventually solve the call the "domestication syndrome": signs that vated wild cereals for thousands of years puzzle of why agriculture arose in the first plants have adapted to humans and that before plants showed physical changes. place—a riddle archaeologists have been researchers eagerly seek at archaeological Meanwhile, new research in the Americas trying to solve for nearly a century. sites. In cereals such as wheat and barley, the

www.sciencemag.org SCIENCE VOL 316 29 JUNE 2007 1831 Published by AAAS I NEWS FOCUS

syndrome includes the tendency for spikelets in collaboration with Piperno, found micro- across the Near East, as large farming vil- to stay on the stalk until they are harvested, as scopic remains of barley and possibly wheat lages sprung up like mushrooms and people seen in the jaggedly scarred specimens found on a large stone implement. They concluded quickly formed trade and communication at Nevali Cori, pius larger seeds and a thinner that the inhabitants of Ohalo II had ground the networks over the entire region. seed coat that allows easier germination. (It grains to make flour and possibly also baked The notion of a long run-up to domestica- also includes less visible traits, such as simul- dough in one of the ovenlike hearths. tion also gets support from new findings by taneous flowering times.) "Ohalo II is an important warning to Willcox and archaeobotanist Ken-ichi Tanno Once humans began to cultivate plants, archaeologists," Fuller says. "We need to of the Research Institute for Humanity and how long did domestication take? In 1990, abandon some of our long-held assump- Nature in Kyoto, Japan. They examined the pendulum swung toward a rapid sce- tions that as soon as people began to use charred wheat spikelets from four sites of dif- nario after archaeobotanist Gordon Hillman cereals, they would begin to [cultivate and] ferent ages in Syria and Turkey. There was a of UCL and plant biologist Stuart Davies domesticate them." clear trend over nearly 3000 years: Earlier of in Wales plugged More recently, some researchers have sites had fewer domesticated spikelets and data from cultivation experiments into a begun taking a second look at just when later sites had more. At 10,500-year-old computer model. They concluded that domesticated plants first showed up in the Nevali Cori, only about 10% of the spikelets domestication might have occurred within Near East. For decades, excavators had were clearly domesticated, whereas 36% were 200 years and perhaps in as few as 20 to pegged this transformation to an archaeolog- domesticated at 8500-year-old el-Kerkh in 30 years, assuming, as many archaeologists ical period that began about 11,800 years Syria and 64% at 7500-year-old Kosak have, that early farmers used sickles to har- ago and is marked by the first permanently Shamali, also in Syria, Willcox and Tanno vest their crops. Sickles presumably would settled villages. There were a few claims for reported last year in Science (31 March 2006, have strongly selected for spikelets that p. 1886). These results suggest that wild stayed on the stalk until harvest, because varieties were only gradually replaced by those that dropped earlier would be lost and domesticated ones, they say. not replanted. "It was possible to put "Domestication was the culmination of a together a nice story, that agriculture lengthy process in which plants were culti- appeared fairly abruptly," says botanist vated but retained their wild phenotypes," Mark Nesbitt of the Royal Botanic Gardens, says geneticist Terry Brown of the Univer- Kew, in Richmond, U.K. sity of Manchester in the U.K. "Early farm- Before long, however, new data began to ers were receiving the benefits of agriculture raise doubts about this story. For example, at long before domestication evolved." Even Jales, Willcox and colleagues conducted Hillman says that he is "very impressed" experiments in a nearby field, cultivating with the analysis, although it contradicts his wild varieties of wheat, barley, and to previous work: "[Domestication] probably deduce how quickly domesticated forms did take this long." might evolve. The answer: not very fast. No But why? Fuller, in an article earlier matter how researchers harvested the grains, this year in the Annals of , suggests a good portion of the easy-to-detach wild that humans may have exerted weak rather spikelets fell to the ground and germinated than strong selection pressure on their to sprout a new generation of wild wheat. All in the family. Maize and its wild ancestor teosinte crops. "Weaker selection means domesti- Meanwhile, a remarkable discovery in (left) are closely related despite their differences. cation would take longer, while stronger Israel also suggested a long run-up to selection means it would happen more domestication. In 1989, a team led by Dani even earlier dates, such as a few relatively quickly," he explains. Nadel of the University of Haifa in Israel large seeds of rye at Abu Hureyra in Syria, And there are many ways that early farm- began excavating a site called Ohalo II on dated to about 13,000 years ago, and which ers' behavior might have weakened selec- the southwest shore of the Sea of Galilee. Hillman argued were domesticated. But in a tion. For example, Fuller questioned whether The site was radiocarbon-dated to 23,000 2002 survey, Nesbitt found that the earliest sickles were actually used in early harvest- years ago, when the last Ice Age was still in Near Eastern villages lacked definitive evi- ing. Other methods, such as picking already- full frost and at least 10,000 years before dence of domesticated cereals, although fallen spikelets from the ground, would not the earliest domesticated plants. Excava- wild plants were plentiful. Unambiguous have selected for spikelets that stay on the tors found the remains of huts, plus a burial signs of domestication didn't turn up until stalk. Although sickles date as far back as and several hearths. More than 90,000 indi- about 10,500 years ago, in larger settlements 15,000 years ago, no domesticated plants vidual plant remains were recovered, with different architecture and a much more show up before 10,500 years ago. So the first including acorns, pistachios, wild olives, complex social organization, he concluded. sickles may have been used for other tasks, and lots of wild wheat and barley. But "There is no current evidence for domes- such as cutting reeds for floor matting, rather "there is not a single domesticated species ticated plants in the [first settled villages]," than harvesting grains, Fuller argued. at this site," says team member Ehud Weiss Weiss agrees. "But it was probably a very Willcox favors an alternative explana- of Bar-Ilan University in Ramat Gan, energetic period, when people all across the tion: During hard years, early farmers Israel, nor any evidence that the people of region were playing with cultivation of wild replenished their seed stocks with wild vari- Ohalo II were cultivating the cereals rather plants." And once plants were domesticated, eties, thus slowing domestication. Only than just gathering them. making farming more efficient and inten- when farmers began planting domesticated To their surprise, however, the researchers, sive, this way of life apparently exploded plants farther from the wild stands—

1832 29 JUNE 2007 VOL 316 SCIENCE www.sciencemag.org Published by AAAS NEWSFOCUS

INDEPENDENT CENTERS

Moschata squash 10,000 B.P. Arrowroot 9000 B.P. Yam (D. trifida) 6000 B.P. Cotton 6000 B.P. Sweet potato 4500 B.P. Lima bean 6500 B.P. Leren 10,000 B.P.

Multiple birth. People in many different parts of the world independently began to cultivate and eventually domesticate plants. physically and genetically isolating them served tropical plants, and genetic studies can milling stones in Panama to up to 7800 years from their wild ancestors—did the process date when domesticated lineages split from old, and other Panamanian sites have yielded speed up, he says. Reproductive isolation of wild ancestors. dates for these crops that are nearly as early. domesticated and wild plants could have "We were misled by what was not pre- This week, on page 1890 of this issue of acted as a "trigger," agrees Manchester's served and what we could not see," says Science, a team led by Dillehay reports Brown, spurring increasing proportions of anthropologist Tom Dillehay of Vanderbilt 10,000-year-old squash and 8500-year-old domesticates as farming spread across the University in Nashville, Tennessee. "These peanuts on the floors and hearths of houses Near East. Eventually, says Weiss, sowing, people had a very sophisticated knowledge made of stone and reeds in the Andes tilling, and harvesting "create [d] these arti- of the plants that were out there." Mountains of Peru. Genetic studies and the ficial environments that lead to domestica- Archaeologists began to see more clearly distribution of possible wild ancestors sug- tion. ... It meant totally new ideas and a back in 1997, when the Smithsonian's Bruce gest that these crops were probably domes- totally new way of life." Smith radiocarbon-dated domesticated seeds ticated elsewhere, in South America's low- and other fragments of pepo squash seeds land tropical forests. So these very ancient New World, new paradigm from a cave near Oaxaca, dates show how quickly domesticated crops At the same time that archaeolo Mexico, to nearly 10,000 spread from their original centers of origin, gists are concluding that Old years ago (Science, 9 May the team concludes. But identifying domes- World crops were fully domes- 1997, pp. 894 and 932). The tication is not always easy: Smith questions ticated a little later than once signs of domestication were whether Dillehay's evidence proves that thought, recent discoveries are clear: The seeds were larger squash, peanuts, and other plants had actu- pushing domestication in the and the stems and rinds ally undergone "any of the genetic or mor- New World back, way back. thicker than those of phological markers of domestication." Not so long ago, researchers closely related wild squash All the same, the flurry of early dates in saw little evidence for farming of that still grows in the the New World is "remarkable," says ethno- crops such as squash, maize, and region; indeed the fragments botanist Eve Emshwiller of the University of manioc before about 5000 years found were identical to today's Wisconsin, Madison, because the first ago. "Some archaeologists domesticated pepo squash. Since domesticates appear not too long after thought little of importance had then, earlier dates have steadily humans colonized the Americas, at least taken place in these tropical accumulated for the domestica- 13,000 years ago. That's a contrast to the forests," Piperno says. "We didn't tion of nearly every New World Old World, where people lived for tens of have the data." Researchers now Wild. A 23 000-year- crop. Piperno's team has dated thousands of years before domesticating have new methods to identify old wheat fragment starch grains from domesticated plants. Dillehay agrees: "People between microscopic bits of poorly pre- from Ohalo I manioc, arrowroot, and maize on 13,000 and 10,000 years ago were adapting

www.sciencemag.org SCIENCE VOL 316 29 JUNE 2007 1833 Published by AAAS I NEWS FOCUS

to [changing climatic conditions] more favor- ing teosinte was first domesticated to make domesticated 9000 years ago in Mexico's ably than we had thought before." alcoholic drinks from its sugary stalks Balsa River region inspired Piperno's inter- Genetic data support the early dates, too. rather than for the dinner table. national team to comb the valleys in search For example, John Doebley of the Univer- Maize domestication genes include tbl, of confirmation, for example. In the 30 May sity of Wisconsin, Madison, genotyped which controls the number of stalks, pbf online edition of PNAS, they reported pre- numerous specimens of that New World sta- which controls protein storage in the kernel, liminary evidence that domesticated squash ple, maize, and its wild ancestor, teosinte. and sul, which affects starch storage. and maize were grown on ancient lakesides From the number of genetic changes Recently, Doebley teamed up with ancient probably by 8500 years ago, although the between teosinte and maize, and the likely DNA specialists to track changes in these dates are not yet confirmed. "We think that speed of the "molecular clock," Doebley's genes in ancient maize, using 11 maize cobs before long we will be able to push the team concluded in a paper published in the from Mexico and New Mexico dated from archaeological dates back to match the Proceedings of the National Academy of 5000 to about 600 years ago. The domesti- genetic data," says Piperno. Sciences (PNAS) in 2002 that maize was cated variants of tbl and pbf 'were present in Yet even if people in the New World were domesticated about 9000 years ago. And all the ancient DNA samples, and all the domesticating plants early, they did not nec- they found that maize was probably domes- Mexican cobs had the domesticated variant essarily become full-fledged farmers right ticated only once, in the Balsas River Valley of the sul gene. But 1900-year-old cobs away, some archaeologists argue. "The first of southern Mexico. from New Mexico showed a mix of wild and plant domestication was 10,000 years ago, In an astonishing stream of studies, domesticated variants, the team reported in but the development of village-based agri- Doebley and other researchers have also Science (14 November 2003, p. 1158). cultural economies did not happen until taken a detailed look at the genetic changes If the domesticated variant of sul— more than 5000 years later," says Smith. In a underpinning maize domestication. The which may give corn the properties neces- 2001 paper in the Journal of Archaeological transformation of teosinte to maize was sary for making good tortillas—was not Research, Smith argued that in many parts dramatic, as these plants look so different widespread in maize populations until of the world initial plant domestication was that researchers once doubted their rela- much later, then domestication might have followed by a long period of "low-level tionship. Ears of teosinte are multistalked taken place over an extended period, the food production," during which prehistoric and have only five to 12 kernels, whereas team concluded. "There must be several peoples continued to hunt and gather while single-stalk maize ears have 500 or more. A stages to genetic domestication of plants," slowly adding already domesticated crops tough casing also protects teosinte kernels, says Manchester's Brown. to their diet. whereas maize kernels are "naked" and Doebley's work has spurred the archaeol- "Domestication of a plant is one thing, accessible to humans. Indeed, some archae- ogists to try to keep up. His and fully adopting it is another," agrees ologists have suggested that the unappetiz- finding that maize was Dillehay. But he argues that his new evi- dence from the Peruvian Andes, which includes houses, may indicate that both set- STARCH REVEALS CROP IDENTITIES tled village life and farming economies Until very recently, archaeologists searching for the first domesticated arose earlier than researchers thought, at forms of tropical plants such as yams, manioc, and bananas just kept on least in some parts of the Americas. looking. The humid tropical environments in which these plants grow Piperno agrees that the work of Dillehay destroyed evidence of their existence, leaving archaeologists with "patchy and others may now be providing the "miss- and speculative" accounts of their domestication, says archaeobotanist ing evidence" to fill at least some of that Andrew Fairbairn of the University of Queensland in Brisbane, Australia. 5000-year gap. Then in the mid-1990s, archaeologists realized the potential of starch grain analysis, a technique used for more than a century by botanists to Tell me why identify modern plants. Plants manufacture and store starches in micro- Back in the 1950s, many archaeologists scopic organelles called amyloplasts. Both the size of the amyloplasts thought agriculture was born in only two and the pattern of starch deposition vary from plant to plant, often mak- places: the Near East and the Americas. ing it possible to distinguish species. "This methodology makes things From these two fountainheads of farming, visible that were previously invisible," says archaeobotanist Linda Perry the story went, agriculture spread through- of the Smithsonian Institution in Washington, D.C. That new visibility has out the world. Yet archaeologists now recog- pushed back the dates of domestication for a number of tropical crops, nize at least 10 independent centers, and including squash, manioc, and chili peppers (see main text). When Perry Distinguished. even regions once thought to be agricultural and her colleagues went looking for chili pepper starch grains in Central Starch grains identify backwaters have taken on a new importance. and South America, for example, they found them seemingly every- manioc (fop) and In 2003, a team led by Monash's Denham where: in sediments, on milling stones and stone tools, and on pottery maize (bottom). clinched the case that bananas, taro, and shards. The oldest date back to 6100 years ago. yams were independently domesticated in What's more, in some plants—although not all—starch grains of wild and domesticated New Guinea nearly 7000 years ago {Science, strains are distinct. For example, starch grains of wild chili peppers are 5 to 6 micrometers long, 11 July 2003, p. 180). whereas the domesticated versions are a whopping 20 micrometers. The method is now used So if domestication happened repeatedly, to identify everything from bananas to maize to wild barley and has "breathed new life into what sparked this new relationship between the investigation of early agriculture," says Timothy Denham of Monash University in people and plants? Researchers have pon- Clayton, Australia. -Wl.B. dered the question since the 1920s, when Australian prehistorian V. Gordon Childe

1834 29 JUNE 2007 VOL 316 SCIENCE www.sciencemag.org Published by AAAS NEWSFOCUS pegged the rise of farming to dramatic cli- matic changes now known to have taken Evolution of Food Production From Plants place around 11,500 years ago. That's when FOOD PROCUREMENT FOOD PRODUCTION FROM CROP PRODUCTION the last Ice Age ended and the Pleistocene FROM WILD PLANTS WILD PLANTS DOMINANT DOMINANT period gave way to the much milder Holocene—the geological epoch in which Gathering/collecting Cultivation with Cultivation with Agriculture based we live today, with a warmer, wetter, and including use of fire. small-scale clearance larger-scale land largely or exclusively more stable climate. of vegetation and clearance and on cultivars with Childe's hypothesis sparked a lot of minimal tillage. systematic tillage. greater labor input research. But since his day researchers have into cultivation and swung back and forth between environmen- maintenance of tal explanations and those that focus more on facilities. social changes within increasingly sedentary communities of hunters and gatherers. Decreasing dependence on wild plants for food. All the same, most archaeologists agree that Plant domestication: increasing the origins of agriculture have something dependence on cultivars for food. to do with the broader transition from the Pleistocene to the Holocene. "I am comfortable & seeing this climate change as a precondition for agriculture," says the Smithsonian's Smith. sands of years anywhere in the Near East. land tropics of the New World, as forests But he points out that it can't be the sole Even if the can explain the expanded into once-open areas. Based on explanation for the rise of farming in regions sequence of events at Abu Hureyra, it hasn't the changing availability of both plants such as eastern North America, where been shown to spur farming in other and animals, she calculated that farming squash and several other crops were domes- regions, says David Harris of the Institute of would have been more advantageous than ticated only about 5000 years ago. Archaeology in London. Willcox, in a 2005 foraging right around the time that the Some researchers correlate the origins of review of Near East farming in the journal first domesticated crops appear, about farming not with the early Holocene but with Vegetation History and Archaeobotany, 10,000 years ago. a late Pleistocene global cold snap called the argued that agriculture did not really catch But some archaeologists think that too Younger Dryas, which hit about 13,000 years on until after the Younger Dryas was over much emphasis on environmental expla- ago and sharply reversed warming trends for and the Holocene, with its more stable cli- nations gives short shrift to the less easily more than a millennium. This hypothesis matic conditions, had begun. testable social and symbolic aspects of was prompted by excavations at Abu Indeed, the agricultural lifestyle might human behavior. "We have tended to Hureyra in Syria's Euphrates Valley, led by have been "impossible" during the glacial leave these aspects out and focused on an British archaeologist Andrew Moore, now at conditions of the Pleistocene but "manda- economic paradigm," says archaeologist the Rochester Institute of Technology in tory" during the Holocene, argued ecolo- Joy McCorriston of Ohio State University New York. Abu Hureyra was first occupied gist Peter Richerson of the University of in Columbus. by hunter-gatherers about 13,500 years ago California, Davis, and his colleagues in a In the 1980s, for example, the late and later by early farmers, providing a rare 2001 paper in American Antiquity. One French prehistorian Jacques Cauvin, who window on the transition to agriculture. explanation: Dramatically lower carbon founded the Jales center, proposed that in UCL's Hillman, who analyzed the plant dioxide levels during the Pleistocene the Near East a rise of religious symbolism remains, suggested that the Younger Dryas might have made farming untenable, a changed the relationship between people had a devastating effect on the availability of hypothesis first proposed back in 1995 by and nature and made farming possible. the wild cereals and other plants at the site. botanist Rowan Sage of the University of More recently, archaeologist Brian Hayden Hunter-gatherers eventually disappeared, Toronto. Crops grow more in higher ambi- of Simon Eraser University in Burnaby, and a short time later possible first evidence ent C02 levels. As the Holocene began, C02 Canada, argued that farming had been of farming—larger grains of rye—show up. levels rose by roughly 50%, from 180 parts invented by ambitious hunter-gatherers Hillman and Moore proposed that the per million to 280 ppm in just a few thou- seeking greater prestige and wealth within region's hunter-gatherers invented agricul- sand years, according to polar ice-core their communities. ture to solve food shortages brought on by records. "This would have had a big effect As ideas are batted back and forth, the cold climate. on photosynthesis and plant productivity," some doubt that a global explanation for "Hillman's evidence is convincing," at Richerson says. agriculture will be found. "We are all least for the Near East, says Piperno. "The The Pleistocene-Holocene transition thrashing around, trying to find an expla- Younger Dryas may have been some kind might also have affected decisions about nation for something that is worldwide," of trigger." The worldwide invention of what to eat. Recently, Piperno, Denham, says archaeologist Graeme Barker of the agriculture, Piperno adds, suggests "that Kennett, and others have been studying University of Cambridge in the U.K. "It is there must have been a common set of the choices humans make, borrowing far too simplistic." But that won't stop underlying factors." methods from optimal foraging theory, a researchers from trying. Says Kennett: But not everyone is persuaded by Hillman's Darwinian approach that assumes humans "The transition to agriculture is one of the case for rye domestication. And after its and other animals pursue the most advan- central questions in archaeology. We need possible appearance at Abu Hureyra, tageous strategy for getting food. In a to understand it." domesticated rye doesn't show up for thou- recent study, Piperno looked at the low- -MICHAEL BAITER

www.sciencemag.org SCIENCE VOL 316 29 JUNE 2007 1835 Published by AAAS