W What do Amazonian Shellmounds Tell Us About the Long-Term Indigenous History of South America? Francisco Antonio Pugliese1, Carlos Augusto Zimpel Neto1,2 and Eduardo Góes Neves1 1 Laboratório de Arqueologia dos Trópicos, Museu de Arqueologia e Etnologia, Universidade de São Paulo, São Paulo, SP, Brazil 2 Departamento de Arqueologia, Universidade Federal de Rondônia, Porto Velho, Brazil Entre les temps différents de l’histoire, la longue durée se présente ainsi comme un personnage encombrant, compliqué, souvent inédit. L’admettre au coeur de notre métier ne sera pas un simple jeu, l’habituel élargissement d’études et de curiosités. Il ne s’agira pas, non plus, d’un choix dont il serait le seul bénéficiaire. Pour l’historien, l’accepter c’est se prêter à un changement de style, d’attitude, à un renversement de pensée, à une nouvelle conception du social. Braudel (1958) be understood as meaningful places (Zedeño and Bowser 2009) where human occupation has persisted for millennia (Schlanger 1992). The convergence of factors which led to the emergence and countless waves of reoccupation of these sites has been studied since the origins of archaeology (Steinen 1886, 1894). In the Amazon basin, the earliest shellmounds date at least to the eighth millennium BP and occur in three contexts: in the Lower Amazon (LA), along the Estuary and Coastal Areas (ECA), and in Southwest Amazonia (SWA), along the tributaries of the upper Madeira River (Fig. 1). In common, these sites share some of the earliest ceramics produced in the New World. Outside of the Amazon, there is also a correlation between early ceramic production and shellmound construction in places such as northern Colombia (Reichel-Domaltoff 1985) and Southeastern United States (Sassaman 1993, 2004a, b). The reason why such a pattern developed is one of the most interesting questions of New World archaeology. The Long History of Shellmounds in the Amazon Introduction Shellmounds (also known in Brazilian archaeology as sambaquis), understood as archaeological sites whose stratigraphic matrices are composed predominantly by shells, are found in fluvial and coastal contexts throughout the planet. They can The distance between the areas with known occurrences of shellmounds in the Amazon is considerable, and one could surmise that they represent distinct processes of occupation. However, their common traits in terms of chronology, settlement patterns, and technological histories of their # Springer International Publishing AG, part of Springer Nature 2018 C. Smith (ed.), Encyclopedia of Global Archaeology, https://doi.org/10.1007/978-3-319-51726-1_3030-1 2 What do Amazonian Shellmounds Tell Us About the Long-Term Indigenous History of South America? What do Amazonian Shellmounds Tell Us About the Long-Term Indigenous History of South America?, Fig. 1 Sites with ceramics dating back to 8000 cal BP and the shellmounds’ occurrence areas estimated by the known sites in Amazon ceramics are becoming increasingly more evident with the development of archaeological research in these sites. Early evidence of ceramics in Amazonian shellmounds is reported in the literature (Evans and Meggers 1960; Simões 1981; Williams 1981, 1997; Meggers 1987, 1997; Roosevelt et al. 1991; Roosevelt 1992, 1997; Hoopes 1994), and it is noteworthy that early ceramics are consistently found in the basal strata of Amazonian shellmounds. As will be seen below, these early ceramics always appear as vestiges of simple industries which started to incorporate technological innovations at some point in the Middle Holocene, including the adoption of specific attributes shared among distant contexts: this occurs in Taperinha, Monte Castelo, in the Xingu sites, the coastal sites in Pará and the Guyanas, as well as in the ancient ceramics of the Colombian and Ecuadorian coasts and even in Panama (Meggers et al. 1965; Reichel-Domaltoff 1985; Hoopes 1994; Oyuela-Caycedo 1996). Roosevelt (1995) was the first to show that there are no preceramic strata in Amazonian shellmounds. These sites are always built on topographically salient areas, which gradually received low amounts of sediments until people started to erect an earth mound, after which constructive layers were added with shells. Ceramic remains are present throughout the whole stratigraphic matrix. The scant available data shows that ceramics are present at the basal strata of the mounds, even if in very simple form or in low density. An eventual preceramic component in this type of site would, therefore, be an exception (but see Williams 1997 for another interpretation). This suggests that contexts for early mound building were correlated with the clay experimentation from the beginning. What do Amazonian Shellmounds Tell Us About the Long-Term Indigenous History of South America? Amazonian shellmounds are located in environments where high diversity of resources is available along permanently or seasonally flooded areas such as floodplains and swamps. Shared characteristics are also seen in changes observed in the uppermost stratigraphic layers, where variations in settlement patterns are associated with differences in ceramic technology, moundbuilding activities, and, consequently, changes in the use of these places throughout the Middle Holocene. There are also recurring occupations from the Late Holocene, sometimes after chronological gaps in the sequences. Finally, indigenous people have seasonally occupied some of these places until the present. Thus, shellmounds can be seen as persistent places where archaeology provides data about continuity and change in the long-term indigenous history (LIH) of the Amazon since at least the Early Holocene. Exploring the Similarities Among Amazonian Shellmounds The area definitions proposed here are based on different criteria: LA and ECA for the diverse geoenvironmental contexts where they are localized, whereas SWA is a more isolated from other regions where fluvial shellmounds occur. A divergence in the malacological building materials used in each region exists, since bivalve mollusk shells are predominant in LA and ECA, but gastropod shells are more prominent in SWA. Distance and material differences group these three areas of occurrence into two isolated sets of sites, those in the west and those in the east. Despite this, the attributes shared by known Amazonian shellmounds permit exploring them as a single set of sites under formation throughout the Middle Holocene, allowing an understanding about the history of Amazonian landscape management from then until the present day (Fig. 2). The Taperinha and Lower Xingu Sites (LA) Taperinha is located adjacent to the floodplain of the Amazon River, downstream from the city of Santarém. The site was initially researched by Canadian geologist Charles Hartt in 1870–1871. 3 Hartt recognized that the place was an archaeological site and suggested its old age due to its location next to an abandoned channel of the Amazon (Hartt 1885). More than 100 years later, Anna Roosevelt continued the research started by Hartt, seeking to test the hypothesis that contexts dating back to the Early Holocene could be excavated in the site (Roosevelt 2009). Taperinha is located on a ledge of dry land adjacent to the river’s floodplain, with ready access to lakes, swamps, and natural channels. The site is 50,000 m2 and over 6 m high, built on a beach (Roosevelt 1999), and, like Hartt had already noticed, ceramic fragments were found all the way to the basal layers. These were later dated to around 8000 cal yBP (all ages mentioned below are calibrated. From here on just BP), providing the earliest dates for ceramics in the Americas (Roosevelt et al. 1991; Roosevelt 1995, 1999, 2009). The site was occupied by relatively large and stable groups who carried out intensive fishing of fish and shellfish in the river, lakes, and streams of the region. The technology of early Taperinha ceramics (Fig. 3) was described based on 383 fragments identified in 48 different strata, composed by shells, charcoal, seeds, bones, and lithics (Roosevelt et al. 1991; Roosevelt 1999). The ancient presence of ceramics in the Taperinha region is also supported by similar data from the Caverna da Pedra Pintada rockshelter (Roosevelt 1995), which would indicate the presence of early ceramics in other kinds of archaeological sites besides sambaquis, although other sites with ceramics dated to the transition from Early to Middle Holocene have yet to be found. Ceramics of the basal strata are tempered with grit, sand, and, in a few cases, shell (Roosevelt et al. 1991; Roosevelt 1999). Vessel shapes are gourd-like and simple open bowls. Vessel shape and the presence of carbon soot in some sherds may imply the preparation and consumption of food (Oliver 2008). Eleven fragments evidenced incised or punctuated decoration (Roosevelt 1992, 1999). Lithic instruments in the same strata include hammerstones, flakes, grinders, and firecracked rocks. Other artifacts, such as scrapers of turtle bone and shell, are also reported (Roosevelt 4 What do Amazonian Shellmounds Tell Us About the Long-Term Indigenous History of South America? What do Amazonian Shellmounds Tell Us About the Long-Term Indigenous History of South America?, Fig. 2 Chronology of Amazonian shellmounds from the 150 available radiocarbon ages (Simões 1981; Williams 1981; Roosevelt et al. 1991; Perota 1992; Perota and Botelho 1992; Hoopes 1994; Roosevelt 1995; Gaspar and Silveira 2000; Bandeira 2008, 2012; Miller 2009; Lombardo et al. 2013; Canto Lopes 2016), calibrated (Hogg et al. 2013) and quantified by occurrence areas What do Amazonian Shellmounds Tell Us About the Long-Term Indigenous History of South America?, Fig. 3 Taperinha ceramics from (a) Hartt (both sides) and (b) Roosevelt excavations. Collection of the Technical Reserve of Archaeology Mario Ferreira Simões, Museu Paraense Emílio Goeldi/MCTIC. (Photos: Elis^angela Oliveira) What do Amazonian Shellmounds Tell Us About the Long-Term Indigenous History of South America? et al. 1991). Other than ceramic chronology and some stratigraphic details, there is little further information about the site. From the profile drawings, it can be seen that the basal strata shows the record of an earth mound upon which the shellmound was built. The switch from earth to shellmounding marks also a change in habitation and funerary patterns (Roosevelt et al. 1991; Roosevelt 1992, 1995, 1999, 2009). Taperinha’s occupation persisted for a long period after early ceramic production. Overlaying early ceramic strata related to the earth matrix, there is a thick accumulation of shells composed of multiple strata dated to the Middle Holocene and, above it, a later stratum of Santarém ceramics associated with anthropic dark earths, or Terras Pretas de Índio, the latter dated to the early second millennium AD (Roosevelt 2009). Hartt had already recorded that the shellmound was covered by a stratum of dark soils with ceramics similar to ones found on other sites in the area. Following earlier hints by Bates (1863) and Agassiz (1868), Hartt (1885) also mentioned the presence of other shellmounds in the LA region. Hilbert (1959, 1968), following Protasio Frikel, published data on Ponta do Jauari, also a shellmound located in the Santarem region, where Zone-Hachured ceramics were found overlapping Mina tradition layers. In the 1970s, excavations took place in Guará I and II sites in the lower Xingu River (Perota and Botelho 1992, 1994), and the available data indicates that around 3000 years BP ceramic materials related to the Mina tradition can be found in these shellmounds. They are broad-based sites, formed by the interdigitation of clay and sandy layers. Mina ceramics persist on the intermediate layers, where bivalve shells abound. The chronology also points to a more recent Mina tradition in the region. In fact, the most recent Mina tradition dates were recorded at the Guará I (Perota and Botelho 1992, 1994) and Uruá sites and are situated around 550 years BP (Silveira et al. 2008). The data reported seems to indicate that the old pattern of reoccupations verified at the base of other sites also occurred there. These sites, therefore, are also palimpsests that document the regional extent and persistence of the Mina tradition in the Lower Amazon. 5 The Mina Tradition Sites (ECA) Shellmounds with Mina tradition ceramics were first cited (Baena 1839; Barbosa Rodrigues 1876) and described (Ferreira Penna 1876, 1877) in the nineteenth-century literature, from the lower Tocantins River to the mouth of the Amazon River. Further work done in Guyana in the 1950s identified similar sites as well with ceramics denominated as Alaka (Evans and Meggers 1960; Williams 1997). In the 1960s and 1970s, dozens of shellmounds and open-air sites with Mina ceramics were mapped and excavated at the Atlantic coast east of the mouth of the Amazon (Simões 1981), where ceramics were classified in five phases (Mina, Uruá, Areão, Castália, and Macapá), mainly differentiated by geographical location and by settlement patterns, rather than technological variability (Simões 1981; Oliveira and Silveira 2016). Today, shellmounds from the Mina tradition are known to be found in a large region that extends from the archipelago of Marajó to almost the Parnaiba River delta, as well in the Guyana (Evans and Meggers 1960; Simões 1981; Silveira and Schann 2005; Bandeira 2016), maybe reaching Venezuela (Meggers, apud Simões 1981: 78) and Caribe (Bel 2012). Such distribution, however, is not continuous and there are no records in Amapá or French Guiana. The sites occur in estuarine areas, along rivers and bays, as well as on islands, and are often surrounded by mangrove forests. Their dimensions vary: Porto da Mina site has 40  30 m at the base and reaches 4 m of height, while the Ponta das Pedras is 145  70 m at the base and stands 9 m tall. Mina ceramics are also starting to be found in other types of open-air sites, such as Jabuti, located further inland and where they are associated with anthropic dark earths (Silveira et al. 2011), and possibly Eva 2, a San Jacinto-like (Oyuela-Caycedo and Bonzani 2005) site located near the forested border of the coastal savannas of French Guiana (Bel 2012; Pagán-Jiménez et al. 2015). The chronology of the Mina tradition extends from 5570 to 1245 BP, making it one of the longest and persistent cultural complexes (Fig. 2) in the continent (Simões 1981; Bandeira 6 What do Amazonian Shellmounds Tell Us About the Long-Term Indigenous History of South America? 2012). A common technological trait spanning the millennia is the constant presence of shell tempering, albeit with varying proportions. Firing appears to have been low temperature. Vessel surface is smooth and red slipping is frequently applied. Shapes are simple and open, with flat or slightly rounded bases and flat or round rims. Plastic decoration includes incision, coiling, and brushing (Simões 1981; Bandeira 2008). The Mina tradition is recognized due to its antiquity and the huge spread of its technological characteristics on ceramics from Mid- to Late Holocene assemblages, such as the Alaka phase in Guyana and the Guará phase in the lower Xingu River. Taken together with the ceramics from Maranhão shellmounds, this makes it not only very ancient and long-lasting, but also widereaching in Eastern Amazonia, encompassing an area ranging from the Lower Amazon to the coast several hundred kilometers southeast and northwest of the mouth of the Amazon. However, the excessive attention given to the early age of Mina ceramics outshined the long-term history recorded in these contexts, which have only recently been approached by regional ceramic variability analyses (Bandeira 2009, 2016) and in the variation of settlement patterns (Silveira et al. 2011). Bandeira (2008) presents the best description of a Mina tradition shellmound. The author divides its history in different cultural periods, stressing that ceramics of this kind are found in all of them. In the Bacanga shellmound, east of the mouth of the Amazon, Mina tradition may be found even before the beginning of shellmound construction, between 6600 and 5800 BP. This event is revealed by combustion structures composed by blocks of laterite disposed in a circular fashion, associated with charcoal, food remains, and ceramics that share few technological and morphological attributes with the Mina tradition, as they are thin and well fired and don’t feature Mina’s characteristic shell-tempered paste. The analysis of the vestiges collected in these basal layers points to a scenario in which the diet was not very dependent on aquatic resources and in which the presence of ceramic material is less dense than in relation to overlying strata (Bandeira 2009). After that period, strata dated between 4800 and 1830 BP show significant technological changes marked by the start of the construction of the shellmound and contexts pointing to a wider utilization of resources, based on fishing and shellfish gathering, as well as significative increase in the quantity of ceramics, lithics, and bone remains. Vessel form becomes predominantly globular, with mouth diameters around 30 cm and some surface treatment, including smooth, brushed, simple, and parallel incised lines. Occupational strata are composed by lenses of shells of different species, associated with soils of varying colors. Besides ceramics, remains of terrestrial and aquatic vertebrate bones, charcoal, laterite blocks, lithics, and ornaments are also found. Mina tradition shellmounds also feature human burials with individuals in flexed position in right lateral decubitus, frequently accompanied by shells and ornaments. Just as in the preshellmound period, combustion structures and evidence of postholes are found (Simões 1981; Bandeira 2012). Around 5000 years ago, various shellmounds were already occupied throughout the ECA: Bacanga (Maranhão), Porto da Mina, Ponta das Pedras and Uruá (Pará), and Barambina (Guyana) (Simões 1981; Williams 1981; Hoopes 1994; Bandeira 2012). Between 5000 and 2500 BP, after this initial phase of coastal occupation, different Mina tradition sites are formed both on the coast and further inland, while the ancient shellmounds of initial sequence were still being occupied (Bandeira 2009). Mina phase ceramics are also associated with open-air non-mounded sites located further inland, such as Jabuti, who is found on high ground on an island at the west bank of the Caeté River, 36 km from the coast. Jabuti has anthropic dark earths with deposits ranging from 40 to 100 cm without signs of accumulation of shells, fishbones, or soil as in other Mina sites. There is a single date of 2900 BP for the early formation of this site (Silveira et al. 2011). Above Mina phase occupations, ceramic fragments tempered with different materials are found on the surface of various shellmounds. These are What do Amazonian Shellmounds Tell Us About the Long-Term Indigenous History of South America? associated with different occupations, including colonial Afro-Brazilian sites, but unfortunately not much is known about them (Simões 1981:14) Likewise, in Maranhão, two cultural assemblages were chronologically situated after the end of the Mina tradition at Bacanga site, the first associated with anthropic dark earths contexts and the most recent related to ceramics of the Tupinambá subtradition (Bandeira 2008, 2012) (Fig. 4). Anthropic dark earth contexts associated with occupations above shellmound strata are common in all shellmounds studied in Maranhão. Such upper strata are 30–50 cm thick showing abundant ceramics, lithics, fish remains, human burials, and no shells and dating from 900 BP at Bacanga and Paço do Lumiar to 760 BP at Panaquatira (Bandeira 2012). These later ceramics are totally distinct from earlier Mina phase vessels, being similar to other Amazonian complexes such as the incised rim and Tupinambá traditions (Bandeira 2008, 2012). Such occupations persisted until the 1700s, after initial European settlement, and is associated with the Tupinambá Indians described historically (Bandeira 2016). Shellmounds from Southwestern Amazonia (SWA) The Monte Castelo shellmound is located on the floodplain of the Branco River (Fig. 5), next to the border of Bolivia and Brazil, and is currently 120  100 m long and 6.5 m tall. It had been previously excavated in the 1980s, and such previous research revealed that the shellmound is a monument built and occupied since the Early Holocene (Miller 2009, 2013). Research at the site was restarted in July 2013 (Pugliese 2018). Excavations at MC have shown the presence of ceramics from the deepest strata to the surface of the site. The first samples of this technology were found in association with faunal and charcoal remains in occupational strata lacking shells as building material. There, an impressive volume of fragments of ceramic blocks and slabs were found, perhaps comparable to the clay balls found in Terminal Classic Maya sites in Central America or in the end of the Archaic period in the United States (Ford and 7 Webb 1956; Gibson 2001). However, a few samples show walls or other attributes characteristic of pottery. The predominant temper is composed of quartz grains, which may have been added to the paste or might originate from a selection of a locally available sandy clay matrix for the production of these artifacts. Only a few diagnostic sherds were detected. Evidence of carbon soot on these pieces is rare and their use is still under investigation. The strata that cover these initial occupations contain a larger investment in pottery making (Fig. 6). Likewise, the context in which these later ceramic materials are found differs notably from earlier strata. What before was composed by a matrix similar to an earthwork, in which structured features are not common, turns into a context in which intense domestic activities were recorded in firepits and other remains of food production and consumption. These horizontally overlaid strata, each about 10 cm thick, record a series of occupational contexts covering the initial occupation of the site. Stratum L is composed of a “floor” cut by negative features such as stake and postholes which seem to mirror the malacological contexts of strata K and M; the excavation of a large hearth reached through several of these strata, and the material which were found suggests that a combustion area used for processing food existed on top of the shellmound for a long period. The use of cauixi (freshwater sponge spicules) tempering on clay found in these intermediary layers of the oldest strata (>5200 BP), together with a change in the structure of ceramic sherds, shows that firing was better controlled, reaching higher temperatures. Besides the inclusion of cauixi and quartz, some shell temper is also found. After ceramics had been present at the site for a period that may reach 3000 years, the Bacabal phase (Miller 2009, 2013) shows several technological innovations at Monte Castelo shellmound. These ceramics, which are deposited over the last strata associated with the massive accumulation of gastropod shells (Stratum F), are found in high density and in association with lithics and different kinds of botanical remains. Strata A–D, where Bacabal materials are found, also contain a significant amount of zooarchaeological remains, with a 8 What do Amazonian Shellmounds Tell Us About the Long-Term Indigenous History of South America? What do Amazonian Shellmounds Tell Us About the Long-Term Indigenous History of South America?, Fig. 4 Ceramics from coastal shellmounds in the state of Maranhão (Bandeira 2012): (a) Tupinambá, (b) associated with TPI, (c) Mina tradition, (d) pre-shellmound culture prominence not only of whole and fragmented gastropod shells but also turtle, cervid, rodent, reptile, and fish bones and some bivalve mollusk shells as well. These evidences not only suggest the consumption of these animals but also the use of skeleton parts for the preparation of utensils such as ornaments, points, needles, and hooks. There are also recurring burials marked with contours or agglomerations of shells and contain associated funerary paraphernalia. Analyses of the ceramic material have revealed that these vessels were produced with a paste What do Amazonian Shellmounds Tell Us About the Long-Term Indigenous History of South America? What do Amazonian Shellmounds Tell Us About the Long-Term Indigenous History of South America?, Fig. 5 Aerial view of Monte Castelo shellmound (photo 9 by Carlos Zimpel), in front of the Branco River and the Guaporé wetlands What do Amazonian Shellmounds Tell Us About the Long-Term Indigenous History of South America?, Fig. 6 First pottery (>5200 BP) in Monte Castelo (depth): (a) Layer J (461 cm), (b) Layer K (473 cm), (c) Layer N (507 cm), (d) Layer T (610 cm). (Photos Francisco Pugliese) tempered primordially with spicules, mixed with fine sands. Some fragments in the older strata also present shells as temper. The homogeneous structure and hardness of the ceramics shows that high temperatures were achieved at firing. Pots were manufactured by coiling with the addition of modeled apliqués. Surfaces were smoothed with several different techniques that can be identified by the traces of soft (such as bamboo and gourd) or hard (such as pebbles, seeds, and shells) 10 What do Amazonian Shellmounds Tell Us About the Long-Term Indigenous History of South America? utensils. Polished samples are also found, but the most recurring types of surface treatment are slips and red slip. Excision and incision were the most recurring types of plastic decoration used. Painted motifs are not to be found. On the other hand, plastic decorations are plentiful, with the most frequent type featuring motifs formed by zigzagging excisions, creating triangular areas filled by thin parallel incisions. This same motif is also found in decorations which are based on incise lines, in which the created areas are filled with dotted lines. Large, parallel, and straight excisions abound, as do different types of incision and dotted lines, as well as two kinds of brushed decoration (by corncob or corn ear). This surface treatment confirms the presence of maize (Zea mays) in this context dating about 4000 BP. Applied handles are commonly found, from the simplest varieties to the most stylized, including zoomorphic and anthropomorphic pieces. The pot shapes are simple, with plane to concave bases and rims with simple contours. Mouth diameter and the thickness of the walls indicate that the pots had a wide range of dimensions, from small serving pots to large cooking pots (Fig. 7). This is an archaeological culture with dates reaching back more than four millennia, sharing elements with the first known ceramic complexes, notably in relation to settlement patterns that relate to the exploration of aquatic resources described in the Mina phase, the lower Xingu sites, and the Taperinha site. The zoned-hachured pattern of the decorative elements has been used by Miller (2009, 2013) to link these ceramics to processes of diffusion starting with Valdívia phase sites from coastal Ecuador, whose early dates go back to ca. 6400 BP (Zeidler 2003). In Bolivia, shellmounds are a part of the earthwork landscapes of the Llanos de Mojos (Denevan 1964, 1966; Erickson 2006; Prümers 2012; Lombardo et al. 2013), but little specific information has so far been published. Recently, three sites located south of the flooded savannahs, near the municipality of Trinidad in the Beni River, were excavated. These sites are similar to Monte Castelo in terms of their implementation and stratigraphic matrix. However, they have much smaller dimensions, currently reaching only less than 2 m of height. Initial mound building seems to have started 10,000 years ago, as small mounds were built at the edge of the savannahs bordering the Mamoré River. Just as in Monte Castelo, since the Middle Holocene, these sites witnessed a greater investment in construction and accumulation of shells as well as other technological novelties, such as ceramics (Fig. 8). Ceramics were found there with dates starting ca. 6000 BP, bearing similarities to the intermediary layers in Monte Castelo. Occupations relating to the constructive period of the site extend to ca. 4000 BP, which begins the chronological hiatus which extends until 400 BP. These recent reoccupations also coincide with the reappearance of ceramics and faunistic remains, which were associated with the well-known Late Holocene Earthmover societies (Lombardo et al. 2013). There are also reports of sambaquis in the northwestern Llanos, near the Rogoaguado lagoon (Echevarría 2008). Considered beside the (little) information available about those sites, paleoecological data (Brugger et al. 2016) indicates that they may also corroborate the pattern of association between ancient pottery and malacological material accumulation that is beginning to be recognized in the region’s early settlements. Shellmounds Landscape Formation The monumental character of shellmounds is still unexplored in Amazonian archaeology, differently from what happens with this type of site in the southern and southeastern coasts of Brazil (Fish et al. 2013). The smooth topography of the areas where these sites were built in the Amazon highlights the constructed sites – which can surpass 6 m of height and have dozens of thousands of square meters in area – as imposing features of the local landscape. In the floodable savannahs of SWA, during some seasons, the shellmounds are the only features rising above the water in a radius of kilometers. The significance and persistence of these places must be considered under this perspective. What do Amazonian Shellmounds Tell Us About the Long-Term Indigenous History of South America? 11 What do Amazonian Shellmounds Tell Us About the Long-Term Indigenous History of South America?, Fig. 7 Bacabal phase ceramics from the Monte Castelo shellmound: (a) zoomorphic and anthropomorphic representations; (b) excise and incise motifs; (c) brushed with corncobs and ears. (Photos Carlos Zimpel) What do Amazonian Shellmounds Tell Us About the Long-Term Indigenous History of South America?, Fig. 8 Pottery from Isla del Tesoro, Llanos de Mojos, Bolivia (Lombardo et al. 2013): (a) Incised ceramic sherd, layer dates (6235  62 BP); (b) ceramics from Late Holocene earthworkers In the three areas of occurrence of known sambaquis in the Amazon, the same general structure in the stratigraphic composition (Fig. 9) – in simpler or more complex versions – seems to be visible from the available data, indicating parallels in their histories. The first settlements featured earth mounds dating to the Early Holocene that reach a maximum height of around 2 m, such as in 12 What do Amazonian Shellmounds Tell Us About the Long-Term Indigenous History of South America? What do Amazonian Shellmounds Tell Us About the Long-Term Indigenous History of South America?, Fig. 9 Synthetic stratigraphy from Amazonian shellmound profiles with available chronological information from calibrated radiocarbon dates (BP). I, Monte Castelo (Beta408413, Beta408414); II, Isla del Tesoro (Lombardo et al. 2013: Poz-34228, Poz-34229, Poz34230, Poz22902, Poz34232, Poz36136, Poz34301, Poz36135); III, Taperinha (Roosevelt 1995); IV, Porto da Mina (Canto Lopes 2016: Beta439361, Beta439362, Beta439363, Beta439364); V, Panaquatira (Bandeira 2012). * Estimated layer ages after Miller (2009, 2013), Roosevelt et al. (1991), and Roosevelt (1995) the case of Monte Castelo. Subsequently, throughout the Middle Holocene, the sites undergo changes in their pattern of occupation, marked by the use of massive accumulations of shells as a constructive material, forming thick layers that intercalate occupational structures and mounds, where the variability of vestiges is much more expressive. It can be considered that from this period on the sambaquis acquire a truly monumental character, standing out in the landscapes with proportions like no other sites in the basin. During this period, an expressive increase in the species undergoing management occurs. This was archaeologically recorded mainly by the presence of new botanical remains on the sites, often carbonized and associated with combustion contexts. These changes are followed by a long period of continuity in this type of constructive activity, and the populations which carried it out also produced ceramics with more sophisticated technologies than those used by their predecessors. These contexts also present a wider variability of types of lithic and bone instruments. In fact, between 5000 and 4000 BP, the sambaquis are practically the only sites in the Amazon and in much of the continent where the production of ceramics is so systematically and expressively confirmed. Taken together, these three facts (the construction of monuments, the appearance of new plants, and the development of more sophisticated technologies) seem to indicate that the changes observed in Amazonian shellmounds along the Middle Holocene are related to the transition from low-density landscape management to the beginning of a new indigenous history, which would culminate in the high-density landscape management which was later dispersed What do Amazonian Shellmounds Tell Us About the Long-Term Indigenous History of South America? throughout the basin (Neves and Petersen 2006). When searching for historical meaning in the continuities and variations in settlement patterns recorded in structural changes in the shellmounds layers, we can also take into account a series of occurrences taking place further south in the continent, such as the construction and occupation of earthworks with malacological materials in the Paraguay River (Schmitz et al. 1998; Eremites De Oliveira 2003; Migliacio 2006; Schmitz et al. 2009). The construction of small shellmounds in the Ribeira do Iguape river valley since the Pleistocene transition is also noteworthy. Starting in the Middle Holocene, burials related to the known burial patterns of coastal shellmounds start appearing in these sites, followed by reoccupations of the region during the Late Holocene (Figuti et al. 2013). The explosive increase in dates for coastal shellmounds related to the period between 4000 and 2000 BP, 4ky after the first occurrences of this type of site (Gaspar et al. 2008), coincides with the construction of shell strata in Amazon sites, also founded in the Early Holocene and occupied until the present or the recent past. Other than the well-known Tupinambá reoccupations of the coastal sites, which may have been extinguished by the advancement of European colonization, dates in superior layers of shellmounds surpassing the first millennium of this era are common. If we expand the limits of this approach, one could observe congruences between the contexts of emergence and development of shellmounds in the Amazon and in Colombia (Reichel-Domaltoff 1985), Venezuela (Cruxent and Rouse 1961), Panamá (Hoopes 1995), Southeastern United States (Sassaman 1993, 2004a, b), and in the Lower Mississippi Valley (Russo 1996; Sassaman 2004c), or even in various parts of the Pacific coast, where these sites occur from Chile to the United States. This is an enormous history contained in a long period of construction of shellmounds and associated landscapes in the continent. 13 Why Long-Term Indigenous History? L’intérêt de ces enquêtes pour l’enquête, c’est, au plus, d’accumuler des renseignements; encore ne seront-ils pas tous valables ipso facto pour des travaux futurs. Méfions-nous de l’art pour l’art. Braudel (1958) The construction of a long-term history of the indigenous populations of the South American lowlands has been the object of archaeological research throughout the Amazon. Historical ecology approaches in particular have changed our understanding of the variability in the relation between mankind and the environment, adding to the understanding of how this took place in the long-term occupation of the continent. A lot of this research has revolved around investigating the periods in which the most intense changes in the landscape are perceived archaeologically, proposing interpretations for the emergence and maintenance of cultural traits (Heckenberger 2005; Ballé and Erickson 2006; Thompson and Waggoner 2013). The long-term indigenous history proposed here consists in the analysis of similarities and differences, continuities and changes in order to interpret the way in which the dialectic between the structures and events took place in the past, seeking to contribute to the understanding of the historical contexts in which they were produced and have been signified. To this end, the evaluation of whether an event is relevant in relation to long duration structures partially depends on the analysis of similarities and differences in the artifactual variability and on the archaeologist’s ability to provide a plausible theory based on their internal perceptions, motivations, and cultural standards (Hodder 1987). Archaeologically speaking, we can state that much of what can be observed in present-day cultures is the result of long-term history, which is represented in the present by the continuation of traits that appeared in remote times. In the Amazon, the persistence of certain cultural traits in artifactual assemblages and settlement patterns may represent the continuation of well-adapted lifestyles which were developed by the pioneer 14 What do Amazonian Shellmounds Tell Us About the Long-Term Indigenous History of South America? populations that colonized the region between the Late Pleistocene and the Middle Holocene (Neves and Petersen 2006) – not to be confused with a stagnation in time and place. In this sense, archaeological investigation of persistent places does not reject contemporary indigenous history. Rather, these studies should integrate themselves with it so that the places where long-term indigenous occupation can be archaeologically accessed may be researched ethically (Meskell 2005) and so that their results reach expected representativeness in the problematic political and social contexts in which they are carried out (e.g., Pugliese and Valle 2015). This stems from the recognition of the immeasurable worth of these sites for the construction of LIH and that the rights of indigenous people of these places must prevail over the territorial rights imposed by national states. In relation to this, it is relevant to recall the potential archaeology has to change paradigms in contemporary communities. Archaeology must address the construction of scientific knowledge about the past, which aids our understanding of the present and which may substantiate the construction of a future in which human occupations not only contribute to the maintenance of biodiversity but also sustain a fairer and more diverse society. There is growing consensus among researchers that archaeology must take a central role in the investigations of the relationship between humans and the environment, building insights into wider issues, related, for instance, with the management of tropical forests (Ballé and Erickson 2006; Iriarte et al. 2016; Watling et al. 2015, 2017). Furthermore, the absence of socially useful approaches to archaeology may lead to the disengagement of talented researchers, and maybe the very survival of the discipline is conditioned by its intellectual engagement with great challenges faced by present-day communities (Van der Leeuw and Redman 2002). In this sense, exploring the connections between the ethnographic present and the archaeological past in the search for a new ontology of the archaeological present (Hicks 2016) of the ancient occupations of the Americas allows knowledge to be used so that remaining meaningful places are respected and territories are kept in existence. In fact, many of the most important archaeological sites in the Amazon are located in areas occupied and reoccupied throughout the millennia, which now house contemporary occupations, be they villages, towns, or even large cities. Unfortunately, it is still common to find the descendants of ancient inhabitants residing in the periphery of urban agglomerations, in situations of social risk, used as cheap labor to sustain local economies. In all such places, archaeology has the potential to create a foundation for the insubordination of traditional populations to the destructive drive of nation-states, if it can overcome its traditional colonial role to be appropriated as an instrument in these peoples’ struggle for sovereignty (McGuire 2004). The Indigenous History of Middle Guaporé River Basin Shared between Brazil and Bolivia, the basins of the Guaporé and Mamoré rivers are known as one of the regions with the highest linguistic variability in the planet. Over 50 indigenous languages have been found here, belonging to 7 families (Arawak, Chapacura, Jabuti, Nambikwara, Pano, Tacana, and Tupi), as well as 11 isolated languages (Aikanã, Canichana, Cayubaba, Iranxe, Itonama, Kanoê, Kwazá, Leko, Mosetén/ Chimané, Movina, Yurakaré), sharing various lexical elements among them. For historical linguistics, this complex patchwork of oftentimes genetically distant languages in the present is probably a result of a number of waves of settlement in the region, whose descendants have been living together in the area for millennia (Crevels and van der Voort 2008). The first pieces of written information referring to the occupation of the Guaporé were produced in the eighteenth century, in a context of dispute of this region between the Spanish and Portuguese empires. In the colonial period, the Guaporé was the fluvial connection between Villa Bella, in Mato Grosso, and Belém, in Pará, a route which gained importance at that time through commerce and capture of Indians. Portuguese and Spanish both claimed the river, trying to control the dense What do Amazonian Shellmounds Tell Us About the Long-Term Indigenous History of South America? indigenous populations that inhabited the region concentrated in religious missions (Lucidio 2013). Next to the Branco River, missions and old colonial settlements were installed, and the information available about these places help us identify some of the indigenous occupations of the region in the eighteenth century (Fonseca 1874, 1881). The first mission in the region, Santa Rosa, was founded by Spanish Jesuits in 1743 in an area on the right margin of the Guaporé where the Aricoroni, speakers of Chapacura, lived. In 1744, over 4000 members of the Moré indians, also Chapacura speakers, were subjugated in San Miguel, located on the river of the same name. San Simón was founded in 1746, in the Branco River, and around 700 people lived there between Morés and Mekéns, a part of the Tupari branch of the Tupi language family. The Mekéns are also mentioned as part of the population of the Portuguese settlements of Ilha Comprida (1742), located in the mouth of the River Mequéns, and Casa Redonda (1752), in the mouth of the Corumbiara River (cf. Fonseca 1749; Amado and Anzai 2006). The decline in mining, the discovery of more efficient commercial routes, and the end of the territorial dispute between Portugal and Spain resulted in a diminished imperial interest in the Guaporé region. The people who inhabited its margins, who had been enslaved, killed, or concentrated in missions, could finally go back to living in relative freedom, and many of them headed up the river toward the head of the tributaries on the right margin, since following those on the left could mean encountering Jesuits in Bolivia (Lucidio 2013). The Tupari linguistic family was rediscovered by Brazilian civil servants and ethnologists in the twentieth century (Snethlage 1937; Caspar 1953, 1957). It has four languages: (1) Makurap, centered around the head of the Branco River; (2) Ayuru, around the Colorado River; (3) Mekéns, around the Mequéns River; (4) and Tupari, in the heads of Machado River tributaries (Moore and Galucio 1994). When found in the twentieth century, speakers of Tupari shared a tragic history of contact with capitalism due to the rubber boom 15 that spread throughout the entire Amazon, ravaging entire populations. The first rubber plantations in the region were installed near the Branco River circa 1910, at a place known as Laranjal, where a Bolivian owner concentrated 600 individuals for rubber work (Snethlage 1937). Their exact identity is unclear, but in 1924, Aluízio Ferreira, who would become the governor of Rondônia in later years, took refuge in this area and described the presence of Makuraps there. He attests the same for the Paulo Saldanha plantation, near the head of the Branco River, founded by the Guaporé Rubber Co., which soon after would install the São Luiz plantation downstream. These plantations were responsible for incorporating the Makurap, Ayuru, Jabuti, Arikapú, and Aruá into rubber exploitation work, as well as poaia (Psychotria ipecacuanha) and Brazil nut (Bertholletia excelsa) gathering. In 1927, a plantation employee accompanied by a retinue of Makurap visited the Tupari malocas (T.N.: indigenous Amazonian long house) in the headwaters of the Branco River, convincing these groups to work at Paulo Saldanha (Caspar 1953). The Tupari had heard of this new non-indigenous presence in the area due to reports by neighboring Makurap, who had acquired tools that the Tupari greatly valued after the contact. Soon, the plantations become a local hub, consequently spreading disease (Tupari 2013). Caspar (1953) estimates that in the late 1920s, at least 3000 people inhabited the headwaters of the Branco River, due to a 1948 report by an elder relating the existence of at least 30 malocas in the area before contact with the rubber trade. In 1934, Emil Heinrich Snethlage was in the Branco, and before reaching the Tupari malocas, he visits the São Luis, reporting a complete absorption of indigenous labor in the plantation as well as cases of physical punishment suffered by the Tupari. Upon reaching the malocas, he only finds three of them, therefore estimating a population of 250 inhabitants (Snethlage 1937). In 1948, Caspar lived among the Tupari where he notes the existence of temporary relations with the rubber trade when men moved temporarily to the São Luís plantation to work in exchange for 16 What do Amazonian Shellmounds Tell Us About the Long-Term Indigenous History of South America? manufactured products. At this time, they lived in two malocas, around 200 under the leadership of Waitó and Kuarumé, Caspar’s interlocutors. According to a report by Konkwat Tupari, a daughter of Waitó Tupari, the Tupari people were forced by the plantation owner to leave their malocas and move indefinitely to the São Luiz in 1953. He also moved other families to the Laranjal plantation, preventing them from returning to their homes after working in the plantation (Fonseca 2008). Amnin Tupari describes Laranjal as a small town, formed by families, both Tupari and non-indigenous, involved in the production of foodstuffs to supply the great huts of the rubber plantation (Tupari 2013). This moment of intense contact with the São Luis is crucial to the history of the Tupari. In 1954, a measles epidemic would irradiate from here, leaving the Tupari on the brink of extinction. In 1955, Franz Caspar visits the Branco River again and meets a group of refugees from this group, amounting to no more than 66 people living in a single maloca, probably in Laranjal (Caspar 1957). Up until 1980, the plantations are sold between owners, many of whom kept Indians working in conditions tantamount to slavery. That year, the Branco River Indigenous Station is created in the old headquarters of the São Luis. Ethnologist Apoena Meireles works on freeing many people from this regime of slavery (Meireles 1983). Finally, in 1983, the Rio Branco Indigenous Reservation is demarcated. However, a few settlements were left outside its limits, including three of them, inhabited mostly by Tupari, inside the Guaporé Biological Reserve, demarcated the year before. The Contemporary Indigenous History of Monte Castelo Shellmound The migratory movement of the Tupari from northern Mato Grosso to the Rio Branco basin and the reoccupation of ancestral areas in the Guaporé River are known in the literature since the publication of Franz Caspar’s ethnography in 1953. According to the Tupari, their ancestors reoccupied archaeological sites when they arrived in that region, a common practice when human settlements move. This is due to privileged characteristics of these places in their environment, such as the topographical advantages of wetlands “islands” or the fertile and fish-rich areas adjacent to fluvial terraces. However, over the last two decades of the twentieth century, the populations that traditionally occupied the middle Guaporé underwent a moment of their history in which new displacements were needed due to the creation of various conservation units in the region. The demarcation of indigenous lands and establishment of federal and state reservations restricted their mobility and precluded their access from wide areas. In the specific case of the Rio Branco Indigenous Land (RBIL), the demarcation process left out central components in the history of occupation of that indigenous territory out of its limits. Like the Laranjal “island,” the Palhal village, situated downstream of the RBIL, a very old settlement recognized for the worth of its fishing and agricultural lands, was left outside of the RBIL and inside the Guaporé Biological Reserve (Rebio Guaporé). The same fate befell the Monte Castelo shellmound, which was often used as a seasonal settlement, mainly for hunting and gathering activities in the wetlands during the dry season and for fishing and gathering during the wet season, and a strategic place in the route to Versailles, an Indigenous village in the Bolivian margin of Guaporé river, where the Tupari trade boats and forest resources. During the dry season, the site converts into a privileged spot for big game hunting, especially the cervids which populate the fields and the tapirs which inhabit the Branco’s swamps at that time of the year. In addition, the site is part of a settlement system which encompasses literally all areas above the waterline during the wet season (locally called “islands”), places used as seasonal logistical camps, but also as cemeteries. The current resistance to expropriation of the lower Branco River settlements, explicit in the continuing occupation of sites outside the RBIL, creates a case in which archaeological heritage can act on people’s present, supporting a relationship of belonging not to be overwhelmed by Western territorial rights. Going beyond this case, this What do Amazonian Shellmounds Tell Us About the Long-Term Indigenous History of South America? notion of archaeological heritage geared toward what these sites represent to the people who make use of it may be the most combative to the destruction caused by the advancement of national societies on traditionally occupied territories in the Amazon, if approached as an effective form of protection of places that hold significance to the present occupants of this area. Understood in its present subjectivity, this idea of archaeological heritage can contribute to the construction of a LIH that can be a tool for the transformation of the social relations which took hold with the advancement of capitalist society, helping to reclaim places with meaningful roles in the formation of past and present indigenous territories. In this sense, the contemporary occupation of the Monte Castelo site constitutes an emblematic case for the construction of the long-term history of the indigenous populations of the Amazon based on shellmound archaeology. Right when we first took up research at the site, in 2013, we noticed the presence of recent remains, indicating that although the site was located inside a federal reserve, prohibiting the development of activities by local populations, it still received regular visits by hunters. Sometime later, when our presence was intensified with new excavations in 2014, signs which intentionally warned us about the ownership of this place started cropping up. Name of persons and villages started to be carved on trees, and there was an increase in the area of activity of the current occupation, previously centered on small clearings in the northern and southeastern summits of the site. During the 2016 excavations, besides new dates carved on trees, we found true signatures, with name and ethnic affiliation. These signs seemed to be strong acts of demarcation, a reaction to the invasion of a meaningful place (Fig. 10). However, our attempts to enter into direct contact with the Indians, heading toward their villages by land and water, had come to no fruition. Nowadays, after the shellmound was isolated by the demarcation of the reserve and their traditional occupations were restricted to ILs and maroon areas, access between the site and the upper Branco River is practically impossible during most of the year. The drastic reduction in 17 circulation of boats allowed hydrophilic vegetation to take over the area between the shellmound and the nearest village, Palhal, and the herds of buffalo which multiplied with the extinction of nearby farms have made the overland route truly deadly. Despite this, the inhabitants of the Rio Branco IL, who never truly abandoned their settlements outside the IL, continue to use the shellmound, but mainly in periods of extreme flood, when it is possible to overtake the vegetation coverage, which becomes a little less dense during this period, allowing fluvial paths to be made. Paradoxically, the indigenous population has been growing over the past few years due to the protection granted by the demarcation, but productive areas inside the RBIL are not plentiful. The exhaustion of resources in the most explored parts of the IL have forced an intensification of use of external areas and a growing movement to reclaim places inside the ReBio Guaporé which were traditionally occupied. Furthermore, isolated populations from the Massaco IL, located on the left margin of the Branco River, have been approaching the southern villages of the Rio Branco IL, and they were even seen by inhabitants in some recent cases, increasing the pressure on them and making exploration of the surroundings of some villages very dangerous. Contact with the inhabitants of the RBIL was finally established in October 2016 when we sailed downstream from nearby villages into the IL and met the people who communicated with us in the shellmound’s trees. At that time, we were introduced to the territorial claims of the various communities we visited, and collaborative projects were proposed to us, mainly aiming to reclaim Monte Castelo and others sites in lower Branco River for their land. The archaeological project in the shellmound and other sites, all places of ancestral indigenous occupations, is producing data that must be used to repatriate these places to modern indigenous people. In relation to this, even if linguistics and the relatively recent character of the current occupation of the villages of the Rio Branco IL may allow for questioning the phylogenetic relation between its inhabitants and the people who produced the archaeological 18 What do Amazonian Shellmounds Tell Us About the Long-Term Indigenous History of South America? What do Amazonian Shellmounds Tell Us About the Long-Term Indigenous History of South America?, Fig. 10 Site’s current occupation signs (b–d) and activity areas (a, e). (Photos by Francisco Pugliese) packets of the shellmound, the return to the region of middle Guaporé River is described by the local elders as a return to a land that had already been occupied by ancestors, and that was the reason why they chose the Branco River in their last migration. Currently, these populations consistently occupy this place in a similar manner to how their predecessor occupied it (even producing analogous contexts of activity). The indigenous history of the region, associated with the meanings held by that place, makes the shellmound landscape a fundamental component of their territoriality. It is important to remember that the search for correlations between the history of occupation of the archaeological sites and current indigenous occupations must consider the multiethnic and multilinguistic pattern that became predominant in various parts of the Amazon from, at least, the tenth century on. If on the one hand this pattern What do Amazonian Shellmounds Tell Us About the Long-Term Indigenous History of South America? complicates the establishment of strict correlations between the past of an ethnicity and the material culture of the sites, because the degree of resolution which can be obtained from the archaeological record only allows for direct correspondences to be observed in specific contexts (Neves 2011), on the other, the case of Monte Castelo demonstrates that for the archaeology of persistent places, this search for strict correlations is not the way to understand its territorial characters. What we propose here, based on the case of the indigenous occupation of the Guaporé wetlands, is that shellmounds are fundamental for the long-term history of occupation of the region where they are implemented, and, in this way, they are a part of the formation of all indigenous territories who have embraced them, independently of the identity which might be conferred to the archaeological record (Meskell 2001). The Long-Term Indigenous History of Shellmounds in the Amazon For millennia, monumental sites all over the world have shaped landscapes which reflect histories of colonization, diffusion, and migration which connect communities in faraway regions and through countless generations. Long continuities are frequent in these places, and some characteristics of these occupations persist, transcending the hiatuses which commonly exist in these sites’ chronologies, denoting the permanence of structural traits in its history (Sassaman 2005). In the Amazon, the advance of shellmound research has demonstrated covariations between chronology (Fig. 2), ceramic technology, and settlement patterns which are related to the history of occupation of a very wide region, which encompasses three known areas of occurrence. In these places, characteristics which can be considered structural in the human occupations which produced the sites present shared elements, and related events can be accessed through the archaeological record (e.g., changes between constructive and occupational layers), pointing to a common history. The structure of this history appears in the sequence of periods of occupation 19 of the sites, which was shaped by recurrences in relation to the emergence and complexification of ceramic technology, to the morphology of stratigraphic layers, and to the distribution of remains in the sites in the Middle Holocene, as well as by the reoccupations of these places during the Late Holocene. Some elements shared by the pottery, such as the use of sand and shell tempers in older technologies, seem to indicate that the relative synchronic appearance of ceramic assemblages in several sites was derived from a process of cultural diffusion; nevertheless, the change in the artifacts from modeled to coiled technologies marks a scenario of intense cultural change starting in the Middle Holocene. The contexts where these changes are recorded, with high variability in plant and animal use remains, may be related to the development of a productive model based on agroecological diversity, a pattern dispersed throughout the Amazon in very ancient times and that can still be found today (Neves 2011). In the search of a LIH of the Amazon based on shellmound archaeology, the Monte Castelo site constitutes a privileged chronological record for the recognition of persistence and innovations, notably in relation to the presence of dates distributed from the end of the Early Holocene to the thirteenth century (Miller 2009, 2013). The site is inserted in the southwestern Amazonian regional context, in which the exceptional language dispersion patterns of the past may be associated with patterns found in archaeological sites (Neves 2011). The recognition of these patterns may contribute to construct a new perspective on the archaeology of Amazonian shellmounds and about the appropriation of these places by present-day communities. It is important to highlight that in other parts of the globe, shellmounds are archaeologically identified as aboriginal ancestral sites (Australia, Ulm 2013), monuments from ancient indigenous complex societies (Southeastern North America, Sassaman 2004c), or even as historical references for the construction of modern national identities (Japan, Mizoguchi 2004). In California, the indigenous history of shellmounds of the San Francisco bay area has been investigated by archaeology 20 What do Amazonian Shellmounds Tell Us About the Long-Term Indigenous History of South America? since the very beginning of the twentieth century (Uhle 1907; Nelson 1909). Today, they are considered sacred places with a history charged with meanings in mythology, and the indigenous movements are using archaeology to support the reclaim of these sites as part of their ancestral territories. On the other hand, in the Amazon, shellmounds were never approached by archaeology in a similar vein. Paradoxically, if on the one hand the region hosts one of the largest extensions of officially recognized indigenous territories in the world, on the other, the debate about the antiquity of its chronology has hindered attempts to approach ancient sites as persistent and meaningful places for current communities. Together with sites along the margins of rapids, big waterfalls, and the meeting of the main tributaries of the basin, Amazonian shellmounds are part of landscapes which have been built and reoccupied for millennia, and they must be explored in relation to the true spatial and temporal reach of their archaeology. This is of special relevance for this chapter’s objectives, since the three areas of shellmound occurrence in the Amazon are located in old indigenous territories, now severely changed by recent colonization processes in the Amazon. In the EAC, extractivist mining of malacological lime deposits has partially or completely destroyed many shellmounds, and the advance of urban perimeters has intensified the modification of the landscapes to which these sites belonged, to the point that today many of them have completely disappeared. In the SWA, true massacres of indigenous peoples occurred in the second-half of the twentieth century, and the expansion of the agricultural frontier still is, without a doubt, the current biggest problem for the conservation of the sites. If the creation of various conservation units in the middle Guaporé River basin insulated important areas from this process, they also restricted local communities from accessing various features of their territories, leading to an unnecessary and harmful opposition between conservation and traditional occupation of these places. In the LA, the Taperinha site has been partially destroyed by shell mining, and many sites recorded in the nineteenth century may never be found again for the same reason. However, the most extreme deterioration of indigenous territories in the region can be found is the lower Xingu River. The construction of the Belo Monte dam and other associated developments have set loose a true cataclysm in the region. The local shellmounds are located only a few kilometers downriver from the Xingu’s “Big Turn,” where the dam’s axis was dramatically erected. In some of these areas, initiatives for the protection and recovery of these places have been put into place by indigenous people and other traditional communities, similarly to how the inhabitants of the Rio Branco Indigenous Land currently relate to Monte Castelo. The results of archaeological research should be used to support the recovery of meaningful components of the deep indigenous history of the continent. Final Remarks: Shellmounds and Ceramics in Amazon Recently obtained dates from Monte Castelo shellmound confirm that ceramic remains found on its basal layers are related to occupations dating before 5200 BP, an assertion which is also supported by Miller’s (2009) dates for these layers, reaching 8000 BP. This implies that the ceramic assemblages found in these layers have the same age of the earlier ceramics of the Americas, resulting in a new interpretation about these ancient contexts. However, differently from what has been observed in sites like San Jacinto 1 (Oyuela-Caycedo and Bonzani 2005), for instance, ancient ceramics from Amazonian shellmounds present evident stratigraphic correlation with contexts of food processing, being artifacts used for the ordinary preparation and consumption of food. The analysis of residues and other types of microremains will assess this hypothesis, but early Amazonian ceramics are also distinct from the typically highly decorated ancient ceramics found elsewhere in South America. In Amazonian shellmounds, ceramics characterized by intense decoration normally appear in more recent contexts (Betancourt 2013; Zimpel and Pugliese 2016), from the Middle Holocene What do Amazonian Shellmounds Tell Us About the Long-Term Indigenous History of South America? on, and, in some cases, in layers covering older occupations, which present a simpler ceramic technology coupled, like in Monte Castelo, with a high density of remains. Despite the lack of a correlation between the emergence of ceramics and the advent of agriculture in the lowlands of South America (Neves 2016), data about the first vestiges of this technology in the continent reveal an association with food processing and consumption. Roosevelt (1992) reports proposed a culinary use in the oldest ceramics in Taperinha, but she relates its use to a diet specialized in fish and mollusks, where there is no use of plants such as corn (1999). What Monte Castelo has demonstrated is that, moving beyond the debate about whether societies specialized in fishing, hunting, and gathering adopted agriculture or not, dietary patterns among early ceramists already leaned toward generalism. As has been observed in a series of Early-to-Middle Holocene sites in various parts of South America (Dillehay 2008), diet diversification was also the choice of most Amazonian settlers (Roosevelt et al. 1996), even with the precocious presence of various plants which would become important for the emergence of agriculture elsewhere in the continent. Such generalist strategies are also characteristic of early ceramic technologies associated with foodprocessing remains. The chronological correlation (Figs. 2 and 9) between innovations observed in the ceramic assemblages of the Middle Holocene suggests that their emergence is related to structural changes in LIH (sensu Sassaman 2004b, c, 2005), which seem to be a good alternative hypothesis to explain artifactual variability without relating it exclusively to processes connected to the emergence and diffusion of early ceramic technologies. This is because that assumption is invalidated by the chronology verified in Amazonian sites where the most ancient ceramics were actually found, like Taperinha and Monte Castelo. A deeper understanding of historical contexts of the tropical lowlands around 8000 BP is fundamental to appropriately pose better questions about the advent of ceramic technologies and to test the origin hypotheses. Hypotheses relating to 21 the emergence of ceramic technology should focus on the same chronological framework, be they based on diffusionist or multiple origin hypotheses and novelties in artifactual variability in Middle Holocene assemblages addressed by more specific regional investigations, such as those involving ethnogenetic processes. Acknowledgments We would like to thank all the people that helped in the development of the archaeological research in the middle Guaporé River basin, especially the amazing team of the Laboratory for Tropical Archaeology from the Museum of Archaeology and Ethnology of São Paulo University and the people from government agencies (Funai, ICMBio, Idaron) who gave us crucial assistance to carry out the fieldworks. The Laboratory for Geochronology of Brasilia University hosted the analysis of early ceramic materials from Monte Castelo and we have a special thanks for the support of Dr. Roberto Ventura Santos. The pictures from Dr. Elisangela Oliveira of Taperinha ceramics brought new colors to the discussion, so we want to thank her and Museu Paraense Emilio Goeldi’s crew for giving us the rights to use the images. Diogo Lima Saraiva from Brasilia University was responsible for the translation from Portuguese to English. PhD fellowships from Capes (FAPJ) and CNPQ (CAZN) and grants from NGS and CNPQ (EDN) funded this research. Communities from the municipality of Costa Marques to the Rio Branco IL have kindly hosted us in their lands and helped us in so many ways, and this chapter is dedicated to them. References Agassiz, L. 1868. A journey in Brazil. Cambridge, UK: University Press: Welch, Bigelow, & Co., 540 p. Amado, J., and L.C. Anzai. 2006. Anais de Vila Bela 1734–1789. Cuiabá: EdUFMT. Baena, A.L.M. 1839. Ensaio Corográfico sobre a Província do Pará, Edições do Senado Federal. Vol. 30. Belém: Typ. de Santos & Menor. Ballé, W., and C. Erickson. 2006. Time and complexity in historical ecology: Studies in the neotropical lowlands. New York: Columbia University Press. Bandeira, A.M. 2008. Ocupações humanas pré-históricas no litoral maranhense: um estudo arqueológico sobre o sambaqui do Bacanga na Ilha de São Luís-Maranhão. Unpublished Master’s thesis, Museu de Arqueologia e Etnologia, Universidade de São Paulo, São Paulo. Bandeira, A.M. 2009. Arqueologia no sambaqui do Bacanga, São Luís, Maranhão: Reflexões sobre a ocorrência de cer^amica em sambaquis do litoral equatorial amazônico. Revista Amazônica 1 (2): 484–511. Bandeira, A.M. 2012. Ocupações humanas pré-coloniais na Ilha de São Luís-MA: inserção dos sítios 22 What do Amazonian Shellmounds Tell Us About the Long-Term Indigenous History of South America? arqueológicos na paisagem, cronologia e cultura material cer^amica. Unpublished PhD thesis, Museu de Arqueologia e Etnologia, Universidade de São Paulo, São Paulo. Bandeira, A.M. 2016. A cer^amica Mina do Maranhão. In Cer^ a micas arqueológicas da Amazônia: rumo a uma nova síntese, ed. C. Barreto, H.P. Lima, and C.J. Betancourt, 147–157. Belém: IPHAN: Ministério da Cultura. Barbosa Rodrigues, J. 1876. Antiguidades do Amazonas. Sernambys. Ensaios de Sciencia, Rio de Janeiro 4: 23–34. Bates, H.W. 1863. The naturalist on the river Amazon. London: Murray. van den Bel, M. 2012. Archaeological investigations at the Saint-Louis on the lower Maroni river: The Early Ceramic series. Archaeology and Anthropology, Journal of the Walter Roth Museum of Anthropology 17 (2): 1–52. Betancourt, C. 2013. Diversidad cultural en los Llanos de Mojos. In Arqueología Amazónica. Las civilizaciones ocultas del bosque tropical, ed. F. Valdez. Quito: Instituto Francés de Estudios Andinos; Institut de Recherche pour le Développement; Abya-Yala. Braudel, F. 1958. Histoire et Sciences sociales: La longue durée. Annales Économies, Sociétés, Civilisations 13 (4): 725–753. Brugger, S.O., E. Gobet, J.F. van Leeuwen, M.-P. Ledru, D. Colombaroli, W. van der Knaap, U. Lombardo, K. Escobar-Torrez, W. Finsinger, and L. Rodrigues. 2016. Long-term man-environment interactions in the Bolivian Amazon: 8000 years of vegetation dynamics. Quaternary Science Reviews 132: 114–128. https://doi. org/10.1016/j.quascirev.2015.11.001. Canto Lopes, P.R. 2016. Caracterização do Modo de Vida dos Sambaquieiros que Ocuparam o Litoral Paraense: Quatipuru, Pará, Brasil. Unpublished PhD thesis, Museu Nacional, Universidade Federal do Rio de Janeiro, Rio de Janeiro. Caspar, F. 1953. Tupari: entre os índios, nas florestas brasileiras. São Paulo: Melhoramentos. Caspar, F. 1957. A aculturação da tribo Tupari. Revista de Antropologia 5 (2): 145–171. Universidade de São Paulo. Crevels, M., and H. Van Der Voort. 2008. The Guaporé-Mamoré region as a linguistic area. In From linguistic areas to areal linguistics, ed. P. Muysken, 151–179. Amsterdam/Philadelphia: J. Benjamins. Cruxent, J.M., and I. Rouse. 1961. Arqueología cronológica de Venezuela. Washington, DC: Unión Panamericana. Denevan, W.M. 1964. Pre-Spanish earthworks in the Llanos de Mojos of north-eastern Bolivia. Revista Geografica 34 (60): 17–25. Denevan, W.M. 1966. The aboriginal cultural geography of the Llanos de Mojos of Bolivia. Berkeley: University of California Press. Dillehay, T.D. 2008. The settlement of the Americas: A new prehistory. New York: Basic Books, 352 p. Echevarría, G.T. 2008. Excavaciones arqueológicas en la cuenca del Lago Rogoaguado, provincia de Yacuna (Beni, Bolivia). Estudios Amazónicos 5 (7): 87–150. Eremites De Oliveira, J. 2003. Da pré-história à história indígena: (re) pensando a arqueologia e os povos canoeiros do Pantanal. Revista de Arqueologia (Sociedade de Arqueologia Brasileira. Impresso), São Paulo 16: 71–86. Erickson, C. 2006. The domesticated landscapes of the Bolivian Amazon. In Time and complexity in historical ecology: Studies in the neotropical lowlands, ed. W. Ballé and C. Erickson, 235–278. New York: Columbia University Press. Evans, C., and B.J. Meggers. 1960. Archeological investigations in British Guiana, Smithsonian Institution Bureau of American Ethnology Bulletin, 177. Washington, DC: U.S. Govt. Print. Off. Ferreira Penna, D.S. 1876. Breve notícia sobre os sambaquis do Pará. Archivos do Museu Nacional do Rio de Janeiro 1: 85–99. Ferreira Penna, D.S. 1877. Apontamentos sobre os Cer^amios do Pará. Archivos do Museu Nacional do Rio de Janeiro 2: 47–67. Figuti, L., C.R. Plens, and P.A.D. Deblasis. 2013. Small sambaquis and big chronologies: Shellmound building and hunter-gatherers in neotropical highlands. Radiocarbon 55: 1215–1221. Fish, P.R., S.K. Fish, P.A.D. Deblasis, and M.D. Gaspar. 2013. Monumental shell mounds as persistent places in southern coastal Brazil. In The archaeology and historical ecology of small scale economies, ed. V.D. Thompson and J.C. Waggoner Jr., 120–140. Gainesville: University Press of Florida. Fonseca, J.G. 1748 [1874]. Primeira Exploração dos Rios Madeira e Guaporé feita por José Gonçalves da Fonseca em 1749 por Ordem do Governo. In Memorias para a historia do extincto estado do Maranhão cujo territorio comprehende hoje as províncias do Maranhão, Piauhy, Grão-Pará e Amazonas, Tomo Segundo, ed. C.M. de Almeida. Rio de Janeiro: Nova Typographia de J. Paulo Hildebrand. da Fonseca, J.S. 1881. Viagem ao Redor do Brasil (1875–1878). Rio de Janeiro: Typographia de Pinheiro & Co. Fonseca, M.G. 2008. Casa de escrever no papeo: a escola Tuparí da terra indígena Rio Branco. Unpublished Master’s dissertation. Universidade Federal de Rondônia, Porto Velho, 103 p. Ford, J. A, and C. H. Webb. 1956. Poverty point, a late archaic site in Louisiana, Anthropological Papers, Vol. 46, Pt. 1. New York: American Museum of Natural History. Gaspar, M.D., and M.I. Silveira. 2000. Os pescadorescoletores-caçadores do litoral Norte brasileiro. In Pré-história da Terra Brasilis, ed. M.C. Tenório, 247–260. Rio de Janeiro: UFRJ. Gaspar, M.D., P.A.D. Deblasis, S.K. Fish, and P.R. Fish. 2008. Sambaqui (shell mound) societies of coastal Brazil. In Handbook of South América What do Amazonian Shellmounds Tell Us About the Long-Term Indigenous History of South America? archaeology, ed. H. Silverman and W.I. Isbel, 319–335. Lisboa: Springer. Gibson, J.L. 2001. The ancient mounds of poverty point: Place of rings. Gainesville: University Press of Florida. Hartt, C.F. 1885. Contribuições para a ethnologia do Valle do Amazonas. Archivos do Museu Nacional 6: 1–174. Heckenberger, M.J. 2005. The ecology of power: Culture, place and personhood in the Southern Amazon, A.D. 1000–2000. New York: Routledge. Hicks, D. 2016. The temporality of the landscape revisited. Norwegian Archaeological Review 49 (1): 5–22. https://doi.org/10.1080/00293652.2016.1151458. Hilbert, P. P. 1959. Achados arqueológicos num sambaqui do baixo Amazonas. Publicação do Instituto de Antropologia e Etnologia do Pará, Belém 10: 1–22. Hilbert, P.P. 1968. Archäologische Untersuchungen am mittleren Amazonas: Beiträge zur Vorgeschichte des südamerikanischen Tieflandes. Berlin: Reimer. Hodder, I. 1987. Archaeology as long-term history. Cambridge, UK: Cambridge University Press. Hogg, A.G., Q. Hua, P.G. Blackwell, M. Niu, C.E. Buck, T.P. Guilderson, T.J. Heaton, J.G. Palmer, P.J. Reimer, R.W. Reimer, C.S.M. Turney, and S.R.H. Zimmerman. 2013. SHCal13 Southern Hemisphere calibration, 0–50,000 years cal BP. Radiocarbon 55 (4): 1889–1903. https://doi.org/10.2458/azu_js_rc.55.16783. Hoopes, J.W. 1994. Ford revisited: A critical review of the chronology and relationships of the earliest ceramic complexes in the New World, 6000–1500 B. C. Journal of World Prehistory 1 (8): 1–49. Hoopes, J.W. 1995. Interaction in hunting and gathering societies as a context for the emergence of pottery in the Central American Isthmus. In The emergence of pottery. Technology and innovation in ancient societies, ed. W.K. Barnett and J. Hoopes, 185–198. Washington, DC: Smithsonian Institution Press. Iriarte, J., P.A.D. Deblasis, J.G. Souza, and R. Corteletti. 2016. Emergent complexity, changing landscapes, and spheres of interaction in southeastern South America during the Middle and Late Holocene. Journal of Archaeological Research 24: 1–63. https://doi.org/ 10.1007/s10814-016-9100-0. Lombardo, U., K. Szabo, J.M. Capriles, J. May, W. Amelung, R. Hutterer, E. Lehndorff, A. Plotzki, and H. Veit. 2013. Early and middle holocene huntergatherer occupations in western Amazonia: The hidden shell middens. PLoS One 8 (8): e72746. https://doi.org/ 10.1371/journal.pone.0072746. Lucídio, J. 2013. A Ocidente do Imenso Brasil: as conquistas dos rios Paraguai e Guaporé (1680–1750). Unpublished PhD dissertation. Universidade Nova de Lisboa, Lisboa. McGuire, R. 2004. Contested pasts: Archaeology and Native Americans. In A companion to social archaeology, ed. Lynn Meskell and Robert W. Preucel, 374–395. Oxford: Blackwell. Meggers, B. 1987. El origen transpacífico de la cerámica Valdivia: una reevaluación. Boletín del Museo Chileno de Arte Precolombino 2: 9–31. 23 Meggers, B. J. 1997. La cerámica temprana en América del Sur: ¿invención independiente o difusión? Revista de Arqueología Americana, México 13: 7–40. Meggers, B.J, C. Evans, and E. Estrada. 1965. Early formative period of coastal Ecuador: The Valdivia and Machalilla phases. In Smithsonian contributions to https://doi.org/10.5479/ anthropology, 1–234. si.00810223.1.1. Meireles, D.M. 1983. Populações Indígenas e a Ocupação Histórica de Rondônia. Unpublished monography. Universidade Federal de Mato Grosso, Cuiabá. Meskell, L. 2001. Archaeologies of identity. In Archaeological theory today, ed. Ian Hodder, 187–213. Cambridge, UK: Polity Press. Meskell, L. 2005. Sites of violence: Terrorism, tourism and heritage in the archaeological present. In Embedding ethics, ed. L.M. Meskell and P. Pels, 123–146. Oxford: Berg. Migliacio, M.C. 2006. O doméstico e o ritual: cotidiano Xaray no Alto Paraguai até o século XVI. Unpublished PhD dissertation. Museu de Arqueologia e Etnologia, Universidade de São Paulo, São Paulo. Miller, E.T. 2009. Pesquisas arqueológicas no Pantanal do Guaporé-RO, Brasil: a sequência seriada da cer^amica da Fase Bacabal. In Arqueologia interpretativa. O método quantitativo para estabelecimento de sequências cer^ a micas: estudos de caso, ed. B. Meggers, 103–117. Porto Nacional: Unitins. Miller, E.T. 2013. Algumas culturas ceramistas, do noroeste do Pantanal do Guaporé à Encosta e Altiplano Sudoeste do Chapadão dos Parecis. Origem, Difusão/ Migração e Adaptação – do Noroeste da América do Sul ao Brasil. Revista Brasileira de Linguística Antropológica 5 (2): 335–386. Mizoguchi, K. 2004. Identity, modernity, and archaeology: The case of Japan. In A companion to social archaeology, ed. Lynn Meskell and Robert W. Preucel, 396–414. Oxford: Blackwell. Moore, D., and A.V. Galucio 1994. Reconstruction of Proto-Tupari consonants and vowels. In Proceedings of the meeting of the society for the study of the indigenous languages of the Americas July 2–4, 1993 and the Hokan-Penutian Workshop July 3, 1993, 119–137. Nelson, N. 1909. Shellmounds of the San Francisco Bay region. University of California Publications in American Archaeology and Ethnology 7 (4): 310–356. Neves, E.G. 2011. Archaeological cultures and past identities in Precolonial Central Amazon. In Ethnicity in ancient Amazonia: Reconstructing past identities from archaeology, linguistics, and ethnohistory, ed. A. Hornborg and J. Hill, 31–56. Boulder: University of Colorado Press. Neves, E.G. 2016. Não Existe Neolítico ao Sul do Equador: as Primeiras Cer^amicas Amazônicas e Sua Falta de Relação com a Agricultura. In Cer^ a micas arqueológicas da Amazônia: rumo a uma nova síntese, ed. C. Barreto, H.P. Lima, and C.J. Betancourt, 32–39. Belém: IPHAN: Ministério da Cultura. 24 What do Amazonian Shellmounds Tell Us About the Long-Term Indigenous History of South America? Neves, E.G., and J.B. Petersen. 2006. Political economy and pre-Columbian landscape transformations in Central Amazonia. In Time and complexity in historical ecology: Studies in the neotropical lowlands, ed. W. Ballé and C. Erickson, 279–310. New York: Columbia University Press. Oliveira, E., and M.I. Silveira. 2016. A Cer^amica Mina no Estado do Pará: Oleiras das Águas Salobras da Amazônia. In Cer^ a micas arqueológicas da Amazônia: rumo a uma nova síntese, ed. M.C. Barreto, H. Lima, and C.J. Betancort, 125–146. IPHAN/MPEG. Oliver, J.R. 2008. The archaeology of agriculture in ancient Amazonia. In The handbook of South American archaeology, ed. H. Silverman and W. Isbell, 185–216. New York: Springer. https://doi.org/ 10.1007/978-0-387-74907-5_12. Oyuela-Caycedo, A. 1996. The study of collector variability in the transition to sedentary food producers in northern Colombia. Journal of World Prehistory 10: 49–93. https://doi.org/10.1007/bf02226071. Oyuela-Caycedo, A., and M. Bonzani. 2005. San Jacinto I a historical ecological approach to an archaic site in Colombia. Tuscaloosa: University of Alabama Press. https://doi.org/10.1017/s0003598x0009428x. Pagán-Jiménez, J.R., R. Rodríguez-Ramos, B.A. Reid, M. van den Bel, and C.L. Hofman. 2015. Early dispersals of maize and other food plants into the Southern Caribbean and Northeastern South America. Quaternary Science Reviews 123: 231–246. https://doi.org/ 10.1016/j.quascirev.2015.07.005. Perota, C. 1992. Adaptação Agrícola no baixo Xingu. In Prehistória Sudamericana: nuevas perspectivas, ed. B.J. Meggers, 221–218. Santiago: Taraxacum. Perota, C., and W.C. Botelho. 1992. Les “Sambaquis” de Guara et des variations climatiques pendant l’Holocène. In Evolution des littoraux de Guyane et de la zone caraı̈be méridionale pendant le quaternaire, ed. M.-T. Prost, 379–395. Paris: ORSTOM. Perota, C., and W.C. Botelho. 1994. Os sambaquis do Guará e as variações climáticas no Holoceno. Revista do Departamento de Geografia 7 (1): 49–59. São Paulo: Universidade de São Paulo. Prümers, H. 2012. El Proyecto Lomas de Casarabe: investigaciones arqueológicas en los Llanos de Moxos, Bolivia. In The past ahead. Language, culture, and identity in the neotropics, Studies in Global Archaeology 18, ed. C. Isendahl, 139–159. Uppsala: Acta Universitatis Upsaliensis. Pugliese Jr., F.A. 2018. A História Indígena Profunda do Sambaqui Monte Castelo um ensaio sobre a longa duração da cer^amica e das paisagens no sudoeste amazônico. Unpublished PhD Thesis, Museu de Arqueologia e Etnologia da Universidade de São Paulo, 389 p. Pugliese, F.A., Jr., and R.B.M. Valle. 2015. A Gestão do Patrimônio Arqueológico em Territórios Indígenas: a resistência Munduruku e a preservação do patrimônio cultural frente ao licenciamento ambiental de empreendimentos em territórios tradicionais. Revista de Arqueologia 28: 30–51. Reichell-Domaltoff, G. 1985. Monsú, un sitio arqueológico. Bogotá: Biblioteca Banco Popular. Roosevelt, A.C. 1992. Arqueologia amazônica. In História dos índios no Brasil, ed. M. Carneiro Da Cunha, 53–86. São Paulo: Companhia das Letras; Secretaria Municipal de Cultura; FAPESP. Roosevelt, A.C. 1995. Early pottery in the Amazon: Twenty years of scholarly obscurity. In The emergence of pottery. Technology and innovation in ancient societies, ed. W.K. Barnett and J. Hoopes, 115–131. Washington, DC: Smithsonian Institution Press. Roosevelt, A. 1997. Amazonian Indians from prehistory to the present. Tucson/London: University of Arizona Press, 441 p. Roosevelt, A.C. 1999. The Maritime, highland, Forest dynamic and the origins of complex culture. In The Cambridge history of the native peoples of the Americas, ed. F. Salomon and S.B. Schwartz, 264–349. Cambridge, UK: Cambridge University Press. Roosevelt, A.C. 2009. A historical memoir of archaeological research in Brazil (1981–2007). Boletim do Museu Paraense Emílio Goeldi Ciências Humanas 4 (1): 155–170. https://doi.org/10.1590/s198181222009000100013. Roosevelt, A.C., R.A. Housley, M.I. Silveira, S. Maranca, and R. Johnson. 1991. Eighth millennium pottery from a prehistoric Shell Midden in the Brazilian Amazon. Science 254 (5038): 1621–1624. Russo, M. 1996. Southeastern Archaic mounds. In Archaeology of the mid-Holocene southeast, ed. K.E. Sassaman and D.G. Anderson, 259–287. Gainesville: University Press of Florida. Sassaman, K.E. 1993. Early pottery in the Southeast: Tradition and innovation in cooking technology. Tuscaloosa: The University of Alabama Press, 305 p. Sassaman, K.E. 2004a. Structure and practice in the Archaic Southeast. In North American archaeology, ed. T.K. Pauketat and D.D. Loren, 79–107. Oxford: Blackwell. Sassaman, K.E. 2004b. Common origins and divergent histories in the early pottery traditions of the American Southeast. In Early pottery: Technology, style, and interaction in the lower Southeast, ed. Rebecca Saunders and Christopher Hays. Tuscaloosa: University of Alabama Press. Sassaman, K.E. 2004c. Complex hunter–gatherers in evolution and history: A North American perspective. Journal of Archaeological Research 12 (3): 227–280. Sassaman, K.E. 2005. Poverty point as structure, event, process. Journal of Archaeological Method and Theory 12: 335–364. Schlanger, S. 1992. Recognizing persistent places in Anasazi settlement systems. In Space, time, and archaeological landscapes, ed. J. Rossignol and L. Wandsnider, 91–112. New York: Plenum Press. What do Amazonian Shellmounds Tell Us About the Long-Term Indigenous History of South America? Schmitz, P.I., J.H. Rogge, A.O. Rosa, and M.V. Beber. 1998. Aterros Indígenas no Pantanal do Mato Grosso do Sul. Pesquisas Antropologia 54: 1–271. Schmitz, P.I., J.H. Rogge, A.O. Rosa, M.V. Beber, and E.A.V. Freitas. 2009. Aterros da Tradição Pantanal nas fazendas Sagrado Coração de Jesus e Bodoquena, Corumbá, MS. Pesquisas Antropologia 8 (1): 321–374. Silveira, M. I., and D. P. Schann. 2005. Onde a Amazônia encontra o mar: estudando os sambaquis do Pará. Revista de Arqueologia, São Paulo 18: 67–79. Silveira, M.I., M.C.L. Rodrigues, E. Oliveira, and L.M. Losier. 2008. Sequência cronológica de ocupação na área do Salobo (Pará). Revista de Arqueologia, Belém 21 (1): 61–84. Silveira, M.I., E. Oliveira, D.C. Kern, M. Costa, and S. Rodrigues. 2011. O sítio Jabuti, em Bragança, Pará, no cenário arqueológico do litoral amazônico. Boletim do Museu Paraense Emílio Goeldi. Ciências Humanas 6 (2): 335–345. Simões, M. F. 1981. Coletores-pescadores ceramistas do litoral do Salgado (Pará). Boletim do Museu Paraense Emílio Goeldi, Nova Série: Antropologia, Belém 78: 1–26. Snethlage, E.H. 1937. Atiko y: meine Erlebnisse bei den Indianern des Guaporé. Berlin: Klinkhardt & Biermann. von den Steinen, K. 1886. Durch Central-Brasilien: Expedition zur Erforschung des Schingú im J. 1884. Leipzig: Brockhaus, 372 p. von den Steinen, K. 1894. Unter den Naturvölkern ZentralBrasiliens. Reiseschilderungen und Ergebnisse der zweiten Schingú-Expedition 1887–1888. Berlin: Geographische Verlagsbuchhandlung von Dietrich Reimer, 648 p. Thompson, V.D., and J.C. Waggoner. 2013. The archaeology and historical ecology of small scale economies. Gainesville: University Press of Florida. Tupari, I. 2013. Puop’orop Toap, Um Estudo Sobre A Educação Indígena Tupari. Unpublished monography. Fundação Universidade Federal de Rondônia, Ji-paraná. 25 Uhle, M. 1907. The Emeryville shellmound. University of California Publications American, Archaeology and Ethnology 7 (1): 107. Ulm, S. 2013. ‘Complexity’ and the Australian continental narrative: Themes in the archaeology of Holocene Australia. Quaternary International 285: 182–192. https://doi.org/10.1016/j.quaint.2012.03.046. Van der Leeuw, S., and C.L. Redman. 2002. Placing archaeology at the center of socio-natural studies. American Antiquity 67 (4): 597–605. Watling, J., J. Iriarte, B.S. Whitney, E. Consuelo, F. Mayle, W. Castro, D. Schaan, and Ted R. Feldpausch. 2015. Differentiation of neotropical ecosystems by modern soil phytolith assemblages and its implications for palaeoenvironmental and archaeological reconstructions II: Southwestern Amazonian forests. Review of Palaeobotany and Palynology. https://doi.org/10.1016/ j.revpalbo.2015.12.002. Watling, J., J. Iriarte, F. Mayle, D. Schaan, L.C.R. Pessenda, N.J.F. Loader, A. Street-Perrott, R.E. Dickau, A. Damasceno, and A. Ranzi. 2017. Impact of preColumbian “geoglyph” builders on Amazonian forests. PNAS. https://doi.org/10.1073/pnas.1614359114. Williams, D. 1981. Excavation of the Barambina shell mound northwest district: An interim report. Archaeology and Anthropology, Georgetown 4(1–2): 13–38. Williams, D. 1997. Early pottery in the Amazon: A correction. American Antiquity 62 (2): 342–352. Zedeño, M.N., and B.J. Bowser. 2009. The archaeology of meaningful places. In The archaeology of meaningful places, ed. M.N. Zedeño and B.J. Bowser, 1–14. Salt Lake City: The University of Utah Press. Zeidler, J.A. 2003. Settlement process and historical contingency in the western Ecuadorian formative. In Archaeology of formative Ecuador, ed. J.S. Raymond and R.L. Burger, 69–123. Washington, DC: Dumbarton Oaks. Zimpel Nt, C.A., and F.A. Pugliese Jr. 2016. A fase bacabal e suas implicações para a Interpretação do Registro Arqueológico no Médio Rio Guaporé, Rondônia. In Cer^ a micas arqueológicas da Amazônia: rumo a uma nova síntese, ed. M.C. Barreto, H. Lima, and C.J. Betancort, 420–434. IPHAN/MPEG.