The Social Nature of Agrobiodiversity in Central Amazonia


Material Information

The Social Nature of Agrobiodiversity in Central Amazonia
Physical Description:
1 online resource (203 p.)
Kawa,Nicholas C
University of Florida
Place of Publication:
Gainesville, Fla.
Publication Date:

Thesis/Dissertation Information

Doctorate ( Ph.D.)
Degree Grantor:
University of Florida
Degree Disciplines:
Committee Chair:
Oyuela-Caycedo, Augusto
Committee Members:
McCarty, Christopher
Schmink, Marianne C
Heckenberger, Michael J
Smith, Nigel J


Subjects / Keywords:
agriculture -- agrobiodiversity -- amazonia -- anthropology -- conservation
Anthropology -- Dissertations, Academic -- UF
Anthropology thesis, Ph.D.
bibliography   ( marcgt )
theses   ( marcgt )
government publication (state, provincial, terriorial, dependent)   ( marcgt )
born-digital   ( sobekcm )
Electronic Thesis or Dissertation


The Amazon is recognized as a region of extraordinary botanical diversity harboring a wide array of useful plants. While the conservation of both crop genetic and species diversity in Amazonia has become an increasing concern in recent years, the factors that influence the distribution and maintenance of plant diversity in rural communities are poorly understood. Utilizing data from surveys and ethnographic interviews at 138 rural households in the municipality of Borba, Amazonas, Brazil, this dissertation examines the primary factors that influence the diversity of plant species (i.e. species richness) managed by smallholder farmers. Multiple regression analyses demonstrate that the independent variables ?age of household head,? ?area of land under cultivation?, and ?environmental habitat? (floodplain or highlands) are statistically significant determinants of species richness, accounting for 31.5% of variation in species richness among households. Ethnographic data point toward several other socio-cultural influences that shape the contemporary patterning of plant diversity, including social networks of exchange, gender roles in agricultural management, cultivation of magic plants, and the individual life histories of farmers and the lands they manage. In short, this dissertation demonstrates how history, social relationships, and cultural beliefs have important bearing on the management of agrobiodiversity in rural Amazonian communities, and thus demand consideration in any initiatives aimed at agrobiodiversity conservation.
General Note:
In the series University of Florida Digital Collections.
General Note:
Includes vita.
Includes bibliographical references.
Source of Description:
Description based on online resource; title from PDF title page.
Source of Description:
This bibliographic record is available under the Creative Commons CC0 public domain dedication. The University of Florida Libraries, as creator of this bibliographic record, has waived all rights to it worldwide under copyright law, including all related and neighboring rights, to the extent allowed by law.
Statement of Responsibility:
by Nicholas C Kawa.
Thesis (Ph.D.)--University of Florida, 2011.
Adviser: Oyuela-Caycedo, Augusto.

Record Information

Source Institution:
Rights Management:
Applicable rights reserved.
lcc - LD1780 2011
System ID:

This item is only available as the following downloads:

Full Text




2 2011 Nicholas C. Kawa


3 In memory of Sandy Davis, Mariete Alencar, and Clara Gubbins


4 ACKNOWLEDGMENTS First, I must thank my s upervisory committee chair, Augusto Oyuela Caycedo He gave me acc ess to his personal library, continually encouraged me in my research, and provided important motivation during the writing of this dissertation. I also thank my supervisory committee members, Mike Heckenberger, Chris McCarty, Marianne Schmink and Nigel Smith for their help and support. In addition, I thank Ignacio Porzecanski Hugh Popenoe and Florence Babb for their useful insights at different stages of my graduate career. I must also acknowledge the United States Department of Education for making this research possible through a Fulbright Hays Doctoral Dissertation Research Award. At the Instituto Nacional de Pesquisas da Amaznia I extend my gratitude to Charles Clement who served as an advisor and mentor during my time in Brazil. I also greatly appreci ate the support of fellow students and colleagues at INPA, especially James Fraser Andr Junqueira Joana Maia Salomo Claide de Moraes and Anne Rapp Daniel I must additionally thank Newton Falco and Jos Francisco Gonalves who were sources of support since my early days at INPA in 2003. In Manaus, I warmly thank Rafa, Vinicius, Rodrigo, Roberto, and Suzan Padilla for opening their home to me I also thank all my friends at Vi la do Sol Maior, especially Jo o, Dudu, Daniel, Vanessa, Maria Iva ni, Marcelo, Augusto, Danielzinho, and Francimar. I extend a particulary warm thanks to the Campos family, especially Dona Le for her continued friendship and invitations to eat bolo podre In Borba, I give special thanks to Denise Barata for being a gre at friend and providing me with a room to stay I extend a warm abrao to the Pleplecada many barbeques and evenings of dominoes. I a lso thank Valdo and Jaime who were


5 genuine friends who made me feel at home in Borba The same can be said for the Alencar family especially Tlio and Mariete. I thank all the families that participated and helped me in my research, especially those who became good friends. I want to thank in particular the families at the communities of Puruzinho, Guariba, Auar Grande, and Vila Gomes. Their understanding, patience, and willingness to welcome me were deeply felt. I also thank the many individuals at the extension agency IDAM who supported me in my research, especially Tarcsio and Dorinha. Lastly, I thank my f amily: Mom, Dad, Jim, Nate and Kolleen I also thank my friends from the University of Florida, the University of Arizona, and Batavia who I conside r my larger, t name any names but they know who they are I do truly appreciate every last one of them.


6 TABLE OF CONTENTS page ACKNOWLEDGMENTS ................................ ................................ ................................ .. 4 LIST OF TABLES ................................ ................................ ................................ ............ 9 LIST OF FIGURES ................................ ................................ ................................ ........ 10 LIST OF ABBREVIATIONS ................................ ................................ ........................... 12 ABSTRACT ................................ ................................ ................................ ................... 13 CHAPTER 1 AGROBIODIVERSITY AND RURAL AMAZONIAN SOCIETY ................................ 15 Introduction ................................ ................................ ................................ ............. 15 Research Problem ................................ ................................ ................................ .. 17 Dissertation Framework ................................ ................................ .......................... 17 Agrobiodiversity Studies ................................ ................................ ................... 18 Historical Ecology ................................ ................................ ............................. 20 Social Networks ................................ ................................ ................................ 21 The People: Riverine Peasants and Caboclos ................................ ........................ 23 Study Area ................................ ................................ ................................ .............. 24 Sampling ................................ ................................ ................................ ................. 25 Research Methods ................................ ................................ ................................ .. 27 Botanical surveys ................................ ................................ ............................. 27 Household questionnaires ................................ ................................ ................ 2 7 Social Network Analysis ................................ ................................ ................... 28 Ethnographi c life histories ................................ ................................ ................ 29 Contribution of Dissertation ................................ ................................ ..................... 30 2 HISTORY AND AGROBIODIVERSITY IN BORBA AND THE LOWER MADERIA RIVER ................................ ................................ ................................ ... 33 Introduction ................................ ................................ ................................ ............. 33 Pre Columbian Agriculture and Agrobiodiversity ................................ .................... 33 Col onial History ................................ ................................ ................................ ....... 37 The Rubber Boom ................................ ................................ ................................ .. 44 ................................ ................................ ........... 46 Jute Agriculture on the Madeira River ................................ ................................ ..... 51 Conclusions ................................ ................................ ................................ ............ 52 3 CONTEMPORARY AGROBIODIVERSITY AT THE HOUSEHOLD LEVEL ........... 56 Introduction ................................ ................................ ................................ ............. 56


7 Agriculture and Agrobiodiversity in Smallholder Communities ................................ 56 Methods ................................ ................................ ................................ .................. 59 Results ................................ ................................ ................................ .................... 61 Discussion ................................ ................................ ................................ .............. 65 Age of Household H ead ................................ ................................ ................... 65 Area of Cultivated Land ................................ ................................ .................... 67 Environmental Habitat: The Uplands and Floodplain ................................ ....... 67 Household Wealth, Market Orientation, and the Barge Economy .................... 70 Oral Histories and Magic Plants ................................ ................................ ....... 73 Conclus ions and Future Research ................................ ................................ .......... 74 4 SOCIAL NETWORKS AND MANIOC VARIETAL DIVERSITY ............................... 87 Introduction ................................ ................................ ................................ ............. 87 Background ................................ ................................ ................................ ............. 88 Manioc Diversity and Social Exchange in Amazonia ................................ .............. 89 Methods ................................ ................................ ................................ .................. 91 Municipal Manioc Varietal Diversity and Exchange ................................ ................ 93 Network Analysis of Municipal Manioc Diversity ................................ ..................... 94 Manioc Varietal Diversity, Exchange, and Selection in Puruzinho Lake Communities ................................ ................................ ................................ ........ 98 Network Analysis of Communities on Puruzinho Lake ................................ .......... 102 Manioc Exchange Network for Communities at Puruzinho Lake .......................... 103 Knowledge and Manioc Agriculture ................................ ................................ ...... 104 M anioc Diversity Loss ................................ ................................ ........................... 106 Conclusions ................................ ................................ ................................ .......... 109 5 MAGIC PLANTS AND THEIR CONTRIBUTION TO AGROBIODIVERSITY ........ 118 Introduction ................................ ................................ ................................ ........... 118 The Anthropology of Magic and Magic Plants ................................ ....................... 118 Categories of Magic Plant s in Amazonian Peasant Communities ........................ 121 Plants with Power to Repel (and Protect) ................................ ....................... 121 Plants with Powers of Attraction ................................ ................................ ..... 127 Plants with Power to Heal ................................ ................................ ............... 129 Contribution of Magic Plants to Overall Species Diversity ................................ .... 131 The Ambiguous Nature of Plants Used for Magic and Healing ............................. 132 Magic Plants and Religion ................................ ................................ .................... 134 Conclusions ................................ ................................ ................................ .......... 135 6 GENDER AND AGROBIODIVERSITY MANAGEMENT IN RURAL AMAZONIA 143 Introduction ................................ ................................ ................................ ........... 143 Background: Gender, Inequality, and Agrobiodiversity ................................ ......... 143 Species Diversity by Gender of Household Head(s) ................................ ............. 145 T he Roa and the Homegarden ................................ ................................ ............ 146 Case 1: Dona Nanda and Nilo ................................ ................................ ........ 146


8 Case 2: Cndida and Jos ................................ ................................ ............. 149 Case 3: Alcia ................................ ................................ ................................ 151 Discussion and Conclusions ................................ ................................ ................. 153 7 AGROBIODIVERSITY, AMAZONIAN HISTORICAL ECOLOGY, AND FINAL CONCLUSIONS ................................ ................................ ................................ ... 158 Inherited Agrobiodiversity ................................ ................................ ..................... 160 Historical Ecology, Biodiversity, and Anthropogenicity ................................ .......... 162 ............................... 165 ................................ ..................... 168 Cobra Grande and the Evolving Amazonian Landscape ................................ ...... 170 Concluding Remarks: The Anthropology of (Agro)biodiversity .............................. 174 Research Questions Posed in Dissertation ................................ ........................... 175 Primary Factors Influencing Diversity at Household Level ................................ .... 175 Social Networks ................................ ................................ .............................. 177 Gender ................................ ................................ ................................ ........... 178 Local and Regional History ................................ ................................ ............. 178 AGROBIODIVERSITY AND HOUSEHOULD SURVEY #1 ................................ ......... 183 AGROBIODIVERSITY SURVEY #2 ................................ ................................ ............ 185 LIST OF REFERENCES ................................ ................................ ............................. 187 BIOGRAPHICAL SKETCH ................................ ................................ .......................... 203


9 LIST OF TABLES Table page 1 1 Sampling of communities and households surveyed in Bor ba, AM, Brazil ......... 32 3 1 Descriptive statistics of households surveyed in Borba, AM, Brazil .................... 76 3 2 List of species identifed in bot anical surveys at 138 households in Borba, AM, Brazil ................................ ................................ ................................ ........... 77 3 3 Regression model of number of species per household in Borba, Amazonas, Brazil. ................................ ................................ ................................ ................. 86 4 1 Varieties of bitter and sweet manioc managed by households surveyed in Borba, AM, Brazil ................................ ................................ .............................. 111 5 1 List of species used in magic and symbolic practice in Borba, AM, Brazil ........ 140


10 LIST OF FIGURES Figure page 1 1 Map of communities surveyed in the municipality of Borba, Amazonas, Brazil .. 31 2 1 Archaeologist Claide Moraes standing above the defensive ditch at Vila Gomes (August 2010) ................................ ................................ ........................ 54 2 2 Claide Moraes standing in the trough of the defensive ditch a t Vila Gomes (August 2010) ................................ ................................ ................................ ..... 54 2 3 Excavation at Vila Gomes in terra preta soils (>1 meter depth) with transitional Oxisols in the lower 10 cm (August 2010) ................................ ........ 55 2 4 Brazil nut trees surrounding a stretch of terra preta that was converted to cattle pasture (Vila Gomes; August 2010) ................................ .......................... 55 3 1 Floodplain homegarden in the comm unity of So Pedro, Autazes Au River (August 2010) ................................ ................................ ................................ ..... 75 3 2 Frequency distribution of species richness managed by households in Borba, Amazonas, Brazil ................................ ................................ ................................ 84 3 3 Relationship of household head age to number of species managed by the household ................................ ................................ ................................ ........... 85 3 4 Relationship of area of cultivated land (log transformed) to number of sp ecies managed by the household ................................ ................................ ................ 86 4 1 Municipal network of households tied by manioc varieties planted in common and colored by geographic location in the municipality ................................ ..... 113 4 2 Municipal network of households tied by manioc varieties in common with 4 factions calculated ................................ ................................ ............................ 114 4 3 Municipal network of households tied by ma nioc varieties, nodes sized by betweenness centrality and colored by exchange. ................................ ........... 114 4 4 Municipal network of households tied by manioc varieties, households colored by community. ................................ ................................ ...................... 115 4 5 Puruzinho Lake network of households tied by 1 or more manioc varieties planted in common, households colored by community. ................................ .. 115 4 6 Pu ruzinho Lake network of households tied by 2 or more manioc varieties planted in common, households colored by community, nodes sized by betweenness centrality. ................................ ................................ .................... 116


11 4 7 Puruzinho Lake network with households tied by manioc varieties exchanged 116 4 8 Puruzinho Lake network with households tied by manioc varieties exchanged, nodes sized by betweenness centrality.. ................................ ....... 117 5 1 Pio Roxo, a plant used in healing baths and to ward off the evil eye .............. 136 5 2 Rue, an Old World plant used to ward off the evil eye ................................ ...... 136 5 3 Poraqu, the electric eel plant, used to protect oneself when playing soccer ... 137 5 4 Mucuraca, the possum plant, commonly use d in healing baths ...................... 138 5 5 Chumbinho receives a bath from his cousins (Puruzinho, August 2010) .......... 13 9 6 1 olds ................................ ................................ ...... 155 6 2 ................................ ....... 156 6 3 Canteiro on the Autazes Au River (August 2010) ................................ ........... 157 7 1 The silhouette of a jackfruit tree with its toothed leaves (Vila Gomes, August 2010) ................................ ................................ ................................ ................ 180 7 2 Nilo standin g in the plot inherited from his grandfather, Manoel (Auar Grande, November 2009) ................................ ........................ 181 7 3 Sapo holding the small anaconda that we had stepped on in a creek bed near Lago do Comprido (November 2009). ................................ ...................... 182


12 LIST OF ABBREVIATION S ADE Amazonian Dark Earth, anthropogenic soils known as terra preta do ndio in Portuguese. CEPLAC C omiss o Executiva do Plano da Lavoura Cacaueira ( Brazilian Cacao Production Planning Comm ission ) IDAM Instituto de Desenvolvimento do Amazonas (Amazonas State Development Institute) INPA Instituto Nacional de Pesquisas da Amazonia ( National Institute of Amazonian Research )


13 Abstract of Dissertation Presented to the Graduate School of the U niversity of Florida in Partial Fulfillment of the Requirements for the Degree of Doctor of Philosophy THE SOCIAL NATURE OF AGROBIODIVERSITY IN CENTRAL AMAZONIA By Nicholas C. Kawa August 2011 Chair: Augusto Oyuela Caycedo Major: Anthrop ology The Amazon is recognized as a region of extraordinary botanical diversity harboring a wide array of useful plants. While the conservation of both crop genetic and species diversity in Amazonia has become an increasing concern in recent years, the f actors that influence the distribution and maintenance of plant diversity in rural communities are poorly understood. Utilizing data from surveys and ethnograp hic interviews at 138 rural households in the municipality of Borba, Amazonas, Brazil, this diss ertation examines the primary factors that influence the diversity of plant species (i.e. species richness) managed by smallholder farmers Multiple regression analyses demonstrate that the unde floodplain or highlands) are statistically significant determinants of spe cies richness account ing for 31. 5 % of variation in sp ecies richness among households E thnographic data point toward several other soci o cultural influences that shape the contemporary patterning of plant diversity, including social networks of exchange, gender roles in agricultural management, cultivation of magic plants, and the individual life histories of fa rmers and the lands they ma nage In short, t his dissertation d emonstrates how history social relationships, and


14 cultural beliefs have important bearing on the management of agrobiodiversity in rural Amazonian communities, and thus demand consideration in any initiatives aimed at a grobiodiversity conservation.


15 CHAPTER 1 AGRO BIODIVERSITY AND RURAL AMAZON IAN SOCIETY Introduction In recent decades, concern has grown over the loss of global biodiversity with particular attention toward agrobiodiversity. The FAO claims that since the 1 900s more Furthermore only 12 species of plants and 5 species of animals have come to represent 75% o f global food consumption (FAO 1999 ). This loss in agrobiodiversity has largel y resulted from the rise of industrial agriculture, in which high yielding uniform crop varieties have been widely adopted over genetically diverse traditional crop varieties, and monocultural plantations dependent on high inputs (e.g. fertilizers and pest icides) have been promoted over diverse, low input polycultures ( Altieri 1999; Thrupp 2002 ). Smallholder farmers, particularly in the developing tropics, are responsible for managing a significant proportion of the remaining global agrobiodiversity, and various movements have begun to work towards the conservation of genetic and species diversity of useful plants in developing areas across the globe. Some efforts have focused on ex situ conservation, preserving germplasm in seed banks (e.g. Schoen and Br own 2001). Many scholars, however, have argued for increased focus on maintenance of in situ crop diversity, or in other words, crop diversity in managed agricultural landscapes (Altieri 1999; Brush 1995, 1999 2000, 2005 ). Heterogenous agricultural lan dscapes with high plant diversity are not only important for the conservation of useful plant species and varieties, but they also provide key benefits for smallholder farming families, including heightened food security, improved nutrition, and mitigation of pests and disease that attack crops (Thrupp 2002;


16 Zhu et al. 2000). In general, more heterogenous landscapes are viewed as being in line ng term productive potential, are less likely to lead to environmental degradation and are overall more ecologically landscapes ( Loreau et al. 2003; Jackson et al. 2007 ) Since smallholder farmers managing more diverse landscapes generally need less agrochemical inputs for agricultural prod uction, they also tend to pollute less, and can help to maintain important ecosystem services (Altieri 1999). In addition, farmers that manage a great diversity of plants create more niche habitats that support diversity of other organisms inc luding birds mammals, and insects ( Duelli and Obrist 2003; Estrada et al. 1997 ; Jackson et al. 2007; Thomas and Marshall 1999 ; Thrupp 2002 ). This diversity also has been said to have socio economic advantages fo r smallholder farmers, including food s ecurity and heig htened risk management ( DiFalco and Perrings 2005; Perreault 2005). Despite the described benefits of agrobiodiversity, great variation exists in the diversity of plants managed by rural smallholder household s The broad objective of this dissertation i s thus to understand why some households manage a greater diversity of plants than others Although this research focuses on agrobiodiversity in Amazonia specifically the larger questions investigated in this dissertation are of universal appeal: What a re the primary factors that influence the diversity of plant s managed by smallholder farming households ? What is the role of social networks, cultural beliefs, and gender roles in the maintainance and patterning of s uch diversity ? Lastly, h ow does an exa mination of local and regional history inform our understanding of the plant diversity seen at present?


17 Research Problem Many studies have demonstrated that the diversity of plant species managed by smallholder farmers is distributed unevenly both amon g communities and within them (Coomes and Ban 2004; P adoch and de Jong 1991; Perrault Archambault and Coomes 2008). Increasing articulation of smallholder farmers with market economies is often considered to have a negative influence on the agrobiodiversi ty of smallholder farms, but empirical studies have demonstrated mixed results when examining the relationship between market participation and on farm crop diversity ( Major et al. 2005; Vadez et al. 2004; Zimmerer 1996: 24 25 ) Studies in the Peruvian Am azon suggest that household demographic and socio economic factors have important bearing on crop diversity including land wealth and number of available laborers (Coomes and Ban 2004; Perrault Archambault and Coomes 2008). Most recently, researchers inte rested in agrobiodiversity have identified that social networks may also contribute significantly to the distribution of crops within communities ( Coomes and Ban 2004; Perrault Archambault and Coomes 2008; Subedi et al. 2003). Since many rural communit ies access seeds and planting materials through local social networks, the position of a household in networks of exchange can influence the variety of crops or crop varieties accessible to the household. Through the examination of myriad socio cultural, dem ographic, environmental, economic, and historical data, this dissertation attempts to identify the primary factors that influence the diversity of useful plants found on smallholder farms in the municipality of Borba, Amazonas, Brazil Dissertation Framewo rk To develop a framework for investigating factors influencing useful plant diversity in rural Amazonian communities I have chosen to draw on empirical and theoretical


18 insights from agrobiodiversity researc h, historical ecology, and social network studie s interaction of environmental economic, and demographic variables as well as social networks and related social histories for understanding the contemporary distribution of useful plant species and crop varieties (Zimmerer 1996). Agrobiodiversity Studies A wide array of variables has been examined for explaining variation in plant n smallholder farming communities. At the household level, key variables that have been identified include ethnicity of the household, age and gender of household head and number of available laborers in household ( Coomes and Ban 2004, Cromwell and van Oosterhout 1999 Perrault Archambault and Coomes 2008) Age is considered an important variable because crops species are continually acquired over the lifetime s of r ural smallholders (Perrault Archambault and Coomes 2008: 118). Gender can also influe nce the diversity of plant species and crop varieties man aged by a household particularly in Amazonian societies where management of species diverse homegardens is t he responsibility of women (e.g. Descola 1994 ). Also, when comparing households among communities, the size of the community is considered an impo rtant control variable since larger communities tend to pool a large r diversity of plants (Perrault Archambault and Coomes 2008). I mportant economic variable s influencing species richness include t he a rea and number of land holdings. In general, househol ds with greater land wealth tend to manage a higher number of plant species among smallholder farming communities (Coomes and Ban 2004; Cromwell and van Oosterhout 1999: 227). In several studies, household wealth has also been linked to agrobiodiversity. In the Peruvian Amazon it is


19 suggested that access to wealth is closely related to access to planting material (Perrault Archambault and Coomes 2008). However, the source of household wealth can be a critical factor as off farm income has been shown to ha ve a negative effect on agrobiodiversity ( Reyes Garcia et al. 2008; Yongneng et al. 2006). The influence of markets is also commonly discussed in studies of agrobiodiversity management, but little consensus exists as to how arket affect s agrobiodiversity management and the diversity of species managed at the household level Some studies argue that increased market orientation leads to degradation of agrobiodiversity in smallholder farming communities (Kehlenbeck and Maass 2 004 ; Michon and Mary 1994). agrobiodiversity is ambiguous and does not necessarily result in agrobiodiversity loss (Major et al. 2005; Vadez et al. 2004; Zimmerer 1996: 24 25). Still other studies have shown that communities that are more isolated from urban areas and markets actually manage less species diversity than those closer to urban areas (Padoch and de Jong 1991; Wezel and Ohl 2005). These contradictory findings may result in part from variability in individual communities and households respond to market influences differentl y A grobiodiversity s tudies also recognize that environmental factors play an important role in the patterning of useful plant diversity In Amazonia, floodplains (v rzea) usually attract a form of agriculture that is very distinct from that conducted on the uplands (terra fi rme). However numerous ecotones and microen vironments exist within the generalized floodplain/upland dichotomy (Moran 1993 : 21 ). Moreover, many


20 smallholder farmers exploit different environmental niches to cultivate different suites of crops. While cer tain habitats may be bes t suited for particular crops, great variation exists in the number of species that farmers can potentially manage in a given habitat For this study, data were collected from communities in distinct environmental habitats to accou nt for differences particularly in the uplands and floodplains. Historical Ecology Within environmental anthropology, historical ecology has emerged as a new research program for understanding human environmental interaction. Although historical ecology is not a true t heoretical paradigm per se (Bal e 2006), the key questions that drive research in historical ecology are enabling new ways of understanding human environmental interaction. It argues that humans not only adapt to the natural environment, bu t also play a critical role in shaping it and even enriching it through time. This project is oriented in part by perspectives from the research program of historical ecology, which has sparked much new anthropological research in Amazonia and South Amer ica in general ( Bal e 1994 ; Bal e and Erickson 2006 ; Denevan 1992 b ; Erickson 2005 ; Heckenberger et al. 2007 ; Heckenberger et al. 2008 ; Oyuela Caycedo 2008 ; Rival 2002 ) While much of historical ecological research has been aimed at looking at what the con temporary landscape can tell us about the human past, this project is more concerned with examining the specific historical processes and social relations that were influential in shaping the contemporary landscape. In this sense, this project is an attem pt to develop a historical ecology of the present with the aim of describing the fine grained aspects of landscape evolution vis a vis gender, exchange networks, and individual life histories. How the environment restricts, confines, and


21 reacts to these s oci al and historical conditions is examined in turn, so as to appreciate the recursive and dialogical nature of such relationships. Historical ecological perspectives have particular import for the study of agriculture and contemporary agrobio diversity in Amazonia. European colonization introduced a large number of exotic economic plants into Amazonia, including bananas ( Musa spp.) mangoes ( Mangifera indica ) coffee ( Coffea arabica ) citrus ( Citrus spp.) and sugar cane ( Saccharum officinale ) which were widely adopted During the rubber boom in the late 19 th and early 20 th century, much of the Brazilian Amazon was flooded by migrants from Northeastern Brazil who also brought with them crops from that region ( Dean 1984). In Borba, as in many other munici palities along the Madeira River, the rubber boom led to a major repopulation of the area after the initial decline of the indigenous populations in the post contact period ( Weinstein 1983: 53). Although migrants who moved to the region to tap rubber were often bound by a debt peonage system referred to as aviamento others were relatively independent and able to farm when not gathering seasonal extractive resources ( Weinstein 1983: 243 ). Many contemporary farming communities in the Central Amazon trace b ack to this period of occupation, as Northeastern immigrants and native Amazonians intermarried and established settlements. Today, this history is still reflected in the stands of aging rubber trees ( Hevea spp.) and cacao ( Theobroma cacao ) scattered in c ommunities, which once served as the basis of the regional economy. In many ways, this history continues to influence agricultural production and exchange in the region today. Social Networks Social network analysis is an increasingly popular method withi n the social sciences for examining the patterning of social relationships and the influence of those


22 relat ionships on a host of phenomena (Borgatti et al. 2009). Social network analysis thus focuses on relations between individuals and the ways individua ls are connected to form a system or a network. There are two primary forms of social networks: whole certain kind among all members of a population (Wellman 1983 : 160). In this research, I utilized whole networks at a regional scale (farmers tied by manioc varieties planted in common) and local scale (farmers who exchanged manioc cuttings within three cl ose knit communities) to examine the influence such networks had on the patterning and distribution of manioc ( Manihot esculenta ) landraces. Although I originally intended to track generalized netw orks of plant and seed exchange I decided to focus strict ly on the exchange of manioc, the pr imary staple crop of the region, because it allowed for more precise data collection. Prior research suggests that social networks have important bearing on crop diversity (Coomes and Ban 2004; Perrault Archambault and C oomes 2008; Subedi et al. 2003), but few studies at present have applied concepts derived from network theory in the study of crop diversity. One of the few studies to investigate the relationship of social networks to agrobiodiversity in Amazonia found t hat a large part of exchange of planting materials occurred between community members rather than with outsiders, and much of it was shaped by kin ties (Coomes and Ban 2004). The researchers also viewed e xchange as important contributor to diversity as th exchange more seeds, cuttings, etc. with other households have more diverse gardens


23 and vice versa diversity begets exchange, as exchange begets diversity Coomes and Ban 2004 : 430 ). In the Oaxaca Valley of Mexico a study by Badstue and colleagues investigated whether or not social networks had developed specifically around the ex change of seed in farming communities focused on maize ( Zea mays ) production. They found that no collective action existed specifically for seed exc hange, but rather the majority of exchange occurred through existing social networks, particularly among kin (Badstue et al. 2006). A study by McGuire (2008) in eastern Ethiopia examined the role of social networks in sorghum ( Sorghum spp.) seed exchange and observed that access to seed varied considerably among households within communities. They argued that both quantities supplied by donors and the terms of exchange depended greatly on positions in social networks and local hierarchies. Lastly, in Tib et, Subedi and colleagues ( Oryza sativa ) varieties and found that or those who represented hubs within networks, were often wealthier and maintained a greater number of rice varieties. These no dal farmers were also described as knowledgeable persons in the community and played an important role in dissemination of information (Subedi et al. 2003). The People: Riverine Peasants and Caboclos This study was conducted with riverine peasants who are commonly referred to as caboclos. Although the term caboclo is still often generally used to describe an emergence is tied to racial classification in the colonial period. T he first appearance of the term in dictionaries describes caboclos while later definitions describes caboclos as individuals of mixed Portuguese and


24 indigenous descent (Parker 1985). In the colonial period urbanites and white c ollar workers applied the term caboclo to the illiterate and semi literate agriculturalists and collectors of the rural zones (Wagley 1976: 289). In this manner, caboclo became not only a racial category, but a term used by elite cl asses to establish social differentiation from the Amazonian peasantry. Amazon Town was the first full length ethnography describing caboclo culture and in many ways his description s of caboclos and caboclo livelihoods are still largely a pplicable today. Yet Wagley also acknowledged that the term had a pejorative nature ( Wagley 1 976 : 140). In my personal involvement with rural Amazonians, I tended to avoid usage of the word because of my own ambiguous feelings regarding the term. Noneth eless, academics employ the term as a general social category referring to the Amazonian peasantry (see Harris 2000, Ioris 2005, and Nugent 1994 for further discussion). Some have suggested that academics should refrain from using the term to describe rur al Amazonian peasants (Lima 2004, Pace 1997), yet others argue that despite its problematic nature, other options are equally unattractive or unsuitable (Harris 2000: 15). Study Area The Madeira River is the longest tributary of the Amazon and one of the largest rivers in the world. The Madeira cuts a large gash across South America, descending from the Peruvian Andes in a Northeasterly direction, eventually dumping into the Amazon River in the heart of the basin. As su ggested by its name (literally the wood river raw materials and goods, most recently soy from the southern rim of the region. Between Manaus, the primary economic center of the Central Amazon, and soy coun try


25 to the south, the Madeira courses through a large expanse of land populated largely by riverine peasant communities. The municipality of Borba, located on the lower Madeira River was one of the earliest settlements established in the Central Amazon ( C omisso de Estudos da Estrada de Ferro do Madeira e Mamor 1885: 73 ). It lies 150 kilometers southeast of the city of Manaus (215 km by waterway), the capital of Amazonas state. Today, Borba covers an area of 44,251 km 2 and has a population of 31,098 inh abitants (IBGE 2005 ). The economy of Borba depends largely on agriculture and extractive activities. Manioc, bananas, watermelon ( Citrullus lanatus ) beans ( Vigna spp.), papaya ( Carica papaya ) and citrus are among the primary agricultural products produ ced in the area, while rubber, timber, Brazil nuts ( Bertholletia excelsa ) copaiba oil ( Copaifera multijuga ), and rosewood essential oil ( Aniba rosaeodora ) represent major extractive resources. Soon Borba and other municipalities of the Lower Madeira River are likely to undergo radical changes in their relationship to regional markets. The imminent paving of the BR 319 highway will connect Borba and nearby municipalities by an all weather road to the urban center of Manaus ( Fearnside and Graa 2006). T his development is likely to have a profound effect on agricultural production in the region, and for this reason, resea rch at this site represented a unique opportunity to assess the factors influencing agrobiodiversity in the region before such changes occu r. Sampling Research was conducted at 17 rural smallholder communities in the municipality of Borba (Figure 1 1). Contact had been established in previous research with 5 of the communities (Kawa 2008 ). Because of the difficulties and problematics of ind ependently contact ing communities as a foreign researcher, s tate agricultural


26 extension agents operating in Bo rba facilitated introductions at the other 12 communities. Several factors were considered in the sampling of communities, including environmenta l habitat (floodplain, uplands), community history, community size, geographic location, and market access. 9 of the communities sampled w ere located on the floodplain, 7 wer e located on the uplands, and one community stra d dled both the uplands and floodp lain. In terms of access and geographic location in the municipality, 10 communities wer e located on the Madeira River the prinicipal waterway in the area. Of the 7 remaining communites, 3 were localized on a flo odplain lake near the Madeira, one commun ity was loc ated on the Autazes A u River, one on the Madeirinha River, one on the Para n do Mand channel (off the Madeira River) and one community was a government settlement located outside of the town of Borba on an unpaved road. T his served as a rela tively good representation of the diversity in commun ities in the municipality (numbering nearly 200 in all) when considering environmental, social, and historical factors. From the 17 sampled communities, 138 households were surveyed in all ( Table 1 1). A sampling frame from which to draw random samplings of households in the communities was not available (Bernard 2006: 149). Since most of the communities were relatively small, a sampling of 90% of households was sought to sufficiently account for varia tion among households However, this was achieved at only 5 communities. Because of high mobility of residents and only isolated visits to the more remote communities, it was very difficult to attain more complete sampling for the remaining 12 communitie s. In the case of Puxurizal, which is a large governmental settlement with 100 properties, not all properties serve as primary residences. Despite


27 the relatively low number of households surveyed there, they were included to explore any potential differe nces in the agricultural practice and agrobiodiversity management of government settlers when compared to residents of historically established communities Research Methods Botanical surveys At each farm visited, I conducted b otanical surveys after receiv ing the household At all households, I invited the household head (or heads) to name all the species of plants under management in homegardens, orchards, and annual crop plots. In the case of the primary staple crop, manioc, I asked what v arieties were planted by the household. I walked around the homegarden to view as many plants as possible and made visits to annual fields when it was not inconvenient for the household. I collected attribute data of individual crops including approximat e area of cultivation, use(s) of the crop, and duration (i.e. perennial or annual). Some plants included in the survey were not directly planted by households, but were cared for (i.e. wee ds were cleared from their base ) by the household. Other plants th at were spontaneous volunteers, but were utilize d by the household (e.g. papaya or star nut palm) were also counted. For each household, I tallied the total number of species under management and total number of manioc varieties cultivated. Household ques tionnaires I employed structured questionnaires to elicit information from household heads regarding number of household members number of adults in the household length of occupation of the property number of cultivated fi elds belonging to the househol d, the area of land under management and the total area of landholdings (Appendix A) I also


28 collected data on the sources of household income, use of agricultural inputs, access to credit, and participation in agricultural projects. I collected environ mental data, inquiring specifically about the soils in which the household planted their crops. For a portion of the households, I read them a list of consumer goods and asked which items they had in their possession. The intention of this exercise was to develop a Guttman scale of household wealth for the study (a more detailed description of the method is provided in Chapter 3). Social Network Analysi s I collected social network data related to the planting of manioc varieties at the community and muni cipal level in order to account for the different scales at which social networks could affect the distribution of this staple crop. For all households surveyed in this study household heads were asked what varieties of manioc they planted and the area o f land dedicated to manioc. A 2 mode matrix was developed of households by manioc varieties planted to analyze patterns of households that had manioc varieties planted in common. This matrix was used to create a w hole social network for all the househ olds in the 17 communitie s visited in the municipality which served to explore the broad geographic distribution of manioc varieties and their management in the area (see Chapter 4 for further description) At 45 households in three closely related commu nities, a second structured interview was implemented inquiring specifically about manioc cultivation. In these three communities, household heads were asked to name the manioc varieties they had planted and with whom they had exchanged manioc cuttings in the 12 months prior to the interview. From these data, a 1


29 developed to analyze patterns of manioc varietal exchange (or loaning) at the community level. Both the 1 mode and 2 mode matrices were analyzed fo r structural information of the networks, looking specifically at factions and their relation to varietal distribution. The measure of etweenness centrality defined as the number of times a node (or a household) lies along the s hortest path between tw o others, was calculated for these networks (Scott 2000: 86) This measure was used to explore the relationship between household centrality and manioc vari etal diversity and distribution. This measure also served to identify key households that stood ou t in the networks, and pointed to individual cases that deserved complementary ethnographic detail. Ethnographic life histories I employed ethnographic methods to gather indispen sable qualitative information. Specifically, I collected life histories of i ndividual farmers to develop a more nuanced vision of agrobiodiversity management over time. Along with the life histories of farmers, I gathered information on the history of the farms themselves to discuss the development of diversity at the household l evel. As I found in previous research, diversity encountered on Amazonian farms is not always the product of the management by contemporary farmers. Many individuals inherit land with an existing diversity of economic plants introduced either in the Pre columbian era or in the late colonial period. Others farmers clear lands near forests that have pre existing concentrations of economic species without known histories. By acknowledging the complexity of these relations, I hope d to approach this on going dialogue betwe en humans and the environment


30 Contribution of Dissertation Although human activity has incredible bearing on biodiversity, particularly agrobiodiversity, socio cultural factors influencing this diversity are poorly understood. In this dis sertation, I present social, demographic, economic, environmental, and historical data to examine the factors have primary bearing on the diver sity of plants managed on rural Amazonian farms. I also implement social network analysis, a useful form of quan titative data analysis for examining social relationships, to test the degree to which social networks of exchange have bearing specifically on manioc varietal diversity. Supplementing quantitative analyses, I discuss ethnographic data and life histories in order to qualify statistical analyses and strengthen the understanding of F ood security is threatened in many areas of the world and the demand for alt ernative sources of fuel is on the rise. For these reasons and more, conservation of both genetic and species diversity of economic plants has become a global imperative. In identifying the key social, economic, environmental, and historical factors that influence the di versity of useful p lants, this dissertation provides information critical for informing policies and initiatives aimed at sustaining agrobio diversity in Amazonia and the rural tropics in general.


31 Figure 1 1 Map of communities surveyed in the m unicipality of Borba, Amazonas, Brazil


32 Table 1 1 Sampling of communities and households s urveyed in Borba, AM, Brazil Community Locale (River) Enviroment # Households # Households surveyed % Households surveyed Auar Grande Madeira (upstream) High floodplain 23 21 91.3 Boa Esperana (Mandi) Paran do M andi Uplands 6 4 66.7 Floresta Madeira (upstream) High floodplain 4 1 25.0 Guariba Madeira (downstream) Floodplain 15 10 66.7 Puruzinho vrzea Madeira (downstream) Floodplain 17 17 100.0 Puruzinho centro 1 Puruzinho Lake Uplands 23 21 91.3 Puruzinho c entro 2 Puruzinho Lake Uplands 10 9 90.0 Puxurizal None city settlement Uplands 100 14 14.0 Vila Gomes Madeira (upstream) Uplands 6 4 66.0 Sempre Viva Santana Madeirinha Uplands 14 5 35.7 So Pedro (Autazes Au) Autazes Acu Floodplain; Uplands 12 9 7 5.0 (Anuma) Madeira (downstream) Floodplain 12 9 75.0 So Pedro (Anuma) Madeira (downstream) Floodplain 6 5 83.3 Monte Horebe Madeira (upstream) Floodplain 3 3 100.0 Alexandre Madeira (downstream) Floodplain 9 2 22.2 So Sebastio (Anuma) Madeira (downstream) Floodplain 6 4 66.7 Puru Grande Puruzinho Lake Uplands -1 -TOTALS -265 138 -


33 CHAPTER 2 H ISTORY AND AGROBIODI VERSITY IN BORBA AND THE LOWER MADERIA R IVER Introduction ntemporary agrobiodiversity can not be und erstood without first examining the In t his chapter I provide a general overview of the history of Borba and the Lower Madeira River in relation to agriculture and extractivism, and discuss the consequences of such history for und erstanding contemporary agrobiodiversity in the region It begins by overviewing agricultural management in the Pre Colombian era based on the limited archaeological surveys of the past, and I draw on recent research conducted at the community of Vila Gom es, which was surveyed in this study. Following this, I describe the effect of the Columbian exchange and the opening of Amazonia to global markets during the colonial period, and then examine latex extractivism during the Rubber Boom (late 19th and early 20th centuries) and World War II. In describing this latter period, I include oral histories collected from elder residents of Borba to provide a more nuanced discussion of latex extraction and its surrounding social historical context. Lastly, I discus s the brief period of jute agriculture on the Madeira River and the more recent history of agricultural production and management in the mun i cipality. Pre Columbian Agriculture and Agrobiodiversity In the late Pre Columbian e ra, the Lower Madeira River is believed to have been relative ly densely populated, although archaeological evidence has been limited until recently The abundance of Amazonian Dark Earth (i.e. anthropogenic soil) sites in the area, paired with the findings of the brief archaeological s urveys by Nimuendaj, Hilbert, and Simes and Lopes, support this notion (Nimuendaj 2004; Hilbert 1968; Simes


34 and Lopes 1987). A recent survey conducted in the municipality demonstrates that prior archaeological surveys only began to scratch the surface of the wealth of archaeological sites in the municipality ( Moraes 2010; Fraser et al. 2011). Before European arrival at least 138 different species of plants were under cultivation or management in Amazonia (Clement 1999). Manioc likely played an impo rtant role in diets of the past as it does today. An Amazonian domesticate well adapted to the highly weathered and acidic soils of the region, it is argued that manioc and manioc flour probably gained prominence in the regional diet as more sedentary pop although archaeological evidence is scant ( Arroyo Kalin 2008 : 46 ) In addition to manioc, other key domesticat es included New World yam ( Dioscorea trifida ), sweet potato ( Ipomoe a batatas ), ho t pepper ( Capsicum spp.), peanut ( Arachis hypogaea ) cotton ( Gossipyum spp.) and tobacco ( Nicotiana spp.) (Miller et al. 2006). Domesticated and semi domesticated palms including most notably peach palm ( Bactris gasipaes ), but also American oil palm ( Ela eis oleifer a ), patau ( Oenocarpus bataua ), and Star Nut palm ( Astrocaryum spp.), made important contributions to the indigenous diet as well ( Clement 1999; Morcote Rios and Bernal 2001 ) Important fruit trees likely to have been planted in agroforestry sy stems would ( Psidium guajava ) cashew apple ( Anacardium occidentale ) hogplum ( Spondias mombi n ), and various fruits from the Sapotaceae and Annonaceae families (e.g. Annona and Pouteria spp.) (Miller and Nair 2006). Research on Amazonia n Dark Earths ( known locally as terra preta de ndio ) suggest s that large sites of anthropogenic soils were the product of sedentary


35 agricultural villages dating from 2500 to 500 year s ago (Neves et al. 2003) A relationship between the rise of manioc agr iculture, sedentism, and the forma tion of anthropogenic soils is being proposed in recent archaeological and ethnoarchaeological research in the region: over 2000 years ago with concomitant inc rease in population and the size of sedentary increase in the disposal of manioc byproducts as a source of bulk organic matter is one of the key ingredients in the forma tion of darkened a nthrosols 882 ; see also Arroyo Kalin 2008: 174 175 ) In addition to manioc byproducts, the deposition of fish and animal remains, human waste, palm thatch, charcoal from village fires, and other organic waste contributed to the formation of these environments of heightened soil fertility ( Glaser et al. 2001: 40 ). The present day community of Vila Gomes in Borba is situated on one such archaeological site with nearly 20 hectares of anthropogenic soil (Moraes 2010 ) Surroun ding much of the community is a defensive ditch that was dug by the indigenous population in the late Pre contact era ( Figures 2 1 and 2 2 ). The ditch is nearly 1100 meters in length, approximately 2 meters in depth, and ranges in width between 2 and 3 met ers. While dark earths and transitional soils extend somewhat beyond the ditch, these anthropogenic soils extend no more than 10 cm below the surface horizon at the areas beyond the ditch while in central portions of vila, they reach as de ep as 1 meter o r more (Figure 2 3). At the height of its occupation, Vila Gomes would have had a relatively large population, inferred from calculations of necessary labor to construct the genic


36 soils also suggests that in the past as today, homegardens and central living space were concentrated on the area of dark earth while manioc fields were likely planted in transitional soils and non anthropogenic Oxisols in the land beyond the ditch ( Fieldnotes August 20, 2010) It is notable that large concentrations of Brazil nut trees are found in the forests surrounding the community of Vila Gomes ( Figure 2 4 ). These are probably vestiges of historical Brazil nut groves cultivated or nurture d by indigenous populations ( Bale 1989 ; Shepard and Ra mirez 2011). This patterning has been described as a form of i ndigenous populations managed forests surrounding villages, planting and selecting useful palms and fruit t rees that rendered (Clement 1999 : 191 ; see also Posey 1985 ) The effects of this management continue to show as forests situated on and around archaeological sites tend to harbor unique concentration s of useful pl ant species. A recent study in the municipality of Manicor on the middle Madeira River also demonstrates that secondary forests on anthropogenic soils have distinct species composition when compared with those on adjacent non anthropogenic soils (Junqueira et al. 2010 ) Sev eral useful tree species serve as indicators for anthropogenic soils including hogplum ( Spondias mombin ), murumuru ( Astrocaryum murumuru ), and urucuri ( Attalea c.f. phalerata ; Junqueira et al. 2009 ). In the municipality of Borba, it was also shown that contemporary farmers managing anthropogenic soils typically plant different market crops than those managing Oxisols (Kawa et al. 2011) Despite the centuries that have passed, Pre Columbian agriculture and natural resourc e


37 management thus continue to have a pronounced influence on Amazonian forests, soils, and even contemporary agrobiodiversity and its management. Colonial History The arrival of the Portuguese and the Colonial period brought radical changes to the Amazonia n landscape and its agrobiodiversity T he region was colonized not only by new populations of people but also a wide array of foreign flora and fauna in what has been describe d as a form of Crosby 1986 ) A number of annual crop s were introduced during this period including sugar cane, indigo ( Polygonum tinctorium ) bananas and rice (Miller et al. 2006). Many perennial fruit trees from the Old World were also later brought to Amazonia and became naturalized including mango bre adf ruit ( A rtocarpus altilis ) jackfruit ( Artocarpus heterophyllus ) malay apple ( Syzygium malaccense ) several varieties of citrus trees, and later coffe e (Dean 1987: 23) 1 During this time, new agricultural implements were also adopted. Steel tools that arrived with the Portuguese eventually came to replace indigenous technologies like stone axes and digging sticks, and thus changed cropping patterns and agricultural management systems (Denevan 1992 a ; Miller et al. 2006). Despite the new crops and techno logies that were introduced into Amazonian agriculture, much of the colonial economic production relied on the rather unsophisticated collection of a select number of native plants, including cacao, sarsaparilla ( Smilax spp.), wild clove or Dicype llium caryophyllatu ) Brazil nuts, and later, rubber. 1 Dean mentions that coffee was smuggled into Brazil, a n event that is often overlooked in discussions about biopiracy in modern Brazil.


38 The lower Madeira River remained relatively isolated from colonial influence until the 18 th century 2 Borba, first known as Aldeia do Trocano was founded by the Jesuit Priest Joo de Sampaio around 1728 (Comisso de Estudos da Estrada de Ferro do Madeira e Mamor 1885: 73 ) 3 However, after its establishment, the mission was moved downstream because of incursions by attacking Indians and the threat of malaria (Santos 1999: 78; Marcoy 2001; Keller 197 5: 54). In 1755, the settlement was named the first vila ( i.e. Portuguese town) of the Rio Negro Captaincy (which would later a As Borba was being established, settling colonists w ere engaged i n on going conflict with the Mura Indians (Leite 1943 : 43 ; Marcoy 2001: 207; Santos 1999: 78). In the late 2010:148), and the life of the Mur a was described as being r elayed in the form of a repet it ivo notici rio de Guerra Amoroso 1992: 297). At the beginning of the 18 th century, the Mura population expanded, occupying territory left by other ethnic groups that suffered from depopulation b ecause of rampant disease introduced by the colonizing Europeans. The Mura thus became a hegemonic ethnic group, dominating at a time when other groups were weakened and living in vilas and villages under the 98 ; Nimuendaju 1 9 48 control on the Madeira River in the 18 th century (Davidson 1970: 12). Colonial 2 potential of the Madeira valley, which had already yielded some cacao, cr avo, and timber, and where the 264). 3 uncertain. Francisco Jorge dos Santos (1999) cites the date 1724 while other authors provide later estimates (e.g. Keller 18 75: 44)


39 documents presented in 9 that the Mura had killed many oarsmen of canoe collecting expeditions that sought out cacao along the Madeira River. In one episode it was reported that they had killed a white man, head of one canoe, and in another instance they sacked the fie lds of the village of Santo Antonio and attacked the indios aldeados or village I ndians (Autos 1986: 10 11; also cited in Amoroso 1992: 300). These d enunciations of the Mura are seen, however, as feeble justification s by the colonial and religious leaders in an attempt to formalize a war against the Mura and thus o pen the Madeira River region for extraction of the major export crop of the time: cacao (Amoroso 1992: 300). Cacao became a major export during the colonial period because of its demand in Euro pe for both culinary and medicinal purposes. In 1679, King Pedro II issued a carta r gia (royal directive) that encouraged all Brazilian landowners to plant cacao trees on their property as the Portuguese Crown hoped it could take advantage of this lucrat ive crop through taxation (Walker 2009) In the Captaincy of Gro Par and Maranho, in Eastern Amazonia, cacao even served as the official form of money in the early to mid 18 th century (Alden 1976; Sweet 1974: 65 ). Systematic cacao cultivation began in Bahia in Northeastern Brazil, but Amazonia was the cacao 4 along with other drogas do serto including sarsaparilla and wild clove The Madeira River in particular, was described as being abundant in cacao 5 es that search for cacao on the banks of the Madeira River are ordinarily always successful 4 eira Rivers were the product of Amerindian management either in agroforestry systems or forest gardens. At the very exploitation and deposition of se eds. 5 In 1699, a si milar claim is made by Padre Jo o Felipe Betendorff in Cronica dos Padres da Miss o da Companhia de Jesus no Estado do Maranho ( 1910: 355; also cited in Davidson 1970: 249)


40 paiol do cacao the Solim es River (Amazon R ( Daniel 2004: 86 87; translation mine). It was noted that the Indians who were sent out clove because as they collected, they could eat cacao pulp and also make vi nho 6 from it (Daniel 2004: 85) Cacao was also favored as it was often found growing on the margins of lakes and rivers, which were also rich in aquatic resources including manatees Sar saparilla on the other hand, was typically found deeper into upland forest, and because of its thorns was much more difficult to collect, while clove also required more work because only its bark was collected ( Daniel 2004: 85 86; also cited in Roller 2010). Collecting expeditions began early in colonial Amazonia as des cribed in a recent historical review of the period parties, had operated extensively but sporadically in the Portuguese Amazon since at least the mid seventeenth century, recruiting crewmen from tho se same settlements that would later become Directorate villages. Exports of uncultivated products (the so called drogas do serto ) fluctuated during this earlier period in response to labor shortages, disruptions in transatlantic shipping, overexploitatio n of collecting grounds, and price instability ( Roller 2010: 438). As noted here, missionaries played an important role in motivating early collecting expeditions. To finance mission activities, missionaries often tended small cacao plantations while en couraging (or in some cases, 6 rn Brazilian Amazonia drinks made from the Roller (2010) citing Padre Jo ferment ed cacao to create an alcoholic beverage.


41 coercing) their new converts to collect in the serto or backlands (see Walker 2009: 546 547) However, this development created competition bet ween missionaries and colonists, who experienced a strained relationship the colo nial period: between the settlers and the missionaries had existed since the beginnings of colonization in the Amazon and concerned such issues as the missionaries' opposition to the enslavement of the Amerindians, their segregation of Indians i n remote mission compounds from which seculars were excluded, their efforts to supervise labor relations between the Indians and the settlers, and the special royal privileges they enjoyed, exempting them partially or entirely from payment of customs dutie s on the goods they 121). Colonists complained to the Portuguese Crown that by participating in the spice trade the missionaries were abandoning their spiritual duties and taking away workers that th ey, the colonists, needed (Alden 1976 ). To address the issue of missionary competition with colonists and the labor shortage that inhibited the production of cacao and other collected products, the Portuguese Crown enacted directorate legislation on the administration of Indians (1757 179 bodied, nonelite men were employed in either royal service, communal agriculture, service to private parties, or oller 2010: 4 36). Beyond labor regulation, R oller pinpoints the significance of this legislation for the colonial economy: serto: to establish a set of standard procedures by which collecting canoes would be dispatched ann ually from each Indian village, to regulate participation in the expeditions and compensation of those involved, and to curtail contraband trade and illegal labor practices, both of which had run rampant during the missionary era. Around the same time (17 55), the Crown established a royal monopoly trading company, the Companhia Geral do Gro Par e Maranho with the goal of promoting economic


42 development in the Amazon region through regular transatlantic shipping and the supply of African slav es for agric ultural enterprises (p.438). Once the Directorate legislation was passed, collecting expeditions became more formalized and cacao exports expanded. In the Rio Negro Captaincy, about 3 0 men were employed per village and directorate villages sent an average of 250,000 pounds of cacao beans to Lisbon every year between 1772 1788. By the end of the 18 th century, the massive quantities of cacao beans collected in Amazonia helped Brazil become the second largest New World exporter of cacao ( Roller 2010 : 439 ) T he Madeira valley remained an important focus of cacao collection moving into the 19 th century, but conflicts continued between colonists in Borba and the Mura. In 1833, after an invasion by the Mura, Borba lost its designation as vila Two years later, the Cabanagem Revolt began in which detribalized Indians ( tapuios ) and escaped slaves united in rebellion against Portuguese settlements in the region. During this time, Borba was one of the few settlements that resisted the cabano rebels and eventually r egained its status o f vila although not without a significant loss in population. ts in 1833 to 1,075 in 1849 (Harris 2010: 281). In 1852, U.S. Naval Officers William Lewis Herndon and Lardner Gibbon passed through Borba while conducting a survey of the Amazon region. They described Borba as a small town which they estimated as having 300 inhabitants 7 ; the majority of the popula half of which were slaves (Herndon & Gibbon 7 This figure should be approached skeptically as Harris (2010) cites a population of more than a thousand inhabitants in 1849 and Bastos (1866) claims a population of 2,335 inhabitants in the district in 1864.


43 185 4: 311). In terms of agricultural production, sugarcane was produced in farms of the area, principally for manufacture of rum (cachaa) while oranges, limes, and watermelons (all exotic crops) were cultivated for local consumption (Herndon & Gibbon 1854: 312). The tobacco produced in the area was claimed to be the best in Brazil, and was traded to the Atlantic Coast along with cacao, sarsaparilla, coffee, and Brazil nuts (Herndon & Gibbon 1854: 311). Despite the portrayal of Borba as a minor settlement, in the year 1862 it was t he leading exporter in Amazonas, having even greater revenue than Manaus (Loureiro 2007: 237). The economic potential of the lower Madeira is noted by several visitors in the 19th century (e.g. roteiro 1853), mentioning the abunda nce of cacao and rubber trees fou nd in the valley : Above Borba, which is said to have formerly produced good tobacco, there are some cacao plantations, whose fruit at the time of our passing there (in June) were almost ripe and of a bright yellow colour; and about this part of the river also the first high trunks of the caoutchouc tree are seen the Siphonia elastica or Seringa as it is called here. On the Amazon and Lower Madeira these valuable plants are almost destroyed by continuous withdrawal of t heir milky sap. The huts of some caoutchouc gatherers (Seringueiros) are seen now and then low roofs of palm leaves, beneath one end of which there is a raised floor or framework of lath, one or two yards from the ground, to which the inhabitants retire at high water, when necessity obliges them to lead almost an amphibious life. (Keller 1875: 45) As alluded here, rubber tapping also became a notable economic activity in the region during this period. When the price of rubber rose in the mid 19 th century more and more individuals were attracted to the Madeira River since rubber trees occurred in relative concentration along its banks (Weinstein 1983: 53). The demand for rubber led to an eventual repopulation of the region following the devastation of mos t of the Columbian indigenous populations that were largely concentrated in agricultural villages, the Madeira during


44 ling middlemen. The Rubber Boom Between 1850 and 1880, Borba lost and gained its status as vila numerous times until finally in 1888, it was e stablished definitively as a municipality. This period oom when the Madeira River v alley was flooded by immigrants mostly from the Brazilian Northeast who were seeking out the quality rubber ( Hevea brasiliensis ) found in the region. Maria Mitouso de Melo, one time resident of Borba and mother of the acclaimed Amazonian poet Thiago de Me llo, tells a story similar to many of those migrants arriving to the Amazon from the Brazilian northeast during the Rubber B oom with the hope of making themselves rich. On t he first page of everal houses that Ceara where we resided. With the news that in Amazonas, rubber was making lots of money, de Melo 1 983: 9; translation mine). During the Rubber Boom, the vast majority of the rubber that was tapped along the Madeira came from wild stands of rubber trees or seringais Typically, the extraction of rubber occurred under a patro n client system (also known as aviamento ) in which an owner of a rubber stand would provide supplies on credit for rubber tappers who were granted access to tap trees that were usually connected by a network of paths in the seringal Under this system, rubber tappers had to agree to deliver all their rubber to the patron and buy their supplies from him (Dean 1987: 40). This relationship was highly exploitative since the patron could manipulate the price of goods given on credit,


45 yet it did afford rubber tappers a certain degree of f reedom as the work was seasonal and they could work at their own pace (ibid.). While the boom brought migrants from other reaches of Brazil to Amazonia, it also absorbed the local labor force, which led to a sharp decline in agricultural production. The se arch for rubber also led to a penetration of more isolated areas of the region, resulting in conflicts with indigenous groups and in some cases forced indigenous labor. Another consequence of this development was further disruption of indigenous agricult ural systems including homegarden management and production of regional staples to the point that food prices shot up in many regional centers (Miller et al. 2006 : 48 ). While the Rubber Boom disrupted regional agricultural production, the arrival of Northe asterns to the Amazona also led to the introduction of some new species of plants along with distinct varieties of already existing species. Maria Mitouso de Melo casually mentions in passing a number of plants in her yard in Borba that had been brought f rom the Northeastern state of Cear by her landlord, Coronel Vitor Coutinho: whose branches had interwoven and taken the shape of a chalice. When you pulled one of its branches, it would rock. On the same side, there was a mussambe There was also a teeth and whiten them. (de Melo 1983: 100; transl ation mine). In this manner, the Rubber Boom had important implications for agriculture and agrobiodiversity in the region as it brought new species and varietie s of plants with arriving migrants while also disrupting local patterns or systems of agricult ural production. Eventually rubber prices crashed, leading to a diversification of the Amazonian economy beginning in the 1920s (Schmink and Wood 1992: 46 50) The


46 mark left by the Rubber Boom, however, would remain, and the effects it had upon economy, society, and environment continue to reverberate up to the present. During World War II, the Allied Forces experienced a rubber shortage as the 8 and had come und er control of the Japanese. During that period, the demand for rubber in Brazil peaked once again and a second wave of migrants swept through the Amazon region. The migrants, who were typically men from the Brazilian Northeast, became known as the Soldado s d e Borracha A large number of migrants moved to the Lower and Middle timer in Borba named Nini, had put it. Several individuals that I met and interviewed mentioned that the ir father or Borba traced their roots back to this period. In addition to rubber extracted from trees of the genus Hevea other forms of latex Madeira River and its tributaries. Balata, known in Brazil as maaranduba, was the name given in the British market to latex derived from several species of the genus Manilkara (Dean 1987: 38). It y ielded durable and less elastic latex that was used for machi however, was also applied more generally by 8 Gardens in 1876, which were then propagated and planted in Malaysia. Unlike Amazonia, where leaf blight prevented the establishment of a plantation system, Southeast Asia proved to be an excellent region to plant rubber trees in high densities. This development helped lead to the collapse of the Amazonian Rubber Boom, as Malaysian rubber severely undercut the price of Amazonian rubber (see Dean 1987).


47 locals, and one old timer named Seu Hernando told me that the latex of several different trees (garrote, parajuba, rosa da, and sorva) were mixed together to make balata. Other trees including coquirana ( Ragala spp.) were also tapped for their latex, according to Seu Hernando (Audio recordings, February 2010) Oral histories shed much light on this period when balata and s orva were widely collected in the region. In my conversations with Dona Clia an 81 year old woman from the community of Puruzinho, I learned that her husband and other men from the area made trips into the centro, or the deep isolated reaches of the for est, in the months of December and January and would spend up to 4 or 5 months collecting latex. The men (I heard no stories of women collecting balata) would carry basic provisions, which consisted mostly of manioc flour and munitions for hunting game for their subsistence. Seu Laudelino, another Puruzinho resident, said on one trip he and his crew lost their manioc flour when their canoe capsized and they were forced to rely strictly on hard, dried meat. For most Amazonians, a meal without manioc flour is not a meal. As Dona Clia Fome de farinha a pior coisa que existe! ( Hunger for ). Despite their suffering, men who worked with latex often recall the era with a bitte rsweet nostalgia. They were tough times, but times they were proud to have lived through. Men paddled to far reaches of the municipality that today in an era of motorboats still seem distant, far removed, isolated. T hey went up the Sucunduri, Igap Au, Tupana, Acari, and Abacaxi Rivers, which all ran far from the main artery of serves as the center of commerce and activity, the centro in rural Amazonia lies deep in


48 the mata or the f orest (see also Raffles 1999 : 222 223 ). Stories of hardship and near scrapes with death are often recounted from this period, and many legends and oral traditions are told in the context of this time when men would get lost in the woods, en countering Curupira, Juma, and other creatures of Amazonian legend. As was the case during the Rubber Boom, many men collected balata and sorva under a patron client or aviamento system. Before men would leave for the centro, their patro would provide them with provisions for their trip and for their families during the period they would be absent from the home. On return from their trips, men would then remain in debt. Those who were able to collect beyond what they owed would then be paid the difference by the patro. In some cases, competition among men collecting latex could be fierce. One friend, Sapo brother h ad told him about his days working with balata. The story illustrates the perils, both physical and psychological, real and imagined, of balata extraction at the time: passed away. He had worked with balata in Borba and had reached the point where he was pretty well off. He was worried, however, that the men working with him were jealous of his success. One night he dreamt of a vulture picking away at his bones and when he woke th e next day, he was afraid that they would come to kill him. He decided to take off for Mapi (a river east of Borba), but just as he suspected, a group of men that worked with balata followed him out of town and surrounded him. He was worried that his prem onition was going to come true and that they would kill him, but the men said that since they had been his friends, he could escape with his life. Then t hey demanded that he leave town and all of his possessions behind. So he took off into the woods of Map i and spent six months living off fruits and other foods he could find in the forest. Every day he prayed and rec several months in the forest, he realized that he was being accompanied by a small man all in white who told him that he was entering hell. There he found the souls of people who he had known in Borba, but who were still alive. He found his sister, Zuza, on all fours with a dog in her mouth. The


49 man in white said she had sinned by aborting unwanted children. And then time to die, but that he had to repent for his sins (as he had had relations ave to return to his body on Earth, which then appeared before him like an animal and it disgusted him. Shortly thereafter, he said he was found on the banks of a small river and brought back to the city and was given clothes. Only then did he realize he w as completely naked. ( Sapo Puruzinho, Nov. 1 st 2009 Field Notebook#2; Translation mine ) Stories of treachery and back stabbing among latex collectors and rubber tappers are rimes committed against rubber tappers and extractivists during the period often came from the orders of local patrons. Some residents of Borba recall stories of their parents and grandparents who claimed that when individuals had been too successful at r epaying their patro and were able to profit off rubber or balata extraction, they would be threatened, or in other cases, murdered. Seu Laudelino said that although his patr o treated him fairly, in Nova Aripuan (the next municipality upriver from Borba ) many his patr o March 2010) Horror stories of the exploitative nature of the aviamento system appear in other areas of the B razilian Amazon (e.g. Stanfield 1998). Death threats were not uncommon the seringal because t he patr o trans. mine). De Melo goes on to write that, That was how many rubber tappers and workers died in the distant seringais of our Amazonas! They came from the Northeast full of hope and


50 illusion, thinking about making lots of money, but what waited them was death, either from malaria or the crimes committed by their patrons. How they suffered, these p oor Brazilians that came to brave our forests in search of rubber! And how much evil was in the hearts of some men of this time, and h ow many crimes went unpunished! (de Melo 1983: 53 ; translation mine ). However, many of the older residents of communities in Borba said that local patrons o n the Madeira River paid fairly, River between the 1920s and 50s. Manoel Saraiv a, from the community of Caiara in the municipality of Borba, was the main patron in the area during the time. Many interviewees spoke highly of the man, as a good and decent patro. Seu Zizi, Seu Laudelino and other residents of Puruzinho, all agreed t hat Manoel Saraiva paid them when they returned after their months of labor in the centro Their balances were never denied and they said that despite the hard work, the money was usually enough to buy several months of provisions for their families when t hey returned home. In the off extractive and agricultural activities. Some men would cut firewood for the steamships barraco (general st ore or depot), before it chugged upstream to the municipal seat of Borba. Many men also helped at home widely. For household consumption, some fruits and vegetable s were planted to complement the household diet including watermelon, corn, and beans. In addition, Brazil nuts and andiroba nuts ( Carapa guianensis ) were collected by many families, and these were often sold to the local patrons.


51 In the 1960s and 70s, a market left for such products. The decreased dependence on extractivism resulted in a concomitant increase in agricultural production in much of Central Amazonia (Fraser et al. 2009 : 232 233 ) With this transition to agriculture, one product in particular began to appear in the floodplains of the Amazon and the Madeira: jute. Jute Agriculture on the Madeira River Japanese entrepreneurs brought jute ( Conchorus capsularis ) a species native to Asia, to the Amazon in the late 1930s as it was used to make sacks for coffee produced in the south of Brazil by Japanese immigrants. Most of the jute production began on the floodplains of the lower Amazon, but later spread to other major tributaries in cluding the residents recalled that Japanese middlemen taught community members how to plant the crop and shortly thereafter, they began producing jute in the community in the 1940s and 50s. Two older residents, Seu Zizi and Dona Clia both worked with jute on the floodplain and recall the time clearly. Jute agriculture requires intense labor and since jute is produced in inundated areas, Seu Zizi and Dona Clia said tha t they spent much of the day standing in water. Harvesting jute was particularly demanding as WinklerPrins (2006) describes here: To assure some level of marketable fiber quality, jute needed to be ering and fruiting (which are separated by only 1 to while the flood was just beyond its peak, in June or July. Harvesting involved arduous physical labor as the plant had to be cut (using a scythe) while laborers sto od up to their waists in water. This exposed workers to stingrays, piranhas, and other dangerous river fauna. It has also led to a legacy of arthritis in the knees, limiting the mobility of aged jute cultivators (p 827)


52 To add to the many perils of worki ng with jute described above surpr ises from anacondas wa s a common theme in stories told to me by form er jute workers in Borba since the massive snakes are commonly attracted to wallows Despite the intense labor and occasional scares from anacondas, jute was seen stron g and then eventually tailed off (Fieldnotes, June 2010). In t he 1970s, when Brazil began to maximize jute production, world demand for the product fell as jute sacks became replaced by other products used for the transport of agricultural commodities communities that plant jute. Instead, cacao remains one of major agricultural product s as well as a ai, a newcomer to the agricultural export sector. With c hanging trends in markets, agricultural production in Borba and the lower Madeira continues to shif t and as a result, the patterns of biodiversity found on Borb to evolve. Conclusions In this brief historical sketch of the history of Borba and the lower Madeira River, ning of Amazonia to global economic markets had radical effects on the region and its agrobiodiversity. However, it is also clear that contemporary agricultural landscapes and agrobiodiversity continue to exhibit distinct patterns inherited from managemen t by indigenous groups during the pre contact period. With the Rubber Boom and the later arrival of the Rubber Soldiers, agricultural production dropped in the region as the economic focus shifted to the extraction of rubber and latex, and to lesser exten t, Brazil nuts and non timber forest products. Following the period of rubber extraction with the introduction of jute,


53 agriculture became as an important area of economic development once again in the municipality of Borba. Today, as both Brazilian sta te and federal governments are trying to promote growth in agricultural production in Amazonia, greater attention is being drawn to how smallholder agricultural communities manage agricultural lands and the biodiversity associated with those lands. Althou gh Borba is only one small municipality in the expansive Amazonian territory, it serves as an important case for illustrating the manner in which the process of globalization beginning in colonial period had pervasive effects on the Amazonian landscape and its management, and continues up to the present. Taking into consideration the agricultural history of the region outlined here the following ch apter examines the factors that shape the diversity of plant species managed by households in contemporary sm allholder communities in Borba.


54 Figure 2 1 Archaeologist Claide Moraes standing above the defensive ditch at Vila Gomes (August 2010) Figure 2 2 Claide Moraes standing in the trough of th e defensive ditch at Vila Gomes (August 2010)


55 Figure 2 3 Ex cavation at Vila Gomes in terra preta soils (>1 meter depth) with transitional Oxisols in the lower 10 cm (August 2010) Figure 2 4 Brazil nut trees surrounding a stretch of terra preta that was converted to cattle pasture (Vila Gomes ; August 2010 )


56 CHAPT ER 3 CONTEMPORARY AGROBIO DIVERSITY AT THE HOU SEHOLD LEVEL Introduction This chapter explores the diversity of plants managed in contemporary smallholder communities in Borba. Here I present findings from botanical and household surveys regarding the diver sity of useful plant s managed in 17 communities in the municipality, sampled multiple regression analyses demonstrate that the and ( floodplain or highlands) are statistically significant determinants of spe cies richness while h ousehold wealth and indicators of market oriented farming demonstrate no significant bearing. The three statistically significant variables identified in this analysis account for 31. 5 % of variation in species richness among households, which leaves room for many other potentially influential factors. I draw on ethnographic data to discuss these results and point towards other influences on contemporary patterning of plant diversity, including the life histories of farmers and the cultivation of magic pl ants in the household Agriculture and Agrobiodiversity in Smallholder Communities In Borba today, the majority of households in rural communities rely on hunting, fishing, and farming as their primary forms of economic production as is the case in much of rural Amazonia (Hiraoka 1992; WinklerPrins 2002 ). Very few households rely tapping), however many individuals do manage extractive forest resources including


57 Brazil nut puxuri ( Licaria puchury major ), andiroba, and rubber. In the most communities, a few enterpreuneurial households also sell consumer goods including industrialized food products, fuel, tobacco, and alcohol. For many households, though not all, agricultu re is the principal economic activity. Manioc ( Manihot esculenta ) is the primary staple crop in most of rural Brazilian Amazonia. Manioc is cultivated on both the floodplain and the uplan d s, although different varieties of manioc are planted in these diff erent environments. In the floodplain, fast growing varieties of manioc are planted when the river recedes at the onset of the dry season. In the uplands, slow growing varieties, which often yield larger tubers, are planted, usually after maturing anywher e between 8 and 18 months. On both the floodplain and the uplands farmers cultiva te both bitter and sweet forms of manioc (see Chapter 4 for further discussion). Aside from manioc, rural households manage a wide range of other crops. The more fertile soils of the floodplain are exploited for production of d emanding annual crops like corn, cow peas ( Vigna unguiculata ), and watermelon. Also in the uplands, a number of households take advantage of the heightened fertility of terra preta do ndio or Amazo nian Dark Earth. These areas of anthropogenic soil are often used for the production of market crops including watermelon, West Indian Gherkin ( Cucumis anguria ), papaya and cacao, but are also used for production of subsistence crops like manioc, corn, a nd beans. In both the floodplain and the uplands, most households manage large homegardens and perennial fruit orchards ( Figure 3 1 ) Aai ( Euterpe spp.), cacao, soursop ( Annona muricata ), citrus, cashew ( Anacardium occidentale ), guava, cupuassu ( Theobro ma grandiflorum ), mango and banana are all commonly


58 found in the gardens and orchards of most households. However, older and more established households usually invest more in the production of perennial crops while younger households or those they have only been recently established focus production on annual crops (Perz 2001: 163 164 ; Walker and Homma 1996). Annual crops provide a quicker turn around of profit, but also require more investments in labor and the clearing of fields. Perennial crops, on the other hand, take longer to produce, but are usually more valued in the markets and can provide more long term benefit. Research shows that the distribution of useful plant species tends to be uneven, both among communities and within them (Coomes and B an 2004; P adoch and de Jong 1991; Perrault Archambault and Coomes 2008). participation in market economies is often considered to have a negative influence on agrobiodiversity of smallholder farms, but empirical studies have demonstrated mixed results when examining the relationship between market orientation and species richness of useful plants found on farm ( Major et al. 2005; Vadez et al. 2004; Zimmerer 1996: 24 25 ) Studies in the Peruvian Amazon suggest that household demographic and economic variables have important bearing on crop diversity including land wealth and number of available laborers (Coomes and Ban 2004; Perrault Archambault and Coomes 2008), but further research is necessary to measure the influence of t hese variables in other areas of the Amazon region Drawing on socio demographic, economic, and environmental data collected from rural communities in the municipality of Borba, this chapter identifies the household level variables that have the most signi ficant bearing on plant species richness, and the degree of variation among households that these independent variables explain. Here,


59 a model is presented based on analyses of the following variables: age of household head, number of individuals in house hold, number of adults in household, leng th of residence on property, area of land under cultivation and environmental habitat (upland or floodplain). Additional economic analyses are presented, testing the relationship between species richness and house hold wealth as well as the influence of agricultural management practices associated with market oriented farming (e.g use of agro chemicals ). Methods I collected data over a period of 12 months visiting 138 farms in seventeen different communities locate d in the municipality of Borba. At each household, I invited the household head(s) to list the number of plant species under management in his/her homegardens, orchards, and swiddens. I also collected attribute data of individual plants including approxi mate area of cultivation, use(s) of the plant, and duration (i.e. perennial or annual). I employed structured questionnaires to elicit information from household heads regarding economic, demographic, and environmental variables that were recognized by pr evious studies as influencing useful plant species diversity. Specifically, I inquired about the age of the household head(s), size of household, number of adults in the household, length of residence at the property and sources of income for the househo ld I also collected data on the general environment on which the property was located (floodplain or upland environment), the soil orders found on the property, the number of cultivated fi elds belonging to the household, and the area of land under managem ent Collecting accurate economic data proved difficult as many households did not


60 production figures. To address this problem, I developed a Guttman scale to rank household wealth based on salient consumer items identified during initial visits and interviews (Guest 2000; Bernard 2006: 321 327). At 74 household s I asked the household head(s) about the presence or absence of the following items: 5 hp motor ( rabeta ), canoe, television, DVD player, satellite dish, fan, bed, sofa, freezer, refrigerator, food processor, radio, oven, stereo, generator, and cell phone. These items represented a range of goods from ones that were fairly ubiquitous (e.g. canoe, 5 hp motor) to other s that were only found in a few wealthier households (e.g. sofa, freezer/refrigerator). To minimize errors, the unidimensionable scale was reduced to 7 items: canoe, television, DVD player, food processor, freezer/refrigerator, generator, and sofa. This scale produced a high coefficient of reproducibility 9 of .936 and high Because of difficulties in collecting consistent economic data, information regarding agricultural management techniques and use of agricultural in puts was also collected in an attempt to distinguish households with more market oriented profiles from those that were less market oriented All ho usehold heads were asked about use of herbicides, pesticides, chemical fertilizers, and organic fertilizers following the assumption that more market oriented household s would rely on chemical inputs. In addition, household heads were asked if they purchased or saved seeds since market oriented households are more likely to purchase seeds while those that ar e less market oriented typically save seeds. 9 Calcu lated as: [1 (# errors / # entries)] (Guest 2000).


61 With some household heads, I collected life histories to gather qualitative information that helped provide a more nuanced vision of agrobiodiversity management over time. In collecting these ethnographic data, I sought to identify distinct themes in the life histories of individual farmers that appeared to have important bearing on their occupational histories, amount of time spent in wage labor, migration, family origin, and children. Along with the life history of the farmer, I gathered information on the history of the farm itself and used this information to discuss the development of div ersity at the household level. Pe was used to test the relationship between the Guttman scale of household wealth (based on their score of household goods present) and the total number of species managed by the household. I ndependent t tests were conducted to compare s pecies richnes s for households that used chemical fertilizer, herbicide, and pesticide with those that did not The same was done to test differences between households that used organic fertilizers and saved seeds and those that did not. M ultiple reg ress ion analysis was conducted to test the significance of the following independent variables as predictors of household species richness: age of household head, number of household members, length of residence on the property, area of land under management, and environment (floodplain =0 highlands =1 ). Data for the Results The age of the household heads interv iewed ranged from 18 to 90 years old with a mean age of 43 years old Households ranged in s ize from 1 to 16 members with an


62 average of 5.3 members per household The number of years that household heads occupied their farms varied from 1 to 77 years with a mean of 23 years and a median of 18 years. In terms of agricultural management, the mean area of land actively managed by households was 2.5 hectares ( Table 3 1 ). From the 138 households surveyed, 185 different useful plant species were identified fro m 54 botanical families ( Table 3 2). The mean number of species per household was 19.5 with a standard deviation of 8.3 species, reflecting high variation in species richness among households ( Figure 3 2). The minimum number of species managed by a house hold was 2 and the maximum was 48. The most commonly cultivated (or managed) species were manioc (90.6% of households), aai (83.3%), cacao (69.6%), banana (68.8%), guava (63.8%), mango (59.4%), cupuassu (59.4%), lime (58.0%), malay apple ( Eugenia malacce nsis ; 41.3%), chives ( Allium schoenoprasum ; 46.4%), and cashew (41.3%). A large proportion of species were poorly distributed with 28.1% of species ( 52 of the 18 5) occurring at only one household. 61.6% (114 of 185) of the useful plant species found in this study are native to the Neotropics and 32.4% (60 of 185) are considered native specifically to the Amazon region. However, there were a significant number of Old World exotics (62 species or 33.5%) and a few of unknown origin (9 species). Curiously, of the 10 most cultivated species, 5 are of Old World origin (banana, mango, lime, malay apple, chives), showing that introduced species make a significant contribution to rural Amazonian livelihoods and subsistence. Also of the 52 species that occur at only one household, 33 are natives, suggesting t hat many useful endemic species are poorly distributed


63 O f the 138 households surveyed, 60 were situated on the uplands and 78 were located on the floodplain. Upland ho useholds managed a mean of 22.0 ( 8.23 ) species while floodpla in households had a mean of 17.5 species ( 7.85 ), which was significantly different (t=3 244, p=.002 ; equal variances not assumed). Among upland households, 20 were situated on Amazonian Dark Earths (i.e. anthropogenic soils) and 40 were located on non anthropogenic soils. Farmers managing plants on Amazonian Dark Earth had a slightly higher mean number of species (23.0 5.17) than those on non anthropogenic soils (21.6 9.43), but with no significant difference (t=.729, p=.469; equa l variances not assumed ). Market oriented agricultural management practices appeared to have no relationship to the species richness managed by the household The 46 households that reported to use chemical fertilizers managed a mean of 19.33 8.57 species while the 89 households that did not use chemical fertilizers managed a mean of 19.46 8.06 with no significant difference between the means (t= 059, p=.953 ; equal variances not assumed) 10 The 31 households that used herbicide had a mean of 19.55 8. 52 w hile the 105 households that did not not use herbicides had a mean of 19.42 8. 16 species, once again with no significant difference (t= .075; p=.941 ). A siimilar pattern was found regarding the use of pesticides. Th e 76 households that reported to use pe sticide managed a mean of 19.50 8.23 species and the 60 households not did not use pesticide had a mean of 19.38 8.26 species with no significant difference (t= .082 p=.935; equal variances not assumed). 10 Data were missing for use of chemical fertilizers at 3 households, and at 2 households data were missing regarding use of herbicides, pesticides, and organic fertilizer. Lastly, data on seed saving we re missing at 5 households.


64 These trends were also found in analyses of househo lds that used organic fertilizers (i.e. animal manure) and saved seeds, characteristics of less market oriented households. The 56 households that reported to use organic fertilizer managed a slightly higher mean of species (20.13 8.04) than the 80 househ olds that did not use organic fertilizer (mean of 18.98 8.35), but with no significant difference between the means (t=.808, p=.421). The 86 h ouseholds that saved seed (mean of 19.34 8.19) also showed no significant differences from 57 households that did not save seed (mean of 19.74 8.49 ) ( t= .268; p=.789). Lastly, no significant relationship was found between species richness and household wealth based on the Guttman scale (r= .077, p=.51 3). To predict plant species richness across all households, a m odel was developed initially selecting six independent variables: age of household head, number of household members, number of adults in the household, length of residence on the property, area of land under management, and environment al habitat (floodpla in, highlands). Age of household head ( Figure 3 3 ), area of land under management ( Figure 3 4 ), and environmental habitat were all significant (p=.004, p=.000, and p=.007 respectively). However, the independent variable s and which produced a negative coefficient. These three variables were excluded from the household and


65 produced an r square value of .315 thus accounting for 31.5% of variation in species richness observed for 134 11 cases of the study ( Table 3 3). Discussion of variables can influence the number of plant species managed by a household. Analyses from this study identified three significant variables that explained a considerable portion of the variation in plant species richness man aged on rural Amazonian farm s: age of household head, area of land under cultivation, and environment. Household wealth and agricultural management strategies typical of market oriented households on the other hand, did not show any clear bearing on species richness. The si gnifica nce of these findings is discussed below. Age of Household Head These findings show that age of the household head is a significant variable, but have argued, acquisi tion of agricultural knowledge and useful plants is a lifelong project for many farmers. Many families in rural Amazonia are highly mobile, and individuals may change residence numerous times in their life, but they often carry with them plants and seeds Archambault and Coomes 2008). Also, as individuals grow older, they develop a larger network of friends and family from which they can acquire plants and planting materials. From the ethnogra phic record, numerous anecdotes serve to support the above points. Outside of Borba, a government settlement ( assentamento ) was established in 11 Complete data were missing from 4 of the 138 households studied.


66 the late 1990s to provide land for families in need. Despite having been established for a relatively short peri od of time compared to the majority of other farming communities in the municipality, many of these households were botanically diverse as farmers transferred seedlings they had taken from their former homes or from the farms of friends and family where t hey once lived. This ethnographic fact became apparent during my visits to the assentamento with members of the local agricultural extension agency. For example, Pedro, a newcomer to the settlement, proudly showed me a growing avocado tree that he had p ropagated from an impressive specimen that was Hildebrando, who was a prominent figure in the settlement, he informed me that Brazil nut trees he cultivated on his property wer e taken as seedlings from the community where he had grown up. And while interviewing Luiz, who hailed from Rio de Janeiro and resettled in Amazonas, he pointed out to me a tamarind tree that he brought all the way from Southeastern Brazil. Settlers thus seemed to cobble together a variety of useful plants through their ties to former communities and friends with whom they had built relationships over the course of their lives. As individuals grow older they not only have the potential to accumulate a gr eater diversity of plants, but they also build networks that allow them greater access to a diversity of plants. The reason why age is a significant predictor of species richness may also be related to differences in household strategies of younger and old er household heads (Perz 2000; Walker and Homma 1996). It can be argued that younger household heads have a tendency to focus their production on a limited number of market crops in order to support their young families while older household heads typical ly have greater


67 Following Chayanov (1966) and mor e recent research drawing from his work patterns of a household, but also the diversity of useful species managed by the household (Barbieri et al. 2005 ; see also Perz 2001 ). Area of Cultivated Land The second significant variable, total area of cultivated land, is positively correlated with sp ecies richness because more available land allows for the planting of a greater diversity of species. This probably also relates to the household lifecycle as the area of land managed by the household expands from the early stages into more mature stages ( Chayanov 1966). Although the number of adults in the household was not found to be a significant predictor in this study, this is might be a reflection of a large number of households with elderly adult members who did not participate in agricultural act the area of cultivated land is a significant predictor since many households that manage large areas of land have diverse agroforestry systems, which allow for accumulation not only of species that they have actively c ultivated, but also other wild and incipient domesticates that are found as volunteers on their agricultural lands, which can be managed passively (i.e. not cut down). Environment al Habitat : The Uplands and Floodplain From these data uplands ( terra firme ) or floodplain ( vrzea ) has an important influence on the diversity of plants managed by the household. The floodplain is an inherently risky area to manage crops because flooding during th e rainy season is highly unpredictable (Chibnik 1994; Denevan 1984: 320; Moran 1993: 97; Padoch 1999: 348; Padoch and de Jong 1992:


68 163; WinklerPrins 2002). Shortly after I arrived in 2009 to begin research, I learned that communities on the Lower Madeira River had suffered from the highest flood on record (the previous record holding year was 1953). Many families lost large cacao orchards and acai groves. The majority of floodplain households lost a significant portion, if not all, of their homegardens. Dona Nanda a long time resident of a floodplain community, lamented that she no longer had the desire to rebuild her garden into what it once was 2009). Dona Nanda was n ot alone in her frustration, as numerous families interviewed on the floodplains complained of the trees and plants they lost, including families that lost cacao orchards that they had established through a financed agricultural project sponsored by CEPLAC the Brazilian Cacao Production Planning Commission. Despite their losses, they were expected to pay back the loans they had received to establish their cacao plantations. In response to the floods, the government of Amazonas state provided each affecte d household with R$400 in emergency aid to help overcome the initial recovery. Later, loans for R$2000 were also made available, although many of the families that sought out such support were denied because of problems related to credit or official ident ification documentation. While no one had anticipated the extent of flooding in 2009, families living on the floodplain are accustomed to environmental unpredictably, and for that reason they typically farm in a manner that manages risk by often exploiting more stable environments outside the floodplain (Padoch and de Jong 1996; WinklerPrins 2002) 80.8% (63 of 78) of floodplain households that were interviewed in this study reported to also farm in areas of the uplands especially for the cultivation of ma nioc, which usually


69 requires a full year or more to mature. Some floodplain households also managed s tios or perennial orchards on the uplands. In some cases these orchards were composed of a diversity of fruit bearing trees for subsistence while in a few cases the orchards were used for commercial production of specific market crops like cacao. Despite its risks, the floodplain s of whi tewater rivers like the Madeira remain some of the most fertile environments for agriculture in Amazonia, particularly on natural levees ( Denevan 1984; Moran 193: 97; Padoch 1999; Sternberg 1975 ). Because of the the uplands occasionally cultivate areas of nearby floodplain, or in some communities that I visited they plant ed annual cr ops on the edges of receding floodplain lakes. In the community of Puruzinho where watermelon was commonly planted on anthropogenic soils in the months of March and April, several households also planted watermelon on the edges of the floodplain lake in t he months rich soils. Rural smallholders in Amazonia exploit different environments and ecotones for agriculture, but in general upland households benefit from much greater sta bility and as a result they tend to manage a greater diversity of useful plant species. Yet even while households farming on the uplands benefit from stability, they also have to contend with other obstacles including problems with soil fertility. In con trast to the floodplains that are quite fertile environments for agriculture, the dominant soil orders of the uplands are Oxisols and Ultisols, which are generally very acidic and exhibit high concentrations of aluminium and iron oxides that inhibit the gr owth of many crops (Moran 1993: 66 ). In communities that have stretches of Alfisols (known locally as terra roxa ) or anthropogenic soils, these are usually exploited due to their comparative advantages for


70 the production of many crops like corn, beans, We st Indian gherkin, and watermelon. Aside from limitations imposed by soils, households farming in the uplands also suffer from their dependence on rain fed irrigation. In the rainy season, high precipitation can have negative consequences for agricultura l production, particularly for root and tuber crops, that may rot before they are harvested. In the dry season, droughts usually have even more profound consequences. In addition to being a year of extreme flooding that affected floodplain communities, 2 009 was also described by farmers as one of extreme drought on the uplands. Numerous farmers reported having lost entire manioc fields because of the lack of precipitation More than a month had passed without a drop of rain and the orange clays of the u plands cracked from the lack of moisture. The impact resistant to the physical demands of the Amazonian environment, to simply die. Families on the floodplains who had manioc fields in the uplands suffered doubly. As 2010). While differences in the uplands and floodplains have important consequences for the patterning of agrobiodiversity, risk is inherent to farming in either environment. Household Wealth Market Orientation, and the Barge Economy Variables or the use of chemical inputs like herbicides, chemical fertilizers, and pesticides appear to have no bearing on the number of species managed at the household leve The same was found for agricultural management strategies associate d with less market oriented farmers including use of organic fertilizer and seed saving. There are several reasons why this may be the case. First, the quan tity of chemical inputs used by the household and the


71 frequency of their usage were not recorded. Use of chemical inputs is often associated with market oriented farming, but households that focus largely on subsistence production also occasionally use chemical fertilizer to enhance production. To distinguish households that are highly market oriente d from those that are not, more specific data regarding usage of agro chemicals may be necessary. Secondly, practices such as seed saving and use of organic fertilizers can be agricultural strategies of both subsistence farmers and those more oriented tow ard the market. Lastly, it co uld be argued that there is no significant variation in the market orientation of households. As mentioned above, most households enga ge in a form of agriculture that to find rural households that produce strictly for the market. Household wealth, based on the Guttman scale developed for this study also had no significant relationship with the species richness maintained by the household. more oriented toward the market and focus on production of only a few cash crops, yet income as is commonly the case in industrialized agriculture. Furthermore, rural household incomes are not directly dependent on agricultural production since hunting, fishing, wage labor, government loans, and the sale of industrialized products (and illicit drugs) can also make significant contributions to household income and wealth. Lastly, the smallholding farming communities stu died, there is no


72 discernible relationship between household wealth and the diversity of plant species managed by the household Another important factor potentially influencing the relationship between household wealth and species richness is the new rive r barge economy on the Madeira River. Every day massive river barges pass through Borba as they transport soy and industrial goods between Manaus Itacoatiara and Porto Velho, the capital of the neighboring state of Rondnia. W ith this traffic, new econo mic opportunities have arisen for near the river sell fresh fish and agricultural produce, game meat, wild honey, and sometimes even illict drugs to the barge operators, known as balseiros The balseiros occasionally purchase these products, but also frequently trade diesel or sometimes industrialized food products (like frozen chickens) in exchange for the fresh foodstuffs. Diesel is a particularly valuable commodity b ecause it provides energy for community generators, allowing rural communities to watch television in the evening, and for those households that have freezers, to preserve fish and have cold drinks. In other words, barges represent new market opportunitie s that come to the communities, creating different market arrangements than were seen in the past. Before the arrival of the barge economy, households either sold goods to a local patro or middleman, or had to travel to the city to negotiate the sale of their p roducts, which brought on extra costs of transport and vulnerability in negotiation if prices were lower than expected. Furthermore, producers would often have to focus on the production of one or two agricultural products sold in large quantity. With the appearance of barges, however, households have the opportunity to sell or trade small quantities of a wider variety of


73 products without having to travel far from home. Delicate fruits like graviola ripen quickly and become easily damaged in trans it to the city, which complicates the small scale market production of these fruits. Today, however, rural households can pick the fruit from their homegarden and sell it directly to a balseiro In terms of agrobiodiversity, this means that households ca n maintain a wide diversity of plants and still capitalize upon them in the barge economy. This also allows for accumulation of household wealth and potential purchase of consumer goods without necessarily compromising the diversity of species maintained by the household. Oral Histories and Magic Plants In addition to the influences identified from household surveys, oral histories can also be helpful for understanding contemporary plant species richness, and the manner in which lands and the useful plants that populate such lands are passed down over generations. While not all landholders build up a productive agroforestry system with their heirs in mind, others are very conscious of planting for future generations (Perz 2000: 165). An older fellow livin g on Puruzinho Lake pointed out several Brazil nut trees that his father had planted on their property and told me: who knows who else may well tended Brazil nut tree may produce after 8 years and Dad was able to harvest Brazil nuts sev eral years before he passed away. ( Fieldnotes, October 2009 ) Walking around homeg ardens, I also found women who had inherited plants from their mothers, including one of my close friends, Dani who lived in the town of Borba. In her backyard, she pointed out a Starfruit tree, the herb vence tudo ( Caladium bicolor ) and a g inj eira tree ( Eugenia spp. ) among many other plants and she


74 physical plants th emselves (and the successive generations derived from original specimens), knowledge regarding the use, cultivation, and management of plants is also passed down. And along with the inheritance of physical plants themselves and knowledge of their manageme nt comes the symbolic import ance of such plants and their use in social practice. The contribution of magic plants to agrobiodiversity and their role in the social life in rural communities in Borba is explored further in Chapter 5. Conclusions and Futur e Research The age of the household head, the area of land under management and environmental habitat are all significant variables that have important bearing on the diversity of plant species managed by rural households. Older household heads typically manage a greater diversity of species because they are able to accrue a greater diversity of plants through social interactions and relationships built over their lifetime. Households with more land available also manage a larger number of species becaus e of the greater space available for experimentation and cultivation. Furthermore, households in a mature phase of the household lifecycle typically have more available laborers and can cultivate large expanses of land. Lastly, environmental habitat has very important bearing on the species richness managed by rural households. Households located in the uplands face less crop losses than those in the seasonally unpredictable floodplains, and manage a significantly high number of species as a result. Eth nographic data point to o ther potentially influential factors including the life hi st ories of household heads, the biographies of individual farms, and the u se of medicina l and magic plants, which all can make important contribution s to the o verall species diversity at the household level.


75 Future analyses will require closer examination of mar ket influences including access and household income to see if they have a greater influence than the identifiable socio demographic variables discussed here. Future research should further e xamine the relationship between agrobiodiversity and household economic success as well as the degree to which native species contribute to agrobiodiversity on Amazonian households and household income. Finally, th e role of socia l networks in the patterning of agrobiodiversity merit s consideration, and is examined in the following chapter with regards to i ntraspecific diver sity of the staple crop manioc Figure 3 1 Floodplain h omegarden in the commun ity of So Pedro, Autazes A u River (August 2010)


76 Table 3 1. Descriptive statistics of households s urvey ed in Borba, AM, Brazil Age # Household Members # Years Residence Area of Land Under Cultivation N Valid 135 136 131 136 Missing 3 2 7 2 Mean 4 3.44 5.37 23.00 2.4714 Median 40.00 5.00 18.00 2.0150 Std. Deviation 14.455 2.762 19.017 1.90672 Variance 208.934 7.627 361.640 3.636


77 Table 3 2. List of species identifed in botanical surveys at 138 h ouseholds in Borba, AM, Brazil Family Scientific Na me Local Common Name English Name Native or Exotic (N=1) Origin #HH Present Adoxaceae Sambucus sp. Sabogueiro Elderberry 1 Neotropics 1 Agavaceae Sansevieria trifasciata Espada de So Jorge Snake plant 0 Africa 2 Alliaceae Allium schoenoprasum L. Ceboli nha Chives 0 Old World 64 Amaranthaceae Alternathera tenella Cuia mansa 1 South America 1 Amaranthaceae Pfaffia glomerata Emenda osso 1 1 Amaranthaceae Chenopodium ambrosioides Mastruz Wormseed 1 Mesoamerica 17 Anacardiaceae Spondias edulis Cajarana Polynesian plum 0 Asia 2 Anacardiaceae Anacardium occidentale Caju Cashew 1 South America 58 Anacardiaceae Mangifera indica Manga Mango 0 Asia 82 Anacardiaceae Spondias mombim Tapereb; caj Hog plum 1 South America 15 Annonaceae Annona montana Aratic um 1 Amazonia 6 Annonaceae Annona squamosa Ata Sweetsop 1 Central America 1 Annonaceae Rollinia mucosa Birib 1 Amazonia 19 Annonaceae Duguetia spixiana Biribarana 1 Amazonia 1 Annonaceae Annona reticulata Condessa 1 1 Annonaceae Annona muricata Graviola Soursop 1 South America 53 Apiaceae Coriandrum sativum L. Cheiro Verde; Coentro Cilantro 0 Old World 27 Apiaceae Eryngium foetidum Chicoria 1 18 Apocynaceae Thevetia peruviana Castanha da India 1 Andes 4 Apocynaceae Himatanthus sucuuba Sucu ba 1 2 Araceae Caladium lindenii Brasileirinha 1 South America 1 Araceae Caladium sp. Cachorrinho 1 Neotropics 1 Araceae Dieffenbachia seguine Comigo ninguem pode 1 Neotropics 1 Araceae Scindapsus aureus Jiboia Ivy 0 Asia 1 Araceae Unidentified t aja buceta 2 Araceae Caladium bicolor Vence tudo Caladium 1 South America 1 Araceae? Unidentified Poraqu 1 Arecaceae Euterpe spp. Aai Acai 1 Amazonia 116 Arecaceae Euterpe precatoria Aa do mato; Jussara Acai 1 Amazonia 1 Arecaceae Attale a sp ecios a Babau 1 South America 2 Arecaceae Oenocarpus bacaba Bacaba 1 Amazonia 56


78 Table 3 2. Continued Family Scientific Name Local Common Name English Name Native or Exotic (N=1) Origin #HH Present Adoxaceae Sambucus sp. Sabogueiro Elderberry 1 Neotropics 1 Arecaceae Oenocarpus minor Bacabinha 1 Amazonia 1 Arecaceae Mauritia flexuosa Buriti 1 South America 6 Arecaceae Elaeis oleifera Caiau American oil palm 1 South America 6 Arecaceae Cocos nucifera Cco Coconut 0 Asia 26 Arecaceae Attal ea maripa Inaj 1 Amazonia 2 Arecaceae Attalea attaleoides Palha Branca 1 Amazonia 1 Arecaceae Oenocarpus bataua Patau 1 Amazonia 3 Arecaceae Bactris gasipaes Pupunha 1 Amazonia 53 Arecaceae Astrocaryum aculeatum Tucum Star nut palm 1 Amazonia 44 Arecaceae Attalea phalerata Urucuri 1 Amazonia 1 Aristolochiaceae Aristolochia sp. Uec 0 Old World 1 Asphodelaceae Aloe vera Babosa Aloe vera 0 Old World 6 Asteraceae Lactuca sativa Alface Lettuce 0 Old World 4 Asteraceae Vernonia condensata Boldo japons 0 Old World 1 Asteraceae Tagetes spp Cravo; Cravo de defunto Marigold 1 Americas 4 Asteraceae Pectis elongata cuminho cumin 0 2 Asteraceae Spilanthes acmella Jambu 1 Amazonia 7 Asteraceae Ayapana sp. (formerly Eupatorium ayapana) Japana 1 9 Asteraceae Pluchea sagittalis Marcela 1 Americas 4 Bignoniaceae Mansoa alliacea Cip alho Garlic vine 1 Amazonia 9 Bignoniaceae Arrabidea chica crajiru 1 3 Bignoniaceae Crescentia cujete Cuia Calabash gourd 1 Amazonia 45 Bixaceae Bixa orellana U rucum Annatto 1 Amazonia 9 Brassicaceae Brassica oleracaea L. Couve Collard greens 0 Old World 9 Bromeliaceae Ananas comosus Abacaxi Pineapple 1 Neotropics 22 Cactaceae Opuntia ficus indica Palmeira (cactus) 1 Mesoamerica 1 Caricaceae Carica papaya Ma mo Papaya 1 Americas 46 Clusiaceae Platonia insignis Bacuri 1 Amazonia 1


79 Table 3 2. Continued F amily Scientific Name Local Common Name English Name Native or Exotic (N=1) Origin #HH Present Adoxaceae Sambucus sp. Sabogueiro Elderberry 1 Neotrop ics 1 Commelinaceae Callisia repens Dinheiro em penca 1 Neotropics 2 Convolvulaceae Ipomoea batatas Batata doce Sweet potato 1 South America 2 Costaceae Costus spicatus Pobre Velho; cana mansa 1 S. America; Caribbean 2 Crassulaceae Kalanchoe pinnata; K. brasiliensis Coirama; escama de pirarucu 0 Africa 15 Cucurbitaceae Cucurbita spp. Abobora, jerimum Squash 1 Neotropics 1 Cucurbitaceae Luffa operculata Buchinha Luffa 0 Old World Tropics 1 Cucurbitaceae Cucurbita spp. Jerimum; Abobora Squash 1 Neotr opics 35 Cucurbitaceae Cucumis anguria L. Maxixe West Indian Gherkin 0 Africa 23 Cucurbitaceae Citrullus lanatus Melancia Watermelon 0 Africa 11 Cucurbitaceae Cucumis melo Melo Melon 0 Africa 1 Cucurbitaceae Cucumis sativus Pepino Cucumber 0 Old Worl d 4 Dioscoreaceae Dioscorea trifida Car Cocoyam 1 South America 22 Euphorbiaceae Euphorbia milii Coroa de cristo Crown of thorns 0 Old World 1 Euphorbiaceae Euphorbia tirucalli Dedo de Deus Indian tree spurge 0 Africa 3 Euphorbiaceae Ricinus communis Mamona Castor bean 0 Africa 1 Euphorbiaceae Manihot esculenta Krantz Mandioca; Macaxeira Manioc; Cassava 1 Amazonia 125 Euphorbiaceae Jatropha podagrica Pio barrigudo Gout plant 1 Americas 5 Euphorbiaceae Jatropha curcas Pio branco 1 Americas 15 Eup horbiaceae Jatropha gossypifolia Pio roxo Bellyache bush 1 South America 41 Euphorbiaceae Croton cajucara Sacaquinha 1 2 Euphorbiaceae Hevea spp. Seringa 1 Amazonia 45 Euphorbiaceae? Unidentified Caboclo roxo 1 Fabacaeae (Mimosoideae) Inga sp. I ng chato 1 Amazonia 2 Fabaceae Inga edulis Ing 1 Amazonia 48 Fabaceae Cassia leiandra Marimari 1 Amazonia 4


80 Table 3 2. Continued F amily Scientific Name Local Common Name English Name Native or Exotic (N=1) Origin #HH Present Fabaceae (Caesa lpinoideae) Caesalpini ferrea Juc 1 6 Fabaceae (Faboideae) Vigna unguiculata Feijo de corda Surinamese longbean 0 Africa 2 Fabaceae (Faboideae) Vigna unguiculata Feijo de praia Cowpea 0 Africa 11 Fabaceae (Mimosoideae) Inga sp. Ing So Joo 1 Ama zonia 1 Fabaceae (Mimosoideae) Inga spp. Ingazinho 1 Amazonia 5 Fabaceae (Papilionoideae) Dipterex odorata Cumaru Tonka bean 1 Amazonia 3 Icacinaceae Poraqueiba sericea Mari ; Umari 1 Amazonia 7 Iridaceae Eleutherine bulbosa Marupa 1 South America 1 Lamiaceae Ocimum micranthum Alfavaca Amazonian basil 0 Neotropics 22 Lamiaceae Aeollanthus suaveolens Catinga de mulata 1 4 Lamiaceae Leonurus sibiricus Cibalena 0 Asia 1 Lamiaceae Mentha spicata Hortel grande Mint 0 Europe 7 Lamiaceae Mentha x p iperita; Mentha sp. Hortelazinha Mint 0 Europe 11 Lamiaceae Plectranthus amboinicus Malvarisco Cuban oregano 0 1 Lamiaceae Ocimum basilicum Manjerico Basil 0 Old World 10 Lamiaceae Pogostemon cablin Oriza; Uriza Patchouli 0 Asia 1 Lauraceae Persea am ericana Abacate Avocado 1 Neotropics 47 Lauraceae Licania puchuri major Puxuri 1 South America 10 Lecythidaceae Berthol le tia excelsa Castanha Brazil nut 1 Amazonia 51 Malphigiaceae Malphigia emarginata Acerola Barbados cherry 1 C. America/Caribbean 19 Malphigiaceae Byrsonima crassiflia Murici Nance 1 Neotropics 1 Malphigiaceae Byrsonima intermedia Sara tudo 1 7 Malvaceae Gossypium barbadense Algodo Cotton 1 South America 11 Malvaceae Theobroma cacao Cacau Cacao; cocoa 1 Amazonia 97


81 Table 3 2. C ontinued F amily Scientific Name Local Common Name English Name Native or Exotic (N=1) Origin #HH Present Malvaceae Theobroma speciosum Cacau rana 1 Amazonia 4 Malvaceae Theobroma grandiflorum Cupuau 1 Amazonia 83 Malvaceae Guazuma ulmifolia Mu tamba 1 1 Malvaceae Hibiscus rosa sinensis Pampola; Papola Hibiscus 0 Asia 4 Malvaceae Abelmoschus esculentus Quiabo 0 Africa 1 Malvaceae Hibiscus sabdariffa L. Vinagreira 0 Old World 7 Marantaceae Callathea allouia Ari Sweet cornroot 1 South Amer ica 2 Meliaceae Carapa guianensis Andiroba 1 Amazonia 18 Meliaceae Cedrela odorata Cedrinho 1 Amazonia 1 Meliaceae Swietenia macrophylla Cedrinho mogno 1 Amazonia 1 Monimiaceae Peumus boldus Boldo 1 South America 9 Monimiaceae Siparuna guianensis Capitiu; caa pitiu Fevertree 1 Neotropics 2 Moraceae Artocarpus altilis Fruta po Breadfruit 0 Oceania 7 Moraceae Artocarpus integrifolia Jaca Jackfruit 0 Asia 11 Musaceae Musa spp. Banana 0 Asia 95 Myrtaceae Eugenia stipitata Araa boi 1 Amazonia 2 Myrtaceae Eugenia cuminii Azeitona Java plum 0 Asia 27 Myrtaceae Psidium guajava Goiaba Guava 1 South America 89 Myrtaceae Psidium guianensis Goiaba araa 1 Amazonia 3 Myrtaceae Eugenia malaccensis Jambo Malay apple 0 Asia 67 Myrtaceae Eugenia uniflo ra Pitanga 1 South America 1 Myrtaceae Talisia esculenta Pitomba 1 South America 5 Oxalidaceae Averrhoa carambola Carambola Starfruit 0 Asia 6 Oxalidaceae Averrhoa bilimbi Limo caiano Bilimbi 0 Asia 3 Oxalidaceae Oxalis triangularis Trevo Roxo Shamr ock 1 South America 1 Passifloraceae Passifloria edulis Maracuj Passion fruit 1 South America 24 Pedaliaceae Sesamum indicum Gergelim Sesame 0 Old World 4 Phytolaccaceae Petiveria alliacea Mucura ca 1 Neotropics 13 Piperaceae Pothomorphe peltata Caa peba 1 Neotropics 1 Piperaceae Piper callosum Paregorico; Panquile 1 South America 2 Piperaceae Piper nigrum L. Pimenta do reino Black pepper 0 Asia 4 Poaceae Oryza sativa Arroz Rice 0 Asia 1


82 Table 3 2. Continued F amily Scientific Name Local C ommon Name English Name Native or Exotic (N=1) Origin #HH Present Poaceae Cymbopogon citratus Capim cheiroso Lemon grass 0 26 Poaceae Zea mays Milho Corn 1 Neotropics 14 Poaceae Vetiveria zizanioides Patichulim Vetiver(?) 0 Asia 1 Portulacaceae Portul aca pilosa Amor crescido Kiss me quick 1 Americas 11 Portulacaceae Talinum fruticosum Cariru 1 Americas 13 Rubiaceae Coffea spp. Caf Coffee 0 Africa 39 Rubiaceae Genipa americana Jenipapo Genipap 1 South America 30 Rubiaceae Morinda citrifolia Noni 0 Asia 5 Rubiaceae Alibertia edulis Puru 1 Amazonia 8 Rutaceae Ruta spp. Arruda Rue 0 Old World 12 Rutaceae Citrus sinensis Laranja Orange 0 Asia 54 Rutaceae Citrus aurantifolia Lima Sweet lime 0 Asia 9 Rutaceae Citrus aurantifolia Limo Sour lime 0 Asia 80 Rutaceae Citrus sp. Limo tangerina 0 Asia 2 Rutaceae Citrus reticulata Tangerina Tangerine 0 Asia 25 Sapindaceae Nephelium lappaceum Rambutan Rambutan 0 Asia 1 Sapotaceae Pouteria caimito Abiu Abiu 1 Amazonia 11 Solanaceae Solanum melongena Berinjela Eggplant 0 Africa 1 Solanaceae Solanum sessiflorum Cubiu Orinoco apple 1 Amazonia 15 Solanaceae Mimosa pigra Jiquiri; Jequiri 1 11 Solanaceae Capiscum chinense Pimenta chumbinho 1 Amazonia 6 Solanaceae Capiscum chinense Pimenta de cheiro Sweet hot pepper 1 Amazonia 48 Solanaceae Capsicum chinense pimenta de mesa 1 Amazonia 1 Solanaceae Capsicum annuum Pimenta doce 1 Americas 5 Solanaceae Capiscum frutescens Pimenta malagueta Hot pepper 1 South America 23 Solanaceae Capiscum chinense Pimenta murupi 1 Amazonia 19 Solanaceae Capiscum annuum L. Pimento Bell pepper 1 Central America 6 Solanaceae Nicotiana spp Tabaco Tobaco 1 Andes 1 Solanaceae Lycopersicon esculentum Mill. Tomate Tomato 1 Mesoamerica 27


83 Table 3 2. Continued. F amily Scientific Name Local Common Name English Name Native or Exotic (N=1) Origin #HH Present Urticaceae Urtica dioica Urtiga Stinging nettles 0 Europe; North America 1 Verbenaceae Lippia alba Erva cidreira 1 10 Verbenaceae Vitex agnus castus Pau de Angola 0 Europe 2 Verbenaceae Lippia grandis Salva de Maraj 1 Neotropics 3 Zingiberaceae Zingiber officinale Gengibre; Mangarataia Ginger 0 Asia 7 Zingiberaceae Alpinia nutans; A. zurembet Vendica; Vindic Dwarf cardamom 0 Asia 1 Unidentified Unide ntified Capim santo variety 2 ? 1 Unidentified Unidentified Curuminzeiro ? 1 Unidentified Unidentified Iraporanga ? 1 Unidentified Unidentified Jiru ? 2 Unidentified Unidentified Maria bonita ? 1 Unidentified Unidentified Vai vem ? 1


84 Figure 3 2 Frequency distribution of species richness managed by households in Borba, Amazonas, Brazil


85 Figure 3 3 Relationship of household head age to number of species managed by the household


86 Figure 3 4 Relationship of area of cultivated land (log transformed) to number of species managed by the household Table 3 3. Regression model of number of species per h ousehold in Borba, Amazonas, Brazil. Coefficient Standard Error t Sig. Constan t 9.658 2.023 4.775 000 Age of household head 127 .043 2.688 .010 Area of cultivated land (log transformed) 4.296 819 5.249 .000 Environmental habitat (0 floodplain, 1 upland) 3.461 1.252 2.764 007 R .315 -----------------F 19.729 -----------.000 df 134 -----------------


87 CHAPTER 4 SOCIAL NETWORKS AND MANIOC VARIETAL DIVE RSITY Int roduction This chapter examines how social networks influence the diversity and distribution of manioc varieties and discusses why the social patt erning of manioc varietal diversity is important to research on agrobiodiversity First I explore manioc varietal distribution in all 17 surveyed communities in Borba, presenting a network of households based on mod geography and community residence have important influence on the distribution of manioc varieties at the municipal level. I also highlight the ways in which manioc diversity is distributed socially, contrasting the cases of two households in the network, one with high centrality and another that is in a peripheral position. To provide a more fine grained analysis of manioc cultivation, I focus on three closely related communities situated around a floodplain la ke in Borba Examining both a 2 mode network of households tied by varieties planted in common and a 1 mode network based on reported exchange of manioc varieties among households, I describe how the two networks differ and discuss the relevance of thes e differences for understanding manioc varietal distribution Looking specifically at the network based on reported exchange, I demonstrate that households seen as experts in manioc agriculture play an important role in the sourcing of manioc varieties to other households in the area H as households with peripheral positions in the exchange network often manage less circulated varieties that contribute significantly to the overall diversity found in the communities.


88 Background Manioc ( Manihot esculenta Crantz), also known as cassava, is a biannual shrub endemic to the Amazon basin that produces starchy roots that provide the primary oc is locally classified in two general categories: macaxeira mandioca Sweet varieties of manioc contain little cyanogenic compounds and can be eaten after being peeled and boiled. Bitter varieties, on the other hand, contain higher concentrations of such compounds, and must be processed to extract their poison. However, the generalized categories of bitter and sweet manioc actually represent different extremes of a continuum since cyanogenic glucoside compound co ncentrations range widely in manioc cultivars (Wilson and Dufour 2002: 50; cf. Nye 1991 ). Manioc is one of the few major world crops in which more toxic cultivars are selected over less toxic ones (Wilson and Dufour 2002 : 50 ). B itter manioc has a tendenc y to produce greater yields than sweet manioc, perhaps because of bitter (Wilson and Dufour 2002 : 54 55) Since bitter manioc is processed as farinha (manioc flour), it can also be kept for months wi thout spoiling while sweet tendency to rot more quickly It is argued, however, that the distinctions between bitter and sweet varieties are often culturally derived and not necessarily reflective of actually c yanide contents (Nye 1991 : 47 48 ). It is further argued that manioc is not necessarily required for detoxification, but rather implemented for the production of derived food products (Nye 1991).


89 Bitter manioc varieties are typic ally used to make farinha as well as the by products tucupi and tapioca Tucupi is used as a sauce, oftentimes prepared with hot peppers ( Capsicum spp. ), while tapioca is typically prepared in form of manioc cakes known as beiju In much of Brazilian Amazo nia, bitter manioc provides the primary calories of the rural diet while sweet manioc is usually a secondary staple, oftentimes eaten at breakfast or as a snack. Modern Amazonian farming communities also typically plant a greater diversity of bitter manio c cultivars than sweet ones as is evidenced in the data collected in this study. Manioc Diversity and Social Exchange in Amazonia A number of studies have been conducted on manioc crop diversity among different ethnic groups across Amazonia ( e.g. Boster 1 984, 1986 ; Carneiro 1983 ; Chernela 1986 ; Salick et al. 1997 ) and much variation appears to exist in the distribution and patterning of varieties managed both within and among groups. In a recent study among the Piaroa of Venezuela 113 varieties of manio c were identified by more than 90 primary manioc farmers in 10 communities (Heckler and Zent 2008 ). This was attributed to selection related both to ecological factors (e.g. crop resistance to rotting and pests) and social factors (e.g. s ocial importance of manioc exchange, social status, and local aesthetics of diversity) Similarly a study in a Makushi community in Guyana with 27 households found 86 different landrace names of manioc corresponding to 76 different varieties, and an aver age 16 varieties of manioc per household ( Elias et al. 2000) In this latter study, the diversity of manioc varieties found among the Makushi was associated with both natural and human selective pressures, but it was note d that social factors including so cial exchange were important for maintaining diversity in the community.


90 In a comparative study of manioc cultivar diversity in Northwest Amazonian communities and Caiara communities of Southern Brazil, it was observed that manioc varieties circulate in s ocial networks that vary in size and composition (Emperaire and Peroni 2007) In Eastern Tukano groups of Northwest Amazonia, for example, women play the primary role in exchanging varieties, which are shown to circulate widely across the region because o f the broad kin networks resulting from the practice of linguistic exogamy (Emperaire and Peroni 2007; Chernela 1986). It is argue d and diffuse exchange networks paired with a general curiosity in varietal novelty explains the high numb er of manioc varieties cultivated in the area (Emperaire and Peroni 2007) In contrast, a lower diversity of cultivars is found among Caiara farmers in S o Paulo state, where exchange is usually restricted to neighbors and close relatives (Emperaire and Peroni 2007 ). In Piaroa communities, both men and women exchange manioc varieties, particularly among neighbors and kin (Heckler and Zent 2008) However, it is mentioned that similar to Eastern Tukanoan groups, Piaroa women are usually the ones who provi de manioc cuttings as gifts or in exchange for labor, which serves to establish and maintain productive relationships among women in their communities ( Heckler and Zent 2008 ). In another case study looking at manioc cultivar maintenance and exchange amon g the Aguaruna i n the n orthern Peruvian Amazon, the social aspects of exchange were shown to be important for identification of manioc varieties (Boster 1984, 1985) It was observ ed that most Aguaruna women exchanged varieties with their close kin, and on ly a few exchanged widely with other women in the community. I ndividuals who were


91 more well connected in manioc exchange ne tworks were the ones who had a greater understanding of the Aguaruna cultural system of identification and of local manioc diversity (Boster 1986). In contrast to prior studies that focused largely on indigenous groups of Amazonia, this research examines manioc diversity among rural smallholder communities of mixed s have long been seen to have an important influence on the distribution of crop diversity in smallholder farming communities, few studies have utilized methods from social network analysis to examine the distribution of manioc varieties both within and am ong communities. This chapter thus seeks to broaden the general understanding of manioc cultivation in the region and its social patterning in rural Brazilian Amazonia. Methods Household surveys and structured interviews were conducted at 138 households i n 17 communities in the municipality of Borba, Amazonas, Brazil. During the survey and interviews, household heads were asked what varieties of manioc they planted and the area of land dedicated to manioc. A 2 by manioc varieties in common. A t three closely related communities, a second structured interview was implemented, inquiring specifically about manioc cultivation at 45 households (represe nting 90% of the households) Household heads were asked to name the manioc varieties they had planted and with whom they had exchanged manioc cuttings in the 12 months prior to the interview. Because the majority of people interviewed were able to recall from whom they had acquired cuttings, but not necessarily to whom they


92 had donated cuttings, network data were collected on acquisition of cuttings and not developed to a nalyze patterns of manioc varietal exchange (or loaning). Household heads were also asked in what type of soil they planted manioc and what characteristics t hey considered most important in manioc varieties. Finally, to develop an understanding of which trabalhava bem com mandioca ), or in other words, were most skilled and knowledgeable in manioc agricul ture, and more than one household could be named. Expert households were ranked by a simple tallying of the number of times they were cited by other households. For b oth the 1 mode and 2 mode matrices centrality measures were calculated to distinguish differences in positionality of households in the networks In general terms, measures of centrality reflect the relative position of a node (in this case, a household ) within a network, and many studies show a strong relationship between centrality and p ower or influence in a network. There are numerous measures of centrality that can Betweenness centrality is defined as the measure of the number of times a node (or a household) lies along the shortest path between two others (Scott 2000: 86) For this study, the measure was calculated to explore the relationship between household centrality and manioc vari etal diversity and distribution. This measure also serve d to identify key households that stood out in the networks, and pointed to individual cases that deserved com plementary ethnographic detail. Lastly, analyses


93 were conducted to acquire structural information regarding the networks, looking particularly at factions and their relationship to manioc varietal distribution. Municipal Manioc Varietal Diversity and Exchange Of 138 household heads surveyed from 17 communities i n the municipality of Borba, 125 households planted manioc with an average of 3.7 v arie ties planted per household At 109 households at least one variety of manioc was named while at 16 households the names of manioc varieties planted were not known. In all, 74 different varieties of manioc were named (Table 4 1) It is likely that some of these varieties may be duplicates with distinct names applied in different communities while others are different varieties with homonyms applied. An example of the l atter is found with the residents claim that a woman from the community had brought the variety from the city of Nova Olinda around 20 years ago (see more discussion below) a nd named it after its city of origin. However, residents of another community, nearly 4 hours away from oth communities planted the same variety, the sources cited by these communities would suggest that they plant distinct varieties that share the same name. Of the 125 households that planted manioc, 77 household heads claimed to have borrowed (or been loan ed) manioc cuttings in the 12 months prior to the study while 42 household heads reported to have replanted cuttings taken solely from their own fields. The provenance of manioc cuttings was not ascertained at 6 households. 54 households acquired cutting s from within their community while 16 acquired cuttings


94 from members of outside communities and 7 households acquired cuttings from both within and outside their community. Network Analysis of Municipal Manioc Diversity Examining the municipal network, it becomes evident that households planting varieties in common are largely organized in factions based on community membership or geographic location. In Figur e 4 1, it can be seen that households in communities downriver from the city of Borba (in red) la rgely form factions apart from those upriver (in black), as do households in communities on t he tributaries of the Autazes A u and Madeirinha Rivers (in gray) as well as those from the settlement near the city of Borba (in blue). When four factions are c alculated on the visual application Netdraw from UCINET, a strikingly similar image appears to the one colored b y municipal geography (Figure 4 2), demonstrating that manioc cultivar diversity is very distinctly shaped geographically. This also suggests t hat networks of exchange are largely restricted to communities and their nearby neighbors. Cases of households seeking out unique manioc cultivars from more distant communities appear to be somewhat uncommon. In the cases of the households that did acqu ire manioc varieties from both within and outside of their community, they show particularly high betweenness centrality in the network as would be expected. In fact, the two households with the highest betweenness centrality are also two of the seven ho useholds that acquired manioc cuttings both from within and outside of their community. since they negotiate between groups. Brokers thus have a high capacity to control the flow of in


95 within a network where there would otherwise be separate components or isolated groups (Burt 2004). Jos and Cndida are the heads of the household with the highest betweenness cen trality in the municipal network, and therefore occupy a critical position in the network (Fig ure 4 3 : ). Jos was born in the floodplain community of Puruzinho where they reside while Cndida was born in a neighboring community upriver, alt hough her mother and step father now live in yet a third community across from Puruzinho on the opposite bank of the Madeira River. Jos is currently the president of his community cial activities, including the sale of diesel, gasoline, fish, Brazil nuts, cacao, and other minor market crops. Although their household has a high centrality in this manioc network, they do not rely primarily on manioc farming for their livelihood. Jos in fact, planted watermelon in te rra firme and squash in the floodplain while Cndida was the one who wanted to plant a few hectares of manioc through a project for which they had received credit through the state agricultural extension agency. Because of the flooding in the year prior, and the general lack of manioc cuttings in the area, Cndida and Jos had to scrape together cuttings in the community to plant their field, but fire from a neighboring plot destroyed a large portion of their crop. To replant the field, Jos and Cndida p urchased cuttings from contacts in Axinim and Cai ara, two of the largest communities in the municipality. Buying manioc cuttings, however, is an uncommon practice in the region. While some people may trade other goods for manioc cuttings, cash payment i s rare. Nonetheless, because of massive flooding and intense drought, source material for planting manioc became extremely scarce in 2009. Since Jos and Cndida are


96 heads of one of the wealthier households in their community and draw income from a varie ty of sources, the purchase of manioc cuttings was a simple solution to their problem. By cobbling together cuttings from within the community and also taking outside thei r community, they become a central hub of this network by bridging different pools of manioc varietal diversity in the municipality. Unlike network brokers like Jos and Cndida households with low betweenness centrality hold more peripheral positions in the network. But, these households reveal equally important information about regional manioc varietal diversity and its B Figure 4 3, for example is head of one of the households with the lowest b etweenness centrality (score=0) in the network. He also has characteristics that set him apart from many other household heads. First and foremost, Edmilson is not an Amazonian native, but rather a migrant from the Northeastern Brazilian state of Maranho He moved to Amazonas state and was able to acquire land in a farming settlement outside of the city of Borba, where he has lived for the past 10 years with his wife. He said that he manioc cuttings from anyone in the 12 months prior to the interview but rather rep lanted cuttings from manioc he already had. Although he claimed to have 5 different varieties that had al research corporation. This variety, which was distributed by agricultural extension agents in Borba was planted by only one other farmer in this study. Although the variety was developed for resistance to pests and drough t, the few farmers and agents who talked about this particular variety


97 because of its white roots which produced a less appealing manioc flour than yellow rooted varieties according to the preferences of most farmers and consumers in th e region. Since Edmilson is not an Amazonian native, however, his preference of manioc flour differs since in the Brazilian Northeast, whiter and finer grained manioc flour is common, oftentimes produced from sweet varieties of manioc. Reflecting on the cases of Edmilson and Jos and Cndida it appears that the biographies of household heads and their socio economic relationships with other households in their communities and the region have real bearing on the varieties of manioc that they acquire and p lant. O n a municipa l scale, geography and its constraints on access are important for determining patterns of manioc varietal distribution (Figure 4 4) however general attributes of the household heads appear to also be influential in the types of variet ies accessed, the ability to acquire new varieties and the preference of particular varieties cultivated. To understand manioc varietal diversity at a smaller scale, the remainder of the chapter presents a case study from three neighboring communities sur rounding a floodplain lake in Borba (colored in yellow in Figure 4 4) Unlike the municipal network in which households were studied based strictly on their ties through the manioc varieties held in common, this case study presents networks based on data from reported exchange among households as well as manioc varieties held in common. This section also provides more detailed descriptions of the varieties selected by farmers and how they circulate socially, allowing for a more nuanced view of manioc vari


98 Manioc Varietal Diversity, Exchange, and Selection in Puruzinho Lake Communities Of the 45 permanently occupied households in the three closely related communities on Puruzinho Lake, 44 households cultivated manioc. All of these 44 households planted their manioc fields in terra firme including14 households that reside on the floodplain. Some of the households that resided on the floodplain planted sweet manioc in floodplain homegardens, but their primary manioc fields wer e planted on terra firme. Of the 44 households planting manioc, 24 households cultivated manioc in anthropogenic (typically Ultisols). Because of the relatively low floodplain at Puruzinho, most families in large part because of the risk, but also because of less available land. 20 households (46%) repor ted to acquire manioc cuttings from at least one other household around Puruzinho Lake. 16 households (34%) did not acquire manioc cuttings from other households, simply replanting cuttings from their older fields. 4 households (9%) acquired manioc cuttin gs strictly from households from communities outside Puruzinho Lake and 4 (9%) households acquired manioc cuttings from both within and outside the Puruzinho Lake communities. Households studied in the 3 communities in Puruzinho actively cultivated up to 8 varieties with an average of 3.5 varieties planted per household (based on reporting from the 38 households that could name the varieties planted ) Knowledge of manioc varieties and their names varied greatly among household heads, and 7 of the 45 (15. 6%) interviewed households reported to not know any of the names of the varieties


99 in their fields. It is evident that although manioc is the primary staple of the region, not all household heads are necessarily expert manioc farmers. Some individuals foc us on planting different market crops while others may spend little time in agriculture at all, dedicating themselves instead to hunting and fishing. Nearly all households on Puruzinho Lake planted manioc, at least to avoid buying farinha but some househ old heads seemed less concerned about the varieties that they planted. This attitude farmers. One of the community experts told me point [var thought made the best farinha particularly varieties with bright yellow roots that produced yellow manioc flour, which is favored locally. 29 different manioc varietie s wer e cited by residents, a fairly low number when compared to studies of indigenous groups in other areas of Amazonia. The most common varieties were: Nova Olinda ( 71% of households), Branquinha (49% of households), and Jabuti (40% of households). 17 variet ie s were cultivated by only one or two households, demonstrating that a small number of varieties dominated in the community while much of the varietal diversity was poorly distributed. It is possible that more varieties were cultivated than those cited, since some informants in second round interviews revealed that they propagated some of their own called Guia Roxa (or Maniva do Bodeco ) was planted by several households. In other ca ses, some farmers selected and propagated varieties that only they themselves planted as was the case of

PAGE 100

100 one woman named Maria who planted a variety Nova Olinda, the most common cultivar planted in the comm unity, was brought to Puruzinho around 20 years ago according to local oral hist ory. The variety was known ata killer), but after it was brought from the town of Nova Olinda by a Puruzinho resident, it became known by the name of the town from which it originated. Dona Antonia, the woman who brought the variety to the community, planted a small patch near he r home and when the variety was shown to produce large roots that were deep yellow, other families requested cutti ngs from her so they could propa gate the variety as well. From that point on, the variety came to be planted widely in the community. produces large yellow roots that mature relatively quickly Many farmers in Pu ruzinho commented on the maturation of roots as an important aspect of varietal selection. Slower maturing varieties (sometimes referred to farinha while quicker growing varieties ( ( aguada ) and produce more tucupi juice which is used as a base for hot sauce Some farmers favor fast growing varieties because they yield farinha more quickly while others prefer slower growing varieties that provide more dry mass, and according to some farmers, better quality farinha Slow growing varieties tend to be more resistant to rot ting and can thus be left under ground for longer periods of time while quick growing varieties have a smal ler window of time for harvesting. One of the popular slow growing varieties is known regionally as name d after the Amazonian tortoise because it

PAGE 101

101 custa a morrer ) while a quick gro mes whic h suggests that it can mature in as fast as 4 months. Still, some farmers prefer varieties that have intermediate maturation rates that provide plenty of dry mass from their roots, but also have enough moistur e content to be easily grated ( cevado ) and pro cessed. When farmers in Puruzinho were asked about the most important characteristics that they looked for in manioc varieties, the majority said that they wanted variet ies that produced large roots ( d bem batata ). Many farmers also reported to seek out or select farinha has a distinctive taste, it is nearly always favored over lighter colored varieties in the Central Amazon region. To highlight this point extensi on agents told me that some producers in Borba even added different products to dye their farinha yellow because it made their product more attractive to consumers and helped boost their sales S weet manioc varieties typically have white roots but there are also varieties of bitter manioc with white roots that are used mostly for making manioc cakes, the most gostosa ) manioc flour, but because of the color, few people used the variety for that purpose. She went on to say that out of custom (or perhaps aesthetic motivations) varieties with yellow roots were the most coveted for making farinha This was also found among the M akushi of Guyana ; white varieties wer e used to make manioc cakes while yellow varieties we re prefer red for making manioc flour (Elias et al. 2000).

PAGE 102

102 Network Analysis of Communities on Puruzinho Lake The network of households tied by manioc varieties planted in common shows no distinct factions for the three individual communities in the Puruzinho Lake area. This suggests that geography and community residence have less of an impact on varietal distribution at this smaller scale network than at the larger mu nicipal level network. Although one of the three communities sits on a stretch of floodplain in between the Madeira River and Puruzinho Lake, and is separated from the other two communities that reside on terra firme, floodplain residents are in frequent contact with members of the other two communities. Because all of the households living on the floodplain plant manioc on the terra firme and members of all three communities have close kin relations, interaction is relatively constant. When the network is manipulated so that nodes (i.e. households) are sized by betweenness centrality, several households have high betweenness and no one particular household stands out (Figure 4 5 ). However, when the network is altered to only show ties among households t hat share 2 or more varieties of manioc, a very disti nct structure emerges (Figure 4 6 ). In this configuration, one household stands alone with the highess betweenness centrality, which is that of an elderl y widow, Dona Clia (Figure 4 6 ; household A). Do na Clia was born in the community of Caiara in the municipality of Borba and moved to Puruzinho to live with her husband in 1942, at the age of 14. Today she lives alone, but has children and grandchildren living only a few houses down from her. Despit e her age, she continues to plant manioc and eagerly weeds her plot. Even after two different incidents in which she was bitten by snakes (one of which was venomous)

PAGE 103

103 while working in her manioc field in the past few years, she refuses to quit planting man ioc and tending her field. She takes pride in staying active at her age. I spent many afternoons with Dona Clia on her front porch listening to her stories about Puruzinho and life in the Amazonian countryside. She laughed and told me that one of her gr Dona Clia has children who live in the capital of Manaus who have also urged her to move to the city and live with them, but she told me that she prefers to be out in the Dona Clia cited 8 different varieties of manioc that she planted, more than any other household in Puruzinho except for one (that also planted 8). Compa red to the many individuals whom I interviewed about manioc varieties and their diversity, Dona Clia expressed a heightened interest in the subject and eagerly discussed at length the nuanced differences of local varieties, including their rates of maturation, the color of their roots and leaves, and the quality of the farinha they produce. As she sai d herself, and it c an be argued that her unique enthusiasm for planting manioc and tending to her fields day after day despite the intense Amazonian sun and threatening snakes can explain in part her centrality in the network. Manioc Exchange Network for Communities at Pur uzinho Lake In addition to the network of households tied by manioc varieties in common, data were collected to develop a network based on reported exchange of manioc varieties among households in the three communities surro unding Puruzinho Lake (Figure 4 7 ). This network of exchange presents a very different structure from the one based on

PAGE 104

104 manioc varieties planted in common. This difference is seen statistically when comparing the structures of these networks using a QAP (Quadratic Assignment Procedure), which shows only a minor significant correlation (r=.07; p=.012). Discrepancies between the two matrices are not necessarily because household heads were not accurately reporting from whom they had acquired cuttings (although this may contribute in part) but rather two households may plant the same varieties without necessarily acquiring the varieties from one another in the reported year. Therefore, many households are seen as tied in the first network based on varieties planted in common, but not in t he network based on reported exchange. As mentioned above, 16 household heads reported not to borrow or exchange manioc cuttings during the 12 months prior to interviews, and because of the general an e ven higher number of households would replant cuttings taken from their own fields than interviews, a common theme in discussions about manioc agriculture was the general lack of available manioc cuttings after the record floodi ng and drought in 2009, and many household heads admitted that they had to scramble to find cuttings to plant their getting cuttings and many times they have plenty left over from the own fie lds, so they the exchan ge network observed in Figure 4 7 is not a reflection of typical year, but rather reflects a year of shortage of planting materials. Knowledge and Manioc Agriculture During interviews, household heads were asked to name the households that they trabalha bem com mandioca ), or in other words, the

PAGE 105

105 households that were most skilled and knowledge able in relation to manioc agriculture. One household stood out in particular, which was referenced 23 times. Not coincidentally, this household that was also the household with the highest betweenness centrality in the Puruzinho network of exchange (Fig ure 4 8 ; household A). The second most cited household, which was named 10 times, had the second highest betweenn ess centrality in the network (Figure 4 8 ; household B). This would suggest that households that are seen to have more knowledgeable or exper t manioc farmers are those that are often sought out when individuals are seeking to acquire manioc cuttings. In their research with the Makushi in Guyana, Elias et al. (2000) eholds that manage their cuttings efficiently and have enough to share with other community members while the latter refers to households that suffer from shortages of cuttings and Puruzinho, a similar tendency appears in which more centralized households in exchange not only donate to more households, but they are also viewed as the mos t knowledgeable or expert manioc producers. most centralized households in the manioc exchange network in Puruzinho do not maintain a high number of manioc varieties, and there was no significant correlation .151 ; p=.373). In general, only a few varieties are distributed widely through the community, and in interviews, the impression was given by many that a high diversity of varieties is

PAGE 106

106 not necessarily sought out by most manioc farmers. Instead, it seems that most individuals (and households) seek only a few varieties that are reliable or known to produce large, yellow roots. It could be further ar gued that the general lack of varietal diversity in the area may be related to the network structure as the majority of households are reliant on only a few households as sources for manioc cuttings, and varieties to those households in need of loaned planting material. If the most centralized households in exchange networks are typically those recognized in the community as having the most knowledgeable manioc farmers, s ocial network analyses do demonstrate is that households that are not necessarily viewed as change, community. In other words, social network analysis can be helpful to demonstrate how households that are peripheral to the network harbor a considerable portion of var ietal diversity, diversity that is otherwise ignored, or hidden. Manioc Diversity Loss The overall number of manioc varieties as well as the mean number of varieties planted per household in the communities around Puruzinho Lake are relatively low when com pared to most manioc diversity studies from other areas of the Amazon region. It would appear, at least from these data, that caboclo communities simply plant less varieties of manioc compared to the Amazonian indigenous groups that have been more commonl y studied (c.f. Boster 1984; Heckler and Zent 2008; Emperaire and Peroni 2008). This may also suggest a loss in indigenous knowledge regarding manioc

PAGE 107

107 cultivation among mixed descent peasants ( cf. Atran and Medin 2008 : 172 173 ), or perhaps a comparative la ck in social or cultural value attached to manioc varietal diversity in caboclo communities. On the other hand it is also possible that this difference in overall diversity is related at least in part to data collection and the specific period of time wh en research was undertaken. Nearly all households that were visited commented on the extreme nature of the climate and its impacts on agriculture prior to my arrival to the field site in 2009. Record floods led to a general loss of manioc production in B razilian Amazonia and a concomitant general shortage of manioc cuttings throughout the region. Although cuttings and different manioc cultivars from other communitie s that plant in the floodplain was greatly reduced. Moreover, many farmers in Puruzinho commented that the intense summer in 2009 also led to massive manioc crop failure. Families that usually replanted fields with cuttings taken from older fields had to scramble to find cuttings elsewhere, potentially losing varieties they had collected and adopted over time. Curiously, m any informants chuckled when asked what varieties they planted in their manioc swiddens. During initial interviews, a farmer known as Tra ra, who was the said that people invented all kinds of names for manioc varieties nly after being asked twice did he name a variety that he had planted. It was diffic ult to discern whether he and others did no t take the varietal names seriously or the relevance or need of discussing them with a researcher Later dur ing a second

PAGE 108

108 interview with Tra ra, he provided more manioc varietal name s and also mentioned several others that were planted in the community in the past, but have since begun to disappear in the community When asked why such varieties were disappearing, he said other varietie s had been introduced that produced bigger roots and were yellower and became more wide ly adopted and favored in the community. It must also be noted that an exhaustive catalogue of manioc varieties was not attained. Many household head s cited the names of the v arieties that they knew best while info rming that several ot hers existed in their fields, but t hey did not know the specific name s of the other lesser known varieties Also, while names of manioc varieties were solicited from all households, field visits were rarely conducted during interviews because manioc fields were oftentimes located very distant from the home. Certainly a greater number of varieties would have been elicited if manioc fields were visited, thus allowing farmers more direct identification of varieties beyond those they reca lled strictly from memory. Finally, while manioc remains the primary staple for the majority of rural families in the municipality of Borba, it seems that many households in past years shifted the focus of their production to other market crops and opted t o buy or trade for manioc flour. This strategy was re evaluated by many farmers however, as the flooding and drought in 2009 led to major losses in manioc production and prices for manioc flour soared as a result. In 2007, a sack of manioc flour (approxi mately 80 liters) reached a price of about 40 45 reais (approximately 25 U.S. dollars at time of study) in Borba. In 2009, prices reached 160 reais for a sack (around 95 U.S. dollars), and many people began selling manioc strictly by the liter (at rates o f 2 to 2.50 reais). Some farmers who had adopted

PAGE 109

109 other market crops that were once more valuable in the market reverted to manioc production not only because prices improved, but also because it remained a critical afford to pay for manioc flour at such elevated prices. Conclusions From these analyses at both the municipal and local scale, several important conclusions can be drawn. First, geography (or location within the municipality) and community residence have important influences on the patterning of manioc varietal distribution at the municipal level. However, when looking at the individual cases of households at different positions in the network, it becomes apparent that general attributes of the household heads also have crucial influence on the types of varieties accessed, the ability to acquire new varieties, and the preference of particular varieties cultivated. At the local scale, household location and environmental zonification plays less of a role on varietal distribution as its patterning appears to be related more closely to key households that drive exchange and the sourcing of manioc varieties. In the three communities at Puruzinho, households that were seen as the best manioc farmers or largest impact on manioc varietal distribution. However, expert households did not manage the highest diversity of manioc varieties. Instead, households that were more perip heral in the exchange network and highly centralized in the 2 mode network of households (tied by varieties planted in common), served important roles in managing and maintaining rare and poorly distributed varieties. For this reason, it can be argued

PAGE 110

110 tha t peripheral households in exchange networks are key players in managing unique varieties important to overall varietal diversity. This study demonstrates some of the ways in which social network analysis can be a useful tool for understanding the social d istribution of crop diversity, and future studies examining the distribution of other crops in different areas of the world may benefit from this model. Future research could be improved by incorporating GIS into the analysis to further investigate the re lationship between spatial distance and social networks Also, longitudinal studies that collect data on patterns of exchange networks could be particularly useful for understanding how such networks may respond or adapt to changes over time, specifically in response to crises like flooding and drought as was witnessed in this study. Furthermore, the comparison of networks of exchange among different social groups and ethnicities could provide useful information for understanding how specific social stru ctures or social networks are more conducive to the conservation of crop diversity than others. Lastly, the influence of gender requires on going consideration with regards to circulation of planting materials in social networks. Gender roles and the gen dered division of labor vary in Amazonian societies, having potentially distinct effects on the exchange and distribution of crops (see Chapter 6)

PAGE 111

111 Table 4 1. Varieties of bitter and sweet manioc managed by household s survey e d in Borba AM, Brazil Commo n Name Sweet/bitter #H ouseholds %H ouseholds Amarelinha bitter 23 18.4 Anuma bitter 1 0.8 Aripuana bitter 4 3.2 Aruana de branca bitter 1 0.8 Aruana de vermelha bitter 1 0.8 Arrepiada bitter 1 0.8 Arroz bitter 1 0.8 Aruari bitter 16 12.8 Aruazinha bitter 1 0.8 Azulona bitter 3 2.4 Batatinha bitter 5 4 Boazinha bitter 1 0.8 Branquinha bitter 24 19.2 BRS Embrapa bitter 2 1.6 Cabral bitter 14 11.2 Cacaia bitter 4 3.2 Camarao bitter 2 1.6 Curua' bitter 1 0.8 Castanha bitter 1 0.8 Flechinha b itter 3 2.4 Folha fina bitter 1 0.8 Geralda bitter 3 2.4 Giovana bitter 1 0.8 Guia_roxa bitter 5 4 Hermosa bitter 1 0.8 Jabuti bitter 20 16 Jabuti de Ouro bitter 9 7.2 Jabutirana bitter 12 9.6 Japiim bitter 1 0.8 Japonesa bitter 3 2.4 Ligeirinho bitter 1 0.8 Macaxeira amarela sweet 2 1.6 Macaxeira amazonas sweet 1 0.8 macaxeira bolacha sweet 1 0.8 Macaxeira braco vermelho sweet 1 0.8 Macaxeira cardoso sweet 2 1.6 Macaxeira pele b ranca sweet 1 0.8 Macaxeira pele vermelha sweet 29 23.2 Maca xeira manteiga sweet 5 4 Macaxeira po sweet 13 10.4 Macaxeira peruana sweet 3 2.4 macaxeira piranguinha sweet 2 1.6 Macaxeira requebra sweet 3 2.4 Macaxeira 4 mes sweet 1 0.8 Mandioca branca bitter 2 1.6 Maniva baixinha bitter 1 0.8 Manjiona bitte r 3 2.4

PAGE 112

112 Table 4 1. Continued Common Name Sweet/bitter #H ouseholds %H ouseholds Mantegueira bitter 1 0.8 Nariquinha bitter 1 0.8 Nova Olinda bitter 34 27.2 Onca bitter 1 0.8 Papinhana bitter 1 0.8 Paraiso bitter 9 7.2 Piramiri bitter 3 2.4 Pirar ucu bitter 10 8 Pirarucu amarelo bitter 5 4 Pirarucu branco bitter 3 2.4 Pirarucu de ouro bitter 5 4 Pirarucu de prata bitter 1 0.8 Pirarucuzinho bitter 1 0.8 Purupuru bitter 9 7.2 Roxinha bitter 5 4 Tartaruga bitter 5 4 Tauazinha bitter 4 3.2 Ti ririca bitter 20 16 Tracaja' bitter 2 1.6 Tucum bitter 7 5.6 Tucunare bitter 3 2.4 Zuiuda bitter 2 1.6 4 mes bitter 19 15.2 4 mezinha bitter 2 1.6 6 mes bitter 5 4 1 ano bitter 2 1.6 21 bitter 1 0.8

PAGE 113

113 Figure 4 1 Municipal network of hous eholds tied by manioc varieties planted in common and colored by geographic location in the municipality Madeira downriver Madeira upriver Autazes Au INCRA settlement

PAGE 114

114 Figure 4 2 Municipal network of households tied by man ioc varieties in common with 4 factions calculated Figure 4 3 Municip al network o f households tied by manioc varieties, nodes sized by betweenness centrality and colored by exchange A) Jos and Cndid B A Exchange in comm. Exchange in & out comm,

PAGE 115

115 Figure 4 4 Municipal network of house holds tied by manioc varieties, households colored by c ommunity Figure 4 5 Puruzinho Lake network of households tied by 1 or more manioc varieties planted in common, households colored by community.

PAGE 116

116 Figure 4 6 Puruzinho Lake network of households tied by 2 or more manioc varieties planted in common, households colored by community nodes sized by betweenness centrality A) Dona C Figure 4 7 Puruzinho Lake network with households tied by manioc varieties exchanged A

PAGE 117

117 Figure 4 8 Puruzinho Lake network with househol ds tied by manioc varieties exchanged nodes sized by betweenness centrality A) The household with the higest betweenness centrality. B) The household with the second highest betweeness centrality. A B

PAGE 118

11 8 CHAPTER 5 MAGIC PLANTS AND THE IR CONTRIBUTION TO AGROBIOD IVERSITY Introduction This chapter discusses how s ocio cult ural beliefs attached to plants and the use of plants in the practice of magic play an important role in agrobiodiversity maintenance. I n many rural households in Central Amazonia, one can find pl ants that are actively adopted and cultivated because of the ir perceived supernatural power or influence Some plants are cultivated to ward off the "evil eye" ( mau olhado ) or deter unwelcomed spirits and people from a family's home. Other plants are use d for healing baths ( banhos ) when individuals are afflicted by folk illnesses or sicknesses of ambiguous origins. Still others are carried by farmers as they work in their fields to protect them from snakes and other physical dangers. In this chapter I outline three categories of magic plants and discuss their importance to the study of Amazonian agrobiodiversity. Based on homegarden surveys and ethnographic data collected at households in the municipality of Borba I show that approximate ly 25 % of the 185 species identified in this study have uses related to magic and supernatural beliefs To conclude I discuss some of the ambiguities inherent to the categor ization of "magic" plants that stem from the ir multiple uses and the varying perceptions of th eir benefits The Anthropology of Magic and Magic Plants The study of magic has received attention from anthropologists since the birth of The Golden Bough: A Study in Magic and Religion was one of the most influe ntial early anthropological texts to outline both

PAGE 119

119 later received criticism 12 it nonetheless represented an important contribution to the early anthropological study of magic, inspiring many other works In fact, after reading The Golden Bough Malinowski then that anthropology, as presented by Sir James Frazer, is a great science, worthy of as much devotion as any of he r elder or more exact sister studies, and I became bound In his book Coral Gardens and their Magic Malinowski dealt specifically with the agriculture. In this work, he described the use of magic in the clearing of fields, the growth of plants, and the harvesting of crops. that were used in garden magic in Trobriand society, and he noted that only three non plant materials were used in magic, demonstrating the dominant role of vegetative materials in magical ritual and practice among the Trobrianders (pp. 152 154). In many different instances, Malinowski de scribed garden magic rites and applied further wrote that such leaves a are consumed or destroyed i n the rite; and though they are in a way instruments, [they] stand in special relation to magic in that they are always being the r ecipients of the magical virtue 12 (Frazer 2009: 11). La ter scholars took issue with this view, including Levi Strauss 1966: 13).

PAGE 120

120 Many other anthropologists followed Malinowski in the study of magic and witchcraft including E.E. Evans Pritchard (1958) who also described the use of magic plants in his classic work on the Zande. Because of the close relationship between magic and healing, Evans and pro vided a revealing justification for this translation: to the context is ngua Ngua ask a Zande what medicine is used for a certain acti vity we are asking him what tree or plant is used, though our question may also have a general sense of what magical rite is performed. (p. 440). Pritchard also alluded to the great diversity of plants used in Zande magic: Zande homesteads there are medicines growing at the foot of shrines, through the bush with A zande the y would often point out to me trees and plants with magical uses since they knew I was interested in the subject. There are a vast number of such plants and trees, and they are employed in a great variety of ways. A full description of even the f ew medicines of which I have records is out of the question (p.433) Evans Pritchard describes several categories of medicines, which he outlined as the to avenge adu Pritchard revealed the pervasive use of plants in Zande magic.

PAGE 121

121 C ategories of Magic Plants in Amazonian Peasant C ommunities Among contemporary rural Amazonians, a wide number of plants is believed to However, the ways in which such magic plants are used and the reasons for their usa ge can vary greatly. In contrast to other scholars who have described magic plants related to specific activities, I discuss the practice of magic and the usage of magic plants in relation to three distinct powers: the power to repel (and protect) the po wer to attract, and the power to heal. I argue that although the use of plants to ward off the evil eye or draw money into the household may seem curious to an outside observer individuals in these communities rely upon these socio cultural symbolic prac tices to c ope wi th preoccup ations including martial infidelity, social anxieties health concerns, and personal finances, among others. Furthermore, the symbolic power imbued in magic in these communitie s, including healing a loved one suffering from a folk illness or protecting the household from unwanted visitors. Plants with Power to Repel (and Protect) After my first few weeks in the city of Borba, I began to notice a species of ornamental plant that appeared consistently in the fronts of houses on the street where I lived. I asked Dani the owner of the house where I was living, the name of the plant, and she told me it was pi o roxo ( Jatropha gossypiifolia ) (Figures 5 1 and 5 2 ). She said it had sta rted growing in her yard voluntarily and she left it because she liked the way it looked. Since then several filhos as well. I asked if the plant had any use and she said that some people used the latex as a remedy to help heal wounds or scars. Later, I learned, however, that other people

PAGE 122

122 planted it to protect the household from the evil eye ( mau olhado or olho gordo in Portuguese) From that day on, I began to notice pi o roxo in front of residences both in the city and rural areas where I was conducting interviews and collecting data on plant diversity. It seemed to crop up everywhere (Fieldnotes October 4, 2009) I also found pi o roxo to crop up in many studies on the use of medicinal and magic plants in the Americas. According to the Handbook of Medici nal Plants of Latin America pi o roxo Afro Brazilian Candombl A long list of medicinal uses for the plant are cited in the text from a wide range of cultures through the Caribbean and across the Amazon. In another instance pi o roxo is said to be planted in the front yards of followers of the A fro Brazilian religion Candombl to avoid the evil eye (Voeks 1997: 106). In the text Medicina Popular do Centro Oest e (Popular Medicine of West Central [Brazil]), numerous varieties of pio and distinguishes pio roxo as a potentially deadly poison (Or tencio 1 997 : 337). The informant g oes on to say, however, that pi o roxo is a leaves for sale in the fairs, which he learned were u sed in baths to relieve feiti o or spells (O rtencio 1997: 337 ). In a r eview of Amazonian homegardens and their historical evolution, a reference is made to pi o roxo as a representative example of an Amazonian magic plant that became adopted in folk Catholicism during the Colonial period: Life in t cultural result was a patchwork of beliefs, the syncretism of shamanism with practices and beliefs and the

PAGE 123

123 colonial influences of the Portuguese, as well as African slaves. Some of these beliefs are associated with a variety of magical/medicinal plants (e.g., pio roxo Jatropha gossypiifolia ) often cultivated in modern homegarden s, and which along with ornamentals, are often seen even in diminutive front yards in cities such as Manaus (Miller et al. 2006: 49 ; boldface print mine ). As pi o roxo i s established i n a category made to the c lose relationship between plants used in magic and healing Therezina Fraxe (2004) illustrates this relationship as she describes a midwife using the leaves of pi o roxo when praying for a woman or child who may be suffering from quebranto vento caido v ermelho and other spells or folk illnesse s typical in rural Amazonia (Fraxe 2004: 2 09). An excerpt from The Enchanted Amazo n Rain Forest explores this further: Many people in the region believe in the evil eye known in the Brazilian Amazon Young children are especially vulnerable to the evil eye. If a child is the object of a great deal of admiration, for example, he or she may become weak or feverish. A victim ward off its ha rmful influence, women plant pio roxo, pi o branco, vassoura, and rue in their yards, especially by the front entrance of their home (Smith 1996: 126). In Brazilian Amazonia as much of Latin America, emotions are considered to have the pow er to cause sickness (Rebhun 1994). The evil eye, which is believed to be rooted in feelings of envy, is commonly seen as responsible for folk illnesses in contemporary rural Amazonia. Although a wide number of anthropological analyses of the evil eye ex Americas, but was brought over from the Old World. Yet regardless of its origins, the evil eye is a concept that reflects an arguably universal preoccupation in humanity with envy and its potentially dangerous or injurious implications. In a classic review entitled this preoccupation is pinpointed : in every society people use symbolic and nonsymbolic cultural forms who se function is

PAGE 124

124 to neutralize, or reduce, or otherwise control the dangers they see stemming from envy, and especially the ir fear of envy (Foster 1972 : 165 ) While pi o roxo is sometimes viewed as a medicinal plant, it must also be recognized as a symboli c weapon that serves to fight the power of envy. Many other plants are cultivated in Amazonian homegardens for similarly symbolic reasons. Rue ( Ruta graveolens ), known locally as arruda is another common species found in Amazonian homegardens and it is a lso to ward off the evil eye (Figure 5 3) Rue is of an Old World origin and the beliefs attached to it are rooted in African religions (Voeks 1997). Although assumed that use of magic plan ts in Amazonia communities is the product of Amazonian indigenous tradition, several magic plants found in modern rural Amazonian communities are actually of Old World origin. With the Columbian exchange, several species were introduced and adopted because of their cachet as exotic plants, and believed to po tentially have greater powers for healing and magic. During a visit in the community of S o Joaquim, I learned that cowboys sometimes placed a stalk of rue behind their ear to protect them while working In the community of Auar Grande, an affable fe llo w named Mario told me that rue was used to protect people from mau olhado and olho gordo. Mario also said that the oldtimers ( os antigos ) believed that you had to steal a cutting of rue if you wanted to plant it in your yard, otherwise it would lose its e ffect if it was knowingly loaned from someone (Fieldnotes October 2009) This presents a unique irony that could be appreciated in a Levi Straussian structural analysis: you have to steal a magical plant that protects you from covetousness.

PAGE 125

125 A plan t also cultivated because of its power to repel is known as comigo ninguem pode Dieffenbachia sp.). Comigo ninguem pode is a common ornamental that can be purchased even at garden stores in the Unit d can be kept in doors, although in Amazonian homegardens it is often kept close to the home. On a visit to the community of Floresta I asked a young woman about some of the magic plants in her garden, and I learned that comigo ninguem pode was not just cultivated for aesthetic reasons, but also used to keep unwanted guests or enemies away from the household (Fieldnotes : January 28, 2010 ) I shared with her that it also contained potent toxins. Referred to a potential toxicity of the plant is most likely linked to its symbolic power. I initially foun d it found that problems with thieves were common in many communities. I real ized that plants like comigo ninguem pode represented ways of coping with threats from un desirable characters in the community at least in an indirect and deflected manner. And as I found in a brief visit to the Peruvian Amazon, individuals who command ed a heig htened knowledge of plants and we re considered curers or witches, we re less likely to be crossed by other community members. Therefore, the ability to command or exploit the symbolic power of plants can sometimes translate into real power. Another c urious plant that fits in this categ ory is known locally as poraqu (Unidentified sp., Araceae family) named after the Amazon i an electric eel (Figure 5 4) I

PAGE 126

126 learned about the plant while visiting with D ona Maria Guadalupe in the community of Vila Gomes, who was one of the few people who talked openly about magic plants without a tinge of self conciousness (Fieldnotes : January 29, 2010) Most other people I interviewed chuckled awkwardly when I asked about the subject either because they found it to be si lly or they were embarrassed to discuss what may be perceived as the different plants, but also helped with a few. Her son Jorginho sho wed me the plant they knew as pora qu which they said you could use when playing soccer (Figure 5 5) He said that you simply had to tuck a leaf in your belt and if some other player tried to body check you, you poraqu would repel them! Beyon d repelling soccer defenders or the evil eye, other plants like castanha da ndia ( Thevetia peruviana ) are used to repel the physical threats of animals. The fruit of castanha da ndia is said to have power to scare off snakes. Farmers pierce a hole thro them as the y work in the fields, either clearing a new piece of land or weeding their manioc plots. Poisonous snakes like the surucucu are real threats to farmers, and friends told me of numerous horror stories about snake bites. The surucucu pico de jaca ( Lachesis muta ) is considered the most dangerous, and people commonly say if it ( Quando n o mata, aleija ) Like comigo ni nguem pode, castanha da india is a toxic plant and its toxicity is perhaps conceptualized as countering that of a poisonous snake. Moreover, it provides comfort to farmers when they must work in fields to know that they have a charm that may protect them from

PAGE 127

127 harm. A n unidentified aroid known as taj buceta (curiously named after a colloquial term for vagina in Portuguese) was found at two households and said to be used to prote ct against snakebite much like castanha da ndia. Although the actual physi cal capacity of castanha da ndia to ward off snakes is not known, the plant, like others mentioned before, is used symbolically to cope with the fear of poisonous snakes. Plants with Powers of Attraction Sometimes interviewees pointed out plants that were kept close to the home not to repel threatening animals or the evil eye, but rather to attract money, good luck, or even people. One example is the plant dinheiro em penca ( Callisia repens ) which appeared in two homegardens and was said to be used to at tract money to the household. The plant is typically kept in a small hanging basket, spilling out over the sides with its numerous small leaves, much like an overflowing basket of money. Sir James Frazier omeopathic magic as a form in which one attempts to attain a result through its mimicry, and this plant is employed symbolically following the same rationale Just as some herbalists follow the doctrine of signatures, using plants that resemble parts of the body to heal those body parts, an overflowing pot of green leaves can similarly be used to bring money into the household. The plant known as vai buscar ( Unidentifed sp. ) which I found in Dona garden, was also designed to attract money. O ther plants are kept as go od luc k charms. U ec ( Aristolochia sp.) is one of these plant iraporanga (Unidentified sp.) has more specific purposes as it is used by fisherman to ensure a successful outing on the In the rainy months of January,

PAGE 128

128 February, March, when lakes are flooded and fish are rela tively scarce, fishermen can always use a little luck Some magic plant s are used to attract people. Jiboia ( Scindapsus aureus ) is a smal Since snakes are widely hated or feared by Amazonians, I thought it to be somewhat counterintuitive that a plant named after a boa w as believed to attract people. I considered that the plant may have b e e n given the name because of it s winding, snake the fact that the boa is known for hypnotizing its prey (N.J.H Smith, pers. com m.) Curiously I found that the owner of a general store in the town of Borba kept a small dead boa constrictor (the animal) preserved in a jar because he said it also helped attract customers. I jokingly wondered to myself if lowering his over inflated prices may have be en a more effective tactic One of the more comically powerful plants de scribed to me was cachorrinho ( Caladium spp.) yard and s aid it was good for a married man to k eep around his house. I inquired as to why and she told me around l ike a puppy, responding (Fieldnotes January 28, 2010). Although we laugh ed about this, later reflection mad e me consider a more se rious reason for its symbolic us ag e. In rural Amazonian communities as in most societies, infidelity is a major concern among couples. For t his reason, the cultivation of cachorrinho may be a way for a concerned husband to indirectly confront his doubts of infidelity, or perhaps serve as a guilty reminder to his unfaithful spouse.

PAGE 129

129 Plants with Power to Heal When asking about pi o roxo and plants us ed against the evil eye, I found that many respondents also discussed plants they kept in their gar dens that were used in healing b aths, or banhos Many of the same plants used to ward off the evil eye we re used for banhos including pio roxo, pio branco ( Jatropha curcas ) cip alho ( Mansoa alliacea ) mucura ca ( Petiveria alliacea ) and rue In additi on, japana ( Ayapana spp ) manjeric o ( Ocimum basilicum ) cuia mansa ( Alternathera tenella ) alfavaca ( Ocimum micranthum ) cravo ( Tagetes spp.) hortel ( Mentha spp.) jambu ( Spilanthes acmella ) and uec were used in healing bath s. These banhos are usual ly used to treat infants and children either when they catch colds or are cranky, but they are also commonly employed when children are affl icted with folk illnesses like quebranto or v ento caido (see Fraxe 2004; Smith 1996) Much in the same way that the se plants protect the se plants are used to heal children that may be affected by envious or excessively affectionate neighbors and strangers, the sources of these folk illnesses (Figure 5 7) When I asked my friend Dona Nanda about how she learned to use plants in When we have kids, we have [to learn] to do everything (Fieldnotes: April 2010). L acking medical assistance, l earning home remedies for common problems and illnesses is a critical skill for young mothers. I later heard other women, report similar stories, explaining that they learned from their mothers when they had children In some cases, magic plants are used to bathe adults, particularly those that are acting irrationally or in a bad mood. One woman told me that cuia mansa was used to and supposedly it has a calming effect.

PAGE 130

130 I mily held a birthd ay party for him, he spent the d ay drinking and began acting irrationally and a ggressive His m other grabbed some leaves of pi o roxo from the yard in attempt to give him a bath in the hope that it would calm him down. In other cases, ban hos are used to treat panema or panemice Panema is an Amazonian cultural concept that can be understood as a form of bad luck. In rural Amazonian, panema is most commonly described when an individual has difficulty hunting game or catching fish (Smith 1 996: 101) A number of practices and habits are performed eithe r to prevent one from getting p anema or to rid oneself of it. I learned, for example, that I was not supposed to walk on the scales of the pirarucu ( Arapaima gigas ) or the aru an ( Osteoglossu m sp p.) fish because they could give me panema To free oneself of panema some women told me that baths made with the vine known as c ip alho we garlic smell when crushed which give vine A plant with a similar odor is mu cura c a ( Petiveria alliacea ) which is used for the same purpose (Figure 5 6) In We s tern Amazonia, it is sometimes prepared with the hallucinogenic plant ayahuasca in a ritualistic cleansing practice against witchcraft ( McKenna et al. 1995 ). One common characteristic of many of the plants used in banhos is their distinct smells and aromatic qualities when crushed and mixed in water. After interviewing a woman n amed Ivanice and asking several questions about such plants, she did her best to sum up the tradition of banhos

PAGE 131

131 were sick. They bathed us with stinky plants! Well, I guess they smell good to some Con tribution of Magic Plants to Overall Species Diversity Of the 18 5 total useful species identified in this survey, 48 species of plants could be considered to have supernatural infl uence or, at the very least, are used in practices related to supernatural b eliefs (Table 5 1) In other words, 26 % of the plants identified in this research had uses that were symbolic or related to magic 28 of these species are native to the Americas while 14 are exotics, and 6 are of unknown origin. With regards to their u sage, 12 species were cultivated because of their powers to repel with 8 plants used specifically to ward off the evil eye. 6 species of plants were used because of their powers of attraction, 3 of which were used to attrac t luck or wealth while the rema ining 3 were used to attract people. 31 of these species were used in healing baths to treat folk illnesses. A large portion of these plants were identified as having multiple purposes. The most common magi c plant from this survey was pi o roxo, found at 30% percent of those households surveyed (41 out of 138). The other most co mmon plants were: capim c heiroso (18.8% households ), pimenta malagueta (16.7%), alfavaca (15.9% ), chicoria (13.0%), mastruz (12.3%), pio branco (10.9%), and coirama (10.9% ). 14 o f the species identified occurred at only one household. Of these most common species, it must be noted that the majority of these species are also used in contexts unrelated to healing or magic. For example, capim cheiroso (lemon grass ) is often prepared with hot water to make tea. Pimenta malagueta the third most common species, is most frequently used as a condiment or prepared in hot pepper sauce. Chicoria and alfavaca are also used as spices in soups and fish stews.

PAGE 132

132 To describe these plants as str other roles and uses in daily Amazonian life. The varied use of these species speaks to ambiguous nature of many magical or symbolically signific ant plants The Ambiguous Nature of Plants Used for Magic and Healing When I first learned about pio roxo and began questioning farmers who had the plant in their yard, I found that the reasons for its existence on the property often varied. While visiti ng Pedro, a newcomer to the agricultural settlement outside of the town of Borba, I started chatting with his wife about magic pl ants to see if she had any interesting insights. She said that pi o roxo was one of the pl ants that she knew was used to ward off the evil eye but she told me that had simply appeared there When I asked other informants about pio roxo after spotting it in their yards, many also told me that the plant appeared without their having intent i onally cultivated it. Since pi o roxo is a weed y species planted, and can often be found as a volunteer in homegardens. However, if the plant was completely unwanted, it would likely be re moved since most households keep the are a surrounding the house ( known as the terreiro ) cleared to prevent snakes or other pest s from having a place to hide. In this way, pio roxo is often adopted through a passive form of cultivation that is common practice in much Amazonian gardening. The ambiguity tha t exists in the way in which pi o roxo is cultivated mirrors the same ambiguities that exist in the re asons behind its propagation. Some fa milies said they had planted pi o roxo in their yard specifically to protect the home from the evil eye while others said that they used i t in banhos or in home remedies Some people rejected the idea that they would have planted it to protect the

PAGE 133

133 although I suspected that it might have been one of the reasons the y planted it I later found several other plants that I placed in these categories that were cultivated for more than one reason or purpose. A s described above plants that were used to protect the household can also be used medicinally or in healing baths It can be argued that the power of a plant to heal is directly linked to its power to ward off the evil eye. Mucura ca for example, serves many purposes as it is used in healing baths, but also kept to ward of the evil eye mpress on a wound. Because of the ambiguities described above, it is difficult to create clear cut categories of plants that are used strictly utilized for magical reasons and others for homeopathy. These ambiguities als o relate to the fact that traditi onal healing and magic are interchangeable in many societies in the world. Since the causes of many illnesses are not easily observable outside of Western science they are attributed to supernatural forces. And in many societies, such illnesses are comb ated with magic worthwhile to recall Evans among the Zande : the word for (magic) plant and medicine was one and the same. As I was writing notes on an ea rlier draft of this chapter, a neighbor ca me over to pick some le aves of j apana from the backyard. The owner of the house and my friend, Dani had numero us plants in the backyard that we re used in baths and healing practices, many of which she had inherited from her mother. Dani denied using plants to ward off the evil eye, and said that she used them simpl y as herbal remedies. But I reminded myself that Dani was a nurse in town and probably had a d ifferent relationship to healing and medicine which perhaps made her feel uncomfortable entertaining the notion that her use of plants could be more closely related to ritualistic and spiritual

PAGE 134

134 practice than to Western biomedicine Although s ome of the plants described in this chapter have known medicinal benef its in Western science I would argue that most are used today b ecause of their socio cultural value in the practice of healing. Yet re gardless of the way they work their magic, on numerous occasions, I found friends, neighbors, and acquaintances stopping by Dani to pick a few leaves or spri gs of plants to bat he sick children or help an ailing relative Magic Plants and Religion The relationship between religious belief and the use of magic and healing plants is also critical to this discussion. At the onset of this chapter, reference was made to a description of the use of magic/ medicinal plants in Folk Catholicism a blend of Catholicism with Afro Brazilian and Amerindian relgious belief and practice Ma ny of the communities I visited had a loose affiliation with the Catholic C hurch and arguably practiced some form of In these communities, the use of plants to ward off the evil eye, for example, was not uncommon. However, on a visit to a community that had adopted an Evangelical Christian church I found a very different relationship to magic plants. When I asked about their use one household head told me matter of factly: that we know the true word of the Lord, we no longer beli eve in those things (Fieldnotes June 2010). In my interactions with other Evangelical Christians, I noted a similar attitude in which the use of magic plants was deemed superstitious, or perhaps in some cases, even un Christian. Rifts between Evangelical Christians and Catholics led to the fissuring of more than one community I visited and the use of magic plants is just one of many practices that demonstrate such divisions. Much could be written about tensions between these groups, but what is of particular interest here in the potential effect religious belief and practice may have on

PAGE 135

135 agrobiodiversi ty. If the tendency of Evangelical Christians is to abandon the use of plants associated with folk Catholicism it could lead to a serious decline in the diversity of species managed in rural communities. Yet it is also equally possible that those specie s of plants associated with Folk Catholicism will continue to be cultivated by Evangelical Christians while their uses, and the rationale behind such uses, are adapted to their new beliefs. Conclusions The use of magic plants in Amazonian communities repr esents more than an anthropological novelty; it has important implications for understanding the maintenance of Amazonian biodiversity. This chapter demonstrates that many of the plants managed on smallholder farms in Borba are valued not strictly because of their direct economic or material benefits, but because of their symbolic use and role in socio cultural practice. If conservation of Amazonian agrobiodiversity is truly a concern of non Amazonians, then the socio cultural value that Amazonian peoples attach to such useful plants demands recognition and consideration. Although some conservation minded entrepeneurs have introduced products derived from Amazonian plants into foreign markets in the hope that they may support and sustain their use, this is a less than reliable solution for supporting the bulk of Amazonian agrobiodiversity. A more reasonable focus may be a turn toward understanding and encouraging practices of Amazonian communities that sustain agrobiodiversity, including the use of magic p lants. It is reasonable to believe that the social practices and cultural values associated with magic plants are critical to the continued maintenance of Amazonian agrobiodiversity.

PAGE 136

136 Figure 5 1 Pi o Roxo a plant used in healing baths and to ward off t he evil eye Figure 5 2 Rue, an Old World plant used to ward off the evil eye

PAGE 137

137 Figure 5 3 Poraqu the electric eel plant used to protect oneself when playing soccer

PAGE 138

138 Figure 5 4 Mucuraca the possum plant, commonly used in healing bat hs

PAGE 139

139 Figure 5 5 Chumbinho receives a bath from his cousins (Puruzinho, August 2010)

PAGE 140

140 Table 5 1 List of species used in magic and symbolic practice in Borba, AM, Brazil Family Scientific Name Local Common Name English Name Native / Exotic (1/0) Origin Magic Use Other Uses #HH Present Agavaceae Sansevieria trifasciata Espada de So Jorge Snake plant 0 Old World (Africa) repel ornamental 2 Amaranthaceae Alternathera tenella Cuia mansa 1 South America banho 1 Amaranthaceae Chenopodium ambrosioides Mastruz Wormseed 1 Mesoamerica banho medicinal 17 Annonaceae Annona montana Araticum 1 Amazonia banho food 6 Apiaceae Eryngium foetidum Chicoria 1 barnho condiment 18 Apocynaceae Thevetia peruviana Castanha da India 1 Andes repel (snakes) ornamental 4 Araceae Caladiu m sp. Cachorrinho 1 New World Tropics A ttract (people) ornamental 1 Araceae Dieffenbachia seguine Comigo ninguem pode 1 Neotropics repel ornamental 1 Araceae Scindapsus aureus Jiboia Ivy 0 Old World (Asia) attract (people) ornamental 1 Araceae Unident ified taja buceta repel (snakes) ornamental 2 Araceae Caladium bicolor Vence tudo Caladium 1 South America banho ornamental 1 Araceae? Unidentified Poraqu ? ? repel ornamental 1 Aristolochiaceae Aristolochia sp. Ueca' 0 Old World attract (luck) or namental 1 Asteraceae Vernonia condensata Boldo japons 0 Old World banho tea 1 Asteraceae Tagetes spp Cravo; Cravo de defunto Marigold 1 New World banho ornamental 4 Asteraceae Spilanthes acmella Jambu 1 Amazonia banho food 7 Asteraceae Ayapana sp. (formerly Eupatorium ayapana ) Japana 1 banho 9 Asteraceae Pluchea sagittalis Marcela 1 Americas banho medicinal 4 Bignoniaceae Mansoa alliacea Cip alho Garlic vine 1 Amazonia banho 9 Bignoniaceae Arrabidea chica C rajiru 1 banho medicinal 3 Comm elinaceae Callisia repens Dinheiro em penca 1 Neotropics attract (money) ornamental 2 Costaceae Costus spicatus Pobre Velho; cana mansa 1 South America; banho 2

PAGE 141

141 Table 5 1. Continued Family Scientific Name Local Common Name English Name Native / Exot ic (1/0) Origin Magic Use Other Uses #HH Present Crassulaceae Kalanchoe pinnata; K. brasiliensis Coirama; escama de pirarucu 0 Old World (Africa) banho medicinal 15 Euphorbiaceae Euphorbia milii Coroa de cristo Crown of thorns 0 Old World repel ornament al 1 Euphorbiaceae Jatropha podagrica Pio barrigudo Gout plant; Buddha belly plant 1 Mesoamerica repel medicinal 5 Euphorbiaceae Jatropha curcas Pio branco 1 Mesoamerica Repel ; banho medicinal 15 Euphorbiaceae Jatropha gossypifolia Pio roxo Bellyach e bush 1 South America Repel ; banho medicinal 41 Euphorbiaceae? Unidentified Caboclo roxo ? ? repel 1 Lamiaceae Ocimum micranthum Alfavaca Amazonian basil 0 Neotropics banho condiment 22 Lamiaceae Aeollanthus suaveolens Catinga de mulata 1 banho med icinal 4 Lamiaceae Mentha x piperita; Mentha sp. Hortelazinha Mint 0 Old World (Europe) banho food; medicinal 11 Lamiaceae Ocimum basilicum Manjerico Basil 0 Old World banho food 10 Lamiaceae Pogostemon cablin Oriza; Uriza Patchouli 0 Old World (Asia) banho 1 Malvaceae Gossypium barbadense Algodo Cotton 1 South America banho utilitarian 11 Monimiaceae Peumus boldus Boldo 1 South America banho tea 9 Monimiaceae Siparuna guianensis Capitiu; caa pitiu Fevertree 1 Neotropics banho medicinal 2 Phytola ccaceae Petiveria alliacea Mucura ca 1 Neotropics banho 13 Piperaceae Piper callosum Paregorico; Panquile; Oleo eletrico 1 South America banho medicinal 2 Poaceae Cymbopogon citratus Capim cheiroso; Capim santo Lemon grass; Citronella 0 Old World banh o tea 26 Poaceae Vetiveria zizanioides Patichulim Vetiver(?) 0 Asia banho 1 Portulacaceae Portulaca pilosa Amor crescido Kiss me quick 1 Americas banho 11 Rutaceae Ruta spp. Arruda Rue 0 Old World repel 12 Solanaceae Capiscum frutescens Pimenta mala gueta Hot pepper 1 South America repel food 23

PAGE 142

142 Table 3 2. Continued Family Scientific Name Local Common Name English Name Native / Exotic (1/0) Origin Magic Use Other Uses #HH Present Unidentified Unidentified Jiru ? ? repel ornamental 2 Uni dentified Unidentified Maria bonita ? ? attract ornamental 1 Unidentified Unidentified Vai vem ? ? attract 1 Urticaceae Pilea nummulariifolia Mutuquinha 1 Neotropics banho 4 Zingiberaceae Zingiber officinale Gengibre; Mangarataia Ginger 0 Asia ban ho food; medicinal 7

PAGE 143

143 CHAPTER 6 GENDE R AND AGROBIODIVERSIT Y MANAGEMENT IN RURA L AMAZONIA Introduction Among the many socio cultural factors influencing agrobiodiversity, the relationship of gender to the maintenance and distributio n of useful plant speci es is of critical importance Gender can have a particularly strong influence on plant diversity in Amazonian societies where management of species diverse homegardens is often the responsibility of women (e.g., Descola 1996 ). While data presented here s how that th e gender of the household head had no significant bearing on the species richness at the household level, agrobi odiversity management within households is still very much influenced by gender and gender roles. In this chapter I highlight the c es to agrobiodiversity while discussing the relationship of agrobiodiversity management to gender divisions of labor and economic production. I point out that even though the generalization that roa (manioc can apply to many households, this generalization also obscures a great deal of variation that exists among households. Looking specifically at the cases of three housesholds, I explore and problematize the relatio nships between gender, household division of labor and agrobiodiversity management. Background: Gender Inequality, and Agrobiodiversity In many analyses of gender in agricultural societies in Latin America and other parts of the world, it is shown that women are typically bou nd to the domestic sphere where as men often have greater liberty to operate in the public sphere. For this

PAGE 144

144 management has been said to go unseen: e most plant use, management and conservation occurs within the domestic realm, and because the principal values of plant genetic resources are localised and non (Howard 2003a: 2 ). It is argued that scientific resear ch on agrobiodiversity has also suffered from management in the domestic sphere (Howard 2003a : 2 ). Gender hierarchies and inequalities within societies often obscure, underp contribution to agricultural production, agrobiodiver sity, and agricultural knowledge. As societies where men po w representatives of the household in the public sphere. It is said that, subsistence work related to plants is intended 2003b: 6; italics from original text). Looking specifically at gender a nd agrobiodiversity management in Brazilian Amazonia, a case study by Murrieta and Winkler Prins (2003) in the Eastern Amazonian community of Ituqui showe d that men dedicated most of their efforts to cultivating large fields of manioc and other staple crop s like bananas while w omen were largely responsible for management of the quintal (backyard) and jardim (garden) beside their homes It was observed that although women helped men in the manioc fields, homegardens were nearly exclusively managed by women which they described as the see also WinklerPrins and Souza 2005).

PAGE 145

145 In another study from rural Brazilian Amazonia by Siqueira et al. (2007) it was reporte d that women worked in the mani oc fields, planting, harvesting, and processing manioc flour, however their participation in commercialization or marketing was It was further noted that even though the majority of women were active in the household economy, 65.6 percent stated that their main activity was (Siqueira et al. 2007: 188 ) In many studies of contemporary Amazonia, the partriarchal norms characteristic of rural Amazonian c ommunities are shown to in the shadows Yet despite t hese I ural production, and their management of agrobiodiversity, varied considerably in the households that I visited in Borba. Rather than finding one generalizable pattern, I found many different patterns of agricultural management in relation to gender that could be described and analyzed Th is chapter serves as an attempt to sketch out some of these patterns and their relationship to the broader management of Amazonian agrobiodiversity. Species Diversity by Gender of Household Head(s) Of the households inte rviewed, female household heads were interviewed at 33 households, male household heads were interviewed at 78 households, and both male and female household head s were interviewed at 27 hou seholds. Men were clearly over represented in surveys and intervi ews in large part because it was more common for men to address me when visiting communities and individual household s Despite this skewed sample descriptive statistics demonstrated little difference in the mean

PAGE 146

146 number of species managed by the househol d s based on the gender of the household representative Households with female heads interviewed average 19.4 species 7.6 while at households with male heads the average was 19.6 species 8.6 with no significant difference between the means (t=.145; p=.866). At the households with both male and female heads interviewed, th e average number of species was 19.3 8.3 with n o significant difference between households with only males interviewed (t=.144; p=.886) or only females interviewed (t= 0.015; p=.988). This simply demonstrates that the gender of the household head interviewed had no significant effect on the documented species richness managed by the household. Despite this it was made apparent that within many households, women and men often managed different plants and usually did so in different spaces or areas of management Relying on et hnographic data and observ ation I explore the variation in patterns of management in the roa (manioc field) and the homegarden by men and women in three distinct households The Roa and the Homegarden Case 1: Dona Nanda and Nilo While Nilo worked to clear weeds from his roa ( ma nioc field ) Dona Nanda who suffered fr om the onset of cataracts, stayed at home. Despite not being able to get around as well as she once did, Dona Nanda still tended to several species of plants near her home including several medicinals and ornamenta ls that she planted and managed in her yard. Unfortunately, I arrived after a record flood and Dona Nanda lamented that many of the plants she had acquired over the years had died. She said she had nurtured a colorful variety of flowering plants, some o f which she had brought

PAGE 147

147 back from the city of Borba, but the river had drowned nearly all of them. She confessed that she felt like giving up on her garden, although I suspected that such a declaration was made more out of frustration than actual intent. destruction, I still noticed a few bright ornamentals surrounding Nilo and home including a pink crape myrtle in the front of the house and the intensely orange French marigolds that caught my eye along the side of the yar d ( Fieldnotes, October 2009; Figure 6 1) Chatting with Dona Nanda over the course of several days, it seemed that she often downplayed her activ ities in the home significant or me rit the interest of a foreign researcher. Because she often discussed her cultivation of plants in a casual manner, her contribution to agrobiodiversity demanded greater attention to fully grasp and appreciate. A fter a day or two of poking around in her garden, I realiz ed that many of the plants that she nurtured had been tucked away or hiding underneath my nose. I learn ed, for example, that the plants in the backyard growing out of paint buckets and small clearings were not opportunistic weeds, but in f act medicinal herbs, some of which were being used to treat a visiting friend that was ill ( Figure 6 2 ). Nilo casually pointed out that Dona Nanda manjava um pouco their use in healing. When I heard this, I asked Dona Nanda more about plants she used in healing, and only then did I learn i that I had overlooked in my original sur vey when I had tallied species in the garden with Nilo Although I originally intended to avoid ornamentals and medicinal plants in my

PAGE 148

148 but recognize their contribut medicinals in the homegarden can be overshado crops that are typically planted in m ore conspicuous roas and orchards. I also noticed that during interviews and surveys, my interlocutors often assumed that I was only interested in the cultivation of the primary economic crops typically managed in large, less species diverse fields even when I tried to impress upon them that any little plant homegardens. This tendency in reporting made me think more about the attention and prestige given to cash crops managed in large fields while less lucrative useful plants, many of which were cultivated by women closer to the home were ignored. I also considered further how gender and household economics articulated with agrobiodiversity and my perception of its ma nagement. As in the case of Dona Nanda and Nilo I often observed that species rich homegardens were managed by wome n, while men were primarily responsible for the production of market and staple cro ps in large fields and orchards. Yet I also noted that w omen commonly planted manioc even when faced with other responsibilites around e household economy varied depending on the social dynamics within the household and the degree to which men were comfortable with roa and (or that men are the breadwinners and women take

PAGE 149

149 care of the home ) coul d be applied in many cases, it also obscure d the great deal of variation that existed in the communities I visited and the households with who m I stayed Case 2: Cndida and Jos One of the households where I spent the most time during my 12 months of re search was in the home of Cndida and Jos who lived in a floodplain community on Puruzinho Lake. Thinking about gender roles in agriculture one afternoon, I asked Cndida about her perspectives on women working in the homegarden and the roa : Cndida : I boring. I prefer to work in the roa . Nick : Do most women work in the roa or do they stay at home? Cndida : M ost work in the roa Nick: What do you like to plant? Cndida : I li ke to plant things in the vegetable garden O nions, cilantro, bell peppers. This year I had a nice little canteir o but the flood carried it away. ( Fieldnotes December 9, 2009 ; translation mine ) Cndida that the roa is a predominately male space. She herself even claimed to prefer work in the roa over Cndida while she said that she preferred to be out working in the manioc f ields, when asked about what things she liked to plant, she listed crops that are typically cultivated close to the home in a canteiro The canteiro is a common feature of most Amazonian homegardens, often an old canoe that is filled with rich black soil s to plant chives, cilantro, leafy greens, and spices ( Figure 6 3 ). At the time I interviewed Cndida her husband Jos was building another

PAGE 150

150 canteiro so that she could plant chives. She also had tomatoes planted in an area cordoned off by fishnet to prot ect them from her chickens. She said that on the grows, all you have to do is During my stays at Cndida and Jos Cndida adopted new plants from friends and kin that she plant ed in the yard or the canteiro take root. In the academic literature, homegardens are justifiably seen as important sites of experimentation ( Murrieta and WinklerPrins 2003 ; Stoen et al. 2009 ; Trinh et al. 2003 ) Unique plants acquired t hrough friends or relatives are planted in areas where they can be closely monitored. Women may also experiment with use of mulch and other soil amendments in canteiros and thus continually build on their knowledge of agriculture. In some instances, suc cessful experimentation with a particular crop or soil amendment can translate into a wider application in fields or orchards, thus making homegardens important agricultural laboratories. As a result of this process of tend to harbor a high degree of plant species (WinklerPrins and Souza 2005; Zimmerer 2006). Cndida seemingly derived pleasure from experimenting in her canteiro and tending t o her vegetable garden, but manioc field s Jos Cndida in her work, but he pointed out to me that it was Cndida who had received a loan through the state agricultural extension agency to pl ant manioc and she w as the primary caretaker of the ro a not him. Since Jos was successful selling other agricultural produce and diesel fuel in the community and served as the community president, he

PAGE 151

151 manioc produc tion and her potential economic independence. Instead, he seemed to encourage it, but clearly from a privile ged economic position that allowed him to be comfortable in doing so. In Cndida thus be seen that even though women typically take care of little gardens and canteiros with chives, cilantro, peppers, and tomatoes, women can also be the primary caretakers of the household ro a The converse of this case also exists as there are many men that I interviewed who took to planting herb ga rdens to sell cheiro verde (a mix of cilantro, chives, and chicoria ) and cab bage in the city of Borba For these reasons, I found that gender division s in a griculture among households in Borba could n ot be easily pinned down S ome women worked with their husbands in the ro a while others stay ed close to home and tend ed to a vegetable garden or medicinal plant garden; still many ot hers, like Cndida did both. Case 3: Alcia Al c ia who lived on the opposite end of Puruzinho Lake from Cndida provided a di fferent perspective on gender roles in agriculture and the inequalities found in the division of labor in many households Alcia said that women and men work ed but she emphasi zed the point that their work was also different. According to Al cia men cleared the forest for new fields, burned the dried brush and prepared the fields for planting. Women, on the other hand, planted manioc, weeded the fields, and harvest ed the manioc. Alcia told me that while men were out fish ing women ha d domes tic activities: preparing food, taking care of children, cleaning clothes, and doing dishes. And finally when men ca work d id Feb 2010).

PAGE 152

152 Alcia alluded to having problems with do mestic abuse, and I later found that it was common in many households, Without hestitation, Alcia nonchalantly told me that me n often beat or threaten ed to beat their wives if they ever suspect ed th em to be unfaithful even while the men themselves were guilty of their own infidelities. Alcia need af fection. Sometimes a man thinks it s enough to just provide food and work then he goes out at night. e his wife any affection or attention that she na rua ) to find it. Men have Alcia said she had problems in the past with her husban d beating her or trying to beat her but she said that things had improved over time, old er now (Audio recordings Feb. 2010) In direct contrast with Dona Nanda who m I described at the beginning of the chapter as spending the majori ty of her time close to hom e tending to her medicinal plants and herbs, Alcia dedicated herself to planting manioc and working out in the roa When I asked Alcia about her homegarden, she said she planted a few tomatoes and spices but otherwise she in tak little plants the house ro a Talking with Alcia I could tell that she took pride in her work in the ro a After all, the farinha she produced constituted the bas ic foundation of subsistence Although Alcia never articulated this directly, I suspected that her independent character and self reliant nature may have also threatened her husband, and could in some part explain h is desperate use of vi ol ence a gainst her I n her daily work in the manioc fields, she demonstrat ed her ability to successfully support her family, which

PAGE 153

153 may have contributed to a struggle for power within the household both psychological and physical This struggle may have been exacerbated by Alcia dissatisfaction with domestic chores and her willingness to challenge the notion that the She, like Cndida seemed to prefer to be out in the ro a Discussion and Conclusion s In most rural Amazonian communities like those examined here, it can be generalized that women have greater responsibilities in domestic activites and as a result will often work i n the homegarden, cultivating ornamentals, medicinals, magic plants, and s pices However, in many households, women also plant manioc and other staple crops that can contrib ute to household subsistence and potentially provide important cash income In some households economic production by women is downplayed or overlooked as men may attempt winners women themselves may fail to recognize their contributions to household agricultural and economic production. However, in other households men and women can play complementary roles helping one ano ther, or encouraging independent projects Considering the different ways that Dona Nanda Cndida and Alcia manage plants and divisions of labor vary greatly in rural Amazoni an households. These findings have important implications for the study of agrobiodiversity in rural and their contribution to overall agrobiodiversity requires greater attention and investigation. In th e past, agricultural development programs and conservation policies focused on market production of cash crops that was typically dominated by men, but it is recognized that

PAGE 154

154 women can often manage a larger portion of diversity in homegardens. Although hom e gardens are typically less designed for market production (although not always), they nonetheless represent important resources for household subsistence, construction materials, and medicines. Moreover, homegardens can serve as sites of agricultural exp erimentation that can later yield useful varieties that have market potential. Second, since s roles vary, their management of agrobiodiversity can not be easily generalized. Many case studies argue that women are responsible for managing a greater portion of household plant diversity than men, but in rural Amazonia this is true in some cases and not in others. Futures i nvestigation should attempt to understand this variation in the gendered patterning of agrobiodiversity management and how it may be changing Beyond variation in roles related to agrobiodiversity management, gendered division of labor in agriculture and its relationship to household economics and gender inequalities is critical to future research. In many areas of the developing world, the production have led women to move out of agriculture as Howard (2003b) notes: Women throughout the global South and in traditional communities in the North str uggle to maintain their livelihoods, their cultural integrity, status and biological wealth, but are very often disadvantaged by gender inequalities and bias with households, economies and political systems. It is little wonder that many mothers want their daughters to abandon traditional occupations and knowledge, or ostensibly to enjoy the advantages of the abandoning the countryside to improve their status (p.42). If current concern over agrobiodiversity loss is to translate into effective conservation, the gendered nature of agriculture and agrobiodiversity management and their relationship to gender inequality must be addressed Where women have the tendency

PAGE 155

155 to manage a higher diversit y of plant species than men, they must receive appropriate recognition and potential compensation for their contribution to agrobiodiversity maintenance. If developed industrialized nations are truly concerned over the loss of biodiversity and agrobiodive rsity, then they must be willing to provide resources and implement programs that provide incentives, economic or otherwise, for rural farmers to maintain such diversity. Lastly, if such programs are to be effective, they must recognize and address existi ng social and gender inequalities in ways that are flexible and culturally appropriate. Figure 6 1 Dona French marigolds

PAGE 156

156 Figure 6 2 Rue planted in a paint bucket in Dona yard.

PAGE 157

157 Figur e 6 3 Cante iro on the Autazes A u Ri ver (August 2010)

PAGE 158

158 CHAPTER 7 AGROBIODIVERSIT Y, AMAZONIAN HISTORI CAL ECOLOGY, AND FINAL CONCLUSIONS I made my first trip to Borba in July of 2003. At the time, I found nothing more interesting about the trees lining the Madeira Rive r than those found on a wooded road was wide like the Mississippi, made the forest on its banks appear rather insignificant by Amazonian forest to be particularly awe inspiring, at least not from the deck of a boat. However, on my second trip in 2007, after I had taken a general course in botany and learned more about the history of human occupation in Amazonia, the seemingly monotonous and unremarkable landscape began to unfold and reveal more about itself. I started to notice the Breadfruit trees with their distinctly toothed leaves that appeared in seemingly undisturbed tracts of floodplain forest ( Figure 7 1). I learned that these were living artifacts of the Columbian exchange, having been introduced from the Old World along with mangos, bananas malay apples and numerous other fruit trees that dot the contemporary Amazonian la ndscape. Originally hailing from New Guinea B readfruit trees were naturalized in the Amazonian forest, and they reminded me of the earth shattering revelation, it repr esented one tendril of the many tangled and overlapping histories found within the forest that were appealing to me. In this way I began to see beyond the brown and green monolithic forest, and I developed interest in the stories of individual trees, thei r habits and lineages, as well as the suites of economic plants that had arrived with Portuguese colonists, immigrants from Northeastern Brazil, and even early indigenous migrants that hailed from other reaches

PAGE 159

159 of the Americas. I was intrigued by the hist orical and social forces that helped shape the contemporary Amazonian landscape, and I was thus exposed to the ways in which River itself. As I explored the region by l orchards, and terra firme palm stands, I found subtle clues of past human occupation and disturbance that dated back in some instances to more than a thousand years ago. Contemplating the deep historical re lationship between humans and the Amazonian environment, and the unique ways in which humans and the natural environment interact I became more and more interested in studying Amazonian historical ecology and its relationship to the contemporary botanical diversity manag ed on rural Amazonian farms. Building off this interest in deep rooted h uman environmental relationships, I describe here the biographies of rural smallholders and the histories of the fores ts and gardens managed by their familie s in the mu nicipality of Borba. I stress the point that agrobiodiversity is inherited and passed on wi thin families. In some cases, inherited botanical diversity is exploited for subsistence or commercial production by contemporary smallholders while in other cases it is overlooked or even c leared to make room for new fields. In discussing the inherited and historical nature of agrobiodiversity in Amazonian forests, I dissect the common dichotomy between fore sts implied in many studies of historical e cology to describe forests without known histories Using an alternative perspective, I draw on

PAGE 160

160 Deleuze and Gua rgues f or the analysi s of landscapes in state s Translating these perspectives into practical application, I argue that historical ecological research should draw on lines of intersecting evidence from archaeology, ecology, o ral history, and even mythology to gather information about the known histories of Amazonian forest s that can contribute to dynamic maps of knowledge To conclude, I expand on these perspectives for new mazonian landscape and its agrobiodiversity. Inherited Agrobiodiversity After spending a hot morning conducting surveys with some members of the floodplain community of Santana, I took a break with my research assistant Maciel. Nearby, we s potted Antoni o a young man we had met on our first visit to the community a few days prior. We invited Antonio to drink some soda and eat a few cookies with us front of us stood a cacao grove with intermingled rubber trees that shaded the cacao from above. Out of curiosity, I asked Antonio if he knew how old the cacao stand was. He said it was more than 20 years old and that his grandfather had planted the cacao trees, more tha n 2 hectares in all, as well as the rubber trees that loomed over them. Since Antonio Antonio made part of his living off the cacao his grandfather had planted. He collected the pods, removed and sun dried the beans, and th en sold them to a local buyer that shipped them on to the regional capital of Manaus. My assistant, Maciel, jokingly said that his grandfather had been lazy Antonio had inherited. We all laughed. Maciel continued by saying that when he died, he hoped that

PAGE 161

161 his grandkids would remember him because he will have left them something behind, maybe even a large orchard from which his grandchildren could make a living from. Then, perhaps somewha t ironically, Maciel mused aloud that his grandfather had probably been more preoccupied by the thought of his own death than with the future of his family (Fieldnotes: June 10, 2010) The contrasting cases of Antonio and Maciel show that while the inherit ance of land is important to rural Amazonians, the plants that are being passed down with the land can be of equal or even greater importance. Snooping around homegardens, I found than many women had inherited plants from their mothers, including one of m y close friends, Dani In her backyard, she pointed out a Starfruit tree, the herb vence tudo and a gingeira among several other plants, and told me that all of the plants had Yet b eyond the physical plants themselves and the successive generations derived Dani also inherited knowledge regarding the use, cultivation, and management of plants Inheriting a diversity of plants is clea rly useless unless the requisite knowledge to properly manage and utilize them is also passed on to sons, daughters, and grandchildren. Like Dona Nanda in chapter 6, Dani told me that she learned how to use medicinal and healing plants from her mother whe n she herself had her own children. Today Dani continues to use the specialized knowledge of their usage, sharing it with neighbors and friends while nurturing the plants that were passed ignificant effect on the larger Amazonian landscape, but over time, continual selection and cultivation of

PAGE 162

162 individual species can eventually lead to a distinct patterning of agrobiodiversity, and thus requires a historical reading. Historical Ecology Biod iversity, and Anthropogenicity In response to the environmentally deterministic bent of earlier Amazonian studies (e.g. Meggers 1996), a great deal of current Amazonia research has investigated the specific ways in which the environment reflects historical management and intervention by humans ( Bale 1994, Clement 1999, Erickson 2005 Heckenberger et al. 2008 Neves et al. 2003 ). The research program of historical ecology has played a critical role in this movement as it emphasizes the recursive nature of human environmental interaction, and argues that humans do not simply adapt to their surroundings, but play an active role in shaping them through time (Bale 2006) Human e nvironmental interaction is conceived therefore as a dialogical and co evolutionar y process. One of the key questions that has driven research in historical ecology is the effect of past human activity on contemporary biodiversity. Within this research the historical influence of human activity on botanical diversity has received pa rticular attention. In Amazonia and neighboring regions, research by Bal e has shown that historical human activity had a profound effect on the species composition of contemporary forests. In Bale outline d different forms of anthropogenic fores ts that can be found across Amazonia including palm forests, bamboo forests, B razil nut groves, and liana forests (Bale 1989). Bal e argues that historical human management and selection of economic plants contribu ted to the formation of anthropogenic forests in more than 11% of Amazonia (ibid.). Studies by the late Darrell Posey among the Kayapo have also been hugely

PAGE 163

163 management of sava (Posey 19 85 ). Similarly, i n Central America it has been demonstrated that the Yucatec Maya created forest gard species for management while excluding others (Gomez Pompa et al. 1987) In another case study in the Peten of Guatemala, that much of the contemporary forests in that region are anthropogenic, resulting from active encouragement and selection of valued economic species by the Late Classic Maya (Campbell et al. 2006 ). Much of the research in historical ecology has demonstrated that environments fact reflect evidence of deep historical relationships with humans (e.g. Denevan 1992 b ). This research, however, has been critiqued for over emphasizing human influence on the environment and making the related assumption of human intentionality in enviro nmental transformation (see Bal e and Erickson 2005). In their rejection of environmental determinism, historical ecologists have occasionally overextended their argument in support of human agency by characterizing environmental phenomena that are potent ially by products of human activity as For example, Erickson suggests the anthropogenic soils known as Amazonian Dark Earth are the result of (Erickson 2008: 171). Yet few researchers dedicated to th e study of Amazonian Dark Earth have ventured to make this cl aim unequivocally (see Eidt 1977 ; Neves et al. 2003 ). In a separate publication, Erickson further argu es

PAGE 164

164 environment as an indigenous creation is much more useful and accurate than the more Erickson 2005; italics from the or iginal text). Curiously, Erickson criticizes other anthropologists for treating the environment as static, yet his own research privileges the role of humans in shaping the environment s to describe enviro nments as he overlooks the fact that many indigenous groups do not necessarily view their management as b eing intentional (Posey 1992). s in which human actions have uniquely shaped the environment, I also fear that such fascination with the human footprint on the Amazonian environment can mislead, or g For example, th and ignor es the roles of other organisms in the creation and existence of the ecological community found in such forests This denie s the work of pollinator s, soil microbes, and decaying bodies that contribute humic matter and release nutrients crucial to the forest human relationship to the Amazonian landscape and the agrobiodiversity found therein, and problematized by many anthropologists working in Amazonia (e.g. Descola 1996), but nonetheless continues in anthropological analyses. One November morning while visiting a plot of fores t with my friend Nilo I came t o revisit this problem.

PAGE 165

165 As the sun began to rise over the Madeira River, Nilo and I took a small canoe with a noisy 5 hp rabeta motor to look at the future manioc field he was clearing on the uplands across from the strip of floodplain where he and his family resided. After the 20 minute canoe ride acro ss the floodplain lake, we arrived at an area that was occupied by high old growth forest. Climbing up a steep, gravely bank, we arrived to his plot. Surrounding us were a number of towering rubber trees mixed in with old mango trees and a host of palms c lumped together. Further in the background was a small clearing where Nilo had planned on planting manioc. Nilo explained to me that the land we were standing on had been owned by his grandfather who passed the land down to his father and later on to him ( Figure 7 2 ). He said he wanted to post a small sign in front that rea of the trees that we gazed up at that day. H e showed me stands of andiroba from which h e was eager t o collect fruits so he could make the oil that is highly valued in regional homeopathic medicine and the international cosmetics market The same andiroba oil had also been used by Indians in the past as an equally valuable mosquito repellent. Next to th e future manioc plantation, Nilo also pointed out the clumps of bacaba (bacaba palm stand). Alongside the bacaba palms, we found Brazil nut trees, bacuri, and hogplums ( tape reb ) as well as caiau and urucuri palms, all signs of past human Pondering the diversity of useful plants on Nilo land, I began to ask him which species to his knowledge had been planted by his father and grandfather. He said that his father and grandfather had planted the andiroba and some of the rubber trees which we examined, scarred from years of tappings. Other

PAGE 166

166 cultivated and whic h had not. When I asked about the bacaba palm stands, he said that they were native or naturally occurring, but because of their high concentration, I seem to discount the idea, and contemplated the possibility. How was one to know for sure? Considering what Nilo knew of the biography of that swath of particular forest, it was fair to say that many of the economic species that he utilized and managed on the land were a direct inheritance from his father and grandfather. Deeper into the history indigenous groups of the region. Considering that several stretches of Amazonian Dark Earth e this possibility all the more likely. It was possible that Nilo attracted to the land because of the concentration of valuable plants that had been con centrated in the area because of past management. Pondering the trees before me and thinking about their origins, I began to Historical ecologists have admirably sought to describe and define anthropogenic forests, but the intertwining and complex histories. Was Nilo from only the past 100 or 1000 years? Could we (or should we ? ) distinguish clumps of bacaba palms that were the result of indigenous management or disturbance from those that had no known history?

PAGE 167

167 In examining the biography of a particular tract of forest, it becomes incredibly difficult to excise the natural from the cultural as the processes by which the forest came into being at its present state are inextricable and intertwined. For the study of Antonio Dan i and Nilo have inherited a number of useful plants from the past and that such diversity can be passed over generations like any form of inherited wealth. The degree to which we can usefully distinguish the accreted diversity of social history from that Rather than attempt to distinguish cultural and natural forests anthropogenic forests whether it be archaeological ( known material evidence or archaeological soils ), ecological ( distinct patterns of species richness or species oligarchy), or histo rical ( oral histories associated with the forest, and its use and management over time). These lines of investigation rev eal more about the specifics and eliminate the need for a dichotomous view of forests, other than perhaps forests o f known histories and forests without history. can no longer be assumed that a given tract forest was not managed or touched by human presence in some way, it can only be acknowledged that any knowledge of such interaction is unknown. In our r eliance on the limited historical knowledge of human interaction with the forest, oftentimes the agency of individual trees and lianas that live within the forest are ignore d. This post humanist argument is taken up in a recent analysis of Australian abor ginal use of fire and the adaptation of gum trees to fire :

PAGE 168

168 gatherers made to the landscape must be seen not merely as something that happened to gum trees but also as something that they did Franklin goes on networks of objects and other organisms that define it, and although these may best be thought of as bundles or rhizomes, even these metaphors may be too fo reclosing of the essentially open and fluid nature of its ecological being. This shift in emphasis undermines our tendency to see organisms like trees as mechanisms, suggesting instead that they respond to ev ents and changes in their world 26). Ex panding on this line of reasoning I draw on Dele uze and Guattari metaphor of the rhizome ( referenced her e by Franklin) to offer a different approach to thinking about historical ecology. In the introduction to their book Thousand Plateaus: Capitalism and Schizophrenia Gilles Deleuze and Felix Guattari employ the botanical metaphor of the rhizome as a model for research and theoretical inquiry. They point out that u nlike roots or branches that bifurcate, and are thus dichotomous, rhizomes have the potential to move in a multitude of ways and directions: expanding and contracting, occupying, abandoning, and reoccupying spaces. In outlining the characteristics of the rhizome, Deleuze and Guattari stress princi ples of connection, heterogeneity, and multiplicity: he wisdom of the plants: even when they have roots, there is always an outside where they form a rhizome with something else with the wind, an animal, human beings (and there is also an aspect under which animals themselves form rhizomes, as

PAGE 169

169 plateau is always in the middle, not at the beginning or the end. A rhizome is made of plateaus. Gregory a continuous, self vibrating region of intensities whose development avoids any plateau are metaphors t hat can be employed to remind us that while we may look for n relationships with it, we can not approach our inquiry into these histories as having either a beginning or an end. of intersecting, overlapping, and on going histories. Drawing on distinct lines of evidence as outlined in this chapter is helpful for illuminating some of these histories, but we can never naively assume that we will approach a total understanding of hu man relationships with a given tract of forest (if we were to arbitrarily cordon one section off) since our knowledge is always partial and fragmentary. To aspire to such knowledge would miss the point, as Deleuze and Guattari note that there Unlike the graphic arts, drawing, or photography, unlike tracings, the rhizome pertains to a map that must be produced, constructed, a map that is always detachable, connectable, reversible, modifiable, and has multiple entryways and exits and its own line General and without an organizing memory or central automaton, defined solely by a scussing the work of Deleuze Escobar (2008) reite a static collection of beings that know 127). Pon dering the

PAGE 170

170 I was eventually struck by the fact that Amazonians ha ve their own folklorical variation on this metaphor: the mythical snake known as Cobra G rande Cobra Grande and the Evolving Amazonian Landscape On a Saturday morning in Novem ber, I accompanied friends on a fishing trip to Lago Comprido, a seasonal lake f ound on an island in the middle of the Madeira River. When we arrived, we crossed swamps on fallen lo gs and then made a short canoe trip through a shallow flooded section of forest before arriving to a temporary camp that was situated on relatively stable ground. After getting settled in camp and starting a small fire, some of us went to search for a seasonal poo or small pond that we hoped would be full o f prehistoric looking bod fish for our afternoon meal As we set off towards the poo, we entered a gnarly bamburral full of fallen palm fronds with needle like spines that stuck in our feet. We then came upon a deep dried creek bed that looked like a canal. We descended into the creek bed and followed its course. While walking my friend Sapo off h andedly mentioned that the canal we were passing through was a trail created by Cobra Grande, the mythical snake of caboclo and indigenous oral tradition. Then Sapo stepped on a sucuri anaconda. I laughed, thinking Broca was joking with me, the gullible gringo. I turned around, slowly walking back towards Broca, but before I could appreciate his joke, he stuck his hand down in the mud and unearthed a snake, about 3 feet long. It was, after all, a small anaconda. Sapo grabbed the snake from his brother, held it up, and told hind by their mom, Cobra Grande (Figure 7 3).

PAGE 171

171 Cobra Grande is a massive snake, much larger than the largest anaconda, and in shifting topography and hydrology. Seasonal fluctuations in precipitation results in constant shifts in the size and form of regional waterways, and the accumulation of silt on an odd sandbar can sometimes lead to the formation of an island over night. In other cases, a shallow spot on the edge of the floodplain may some times form an enclosed pond, trapping fish like the bo that hangout in muddy bottoms. As Hugh Raffles documented in his thoughtful book In Amazonia a number of Amazonian waterways are carved out by rural Amazonians, some being simply cleaned or widened while others are laboriously excavated by community work parties. Amazonian waterways are thus defined by processes that are both natural and cultural. Curiously, when the origin of a channel or furo is ambiguous, or no known history is associated with its formation, then it is sometimes recognized as the work o f Cobra Grande, a supernatural( cum cultural) force. When I talked to one affable farmer about the floodplain community where he had been born and where I furo that connected the floodplain lake to the Madeira River was created by Cobra Grande. I heard the story on other occasions from other interlocutors when discussing the origins context, was told me to me in numerous occasions. Many variations of the account exist, but in most versions it is told that during a large party, a reveler commits an of fensive act that provokes the ire of Cobra Grande. With a whip of its tail it beats the earth causing a major crack to form across the front half of the community. The crack rapidly widens

PAGE 172

172 until a massive slab of land breaks off into the Madeira River, h url ing unfortunate partygoers who were per are subsequently swallowed up by the muddy water of the river below. If our concern is to understand the Amazonian environment, and the relationships that exist within it (i reason to believe that mythologies that speak directly to the shifting nature of Amazonian landscapes are important, and must also be considered in our study of historical ecology. Cobra Gran de does not exist in the biological way that anacondas do, but it exists in the Amazonian imaginary, and the way that many Amazonian peoples understand their surroundings. Jimison, a history student from Borba provided me with this insight while sharing s tories during a long boat ride from Manaus. He pointed out that Cobra Grande is important for understanding the on going transformation of the topography and hydrology of the region as it lifts islands, pulls banks into the river, cuts furos and weaves n ew water courses for the river to travel. As a student of history, Jimison argued that oral history complemented science or natural history, and even islands or cuts in the river, he recognized that it remained a powerful symbol of the evolving Amazonian landscape. Just as oral histories, archaeological artifacts, and patterns of botanical species are important sources of information for dissecting the history of swath s of Amazonian forest, mythologies that explain the seasonal creeks that run through them, the rhizomatic waterways that intersect and overlap should also be considered pertinent details for the study of the Amazonian landscape. The story of Cobra Grande is a

PAGE 173

173 compelling narrative, much like Curupira and other mythical beings of Amazonian folklore. But the story of Cobra Grande should be recognized as more than just a quaint folktale; it is as an important Amazonian metaphor that embodies the idea that our surroundings are constantly changing and that the landscape is in constant flux. Although we can not always identify the forces responsible for those changes, an imaginary snake can be a useful reminder that they are always present. In other words, Cobra Grande acknowledges what Deleuze and Guattari also point out with their metaphor of the rhizome: things are in a constant ; they are emergent (p21). banal ly point out the endless complexity of the Amazonian environment and the relationships that exist within it over time, but rather to illustrate the importance of drawing on diverse lines of inquiry to build a br oader (although never complete) and more dyna mic map of the landscape and its ecology sensu lato Returning to the study of agrobiodiversity, it is clear that diversity seen at present in the Amazonian landscape reflects in some part the history (socio cultural, econ omic, ecological) of the locale; however we must remind ourselves that the landscape and the farms that make it up change over time along with the evolving surroundings carved by rivers, fish, people, and Cob ra Grande, the diversity of plants encountered in those landscapes also change in turn. Varieties of cacao are gained and lost with every major flood; some landraces of manioc circulate more widely across a region while others disappear altogether. Trend y agricultural exports like aai slowly begin to populate the floodplains in little

PAGE 174

174 green armies, while other palms like bacaba hide out in the back of the homegarden, waiting for their chance to make it big in overseas markets. Papayas crop up over field s of dark soils littered with artifacts while rubber trees, scarred from productive years past, wai t nearby in the shadows, As pointed out above, botanical d iversity encountered on Amazonian farms is rarely the sole produc t of management by its contemporary occupants Many individuals inherit land with an existing diversity of economic plants introduced as early as the Pre columbian era or the late colonial period. Others farmers clear lands near forests that have pre exis ting concentrations of economic species without known histories. By acknowledging the complexity of these relations, and utilizing diverse lines of inquiry to understa nd them, we can begin to build more complete map s of such landscapes. Concluding Remarks : The Anthropology of (Agro)biodiversity Studies by anthropologists and social scientists focused on biodiversity and agorbiodiversity management are important not only because they provide more nuanced information regarding the nature of biological divers ity and human management of such, but because biodiversity loss is an intensely human problem and one that has extreme consequences for our species, and potentially our own survival. Biodiversity loss is alarming not simply because of our appreciation for other biological organisms or our cultural values attached to other species. It is deeply concerning because the endangerment and extinction of other species reminds us of our own inherently precarious situation in the world. If, as the FAO claims, 75% o f global food consumption is attributed to 12 species of plants and 7 species of animals, then the study of agrobiodiversity is more than a simple academic exercise (FAO 1999). Rather, it is an attempt to understand the risk that humanity potentially face s in its increasing

PAGE 175

175 focus on a shrinking number of biological species. Studies such as the present one are critical for understanding the influences that contribute to the maintenance and/or loss of the many other species beyond those 19 upon which hum ani ty has become so reliant Research Questions Posed in Dissertation Through the examination of myriad socio cultural, demographic, environmental, economic, and historical data, this dissertation sought to identify factors that influence the diversity of us eful plants found on smallholder farms in the municipality of Borba, Amazonas, Brazil Three general questions were posed in the introduction of this dissertation : 1. What are the primary factors that influence the diversity of plants managed by smallholder farming households? 2. What is the role of social networks and gender roles in maintainance and patterning of such diversity? 3. How does an examination of local and regional history inform our understanding of the pla nt diversity seen at present? Primary Factors Influencing Diversity at Household Level The chapters of this dissertation aim toward a common denominator in understanding the management of agrobiodiversity in Amazonia, namely that it is heavily influenced by social phenomena. As was demonstrat ed in chapter 3, the age of the household head is one of the significant predictors of species richness in part interactions with other people, including friends, family, and a cquaintances. In other greater number of opportunities to acquire new plant species through social interaction.

PAGE 176

176 Archambault and Coomes 2009: 118) In addition to the age of the household head, area of land under cultivation is a second significant predictor of species richness at the household level. Simply put, households that manage a greater area of land typically manage a higher number of species. This can also be potentially related to maturity in the household lifecycle as households with older household heads typically have more available laborers in the household which can contribute to house hold agricultural production, managing a larger area and greater diversity of crops. Furthermore, older and more stable households have a tendency to invest in more perennial crops in diverse agroforestry systems, whereas younger households focus on annua l crops, which limits the diversity of species that can accumulate. Despite the importance of social and demographic factors, environmental influences should not be overlooked. Environmental habitat was the third primary factor identified in this research that had critical bearing on the diversity of species managed at the household level. Intense flooding in the Central Amazon region in 2009 had a major impact on much of the agriculture along the floodplains of the Amazon, the Madeira, and other major tr ibutaries. Because of the risky nature of agriculture on the floodplain, floodplain households showed significantly lower means of plants species under management than households in the more stable uplands. Upland households pointed out, however, that dr ought in the uplands is much more extreme than in the floodplain and can also lead to major crop losses. Risks are thus inherent to agriculture in both environments, but upland households nonetheless appear to benefit from

PAGE 177

177 greater stability, which is refl ected in the significantly higher mean for species richness. In years with less extreme flooding, this difference may not be as pronounced, however. A final factor that appeared to be influential in household agrobiodiversity surveys is the usage of magic or symbolic plants. Approximately 25% of the total species identified had uses related to the supernatural or symbolic. This demonstrates that a significant portion of plants managed on smallholder farms in Borba are valued not strictly because of their direct economic or material benefits, but because of their role in social and cultural practice. These plants are used symbolically in rural communities to cope with social preoccup ations including infidelity, theft health and personal finances. Based on ethnographic data and observation, it seems that households that maintain these practices are likely to manage a larger variety of species than those that do not. Social Networks Looking specifically at the diversity of manioc varieties in the municipa lity of Borba, social network analysis was useful for demonstrating the social distribution of such diversity. At the municipal scale geography and community residence have important influences on the patterning of manioc varietal distribution. However when looking at the individual cases of households at different positions in the network, it becomes apparent that general attributes of the household heads also have crucial influence on the types of varieties accessed, the ability to acquire new variet ies, and the preference of particular varieties cultivated. At the local scale, geography plays less of a role in varietal distribution as its patterning appears to be related more closely to key households that drive exchange and the sourcing of manioc varieties. Households that were seen as expert manioc farmers played central roles in the loaning of manioc source material, and had the

PAGE 178

178 largest impact on manioc varietal distribution. However, expert households did not manage the highest diversity of ma nioc varieties. Instead, households that were more peripheral in the exchange network, served important roles in managing and maintaining rare and poorly distributed varieties. Gender The diversity of plant species managed in the household is also very much shaped by gender roles and gendered divisions of labor. Women are often responsible for the maintenance of species diverse homegardens, in which they cultivate ornamentals, medicinals, magic plants, and spices. Nonetheless, many women also plant man ioc and other staple crops that can contribute to household subsistence and potentially provide important cash income. While in some households men seek credit as the women s hare roles in agricultural production, working together in a complementary manner. Recognizing this diversity in the gendered nature of agricultural practice is thus critical for understanding the distribution and patterning of agrobiodiversity. Local and Regional History As articulated in Chapter 2, contemporary Amazonian agrobiodiversity can not be history of indigenous food production and management of Amazonian forests as well as the arrival of the agro export market that commenced in the colonial period and continued on with the Rubber Boom, and later jute production. On a smaller scale, the micro histories of individuals and their families are also crucial to this und erstanding as was found in the cases of Nilo Jos Dona Clia and other individuals that appeared in ethnographic vignettes included in these chapters. The lands these individuals inherited

PAGE 179

179 and the knowledge that they accrued working their lands with th eir families are equally inheritance of agrobiodiversity and its relationship to reginal historical ecology. This dissertation d emonstrates how history social relationships and practice, and even cultural beliefs (in the case of magic plants) have important bearing on the management of biodiversity in rural Amazonian communites. In an atte mpt to steer away from potential environmental disaster at the global scale the significance of such findings must be creatively adapted in order to perpetuate socio cultural practices that are useful to agrobiodiversity maintenance and correct tho se that are largely detrimental. In some ways this is already happening as can be seen in the of ). What is being demanded by acti vists and writers, such as Pollan, the environment, and our management of such. This has opened up a discussion regarding the restructuring of food systems in the developed world, which are t he same ones being implemented in the developing world However, in order to restructure these systems and promote new models, the nature of the problems and threats that we face at present must also be confronted and carefully studied. With this in mind this dissertation represents a step toward addressing these problems by highlighting the primary influences that shape contemporary agrobiodiversity and its social nature in the context of rural Amazonia.

PAGE 180

180 Figure 7 1 The silhouette of a jackfruit tree with its toothed leaves (Vila Gomes, August 2010)

PAGE 181

181 Figure 7 2 Nilo standing in the plot inherited from his grandfather, Manoel ( Auar Grande, November 2009)

PAGE 182

182 Figure 7 3 Sapo holding the small anaconda that we had stepped on in a creek bed near Lago do Comprido (November 2009)

PAGE 183

183 APPENDIX A AGROBIODIVERSITY AND HOUSEHOULD SURVEY #1 DATA: No. de entrevista: Roteiro de entrevista 1. Nome : 2. Idade: 3. Sexo: 4. Origem (onde nasceu?) 5. Quantas pessoas moram na sua casa? 6. Quantos adultos? 7. Quantas pesso as trabalham na agricultura (na roca)? 8. Quantos anos voc mora na propriedade? 9. Qual a rea total do seu lote? a. rea de roa : b. rea de capoeira: c. rea de fruteiras e perenes: d. rea de pastagem: e. rea de mata : 10. Tem outra propriedades? No caso sim, quantas hectare s em total? Economia 11. Quais so as principais atividades de renda? 12. Qual a renda mensal da casa? 13. Vende produtos nas balsas? 14. Tem acesso a credito? No caso sim, participa em qual programa? 15. Agricultura 16. Voce usou fertil izante qumico no ultimo ano? 17. Adubo orgnico? 18. Herbicida? 19. Inseticida? 20. Tem terra preta na sua propiedade? O que voc produz nesses solos? 21. Voc participou de mutiro ou puxurum quantas vezes no ultimo ano? 22. Voce pagou alguem para trabalhar na sua roca no ulti mo ano? 23. Qual a rea de culturas produzidas para o mercado e a rea total sob cultivao? 24. 25. Como voc leva produtos para o mercado? Qual cidade ? (Borba, Manicore, etc.) Troca de Se mentes e Mate riais Vegetativas 26. Voc guarda sementes? 27. Voce faz suas proprias mudas? No caso sim, de quais plant as? 28. Voc compra sementes e mudas? 29. De quem voce comprou (ou pegou) sementes no ultimo ano? 30. Com quem na comunidade voc trocou ou pegou sementes ou materiais veg etativas nos ltimos 12 meses? 31. Com quem fora da comunidade voc trocou sementes ou materiais vegetativas nos ltimos 12 meses? Agrobiodiversidade 32. Quantas culturas diferentes voc tem na sua(s) roa(s)? Quintal? 33. Quantas cultivares (variedades) de mandioca/ macaxiera? 34. Tem plantas que voce tem para remedios ou associadas com certas cren as (ex. olho gordo)?

PAGE 184

184 LISTA DE CULTURAS (com nmero de indivduos de cada uma quando possvel) R= roa Q= quintal Nome Location (R,Q) Nmero indiv. Nome Location Nmero 1 2 6 2 27 3 28 4 29 5 30 6 31 7 32 8 33 9 34 10 35 11 36 12 37 13 38 14 39 15 40 16 41 17 42 18 43 19 44 20 45 21 46 22 47 23 48 24 49 25 5 0

PAGE 185

185 APPENDIX B AGROBIODIVERSITY SURVEY #2 MANIVA 1. Quantas qualidades de maniva (e macaxeira) tem plantado? Quais so? 2. Quantos feixos tem plantado de cada qualidade? 3. Qual rea total de maniva plantado? 4. Voc tem maniva plantado em que tipo de solo? 5. Porque voc plantou essas qualidades que tem? (e.g. sabor, cor da farinha, tamanho da batata, adaptao a vrzea ou terra firme, crece rpido) 6. Quais so as caractersticas mais importantes de uma qualidade de maniva para voc? a. Cor da farinha b. Sabor da fari nha c. Tamanho da batata d. Resiste pragas e doenas e. Atura muito tempo na terra f. Ligeiro (produz batata em poucos meses) g. Adaptao a vrzea, terra firme 7. Quando aparece filhos de maniva, voc tira ou mantem na roa para ver se presta? 8. Voc planta maniva da vrzea na terra firme? Ou da terra firme na vrzea? 9. Voc usa as palavras `fracae fortepara distinguir qualidades de maniva? 10. Voc j plantou maniva na terra preta? 11. Qual a diferena da produo da mandioca na terra preta comparado com barro vermelho? 12. Quem s o as pessoas na comunidade que tem mais conhecimento de maniva? 13. Voc comprou maniva esse ano? 14. Com quem voc trocou (pegou ou emprestou) maniva nos ltimos 12 meses? USE ROSTER!!! AGROBIODIVERSITY 1. Homegarden species abundance. Faz uma lista de especies e # de cada um: 2. # de especies na roa e rea de roca:

PAGE 186

186 MAGIC PLANTS 1. Tem plantas que usa para banhos? 2. Tem plantas que usa para remdio? 3. Tem plantas que usa para evitar mau olhado? 4. Onde (ou com quem) voc aprendeu usar essas plantas? 5. Onde voc conseguiu essas plantas (origem de cada um)?

PAGE 187

187 LIST OF REFERENCES Aguilar Stoen, Mariel, Stein R. Moe, Sara Luc a Camargo Ricalde 2009 Home Gardens Sustain Crop Diversity and Improve Farm Resilience in Candelaria, Loxicha, Oaxaca, Mexico. Economic Botany 37: 55 77. Al den, Dauril 1976 The Signif icance of Cacao Production in Amazon Region during the Late Colonial Period: An Essay in Comparative Economic History. Proceedings of the American Philosophical Society 120(2): 103 135. Altieri Miguel A. 1999 Th e Ecological Role of Biodiversity in A groecosystems. Agriculture, Ecosystems and Environment 74(1 3): 19 31. Amoroso, Marta A 1992 Corsrios no Caminho Fluvial: Os Mura do R io Madeira. In Historia dos ndios no Brasil. M.C. da Cunha, ed Pp. 297 310. S o Paulo: Companhia das Letras. Arroyo Kalin, Manuel 2008 Steps towards an Ecology of Landscape a Geoarchaeological Approach to Study of An thropogenic Dark Earths in the Central Amazon R egion, Brazil PhD Dissertation, Dept. of Archaeology, University of Cambridge. Atran, Scott, and Douglas L. Medin 2008 Native Mind and the Cultural Construction of Nature. Boston: MIT Press Auto s da D evassa contra os Indio Mura do R io Madei ra e Naes do R io Tocantins 1986 Fac similes e Transcries `P a leogrficas. Manau s: Universidade do Amazonas. Badstue, Lone R., Mauricio R. Bellon, Julien Berthaud, Xchtil Juarez, Irma M. Rosas, Ana Maria Solano, and Alejandro Ramirez 2006 Examining the Role of Collective Action in an Informal Seed System: A Case Study from the Cent ral Valleys of Oaxaca, Mexico. Human Ecology 24(2): 249 273. Bale, William L. 1989 The Culture of Amazonian Forest. In Resource Management in Amazonia: Indigenous and Folk Strategies. W. Bale and D. Posey., eds. Pp. 1 21. Bronx, NY: New York Botanical Garden. 1994 the Historical Ecology of Plant Utilization by an Amazonian People. New York: Columbia University Press. 1998 Historical Ecology: Premises and Postulates. In Advances in Historical Ecology, W illiam Bale, ed. New York: Columbia University Press.

PAGE 188

188 2006 The Research Program of Historical Ecology. Annual Review of Anthropology 35 : 75 98. B ale, William L., and Clark L. Erickson, eds. 2006 Time and Complexity in Historical Ecology: Studies in the Neotropical Lowlands. New York: Columbia University Press. Barbieri, Alisson F., Richard E. Bilsborough, and William K. Pan. 2005 Farm Household Lifecycles and Land Use in the Ecuardorian Amazon. Population and Environment 27(1): 1 27. Bastos, Aurel iano Candido Tavares 1866 O Valle do Amazona s. Rio de Janeiro: B.L. Garnier Livreiro Editor. Bengtsson, Janne, Per Angelstam, Thomas Elmqvist, Urban Emanuelsson, Carl Folke, Margareta Ihse, Fredrik Moberg and Magnus Nystrm 2003 Reserves, Resilience an d Dynamic L andscapes. Ambio 32 : 389 396. Bernard, H. Russell 2006 Research Methods in Anthropology: Qualitative and Quantitative Approaches, 4 th edition. Lanham, MD: Altamira Press. B etendorff, Joo Felipe 1910 Cronica dos Padres da Misso da Companhi a de Jesus n o Estado do Maranho Rio de Janeiro: Revista do IHGB. Borg atti, Stephen P., Martin G. Everett, and Lynton C. Freeman 2002 Uci net for Windows: Software for So cial Network Analysis. Analytic Technologies. Borgatti, S.P., Ajay Mehra Daniel J. Brass, a nd Giuseppe Labianca 2009 Network Analysis in the Social Sciences. Science 323(5916): 892 89 5 Boster, James S. 198 4 Classification, Cultivation, and Selection of Aguaruna Cultivars of Manihot Esculenta (Euphorbiaceae). Advances in Economic Bota ny I: 34 47. 1986 Exchange of Varieties and Information between Aguaruna Manioc Cultivators. American Anthropologist 88(2): 428 436. Brush, Stephen B. 1995 In Situ Conservation of Landraces in Centers of Crop D iversity. Crop Science 35: 346 354.

PAGE 189

189 1999 The Issues of In Situ Conservation of Crop Genetic R esources. In Genes in the Field: On Farm Conservation of Crop Diversity. S.B. Brush, ed. Pp 3 26. Boca Raton: Lewis Publishers. 2000 Genes in the Field: On Farm Conservation of Crop D iversity. Boca Raton: Lewis Publishers. 2005 Rethinking Crop Genetic Resource Conservation. Conservation Biology 3(1): 19 29. Burt, Ronald S. 2004 Structural Holes and Good I deas. American Journal of Sociology. 110:349 399. Campbell, David G., Anab el Ford, Karen S. Lowell, Jay Walker, Jeffrey K. Lake, Constanza Ocampo Raeder, Andrew Townesmith, and Michael Balick 2006 The Feral Forests of the Eastern Peten. In Time and Complexity in Historical Ecology: Studies in the Neotropical Lowlands. W. Bale and C. L. Erickson, eds. Pp. 21 56. New York: Columbia University Press Carneiro, Robert L. 19 83 The C ultivation of M anioc among the Kuikuru. In Adaptive Responses of Native Amazon ians. R.B. Hames and W.T. Vickers, eds. Pp. 65 111. New York: Academic Pr ess. Deleuze, Gilles and Felix Guattari 1987 A Thousand Plateaus: Capitalism and Schizophrenia. Minneapolis: University of Minnesota Press. Chayanov, Alexander V. 19 66 The Theory of Peasant Economy. Homewood, IL: American Economic Association. Cherne la, Janet 1986 Os Cultivares de Mandioca na rea do Uaups (Tukno). In SUMA: Etnolgica Brasileira 1 D. Ribeiro, ed. P p. 151 158. Petropolis: Etnobiologia. Finep. Chibnik, Michael 1994 Risky Rivers: The Economics and Poltics of Floodplain Farming in Amazonia. Tucson: University of Arizona Press. Clement, Charles R. 1999 1492 and the Loss of Amazonian Crop Diversity. I. The Relation between Domesticatio n and Human Population Decline. Economic Botany 53 : 12 25.

PAGE 190

190 Comisso de Estudos da Estrad a de Ferro do Madeira e Mamor 1885 Do Rio de Janeiro ao Amazonas e alto Madeira. Rio de Janeiro: Soares e Niemeyer Coomes Oliver T., and Natalie Ban 2004 Home G ardens in Amazonian Peru: Diversity and Exchange of Planting M aterial. Geographical Review 94(3):348 367. Cromwell, Elizabeth, and Saskia v an Oosterhout 1999 On Farm Conservation of Crop D iversity : Policy and Institutional Le ssons from Zimbabwe. In Genes in the Field: On Farm Con servation of Crop Diversity. S. Brush, ed. Pp. 217 238. Boca Raton: Lewis Publishers Crosby, Alfred W. 1986 Ecological Imperialism: The Biological Expansion of Europe, 900 1900. Cambridge: Cambridge University Press. Daniel, Joo 2004 T esouro Descoberto no Mximo Rio Amazonas, 1722 1776. Vol 2. Belm: Contraponto. Davidson David M. 1970 Rivers and Empire: The Madeira Route and the Incorporation of the Brazilian Far West, 1737 1808. U npublished Ph.D. dissertation, Dept. of History, Yale University. Dean, Warren 1984 Brazil and the Struggle for Rubber: A Study in Environm ental History. Cambridge: Cambridge University Press. Denevan, William M. 1984 Ecological Heterogeneity and Horizontal Zonation of Agriculture in the Amazon Floodplain. In: Frontier Expansion in Amazonia, M. Schmink and C.H. Wood, eds. Pp. 311 336. Gaine sville: University Press of Florida. 1992 a Stone Axes vs Metal Axes: The Ambiguity of Shifting Cultivation in Prehistoric Amazonia. Journal of the Steward Anthropological Research 20(1&2): 153 165. 1992 b The Pristine Myth: The Landscape of the Americas in 1492. Annals of the Association of American Geographers 82(3):369 385. 1996 The Bluff Model of Riverine Settlement in Prehistoric Amazonia. Annals of the Association of American Geographers 86(4):654 681.

PAGE 191

191 D escola, Philippe 1996 In the Society of N ature: A Native Ecology in Amazonia. Cambridge: Cambridge University Press. Di Falco, Salvatore, and Charles Perrings 2005 Crop Biodiversity, Risk Management and the Implications of Agricultural A ssistanc e Ecological Economics 55: 459 466. Duelli, Pete r, and M artin K. Obrist 2003 Regional Biodiversity in an Agricultural Landscape: The C ontribution of Seminatural Habitat Is lands. Basic and Applied Ecology 4: 129 138. Duke, James A. 2008 FL : CRC Press. E dit, Robert C. 1977 Detection and Examination of Anthrosols by Phosphate Analysis. Science 19 7( 4311): 1327 1333. Elias, Marianne, Laura Rival, and Doyle McKey. 2000 Perception and Management of Cassava (Manihot esculenta) D iversity amo ng Makushi Amerindians of Guyana (South America). Journal of Ethnobiology 20(2): 239 265. Emperaire, Laure and Nivaldo Peroni 2007 Traditional management of agrobiodiversity in Brazil: A case study of manioc. Human Ecology 35: 761 768. Erickson, Clark L 2005 The Domesticated Landscapes of the Bolivian Amazon. In Time and Complexity in Historical Ecology: Studies in the Neotropical Lowlands. W.L. Balee and C.L. Erickson, eds. Pp. 235 278. New York: Columbia University Press. 2008 Amazonia: The Histori cal E cology of a Domesticated Landscape. In T he Handbo ok of South American Archaeology. H. Silverman and W.H. Isbell. Pp.1 57 183. Berlin: Springer Verlag. Escobar, Arturo 2008 Territories of Difference: Place, Movements, Life, Redes Durham: Duke Univer sity Press. Estrada, Alejandro Rosamond Coates Estrada, and Dennis A. Merritt 1997 A nthr opogenic Landscape Changes and Avian D iversity at Los Tuxtlas, Mexico. Biodiversity and Conservation 6: 19 43.

PAGE 192

192 Evans Pritchard, E.E. 1958 [1937] Witchcraft, Oracle s and Magic among the Azande. London: Oxford University Press. Food and Agriculture Organization 1999 Women: Users, Preservers and Managers of Agrobiodiversity. m Accessed 17 April 2009 Fairhead, James and Melissa Leach 1996 Misreading the African Landscape: Society and Ecology in a Forest Savanna Mosaic. Cambridge: Cambridge University Press. Fearnside, Philip and Paulo M.L. Graa 2006 BR anaus Porto Velho Highway and the Potential Impact of Linking the Arc of Deforestation to Central Amazonia. Environmental Management. 38: 706 716. Foster, George M. 1972 The Anatomy of Envy: A Study in Symbolic Behavior. Current Anthropology. 13(2): 165 202. Franklin, Adrian 2008 A Choreography of Fire: Posthumanist Account of Australians and Eucalypts. In The Mangle in Practice: Science, Society, and Becoming. A. Pickering and K. Guzik, eds. Pp. 17 27. Durham: Duke University Press. Fraser, J ames A. 2 010 Caboclo Horticulture and Amazonian Dark Earths along the Middle Madei ra River, Brazil. Human Ecology 38(5): 651 662. Fra ser, James A. Thiago Cardoso, Andr B. Junqueira, Newton P.S. Falcao, and Charles R. Clement 2009 Historical Ecology and Dark Ea rths in Whitewater and Blackwater Landscapes: Comparing the Middle Mad eira and Lower Negro Rivers. In W.I. Woods, W.G. Teixeira, J. Lehmann, C. Steiner, A. WinklerPrins and L. Rebellato, eds. Pp. 229 261. Amst erdam: Springer Fraser, James A., Andr B. Junqueira, Nicholas C. Kawa, Claide P. Moraes, and Charles R. Clement 2011 Crop Diversity on Anthropogenic Dark Earths in Central Amazonia. Human Ecology Online first, DOI: 10.1007/s12231 011 9158 z

PAGE 193

193 Fraxe, Therezinha J.P. 2004 Cultura Cabocla Ribeirinha: Mitos, Lendas e Tr ansculturalidade. Sao Paulo: Annablume. Fraze r, James G. 2009 [1922] The Golden Bough: A Study of Magic and Religion. New York: Cosimo. Glaser, Bruno, Ludwig Haumeier, Georg Guggenberger and Wolfgang Zech 2001 Humid T ropics. Naturwissenschaften (2001) 88:37 41 Godoy, Ricardo Wilkie, D avid S., and Franks, Jeffrey 1997 The Effects of Markets on Neotropical Defores tation: A Comparative Study of F our Ameri ndian S ocieties. Current Anthropology. 38: 875 878. Gomez Pompa, Arturo Jos Salvador Flores, and Victoria Sosa 1987 made tropical forest of the Maya. Interciencia 12(1): 10 15. Guest, Greg 2000 Using Guttman Scaling to Rank Wealth: Integrating Quantitative and Qualitative Data Field Methods 12(4): 346 357. Harris, Mark 2000 Life on the Amazon: Anthropology of a Brazilian Peasant Village. London: British Academy 2010 Rebellion on the Ama zon: The Cabanagen, Race, and Popular Culture in the North of Brazil, 1798 1840. Cambridge: Cambridge University Press. Heckenberger, Michael J., J. Christian Russell, Joshua R. Toney, and Morgan J Schmidt 2007 T he Legacy of Cultural L andscapes in the B razilian Amazon. Philosophical Transactions of the Royal Society B. 362 (1478):197 208. Heckenberger, Michael J., J. Christian Russell, C arlos Fausto, Joshua R. Toney, Morgan J. Schmidt, Edithe Pe reira, Bruna Franchetto, and Afukaka Kuikuro 2008 Pre Colu mbian Urbanism, Anthropogenic Landscapes, and the Future of the Amazon. Science 321(5893): 1214 1217. Heck ler, Serena L. 2004a Tedium and Creativity: The Valorization of Manioc Cultivation and Piaroa Women. Journal of Royal Anthropological Institute. 10: 241 259

PAGE 194

194 2004b Cultivating Sociality: Aesthetic Factors in the Composition and Function of Piaroa Homegardens. Journal of the Ethnobiology 24(2): 203 232 Heckler, Serena L., and Stanford Zent 2008 Piaroa Manioc Varietals: Hyperdiversity or Social Curren cy? Human Ecology 36: 679 697. Herndon, William L. and Lardner Gibbon 1854 Exploration of the Valley of the Amazon. Washington: A.O.P Nicholson, Public Printer. Hilbert, Peter Paul 1968 Archaeologische Untersuchungen am Mittleren Amazonas: Beitrage zur Vorgeschichte des sudamerkanischen Tieflandes. Berlin: Dietrich Reimer Verlag Hiraoka, Mario 1992 Caboclo and Ribereno Resoruce Management in Amazoni a: A Review. In Conservation of Neotropical Forests: Workin g from Traditional Resource Use. K.H. Redford and C. Padoch, eds. Pp. 134 157. New York: Columbia University Press. Howard, Patricia L. 2003a International Institute for Development Gatekeeper Series 112: 1 24. 2003b Women a nd the Plant World: An Exploration. In Women and Plants: Gender Relations in Biodiversity Management and Conservation, P.L. Howard, ed. Pp. 1 48. London: Zed Books. I nstituto B rasileiro de G eografia e E statstica 2005 Brazilian Institute of Geography an d Statistics. Accessed 10 November 2007. Rio de Janeiro: Instituto Brasi leiro de Geografia e Estatstica Ioris, Edviges M. 2005 A Forest of Disputes: Struggles over Spaces, Resources, and Soci al Identites in Amazonia. Unpublished Ph.D. Dissertation, Dept. of Anthropology, University of Florida. Jackson, Louise E., Unai Pascual, and Toby Hodgkin. 2007 Utilizing and Conserving Agrobiodiversity in A gricult ural L andscapes. Agriculture, Ecosystems and Environment. 121(3): 196 210. Junqueira, Andr B., Glenn J. Shepherd, and Charles R. Clement. 2010 Secondary Forests on A nthropoge nic Soils in Brazilian Amazon Conserve A grobiodiversity. Biodiversity Conservation 19: 1933 1961.

PAGE 195

195 Kawa, Nicholas C. 20 08 Use and Management of Amazonian Dark Earth in Borba, Amazonas, Brazil Unpublished Master's thesis. Dept. of Anthropology, University of Florida. Kawa, Nicholas C., Daniel Rodrigues, and Charles R Clement 2011 Useful Species Richness, Proportion of Exotic Species, and Market Orientation on Amazonian Dark Earths and Oxisols. Economic Botany 65(2): 169 177. Kehlenbeck, Katja, and Brigitte L. Maass 2004 C rop diversity and classification of homegardens in Central Sulawesi, Indonesia. Agroforestry Syste ms 63: 53 62 Keller, Franz 1875 The Amazon and Madeira Rivers. London: Chapman and Hall. Leite, Serafim 1943 Histria da Companhia de Jesus no Brasil. Vol 3. Rio de Janeiro: Instituto Nacional do Livro. Levi Strauss, Claude 1966 The Savage Mind. Chic ago: University of Chicago Press. Lima, Deborah M. 2004 The Roca Legacy: Land Use and Kinship Dynamics in Nogueira, an Amazonian Community of the Middle Solimoes Region. In Some Other Amazonians: Perspectives on Modern Amazonia. S. Nugent and M. Harris, eds. Pp. 12 36. London: Institute for the Study of the Americas. L oreau, M., N. Mouquet and A. Gonzalez 2003 Biodiversity as spatial insura nce in heterogenous landscapes. Proceedings of the National Academy of Sciences 100 : 12765 12770. Loureiro, Antoni o 2007 O Amazonas na Epoca Imperial. 2nd Ed. Manaus: Editora Valer. Major, Julie Charles R. Clement, and A. DiTommaso 2005 Influence of Market Orientation on Food Plan t Diversity of Farms Located on Amazonian Dark Earth in the Region of Manaus, Amazon as, Brazil. Economic Botany 59(1):77 86. Malinowski, Bronislaw 1935 Coral Gardens and their Magic: A Study of the Methods of Tilling the Soil and of Agricultural Rites in the Trobriand Islands. Vol. 2. The Language of Magic and Gardening. London: George Allen and Unwin Ltd.

PAGE 196

196 Malin ows ki, Bronislaw, and Robert Redfield ed. 2004 Magic, Science and Religion and Other Essays 1948. Whitefish, Montana: Kessinger Publishing Marcoy, Paul 2001 Viagem pelo Rio Amazonas. (trans. A ntontio Porro.) Manaus: Editora da Universidade do Amazonas Marsden, Peter V. 1 990 Network Data and Measurement. Annual Review of Sociology 16: 435 463. McGuire, S hawn J. 2008 Securing Access to Seed: Social Relations and Sorghum Seed Exchange in Eastern Ethiopia. Human Ecology 36(2 ): 217 229. McKenna Dennis J ., L.E. Luna, and G.N. Towers 1995 Biodynamic Constituents in Ayahuasca Admixture Plants: An Uninve stigated Folk Pharmacopoeia. In Ethnob otany: Evoltion of a Discipline. R.E. Schultes and S. von Reis, eds. Pp. 349 360. Portla nd: Timber Press. Meggers, Betty J. 1996 [1971] Amazonia: Man and Culture in a Counterfeit Paradise, Revised Edition. Washington, D.C: Smithsonian Institution Press. Michon, Genevieve, and Fabienne Mary 1994 Conversion of traditional village gardens and new economic strategies of rural households in the area of Bogor, Indonesia. Agroforestry Systems 25: 31 58. Miller, Robert P., James W. Penn, and Johannes Van Leeuwen 2006 Amazon ian homegardens: thei r ethnohistory and potential contribution to agro forestry development. In Tropical Homegardens: A Time Tested Example of Sustainable Agroforestry B.M Kumar and P.K.R. Nair, eds. Pp. 43 60 Dordrecht: Springer Miller, Robert P. and P.K. Nair. 2006 Indigenous agroforestry systems in Amazonia: from pr ehistory to today. Agroforestry Systems 66: 151 164. Moraes, Claide P. 2010 Sedentarismo e Guerra no Baixo Madeira Pr Colonial. In: II Encontro Internacional de Arqueologia Amaznica (EIAA II), 2010, Manaus. Caderno de Resumos do EIAA II. Moran, Emili o F.

PAGE 197

197 1993 Through Amazonian eyes: The human ecology of Amazonian populations. Iowa City: University of Iowa Press. M orcote Rios, Gaspar, and Rodrigo Bernal 2001 Remains of palms (Palmae) at Archaeological Sites in the New World. The Botanical Review 67( 3): 309 350. Murrieta, Rui S.S. and Antoinette M.G. WinklerPrins 2003 Flowers of Water: Homegardens and Gender Roles in a Riverine Caboclo Community in the Lower Amazon, Brazil. Culture and Agriculture 25(1): 35 47. Neves, Eduardo G., James B. Petersen Robert N. Bartone, and Carlos A. Silva 2003 The Historical and Socio Cultural Origins of Amazonian Dark Earths. In Amazonian Dark Earths: Origin, Properties, Management. D.C.K. J. Lehmann, B. Glaser, and W. I. Woods, eds. Pp. 29 49. Dordrecht: Kluwer A cademic Publishers. Nimuendaju, Curt 1948 The Mura and the Piraha. In Handbook of South American Indians. J. Steward, ed. Pp. 255 269. Washington, D.C.: Bureau of American Ethnology 2004 Pursuit of a Past Amazon: Archaeological Researches in the Brazi lian Guyana and in the A mazon Region. Gteborg: Etnologiska Studier 45. Nugent, Stephen 1994 Amazonian Caboclo Society: An Essay on Invisibility and Peasant Economy. Oxford: Berg Publishers. Nye, Margaret M. 1991 The mis measure of manioc. Economic Bot any 45(1): 47 57. Ortencio, W. Bariani 1997 M edicina Popular do Centro Oeste, 2 nd ed. Brasilia: Thesaurus. Oyuela Caycedo, Augusto 2008 Late Pre Hispanic Chiefdoms of Northern Colombia and the Formation of Anthropogenic Landscapes. In Handboo k of Sout h American Archaeology. H. Silverman and W.H. Isbell Pp. 405 428. Berlin: Springer Verlag. Pace, Richard 1997 Brazilian Review. 34(2): 81 89. Padoch, Christine 1999 Varzea: Diversity, Development, and Conserv ation of Amazonia's W hi tewater F loodplains. New York: New York Botanical Garden Press.

PAGE 198

198 Padoch, Christine, and Wil de Jong 1991 The House G arden s of Santa Rosa: Diversity and Variability in an Amazonian Agricultural S ystem. Economic Botany 45(2): 166 175. 1992 Diversity, Variation, and Cha nge in Ribereno Agriculture. In Conservation of Neotropical Forests: Workin g from Traditional Resource Use. K.H. Redford and C. Padoch, eds. Pp. 158 174. New York: Columbia University Press. Parker, Eugene ed. 1985 A mazon Caboclo: Historical and Contemporary Perspectives. Studies in Third World Societies, No. 32. Williamsburg, VA: College of William and Mary. Perrault Archambault, M athilde, and Oliver T. Coomes 2008 Distribution of Agrobiodiversity in Home G ardens a long the Corrientes River, Peruvian Amazon. Economic Botany 62(2):109 126. Perreault, Thomas 2005 Why Chacras (Swidden gardens) Persist: Agrobiodiversity, Food Security, and Cultural Identity in the Ecuadorian A mazon. Human Organization 64(4): 327 339. Perz, Stephen G. 2001 Ho usehold Demographic Factors as Life C yc le Determinants of Land U se in the Amazon. Population Research and Policy Review. 20: 159 186. Pollan, Michael 2006 The Omnivore's Dilemma: A Natural History of Four Meals. New York: Pengui n Press. 2008 In the Defense of Food: An Eater's Manifesto. New York: Penguin Press. Posey, Darrell A. 1985 Indigenous Management of Tropical Forest Ecosystems: The C ase of the Kayapo Indians of the Brazilian Amazon. Agroforestry Systems 3(2): 139 158 1992 Interpreting the Reality of Indigenous Concepts. In Conservation of Neotropical Forests: Working from Traditional Resource Use. Kent H. Redford and Christine Padoch, eds. Pp. 21 34. New York: Columbia University Press. Raffles, Hugh 2002 In Amaz onia: A Natural H istory. Princeton: Princeton University Press. Rangel, Alberto 2000 [1934] A Bacia do Mar Doce. Coleo de Documentos da Amazn ia No. 11. Mana us: Governo do Estado do Amazonas Secretaria de Estado da Cultura, Turismo e Desporto.

PAGE 199

199 Rebhun, Linda Anne 1994 Swallowing Frogs: Anger and Illness in Northeast Brazil. Medical Anthropology Quarterly 8(4): 360 382. Reyes Garcia, Victoria Vicent Vadez, Neus Marti, Tomas Huanca, William R. Leonard, and Susan Tanner 2008 Ethnobotanical Knowledge an d Crop Diversity in Swidden Fields: A Study in a Native Amazonian Society. Human Ecology 36(4): 569 580. Rival, Laura M. 2002 T rekking through History: The Huaorani of Amazonian Ecuador. New York: Columbia University Press. Rocheleau, Dian n e and David Edmunds 1997 Women, Men and Trees: Gender, Power and Property in Forest and Agrarian Landscapes. World Development 25(8): 1351 1371. Roller, Heather F. 2010 Colonial Collecting Expeditions and the Pursuit of Opportunities in the Amazonian Serto, c. 17 50 1800. The Americas 66(4): 435 467. Roteiro da primeira viagem de vapor at a cidade da Barra 2000 [1 853] Coleo Documentos da Amazonia No. 14. Manaus: Governo do Estado do Amazonas Secretaria de Estado da Cultura, Turismo e Desporto. Salick, Jan Ni coletta Cellinese, and Sandra Knapp. 1997 Indigenous diversity of cassava: generation, maintenance, use and loss among the Amuesha, Peruvian Upper Amazon. Economic Botany 51(1): 6 19. Santos, Francisco Jorge dos 1999 Alm da conquista: Guerras e rebeli es indgenas na Amaznia Pombalina. Manaus: Editora da Universidade do Amazonas. Schmidt, Morgan J. 2010 Reconstructing Tropical Nature: Prehistoric and Modern Anthrosols (Terra Preta) in the Amazon Rainforest, Upper Xingu River, Brazil. Unpublished Ph .D. dissertation, Dept. of Geography, Uni versity of Florida. Schmink, Marianne, and C harles H. Wood 1992 Contested Frontiers in Amazonia. New York: Columbia University Press. Schoen, Daniel J. and Anthony H.D. Brown 2001 The conservation of wild plant species in seed banks. Bioscience (51)11: 960 966.

PAGE 200

200 Scott, John 2000 Social Network Analysis: A Handbook, 2 nd ed. London: Sage Publications. Shepard, Glenn, and Henri Ramirez 2011 Made in Brazil ispersal of the Brazil N ut ( Bertholletia excels a Lecythidaceae ) in Ancient Amazonia. Economic Botany 65(1): 44 65. Sillitoe, Paul 2003 The Gender of Crops in the Papua New Guinea Highlands. In Women and Plants: Gender relations in Biodiversity Management and C onservation P.L. Howard, ed. Pp. 165 180 London: Zed Books. Simes, Mario F., and Daniel F. Lopes. 1987 Pesquisas arqueolgicas no baixo/medio rio Madeira (Amazonas). Revista de Arqueologia. 4: 117 134. Siqueira, Andrea F. Moran 200 7 Embodied Decisions: Reversible and Irreversible Contraception Methods among Rural Women in the Brazilian Amazon. Human Organization 66(2): 185 195. Smith, Nigel J .H. 1996 The Enchanted Amazon Rain Fo rest : Stories from a Vanishing Wo rld. Gainesville, FL: University Press of Florida. Stanfield, Michael E. 1998 Red Rubber, Bleeding Trees: Violence, Slavery, and Empire in Northwest Amaznia, 1850 1933. Albuquerque: University of New Mexico Press. Sternberg, H ilgard O'Reilly 1975 The Amazon River of Brazil. New York: Springer Verlag. Subedi Anil, Pashupati Chaudhary, Bimal K. Baniya, Ram B. Rana, Radha Krishna Tiwari, Deepak K Ri jal, Bhuwon R. Sthapit and Devra I. Jarvis 2003 Who Maintains Crop Genetic Diversity and How? Implications for On Farm C onservation an d Participatory Plant B reeding. Culture and Agriculture 25(2): 24 26 Sweet, D avi d G. 1974 A Rich Realm of Nature Destroyed: The Middle Amazon Valley 1640 1750. Unpublished Ph.D. dissertation, Dept. of History, University of Wisconsin, M adison.

PAGE 201

201 Thomas, C.F. George, and E.J.P. Marshall 1999 Arthropod Abundance and Diversity in Differently Vegetated Margins of Arable F ields. Agriculture, Ecosystems & Environment 72 ( 2 ): 131 144 Thrupp, Lori Ann 2002 Linking Agricultural B iodiversity and Food Security: The Valuable Ro le of Agrobiodiversity for Sustainable A griculture International Affairs 76(2): 265 281. Trinh, Luu Ngoc Jessica W Watson, Nguyen N goc Hue, N guyen Ngoc De, Nguyen van Minh, P. Chu, Bhuwon R. Sthapit, and Pablo B. Eyzaguirre 2003 Agrobiodiveristy Conservation and Development in Vietnamese Home G ardens. Agriculture, Ecosystems & Environment 97(1 3): 317 344. Vadez, Vincent Victoria Reyes Gar cia, Ricardo A. Godoy, V. Lilian Apaza, Elizabeth Byron, Tomas Huanca, William R. Leonard, Eddy Perez and David Wilkie 2004 Does Integration to the Market Threat Agricultural Diversity? Human Ecology 32(5): 635 646. Voeks, Robert A. 19 97 Sacred L eaves of Candombl e: African Magic, Medicine and R eligion in Brazil. Austin: University of Texas Press Wagley, Charles 1953 Amazon Town : A Study of Man in the Tropics. New York: Oxford University Press. Walker, Timothy D. 2009 Establishing Cacao Plantation C ulture in the Atlantic World: Portuguese Cacao Cultivation in Brazili and West Africa, Circa 1580 1912. In Chocolate: History, Culture, and Hertiage. L.E Grivetti and H.Y. Shapiro, eds. Pp. 543 558. Hoboken, NJ: Wiley. Walker, Robert T., and Alfredo K.O. Homma 1996 Land Use and Land Cover Dy namics in the Brazilian Amazon: An O verview. Ecological Economics 18: 67 80. Weinstein, Barbara 1983 T he Amazon Rubber Boom: 1850 1920. Stanford: Stanford University Press. Wellman, Barry 1983 Network Analysis: Som e Basic Principles. Sociological Theory 1: 155 200. Wezel, Alexander, and Julia Ohl 2005 Does Remoteness from Urban Centres Influence Plant Diversity in Homegardens and Swidden Fields? A Case Study from the Matsiguenka in the Amazonian Rain Forests of Pe ru. Agroforestry Systems 65: 241 251.

PAGE 202

202 Wils on, Warren M. and Darna L. Dufour 2002 Why "Bitter" C assava? Produc assava in a Tukanoan Ind ian S ettlement in the Northwest Amazon. Economic Botany 56(1): 49 57. WinklerPrins, Anto inette M.G.A. 2002 Seasonal Floodplain Upland Migration along the Lower Amazon River. Geographical Review 92(3): 415 431. 2006 Jute C ultivation in the Lower Amazon 1940 1990: An Ethnographic A ccount from Santarm, Par, Brazil. Journal of Historical G eography 32: 818 838. WinklerPrins, Antoinette M.G.A. and Perpetuo S. de Souza 20 05 Surviving the City: Urban Homegardens and the Economy of Affection in the Brazilian Amazon Journal of Latin American Geography 4(1): 107 26. Yongneng, Fu, Guo Huijun, Chen Aiguo, and Cui Jinyun 2006 Household D ifferentiation and On Farm Conservation of Biodiversity by Indigenous Households in Xishuangbanna, China. Biodiversity and Conservation 15: 2687 2703. Zhu, Youyong., Hairu Chen, Jinghua Fan, Yunyue Wang, Yan Li Jianbing Chen, JinXiang Fan, Shisheng Yang, L ingping Hu, Hei Leung, Tom W. Mew, Paul S. Teng, Z onghua Wang, and C hristopher C. Mundt 2000 Genetic Diversity and D ise ase Control in R ice. Nature. 406: 718 722. Zimmerer, Karl S. 1996 Changing F ortunes : Bi odiversity and Peasant L ivelihood in the Peruvian Andes. Berkeley: University of California Press. 2006 Introduction: Geographical issues in the expansion of conservation and globalization. In Geographies of Conservation and Globalization: Expanding Dim ensions and Dilemmas K. S. Zimmerer, ed. Pp. 1 43. New York: Columbia University Press

PAGE 203

203 BIOGRAPHICAL SKETCH Nick Kawa was born and raised in northern Illinois. In 1999, he moved to Tucson, Arizona to begin his undergraduate studies at the University of A rizona. In 2003, after graduating from the University of Arizona with a Bachelor of Arts degree in a nthropology he moved to the city of Manaus in the Brazilian Amazon to teach English. While living in Manaus he interned at the National I nstitute of Ama zonian Research for 6 months where he participated in research on anthropogenic soils of the Amazon. In 2006 he began his graduate studies in the Department of A nthropolog y at the University of Florida and in 2008 he received his M s degree with hi Having completed his doctorate he continues his career conducting research in Amazonia and teaching anthropology.