Brazil nut, the Bertholletia excelsa seed, is one of the most important non-timber forest products in the Amazon Forest and the livelihoods of thousands of traditional Amazonian families depend on its commercialization. B. excelsa has been frequently cited as an indicator of anthropogenic forests and there is strong evidence that past human management has significantly contributed to its present distribution across the Amazon, suggesting that low levels of harvesting may play a positive role in B. excelsa recruitment. Here, we evaluate the effects of Brazil nut harvesting by the Kayapó Indigenous people of southeastern Amazonia on seedling recruitment in 20 B. excelsa groves subjected to different harvesting intensities, and investigated if management by harvesters influences patterns of B. excelsa distribution. The number of years of low-intensity Brazil nut harvesting by the Kayapó over the past two decades was positively related to B. excelsa seedling density in groves. One of the mechanisms behind the higher seedling density in harvested sites seems to be seed dispersal by harvesters along trails. The Kayapó also intentionally plant B. excelsa seeds and seedlings across their territories. Our results show not only that low-intensity Brazil nut harvesting by the Kayapó people does not reduce recruitment of seedlings, but that harvesting and/or associated activities conducted by traditional harvesters may benefit B. excelsa beyond grove borders. Our study supports the hypothesis that B. excelsa dispersal throughout the Amazon was, at least in part, influenced by indigenous groups, and strongly suggests that current human management contributes to the maintenance and formation of B. excelsa groves. We suggest that changes in Brazil nut management practices by traditional people to prevent harvesting impacts may be unnecessary and even counterproductive in many areas, and should be carefully evaluated before implementation.
Citation: Ribeiro MBN, Jerozolimski A, de Robert P, Salles NV, Kayapó B, Pimentel TP, et al. (2014) Anthropogenic Landscape in Southeastern Amazonia: Contemporary Impacts of Low-Intensity Harvesting and Dispersal of Brazil Nuts by the Kayapó Indigenous People. PLoS ONE 9(7): e102187. https://doi.org/10.1371/journal.pone.0102187
Editor: John P. Hart, New York State Museum, United States of America
Received: January 22, 2014; Accepted: June 16, 2014; Published: July 16, 2014
Copyright: © 2014 Ribeiro et al. This is an open-access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
Funding: This work was funded by Conselho Nacional de Desenvolvimento Científico e Tecnológico - CNPq (http://www.cnpq.br) (grant number 141697/2007-5); Conservação Internacional do Brasil - CI-Brasil (http://www.conservation.org.br)(grant number CPFY 08/022); Instituto Internacional de Educação do Brasil - IEB/Gordon and Betty Moore Foundation (http://www.iieb.org.br) (grant number B/2007/02/BDP/04); Institut de Recherche Pour le Développement - IRD (http://www.ird.org); and International Conservation Fund of Canada - ICFC (http://icfcanada.org). MBNR was supported by a CNPq scholarship through the National Institute for Science, Technology and Innovation for Amazonian Biodiversity (INCT-CENBAM) during manuscript preparation. WEM is supported by the Program for Biodiversity Research (PPBio) and the INCT-CENBAM. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.
Competing interests: The authors have declared that no competing interests exist.
There is a growing body of evidence supporting the hypothesis that the Amazon is a mosaic of anthropogenic landscapes, managed and domesticated by indigenous pre-Columbian people in various degrees, rather than a pristine and untouched forest –. Bertholletia excelsa (Lecythidaceae) groves, locally known as castanhais, are frequently cited as anthropogenic forests in the Amazon region , . B. excelsa seeds, Brazil nuts, have been used for subsistence by indigenous people in the Amazon forest for thousands of years , and there is increasing evidence that those human populations have significantly influenced the current distribution of this species across the Amazon . The occurrence of B. excelsa groves in many parts of the Amazon is associated with patches of Amazonian anthropogenic dark earths ,  and recruitment of B. excelsa trees is favored in agricultural fallows and disturbed sites , . Genetic and linguistic data also suggest that the dispersal of the species throughout the Amazon Basin is a recent process , –, which cannot be explained solely by the activity of natural dispersers, and Scoles & Gribel  found that B. excelsa stands were younger in sites with more intensive pre- and post-colonial human occupation.
Brazil nuts have been harvested almost exclusively from the wild and their commercialization represents an important source of revenue for many Amazonian indigenous and riverine communities, and one of the most promising alternatives to predatory economic activities , . Due to its uncontestable economic importance, there has been concern over the sustainability of Brazil nut harvesting –. While some studies have shown that medium and even high levels of harvesting may be sustainable over the long term, a meta-analysis conducted by Peres et al. , which was later criticized by Scoles & Gribel , concluded that long-term intensive Brazil nut harvesting has reduced B. excelsa recruitment throughout the Amazon. Despite the possible demographic impacts of high-intensity Brazil nut harvesting, the historical evidence , ,  suggest that past human activities favored B. excelsa and, therefore, that low levels of harvesting may play a positive role in the recruitment of this species.
Much of the current Brazil nut harvesting in the Amazon is still carried out by traditional people, who often conduct non-intensive seed collection , . Many of those traditional harvesters, such as indigenous peoples, have interacted with B. excelsa for centuries  and may be also intentionally and non-intentionally managing B. excelsa stands , . The commercialization of Brazil nuts provides one of the major sources of income for the Kayapó Indians, a Jê speaking indigenous group that maintains periodic semi-nomadic practices  and is known to manage many natural resources , . Kayapó informants say that ancient Kayapó villages and camping sites are associated with Brazil nut, babaçu (Orbignya spp.), açaí (Euterpe spp.) and bacaba (Oenocarpus spp.) plantations, and that the planting of those species is an ancient Kayapó tradition .
The Kayapó group includes approximately 10000 people that inhabit a Brazil nut rich territory on both sides of the Xingu River. The Kayapó territory, together with other indigenous lands and conservation reserves, forms a continuous forest corridor of 28 million ha within the arc of deforestation in southeastern Amazonia. Inhabited by 25 indigenous groups, including the Kayapó, and hundreds of riverine families, the Xingu basin is in large part considered an anthropogenic landscape, inhabited and managed by indigenous people over at least the last 1200–1500 years . The Kayapó ethnic group originated in the savanna biome, in an area between the Tocantins and Araguaia rivers, located about 200 km east of the region in which it is found today . The migration of the Kayapó to the west occurred at least 170 to 250 years ago , although they probably interacted with the Amazonian forests earlier than that. Before their arrival, the Xingu region was inhabited by several Tupi speaking indigenous groups  and there is archeological evidence of ancient occupation of the region, including ceramic remains and sites with anthropogenic dark earths . Since the occupation of the Xingu basin, the Kayapó has been the main group responsible for the active protection and management of the area , , . The peaceful contact with most of the Kayapó groups occurred in the 1950s , . During the 1980s and 1990s, some Kayapó chiefs were engaged in illegal and unsustainable activities, mainly gold mining and selective logging . In the last decades, however, many Kayapó villages have invested in sustainable economic activities, and the Brazil nut trade represents one of the main income-generation initiatives. Brazil nut harvesting in recent years by the Kayapó in the three villages we studied has been mostly low intensive, varying from approximately 7% to 43% of the estimated total seed stocks of the harvested groves of the villages .
This study investigates the contemporary effects of Brazil nut harvesting and management by the Kayapó indigenous people on the recruitment and dispersal of B. excelsa. Specifically, we asked if current seed harvesting by the Kayapó of three villages has affected density of seedlings in 20 B. excelsa groves subjected to different harvesting intensities, whether the Kayapó effectively disperse Brazil nuts along trails used to transport seeds, and if they have intentionally or unintentionally planted B. excelsa seeds and seedlings in their villages.
Material and Methods
We conducted our study in the territories of three Kayapó villages – A'Ukre (07°41′43″S, 51°52′53″W), Moikarakô (07°26′11″S, 51°48′57″W) and Kikretum (07°08′17″S, 51°39′26″W), located along the Riozinho and Fresco Rivers, which are second- and first-order tributaries of the Xingu River, respectively. The villages are located in the Kayapó Indigenous Land, a 3.284 million ha reserve located in southern Pará State, southeastern Amazonia, in the transition between the Amazon Forest and the savannas of central Brazil (Fig. 1). The Kayapó people traditionally collect Brazil nuts for subsistence, and the close relationship between the Kayapó and B. excelsa is a recurrent theme in mythical, historical and present times . Kikretum, A'Ukre and Moikarakô villages were founded in 1976, 1979 and 1996, respectively, and most of their families are involved in harvest of Brazil nut for subsistence and trade. A'Ukre produced Brazil nut oil for a British cosmetic company between 1991 and 2003, and has sporadically traded nuts on the local market . Kikretum has constantly traded Brazil nuts on the local market, except during the gold mining and logging period between the late 1980s and 2002. Moikarakô has traded nuts on the local market only in the last few years. Since 2005, all three communities have been engaged in an initiative of Brazil nut certification and fair trading lead by the Associação Floresta Protegida (AFP), a local indigenous non-profit organization which currently represents 22 Kayapó communities.
The triangles indicate the location of the three Kayapó villages and the points, the location of the 20 B. excelsa groves sampled in this study, in Kayapó Indigenous Land, southeastern Amazonia. Identification numbers of groves correspond to those of Table S1 in the supporting information.
Bertholletia excelsa trees are among the tallest and most long-lived trees of the Amazon Forest. They can reach up to 50 m in height, more than 5 m in diameter at breast height  and may live up to 1000 years . The species occurs in unflooded (terra firme) forests of Brazil, Bolivia, Peru, Colombia, Venezuela and the Guianas , , usually in groves of 50 to more than 300 individuals , , although individuals may be randomly distributed in the landscape at relatively low densities in some areas .
In the study site, B. excelsa groves are usually well defined, and vary greatly in area (from 5 to ≈800 ha), density of B. excelsa trees (1 to 5.1 individuals >60 cm DBH ha−1), and abundance of B. excelsa trees (20 to ≈800 individuals grove−1) . Pollination depends on a few species of bees . Mean B. excelsa annual fruit production estimated for the study site in four consecutive years varied from 103 to 269 fruits per tree (mean 184.3±72.2 fruits tree−1 year−1), but some individuals may produce up to 1000 fruits . However, production varies considerably in other Amazon regions (mean 66±98 and 102 fruits tree−1 year−1 in Acre  and Bolivia , respectively). The extremely hard fruits fall during the rainy season, and contain from 8 to 30 seeds , . Only a few species can open B. excelsa fruits, especially agoutis, which have been considered the main non-human B. excelsa seed predators and dispersers in most areas , . Recruitment and growth of young B. excelsa trees is favored in open areas, such as natural forest gaps and disturbed sites , –
Impacts of Brazil nut harvesting on seedling density
We evaluated the impacts of Brazil nut harvest on the regeneration of B. excelsa in 20 groves with different harvesting histories located in the territories of the Kayapó villages of A'Ukre, Moikarakô and Kikretum, including five unharvested groves. In April to June 2009 and April to June 2010, we estimated the density of B. excelsa seedlings and adult trees from line transects established in the groves. One transect was established in each grove from the border to the interior along the longest axis, usually crossing the entire grove. In very small groves, two perpendicular transects were established to attain a minimum length of 850 m, discounting the overlap area where they crossed. Transect lengths varied from 850 to 1400 m, depending on the size of the grove (more information on each grove is provided in the Table S1). In each line transect, we measured all individuals seen and recorded their perpendicular distance from the transect. Densities were estimated with the program DISTANCE 6.0 . When the number of detected individuals in a grove was lower than 30, we corrected the density by a correction factor calculated with DISTANCE using transects of all groves pooled in a single analysis, with truncated distances of 2 m and 30 m on each side of the transect for seedlings and adults, respectively. At those truncation distances, the probability of detection of individuals calculated by DISTANCE was 100% and the mean detection distances were not different between transects (F19,101 = 1.002, P = 0.466 for seedlings; F19,332 = 0.051, P = 0.539 for adults). We considered only seedlings from 30 to 150 cm in height, since mortality rates of seedlings shorter than 30 cm are considerably higher [MBNR, unpublished data]. Adult trees were considered those with more than 60 cm diameter at breast height (DBH), since trees in the study site start to produce fruits at approximately this DBH (; MBNR, unpublished data).
We defined size of B. excelsa groves and location of line transects with the program ArcGIS  from SPOT 5 HRG panchromatic satellite images with resolution of 2.5 m and SPOT 5 HRG multispectral satellite images with resolution of 10 m (details in ). The SPOT images allowed us to visualize harvested groves informed by the Kayapó, as well as to identify isolated and unharvested groves not known to the harvesters. One of the B. excelsa groves sampled was previously studied by Baider , who mapped all its adult individuals. Because this grove was not inside the SPOT images we had, we used the size of the grove given by Baider and defined the location of the transect from her maps. To control the effect of soil fertility on the density of B. excelsa seedlings, we collected a composite soil sample (1–20 cm depth) at the center of each grove. Soil analyses were conducted at the Laboratory of Soils and Plants of the Instituto Nacional de Pesquisas da Amazônia (INPA), according to standard procedures . The sum of the extractable bases Ca, Mg and K (cmolc kg−1) was used as a measure of soil fertility.
We estimated Brazil nut harvest intensity in each B. excelsa grove from information on recent commercial production of nuts by the Kayapó; Smith's S index of salience, which reflects the importance of each B. excelsa grove for villagers; and from the number of years of commercial Brazil nut harvest in each grove during the last 20 years. Twenty years is the approximate time in which seedlings should attain a height of 150 cm based on a study of growth of 139 seedlings over two years in 12 B. excelsa groves in Kayapó Indigenous Land (mean growth rate of seedlings including regrowth after damage caused by falling branches was 6.3 cm year−1±10.5 (SD) [MBNR, unpublished data]. Recent commercial Brazil nut production in tons by the Kayapó was obtained from AFP records of 2008/09 and 2009/10 harvests, which were based on the number of sacks collected by each Kayapó harvester in each grove. Smith's S is an index of psychological salience which weights both rank and frequency of citations of an item in free listings , . During structured interviews with 37, 32 and 29 Kayapó from A'Ukre, Moikarakô and Kikretum, we conducted free listings of the B. excelsa groves in the territories of each village. During interviews, the Kayapó were asked to cite the groves of their respective village. All interviews were conducted between October 2007 and February 2009 in the Kayapó language, with the help of a Kayapó translator. Smith's S index had a high correlation with commercial Brazil nut production (r = 0.82, P<0.001), and was excluded from the analysis to avoid colinearity. However, the high correlation indicates that the short-term data on commercial seed production reflects long-term harvesting. We estimated the number of years of commercial harvest during the last two decades from historical harvesting information obtained from structured interviews with Kayapó collectors, Kayapó organizations, Fundação Nacional do Índio (FUNAI) and Kayapó-elder informants. Summarized data for each B. excelsa grove is available in Table S1.
We used multiple regression to test the effect of Brazil nut commercial production, number of years of commercial harvest of the grove, area of the grove, soil fertility (sum of bases) and density of adult trees on the density of seedlings. Our evaluation was conducted in a single geographic area, with similar climatic conditions, so there was no variation in rainfall and temperature that might confound our conclusions about the effect of other variables.
Dispersal of B. excelsa by the Kayapó
To determine the amount and frequency of B. excelsa seed dispersal by the Kayapó along trails, we followed 60 harvesters from 36 different families in Brazil-nut-collection trips in the three Kayapó villages, over 19 days, in 2009 and 2010, and visually counted all seeds dropped by them between the groves and camp sites. To test if Kayapó seed dispersal enhances the density of B. excelsa seedlings along the trails, we compared the density of seedlings from 30 to 150 cm in height in one transect along each of the three main trails used for transporting Brazil nuts from the closest B. excelsa groves to A'Ukre village, with the density of seedlings in a similar transect parallel to, and 100 m from each trail. We counted seedlings within 2 m on both sides of each transect. Transects were established only in old growth forests, outside B. excelsa groves, and began 200 m from the edge of the grove. Transect length varied from 700 to 1000 m, which was the maximum distance between groves and agricultural sites around the villages or the river. We used paired t-tests to compare density of seedlings in transects near and far from trails.
We tested if seedling growth and mortality rates were different beside trails (≤2 m) and away from trails (>2 m) from a two year demographic study with 138 seedlings from 30 to 150 cm in height (70 near trails and 68 away from trails) inside eight B. excelsa groves in the three villages, four of them located in the territory of A'Ukre village, one in the territory of Moikarakô village and three in the territory of Kikretum village. In March and April 2008, we measured and marked seedlings found within groves with aluminum tags and registered their geographical coordinates with a GPS. All individuals were surveyed between April and June 2010 and re-measured when alive. We used paired t-tests to compare growth and mortality of seedlings near and far from trails inside each grove.
To investigate if the Kayapó plant B. excelsa seeds and seedlings, we conducted interviews and surveyed for B. excelsa individuals in the villages. During structured interviews, we asked Kayapó villagers if they had planted or knew about ancestors who had planted B. excelsa trees. To determine whether B. excelsa individuals are planted in the villages, we conducted one-day surveys for B. excelsa trees in the central area and home gardens of the three Kayapó villages, which are distant at least 2.0 km from the nearest B. excelsa grove. Agricultural sites and secondary forests near the villages were not surveyed.
Approval for interviewing the Kayapó was obtained from the Instituto Nacional de Pesquisas da Amazônia's Committee for Research Ethics (protocol n° 0154/07). Written informed consents were obtained from the Kayapó after meetings in each village to clearly explain in the Kayapó native language the objectives and importance of this study, as well as to clarify the content of the informed consent in order to guarantee that all villagers fully understood their right to choose to participate or not in the interviews and the confidentiality of their identity. Authorization to access the Indigenous Reserve during field work was obtained from National Indian Foundation (FUNAI, protocol n° 0313/07), the Brazilian government agency responsible for the establishment and implementation of most policies related to indigenous peoples. No traditional knowledge was accessed in this research.
Impacts of Harvesting on Brazil Nut Seedling Density
There was a strong positive effect of the number of years of Brazil nut collecting on recruitment of seedlings, despite seed losses due to harvesting (R2 = 0.576, F5,14 = 3.81, P = 0.022). The multiple-regression model relating density of seedlings (DS) to Brazil nut commercial production in tons (PR), years of harvest of the grove (YH), area of the B. excelsa grove (AG), soil fertility (SF) and density of adult trees (DA) (DS = 5.648 - 1.747PR+0.922YH+0.015AG - 0.152FS+0.793DA) accounted for 58% of the variance in the density of seedlings (Table 1). There was little evidence of colinearity (tolerance >0.51 for all variables). Only years of harvest (P = 0.003) and commercial production (P = 0.027) contributed significantly to the model (Fig. 2). However, models containing only years of harvest and commercial production indicated that the effect of years of harvest was much stronger than that of production. A model with only years of harvest (r2 = 0.36) explained 15% less variance than the model with both variables (R2 = 0.51). A model with only production had an r2 of only 0.04. If we remove the B. excelsa grove in which the most seeds have been collected (Piykôny, number 4 in Fig. 1 and Table S1), the negative effect of commercial production on density of seedlings is no longer significant.
The partial regression to the left shows the relationships between density of B. excelsa seedlings (DS) and years of harvest (YH), and the partial regression on the right shows the relationship between density of seedlings and commercial seed production in tons (PR) in B. excelsa groves in the Kayapó Indigenous Land, southeastern Amazonia. The points in each graph represent the 20 B. excelsa groves sampled. The full multiple regression model has R2 = 0.576.
Management of B. excelsa by the Kayapó
Kayapó collectors were seen to drop a total of 34 B. excelsa seeds (1.8 day−1) along trails on 16% of the days they were followed. Moreover, density of B. excelsa seedlings (Fig. 3) within 2 m of the trails (mean = 28.6 seedlings ha−1 ±7.09 SD) was much higher (paired t-test: t2 = 7.33, P = 0.018) than along the parallel transects 100 m from the trails (mean = 1.2 seedlings ha−1, ±2.1 SD). Estimated density of seedlings along trails outside groves was twice the density of seedlings of all B. excelsa groves sampled in this study (13.4 ha−1 ±8.5 SD) and more than four times the density of seedlings in the five unharvested groves (6.4 ha−1±4.5 SD).
Each point represents one of the transects in the territory of A'Ukre village, Kayapó Indigenous Land, southeastern Amazonia. The three points on the left are transects located along the main trails used to transport Brazil nuts from groves to the village and the three points on the right are transects parallel to those trails and 100m from them. Points with the same symbol represent pairs of along-trail and away-from-trail transects.
Mean growth rate, including reduction in size due to cutting or other damage, and mortality of seedlings within 2 m of trails in groves (mean = 9.7 cm±12.7 SD and 3.3%±5.1 SD, respectively) were not significantly different (paired t-test: t7 = 0.447, P = 0.669 for mean growth rate; t7 = 0.586, P = 0.576 for mortality) from those of seedlings located away from trails (11.2 cm±9.2 SD for mean growth rate; 4.6%±5.4 SD for mortality).
Of the 98 villagers interviewed, 20.4% declared that they had intentionally planted at least one B. excelsa seedling in the village or agricultural fields and 41.7% said that their ancestors or older relatives planted B. excelsa trees. We located six, eight and 51 B. excelsa individuals within Moikarakô, Kikretum and A'Ukre villages, respectively. Most of them were seedlings between 30 and 150 cm in height (41.5%), or saplings with DBH <10 cm (30.8%). Three individuals were already producing seeds. According to information obtained from the Kayapó, most individuals found were intentionally planted by the villagers, while others germinated from seeds unintentionally dropped in their home gardens.
Although we cannot predict the effect of a future possible intensive seed collection on B. excelsa recruitment in the study site, the current low-intensity harvesting conducted by the Kayapó did not reduce seedling recruitment in B. excelsa groves. Despite the large quantities of Brazil nuts collected, the Kayapó have only harvested a small proportion (∼7% to 43%) of the total seed production of harvested groves in the recent years , which is lower than the levels of harvesting described as sustainable in other studies , . In fact, it has been suggested that even high levels of fruit or seed harvesting of some trees may allow persistence of the species in the area over the long term .
Our study also showed that the number of years of harvesting conducted by the Kayapó and/or other activities associated with it increased the density of seedlings in B. excelsa groves. This increase is probably a result of several non-exclusive mechanisms. Demographic mechanisms, such as density-dependent mortality , , could be responsible for the increase in seedling density due to a decrease in seed density and, consequently, a reduction in seed and seedling attractiveness to predators or pathogens , . Furthermore, besides Brazil nut collection, the most accessible B. excelsa groves are also used by harvesters to hunt and collect other forest products, such as fruits, honey, medicines and fibers. Even considering that Brazil nut harvesting and those extractive activities have a low impact in forest structure, they may result in small disturbances (e.g. trails) which could increase light levels in the forest understory, favoring regeneration of B. excelsa (42–44). It is also possible that hunting conducted by the Kayapó in the harvested sites could enhance seedling recruitment by reducing density of B. excelsa seed and/or seedling predators, or influence patterns of B. excelsa seed dispersal in these sites –. Density of game species in the vicinity of one of the studied Kayapó villages (A'Ukre), which includes frequently harvested B. excelsa groves, was previously shown to be lower than in remote areas .
Despite the possible importance of other mechanisms, the extremely high density of seedlings along trails used to transport Brazil nuts to the village indicates that the efficiency of Kayapó harvesters as B. excelsa seed dispersers is one of the reasons for the high density of seedlings within most harvested groves. B. excelsa groves have a complex and dynamic net of trails between harvested trees and, considering that the harvesting season lasts at least two months per year, and that several Kayapó families collect and transport Brazil nuts simultaneously, the annual number of seeds dispersed along trails must be substantial. Kayapó collectors carry B. excelsa seeds long distances (kms) within and outside groves. Although scatter-hoarding rodents, which can move seeds for considerable distances (up to 50 m) from the mother trees , , have been cited as the main agents responsible for the current spatial distribution of B. excelsa , human harvesters can disperse seeds over distances much greater than can be expected for these small rodents with home ranges of a few hectares , . Therefore, seeds dispersed by harvesters could benefit from lower predation due to the higher distance from the mother tree , , , and from greater chance of germination if they are dispersed to suitable sites , . Although seedlings along trails may be more prone to be cut by harvesters, growth and mortality rates were similar near trails and far from trails. Seedlings from seeds dispersed along trails may grow faster due to higher light levels , which possibly compensates for the negative effects of being cut.
The unintentional dispersal of seeds by the Kayapó along trails between groves, camps and villages indicates that seed dispersal along trails is a potential mechanism to expand or initiate new groves. Moreover, B. excelsa is intentionally and unintentionally planted by the Kayapó in villages and agricultural sites, and we found several B. excelsa individuals in the villages. Considering that those villages are less than 35 years old, that the Kayapó have inhabited the Xingu basin region and harvested Brazil nuts for at least 170 years, and that many Kayapó villages have changed their locations several times , it is likely that B. excelsa planting by the Kayapó has significantly influenced the B. excelsa spatial distribution within their territory. This finding corroborates information obtained by Posey  that ancient Kayapó villages are associated with Brazil nut groves and that the Kayapó traditionally plant Brazil nut trees. Brazil nut harvesting by the Kayapó may represent an unusual case of harvesting associated with management, but since there is apparently no specific or complex traditional knowledge used by the Kayapó to manage B. excelsa, similar responses to Brazil nut harvesting by indigenous and non-indigenous people is probably also occurring in several other Amazonian regions , , .
Together, our results support the hypothesis that B. excelsa dispersal throughout the Amazon was, at least in part, influenced by pre-Colombian indigenous groups , , corroborating the theory that the Amazon forest is a mosaic of landscapes modified by humans in the past , . Moreover, current human management probably still contributes to the maintenance and formation of B. excelsa groves in the Xingu basin, and possibly in other Amazonian sites. This may have significant implications for the management of B. excelsa and raises an important question: If B. excelsa groves are indeed anthropogenic formations, is some level of human interference needed for their maintenance in the long term or is the activity of scatter-hoarding rodents by itself ,  enough to maintain the dynamics of groves at a landscape scale? This might have implications for the long-term conservation of B. excelsa, especially inside Amazonian reserves that do not allow the presence of human settlements.
There has been much discussion about how to conciliate the needs of local people and conservation priorities in the Amazon region , , , and one of the priority actions to advance in this question and move towards a more inclusive conservation policy is to assess the ecological impacts of the use of natural resources by traditional populations . Here we have shown that harvesting of Brazil nuts by the Kayapó people of southeastern Amazonia, as practiced today, is an example of beneficial coexistence between people and a natural resource. We still do not know the effects of a possible future increase in Brazil nut market demand, and consequently in harvesting intensity, on the recruitment of B. excelsa inside groves of the Kayapó territories. However, so far, Brazil nut harvesting not only has provided livelihoods for the Kayapó, contributed to the maintenance of their culture and reduced the attractiveness of predatory sources of income, such as logging and mining, but it has also positively affected B. excelsa recruitment and dispersal. We recommend that any suggestions for changes to B. excelsa management practices by traditional people aimed to reduce harvesting impacts (e.g. ), should be carefully evaluated and tailored to different social and environmental contexts. Management prescriptions based on broad generalizations may disrupt the delicate equilibrium between use and conservation of this valuable resource by local Amazonian communities.
We thank all the Kayapó who kindly received us in their villages and B. excelsa groves; Okore Kayapó (in memorium), T. Hoorda and K. Kato for field assistance; M. Barroso for help in identifying B. excelsa groves in SPOT images; and W. Spironello for help with DISTANCE. We are grateful to B. Zimmerman, C. Clement, and T. Haugaasen for valuable comments on the manuscript. SPOT satellite images were provided by SEAS-Guiane/IRD. All logistical support was provided by Associação Floresta Protegida.
Conceived and designed the experiments: MBNR AJ PR NVS BK WEM. Performed the experiments: MBNR NVS BK AJ. Analyzed the data: MBNR WEM. Contributed reagents/materials/analysis tools: TPP. Wrote the paper: MBNR AJ PR WEM. Discussed the results: MBNR AJ PR NVS BK TPP WEM. Reviewed the text of the manuscript: MBNR AJ PR TPP WEM.
- 1. Balée W (1989) The culture of Amazonian forests. In: Posey DA, Balée W, editors. Resource Management in Amazonia: Indigenous and Folk Strategies. New York: The New York Botanical Garden. pp. 1–21.
- 2. Denevan WM (2011) The “Pristine Myth” revisited. Geogr Rev 101: 576–591.
- 3. Clement CR, Junqueira AB (2010) Between a pristine myth and an impoverished future. Biotropica 42: 534–536.
- 4. Heckenberger MJ, Kuiruro A, Kuikuro UT, Russel JC, Schmidt M, et al. (2003) Amazonia 1492: Pristine forest or cultural parkland? Science 301: 1110–1114.
- 5. Levis C, Souza PFd, Schietti J, Emilio T, Pinto JLPdV, et al. (2012) Historical Human Footprint on Modern Tree Species Composition in the Purus-Madeira Interfluve, Central Amazonia. PLoS ONE 7(11): e48559
- 6. Shepard GH, Ramirez H (2011) “Made in Brazil”: human dispersal of the Brazil nut (Bertholletia excelsa, Lecythidaceae) in ancient Amazonia. Econ Bot 65: 44–65.
- 7. Roosevelt AC, da Costa ML, Machado CL, Michab M, Mercier N, et al. (1996) Paleoindian cave dwellers in the Amazon: The peopling of the Americas. Science 272: 373–384.
- 8. Clement CR, McCann JM, Smith NJH (2003) Agrobiodiversity in Amazonia and its relationships with dark earths. In: Lehmann J, Kern D, Glase B, Woods W, editors. Amazonian Dark Earths: Origin, Properties, Management. Dordrecht: Kluwer Academic Publishers. pp. 159–178.
- 9. Cotta JN, Kainer KA, Wadt LHO, Staudhammer CL (2008) Shifting cultivation effects on Brazil nut (Bertholletia excelsa) regeneration. For Ecol Manage 256: 28–35.
- 10. Paiva MP, Guedes MC, Funi C (2011) Brazil nut conservation through shifting cultivation. For Ecol Manage 261: 508–514.
- 11. Buckley DP, O'Malley DM, Apsit V, Prance GT, Bawa KS (1988) Genetics of Brazil nut (Bertholletia excelsa Humb. & Bonpl.: Lecythidaceae). Theor Appl Genet 76: 923–928.
- 12. Kanashiro M, Harris SA, Simons A (1997) RAPD diversity in Brazil nut (Bertholletia excelsa Humb. & Bonpl.: Lecythidaceae). Silvae Genet 46: 219–223.
- 13. Gribel R, Lemes MR, Bernardes LG, Pinto AE, Shepard Jr GH (2007) Phylogeography of Brazil-nut tree (Bertholletia excelsa, Lecythidaceae): Evidence of human influence on the species distribution. Morelia: Association for Tropical Biology and Conservation Annual Meeting, Morelia. pp 281.
- 14. Scoles R, Gribel R (2011) Population structure of Brazil nut (Bertholletia excelsa, Lecythidaceae) stands in two areas with different occupation histories in the Brazilian Amazon. Hum Ecol 39: 455–464.
- 15. Clay JW (1997) Brazil nuts: The use of a keystone species for conservation and development. In: Freese CH, editor. Harvesting Wild Species – Implications for Biodiversity and Conservation. Baltimore: John Hopkins University Press. pp. 246–282.
- 16. Shanley P, Pierce AR, Laird SA, Guillen SA (2002) Tapping the Green Market: Certification and Management of Non-timber Forest Products. London: Earthscan. 456 p.
- 17. Zuidema PA, Boot RGA (2002) Demography of the Brazil nut tree (Bertholletia excelsa) in the Bolivian Amazon: impact of seed extraction on recruitment and population dynamics. J Trop Ecol 18: 1–31.
- 18. Peres CA, Baider C, Zuidema PA, Wadt LHO, Kainer KA, et al. (2003) Demographic threats to the sustainability of Brazil nut exploitation. Science 302: 2112–2114.
- 19. Wadt LHO, Kainer KA, Staudhammer CL, Serrano ROP (2008) Sustainable forest use in Brazilian extractive reserves: natural regeneration of Brazil nut in exploited populations. Biol Conserv 141: 332–346.
- 20. Scoles R, Gribel R (2012) The regeneration of Brazil nut trees in relation to nut harvest intensity in the Trombetas River valley of Northern Amazonia, Brazil. For Ecol Manage 265: 71–81.
- 21. Ribeiro MBN, Jerozolimski A, de Robert P, Magnusson WE (2014) Brazil nut stock and harvesting at different spatial scales in southeastern Amazonia. For Ecol Manage 319: 67–74.
- 22. Robert P (2009) Del pi y-kô al bosque certificado: los varios caminos de la castaña. Anuario Americanista Europeo 6–7: 563–581.
- 23. Posey DA (1985) Indigenous management of tropical forest ecosystems: The case of the Kayapo Indians of the Brazilian Amazon. Agroforestry Systems 3: 139–158.
- 24. Pereira HS (1994) Manejo agroflorestal da castanheira (Bertholletia excelsa H.B.K.) na região do Lago do Tefé (AM). Rev. U.A. Série Ciências Agrárias 3: 11–32.
- 25. Turner T (1992) Os Mebengokre Kayapó: história e mudança social. In: Cunha MC, editor. História dos Índios no Brasil. São Paulo: Companhia das Letras/Secretaria Municipal de Cultura/FAPESP. pp. 311–338.
- 26. Posey DA (2004) Kayapó Ethnoecology and Culture. Plenderleith K, editor. New York: Routledge.
- 27. Schwartzman S, Villas Boas A, Ono KY, Fonseca MG, Doblas J, et al. (2013) The natural and social history of the indigenous lands and protected areas corridor of the Xingu River basin. Phil Trans R Soc B 368: 20120164.
- 28. Viveiros de Castro E (1986) Arawete: Os Deuses Canibais. Rio de Janeiro: Jorge Zahar/Anpocs. 744p.
- 29. Arnaud E (1987) A expansão dos índios Kayapó-Gorotíre e a ocupação nacional (Região Sul do Pará). In: Arnauld E. O Índio e a expansão Nacional. Belém: CEJUP. pp. 427–495.
- 30. Hemming J (2003) The Kayapó. In: Hemming J. Die if you must: Brazilian Indians in the twentieth century. London: Pan Macmillan. pp. 110–132.
- 31. Schwartzman S, Zimmerman B (2005) Conservation alliances with indigenous peoples of the Amazon. Conserv Biol 19: 721–727.
- 32. Morsello C (2002). Market Integration and Sustainability in Amazonian Indigenous Livelihoods: The Case of the Kayapó. PhD thesis. Norwich: University of East Anglia.
- 33. Mori SA, Prance GT (1990) Taxonomy, ecology, and economic botany of the Brazil nut (Bertholletia excelsa Humb. and Bonpl.: Lecythidaceae). Advan Econ Bot 8: 130–150.
- 34. Vieira S, Trumbore S, Camargo PB, Selhorst D, Chambers JQ, Higuchi N (2005) Slow growth rates of Amazonian trees: consequences for carbon cycling. Proc Natl Acad Sci USA 102: 18502–18507.
- 35. Ortiz EG (2002) Brazil Nut (Bertholletia excelsa). In: Shanley P, Pierce AR, Laird SA, Guillen SA, editors. Tapping the Green Market: Certification and Management of Non-timber Forest Products. London: Earthscan. pp. 61–74.
- 36. Peres CA, Baider C (1997) Seed dispersal, spatial distribution and population structure of Brazil nut trees (Bertholletia excelsa) in Southeastern Amazonia. J Trop Ecol 13: 595–616.
- 37. Wadt LHO, Kainer KA, Gomes-Silva DAP (2005) Population structure and nut yield of a Bertholletia excelsa stand in southwestern Amazonia. For Ecol Manage 211: 371–384.
- 38. Maués MM (2002) Reproductive phenology and pollination of the Brazil nut tree (Bertholletia excelsa Humb. & Bonpl. - Lecythidaceae) in Eastern Amazonia. In: Kevan P, Fonseca VLI, editors. Pollinating Bees - the Conservation Link between Agriculture and Nature. Brasília: MMA. pp. 245–254.
- 39. Baider C (2000) Demografia e Ecologia de Dispersão de Frutos de Bertholletia excelsa Humb. & Bonpl. (Lecythidaceae) em Castanhais Silvestres da Amazônia Oriental. PhD thesis. São Paulo: Universidade de São Paulo.
- 40. Kainer KA, Wadt LHO, Staudhammer CL (2007) Explaining variation in Brazil nut fruit production. For Ecol Manage 250: 244–255.
- 41. Tuck Haugaasen JM, Haugaasen T, Peres CA, Gribel R, Wegge P (2010) Seed dispersal of the Brazil nut tree (Bertholletia excelsa) by scatter-hoarding rodents in a central Amazonian forest. J Trop Ecol 26: 251–262.
- 42. Myers GP, Newton AC, Melgarejo M (2000) The influence of canopy gap size on natural regeneration of Brazil nut (Bertholletia excelsa) in Bolivia. For Ecol Manage 127: 119–128.
- 43. Kainer KA, Duryea ML, Macedo NC, Williams K (1998) Brazil nut seedling establishment and autecology in Extractive reserves of Acre, Brazil. Ecol Appl 8(2): 397–410.
- 44. Scoles R, Gribel R, Klein G N (2011) Crescimento e sobrevivência de castanheira (Bertholletia excelsa Bonpl.) em diferentes condições ambientais na região do rio Trombetas, Oriximiná, Pará. Bol Mus Para Emílio Goeldi 6(3): 273–293.
- 45. Buckland ST, Anderson DR, Burnham KP, Laake JL, Borchers DL, Thomas L (2001) Introduction to Distance Sampling: Estimating Abundance of Biological Populations. Oxford: Oxford University Press.
- 46. Johnston K, Ver Hoef JM, Krivoruchko K, Lucas N (2001) Using ArcGIS Geostatistical Analysis. Redlands: ESRI.
- 47. Emprapa (Empresa Brasileira de Pesquisa Agropecuária) (1997) Manual de métodos de análises de solo (Second Edition). Rio de Janeiro: Ministério da Agricultura e do Abastecimento.
- 48. Smith JJ (1993) Using ANTHROPAC 3.5 and a spreadsheet to compute a free-list salience index. Cultural Anthropology Methods Journal 5: 1–3.
- 49. Quinlan M (2005) Considerations for collecting freelists in the field: examples from ethnobotany. Field Methods 17: 219–34.
- 50. Ticktin T (2004) The ecological implications of harvesting non-timber forest products. J Appl Ecol 41: 11–21.
- 51. Janzen DH (1970) Herbivores and the number of tree species in tropical forests. Am Nat 104: 501–528.
- 52. Connell JH (1971) On the role of natural enemies in preventing competitive exclusion in some marine animals and in rain forests. In: Boer PJ, Gradwell GR, editors. Dynamics of populations. The Netherlands: Center for Agricultural Publishing and Documentation. Pp. 298–312.
- 53. Augspurger CK (1984) Seedling survival of tropical tree species: interactions of dispersal distance, lightgaps and pathogens. Ecology 65: 1705–1712.
- 54. Harms KE, Wright SJ, Calderón O, Hernández A, Herre EA (2000) Pervasive density-dependent recruitment enhances seedling diversity in a tropical forest. Nature 404: 493–495.
- 55. Wright SJ, Zeballos H, Domínguez I, Gallardo MM, Moreno MC, Ibáñez R (2000) Poachers alter mammal abundance, seed dispersal and seed predation in a Neotropical forest. Conserv Biol 14: 227–239.
- 56. Dirzo R, Mendoza E, Ortiz P (2007) Size-related differential seed predation patterns in a heavily defaunated Neotropical rain forest. Biotropica 39: 355–362.
- 57. Stoner KE, Vulinec K, Wright SJ, Peres CA (2007) Hunting and plant community dynamics in tropical forests: a synthesis and future directions. Biotropica 39: 385–392.
- 58. Peres CA, Nascimento HS (2006) Impact of game hunting by the Kayapó of Southeastern Amazonia: Implications for wildlife conservation in tropical forest indigenous reserves. Biodivers Conserv 15: 2627–2653.
- 59. Jansen PA, Hirsha BT, Emsensb WJ, Zamora-Gutierrez V, Wikelskia M, Kaysa R (2012) Thieving rodents as substitute dispersers of megafaunal seeds. Proc Natl Acad Sci USA 109 (31): 12610–12615.
- 60. Silvius KM, Fragoso JMV (2003) Red-rumped agouti (Dasyprocta leporina) home range use in an Amazonian forest: implications for the aggregated distribution of forest trees. Biotropica 35: 74–83.
- 61. Jorge MSP, Peres CA (2005) Population density and home range size of red-rumped agoutis (Dasyprocta leporina) within and outside a natural Brazil nut stand in southeastern Amazonia. Biotropica 37: 317–321.
- 62. Peres CA, Schiesari LC, Dias-Leme CL (1997) Vertebrate predation of Brazil-nuts (Bertholletia excelsa, Lecythidaceae), an agouti-dispersed Amazonian seed crop: a test of the escape hypothesis. J Trop Ecol 13: 69–79.
- 63. Howe HF, Smallwood J (1982) Ecology of seed dispersal. Annu Rev Ecol Evol Syst 13: 201–228.
- 64. Schwartzman S, Moreira A, Nepstad D (2000) Rethinking tropical forest conservation: perils in parks. Conserv Biol 14: 1351–1357.
- 65. Peres CA, Zimmerman B (2002) Perils in parks or parks in peril? Reconciling conservation in Amazonian reserves with and without use. Conserv Biol 15: 793–797.
- 66. Brockington D, Igoe J, Schmidt-Soltau K (2006) Conservation, human rights, and poverty reduction. Conserv Biol 24: 605–614.