This outrage reflected a response not only to the crime but also tothe victim. Over the past decade, Petersen, Neves, and their band ofarchaeologists had become local heroes, earning the appreciation of thesurrounding community during seasonal digs conducted on the peninsulathat separates the Rio Negro and Amazon rivers. At more than 100 sitesacross the peninsula, Petersen and his colleagues had unearthedevidence of early civilizations that were far more advanced, far morebroadly connected, and far more densely occupied than that of the smallbands of nomadic hunter-gatherers previously hypothesized for theregion. Before the Europeans arrived, this peninsula in the heart ofthe Amazon was home to communities with roads, irrigation, agriculture,soil management, ceramics, and extended trade. These civilizations,Neves says, were as complex as the southwestern Native Americancultures that inhabited Chaco Canyon and Mesa Verde. But due to thescarcity of stone in the Amazon, the people built with wood, and overtime the structures disintegrated, leaving little evidence of theculture.

One legacy remains, however: their soil. Terra preta de Indio—Portuguesefor “Indian black earth”—is prized among local farmers, and it is adirect contribution of the vanished Amazonian cultures. While mostAmazonian earth is notoriously nutrient poor, yellowish, sterile, andunscented, there are extensive patches of soil that are mysteriouslydark, moist, fragrant, and filled with insects, microbial life, andorganic matter. Scholars have come to realize that by devising a way toenrich the soil, the early inhabitants of the Amazon managed to createa foundation for agriculture-based settlements much more populous thanscholars had thought possible.

Petersen had called the soil a gift from the past; he believed thatstudying it would reveal the region’s past cultures in a new, much morecomplex light. At the time of his death, he and his colleagues had beendeveloping a workshop for teachers in the region on the science andarchaeology surrounding terra preta. The discovery held meaning formore than archaeologists, however. Figuring out the composition of darkearth and how it was formed offers a way to improve soil fertility fortoday’s small farmers and also curb carbon emissions from the firesthat these farmers set to clear the forest. Yet after Petersen’s murderthe project was shut down, and those who had gathered for both theworkshop and the August field season were sent home. Had the Amazonturned too dangerous for its inhabitants to learn their own ancestors’secrets?

No one ever saw those populations again.

The first scientists who studied the Amazon found little evidence tosupport Orellana’s claims of large populations. As productive and lushas the rain forest appeared, the soil it stood on was, literally, dirtpoor. Betty J. Meggers, a Smithsonian archaeologist who worked in theAmazon from the mid-1900s on, referred to the region as “a counterfeitparadise.” The apparent lushness existed only because the vegetationwas so good at sucking up every speck of nutrient released fromdecaying leaves. Any nutrients remaining washed away in the frequentrains. In short, the local soil was ill-suited for agriculture, andwithout agriculture, societies would remain small.

The outcome of the slash-and-burn agriculture that is practicedtoday supports that view. In this method, settlers cut down forests andburn the cover in order to grow crops in the nitrogen- and mineral-richash that is left behind. But the soil in these denuded plots isproductive for only a few years before it reverts to its originalnutrient-scarce state. This lack of agricultural potential helpedpopularize a model of the pre-Columbian occupation that anthropologistMichael Heckenberger of the University of Florida at Gainesvilledescribes as “the myth of Stone-Age savages frozen at the dawn of time.

In the months to follow, Heckenberger and Petersen worked to developa project around the site with Neves, and the Central Amazon Projectwas born. The group wrote up institutional agreements, receivedpermits, got seed money, and began fieldwork in 1995. When their earlyfinds seemed to challenge Meggers’s vision of the pre-Columbian Amazon,they turned to a model proposed by Donald Lathrap, an archaeologist atthe University of Illinois. During the 1960s, he had hypothesized—basedon linguistic and ceramic evidence—that the confluence of the Amazon,Rio Negro, and Madera rivers may have been the heart of anagriculturally based civilization that once extended from the Caribbeanto southern Brazil.

Heckenberger left the Central Amazon Project in 1999 to concentrateon the Kuikuro, but Petersen and Neves continued the effort. Over theyears, excavation expanded beyond the original site, called Açutuba, toinclude other spots in a 20-square-mile area on the peninsula that jutsbetween the Amazon and Rio Negro rivers. Gradually funding switchedfrom U.S. institutions to predominantly Brazilian, and Neves took moreof an oversight role. But Petersen continued to participate everysummer. In a 2005 interview with the Vermont Quarterly, he called hiscentral Amazon research “some of the richest, most exciting archaeologyanywhere on the planet.”

In the afternoon, Neves takes me to the Hatahara site, about eightmiles from Donna Stella, which contains traces of a more sedentary wayof life. Here pottery shards litter the ground. At one of the site’stwo excavations, the broken bits stick out from the earthen walls of alarge square pit. The layers of protruding pottery are so tight andthick they look almost like wall coverings.

The Hatahara site contains traces of four separate occupations byfarmers. The people of the Açutuba phase, who are thought to have comefrom the Caribbean coast, may have been the first to settle the area,living in villages up to five acres in size, each supporting 100 to 200people. Neves believes that they lived on fish, wild meats, palms,fruits, and manioc, a root that can grow in poorer soils year-round.Although this early culture had only rudimentary agriculturalpractices, Neves says, the plant and animal refuse they were leavingprovided the substrate for future terra preta soils.

Helena Lima, a Ph.D. student at the University of São Paulo, istrying to pinpoint when and how this soil was used to support greaterpopulations in the Amazon Basin. She sees a stark difference betweenthe Açutuba phase, from 300 B.C. to A.D. 400, and the later Manacapuruphase, from A.D. 400 to 900. “The Manacapuru were the first people whoreally changed the soil,” she says.

The terra preta soils at Hatahara and the other sites are made froma mixture of plant refuse and animal and fish bones, along with largequantities of charcoal that were deposited after settlers used stoneaxes and slow-burning fires to clear forest. Such smoldering firesproduced more charcoal than ash. The charcoal, soot, and other carbonremains (collectively called biochar) retained nutrients, particularlypotassium and phosphorus, that are limited in tropical soils. Theresulting improvement in soil fertility may have allowed the land tosupport a larger, more stable crop-based population, although studiesof fossilized pollen have not yet revealed the specific plants theycultivated.

The next phase of settlement, the Paredão, occurred from about A.D.700 to 1200. Neves suspects that the Paredão were outsiders from thesouth. The occupation is peaceful, though, and the Paredão and theManacapuru lived and traded with each other.

The Paredão exploited terra preta soils even more than theirpredecessors. While the initial settlers in this area may have createdthe dark soils by accident, William Woods, director of environmentalstudies at the University of Kansas, says that “at some point theyrecognize their importance and start to promote them.” Over time, thevillages of the Paredão become larger, denser, and surrounded byagricultural fields. Populations grow into the thousands at sitesranging from 5 to 40 acres. The lack of fortification strongly suggeststhat the groups lived in peace.

But around 1200, the Guarita people from the east threatened toattack, and the Paredão built defensive structures. This phase lastedinto the time of European occupation. Over the same period, the Paredãovanished. The Guarita apparently moved in from areas near the mouth ofthe lower Amazon, which have even more terra preta soils than theParedão, and brought with them wilder, multicolored styles of pottery.

“They are like the barbarians attacking the Romans,” Nevesspeculates. He suspects that the newcomers may also have had a valuablepossession—corn. This new, more nutritious staple requires bettersoils, and it is not unreasonable to suspect that the Guarita drove theParedão out to take over their valuable cropland, built on terra preta.

Analysis of buried human remains suggests that the inhabitants ofall four occupations were robust—a well-being that extended even afterdeath. In the site’s remains, Anne Rapp and her husband, Claide Moraes,both students at the University of São Paulo, find evidence that hintsat ceremonial procedures, priests, and perhaps a cottage industry offunerary artisans as well.

Although the central Amazon may not have been the heart of themassive empire that Donald Lathrap envisioned, these cultural tracessuggest that the Amazonians managed to flourish in a formidableenvironment—and terra preta may have been an important component ofthat success.

As thrilling as this evidence is to archaeologists, it may also havevery practical importance as a modern weapon against some of our mosturgent ecological problems. Soil scientist Johannes Lehmann of CornellUniversity believes that the mysterious dark earth holds clues tocreating sustainable farming practices and even to combating globalwarming.

Lehmann explains that nutrients from plant and animal remains—likenitrogen, phosphate, and potassium—bind to charcoal or biochar,drastically reducing how much is washed away by the constant rains. Itis a gradual process that begins with the charcoal breaking down in thesoil over time. Tiny pores in the charcoal, along with changes in itschemistry, provide more surfaces for nutrients to adhere to, which inturn encourages microorganisms to colonize the soil. “With a handful ofbiochar you can keep many more nutrients in the soil than with ahandful of mulch or compost. It is like mopping up nutrients with amagnet that looks like a sponge—that is, it has high surface area likea sponge but can attract a thin layer of material like a magnet,”Lehmann says.

Although the complete transformation of soil ingredients into trueterra preta may take several years, soil scientists have shown that themixture can have immediate benefits when added to nutrient-poor soils.Experiments outside of Manaus have shown that the yield in plotstreated with charcoal and fertilizer (whichs contains plant nutrients),a mix similar in composition to terra preta, was double the yield ofplots treated with fertilizer alone.

In 2001 Petersen published a paper reporting one example of a farmerwho had cultivated crops on terra preta soils near Açutuba for 40 yearswithout ever adding any fertilizer. “That’s incredible,” Woods says.“We don’t get that in Iowa.” A few years of Amazonian rains will washaway the nutrient-laden ash from land that was cleared byslash-and-burn techniques, but the charcoal in the terra preta soilspersists. The terra preta soils at the Central Amazon Project goes backin many places as much as 2,500 years.

Creating new terra preta in the Amazon today would have severaladvantages, Lehmann says. First, because the enriched soil remainsfertile for a long time, its use would discourage farmers from movingon and burning more forest to open up new fields. Second, because ofthe added charcoal, terra preta holds up to 10 times as much carbon asunaltered soils. The late Wim Sombroek—a legendary soil scientist whoselong interest in terra preta earned him the epithet “the godfather ofdark earth”—began to wonder if dark earth could be used to sequestercarbon. Lehmann’s studies have shown that it can: Fifty percent of theoriginal carbon in plants and trees used to make biochar remains in theterra preta soils after the conversion.