Traditional knowledge hiding in plain sight – twenty-first century ethnobotany of the Chácobo in Beni, Bolivia

Background The Chácobo are a Panoan speaking tribe of about 1000 members (300+ adults) in Beni, Bolivia. Originally nomadic, the Chácabo were relocated to their current main location in the 1960s. Researchers have visited the Chácabo since 1911. A first more detailed anthropological report exists from the late 1960s, and ecological–ethnobotanical studies were conducted in the 1980s and 1990s. The presented work represents a complete ethnobotanical inventory of the entire adult Chácobo population, with interviews and plant collection conducted directly by Chácobo counterparts. Methods Based on previous reports and our preliminary studies, we hypothesized that twenty-first century Chácobo plant use centered on income generation, and that traditional plant use related to household utensils, medicine and traditional crop varieties had almost disappeared. To test this hypothesis, we started the “Chácobo Ethnobotany Project,” training 10 indigenous Chácobo participants in ethnobotanical interview and plant collection techniques, in order to more fully document Chácobo knowledge and avoid the influence of foreign interviewers. Results Our study found 331 useful plant species in 241genera of 95 plant families, with leaves, roots and bark being the most commonly used plant parts The comprehensive documentation that these methods enabled completely nullified our initial hypothesis of knowledge loss. Traditional crop varieties are still widely grown and traditional knowledge is alive. Moreover, it is being actively recuperated in certain domains by the younger generation. Most Chácobo know, and can name, traditional utensils and tools, although only the older generation has still the skills to manufacture them. While many Chácobo still know the names and uses of medicinal species, the younger generation is however often unsure how to identify them. Conclusions In this paper we illustrate the complexity of perspectives on knowledge at different ages, and the persistence of knowledge over almost a century. We found that traditional knowledge was only partially affected by the processes of exposure to a market economy, and that different knowledge domains experienced different trends as a result of these changes. Overall knowledge was widely distributed, and we did not observe a directional knowledge loss. We stress the importance to not directly conclude processes of knowledge loss, cultural erosion or acculturation when comparing the knowledge of different age groups.


Background
The Chácobo tribe, living in Northeastern Bolivia, were first visited by the European traveler Erland Nordenskjöld in 1911 [1], followed by an anthropologist only in 1956, who published the last account of Chácobo life before the tribe came under the influence of American Evangelist missionaries [2]. The Summer Institute of Linguistics (SIL) worked with Chácobo communities from 1953 to 1980, and produced the first account of Chácobo linguistics [3], and an unpublished work on Chácobo customs, with a strong focus on evangelist development [4]. This account is in interesting juxtaposition to the writings of German anthropologist Kelm [5], who visited the Chácobo in 1970, in the middle of SIL rule. The SIL finally was replaced in 1980 by the Swiss Protestant mission. Missionary rule led to a profound change of lifestyle, and a permanent process of acculturation [6]. From 1983 to 84, Brian Boom (New York Botanical Garden) led the first ethnobotanical study of Chácobo, documenting their knowledge after almost 30 years of cultural change [7]. Boom did however base his work on the plants collected from a single 1 ha forest plot. In 1995 the Institut Franjáis d'Etudes Andines financed a re-survey of Boom's plot, but the results were never released to the public, and a planned publication [8] existed in a single volume in the Institute's main library in Lima. Muñoz et al. [9] published a study on anti-malarial plants used by the Chácobo. Given the availability of previous studies, the Chácobo are an outstanding possibility to study traditional knowledge over time.
Traditional knowledge (TK) has been recognized for its importance for the protection of ecosystem services and biodiversity [10,11]. However, researchers and policymakers have equally expressed concern about its possible loss as societies modernize. A growing number of studies have reported changes and losses in TK (e.g. medical TK [12][13][14][15], nutritional TK [16], and agricultural TK [17][18][19]. The hypothesis that TK systems are able to adapt to external changes and internal pressures has discussed for some time (e.g., [20]). Traditional.
Knowledge is an important part of the adaptive capacity of many rural and indigenous communities that have been conserving biodiversity while enhancing livelihoods and adapting to disturbance and change [21,22]. Few studies have however examined whether and how loss or alteration of TK in fact occurs [17,[23][24][25]. Consequently, our understanding of the resilience of TK systems and their ability to evolve and adapt is highly limited. The assumption of loss of TK, when younger people know less, is one of the common errors in ethnobotany [26]. This problem can be remedied when analyzing the effects of age and age cohorts separately [27,28]. Not all TK domains might however be shared between generations, and as such there can be domains that would be more vulnerable to TK loss, and domains in which new knowledge is generated as an adaptation to environmental change [29].
Traditional knowledge is also seen as an important component in improving the management of natural resources [10,20,30] and practices relating to the protection of ecosystems and species [11]. Factors such as gender, age, ethnicity, birthplace, and level of education have been identified as important on an individual level [31][32][33][34]. Family size, integration into the market economy (e.g. sale of animals and agricultural products), or amount of material goods at family level (e.g., possessions of farm animals, tools, and transport) have been linked to the household levels [35,36]. Access to commercial centers, and to health, education, electricity or water, as well as land tenure systems and settlement history have shown a greater relevance at the community level [37][38][39]. In the absence of a unifying theory or common research methods, it is however difficult to clearly recognize whether or not these patterns exist at broader scales [40]. Several studies have used literature metadata to analyze large-scale usage patterns of plants [41][42][43]. In many cases, however, comparisons are difficult to make, given the diversity of the objectives and methods employed.
Based on previous reports mentioned, and our own preliminary studies [44], we hypothesized that twenty-first century Chácobo plant use centered on income generation through collection of forest products and agricultural production, and that traditional plant use related to household artifacts and medicine, as well as traditional crop varieties, had almost disappeared. We also hypothesized that the "missionary generation"the first age group growing up under restrictive evangelist rule, would report less TK than other age groups. Because access to markets and services has been reported as a major cause for TK loss [28], we also hypothesized that in villages most distant from the main market center (Riberalta), knowledge about the use of plants, and the number of useful species would be more homogeneously distributed through the generations [25], and expected that this TK distribution show different patterns when analyzing the different domains of knowledge about the use of plants [29].
To test our hypotheses, we started the "Chácobo Ethnobotany Project," training 10 indigenous Chácobo participants in ethnobotanical interview and plant collection techniques, to comprehensively document contemporary Chácobo TK and avoid the limiting influence of foreign interviewers.

Methods
The study area --The Chácobo and Pacahuara The Chácobo belong to the Panoan linguistic group, which includes about twelve tribes (Chácobo, Pacahuara, Matis, Matses, Yaminahua, Ese Eja and others). At the end of the 1890s, the Chácobo lived as semi-nomadic hunters and cassava and maize cultivators, probably in two groups, one with six families and one with four, in north Bolivia, between Lake Roguagnado and the river Mamore, south of their current territory. During the rubber boom in the early 1900s, they were forced by more aggressive tribes to move north, where rubber tappers, who also brought disease and epidemics to the tribe, threatened them. While other tribes were enslaved to work in rubber stations, the Chácobo managed to avoid most of the outside influences. Their first permanent contact with the outside world occurred only in 1953 with members of the the Tribes Missions, and in 1954 the Bolivian government established an agency about 15 km from the current location of Puerto Limones. The missionary linguist Gilbert Prost arrived in 1955 under the auspices of the Summer Institute of Linguistics (SIL). According to [4] there were four Chácobo groups living between the Benicito and Yata rivers at that time, numbering about 200 people [7]. Prost and his wife continued to live among the Chácobo until 1980. In addition to translating the New Testament into Chácobo, they made some observations on cultural                                           and linguistic practices [3,4]. In 1964, Prost managed to buy a territory in the north of the Chácobo's ancestral lands, forming the community of Alto Ivón, and most of the remaining population moved there. In 1965, the Bolivian government finally assigned 43,000 ha of land to the Chácobo, although this area was less than 10% of their original territory. The influence of the SIL caused profound cultural change among the Chácobo, including the reported abandonment of traditional costume and dances in 1969 [4]. The official indigenous organization of the Chácobo (Central Indígena de la Región Amazónica de Bolivia (CIRABO) estimates a current population of the Chácobo community of about 1000 people (350+ adults), with Alto Ivón as the largest settlement. The current territory of the tribe according to CIRABO encompasses 450,000 ha, and is roughly equivalent to the original extent of the tribe's ancestral lands (Fig. 1). The elevation of the territory is about 200 m, and much of the vegetation can be classified as humid tropical Amazon rainforest. However, the territory encompasses also large tracts of periodically inundated savannas, dominated by Mauritiella armata, and large, drier, savanna areas with forest islands. The average annual temperature is 26.8°C, with an average annual rainfall of 1560 mm. A distinct dry season lasts from June to November [7]. Today the Chácobo are governed by two indigenous organizations: The Capitanía Mayor Chácobo, closely linked to the evangelists, and the Chácobo-Pacahuara Association, recognized by the Central Indígena de la Región Amazónica de Bolivia (CIRABO), and supported by the Central de Pueblos Indigenas del Beni (CPIB) and the Confederacion de Pueblos Indigenas de Bolivia (CIDOB).

Ethnobotanical and botanical collection
Our project explored the current traditional knowledge (TK) on plant use of the Chácobo and Pacahuara in Beni, Bolivia and had three goals: 1) to discover and document current traditional plant knowledge through interviews and surveys, 2) to inventory the current flora of the region, and 3) to repatriate the acquired knowledge as well as previous data to the community.
After obtaining consent from CIRABO, and before starting fieldwork, we conducted a community meeting in May 2013, involving representatives of all 27 villages in the Chácobo Territory, in order to obtain prior informed consent from all communities. This session included the repatriation of the results of previous studies [45][46][47]. In addition, during the project all available material on Chácobo plant use was translated to Spanish and repatriated [48]. The Chácobo community itself choose 12 local counterparts to be trained as ethnobotanical interviewers and plant collectors. In September 2013 we conducted a two-week workshop on ethnobiological methods and plant collection, training the 12 selected counterparts, 10 of which finally acted as interveiwers. Training was conducted directly in the field in the central village of Alto Ivón, and involved theoretical exercises (overview on methodology of interviews, collection and herbarium techniques), as well as extensive practical exercises (structuring and testing of questionnaires, test interviews among the participants, field interviews with local community members, plant collection in the field, preparation of herbarium specimens, plant and artifact collection in the local community, data-basing, and initial data analysis).
From November 2013 to May 2015, Chácobo interviewers collected ethnobotanical information from 301 Chácobo participants (150 women, 151 men, representing almost the entire adult Chácobo population), and over 1500 plant samples were collected. Prior to starting the interviews, every interviewer obtained prior oral informed consent from each participant. Chácobo participants were divided into five age classes (18-30 years old: 58 men, 52 women; 31-40 years old: 31 men, 36 women; 41-50 years old: 35 men, 36 women; 51-60 years old: 15 men, 7 women; and >60 years old: 12 men, 19 women). Because the study attempted to interview the whole adult Chácobo population, there was originally no emphasis on achieving a balanced age or gender distribution. All interviews were conducted at the homes of the participants by asking participants to freelist their plant knowledge following [49]. All plant uses were categorized following [49]. All interviews were preferably conducted in Chácobo. In a few cases where participants were not fully fluent in Chácobo, interviewers used Spanish as common language. The plant material was collected under permission from the Ministry of Environment and Water of the Plurinational State of Bolivia, and was identified and deposited at the National Herbarium of Bolivia (LPB) under the collection numbers of the Chácobo collectors. Nomenclature follows www.TROPICOS.org. Use descriptions were coded after the fact into subcategories and, for some analyses, into six major categories: fodder, fuel, medical, cultural, construction, tool, and food.
All work was carried out following the International Society for Ethnobiology Code of Ethics [50], and under the framework provided by the Nagoya Protocol on Access to Genetic Resources and Fair and equitable sharing of benefits arising from their use of the Convention on Biological Diversity, the Chácobo community retains the copyright of the traditional knowledge of all informants. Any commercial use of any of the information requires prior consensus with informants and communities, and an agreement on the distribution of benefits.

Data analysis
The total number of unique species reported and unique uses reported for each use category were compared across communities, genders, and age groups (16-30, 31-40, 41-50, 51-60, and 61-82) for 292 informants (dropping for this analysis 8 informants for whom age was not indicated).
To gain a more nuanced look at how these qualities affected not only the number of reports but which species or uses were reported, we ordered informants using non-metric multi-dimensional scaling on distance matrices for plants and uses, and tested how well vectors (age) and factors (gender, ethnicity, community) fit the location of informants in the ordination, using the R package vegan [51]. We used similar methods with plant family fit onto an ordination from distance matrices of plant-use combinations to test whether plant family explains the uses to which plants are put.
We used Indicator Value [52], as implemented in the R package labdsv [53] to combine occurrence frequency and mean abundance of species and uses to elucidate species and uses that had higher fidelity to and/or relative abundance in certain age groups or genders. For this analysis, the P value is the probability of finding an equally high indicator value in random permutations. Species with significantly high indicator values had higher fidelity and relative abundance in certain age groups / genders (were 'indicators'). We further compared age and gender groups by informant consensus factor (ICF) for each use category, calculated as the number of use reports minus the number of taxa over the number of use reports minus one: (Nur − Nt)/(Nur − 1). We also measured consensus on species uses by quantifying what proportion of each species' mentions fall within a specific use category.
Plant species and plant family importance was ranked by four metrics: Community and Informant Cultural Importance (CIcom/CIinf )the sum within species across all plant-uses of the number of informants (for CIinf ) or communities (for CIcom) reporting a plant-use over the number of informants/ communities reporting the plant; Diversity of Uses (Du)the Shannon Index of uses [51]; and Use Value (UV), the number of reports of a species over total number of informants asked in a region [54].
To test whether greater knowledge of Chácobo language was associated with a more similar set of knowledge and / or a larger knowledge set of plants and uses, we used the ordination based on uses to examine whether interviewees who reported more Chácobo names tended to report a more similar set of uses, and used linear regression to test whether the number of Chácobo names reported was significantly greater for those who reported more species or more uses.

Results
The availability of previous field data gives the unique opportunity to study the long-term change in knowledge of an indigenous group in the age of globalization. Our study found 331 useful plant species in 241genera of 95 plant families, with leaves, roots and bark being the most commonly used plant parts ( Table 1).
The larger Chácobo communities showed very similar patterns in the number of species used, with differences within communities usually greater than between, although Nueva Unión stood out in reporting more food species (Fig 2a). Likewise, all communities were similar in plant-uses (use descriptions for a species within each use category), although in this case Nueva Unión reported fewer use descriptions within the Utensils and tools and Cultural categories, while Motacuzal and Alto Ivón reported more medical uses (Fig. 2b). Within these categories, number of species and uses was fairly consistent across age groups, though we observed a trend for some categories of more species and uses known with  increasing age. The age group between 51 and 60 years (i.e. the first age group growing up under missionary rule), showed a slightly lower knowledge, especially evident in the medical and cultural categories but also in food plants (Fig. 3). These metrics are also quite similar across gender, although across most categories the average number of species and uses reported by women was slightly higher (Fig. 4).

Who uses what and how?
Despite the similarities among communities in total species and uses reported, we found that informant community significantly influenced both which plants and which uses individual informants reported ( Table 2 a&b). In contrast, and in accord with the results above, age and gender did not significantly influence either. Ethnicity of the participants influenced which plant species they used, but did not explain what they were used for. Given the very low r 2 values, it is clear that much variety in uses was not explained by any demographic and environmental variables explored (Table 2 a&b). In the ordination, we can see this effect more clearly: although there was much overlap, the communities clearly structure which plants were reported. This difference was however much driven by the reports Nueva Unión (Fig. 5).
While age did not in itself explain the ordination well, we did find certain plants to be associated with age categories. In this we found no indicator plants or uses among the first three age groups (16-30, 31-40, 41-50), which suggests to some extent that plants and uses reported by these groups are less distinct than that of the second two age groups (51-60, >60). The 51-60 age group was associated with by Styrax sp., Iryanthera juruensis, Xylopia ligustrifolia, Hirtella pilosissima, Inga sp. 1, and Piper nigrispicum, while the >60 group was indicated by Gustavia hexapetala, Astrocaryum aculeatum, Phenakospermum guianensis, Attalea phalerata, Apuleia leiocarpa, Bixa orellana, Hancornia speciosa, Zingiber officinale, and Eriotheca sp. Likewise, the use subcategory Firewood was associated with the 51-60 age group while the medicinal use subcategories: Skin and subcutaneous tissue, Sensory system, Respiratory system and Musculo-skeletal system all were associated with the >60 age group.
Likewise, although gender did not fit to the overall ordinations, there was a large number of plants  associated with female, and a much smaller one with male respondents (Table 3). Interestingly, all indicator uses were exclusively associated with women (Table 4).

Informant consensus factors (ICF)
Looking at specific use categories we found broadly similar trends across age categories and genders: tool, construction and food uses usually had the most use reports. We found a lower number of medicinal use reports, although the same number of respondents reported medicinal uses. Food uses consistently had less ICF than tool and construction uses, and medicinal uses even less. Cultural uses, while often reported by fewer informants and with fewer uses, show disproportionately high ICF (Table 5).
Plant relative importance metrics did show a different picture underlining the problems of using diversity indices. The Cultural Importance Index yielded wildly different species sets for Community and Individuals, and both Use Value Index and Use-diversity Index again yielded different sets as species as most important (Table 6).
Because the Cultural Importance Index tends to prioritize species with few informants, we highlighted the species that had both high index values in general, and also a large number of reports to elucidate species that were of high importance in all indices. As result, Vismia macrophylla, Xylopia peruviana, Attalea phalerata, Gossypium barbadense, Attalea maripa and Phenakospermum guianensis were elucidated as the most important species in the daily life of the Chácobo community (Fig. 6). Overall, however, informant consensus was very high in across all age groups and across all use categories (Fig. 7). Arecaceae, Fabaceae, Malvaceae and Rubiaceae were found to be the most important plant families used across most indices, although Moraceae did yield a higher ranking in Use Value (Table 7).
Results also indicated that qualities of plants did to a certain extent explain which uses they were put to. A large number of plant families had specifically Medicinal uses, while other sets of plant families were specifically    used for Food, Utensils and tools, and Construction. Not surprisingly, data also revealed that plant families with high importance in all indices calculated (Arecaceae, Fabaceae, Malvaceae and Rubiaceae) had uses in all categories (Fig. 8). Different use categories also had different levels of fidelity in the species that were reported for them. For instance, relatively few mentions in the construction and tool categories were of species that are uniquely associated with those categories. In contrast, a much greater proportion of mentions for medical uses were of species that were only used for medical uses. This pattern was also true of food plants (Fig. 9).

Does language influence use knowledge?
Interviewees who reported more Chácobo names did indeed tend to report more similar sets of knowledge, and knew more species and uses (Fig. 10). In addition, the number of plants or number of uses reported strongly increased with the number of Chácobo names participants knew (Fig. 11). Although in some degree this was a feature of the study (there was no way to informants to report more names than species), it was clear that very few of those participants with great knowledge of species or uses failed to report a large number of Chácobo names.

Discussion
While other studies found indecisive patterns of the influence of age, or accessibility to markets on traditional knowledge (negative [17,23,44,55]; positive [25,56,57]), our study did not reveal any pattern that would link differences in plant-use knowledge to age or accessibility of a location, but simply to specific location and associated flora in each of the communities. In most communities the contact with nature still remains vital to the acquisition of knowledge [58,59], and the facility to observe and identify the useful plants clearly adds to this.
The observation that local and indigenous languages often package rich traditional ecological knowledge has led to the question in many studies of whether indigenous language abilities influence plant knowledge, i.e. if native language speakers have a higher knowledge than participants only speaking a mainstream language [44,60]. In our study, the link between language proficiency and other metrics of traditional knowledge (plants and uses reported) does support at least the correlation of these variables, and suggest the possibility of simultaneous language and knowledge retention (or erosion).
The general trend found in relation to the difference in intergenerational knowledge suggests that any patterns are most likely a result of both knowledge transmission, as well as in situ learning, and be related to the time during which people acquire and use knowledge, with the older informants taking more responsibility in their households, who have a need to learn and apply their knowledge [27,54,61]. The knowledge of older people might not have been affected by the need to find new subsistence activities, and was thus Most important species in each index in bold preserved without external influence [62]. The fact that the only generation that did show decrease of traditional knowledge (albeit slight) was the generation of 41-50-year old participants, who had grown up under restrictive missionary rule, is noteworthy. The hypothesis that people who are relatively isolated from the market economy share more traditional knowledge than people who live close to cities or larger towns [25], was not met in our study, because in most places the contact to nature still remains vital to the acquisition of knowledge [58,59]. The predominance of the use for Human food in the more widely shared knowledge can be explained as a long and constant learning process that begins in early childhood, and is common in the more remote locations [26,54].
There is no doubt that Chácobo daily life has changed in the course of the last century. Early accounts of the Chácobo all indicate the wide use of bark-cloth, and little enthusiasm for the rather conservative clothing style which missionaries tried to introduce [2,5]. Boom [7] mentions however the complete disappearance of this custom. However, while the Chácobo use western style clothing available in the markets of Riberalta, traditional bark cloth is still widely used for cultural purposes, and most participants knew how to make it.
Changes in the use of traditional implements were very subtle. Most households still use large pounding tubs, as well as the large wooden boards used to pound food, which have not changed over time. Large clay pans for roasting jibe (Manihot flour), and smaller ceramic pots are also widely used. Even little stools from the petioles of Mauritia flexuosa and balsa wood (Ochroma sp.), first documented by Nordenskjöld [1] are still found in many houses, although they were completely missed in all previous studies. The production of burden baskets has not changed since [1], and the same species are still used today. However, only a few older women in the communities still have the skills to weave baskets, and modern implements like backpacks are clearly replacing traditional materials. Similarly, canoes are still an important means of transportation. However, while Nordenskjöld, Haenke and Kelm described canoes made from bark [1,2,5], the modern variety is made of hollowed out tree trunks, which is already indicated in [7]. House construction and roofing have however not changed much in the last 100 years. Bows and arrows are still maintained as hunting implements, especially for fishing, and all arrow types found in previous studies are still used among the population, although 22 caliber rifles and 20 gauge shotguns are favored for hunting.
Based on previous reports, we originally hypothesized that many household artifacts as well as traditional clothing had disappeared from Chácobo life. Many of these artifacts were mentioned in the 1922-1970 accounts, but not in later studies. Boom [7] and Bergeron [8] in particular indicate that traditional tools and clothing had disappeared. This turned out to be an interview artifact. Early anthropologists, who focused on Chácobo daily life [2,5], while Boom and Bergeron focused only on plants collected from one 1 ha forest plot [7,8]. Our combined study indicates that in fact most artifacts of the Chácobo are still known, and also used, by a large part of the population. This includes traditional clothing that is still being prepared and used on important occasions, as well as hunting and household implements. In daily life however, no traditional clothing and ornaments are found anymore, and the large monkey tooth breast-plates mentioned by [1] and [2] have indeed disappeared.
In case of food, market access has indeed had an influence in Chácobo life. In the 1980's cassava (Manihot esculenta, Euphorbiaceae) was clearly the most important food for Chácobo, and seven varieties were planted (Boom 1987). Maize (Zea mays), was planted on 18% of the land, and upland rice (Oryza sativa) was only planted on 7% of the land [7]. Nowadays rice has become the staple food of the Chácobo, leaving cassava and maize in a more secondary role, However, all original traditional maize and cassava varieties, as well as traditional banana varieties, are still grown. In our work we also found all edible species mentioned by Boom (1987) as planted in home-and forest-gardens, but the Chácobo had incorporated many additional species, e.g. lemon (Citrus sinensis, Rutaceae) in home gardens, and Psidium, Myrica sp. and Eugenia sp. in the forest gardens. One noteworthy exception was the palm Huanimá (Bactris gasipaes var. chichagui, Arecaceae), actively sown formerly in abandoned clearings to collect palm fruits [7]. In 2015 the palm was only found rarely around the villages, and was no longer planted.
The Chácobo keep using a large number of plants for medicinal purposes although missionaries of the Summer Institute of Linguistics tried to eradicate traditional medicinal plant use and traditional agricultural practices, because they regarded this as pagan [4]. Early anthropological and missionary accounts mentioned hardly any medicinal species [2,4,5], but this was clearly an interview artifact. Of the 360 plant species collected by Boom, 174 species were of medicinal value [7]. Bergeron recorded 399 useful plant species, of which 166 were classified as medicinal [8]. This compares favorably to the over 331 useful plant species elucidated in the current study. The Chácobo still favor the preparation of remedies by boiling the leaves, bark or fruits to cure diseases. While Boom did not find a true "healer" among the Chácobo [7], several Chácobo healers were identified in the present study. The knowledge of medicinal plants was particularly alive among older informants interviewed, but younger participants still retained much of such knowledge. The use of plant poisons, especially for fishing was mentioned as highly important by [7], and is still practiced today.
One of the most profound changes in Chácobo life seems to be a return to nomadic patterns, now mostly linked to commerce and income generation. The production of oil from the seeds of Brazil nuts (Berthollettia excels) was reported by Boom [7], but is little practiced nowadaysall nuts are now sold to large companies in Riberalta. The Brazil nut harvest takes place from January and March, and during that time now almost the entire Chácobo population migrates to the South of the territory where the largest concentration of Berthollettia is encountered. During the rest of the year Alto Ivón remains the main population center. However, degree to which a species is uniquely associated with a category number of mentions in a category from that species Fig. 9 Number of mentions that of a species plotted against that species' fidelity to the given use category many Chácobo have "second" homes in Tokyo, where most of the fields are located at present, or in Triangulo, closer to their main fishing sources, and conveniently located at the road to Riberalta.

Conclusions
In this paper we illustrate the complexity of perspectives on knowledge at different ages, and the persistence of knowledge over almost a century. We found that traditional knowledge was only partially affected by the processes of exposure to a market economy, and that different knowledge domains experienced different trends as a result of these changes. Overall knowledge was widely distributed, similar to [63]. However, we did not observe a directional knowledge loss, contrasting [64].
We stress the importance to not directly conclude processes of knowledge loss, cultural erosion or acculturation when comparing the knowledge of different age groups. These results should be treated with caution, because they cannot rule out the role of other variables affecting knowledge, including changes in the composition of other important factors that might be affected by the influence of access to a market Chacobo names uses A B Fig. 11 The number of species (a) and uses (b) was significantly greater for those who reported more Chácobo names. Loess regression line (for Chácobo informants only) is shown economy. It is important to remember that learning, and accumulating experiences, require time. For this reason, the alternative explanation that the knowledge of older people tends to have accumulated over time, compared to the younger generation, should also be considered. It also needs to be taken into account that older generations might have different perceptions of their environment, because their points of reference are different from those of younger people. The ability to generate and apply knowledge in human populations enables actions and adjustments in response to current and future changes. Similarly, the ability to generate and apply knowledge, and not the knowledge itself, helps to increase the resilience of socio-ecological systems. The analysis presented here clearly suggests that perceived knowledge "loss" might easily be an artifact of the researcher's presence, of limited time, and of a very limited number of participants. Training local interviewers provides an excellent tool yield more reliable information on traditional knowledge and its potential loss in the future.
In compliance with the Nagoya Protocol, the original field notebooks, as well as the complete dataset, and a guide on useful plants of the Chácobo was repatriated to the Chácobo [65]. All members of the tribe have access to the compiled interview data for purposes of learning and education. The data collected are a valuable resource to the community as a tool to preserve their traditional knowledge, and will encourage the launch of research projects and community activities so the information does not become static. Species identified as being most important to the community can be targeted for conservation and restoration activities.