Chapter 46 The Llanos de Mojos JOHN H. WALKER INTRODUCTION The Llanos de Mojos (hereafter Mojos) is a tropical savanna in the Bolivian Amazon, shaped by cycles of drought and flood and the labor of generations of farmers. The accounts of Jesuit missionaries from the mid 1600s described large villages with powerful chiefs and influential shamans in the savanna. Was this true? And what kind of economy would have supported these societies? The answers first became apparent in the 1950s with the spread of air travel that revealed artificial earthworks, including agricultural fields, causeways and canals on the landscape below. Recent archaeological research confirms that Mojos was one of several areas within the Amazon basin that had large sedentary populations. Mojos is a research frontier where scant archaeological investigation has been conducted compared to, say, Syria or the Yucatan Peninsula, whose areas are similar. Mojos is a fascinating test case to understand relationships between politics and economics in the Amazon basin over the long term. Because of its location between the Xingu, middle Amazon and Andes, Mojos is also relevant to discussions of cultural history and the movement of groups of people across the continent. Finally, study of the region’s landscapes shows how pre-Columbian peoples acted on their understanding of the relationship between nature and culture to build their environment. This chapter reviews the physical and human geography of Mojos, describes previous archaeological research, and then discusses political and social organization, analyses of interregional contacts, and the creation of anthropogenic landscapes. GEOGRAPHY AND ENVIRONMENT Although it shares many characteristics with other parts of the Amazon basin, the combination of strong seasonality in rainfall and the flooded savanna ecosystems makes Mojos a distinctive geographical region. Mojos is a shallow basin covering approximately 110,000 km2 Handbook of South American Archaeology, edited by Helaine Silverman and William H. Isbell. Springer, New York, 2008 927 928 J.H. Walker Figure 46.1. Location of the Llanos de Mojos (shaded area) within the Madeira River basin of South America. (John Walker) between the Andes to the west and south and the uplands of the Brazilian shield to the east and north (Hanagarth 1993; Langstroth 1996; Denevan 2001) (Figure 46.1). The soils derive primarily from Pleistocene lacustrine sediments, more than two kilometers deep near the Andes and a few hundred meters deep near the Brazilian shield. This thick cover makes stone usable for construction or tools very rare. Topographical relief is minimal, with elevations that only vary between 150 and 170 masl across most of the region. Lying entirely within the tropics, Mojos has a climate characterized by high, stable average temperatures, high humidity and high rainfall. Aside from human activity, the determining ecological feature is the seasonal cycle of flood and drought, which affects soils, plants and animals. Mojos is therefore defined as a humid savanna, a grassland environment with a 2- to 7-month dry season and a total annual rainfall varying between 1,500 and 1,800 mm (Harris 1980). The dry season lasts from May through September, when weeks pass without precipitation. During these months the passage of southern cold fronts occasionally breaks the heat and humidity, dramatically lowering temperatures and sometimes bringing heavy rain. The wet season lasts roughly from November through March. During these months widespread flooding changes the landscape in two different ways. Modern inhabitants call these two types of flooding “water from below” and “water from above.” The first kind of flooding results when tributaries of the Mamoré River rise because of high water levels downstream. When the Mamoré is high, these rivers “back up” and inundate large areas. Some rivers that are easily forded in the dry season are several hundred meters across in The Llanos de Mojos Figure 46.2. Schematic diagram illustrating the patterning of local environments in Mojos. (John Walker) the wet season. The second type of flooding is the result of heavy local rainfall. Rains can leave 20 or 30 cm of standing water on poorly drained terrain, even though that terrain may be at a higher elevation. During the dry season, this water drains more quickly, and these areas remain underwater for a short time. During the wet season, rainfall inundates large areas, making them much more difficult to cross on foot. The interaction of these two processes with local topography creates a mosaic of environments. Low-lying gallery forests along rivers and streams contrast with higher river levees and savannas alongside. Topography and drainage influence the spatial arrangement of plant communities because soils vary with drainage and differences in elevation. Broadly speaking, four types of environments are present today: flooded forests, dry forests, dry savannas and wet savannas (Figure 46.2). Flooded forests are often found in areas of lower elevation along rivers and can be underwater for as much as half of the year. Trees that are submerged for months tower overhead in the dry season. Many flooded forests are located in “galleries” along the rivers, on the sloping banks rather than the adjacent levees. Dry forests are elevated above floodwaters, and are well drained so that rainwater does not accumulate. These forests are often located on levees along the rivers, and in forest “islands” along smaller creeks and abandoned river channels. Dry savannas are also elevated and well drained. Many dry savannas are located on river levees. There are many transitional zones between dry forest and savanna, and it is difficult to separate the two categories clearly. Finally, wet savannas are found both in areas that are elevated but poorly drained, and in lower elevation back slopes. Some soils in the wet savannas accumulate sediment and organic material from annual floods. Older rivers have also cut channels across the savanna, leaving behind levees and back slopes. Differences in earthworks have prompted archaeologists to distinguish four “divisions” within the Mojos region (Denevan 1966, 2001; Erickson 2006) (Figure 46.3). Each of these four divisions has a distinct form of landscape modification: ● ● North. Large raised fields comprise the most imposing element of the built environment. Occupation debris suggests that villages with 1,000–2,000 inhabitants could have been present (Walker 2004) (Figure 46.4). South. The built environment is composed of large mounds, some of which were burial mounds. This is where most archaeological work in Mojos has taken place (Nordenskiöld 1913, 1920, 1924; Bennett 1934; Dougherty and Calandra 1981, 1981–82, 1984; Erickson and Balée 2006; Prümers 2004). Many of the mounds contain urn burials, and some of them contain several dozen burials. The largest mounds are close to 20 m tall, and some of them are hundreds of meters across. 929 930 J.H. Walker Figure 46.3. Map illustrating the regional diversity of anthropogenic landscapes in Mojos. (Adapted from Denevan 1966: fig. 4) ● ● East. Earthworks include long causeways, zig-zag causeways and circular ditches (Erickson 1996; Dougherty and Calandra 1985). The circular ditches are distinct from the landscapes modified by the zig-zag causeways, which Erickson interprets as weirs for the harvesting of fish (Figure 46.5). West. Here are perhaps the most complex built environments, where canals, causeways, mounds and several kinds of raised and ditched fields were combined into integrated landscapes (Erickson 1980; Erickson and Walker ms.). Many raised fields seem well suited to the production of root crops, such as manioc or New World taro. Because raised fields reduce the effects of flooding, they help protect root crops from rotting. Botanical remains from a raised field excavation show that Xanthosoma, Annatto and Ilex were being grown: a root crop, a dye, and a stimulant, respectively (Erickson 1995). This roster suggests that multi-cropping was a cultivation strategy, and The Llanos de Mojos Figure 46.4. Reprojected aerial photograph of large raised fields and settlement north of Santa Ana del Yacuma, along the Iruyañez River. (John Walker) Figure 46.5. Long causeways near Baures. a. Aerial photograph; b. Diagram (adapted from Erickson 2000: fig. 4). (John Walker) 931 932 J.H. Walker that industrial products were as important as food crops. Mojos is within the southern Amazon basin, which has been proposed as the “hearth” of both manioc and peanut cultivation (Olsen and Schaal 2001; Jarvis et al. 2002). Deposits of anthrosols indicate large areas of human occupation along the large tributaries and main course of the Amazon River. Appearing as a distinct soil type on national surveys, anthrosols are nuanced evidence both for settlement and for intentional modification of soils for agricultural uses (Lehman et al. 2003; Glaser and Woods 2004). The human geography of Mojos is also complex. According to the Summer Institute of Linguistics, thirteen languages are present in Mojos today, including groups classified as Arawak (Baure, and Mojo, which is split into Ignaciano and Trinitario), Tupi (Guarayu and Siriono), Tacanan (Ese Ejja and Cavineña), Panoan (Chácobo), and isolated languages (Canichana, Cayuvava, Itonama, Movima and T’simane) (Gordon 2005). Several of these are extinct or nearly extinct (Baure, Canichana, Cayuvava and Itonama). This situation suggests that before the European conquest a variety of languages were present, as in other regions within the upper Amazon. A recent study of mitochondrial DNA among modern populations suggests that considerable genetic diversity is represented (Bert et al. 2004). The study is a significant step and continued research will provide an independent line of evidence to compare with linguistic and archaeological data about population history. HISTORY OF RESEARCH The first archaeologist to carry out research in Mojos was Erland Nordenskiöld, who excavated several large mound sites near Trinidad and surveyed material culture throughout the region (Nordenskiöld 1913, 1924). His research was the basis for Métraux’s summary of eastern Bolivian archaeology and the ceramic data from these excavations were later interpreted in light of evidence from across the entire South American lowlands (Métraux 1942; Howard 1948). Denevan’s subsequent geographical studies of earthworks revolutionized the study of pre-Columbian Mojos, introducing and carefully documenting the scale of landscape modification (Denevan 1966, 2001). Although Denevan did not excavate, his aerial and pedestrian survey of earthworks completely changed the understanding of preColumbian Mojeño societies and placed them in a larger context of intensive agriculture throughout the Americas. His book is still the most useful reference on Mojos. Since Denevan drew attention to the region, a number of archaeological projects have conducted research in the area. Some have focused on documenting and analyzing ceramic industries, attempting to establish chronologies (Dougherty and Calandra 1981, 1981–82; Calandra and Salceda 2004). A few projects have carried out fieldwork at a single location (Arnold and Prettol 1998; Bustos 1978; Ryden 1941). Three projects currently active in Mojos are the Finnish Archaeological Project (Siriäinen and Korpisaari 2002, 2003), the KAVA project (Prümers 2004) and the Proyecto Agro-Arqueológico del Beni (PAAB) (Erickson 1995, 2006; Erickson and Balée 2006; Erickson and Walker ms.; Walker 2004). The Finnish project has carefully documented a fortified site in the far north along the Beni River. The KAVA project has used careful analysis of stratigraphy in combination with ceramic analysis to interpret the large Mendoza mound in southeastern Mojos. The PAAB project has focused on archaeological landscapes, using remote sensing, survey and test excavations in locations throughout Mojos. Archaeologists working in Mojos also depend on information from other sources: historical, ethnographic, and geographic. The Llanos de Mojos Mojos culture history continues to be debated by the archaeologists. Their focus is on three critical issues: social and political organization, contact between Mojos and neighboring areas, and the creation of anthropogenic landscapes. SOCIAL AND POLITICAL ORGANIZATION Mojos drew the attention of European explorers in the sixteenth century as they sought opportunities to extract riches from indigenous states very quickly. El Dorado, Paitití and Gran Mojos were all names for a mythical kingdom of gold, or for its leader. Early historical documents describing the area were written in the course of these explorations, and later as part of Jesuit missionary work in Mojos. These documents described large villages and powerful chiefs; one missionary claimed in a matter-of-fact way to have met this king himself. The image of El Dorado has recently been the rhetorical focus of a discussion of social and political organization in Amazonia (Meggers 2004; Stahl 2002; Heckenberger et al. 1999). In Mojos, this discussion considers two lines of evidence, historical and archaeological. Denevan (1966, 2001) and Block (1994) summarize the early historical sources. The documents describe groups of people living in large villages (larger than 200 people, and in many cases larger than 500 people); alliances between large villages; and political leaders with considerable power, including the power to kill offenders. These sources originally led Julian Steward, in the Handbook of South American Indians, to classify the societies of Mojos with “Circum-Caribbean” societies, in contrast to simpler “Tropical Forest” societies that characterized most of Amazonia (see Chapter 1 in this volume). Historical documents must be carefully interpreted, but this does not mean that early descriptions of Mojeño peoples should be discarded. Jesuit accounts in particular, in their comprehensiveness and detail, provide relevant information for the seventeenth century. These documents (which have been analyzed only in part) indicate that a variety of societies were present in Mojos at the time of contact. They describe hunters, accomplished farmers (but say little about earthworks), weavers, potters, feather workers, fishermen and gatherers. They contrast Indians who wore clothes and were “gentle,” with “wild,” naked cannibals. The ethnohistoric record suggests that Mojos was a multilingual and perhaps multiethnic region, with many different forms of social and political organization present at the same time. The outlines of this argument changed in the late 1950s, when a number of different researchers independently recorded the existence of earthworks on a regional scale. Denevan showed that pre-Columbian people had produced many artificial landscapes across Mojos, of many types, and over a great area. This placed Mojos, as an example of raised field agriculture, alongside many important locations around the Americas, including the Andean altiplano, Colombia, Venezuela, the Guianas, West Mexico, the Valley of Mexico, the Maya region, and the upper Midwestern United States (see Chapters 11, 12, 13, 16, 17, 23 in this volume). Mojos is an example of the intensive use of the landscape by indigenous Amazonians, implying greater social and political complexity than generally admitted for pre-Columbian Amazonia. The anthropogenic landscape allows this organization to be studied on three scales: the regional scale of broad divisions in the archaeological record, the local scale of the creation and maintenance of the agricultural landscape, and the intermediate scale of large political units. At the regional scale (across tens to hundreds of kilometers), differences between the four landscape types correspond to differences in how much labor was required to reshape 933 934 J.H. Walker the terrain, and how much coordination would be required to use and maintain it. For example, the amount of earth moved per unit area can serve as a proxy for labor mobilization. In rough terms, the average figure is highest for the west, followed by the east, then the north and finally the south. The coordination of different types of earthworks and water control also would have been the most complex in the west and east, and much less so in the north and south. Differences in labor costs vary within each area, but the general trends suggest that there were significant differences in the organizational solutions devised by farmers. This in turn suggests diversity in the social and political organization of Mojos raised field farmers. At the local scale (across tens to hundreds of meters), there is similar evidence for diversity among farmers. In the northern region, the spatial patterns of large raised fields were analyzed in conjunction with estimates of raised field construction costs and production. This suggests that groups of between 30 and 100 people might have been responsible for the construction of many large fields, in discrete spatial groups. These groups of people are likely to correspond to a local social unit (Walker 2000, 2004). A similar analysis is underway in western Mojos, along the Apere River, where there are spatially discrete areas of raised fields that may correspond to local social units (Erickson and Walker ms.). The complexity of the built environment creates many opportunities for analysis and for the study of spatial pattern. Scholars of many different theoretical perspectives agree that landscapes contain information about the social and political organization of the peoples who produce it. These minimal units within Mojos landscapes are indicators of the smallest units of social and political organization. Finally, attention should be focused on the intermediate scale between the local and the regional, because this is where the evidence of what is usually called social and political organization will be found. If the predecessors of the “chiefdoms” described in the ethnohistoric literature were associated with raised field construction and maintenance, then it will be possible to discern outlines of large units within the spatial patterns of earthworks. Connections between political organization and agricultural organization are to be investigated and not assumed. In one case where settlement evidence is associated with raised field patterning, it appears that large settlements were associated with a comparatively simple agricultural landscape (Walker 2004). CONNECTIONS TO OTHER REGIONS Mojos is near the geographic center of South America, between the Andes mountains and the Brazilian highlands and upstream from the middle Amazon. Discussions of cultural contact and migration in South America have focused on the origins of complex society at the mouth of the Amazon and the movement of language groups throughout the Amazon basin (Lathrap 1970, 1977; Heckenberger 2005; Hornborg 2005). Enough information is available from several other areas to place Mojos in a larger context. Altiplano The well-known societies of the altiplano and southern Andes were close to Mojos in both space and time. Tiwanaku is less than 300 km from western sites associated with agricultural earthworks, and several Mojeño contexts date to Tiwanaku times and earlier (Walker 2004). The ceramics recovered in Nordenskiöld’s excavations made it possible to speculate The Llanos de Mojos on links to the Andes. Although some ceramic similarities exist, possibly indicating that the southeastern mound cultures were derivative from the highlands, strong arguments have been advanced that these ceramics have more in common with Amazonian examples from further downstream and with other upper Amazonian ceramics (Bennett 1936; Howard 1948; Lathrap 1970). Raised field agriculture is well documented in Tiwanaku (see, e.g., Kolata 1993). As the scale and antiquity of raised field use in Mojos becomes clearer, this becomes relevant to the question of the relationship between Tiwanaku and Mojos (Hornborg 2005). Such systems represent the accumulation of labor over centuries, and it is unlikely that the technology diffused rapidly either from Mojos to the altiplano, or vice-versa. The slow process of creating agrodiversity through building agricultural landscapes suggests that there was no single moment of innovation in the development of raised field agriculture. There are many differences in form and function between raised fields, associated with different agricultural problems and different social and political organizations. A Finnish-Bolivian project has recovered evidence of an Inca presence, well dated to the Late Horizon, at a site along the Beni River in far northern Mojos (Siiriäinen and Korpisaari 2002, 2003). Both the ceramic evidence and radiocarbon dating are convincing. The nature of the occupation is unclear, although it seems that material evidence of the Incas is not widespread, and that highlanders were interested in lowland resources. The location along the Beni, a major river that reaches into the highlands is unsurprising. Being near its confluence with the Madre de Dios River would have furthered trade and contact with large areas of the lowlands, both in Mojos and further downstream. Intensive agriculture was widespread in Mojos, and estimates of its extent continue to expand, while those of its antiquity continue to deepen. Improved remote sensing resources in combination with survey will allow a more accurate estimate in the future, but new earthworks continue to be discovered, and it is likely that agricultural earthworks in Mojos were of the same order of magnitude as those in the altiplano. Mojos represents both a comparative case of the development and abandonment of raised field farming and possibly a shared history with the altiplano. Eastern Slopes Between Mojos and the altiplano lie the eastern slopes of the Andes, a region sometimes characterized as a frontier, but which clearly has its own identity. An increasing amount of research is being conducted in the humid valleys of the eastern slopes. Stone axes, which are rare but widespread in Mojos, may indicate contact with the eastern slopes. Material for ground stone tools does not appear naturally in Mojos, and the best sources are in the Andes (although the Brazilian shield is another possibility). Such axes could have had symbolic value, but because they do not occur in large numbers, it is unlikely that they played a pivotal role in agriculture. Stone axes would have been less effective than the use of fire for clearing vegetation. Slash-and-burn agriculture probably was not widespread before the arrival of the Spanish, and there are good reasons to believe that this form of agriculture developed after the advent of metal tools (Denevan 2001). There are similarities between Mojos ceramics and some of the ceramic industries present in the altiplano and the Bolivian eastern slopes before Tiwanaku (Lathrap 1970). These similarities point to the ease with which cultural traits, and ceramic attributes in particular, may have passed between the lowlands and highlands. Environmental differences seem to make strong boundaries, but they were crossed repeatedly in the past, as they are 935 936 J.H. Walker being crossed today. It may be more accurate to think of these regions as all being part of a single area with regard to many characteristics. Xingu More archaeological work is being done in the highlands of Brazil drained by the Xingu River, and there are several parallels to some Mojeño cultures. Heckenberger (2005; see Chapter 47 in this volume) has detailed the antiquity and extent of settlement in the Xingu. Large settlements are associated with ringed ditches and long causeways, and connections between the archaeological evidence and modern Arawak speakers seem strong (see also Hornborg 2005). The landscapes of eastern Mojos also include ringed ditches and several kinds of causeways (Erickson 2000). In addition to being the region within Mojos closest to the Xingu, the Baure speakers from whom the region gets its name are Arawak speakers. Arawak speakers (including both Baure and Mojo speakers) make up a large part of the diversity of cultures in Mojos, but they are not exclusively associated with either earthworks or political complexity. The Movima, Itonama, Canichana and Cayuvava are other modern language groups with similar associations that lived across Mojos and are known from ethnohistoric sources. Middle Amazon There are significant parallels between the ceramics of the middle Amazon and those of Mojos, and this information is relevant to the reconstruction of cultural history throughout the Amazon basin. New research is taking place along the middle Amazon that builds on previous studies that focused primarily on ceramics. Barrancoid style ceramics, first described along the lower Orinoco in Venezuela, bear a resemblance to many ceramic styles across the Amazon basin. Barrancoid traits have been noted among the ceramics excavated by Nordenskiöld, and there may be other Barrancoid assemblages as well (Walker 2004). The relationship between Icotiara, Guarita and other ceramics from the middle Amazon to Mojeño ceramics suggests that Mojeños were a part of regional processes. ANTHROPOGENIC LANDSCAPES Archaeological studies in South America and, specifically, Amazonia have frequently focused on the effects of the environment on human society. Particularly due to preservation in the coastal deserts, the great antiquity of plant domestication in South America is well documented. The agricultural earthworks, savannas and forests of Mojos are examples of a separate but related phenomenon, the domestication of landscape. Domestication of the landscape means the use of tools such as fire and the control of water to change distributions of plants and animals (Erickson 2006). This makes the environment much more useful, with fruits more easily harvested, animals more easily hunted, fish more easily caught, and distances more easily traversed. This depends more on the creation of desirable attributes in the environment, and less on the manipulation of genetic material to push species across the boundary between “wild” and “domesticated.” Mojos is one of many places where this process has taken place, and three examples demonstrate how the landscape was domesticated. The Llanos de Mojos First, the modern Siriono people of southeastern Mojos use a wealth of knowledge of forest species, modifying the distribution of economically useful trees (Erickson and Balée 2006). Concentrations of peach palm trees show that much of the “pristine” forest in eastern Bolivia is, in fact, the result of generations of modification. Peach palms have many economic uses, from consumption of their high-fat fruits to the making of palm wine. Because the life cycle of trees is long, genetic modification of tree crops is unlikely. Instead, the imprint of human activity takes the form of the location of trees within the forest. Balée estimates that a considerable portion of the forests of the Amazon basin may have been modified in this way. Second, the construction of causeways in the flooded savanna modifies the environment in several ways. It makes low-lying areas easier to traverse on foot, and could make higher areas easier to pass in a canoe, using the canal alongside the causeway. In the wet season, when large areas are inundated, causeways can change flows of water over many square kilometers. A well-maintained causeway only two meters high, if it crosses between the levees of two rivers, can impound many square kilometers of water, maintaining inundation and harvesting rainfall for agricultural uses. Third, the built environment may have made hunting and fishing easier. In the wet season, floods force prey animals onto scarce dry land. Modern inhabitants keep prey “on the hoof” in such situations, and hunt them at leisure (as long as the flooding conditions persist). In the north, for example, there is a strong correlation between dry forest and human settlement and it might have been feasible to use the landscape to predict the movements of prey animals (Walker 2004). To a much greater extent, the builders of “zig-zag” causeways in the northeast could have used their flooded savannas to harvest fish (Erickson 1996). The canals in raised fields also create environments in which frogs, snails and insects flourish, attracting waterfowl as well. Finally, raised field crops themselves would attract animals, which in many cases could be resources rather than pests. The modification of the savanna makes Mojos a clear example of how native Amazonians modified their environment to make landscapes. Domestication of the landscape is not unique to Mojos by any means, but this case highlights the importance of processes that change the availability of resources, without necessarily changing the genetic code of plants and animals. CONCLUSION The archaeological record of pre-Columbian Mojos has only been sketched in this chapter [Note 1]. The combination of well-preserved earthworks and well-developed ceramic industries provides an opportunity to combine studies of the archaeological landscape with cultural-historical goals. Such a synthesis can provide common ground between local interests in archaeology and international interests in questions of the Amazonian landscape. The variety of connections that can be drawn to other parts of South America suggests that there were distinctive aspects to a Mojeño way of life, but also shared histories with groups of people from outside the savanna. Evidence for the complexity of anthropogenic landscapes in South America continues to accumulate from widely divergent settings: geoglyphs on the Peruvian coast; fields and terracing in the Andes; and raised fields and earthworks from around the lowlands. Similarly, evidence for the variety and complexity of social organization in South America, well attested in the Andes, is also mounting for the Amazonian lowlands. These two 937 938 J.H. Walker features of the archaeological and ethnohistoric record—agricultural landscapes and complex political organizations—have been the defining characteristics of Mojos. As more evidence is assembled and compiled, the unique position of Mojos will be seen as the norm for much of the Amazon basin: a complex, multilingual mix of different economic strategies and ethnic groups, with a turbulent history over many millennia. At that point, the archaeology of the Llanos de Mojos and of the Amazon basin will have matured. NOTE 1. The reader will find extensive bibliographies for further reading in Erickson (2006) and Walker (2004). REFERENCES Arnold, Dean E. and Kenneth A. Prettol, 1988, Aboriginal earthworks near the mouth of the Beni, Bolivia. Journal of Field Archaeology 15: 457–65. Balée, William and Erickson, Clark L. (eds.), 2006, Time and Complexity in Historical Ecology. Columbia University Press, New York. Bennett, Wendell C., 1936, Excavations in Bolivia. The American Museum of Natural History, New York. Bert, F., A. Corella, M. Gené, A. Pérez-Pérez and D. Turbón, 2004, Mitochondrial DNA diversity in the Llanos de Moxos: Moxo, Movima and Yuracare Amerindian populations from Bolivia lowlands. Annals of Human Biology 31 (1): 9–28. Block, David, 1994, Mission Culture on the Upper Amazon: Native Tradition, Jesuit Enterprise and Secular Policy in Moxos, 1660–1880. University of Nebraska Press, Lincoln. Bustos, Victor, 1978, Investigaciones Arqueológicas en Trinidad, Departamento del Beni. Instituto Nacional de Arqueología, Publicación No. 22, La Paz. Calandra, Horacio Adolfo and Susana Alicia Salceda, 2004, Bolivian Amazonia: archaeology of the Llanos de Mojos. Acta Amazonica 34 (2): 155–163. Denevan, William M., 1966, The Aboriginal Cultural Geography of the Llanos de Mojos of Bolivia. IberoAmericana, No. 48. University of California Press, Berkeley. , 2001, Cultivated Landscapes of Native Amazonia and the Andes. Oxford University Press, Oxford. Dougherty, Bernardo and Horacio Calandra, 1981–82, Excavaciones arqueológicas en la Loma Alta de Casarabe, Llanos de Moxos, Departamento del Beni, Bolivia. Relaciones de la Sociedad Argentina de Antropología 14 (2): 9–48. Buenos Aires. Erickson, Clark L., 1980, Sistemas agrícolas prehispánicos en los Llanos de Mojos. América Indígena 40 (4): 731–755. , 1995, Archaeological perspectives on ancient landscapes of the Llanos de Mojos in the Bolivian Amazon. In Archaeology in the American Tropics: Current Analytical Methods and Applications, edited by Peter Stahl, pp. 66–95. Cambridge University Press, Cambridge. , 2000, An artificial landscape-scale fishery in the Bolivian Amazon. Nature 408: 190–193. , 2006, The domesticated landscapes of the Bolivian Amazon. In Time and Complexity in Historical Ecology, edited by Clark L. Erickson and William Balée, pp. 243–301. Columbia University Press, New York. and William Balée, 2006, The historical ecology of a complex landscape in Bolivia. In Time and Complexity in Historical Ecology, edited by Clark L. Erickson and William Balée, pp. 199–241. Columbia University Press, New York. and John H. Walker, ms. The Archaeology of Landscapes in the Bolivian Amazon. Manuscript in preparation. Faldín, Juan, 1984, La arqueología beniana y su panorama interpretivo. Arqueología Boliviana 1: 83–90. Glaser, Bruno and William I. Woods (eds.), 2004, Amazonian Dark Earths: Explorations in Space and Time. Springer, Verlag, Berlin. Gordon, Raymond G., Jr. (ed.), 2005, Ethnologue: Languages of the World, Fifteenth Edition. Summer Institute of Linguistics International, Dallas. The Llanos de Mojos Hanagarth, Werner, 1993, Acerca de la geoecología de las sabanas del Beni en el noroeste de Bolivia. Instituto de Ecología, La Paz. Harris, David R., 1980, Human Ecology in Savanna Environments. Academic Press, New York. Heckenberger, Michael J., 2005, The Ecology of Power: Culture, Place and Personhood in the Southern Amazon, AD 1000–2000. Routledge, New York. , James B. Petersen, and Eduardo G. Neves, 1999, Village size and permanence in Amazonia: two archaeological examples from Brazil. Latin American Antiquity 10 (4): 353–376. Hornborg, Alf, 2005, Ethnogenesis, regional integration, and ecology. Current Anthropology 46: 589–620. Howard, George D., 1948, Prehistoric Ceramic Styles of Lowland South America, Their Distribution and History. Yale University Publications in Anthropology, No. 37. New Haven. Jarvis, Andy, Luigi Guarino, David Williams, Karen Williams, Israel Vargas, and Glenn Hyman, 2002, Spatial analysis of wild peanut distributions and their implications for plant genetic resources conservation. Plant Genetics Resources Newsletter 131: 28–34. Kolata, Alan L., 1993, The Tiwanaku. Portrait of an Andean Civilization. Blackwell, Cambridge, MA. Langstroth, Robert, 1996, Forest Islands in an Amazonian Savanna of Northeastern-Bolivia. Ph.D. dissertation. Department of Geography, University of Wisconsin, Madison. Lathrap, Donald W., 1970, The Upper Amazon. Praeger, London. Lehmann, Johannes, Dirse Kern, Bruno Glaser and William I. Woods, 2003, Amazonian Dark Earths. Kluwer Academic Publishers, Dordrecht. Meggers, Betty J., 1971, Amazonia: Man and Culture in a Counterfeit Paradise. Aldine, Chicago. , 2004, Response to paradigms in paradise: revisiting standard Amazonian prehistory. Review of Archaeology 25: 31–39. Métraux, Alfred, 1942, The native tribes of eastern Bolivia and western Matto Grosso. Bulletin of the Bureau of American Ethnology, No. 134. Washington, D.C. Nordenskiöld, Erland, 1910, Archaologische Forschungen im Bolivianischen Flachlande. Zeitschrift für Ethnologie 42: 806–822. Berlin. , 1913, Urnengraber und Mounds im Bolivianischen Flachlande. Baessler Archiv 3: 205–255. Berlin. , 1924, The Ethnography of South America as seen from Mojos in Bolivia. Comparative Ethnological Studies, No. 3. Goteborg. Olsen, Kenneth M. and Barbara A. Schaal, 2001, Microsatellite variation in cassava (Manihot esculenta, Euphorbiaceae) and its wild relatives: further evidence for a southern Amazonian origin of domestication. American Journal of Botany 88 (1): 131–142. Prümers, Heiko, 2004, Hügel umgeben von ‘Schönen Monstern’: Ausgrabungen in de Loma Mendoza (Bolivien). Expeditionen in Vergessene Welten, pp. 47–78. AVA-Forschungen Band 10. KAVA, Bonn. Ryden, Stig, 1941, A Study of the Siriono Indians. Elanders Boktryckeri Aktiebolag, Göteborg. Siiriäinen, Ari, and Antti Korpisaari, 2002, Reports of the Finnish-Bolivian Archaeological Project in the Bolivian Amazon. Department of Archaeology, University of Helsinki. and Antti Korpisaari, 2003, Reports of the Finnish-Bolivian Archaeological Project in the Bolivian Amazon, II. Department of Archaeology, University of Helsinki. Stahl, Peter W., 2002, Paradigms in paradise: revising standard Amazonian prehistory. The Review of Archaeology 23 (2): 39–49. Walker, John H., 2001, Work parties and raised field groups in the Bolivian Amazon. Expedition 43 (3): 9–18. , 2004, Agricultural Change in the Bolivian Amazon/Cambio Agrícola en la Amazonía Boliviana. University of Pittsburgh Latin American Archaeology Publications. Pittsburgh. 939