<%BANNER%>

UFIR PETE



EVALUATING SOYBEAN FARMING PRACTICES IN MATO GROSSO, BRAZIL
CITATION PDF VIEWER
Full Citation
STANDARD VIEW MARC VIEW
Permanent Link: http://ufdc.ufl.edu/IR00001345/00001
 Material Information
Title: EVALUATING SOYBEAN FARMING PRACTICES IN MATO GROSSO, BRAZIL
Physical Description: MS Thesis
Creator: CAROLINA MAGGI RIBEIRO
 Notes
Abstract: This study provides an evaluation of the management practices adopted by soybean farmers in northern Mato Grosso, Brazil to promote sustainable agriculture. Since Amaggi Group, the leading soybean operator in Mato Grosso, requires social and environmental responsibility from its pre-financed producers, this study addresses the following research questions: Do management practices differ between pre-financed and not pre-financed farmers with regard to implementation of no-tillage systems, deforestation and pesticide use? Are there differences in yield between farmers pre-financed by the Amaggi Group and those who are not? Moreover, because farms are located in different biomes such as Cerrado and Amazon Forest, additional questions arise: Do management practices differ among the municipalities of Sinop, Sorriso and Tapurah, with regard to no-tillage systems, deforestation, forested area and pesticide use? Are there differences in soybean yield among the municipalities of Sinop, Sorriso and Tapurah? A case study of two soybean farms in the region of Sinop provides indications of the impacts of uncertainties in the exchange rate and the world soybean price on the farmers’ net revenue. A total of 40 soybean producers chosen randomly were interviewed, from which 20 farmers were pre-financed by Amaggi and 20 farmers were not pre-financed. With regard to their location, 10 farms were located in Sinop, 20 in Sorriso and 10 in Tapurah. Data collected with regard to land use, soybean farming practices, and soybean yields were statistically analyzed. Soybean pest and diseases, and pesticide use were also considered. For two farms located in Sinop, net revenue analyses and risk analyses were conducted. Data analyses revealed that the only difference between pre-financed and not pre-financed farms is related to the no-tillage system: 15% of the pre-financed soybean area did not have surface residues from cover crops, compared with 0.5% of not pre-financed farms. Comparisons among the farm locations revealed that an average of 27% of farm area in Sorriso was deforested while farms in Sinop were on average of 37% deforested. The use of desiccant prior to soybean harvest was an exception in pesticide patterns: farms located in Tapurah applied desiccant to 82% of their soybean area while farms located in Sorriso applied it to 50% of their area. Finally farms in Sinop had a higher soybean yield of fifty-two 60kg bags per hectare compared with 56 bags in Sorriso. The case study of two farms in Sinop revealed that the smaller farm is more susceptible to risks and uncertainties in the exchange rate and soybean price. An unfavorable exchange rate and relatively low soybean price increased the risk of the smaller farm losing money.
Acquisition: Collected for University of Florida's Institutional Repository by the UFIR Self-Submittal tool. Submitted by Peter E. Hildebrand.
Publication Status: Unpublished
 Record Information
Source Institution: University of Florida Institutional Repository
Holding Location: University of Florida
Rights Management:
This item is licensed with the Creative Commons Attribution Non-Commerical License. This license lets others remix, tweak, and build upon this work non-commercially, and although their new works must also acknowledge the author and be non-commercial, they don’t have to license their derivative works on the same terms.
System ID: IR00001345:00001

Downloads

This item is only available as the following downloads:

Carolina_thesis ( PDF )


Full Text

PAGE 1

1 EVALUATING SOYBEAN FARMING PRACTICES IN MATO GROSSO, BRAZIL: ECONOMIC AND ENVIRONMENTAL PERSPECTIVES By CAROLINA MAGGI RIBEIRO A THESIS PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLOR IDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF MASTER OF SCIENCE UNIVERSITY OF FLORIDA 2007

PAGE 2

2 2007 Carolina Maggi Ribeiro

PAGE 3

3 To my wonderful parents.

PAGE 4

4 ACKNOWLEDGMENTS I thank my supervisory committee chair, Dr Peter Hildebrand, for his mentoring and assistance, and for helping me to improve writte n skills. I also would like to thank the committee members Dr. Robert Buschbacher and Dr. Robert McSorley for their support and collaboration through the development of my thesis, and for Dr. Buschbachers assistance in Brazil while I was doing my field work. I thank my parents for their loving encourag ement and support. I would like to thank my family for enabling me to write this thesis in collaboration with Amaggi Group, and for their help in facilitating information. I especially thank Joo Shimada, Amaggi Groups Environmental Coordinator, for his assistance and support while I was in the field, for helping me with background information and sharing his knowledge of good farming practices and legal compliance in Mato Grosso. I also would like to thank the Amaggi Groups managers in the cities of Sinop, Sorriso and Tapurah, for finding the time to introduce me to the soybe an farmers in the regi on. I thank Amaggi Group and staff who collaborated with me and c ontributed to the success of my research. Thanks go to the soybean farmers in northern Ma to Grosso who participated in my survey, for their honest and open participation, and thanks to my friends in this region, who helped me to get in touch with local farmers. I also would lik e to thank Ocimar Villela who first encouraged and inspired me to become invol ved in natural resources manageme nt, leading me to University of Florida. Most importantly, I would like to thank Onil Banerjee for his love, encouragement, support, editing, and for being a lovely boyfriend.

PAGE 5

5 TABLE OF CONTENTS page ACKNOWLEDGMENTS...............................................................................................................4 LIST OF TABLES................................................................................................................. ..........8 LIST OF FIGURES................................................................................................................ .........9 ABSTRACT....................................................................................................................... ............12 CHAPTER 1 INTRODUCTION..................................................................................................................14 Soybean Production and Expansion in Mato Grosso, Brazil..................................................14 Sustainable Agriculture........................................................................................................ ..17 Legal Compliance............................................................................................................17 Permanent preserved areas (APP)............................................................................18 Legal reserve (RL)...................................................................................................18 Good Farming Practices..................................................................................................19 Crop rotation and cover crops..................................................................................19 Soil conservation and management..........................................................................20 Integrated production systems..................................................................................21 Integrated pest management (IPM)..........................................................................22 Pesticide use and technology....................................................................................23 Amaggi Group................................................................................................................... .....24 Study Objectives, Research Questions and Hypotheses.........................................................25 Methods........................................................................................................................ ..........27 Description of Study Sites...............................................................................................27 The BR-163 site.......................................................................................................28 Municipality of Sinop...............................................................................................28 Municipality of Sorriso............................................................................................29 Municipality of Tapurah...........................................................................................30 Field Methods..................................................................................................................31 Interviews..................................................................................................................... ...33 2 SOYBEAN MANAGEMENT PRACTICES.........................................................................40 Introduction................................................................................................................... ..........40 Methods........................................................................................................................ ..........41 Data Analyses..................................................................................................................41 Descriptive Analyses.......................................................................................................42 Statistical Analyses..........................................................................................................42 Regression Analysis........................................................................................................42 Weighted Means..............................................................................................................43 Results........................................................................................................................ .............43

PAGE 6

6 Land Use....................................................................................................................... ...43 Comparisons between financial groups....................................................................43 Comparisons among mu nicipal groups....................................................................44 Linear regression models.........................................................................................45 Soybean Planting Practices..............................................................................................47 Comparisons between financial groups....................................................................47 Comparison among municipal groups......................................................................48 Soybean Pests..................................................................................................................49 Comparisons between financial groups....................................................................49 Comparisons among mu nicipal groups....................................................................50 Discussion..................................................................................................................... ..........51 Land Use....................................................................................................................... ...51 Comparison between financial groups.....................................................................51 Comparisons among mu nicipal groups....................................................................52 Soybean Planting Practices..............................................................................................53 Comparison between financial groups.....................................................................53 Comparisons among mu nicipal groups....................................................................54 Soybean Pests..................................................................................................................55 Comparisons between financial groups....................................................................55 Comparison among municipal groups......................................................................57 Conclusion..................................................................................................................... ..58 3 PESTICIDES AND THE ENVIRONMENT.........................................................................71 Introduction................................................................................................................... ..........71 Pesticide Use.................................................................................................................. .........72 Seed Treatment................................................................................................................72 Herbicide Use..................................................................................................................73 Insecticide and Fungicide Use.........................................................................................73 Method......................................................................................................................... ...........74 Data Analyses..................................................................................................................74 Descriptive Analyses.......................................................................................................75 Statistical Analyses..........................................................................................................75 Results........................................................................................................................ .............76 Seed Treatment................................................................................................................76 Fungicide use for seed treatment..............................................................................77 Herbicide Use..................................................................................................................78 Desiccants before soybean planting.........................................................................78 Post-emergent herbicides.........................................................................................78 Desiccants prior to soybean harvest.........................................................................79 Insecticides and Biological Control.................................................................................80 Fungicide Use..................................................................................................................81 Discussion..................................................................................................................... ..........82 Seed Treatment................................................................................................................82 Herbicide Use..................................................................................................................83 Desiccants before soybean planting.........................................................................83 Post-emergent herbicides.........................................................................................84

PAGE 7

7 Desiccants prior to soybean harvest.........................................................................85 Insecticide Use................................................................................................................ .86 Fungicide Use..................................................................................................................87 Conclusion..................................................................................................................... ..89 4 SOYBEAN PROFITABILITY AND RISK MODELING.....................................................98 Introduction................................................................................................................... ..........98 Exchange Rate.................................................................................................................98 Soybean Price (US$).....................................................................................................100 Production Costs............................................................................................................101 Methods........................................................................................................................ ........103 Statistical Analyses........................................................................................................103 Net Revenue Analyses...................................................................................................103 Risk Analyses................................................................................................................104 Results........................................................................................................................ ...........106 Statistical Analyses........................................................................................................106 Net Revenue Analyses...................................................................................................106 Risk Analyses................................................................................................................107 Discussion..................................................................................................................... ........108 Statistical Analyses........................................................................................................108 Net Revenue Analyses...................................................................................................109 Risk Analyses................................................................................................................111 Conclusion.....................................................................................................................112 5 SUMMARY AND CONCLUSIONS...................................................................................118 APPENDIX SEMI-STRUCTURED INTERVIEW....................................................................124 LIST OF REFERENCES.............................................................................................................127 BIOGRAPHICAL SKETCH.......................................................................................................133

PAGE 8

8 LIST OF TABLES Table page 1-1 Required widths for Permanent Preservati on Areas (APP) according to width of the watercourse.................................................................................................................... ....35 1-2 Legal Reserve (RL) modifications since Brazils 1965 Forestry Code implementation................................................................................................................. .36 1-3 Soybean production, planted and harveste d area for Brazil, Mato Grosso, and the municipalities of Sinop, Sorriso a nd Tapurah for year 2004 and 2005.............................37 1-4 Selected farm size distribution accordi ng to INCRA in Sinop, Sorriso and Tapurah.......37 2-1 Descriptive statistics an d comparison of means of land use between financial groups.....60 2-2 Descriptive statistics and comparison of means of land use among municipal groups.....61 2-3 Multiple regression model for natu ral forest (% of farm area)..........................................62 2-4 Descriptive statistics and comparison of means for soybean planting practices between financial groups...................................................................................................62 2-5 Descriptive statistics and comparison of means of soybean planting practices among municipal groups............................................................................................................... .63 3-1 Pesticide used for seed treatment as reported by farmers in Sinop, Sorriso and Tapurah........................................................................................................................ ......90 3-2 Herbicide used in soybean plantation as reported by farmers in Sinop, Sorriso and Tapurah........................................................................................................................ ......90 3-3 Insecticide used in soybean plantation as reported by farmers in Sinop, Sorriso and Tapurah........................................................................................................................ ......91 3-4 Fungicide used in soybean plantation as reported by farmers in Sinop, Sorriso and Tapurah........................................................................................................................ ......91 3-5 Descriptive statistics and comp arison of means of pesticide use......................................92 4-1 Descriptive statistics and comparis on of means of soybean yield in 2005/2006 harvest........................................................................................................................ ......113 4-2 Summary statistics for risk analyses, s howing net soybean revenue per hectare for a larger farm (6150 ha) and a smaller farm (650 ha) in different scenarios.......................113

PAGE 9

9 LIST OF FIGURES Figure page 1-1 Some cover crop options for the Cotton Soybean rotation system.................................38 1-2 Map of the study region. Municipalities of Sorriso, Sinop and Tapurah located northern Mato Grosso........................................................................................................38 1-3 Highway BR-163 and its area of influence........................................................................39 2-1 Relationship between farm size and per centage of natural fo rest on all sites...................64 2-2 Relationship between farm size and percenta ge of natural forest on farms that were pre-financed by Amaggi Group and farms that were not..................................................64 2-3 Relationship between farm size and percen tage of natural forest in Sinop, Sorriso, and Tapurah.................................................................................................................... ...64 2-4 Relationship between year of land purchase and percentage of natural forest on all sites.......................................................................................................................... ..........65 2-5 Relationship between year of land purchase and percentage of natural forest on farms that were pre-financed by Amaggi Group and farms that were not...................................65 2-6 Relationship between year of land purcha se and percentage of natural forest in Sinop, Sorriso, and Tapurah...............................................................................................65 2-7 Relationship between percentage area of the most recent deforestation and last year of deforestation on all sites................................................................................................66 2-8 Relationship between percentage area of the most recent deforestation and last year of deforestation on farms that were prefinanced by Amaggi Group and farms that were not....................................................................................................................... .......66 2-9 Relationship between percentage area of the most recent deforestation and last year of deforestation in Sinop, Sorriso, and Tapurah................................................................66 2-10 Adoption of cover crops fo r financial groups in 2006.......................................................67 2-11 Weighted mean of cover crops for no-t illage system for financial groups in 2006...........67 2-12 Adoption of cover crops for municipal groups in 2006.....................................................67 2-13 Weighted mean of cover crops for no tillage system for municipal groups in 2006.........68 2-14 Farmers perspective about the worst soybean pest in 2006..............................................68 2-15 Farmers reporting pathogenic dis ease in their soybean crops in 2006..............................69

PAGE 10

10 2-16 Farmers reporting insect infestat ion in their soybean crop in 2006...................................69 2-17 Farmers reporting nematodes in their soybean crops in 2006...........................................70 3-1 Pesticide and inoculant us e for seed treatment for fi nancial groups and municipal groups in 2006................................................................................................................. ...93 3-2 Pesticide and inoculant use for seed treatment for forest -1 groups (F1) and forest-2 groups (F2) in 2006............................................................................................................93 3-3 Fungicide use for seed treatment for fi nancial groups and municipal groups in 2006......93 3-4 Fungicide use for seed treatment for forest-1 groups (F1) and forest-2 groups (F2) in 2006........................................................................................................................... .........94 3-5 Desiccant use before soybean planting fo r financial groups and municipal groups in 2006........................................................................................................................... .........94 3-6 Desiccant use before soybean planting for forest-1 groups (F1) and forest-2 groups (F2) in 2006................................................................................................................... .....94 3-7 Post-emergent herbicide use for financ ial groups and municipal groups in 2006.............95 3-8 Post-emergent herbicide use for forest-1 groups (F1) and forest-2 groups (F2) in 2006........................................................................................................................... .........95 3-9 Desiccant use prior to soybean harvest fo r financial groups and municipal groups in 2006........................................................................................................................... .........95 3-10 Desiccant use prior to soybean harvest for forest-1 groups (F1) and forest-2 groups (F2) in 2006................................................................................................................... .....96 3-11 Insecticide use for financial gr oups and municipal groups in 2006...................................96 3-12 Insecticide use for forest-1 groups (F1) and forest-2 groups (F2) in 2006........................96 3-13 Fungicide use for financial gr oups and municipal groups in 2006....................................97 3-14 Fungicide use for forest-1 groups (F 1) and forest-2 groups (F2) in 2006.........................97 4-1 Trend line for the real:US Dollar exch ange rate from April 2002 to April 2007............114 4-2 Biodiesel production with soyb eans in a farm in Tapurah..............................................114 4-3 Screen capture of the Microsoft Excel sheet showing the RiskTrigen formula for exchange rate and dependent variables in scenario 1......................................................115 4-4 Triangular distribution for net soybean reve nue in scenario 1: exchange rate varying 10%............................................................................................................................ ......116

PAGE 11

11 4-5 Triangular distribution for net soybean reve nue in scenario 2: exchange rate varying 10% and directly influencing pe sticide and fert ilizer costs.............................................116 4-6 Triangular distribution for net soybean re venue in scenario 3: soybean price (US$) varying 10%.................................................................................................................... .117

PAGE 12

12 Abstract of Thesis Presen ted to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Master of Science EVALUATING SOYBEAN FARMING PRACTICES IN MATO GROSSO, BRAZIL: ECONOMIC AND ENVIRONMENTAL PERSPECTIVES By Carolina Maggi Ribeiro August 2007 Chair: Peter Hildebrand Major: Interdisciplinary Ecology This study provides an evaluation of the ma nagement practices adopted by soybean farmers in northern Mato Grosso, Brazil to pr omote sustainable agriculture. Since Amaggi Group, the leading soybean operator in Mato Grosso, requires social and environmental responsibility from its pre-financed producers, this study addresses the following research questions: Do management practic es differ between pre-financed and not pre-financed farmers with regard to implementation of no-tillage syst ems, deforestation and pesticide use? Are there differences in yield between farmers pre-fina nced by the Amaggi Group and those who are not? Moreover, because farms are located in different biomes such as Cerrado and Amazon Forest, additional questions arise: Do manageme nt practices differ amon g the municipalities of Sinop, Sorriso and Tapurah, with regard to no-till age systems, deforestation, forested area and pesticide use? Are there differences in soybean yield among the municipalities of Sinop, Sorriso and Tapurah? A case study of two soybean farms in the region of Sinop pr ovides indications of the impacts of uncertainties in the exchange rate and the world soybean price on the farmers net revenue. A total of 40 soybean producers chosen randomly were interviewed, from which 20 farmers were pre-financed by Amaggi and 20 farmer s were not pre-financed. With regard to their

PAGE 13

13 location, 10 farms were located in Sinop, 20 in Sorri so and 10 in Tapurah. Data collected with regard to land use, soybean farming practices, and soybean yields were statistically analyzed. Soybean pest and diseases, and pesticide use we re also considered. For two farms located in Sinop, net revenue analyses and risk analyses were conducted. Data analyses revealed that the only differe nce between pre-financed and not pre-financed farms is related to the no-tillage system: 15% of the pre-financed soybean area did not have surface residues from cover crops compared with 0.5% of not pr e-financed farms. Comparisons among the farm locations revealed that an average of 27% of farm area in Sorriso was deforested while farms in Sinop were on average of 37% defo rested. The use of desi ccant prior to soybean harvest was an exception in pest icide patterns: farms located in Tapurah applied desiccant to 82% of their soybean area while farms located in Sorriso applied it to 50 % of their area. Finally farms in Sinop had a higher soybean yield of fi fty-two 60kg bags per hectare compared with 56 bags in Sorriso. The case study of two farms in Sinop revealed that the smaller farm is more susceptible to risks and uncertainties in the ex change rate and soybean price. An unfavorable exchange rate and relatively low soybean price increased the risk of the smaller farm losing money.

PAGE 14

14 CHAPTER 1 INTRODUCTION Soybean Production and Expansion in Mato Grosso, Brazil According to the United States Department of Agriculture (USDA, 2004), Brazil is the second largest soybean ( Glycine max ) producer in the world, accounting for 24% of world soybean production. Its 2004/2005 harvest produced 50.5 million tons, with an average yield of 2.34 tons/ha in a harvest area of 21.52 million hectar es. Although the United States is the leading producer, accounting for 40% of world soybean production, it is estimated that Brazil can surpass its production by the end of this decade. Brazilian agriculture has benefited from currenc y devaluations, low pr oduction costs, rapid technological advances, and domestic and fore ign investment to expand production capacity (USDA, 2006). Soybean production in Brazil is likely to increase as a result of the growing demand in the national and international market for oil producing grains such as soybean, which can be processed into meal for live stock rations into oil for domestic use, and for biodiesel production. In addition, Brazil has unparalleled arable land re serves, and the technology to efficiently employ them, particularly in the state of Mato Grosso (Agnol, 2006). The state of Mato Grosso accounts for appr oximately 29% of the soybeans produced in Brazil (USDA, 2007), and is the countrys largest soybean pr oducing state. In 2005, Mato Grosso produced 17.76 million tons, with an av erage yield of 2.90 tons/ha in an area of 6.1 million hectares (IBGE, 2006). Most of Mato Grosso s production is concentrated in the central and southern region, in the Cerrado ecosystem, a tropical savanna. The state of Mato Grosso has an area of 906,807 km2. Although the state is part of the Legal Amazon, an administrative area defined by economic purposes to promote the eco nomic development of the region, its natural vegetation is formed by three different ecosyst ems: Cerrado or savanna biome (38.29%), located

PAGE 15

15 primarily in southeastern and center Mato Grosso, tropical fore st biome (55%) in the north formed by the Amazon Rainforest and Semi-deciduous seasonal forests, and Pantanal (7.02%), a wetland located in the southw estern region (Schwenk, 2005). In the 1980s, the state of Mato Grosso e xperienced a rapid expansion of soybean production in the Cerrado region, due to the availa bility of abundant and affordable arable land, economies of scale compared to the southern region of Brazil, technol ogy, mechanization, and the lowest operating costs per hectare (Goldsm ith and Hirsch, 2006). Currently, 58% of Brazils total soybean production comes from the Cerrado (Mittermeier et al., 2004). Soybean expansion into the Amazon biome began in 1997 when new s oybean varieties were de veloped that tolerated the humid and hot Amazon climate. Soybean producti on in the Amazon has increased at a rate of 15% per year since 1999 (WHRC, 2007). Although the state of Mato Grosso has good potential for soybean production, there are some challenges too. The great distance from th e production region to port s in addition to the poor condition of the roadways leads to high tran sport costs. As a comparison, 74% of Brazilian soybeans still travel by road, 23% are transporte d by railways, and 3% by waterways; in the U.S., waterways carry 61%, and roadways tr ansport only 16% (Goldsmith and Hirsch, 2006). The soils characteristic of Mato Grosso are relatively poor. They are acid, poor in nutrients and have elevated levels of aluminum, which dema nd high amounts of fertilizers accounting for 30% or more of soybean production costs. Moreover, environmental concerns exist. The dramatic shifts in land use as native savannas, dryland fo rests, and even certain rain forest sub-regions become potential areas for soybean cultivation may accelerate land clearing (Goldsmith and Hirsch, 2006).

PAGE 16

16 According to Brazils National Institute fo r Space Research (INPE), 30.75% of Mato Grossos total area (approximately 27,860,110 hectar es) had been deforested as of 2003. Higher deforestation rates occurred in northern Mato Grosso in 2003, where the biome is Tropical forest; and there are recently co lonized areas such as the munici pality of Sinop with 71.05% of its area deforested since 1964 (M oreno, 2005). Deforestation has been reduced dramatically since 2003/2004, however, due to a number of economic and policy variables (Souza, 2006). According to Fearnside (2001), soybean is much more environmentally damaging than other crops because it requires massive transpor tation infrastructure. Also, biodiversity loss occurs when natural ecosystems are converted to industrial farm ing systems. Soybean expansion from the Cerrado toward the Amazon forest certainly contributes to land degradation, fragmentation and biodiversity loss es. However, this environmental trade-off is typical of most countries where industrial agriculture has occurr ed. Furthermore, the degree to which soybean expansion is directly responsible for this loss is uncertain since soybean tends to expand to previously converted land, such as pasture (Brown et al., 2005). In order to protect the great biodiversity of the Cerrado a nd Amazon forest and maintain landscape and environmental values, best manage ment practices need to be adopted for the production of any crop. Rather than bring new land under cultivati on through the deforestation of tropical forest and Cerrado, degraded soils and ecosystems must be restored and used more effectively. Achieving food security and impr oving environmental qual ity through sustainable management of soils is an important tool in avoiding additional defo restation and enhancing landscape values (Lal, 2000). Being aware of these challenges, most of th e industries today have changed from very intensive agriculture, harrowing a nd ploughing the soil like the site preparation used in the 1970s

PAGE 17

17 and 1980s, to minimal cultivation techniques to avoi d soil erosion and nutrient losses (Ondro et al., 1995). The soybean farmers in Mato Grosso state claim that th ey are adopting best management practices to respond to the growi ng demand for commodities, produced in a manner that minimizes environmental impacts, in other words, sustai nable agriculture. Sustainable Agriculture The greatest challenge to contemporary agricult ure is to realize the goals of sustainable agriculture in practice. Most definitions of sustainable agriculture include the following elements: economic viability, maintenance of an adequate food supply for all people, conservation of nutrients and resources, mi nimal impact on the environment and natural ecosystems, intergeneration or even indefinite stability and equity (Powers and McSorley, 2000). The search for a sustainable agricultural model implies the promotion of management practices that are environmentally appropriate, socially benefici al, and economically viable. In the context of soybean farming in the Cerrado region of Brazil, some of these management practices include (1) legal compliance with land protection regulati ons and (2) good farming practices such as crop rotation, notillage planting system, integrated systems, and integrated pest management (Shimada, 2006). They are essential to protect and preserve th e range of species and habitats in the countryside, as well as c onserve valuable soil and water resources. Legal Compliance To manage an agricultural business in Mato Gr osso one must take in to account legislation pertinent to the development of private rural pr operties in the Legal Amazon. Critical elements of this legislation originated in Brazils 1965 New Forestry Code (C digo Florestal, 4.771, 1965) and are related to Permanent Preservation Areas (reas de Preservao Permanente, APP) and the Legal Reserve Area (rea de Reserva Legal, RL). These will be discussed in turn.

PAGE 18

18 Permanent preserved areas (APP) APP is forest and other vegetated area that must be preserved, with the objective of protecting rivers, natural landscap e, biodiversity and soil. Accord ing to Article 2 of Brazils 1965 Forestry Code (4.771) and considering the geogra phic characteristics of Mato Grosso, the most common APPs which must be protec ted or reforested are situated: Along rivers or water courses fr om their highest level along ri parian zones, whose minimal width is shown in Table 1-1; Around lakes, ponds or water ta nks, natural or artificial; In springs, either perennial or ephemera l, with a minimum width of 50 meters; and On the edge of steep slopes with a minimum width of 100 meters. Legal reserve (RL) According to the 2000 provisionary measure n 1956-50/00, the legal rese rve is defined as a forested area located inside a rural property, excludi ng the APP, in which biodiversity, flora and fauna are conserved, and ecological proce sses are rehabilitated. A lthough land clearing is illegal in the RL, sustainable forest management for multiple goods and services is permitted. In Mato Grosso as in any other state that form s the Legal Amazon, the percentage of the rural property to be maintained as an RL is: 80% if located in a tropical forest area; 35% if located in the Cerrado; and 20% if located in all other regions. Among the items discussed in provisional measure n 1956-50/00, is the compensation mechanism for legal reserve. This mechanism offers the rural producers who do not have the minimum legal reserve required in their proper ty, the alternative of compensating the area lacking in another property in the sa me micro-hydrological basin and ecosystem.

PAGE 19

19 Since Brazils 1965 New Forestry Code was enacted, it has been modified many times by law and provisionary measures (T able 1-2), serving to demonstrat e the legislators difficulty in reconciling the social and economic interests of the different st akeholders involved (Joels, 2002). The requirement of registering th e legal reserve with the landed property registration, forbidding changing its destination and de segregation was introduced by law 7.803, of July 18, 1989 (Art. 16 2). Good Farming Practices Crop rotation and cover crops Crop rotations or continuous cr op sequences involve replaci ng monoculture cropping with other crops over time. The rotation can be with cash crops where markets often determines the crop sequence and more than one cash crop can be grown per year; and cover crops which are not usually marketable, but are typically used to protect the soil from erosion, improve soil structure, enhance soil fertil ity, and suppress pest and pathoge ns. Cover crops are lower-value crops usually grown in the dry season which is less favorable for cash crop production. If legumes are used, they are planted in the rainy se ason, but are left in the field throughout the dry season. Another option is to allow an area to fallo w (unintentional rotation), because of erosion or when weeds take over an ar ea (Powers and McSorley, 2000). In the Brazilian Cerrado regions, some of the most prominent production systems are the continuous cash crop sequences of cotton ( Gossypium sp) soybean ( Glycine max ), and cotton ( Gossypium sp) soybean ( Glycine max ) corn ( Zea mays ). The most common cover crops are millet ( Pennisetum glaucum L. R. Br.), finger millet ( Eleusine indica ) and Brachiaria ruziziensis (Figure 1-1) In some regions of Mato Grosso, it is common to have cash crops such as corn, sorghum ( Sorghum bicolor (L.) Moech.), castor bean ( Ricinus communis L.), or sunflower

PAGE 20

20 ( Helianthus annuus ) planted after early soybean as a s econd crop. They serve the same purpose of cover crops and can often be sold (Altmann, 2006). Soil conservation and management According to the Food and Agriculture Or ganization (FAO) and the United Nations (2001), soil erosion, accelerate d by wind and water, is resp onsible for around 40% of land degradation worldwide. Soil cons ervation practices that provide some cover to the soil surface are the most efficient against soil erosion. The most frequently used soil conservation methods by soybean producers in Mato Grosso are dire ct seeding with terraces (Shimada, 2006). In direct seeding (also known as direct drilling, no tillage prac tice, no till, and zero tillage), the surface residues from cover crops are left undisturbed. These residues maintain the soil covered and thus help to control erosion and c onserve moisture. Maintenance of surface residue can reduce soil erosion up to 90% or more (P owers and McSorley, 2000). The seeds are placed in the soil without tilling. Additi onal advantages of this system are that it reduces the need for tractor power and tillage equipment, it demands less fuel, and increases soil organic matter (Buchholz et al., 1993). The main disadvantages of dir ect seeding are that this system relies on herbicide use for weed control and it may delay planting because of greater moisture under heavy residue. Direct seeding involves important trad e-offs between the soil conserva tion benefits and the intensive application of herbicide. Tradit ional tillage systems involve the mechanical and biological breakdown of weed species which, given Mato Grossos climate, result in significant increases in erosion and consequent nutrient loss. Thus, herbic ide use in this system substitutes the need for tilling; it improves crop yield by reducing weed density and competition while direct seeding minimizes soil erosion and nutrient loss (Powers and McSorley, 2000).

PAGE 21

21 Terrace, a raised bank of earth having verti cal or sloping sides and a flat top, slow or prevent the rapid surface runoff of irrigation wa ter or rainfall. Being implemented with no tillage, these sustainable agri culture practices provide soci al, economic and environmental benefits. These practices are recognized by national institutions such as the Brazilian Agricultural Research Corporation (Empresa Br asileira de Pesquisa Agropecuria, EMBRAPA), the Ministry of the Environment (Ministrio do Meio Ambiente, MMA) and the National Water Agency (Agencia Nacional de gua, ANA), and by international in stitutions such as the World Bank and the Food and Agriculture Or ganization (FAO) (Shimada, 2006). Integrated production systems Integrated production systems are farm syst ems that combine the production of various goods and services. One common variant of this syst em in Mato Grosso is the agrosilvipastoral system. An agrosilvipastoral system is a type of agroforestry system th at includes the production of trees or shrubs, crops, pasture, and animal s (World Agroforestry Centre, 2006). In tropical environments, both grazing systems of pasture under commercial tree stands, and growing and managing fodder-producing trees on fa rmlands are used (Nair, 1990). The benefits of the agro silvipastoral system repor ted by Russo (1996) include: diversification that stabilizes the agricultural sy stem; economic benefits ob tained from fuelwood, timber, and posts; soil improvement from legumi nous trees and deep nut rient uptake by trees; production of biomass which may be used as fora ge and/or organic matter; livestock production that provides meat, milk, and nutrient cycling th rough manure; and added nut ritional value to the rural familys diet. One of the few disadvantages is the soil compacting and trampling of crops by animals; however, this can be minimized through careful planning. In integrated production systems combining cr op and livestock production, the pasture is introduced following the soybean harvest. This can reduce pest and disease infestation by

PAGE 22

22 breaking the life cycle of specific pest and s oybean diseases. The pasture also provides a significant amount of organic residues which en ables subsequent direct seeding. The soybean that is planted over the existi ng pasture typically has a higher productivity and tends to reduce production costs due to no tillage. Additional income is also generated in this system with the addition of cattle graz ing (Bortolini, 2006). With the implementation of crop-livestock syst ems in regions that produce corn as a cover crop, three harvests in one year can be obtaine d. In Lucas do Rio Verde, in the north central Mato Grosso, soybean is planted from October to February; corn, a small amount of sorghum, and sunflowers are planted from February to June; and from June to September cattle are allowed to graze on the cover crops where th e pasture was planted (Bortolini, 2006). Integrated pest management (IPM) According to Powers and McSorley (2000), in tegrated pest management programs have the objective of reducing pesticide use and enviro nmental impact. It involves integration of multiple tactics for managing a single pest, and in tegration of the management of multiple kinds of pests. Many techniques such as resistant cu ltivars, soil amendment, crop rotation, sanitation, biological control and chemical control can be used in IPM (Shimada, 2006). Biological control is the mana gement of a pest by another living organism. The strategy involves manipulating predators, parasites, and pathogen presence to maintain pest populations in a field at levels below which they may cause economic injury to the crop (Powers and McSorley, 2000). Application of Baculovirus Anticarsia Gemmatalis to control velvetbean caterpillar began in Brazil in the beginning of the 1980s and grew rapidly from 2,000 hectares in 1982/1983 to more than 500,000 hectares in the 1987/1988 harvest period. The use of Baculovirus reduces the use of ch emical insecticide and has envi ronmental benefits. In addition, it does not interfere with other organisms that may help control other pests. For better results in

PAGE 23

23 controlling insect pests, manage ment decisions should be supported by regular monitoring of the crop and life stage, damage levels, insect lifecy cle, and the size of the outbreak (Shimada, 2006). According to EMBRAPA (2004), the foliar blight disease ( Rhizoctonia solani ) is efficiently controlled when IPM methods are ad opted. Since there are no cultivars resistant to this disease, control methods include direct seeding, crop ro tations with non-host plants, reduction of soybean plant density, balanced plant nutrition especially with potassium (K), sulfur (S), zinc (Zn), copper (Cu) and Manganese (Mn), weed control and chemical control. Pesticide use and technology Pesticide purchase and applic ation represent one of the greatest costs in agribusiness. Pesticides are also responsible for major negative environmental impacts and work-related accidents. To minimize environmental impacts and protect workers, a variety of variables need to be considered before applic ation such as weather conditions (moisture, temperature, wind and rain), the product features, and th e target pest (Shimada, 2006). There are a number of stages in the production process in which farmers in Mato Grosso commonly apply pesticides. First, herbicides are applied to the field before soybean planting to eliminate weeds. Second, agrochemicals are used in treating seeds to kill fungal diseases and improve seed germination. Then, during the growth cycle, post-emergent herbicides, insecticides and fungicides are applied to the crop to reduc e potential damage from weeds, insects and fungus. Finally, non-selective herbicides (also de siccants) are applied to facilitate soybean harvest. To avoid pesticide residues in the harv ested soybean grain, desicc ants should be applied seven days before the soybean harvest (EMBRAPA, 2004). According to Bickel and Dros (2003), five to te n liters of pesticide, depending on the level of technology used, are applied pe r hectare of soybean in Mato Grosso. This translates into 4.3 million kilograms of empty pesticide packages collected each year, rendering Mato Grosso the

PAGE 24

24 third largest producer of this type of waste. Therefore, compulso ry collecting and triple washing of empty pesticide packages are very important components of pest management, reducing the risk of groundwater contamination. Amaggi Group Amaggi Group, a privately held company, is th e leading soybean operator in Mato Grosso. In Amaggi Groups farms, there is a total 122 thousand hectares of soybean, 23 thousand hectares of corn (secondary crop), and 16 thousand hectares of cotton planted. In the municipality of Sapezal, there ar e two farms: Tucunar with an area of 57,833 hectares and the Agro-Sam farm with an area of 20,371 hectares. The 47,212 hectare It amarati farm is located in Campo Novo dos Parecis. The first farms acquired by the Group are in southern Mato Grosso and have a total area of 16,989 h ectares. Tanguro farm, 72,600 hectar es in size, is located in northern Mato Grosso (Grupo Amaggi, 2007). Amaggi Group has been solidifying its positi on in the agribusiness, through vertical integration in the production, processing and export ation of soybean and s ub-products such as oil and meal. One of its divisions, Amaggi Export and Import, operates in the states of Mato Grosso, Rondnia, and Amazonas. Its business is to comm ercialize, store, process, transport, and promote soybean production in Mato Grosso by pre-financing farmers. Amaggi Export and Import has silos in the region of the BR-163 such as in Sorriso and Sinop with a capacity of 60,000 tons each, and in Tapurah with a capacity of 18,000 tons. Amaggi Group silos receive both genetically modified soybeans (GM) and non-genetically modified soybeans (non-GM). The GM soybeans are segregated from the non-GM soybeans. The producers, besides paying royalties to the co mpany which sells the patented seed, receive different prices for GM soybeans and non-GM soybeans due to export logistics. The non-GM soybeans are exported through ports in the town of Itacoatiara (Amazonas state), where the

PAGE 25

25 Amaggi Group has a private port, and in the city of Santos (So Paulo state). The GM soybeans are exported through the port in the city of Parana gu (Paran state). Countri es that are willing to pay more for non-GM soybeans pay a premium for this product, compared with GM soybean. This premium is not transfe rred to the producers, however, since it covers the soybean segregation costs. The countries that usually pay premiums to Amaggi are Norway, Ireland, Denmark and Japan (L. M. Ribeiro, personal communication, June 6, 2007). One of Amaggi Groups objectives is to combine the preserva tion of the environment with excellent results in te rms of production and profitability. In working towards this objective, Amaggi Group has developed and disseminates a set of good farming practices for soybean producers in Mato Grosso, to i nduce a gradual improvement in th e levels of legal compliance and the standards of environmental performan ce. In addition, Amaggi Group requires socialenvironmental responsibility from its pr e-financed producers (Grupo Amaggi, 2007). Since 2004, Amaggi Groups credit policy has ha d the objective of promoting an ongoing improvement of the environmental indicators of its pre-financed producers. Based on data collected on the propertie s of pre-financed farmers, reco mmendations for improvement are developed. With regard to legal compliance, Amaggi Group requires legalization of legal reserves, recuperation of riparian areas and no illegal deforestation fo r the duration of the contract. With regard to good farming practices adoption or increase in the area where the notillage system is applied, and implementation of integrated pest management are recommended (Grupo Amaggi, 2006). Study Objectives, Research Questions and Hypotheses This project provides a compre hensive evaluation of the mana gement practices adopted by soybean farmers in northern Mato Grosso, Brazil where the agricultural frontier is pushing into the Amazon. The first objective is to evaluate differences in farmi ng practices adopted by

PAGE 26

26 soybean farmers in the region of BR-163. In order to accomplish this tas k, farmers that are prefinanced by Amaggi Group which are required to demonstrate environmental responsibility, were randomly selected to be compared with farmers who are not pre-financed by Amaggi Group. The second objective is to evaluate differences in farm ing practices between regions, namely the municipalities of Sinop, Sorriso, and Tapur ah due to the fact that they are located in different biomes. Semi-structured interviews w ith soybean farmers pre-financed by Amaggi Group and those not pre-financed by Amaggi Group in the muni cipalities of Sinop, Sorriso, and Tapurah, where Amaggi Group operates, were carried out in J une and July 2006 to address the following research questions: Do management practices differ between pre-fi nanced and not pre-financed farmers with regard to implementation of no-tillage systems, deforestation and pesticide use? Are there differences in yield between farm ers pre-financed by the Amaggi Group and those who are not? Are there differences in management practi ces among the municipalities of Sinop, Sorriso and Tapurah, with regard to no-tillage system deforestation, forested area and pesticide use? Are there differences in soybean yield am ong the municipalities of Sinop, Sorriso and Tapurah? What are the impacts of uncertainties in exch ange rate and world s oybean prices on the net revenue of soybean farmers in the region of Sinop? Given Amaggi Groups credit policy with their pre-financed producers, and the environmental concerns in northern Mato Gross o, the following hypotheses are tested in Chapter 2 and 3: The pre-financed farmers are more li kely to preserve forested area; The pre-financed farmers have a greater percen tage of soybean area in a no tillage system; The pre-financed farmers have greater cover crop diversity;

PAGE 27

27 The pre-financed farmers use fewer type s of pesticide than those who are not; The pre-financed farmers have grea ter yields than those who are not. Moreover, based on the assumption that farms th at are located closer to the Amazon forest would have more forested area on farm and, theref ore, would need less pesticide use, and that soybean productivity in areas that are newly deforested is lower than in areas that have been cultivated for longer periods of time, additiona l hypotheses tested in Chapter 2 and 3 are: The producers located in Sorriso are less likely to preserve forested areas than those located in Sinop and Tapurah; The producers located in Sorriso are more likel y to plant corn as a second crop than those located in Sinop and Tapurah; The producers located in Sinop use less fungi cide and apply it fewer times than the producers located in Sorriso and Tapurah; The producers located in Sinop have lower yi elds than those located in Sorriso and Tapurah. This study is divided into f our parts. Chapter 2 compares land use and soybean farming practices adopted by farmers in the study regi on. Chapter 3 complements Chapter 2, examining the pesticide use strategies implemented by farm ers. Chapter 4 compares yields between groups, and provides a case study of two farms in Sinop. Ri sk analyses are conducted to evaluate farmer susceptibility to uncertainties in the soybean pr ice and the exchange rate. Chapter 5 summarizes the findings and provides conc lusions and recommendations. Methods Description of Study Sites The study was undertaken in the municipalities of Sinop, Sorriso, and Tapurah situated in northern Mato Grosso (Figure 1-2). Benefits su ch as accessible lands offered by the federal and regional governments as part of the regiona l development programs, and infrastructure development enabled large areas to be purchased by the private sector and colonized. From the

PAGE 28

28 middle of the 1970s until the end of the 1980s, entr epreneurs from the southern and southeastern regions of Brazil attained vast extensions of public or private lands to invest in colonization programs, agriculture and cat tle ranching (Moreno, 2005). Characteristic of this frontier region are large-scale cattle ranching operations and mechanized monoculture farming due to th e arable lands and flat topography. Soybean production in 2004 and 2005 for Sorriso, Sinop, and Tapurah are displayed in Table 1-3. The municipalities are located in the region of highway BR; however, only Sorriso and Sinop are actually on the highway. The BR-163 site The BR-163 is one of the main federal highw ays. It was opened during the 1970s through the National Integration Program (PIN) with the objective of integrating the Amazon region with the national economy. The highway is 1,780 kilometers long and extends from Cuiab, capital of Mato Grosso, to Santarm on the Amazon rive r in Par state. Paving BR-163 is not yet completed, with 953 kilometers remaining betw een Matup (MT) to Santarm (ISA, 2005). Due to the fact that the BR163 passes through remote regions of the country that are of both environmental and cultural interest, the pave ment of the BR-163 has been debated since the 1990s. However, it was not until 2003 that it was decided that paving the BR-163 would go ahead as a component of Mato Grosso and Par s sustainable development programs. In 2004 the federal government created the BR-163 Sustai nable Regional Development Plan for the highways areas of influence (Figure 1-3), whic h seeks to resolve stak eholders demands in a participatory manner (ISA, 2005). Municipality of Sinop In 1972, the Real Estate Society of North Para n (SINOP), a coloni zing enterprise, bought an area of approximately 200 thousand hectares in the municipality of Chapada dos Guimares;

PAGE 29

29 successive acquisitions resulted in an area of more than 600 thousand hectares. In the BR-163s area of influence, the projects of Vera, Sinop, Santa Carmem and Cludia were implemented (Moreno, 2005). The city of Sinop was official ly founded September 14, 1974, and after 5 years the municipality of Sinop was created with an area of 3,207 Km2. The city of Sinop is located 500 kilometers north of Cuiab and 80 kilometers south of the city of Sorriso (Assessoria de Comunicao da Prefeitura de Sinop [ASSECOM], 2006). One of the main economic activities in Sinop is timber production. Sinop began to diversify its economy after 1995 by implementing sustainable forest management, conducting research in reforestation, cattl e ranching and agriculture (Pichini n, no date). The settlers of Sinop came from the south of Brazil. Today, migrant people are coming from other regions of Mato Grosso. According to IBGE, there were 74,831 inha bitants in the year 2000. In 2004, there were an estimated 94,724 inhabitants, an increase of 26.58% (ASSECOM, 2006). The topography in Sinop is generally flat with some slightly undulating areas, which is favorable for agriculture and cattle ranching. Most of its soil is clay with some sandy soils and there are some areas that are su sceptible to erosion. Before the occupation and deforestation of its natural vegetation, Sinop was covered by Rainfo rest. The typical climate is hot and humid with an average annual temperature of 28C. The e quatorial rain pattern is characterized by a dry season from June to August, and a rainy season w ith the heaviest rains from January to March (ASSECOM, 2006). Municipality of Sorriso In 1977, a private colonization project was im plemented along the BR-163 and resulted in the city Sorriso (Moreno, 2005). The majority of the migrant people came from southern regions of Brazil, especially from the st ates of Rio Grande do Sul, Para n and Santa Catarina. In May 13, 1986, the district of Sorriso was desegregated fr om the municipalities of Nobres, Sinop, and

PAGE 30

30 Diamantino and it was denominated the municipality of Sorriso with an area of 10,480 Km2. The municipality of Sorriso is loca ted 412 kilometers north of Cuia b. According to IBGE, Sorriso has 48,325 inhabitants, however a more recent estim ate is 65 thousand people (Sorriso City Hall, 2005). Sorrisos soil has excellent water infiltration capacity and medium susceptibility to erosion. In inadequate use conditions or under hea vy precipitation, irrevers ible soil degradation can occur. The climate is tropi cal humid with a defined dry s eason. The difference in average temperature is 15C from the hottest to the co ldest month. The average annual temperature is 30C. The average annual precipitation is around 2,000mm and it is concen trated in the months between October and March. Relative humidity is on average 80%, but it is 22% from June to the end of August (Sor riso City Hall, 2005). The main economic activities are mechanized agriculture, producing cash crops such as rice, soybean, corn, and cotton. L ogging and wood processing is also an important activity. Sorriso is considered the second largest grai n producer in Brazil. In the 2004/2005 harvest period, the planted area was 613,957 hectares wi th approximately 2,485,000 tons of grain harvested. According to research by IBGE (2005), the municipality of Sorriso is the fourth largest corn producer, and the largest soybean pro ducer in Brazil. Cattle ranching is increasing annually and currently there are 40,000 head of cattle on 30,000 hectar es of pasture (Sorriso City Hall, 2005). Municipality of Tapurah The private enterprise Eldorado was responsib le for colonization in Tapurah, the name referencing an Indian chief of the region. The first family settlement in Tapurah occurred in 1969. In 1981 the district of Tapurah was create d, and on July 4, 1988 th e municipality of Tapurah with an area of 11,600 Km2 was disaggregated from the municipality of Diamantino

PAGE 31

31 (Tapurah City Hall, 2006). In 2002, two new muni cipalities Itanhang and Ipiranga do Norte were disaggregated from the municipality of Tapurah resulting in an area of 4,489.60 Km2. The city of Tapurah is seated 414 k ilometers from Cuiab and 100 kilometers northwest of Lucas do Rio Verde, which is also located along the BR-163 south of Sorriso. The population of Tapurah has significantly in creased; according to IGBE, there were 8,816 people in 1996 and in 2004 there were 13,295 pe ople. The average annual growth rate was 6.87% between 1996 and 2000. Today, Tapurah has 13,735 inhabitants, from which 7,300 inhabitants live in the rural areas (Tapurah City Hall, 2006). Since its colonization, large areas were deforest ed for agriculture, cattle ranching, timber harvesting and settlements. The total defore sted area is 219,900 hectares, from which 150,700 hectares are used for agriculture and 43,000 h ectares for cattle ranc hing; there are 228,969 hectares of forest remaining. In 2005/2006, th ere were 109,500 hectares planted with soybean, 19,000 hectares with corn as a secondary crop, 5,000 hectares with cotton, and 2,000 hectares with rice (Tapurah City Hall, 2006). The agricultural areas are excellent for mechan ized agriculture due to their regular soils, although the smallholder agriculture also exists. In its remaining forested area, selective timber harvesting is practiced. The climate is tropical w ith two well defined seasons (Tapurah City Hall, 2006). Field Methods A total of 40 soybean produ cer interviews and surveys we re conducted during June and July of 2006. Amaggi Group staff presented the res earcher to Amaggis branch managers in the municipalities of Sinop, Sorriso, and Tapurah. The managers assisted the researcher in getting in contact with soybean producers in these regions. The first muni cipality visited was Sinop, then Sorriso, and finally Tapurah. A deci sive factor for farm selecti on was that farmers should grow

PAGE 32

32 soybeans. A second criterion was that half of the producers in each region should be pre-financed by Amaggi Group and the other half not. Thirdl y, the number of farms chosen in each municipality was proportional to th e number of soybean farms in that municipality. Farm size was not a criterion, since most fa rms in the study region are medium to large scale. According to Brazils Institute of Colonization and Agrarian Reform (INCRA), medium farms are from 500 hectares up to 2,000 hectares and large farms are from 2,000 hectares up to 10,000 hectares. The size distribution of the studied farm s is displayed in Table 1-4. From the total of 40 farmers interviewed, 10 interviews occurred in Sinop, where 5 farmers were pre-financed by Amaggi Gr oup and 5 farmers were not; 20 interviews occurred in Sorriso where 10 farmers were pre-financed by Amaggi Group and 10 farmers were not; and finally 10 interviews in Tapurah, where 5 farmers were pr e-financed by Amaggi gr oup and 5 farmers were not. Fewer interviews occurred in the municipa lities of Sinop and Tapurah because there are fewer farmers in these municipalities compared with Sorriso. The pre-financed farmers were randomly select ed from 18 farmers that were pre-financed by Amaggi Group in the region of Sinop, from 51 pre-financed farmers in Sorriso, and from 25 pre-financed farms in Tapurah, all of them pr e-financed in the year of 2006. Joo Shimada (Amaggi Groups Environmental Coordinator) facil itated access to Amaggis Branch Managers in each municipality. Shimada accompanied the resear cher in the field from the beginning of the field work in June 7, 2006 until the beginning of July. His assistance and support was extremely important familiarizing the researcher with the study area and providing background on the soybean production process in th e region. The researcher contacted the selected farmers, and depending on their interest and availability, the interviews were scheduled. Due to time constraints, distance to the fa rms, and farmers convenience, some interviews could not be

PAGE 33

33 conducted on the actual farm sites. Therefore, some farmers were also interviewed in homes, offices, and in Amaggis offices. The farmers that were not pre-financed by Amaggi Group were suggested by Amaggis branch managers, since they knew most of the farmers in the study region, by staff of the Mato Grosso Agriculture and Cattle Ranching Foundati on (Fundao Mato Grosso), especially in Sorriso where the foundation headquarters are lo cated; and by the researcher and Joo Shimadas contacts in Sinop, Sorriso, and Tapurah. The re searcher contacted the farmers, and again according to the farmers interest and time availa bility, the interviews were scheduled in their farms, houses or offices. The researcher encountered some challenges during field research. Th e trip to the study region could not begin before June because soybean farmers were on strike in Brazil. Farmers in Mato Grosso set up blockages of main roads, including the BR-163, demanding better soybean prices and government aid. During this time, s oybean prices were depressed, exchange rates were unfavorable for exports, and in some regi ons there were lower soybean yields due to poor weather during that harvest season. In addition, access to farmers was sometimes made difficult due to tight schedules and Brazils soccer teams participation in the World Cup. Interviews The same semi-structured interviews were applied to all farmers. Interview content included questions concerning land cover and land use distribution, soybean agricultural practices, soybean diseases, pest icide use, and yields (see semi-structured interview in Appendix). More specifically, farmers were asked about the size of their properties, if it was leased or owned, the hectares planted with soybeans and other annual crops, hectares with pasture, forested area and reforested area. They were also asked whether or not they were currently

PAGE 34

34 adopting a no tillage system in their soybean area and for how long th ey had practiced this system, how many hectares of soybean area were being covered by different cover crops, and planted with GM soybeans and nematode resistant cultivars. With regard to soybean diseases and pestic ide use, farmers were asked what kind of soybean disease they consider th e most problematic, what kinds of insect infestation, pathogenic disease, and nematodes they had in the 2006 soybean harvest season, and which chemical pesticide was used for each stage of soybean production. Data on the number of fungicide applications and the area applied with desiccant prior to soybean harvest were also collected. With regard to soybean productivity, farmers we re asked about their soybean yields for the 2005/2006 harvest. Since soybean production costs ar e often kept confidentia l, these costs were only obtained for two farms in Sinop and serve as the basis for the case study presented in Chapter 4. These data were obtained for Sinop since the farmers in this muni cipality appeared to be more receptive.

PAGE 35

35 Table 1-1. Required widths for Permanent Preser vation Areas (APP) according to width of the watercourse. Source: Brazil s 1965 New Forestry Code (4.771). Water courses width APP width Up to 10 meters 30 meters* Between 10 and 50 meters 50 meters Between 50 and 200 meters 100 meters Between 200 and 600 meters 200 meters Larger than 600 meters width 500 meters *In Mato Grosso the require d APP width is 50 meters.

PAGE 36

36 Table 1-2. Legal Reserve (RL) modificati ons since Brazils 1965 New Forestry Code implementation. Biome: Forest (F), Cerrado (C), Ecotone (E). Source: Shimada, 2007. Date legislation Main topics established RL for regions* 1934 Decree 23.793 F 25% Reserve of on forested lands; rural properties located close to forests are exempted. C 0% Sep 15, 1965 New Forestry Code law 4.771 F 20% S; 50% N Reserve of 20% for area with shrubs and 50% for north region. C 0% Jul 18, 1989 Law 7.803 F 20% S; 50% N RL definition and registration requi rement; Cerrado region is included. C 20% Jan 17, 91 Law 8.171 F 20% S; 50% N Reforestation requirement for RL of 1/30 for each year. C 20% 1995 Mato Grosso Complementary law 038-MT F 20% S; 50% N RL of 50% in transition areas, which depends on regulation. C E 20% 50% in MT Aug 22, 1996 MP 1.551 F 20% S; 50% N; 80% E Changes in RL percentages for nor thern regions and northern MT C 20% S; 50% N Dec 11, 1997 MP 1.605 F 20% S; 50% N; 80% E Exceptions of the 80% RL for INCRA settlements, areas under than 100 ha and used for family agriculture. C 20% S; 50% N Dec 14, 1998 MP 1.736 F 20% S; 50% N; 80% E Reforestation requirement abolished; RL modified for Cerrados region. C 20% Jul 18, 1999 MP 1.885 F 20% S; 50% N; 80% E Indigenous area is considered APP. APP can be calculated as RL. C 20% Jan 06, 2000 MP 1.956 F 20% S; 50% N; 80% E Compensation of RL permitted in other areas within the same microhydrological basin. C 20% May 26, 2000 MP 1.956-50 F 80% LA; 20% others Deforested areas after 12/14/98 can not be compensated. Agrarian reform projects are forbidden in fo rested areas. Changes in the % RL. C 35% LA; 20% others Jun 26, 2000 MP 1.956-51 F 80% LA; 20% others Cerrado area for Legal Amazon: 20% RL and 15% compensation area. Small properties planted with fruc tiferous, ornamental and exotics species can count towards RL requirement. C 35% LA; 20% others Jul 26, 2000 MP 1.9956-52 F 80% LA; 20% others APP can count for RL if APP + RL > 50%; and 80% for Legal Amazon region C 35% LA; 20% others Aug 23, 2000 MP 1956-53 F 80% LA; 20% others The owner is exempted for paying taxes for 30 years if he/she donates the RL for forest reserve C 35% LA; 20% others Sep 21,2000 MP 1.956-54 F 80% LA; 20% others Reforestation requirement for RL 1/10th of area every 3 years. C 35% LA; 20% others Mar 22, 2001 MP 2.090-61 F 80% LA; 20% others Forest management in indigenous areas is allowed. C 35% LA; 20% others Aug 22, 2001 MP 2.166-67 F 80% LA; 20% others Currently in effect C 35% LA; 20% others *S = South; N = North; E = Ecotone; LA = Legal Amazon states.

PAGE 37

37 Table 1-3. Soybean production, planted and harvested area for Brazil, Mato Grosso, and the municipalities of Sinop, Sorriso and Tapur ah for year 2004 and 2005. Source IBGE Municipal Agricultural Production. Production (tons) Planted area (h ectares) Harvest area (hectares) 2004 2005 2004 2005 2004 2005 Brazil 49,549,941 51,182,074 21,601,340 23,426,756 21,538,990 22,948,874 Mato Grosso 14,517,912 17,761,444 5,279,928 6,121,724 5,263,428 6,106,654 Sinop 243,395 375,417 84,495 130,326 84,495 130,326 Sorriso 1,688,120 1,804,669 547,867 582,356 540,867 578,356 Tapurah 719,808 332,640 260,800 109,500 260,800 108,706 Table 1-4. Selected farm size distribution according to INCRA in Sinop, Sorriso and Tapurah. Farm size Hectares # of selected farms Percentage Very small < 50 0 0 Small 50 499 4 10 Medium 500 1,999 19 48 Large 2,000 9,999 14 35 Very large 10,000 3 8

PAGE 38

38 Figure 1-1. Some cover crop options for the Cotton Soybean rotation system. [Source: Altmann, N. 2006. Rotao, sucesso e consor cio de espcies para agricultura sustentvel. Boletim de Pesquisa de Soja 2006 (Page 237, Figure 2). Fundao Mato Grosso, Rondonpolis, Mato Grosso.] Figure 1-2. Map of the study re gion. Municipalities of Sorri so, Sinop and Tapurah located northern Mato Grosso. [Source: Wikipedia, 2007. Available at: http://pt.wikipedia. org/wiki/Sinop. Modified by the author.]

PAGE 39

39 Figure 1-3. Highway BR-163 and its area of influe nce. [Source: Ministrio do Meio Ambiente, 2005. Plano de Desenvolvimento Regional Susten tvel para a rea de Influncia da Rodovia BR-163 Cuiab Santarm.]

PAGE 40

40 CHAPTER 2 SOYBEAN MANAGEMENT PRACTICES Introduction The purpose of this chapter is to analyze the agricultural management practices adopted by soybean farmers in Sinop, Sorriso, and Tapurah, locat ed in northern Mato Grosso, Brazil. Since the Amaggi Group requires social-environmental re sponsibility from its pre-financed producers, the management practices adopted by farmers who were pre-financed by Amaggi Group in the study region in the year of 2006 will be compared to those who were not pre-financed. Amaggi Group requires legalization of legal reserves, re cuperation of riparian areas, and no illegal deforestation for the duration of the contract. With regard to good farming practices, adoption or increase in the area farmed as a no-tillage syst em and the implementation of integrated pest management are recommended, although not required. Because the study farms are located in differe nt biomes, for example, farms located in Sinop are in the Amazon forest and farms in Sorriso are in the Cerrado regi on, it is expected that farms would differ in their percentage of forest cover. Moreover, the municipality of Sorriso is known for its important contribution to the grain producing industry, which leads to the expectation that producers in Sorr iso designate a higher farm area to plant soybeans, as well as corn as a secondary cash crop. Therefore, the agricultural practices of farms will also be compared according to their respective municipali ties. The following hypotheses will be tested: Hypothesis 1: Farms pre-financed are more likely to preserve forested area than those that were not pre-financed; Hypothesis 2: Farms in Sorriso are less likely to pr eserve forested areas than the farms located in Sinop and Tapurah; Hypothesis 3: Farms pre-financed have a greater per centage of soybean area in a no tillage system; Hypothesis 4: Farms pre-financed have greater cover crop diversity; and

PAGE 41

41 Hypothesis 5: The producers located in Sorriso are more likely to plant corn as a second crop than those farmers located in Sinop and Tapurah. To accomplish this task, this ch apter is structured as follow: First, the methods section describes the data analysis a nd how the descriptive and statis tical analyses, linear regression models and weighted means were conducted. Late r, the results of land use, soybean planting practices and soybean pests (the latter included in this Chapter as an introduction to Chapter 3), are displayed; finally the discussi on section comments on the hypotheses. Methods Data Analyses Data on agricultural practices collected in J une and July 2006 from 40 interviewed soybean farmers were organized according to land use an d soybean planting practices. Land use describes the physical and productive characteristics of the properties. The variables included in land use are: farm size, years since land purchase and the most recent deforestation, distance from municipal seat, number of head of cattle, percen tage of farm area covered by the most recent deforestation, and percentages of farm area in na tural forest, reforestation, soybean cultivation, other annual crops and pasture. Leased land is not considered here due to a lack of data. It is important to note that the variable percentage of natural forest on farm is part of the legal reserve, but many farmers have additional area off-site as compensation areas. However, it is not the objective of this study to verify if this percentage is in compliance with the law. Therefore, what is reported here is natural forest preserved on farm. Soybean planting practices describe the manage ment practices adopted in areas cultivated with soybeans, in this case leased land is in cluded because it is directly under the farmers management. The variables reported are: area plante d with soybeans; percentage of soybean area that is planted with genetically modified (GM) soybeans and resistant cultivars; years between

PAGE 42

42 land purchase and first implement ac tion of a no tillage system; a nd percentages of soybean area with different cover crops: corn, millet, other cover crops, and area left in fallow. With regard to soybean diseases, the study va riables are: the percen t of farmers reporting insect pests, pathogenic diseas es and plant parasitic nematode s. These variables, although not statistically analyzed, provide a sn apshot of the soybean diseases and insect pests that occurred in the 2006 crop. Descriptive Analyses Descriptive analyses with regard to land us e, soybean planting pract ices, cover crops and soybean diseases were conducted between farms pr e-financed and not pre-financed by Amaggi Group named as financial groups, and among fa rms located in the municipalities of Sinop, Sorriso and Tapurah named as municipal groups. Statistical Analyses Comparisons of independent sample means were conducted to test fo r differences in land use and soybean planting practices at a 95% confid ential interval between financial groups and among municipal groups. The statis tical analysis was conducted as follows: first, a pre-test concerning two-population variances (F-test) was conducted to veri fy if the variances between groups were equal (Ott and Longnecker, 2004). If the variances were equal, a two-tailed t-test for equal variances was conducted; if unequal, a two-tailed t-te st for unequal variances was conducted. Both types of tests (equal or une qual variances) were ca rried out at the 95% confidence intervals. Regression Analysis Two simple linear regression models were formul ated to explain the percentage of natural forest on farms and another simple linear regres sion model was formulated to explain percentage of deforested area in the most recent deforest ation. The explanatory variables for each simple

PAGE 43

43 linear regression respectively are: farm size, y ear of land purchase and year of most recent deforestation. Then, each model was formulated in to financial group and municipal group to test the significance of each linear regression. A multiple regression model was developed to test if the variables (1) years on farm, (2) kilometers from municipal seat, (3) farms lo cated in Sinop and Sorriso and (4) farms prefinanced by Amaggi Group affect th e percentage of natural forest left on farm. Dummy variables were assigned for farm location and if the farm s were pre-financed or not by Amaggi Group. The dummy-coded variable 1 was give n to farms pre-financed by Amaggi Group and 0 to the farms not pre-financed by Amaggi Gr oup. The dummy-coded variable 1 was given to farms located in Sinop and 0 to the other farms, in another column the same procedure was done to farms located in Sorriso. Weighted Means The weighted means for area in each cover cr op were calculated to analyze the importance of different cover crops. The total hectares of each cover crop were summed and divided by the total soybean plantation area (on both owned and leased land) and displayed in pie charts. The least common cover crops reported by farmers were summed together for better visualization in the pie chart, named as other cover crops. This procedure was performed separately for financial groups and municipal groups. Results Land Use Comparisons between financial groups There is a great heterogeneity among farms in each group as well as within groups in land use (Table 2-1). Both groups have a greater perc entage of farm area planted to soybeans rather than pasture, other annual crops, natural forest and reforestation. For the pre-financed farms,

PAGE 44

44 60% is in soybeans, 33% is natu ral forest, 1.6% is other annual crops, 1% is pasture, and 0.09% is reforestation. For the not pre-financed farms, 62% is in soybeans, 32% is natural forest, 2.1% is pasture, 1% is other annual crops and 0.13% is reforestation. Comparisons of means between financial groups do not reveal any statistically significant differences. Other means related to land use show that the pre-financed fa rms have occupied the same land for an average of 16 years and are located an average of 53 kilomete rs from the municipal seat, while the non pre-financed farms have occupi ed the area for an aver age of 11 years and are located an average of 32 kilome ters from municipal seat; these differences are statistically significant (P<0.1). On average, pre-financed fa rms have 66 head of cattle while the not prefinanced farms have 270 head of cattle. Six year s has elapsed since the mo st recent deforestation for both groups: the pre-financed farms deforested 19% of the area in the last deforestation and the not pre-financed farms deforested 21% of the area. The differences in head of cattle per farm and percentage area of the most recent defo restation are not sta tistically significant. Comparisons among municipal groups Comparisons among municipal grou ps reveal some similarities and differences in land use (Table 2-2). More than half of the average farm area is used to plant soybeans in all municipal groups. For farms in Sinop, 56% is in soybeans, 37% is natural forest, 2.3% is other annual crops, 1.7% is pasture and 0.12% is reforestation. For farms in Sorriso, 68% is in soybeans, 27% is natural forest, 0.6% is past ure, 0.35% is other annual crops and 0.09% is reforestation. For farms in Tapurah (most remote m unicipality), 52% is in soybeans, 40% is na tural forest, 3.6% is pasture, 2.3% is other annual crops and 0.23 % is reforestation. Statistica l comparisons of means reveal that the percentage of area of soybean cultivation and of natural forest are significantly different (P<0.05) between Sorriso and Sinop and between Sorriso and Tapurah.

PAGE 45

45 Other means related to land use show that farms in Sinop have been occupying the same land for an average of 11 years and they are an average of 35 kilometers from the city of Sinop. Farms in Sorriso have an average of 13 years occupying the same land and are an average of 56 kilometers from the city of Sorriso. Farms in Tapurah have an average of 17 years occupying the same land and they are an average of 22 kilomete rs from the city of Tapurah. Comparisons of means show that there is a sta tistically significant difference (P< 0.05) in distance from municipal seat between farms locate d in Sorriso and Tapurah. On average, there are 37 head of cattle per farm in Sinop, 137 in Sorriso and 338 in Tapurah. On farms in Sinop, it has been an average of 5 years since the most recent deforestation, which deforested 18% of farm area. On farms in Sorriso, it has been an average of 6 years since the most recent deforestation, which deforested 22% of farm area. Finally, on farms in Tapurah, it has been an average of 8 year s since the most recent deforestation, which deforested 22% of farm area. Despite these diffe rences in means, they are not statistically significant. Linear regression models A linear regression shows that there is a pos itive and significant (P<0.05) relationship between the percentage of natural forest a nd farm area (Figure 2-1). However, separate regressions for financial groups (Figure 2-2) reveal that this relationshi p is statistically significant only for the not pre-financed fa rms (P<0.05); considering the municipal groups (Figure 2-3), the regression is significant onl y for the farms located in Sorriso (P<0.05). There is a slight positive relationship between th e percentage of natura l forest and the year of land purchase (Figure 2-4), that is, the mo re recently the land purchase, the higher the percentage of natural forest. However, this is no t a statistically significant relationship. Separate regression for financial groups (F igure 2-5) does not show a si gnificant relationship; however,

PAGE 46

46 considering the municipal groups (F igure 2-6), there is a statistically significant relationship for the farms in Sinop and Sorriso (P<0.05). There is a positive though not statistically si gnificant relationship between the size area of the most recent deforestation and the year of oc currence (Figure 2-7). This appears to be mostly related to the financial groups (Figur e 2-8) and in Tapurah (Figure 2-9). A multiple regression model was developed to describe the percentage of farm area with natural forest. The model is specified as follows: 0123456 (2-1) NFFSYPKMPFSISO Where: NF = Percentage area of natural forest FS = Farm size in hectares YP = Number of years since farm purchase KM = Distance from munici pal seat in kilometers PF = 1 if the farm is pre-financed SI = 1 if the farm is in Sinop SO = 1 if the farm is in Sorriso The model is statistically significant (P<0.05) and shows that 45.6% of the variance in percentage of natural forest can be predicted fr om the independent variables FS, YP, KM, PF, SI and SO (Table 2-3). The variables FS, KM and SI have a positive effect on the dependent variable. An increase in one of these variables, while holding the other variables constant, results in an increase in the percentage of natural forest. However, only the variable FS is statistically significant (P<0.05). The other variables (YP, PF and SO) have a negative effect on the percentage of natural forest. Therefore, an incr ease in one of these variables, while holding the

PAGE 47

47 other variables constant, reduces the percentage of natural forest. Only the variable SO is statistically significant (P<0.1). Soybean Planting Practices Comparisons between financial groups Comparisons between financial gr oups present some similarity and differences in soybean planting practices (Table 2-4). The total soybean area planted on both owned and leased land for the pre-financed farms averages 1865 hectares, of which 8% is planted with genetically modified (GM) soybeans and 6% with nematode resistan t soybean cultivars. Not pre-financed farms average 1707 hectares of soybeans, of which 2% is planted with GM soybeans and 6% is planted with resistant cultivars. However, there are no statistically significant differences between the financial groups. Regarding to no-tillage system for soybean s, this practice was implemented by prefinanced farmers an average of 6 years after land purchase, while not pre-financed farmers implemented it an average of 8 years since land was purchased. However, this difference is not statistically significant. The cover crops for the no-till system adopted by financial groups ar e as follows: the prefinanced farms planted 33% of their soybean area with corn, 44% with millet, 8% with other cover crops and 15% of soybeans we re left in fallow; the not prefinanced farms planted 31% of their soybean area with corn, 64% with millet, 5% with other cover crops and 0.5% of soybeans were left in fallow. Comparisons of means reveal s that percentage of soybean area with millet cover crop and percentage of s oybean area left in fallow are st atistically different (P<0.05) between the financial groups. Figure 2-10 shows the percentage of farmers adopting each type of cover crop for the notillage system. More than 80% of pre-financed farmers reported using corn and millet as their

PAGE 48

48 cover crops, while up to 10% reported using sorghum, Brachiaria ruziziensis sunflower, rice, cotton and finger millet as cover crops. More than 75% of not pre-financed farmers also reported using corn and millet as their favorite choices of cover crops, while up to 20% of them reported using only sorghum and rice as cover crops. Weighted means is another technique used to analyze the cover crops adopted by the financial groups. This method (Figure 2-11) reveal s that for the pre-financed group, 41% of total soybean area is planted with m illet, 31% with corn, 11% with other cover crops, and 17% left in fallow. Cover crop data for the not pre-financed group reveal that 60% of total soybean area is planted with millet, 32% with corn, 8% with other cover crops and no land left in fallow. Comparison among municipal groups Table 2-5 shows soybean planting practices (including owned and leased land) among the municipal groups. Farms in Sinop average 1690 hect ares of soybeans in which 1% is planted with GM soybeans and 6% with resistant cultivars. Farms in Sorriso average 2112 hectares of soybeans in which 9% is planted with GM soybean s and 4% with resistant cultivars. In the case of Tapurah, farms average 1232 hectares of s oybeans in which 2% are planted with GM soybeans and 10% with resistant cultivars. Compar ison of means reveals that soybean plantation area is statistically different (P<0.1) between farms in Sorriso and Tapurah. Farmers in Sinop implemented a no-tillage syst em for soybean plantations an average of 6 years after land purchase; farmers in Sorriso im plemented it an average of 5 years after farm purchase; and farms in Tapurah implemented a notill practice an average of 12 years since the land was purchased. Comparisons of means reveal s that there is a statistically significant difference (P<0.05) between farms in Sorriso and Tapurah. Farms in Sinop planted 26% of their soybean ar ea with corn, 53% with millet, 1% with other cover crops and 19% of the soybean area wa s left in fallow; farms in Sorriso planted 37%

PAGE 49

49 with corn, 50% with millet, 9% with other cover crops and 4% was left in fallow; and farms in Tapurah planted 28% of soybean area with corn, 62% with millet, 7% with other cover crops and 4% of soybean area was left in fallow. Howeve r, none of these differences are statistically significant between the municipal groups. Figure 2-12 shows the percentage of soybean fa rmers adopting a variety of cover crops for the no-tillage system. More than 60% of the farm ers in all municipal gr oups reported using corn and millet as cover crops, while up to 10% of farmers in Sinop reported using sorghum, rice and finger millet as cover crops; up to 10% of farm ers in Sorriso reported using sorghum, rice and cotton; and up to 20% of farmers in Tapurah reported using sorghum, Brachiaria ruziziensis and sunflower as cover crops. Analysis of the weighted means in the muni cipal groups (Figure 2-13) reveals that, in Sinop, 38% of total soybean area is planted with millet, 28% with corn 5% with other cover crops and 29% left in fallow. In Sorriso, cover crops are 52% millet, 33% corn, 12% other cover crops and 3% of total soybean area was left in fallow. In Tapurah, cover crops are 59% millet, 31% corn, 8% other cover crops and 2% of total soybean area was left in fallow. Soybean Pests Comparisons between financial groups Most of the soybean farmers consider that soybean pathogenic diseases such as Asian soybean rust ( Phakopsora pachyrhizi ) or anthracnose ( Colletotrichum truncatum ) are more problematic than insect infe station such as whitefly ( Bemisia spp.) or stinkbugs and plantparasitic nematodes (Figure 2-14). With regard to pathogenic diseases (Figur e 2-15), all farmers reported having Asian soybean rust in their soybean crops, 50% of pr e-financed farmers reported having anthracnose

PAGE 50

50 disease and 20% repor ted foliar blight ( Rhizoctonia solani / Thanatephorus cucumeris ); 45% of not pre-financed farmers reported having an thracnose disease a nd 35% foliar blight. With regard to insect infestation (Figure 216), 85% of the soybean farmers had problems with whitefly. While 80% of the not pre-finan ced farmers had problems with stinkbug and caterpillar ( Anticarsia gemmatalis ) infestation, only 60% and 55% of the pre-financed farmers had stinkbugs and caterpillar infest ation respectively. Almost half of the farmers in both financial groups reported having nematodes in their soybean crops (Figure 2-17). Comparisons among municipal groups Most of the soybean farmers consider that s oybean pathogenic disease is more problematic than insect infestation and plan t-parasite nematodes (Figure 214). With regard to pathogenic diseases (Figure 2-15), all farmers reported havi ng Asian soybean rust in their soybean crops. Farmers in Sinop did not report any foliar blight disease and 30% repor ted anthracnose. In Sorriso, 35% of the farmers had anthracnose and 35% had foliar blight disease, while in Tapurah 90% of the farmers had anthracnose and 40% had foliar blight disease. With regard to insect infestation (Figure 2-16) in Sinop, 80% of th e farmers had whitefly and stinkbug and 70% had caterpilla r in their soybean crops. In So rriso, 95% of the farmers had whitefly, 55% had caterpillar, and 50% had st inkbug. All farmers in Tapurah had stinkbug, 90% had caterpillar, and 70% ha d whitefly infestations. Although a minority of farmers reported having pr oblems with other insect pests such as soybean looper ( Pseudoplusia includens ) and lesser cornstalk borer ( Elasmopalpus lignosellus ), farmers in Sinop did not report th em. While 63% of the farmers located in Sorriso reported having nematodes in their soybean crops, approximately 30% of th e farmers located in Sinop and Tapurah reported having them (Figure 2-17).

PAGE 51

51 Discussion Land Use The preceding analyses demonstrate that there were no significant differences in land use when comparing farms pre-fina nced and not pre-financed by Amaggi Group. However, when comparing municipal groups, some differences can be noticed. Comparison between financial groups Farms pre-financed farms by Amaggi Group did not conserve more forest than the not prefinanced farms, therefore, rejecting the first hypothesis. Since Amaggi Group requires that their pre-financed farms have their Legal Reserve legalized and they do not allow illegal deforestation, it was expected that these standard s would have a positive effect on natural forest left on pre-financed farms. However, statistical analysis revealed that the mean percentage of natural forest on pre-financed farms is not sign ificantly different from the not pre-financed farms. Multiple regression analysis lends support to the result that whethe r or not a farm is prefinanced does not have a signifi cant effect on percentage of natural forest on farm. This result may be explained by the fact that companie s other than Amaggi Group also provide prefinancing policies such as lega l reserve officially recognized. Although the average year s since the land purchase and dist ance from municipal seat were statistically different between pr e-financed and not pre-financed farms, these variables did not have a significant effect on forested area. Having specific percentages of natural forest preserved on the farms is a landowner obligation according to federal law, however it is unknown if the areas comply with current legislation. In Mato Grosso, the calculati on of the legal reserve requireme nt for each farm is based on the map of forest typology from the project RADAM, which was created in the 1970s (J. Y. Shimada, personal communication, May 21, 2007). Acco rding to types of vegetation inside the

PAGE 52

52 property, the required percentage of legal reserve is calculated for each fraction of farm area and the area is summed. The percentage of legal reserve that is applie d also depends on the date that the property was registered, si nce legislation and forest t ypology has change d since the 1965 New Forestry Code. In Mato Grosso, before adopting the RADAM proj ect, the legal reserve requirement was based on the map of economic-ecological zoning map. Therefore, a farms percentage of legal reserve rarely matches with the current legislation, gi ven potential diversity of vegetation types on the farm and th e date the property was registered. Comparisons among municipal groups The producers located in Sorriso have preserved less forested areas than those in Sinop and Tapurah, as asserted in the second hypothesis. The municipality of Sorriso is located to the south of Sinop and less deeply embedded in the Amazon forest. The environmental law says that 80% of the property should be protected natural forest if situated in the forest region and 35% of the property should be protected natu ral forest if located in the Ce rrado. Therefore, it was expected that farms in Sorriso would have less forested ar ea on farm than the farms located in Sinop, since Sinop is located in the Amazon forest biome. Multiple regression analysis also confirmed that if the farm was located in Sorriso, it had a negative and significant effect on th e percentage of natural forest on farm. Therefore, farmers in Sorriso are less likely to preserve natural forest. Even though farm s located in Tapurah are closer to the municipal seat than the farms in Sorriso, the municipality of Tapurah is not situated along the BR-163. This could have contributed to a greater percentage of natural forest on farms in this region compared to those located in Sorriso. The municipality of Sorriso is considered th e largest soybean producer in Brazil. This is reflected in the data that show the mean per centage of soybean area in the studied farms is statistically greater in Sorriso than in Tapurah and Sinop. One pa rticular detail noticed by the

PAGE 53

53 researcher is that there are also a much larger number of soybean farmers in the municipality of Sorriso than in Sinop and Tapura h. Another possibility is that, since farms in Sorriso are less likely to preserve natural forest on farm than th e others municipalities as statistically tested, farmers use this deforested land for planting soybeans. Soybean Planting Practices The analyses demonstrated that there are few significant differences in no-tillage systems and cover crops adopted by soybean farmers when comparing these practices between financial groups, and no difference are reve aled among municipal groups. Comparison between financial groups The pre-financed farms do not have a greater percentage of soybean area in a no-tillage system, therefore rejecting th e third hypothesis. Even though Amaggi Group recommends that their pre-financed farms adopt the no-tillage sy stem as a good farming practice, results revealed that pre-financed farms had more soybean area with no cover crop (left in fallow) than the not pre-financed farms. That is, pre-financed farms have less ar ea with cover crop residues left unincorporated on soil surface (to plant soybean on) than the not pre-financed farms. Another reason for pre-financed farmers havi ng less soybean area planted with cover crops might be problems with soil erosion; four pre-fi nanced farmers and one not pre-financed farmer reported having soil erosion. Where er osion is problematic, these areas are typically left to fallow for a number of years. Although there is no stat istically significant difference in the other cover crops planted (besides millet) between the financial groups, survey data demonstrates that pre-financed farmers are adopting a greater variety of cover crops Growing different cover crops results in diversification of production, improved yields of subsequent crops, a nd reductions in pest infestation as the lifecycle of pests and pat hogens is interrupted. Th e pre-financed farmers

PAGE 54

54 reported using four more types of cover crops th an the not pre-financed farmers, although this was not statistically tested due to a small sample size. For this reason, it was not possible to test the fourth hypothesis that th e pre-financed farmers have greater cover crop diversity. However, this survey data might demonstrate that some soybean farm ers are interested in diversifying their production. Even though the number of farmers adopting these different cover crops is small, it may suggest that soybean farmer s are starting to search for a more sustainable agricultural system as demonstrated by the integr ation of crops and lives tock with the adoption of Brachiaria ruziziensis With regard to GM soybeans, even though Amaggi Group silos receive them, the area planted with GM soybeans by pre-financed farmers is low and not statistically different from the not pre-financed farmers. This may be explained by the fact that pre-financed farmers are receiving better prices for non-GM soybean, sinc e Amaggi export them through a private port, which reduces transportation costs. Comparisons among municipal groups Although the municipality of So rriso is considered the four th largest corn producer in Brazil, the farmers located in Sorriso do not plant mo re corn as a cover crop than those located in Sinop and Tapurah. Results showed that there is no statisticall y significant difference between the municipal groups with regard to corn as cover crop. Therefore, the fifth hypothesis that farmers in Sorriso are more likely to plant corn as a cover crop than those farmers located in Sinop and Tapurah is rejected. According to Piccoli, the president of the Sorriso rural workers union (A Gazeta, 2007), crop production of corn as a secondary crop can be delayed by the weather; too much rainfall can delay the harvest of soybeans a nd consequently delay the corn planting, which has a specific period to be planted. As reporte d by many farmers, especially in Sinop, too much rainfall in 2006

PAGE 55

55 prevented farmers from harvesting the whole soybean plantation and planting corn at the appropriate time. The weather mi ght be the reason why fewer fa rmers in Sinop, reported planting corn as a cover crop for that year. The variety of cover crops reported by the m unicipal groups reveals that at least three different cover crops besides corn and millet ar e being adopted by farmers in each municipality. Although the number of farmers adop ting these other cover crops is small and not statistically tested, it might demonstrate that the search for a sustainable agriculture is not concentrated only in one region. With regard to GM soybean, results sugge st that although GM soybeans are legally permitted to be planted in Mato Grosso (a pproved two years ago by the Brazilian government (USDA, 2007)), farmers in the study region are currently not readily adopting them since the percentage of area planted with GM soybeans wa s low. According to Fundao Mato Grosso (2006), the production costs for a non-GM soybean crop is very similar to a GM soybean crop, because the high price of RR seeds (genetical ly modified herbicide-resistant Roundup Ready soybeans) and the payments of royalties cancel a ny positive effects with the savings gained with the use of only one post-emergent herbicide such as Roundup. Soybean Pests Although there are no hypotheses re lated to soybean diseases, it is important to compare the pathogenic diseases, insect infestation a nd plant parasitic nematodes between financial groups and among municipal groups in order to better understand Chapter 3, which will discuss pesticide use. Comparisons between financial groups The results showed that there are no major differences between financial groups with regard to pathogenic diseases; for example, all farmers reporte d having Asian rust disease in

PAGE 56

56 their soybean crop. According to Yo rinoni et al. (2004), the first re port of Asian soybean rust in Brazil was at the end of the 2001 harvest season in the state of Paran in southern Brazil. Since then, this disease has spread over the soybean producer states Transported by wind, and when favorable weather conditions exist, such as hi gh air moisture and with temperatures between 18C and 26C, this rust can cause substantial losses in productivity (Fundao Mato Grosso, 2006). The foliar blight ( Rhizoctonia solani ) disease which had disappeared from soybean crops in the northern region of Mato Grosso, reappeared in the 200 5/2006 harvest period especially where the area was not covered with cereal c over crops (Folha do Estado, 2006). Some farmers from both groups reported having foliar blight in their soybean crop. All the pathogenic diseases reported by farmers depend on high moisture to proliferate, and according to the farmers interviewed, rainfall was greater than average in 2006. Regarding insect infestation, more not pre-financed farmer s reported having stinkbugs and caterpillar in their soybean crop. Stinkbugs ha ve been more prevalent since 2001, the most common species of which are Nezara viridula Piezodorus guildinii Dichelops melacanthus Dichelops furcatus Euschistus heros and Thyanta perditor (Gassen, 2002). Damage is caused by the nymphs and adults sucking sap from the bean pods, resulting in reduced soybean quality and foliar retention at the end of the soybean cycle, which can complicate mechanized harvest. Together with defoliator caterpillars, they repr esent the principal insects that are controlled through pest management (Degrand and Vivan, 2006). About 85% of farmers in both groups reported having whiteflies in th eir crop. Direct crop damage occurs when whiteflies feed in plant phloem, remove plant sap and reduce plant vigor. Whiteflies also excrete honeydew, which prom otes sooty mold that interferes with

PAGE 57

57 photosynthesis and may lower harvest quality (Fasulo, 2006). Infestation of whiteflies has become common in soybean cultivars in th e last few years. The whitefly species Bemisia tabaci and Bemisia argentifolii are worrying the soybean producers in Mato Grosso due to the difficulty in controlling them in an economically feasible manner (Degrand and Vivan, 2006). Almost 50% of farmers from both groups repor ted having nematodes in their soybean crop and half of these farmers reported having cyst nematodes ( Heterodera glycines ). When an area is infested with nematodes, nothing can be done in the actual harvest period, rather, precautions need to be focused on the next crop on that speci fic site (Dias at al., 2006). For a sustainable production system where there are occurrences of cyst nematodes on site, crop rotation with a non-host cultivar is the best st rategy to manage nematodes. Corn can be an option as a cover crop in regi ons with occurrence of soybean cyst nematode (Altmann, 2006). Crop rotations with corn break th e nematodes reproductive cycle by providing an unfavorable host. Though this strategy does not completely eliminate nematodes, it does reduce their population on site; the objective is to lower numbers enough so that the next susceptible crop is successful. Leaving land fallow is also a po ssible solution, since it starves nematodes if the area is 100% free of weeds. Ot her problems may result, however, particularly soil erosion and surface runoff (Powers and McSorley, 2000). Comparison among municipal groups High levels of precipitation in December 2005 in northern Mato Grosso contributed to the appearance of Asian rust and folia r blight in various soybean pr oducing municipalities (Folha do Estado, 2006). Anthracnose is one of the prin cipal soybean diseases in the Cerrado region (EMBRAPA, 2004) and some farmers in all the study municipalities reported having this disease in their crops.

PAGE 58

58 Many more than half of the farmers in all municipalities reported having whitefly infestation. According to Degr ande and Vivan (2006), the m unicipalities of Sinop, Sorriso, Tapurah, and other regions in mid-north Mato Grosso had huge outbreaks of this pest in the last few years. In places where precipitation is less intense, whiteflies pose less risk to soybean cultivation. The farmers in Sinop reported having root-knot nematodes ( Meloidogyne spp ) in their soybean crop To manage this species of nematode, the most efficient and cost effective strategy is for farmers to plant resistant cultivars. In th e statistical analyses, resu lts showed that a small percentage of soybean area was planted with re sistant cultivars in all three study regions. Currently many soybean cult ivars are resistant to Meloidogyne spp. in Brazil (Dias at al., 2006). More farmers in Sorriso reported having Heterodera glycines than Meloidogyne spp., and farmers in Tapurah reported having only Heterodera glycines As mentioned before, when nematodes are f ound in the soybean crop, it is important to identify the nematode species in order to desi gn an effective management program. For example, one farmer from the municipa lity of Sorriso detected Heterodera glycines nematode in his crop. To combat this pest, the farmer reported that he was going to plant 200 h ectares of his soybean area with a resistant cultivar in the following planting season. Conclusion This chapter shows that although soybean fa rmers in Sinop, Sorriso, and Tapurah located in northern Mato Grosso adopted similar soil conservation practices, su ch as the no-tillage system and the maintenance of forest cover, there are some differences in the percentage of farm area where these techniques are practiced. With re gard to financial groups, the main difference is that pre-financed farms have a higher percentage of soybean area left in fallow than the farms

PAGE 59

59 that were not pre-financed. With regard to muni cipal groups, the main diff erence is that farmers in Sorriso are less likely to preserve fore st cover than farmers in Sinop and Tapurah.

PAGE 60

60 Table 2-1. Descriptive statistics and comparison of means of la nd use between financial groups: farms that were pre-financed (PRE) and farm s that were not. Table shows number of farms (N); minimum, maximum, mean, and standard deviation of values reported; and t-test statistics for comparisons. N Min. Max. Mean Std. Dev. t-value P Farm size (ha) PRE 20 325 14000 3349 3543 0.414 0.682 NOT 20 121 13300 2891 3460 Years since land purchase PRE 19 0 30 11 10 -1.792 0.081* NOT 20 3 28 16 9 % Natural forest (% of farm PRE 20 15% 67% 33% 14% 0.377 0.708 area) NOT 20 14% 75% 32% 15% Years since the most PRE 12 2 14 6 4 0.140 0.890 recent deforestation NOT 18 2 17 6 4 Area of the most recent PRE 12 4% 58% 19% 19% -0.298 0.768 deforestation (% of farm area) NOT 18 4% 83% 21% 21% Reforestation (% of farm PRE 20 0% 1.4% 0.09% 0.31% -0.344 0.733 area) NOT 20 0% 1.7% 0.13% 0.40% Soybean cultivation (% of PRE 20 27% 84% 60% 17% -0.273 0.787 farm area) NOT 20 18% 83% 62% 17% Other annual crops (% of PRE 20 0% 11% 1.6% 3.4% 0.528 0.601 farm area) NOT 20 0% 11% 1.0% 3.0% Pasture (% of farm area) PRE 20 0% 9% 1.1% 2.1% -0.904 0.372 NOT 20 0% 20% 2.1% 4.7% Head of cattle (per farm) PRE 11 5 300 66 82 -1.580 0.140 NOT 12 0 1100 270 438 Distance from municipal seat PRE 13 3 150 53 40 1.809 0.089* (Km) NOT 14 8 60 32 17 *indicates t-value significant at P<0.1.

PAGE 61

61 Table 2-2. Descriptive statistics and compar ison of means of land use among municipal groups: farms located in the municipalities of Sinop (SIN), Sorriso (SOR) and Tapurah (TAP). Table shows number of farms (N); minimum, maximum, mean, and standard deviation of values reported; and t-test statistics for comparisons. Municipality* N Min. Max. Mean Std. Dev t-value P Farm size (ha) 1 SIN 10 121 14000 2960 4161 1-2 -.256 0.800 2 SOR 20 558 13300 3332 3546 2-3 0.367 0.716 3 TAP 10 230 10000 2858 2818 1-3 0.064 0.950 Years since land 1 SIN 10 0 29 11 10 1-2 -0.480 0.635 purchase 2 SOR 19 3 25 13 9 2-3 -1.061 0.298 3 TAP 10 2 30 17 11 1-3 -1.197 0.247 Natural forest (% of 1 SIN 10 19% 52% 37% 10% 1-2 2.560 0.016** farm area) 2 SOR 20 14% 50% 27% 11% 2-3 -2.471 0.020** 3 TAP 10 19% 75% 40% 19% 1-3 -0.413 0.685 Years since the most 1 SIN 7 2 13 5 4 1-2 -0.192 0.850 recent 2 SOR 14 2 17 6 4 2-3 -1.114 0.278 deforestation 3 TAP 9 2 14 8 5 1-3 -1.113 0.284 Area of most recent 1 SIN 7 4% 59% 18% 19% 1-2 -0.484 0.634 deforestation 2 SOR 14 4% 83% 22% 22% 2-3 0.448 0.659 (% of farm area) 3 TAP 9 4% 58% 18% 19% 1-3 -0.077 0.939 Reforestation (% of 1 SIN 10 0% 0.4% 0.05% 0.12% 1-2 -0.382 0.706 farm area) 2 SOR 20 0% 1.4% 0.09% 0.31% 2-3 -0.889 0.382 3 TAP 10 0% 1.7% 0.23% 0.55% 1-3 -1.006 0.338 Soybean cultivation 1 SIN 10 40% 80% 56% 12% 1-2 -2.306 0.029** (% of farm area) 2 SOR 20 27% 84% 68% 15% 2-3 2.612 0.014** 3 TAP 10 18% 80% 52% 20% 1-3 0.561 0.581 Other annual crops 1 SIN 10 0% 11% 2.3% 4.0% 1-2 1.459 0.174 (% of farm area) 2 SOR 20 0% 7% 0.35% 1.6% 2-3 -1.367 0.201 3 TAP 10 0% 11% 2.3% 4.3% 1-3 -0.001 0.999 Pasture (% of farm 1 SIN 10 0% 9% 1.7% 2.8% 1-2 1.583 0.125 area) 2 SOR 20 0% 3% 0.6% 0.91% 2-3 -1.449 0.181 3 TAP 10 0% 20% 3.6% 6.4% 1-3 -0.853 0.410 Head of cattle (per 1 SIN 6 0 90 37 33 1-2 -0.707 0.491 farm) 2 SOR 10 0 1100 137 340 2-3 -1.092 0.292 3 TAP 7 20 1000 338 420 1-3 -1.889 0.107 Distance from 1 SIN 8 8 100 35 32 1-2 -1.401 0.177 municipal seat 2 SOR 13 27 150 56 33 2-3 2.306 0.034** (Km) 3 TAP 6 3 40 22 16 1-3 0.887 0.392 **indicates t-value significant at P<0.05.

PAGE 62

62 Table 2-3. Multiple regression model for natural forest (% of farm area). Dependent variable Predictor B t-value P R2 Natural forest (% of Intercept 35.906 4.586 0.000 0.456** farm area) Farm size (FS) 0.001 2.209 0.039** Years on farm (YP) -0.400 -1.368 0.186 Km from municipal seat (KM) 0.136 1.587 0.128 Farms in Sinop (SI) 0.582 0.086 0.932 Farms in Sorriso (SO) -12.036 -1.837 0.081* Pre-financed farms (PF) -6.169 -1.211 0.240 *indicates t-value significant at P<0.1. **indicates t-value significant at P<0.05. Table 2-4. Descriptive statistics and compar ison of means of soybean planting practices (including both owned and leased land) be tween financial groups: farms that were pre-financed (PRE) and farms that were not. Table shows number of farms (N); minimum, maximum, mean, and standard de viation of values re ported; and t-test statistics for comparisons. N Min. Max. Mean Std. Dev. t-value P Soybean area (ha) PRE 20 170 6000 1865 1753 0.279 0.781 NOT 20 146 7300 1707 1841 % planted with GM PRE 16 0% 65% 8% 20% 1.130 0.274 soybeans NOT 18 0% 17% 2% 5% % planted with nematode PRE 20 0% 36% 6% 11% 0.148 0.883 resistant cultivar NOT 19 0% 40% 6% 11% Time between land purchase PRE 18 0 25 6 8 -1.106 0.276 and no-till system (yr) NOT 20 0 22 8 7 % of soybean area with corn PRE 20 0% 86% 33% 25% 0.309 0.759 cover crop NOT 20 0% 64% 31% 22% % of soybean area with PRE 20 0% 100% 44% 33% -2.105 0.042** millet cover crop NOT 20 23% 100% 64% 26% % of soybean area with PRE 20 0% 60% 8% 17% 0.598 0.554 other cover crops NOT 20 0% 38% 5% 11% % of soybean area with no PRE 20 0% 100% 15% 27% 2.423 0.025** cover crop (left fallow) NOT 20 0% 10% 0.5% 2.2% **indicates t-value significant at P<0.05.

PAGE 63

63 Table 2-5. Descriptive statistics and compar ison of means of soybean planting practices (including both owned and leased land) am ong municipal groups: fa rms located in the municipalities of Sinop (SIN), Sorriso (SOR) and Tapurah (TAP). Table shows number of farms (N); minimum, maximum, mean, and standard deviation of values reported; and t-test statistics for comparisons. Municipality N Min. Max. Mean Std. Dev. t-value P Soybean area (ha) 1 SIN 10 146 6000 1690 2064 1-2 -0.536 0.596 2 SOR 20 400 7300 2112 2016 2-3 1.843 0.078* 3 TAP 10 600 2000 1232 499 1-3 0.682 0.511 % planted with GM 1 SIN 7 0% 7% 1% 3% 1-2 -0.970 0.343 soybeans 2 SOR 17 0% 65% 9% 19% 2-3 1.079 0.291 3 TAP 10 0% 17% 2% 5% 1-3 -0.189 0.853 % planted with nema1 SIN 9 0% 19% 6% 9% 1-2 0.481 0.635 tode resistant 2 SOR 20 0% 36% 4% 11% 2-3 -1.449 0.159 cultivar 3 TAP 10 0% 40% 10% 13% 1-3 -0.878 0.392 Time between land 1 SIN 10 0 25 6 9 1-2 0.436 0.666 purchase and no2 SOR 19 0 16 5 5 2-3 -2.440 0.02** till system (yr) 3 TAP 9 2 23 12 8 1-3 -1.311 0.207 % of soybean area 1 SIN 10 0% 85% 26% 29% 1-2 -1.110 0.277 with corn cover 2 SOR 20 0% 86% 37% 21% 2-3 1.147 0.261 crop 3 TAP 10 0% 50% 28% 20% 1-3 -0.105 0.918 % of soybean area 1 SIN 10 0% 100% 53% 39% 1-2 0.224 0.825 with millet cover 2 SOR 20 0% 100% 50% 28% 2-3 -1.084 0.288 crop 3 TAP 10 25% 100% 62% 27% 1-3 -0.587 0.564 % of soybean area 1 SIN 10 0% 10% 1% 3% 1-2 -1.663 0.111 with other cover 2 SOR 20 0% 60% 9% 19% 2-3 0.279 0.782 crops 3 TAP 10 0% 28% 7% 11% 1-3 -1.493 0.165 % of soybean area 1 SIN 10 0% 100% 19% 34% 1-2 1.319 0.215 with no cover crop 2 SOR 20 0% 45% 4% 13% 2-3 0.184 0.831 (left fallow) 3 TAP 10 0% 25% 4% 8% 1-3 1.409 0.189 *indicates t-value significant at P<0.1. **indicates t-value significant at P<0.05.

PAGE 64

64 Figure 2-1. Relationship between fa rm size and percentage of natu ral forest on all sites (N=40). Figure 2-2. Relationship between fa rm size and percentage of natu ral forest on farms that were pre-financed by Amaggi Group (N=20; R2=0.0417 NS) and farms that were not (N=20). Figure 2-3. Relationship between farm size and percentage of na tural forest in Sinop (N=10; R2=0.1232 NS), Sorriso (N=20), and Tapurah (N=10; R2=0.2313 NS). R2 = 0.1226 0% 20% 40% 60% 80% 0200040006000800010000120001400016000 Total area (ha)Percentage of natural forest; P<0.05 0% 20% 40% 60% 80% 0200040006000800010000120001400016000 Total area (ha)Percentage of natural forest Pre-financed Not pre-financedR 2 = 0.2383; P<0.05 0% 20% 40% 60% 80% 0200040006000800010000120001400016000 Total area (ha)Percentage of natural forest Sinop Sorriso TapurahR 2 = 0.2386; P<0.05

PAGE 65

65 Figure 2-4. Relationship between ye ar of land purchase and the pe rcentage of natural forest on all sites (N=39). 0% 10% 20% 30% 40% 50% 60% 70% 80% 19751980198519901995200020052010 Year of land purchasePercentage of natural forest Pre-fiananced Not pre-financed Figure 2-5. Relationship between year of land purchase and per centage of natural forest on farms that were pre-financed by Amaggi Group (N=19; R2=0.1379 NS) and farms that were not (N=20; R2=0.0229 NS). Figure 2-6. Relationship between ye ar of land purchase and percenta ge of natural forest in Sinop (N=10), Sorriso (N=19) and Tapurah (N=10; R2=0.007 NS). 0% 20% 40% 60% 80% 19751980198519901995200020052010 Year of land purchasePercentage of natural forest Sinop Sorriso TapurahR 2 = 0.4658; P<0.05 R 2 = 0.2432; P<0.05 0% 20% 40% 60% 80% 19751980198519901995200020052010 Year of land purchasePercentage of natural forestR 2 = 0.075 (NS)

PAGE 66

66 Figure 2-7. Relationship between percentage area of the most recent deforestation and last year of deforestation on all sites (N=30) 0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 1988199019921994199619982000200220042006 Last year of deforestationPercentage area of the most recent deforestation Pre-financed Not pre-financed Figure 2-8. Relationship between percentage area of the most recent deforestation and last year of deforestation on farms that were pr e-financed by Amaggi Group (N=12; R2=0.139 NS) and farms that were not (N=18; R2=0.0002 NS). 0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 1988199019921994199619982000200220042006 Last year of deforestationPercentage area of the most recent deforestation Sinop Sorriso Tapurah Figure 2-9. Relationship between percentage area of the most recent deforestation and last year of deforestation in Sinop (N=7; R2=0.0082 NS), Sorriso (N=14; R2=3E-06 NS) and Tapurah (N=9; R2=0.2567 NS). 0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 1988199019921994199619982000200220042006 Last year of deforestationPercentage area of the most recent deforestationR 2 = 0.0194 (NS)

PAGE 67

67 0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100% CornMilletSorghumBrachiariaSunflowerRiceCottonFinger milletPercentage of farmers adopting each type of cover crop Pre-financed Not pre-financed Figure 2-10. Adoption of cover crops for financial groups in 2006. Figure 2-11. Weighted mean of cover crops fo r no-tillage system for financial groups in 2006. 0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100% CornMilletSorghumBrachiariaSunflowerRiceCottonFinger milletPercentage of farmers adopting each type of cover crop Sinop Sorriso Tapurah Figure 2-12. Adoption of cover crops for municipal groups in 2006. Pre-financed 17% 11% 41% 31% Not pre-financed 32% 60% 8% 0% Corn Millet Others Fallow

PAGE 68

68 Figure 2-13. Weighted mean of cover crops fo r no tillage system for municipal groups in 2006 0% 20% 40% 60% 80% 100% Pre-financed Not prefinanced Sinop Sorriso TapurahPercentage of farmers considering each type of pest as the worst insect pathogenic nematode Figure 2-14. Farmers perspective about the wo rst soybean pest in 2006. Comparisons between financial groups and among municipal groups. Sinop 29% 28% 38% 5% Sorriso 12% 3% 33% 52% Corn Millet Others Fallow Tapurah2% 8% 31% 59%

PAGE 69

69 0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100% Pre-financed Not prefinanced SinopSorrisoTapurahPercent of farms with pathogenic diseases Asian rust anthracnose Rhizoctonia foliar blight Figure 2-15. Farmers reporting pathogenic diseas e in their soybean crops in 2006. Comparisons between financial groups a nd among municipal groups. 0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100% Pre-financed Not prefinanced Sinop Sorriso TapurahPercent of farms with different types of insect infestation whiteflies stink bug caterpillar other Figure 2-16. Farmers reporting insect infesta tion in their soybean crop in 2006. Comparisons between financial groups a nd among municipal groups.

PAGE 70

70 0% 20% 40% 60% 80% 100% Pre-financed Not prefinanced SinopSorrisoTapurahPercent of farms with nematode Yes No Figure 2-17. Farmers reporting nematodes in their soybean crops in 2006. Comparisons between financial groups a nd among municipal groups.

PAGE 71

71 CHAPTER 3 PESTICIDES AND THE ENVIRONMENT Introduction The purpose of this chapter is to examine pe sticide use from seed treatment to soybean harvest among soybean producers in Sinop, Sorriso, and Tapurah, all located in northern Mato Grosso. Consistent with the purpose of this chapter, and based on the Amaggi Groups recommendations to their pre-financed farmers regarding pesticide use, the hypothesis that prefinanced farmers use fewer types of pesticide th an those who are not will be tested. Some of Amaggi Groups recommendations are: producers s hould seek specialized technical support, use inoculations in seed treatment, buy agrochemical s according to the agronomists prescription and follow the technical advice, implement integrated pest management, adopt biological control or insect growth regulator insecticid es when possible, and vary the us e of pesticides to avoid insect and weed resistance to the act ive chemicals in particular pesticides (Grupo Amaggi, 2006). While conducting field work, it was observed that farmers whose property was surrounded by natural forest or in close proximity to forest we re reporting that fewer ty pes of pesticides were used. Therefore, based on the expectation that farms closer to the Amazon forest require fewer pesticide applications, it is hypothesized that the producers located in Sinop, in the Amazon forest biome, use fewer types of pesticides th an the producers located in Sorriso and Tapurah. For each stage of soybean pesticid e application, it is expected that fewer Sinop farmers use the corresponding pesticide than farmers in the other m unicipalities. For fungicide use specifically, it is expected that Sinop farmers apply fungicide fewer number of times. Therefore, the hypothesis that farms with a higher percentage of forest c over use fewer types of pesticide will also be tested.

PAGE 72

72 To accomplish this task, background inform ation on pesticide use and reasons for application will be discussed; specific chemical pesticides used by the 40 interviewed farmers in Sinop, Sorriso, and Tapurah regions will also be summarized. Then, descriptive analyses of the percentage of farmers using specific active ingredients of pesticide products will be presented, as well as statistical analyses of fungicide appli cation and desiccants app lied to soybeans. Both descriptions and statistical analyses were comp ared between the studied farms with regard to their financial status, location, and pe rcentage of natural forest on farm. Pesticide Use Soybean pests identified in Chapter 2 are res ponsible for disease and consequent economic losses; therefore, the use of pesticides is ofte n essential to maintain soybean productivity in mediumto large-scale monoculture plantations. Th e soybean pesticides used for seed treatment, and the herbicides, insecticides and fungicides used during cu ltivation are presented below. Seed Treatment Most of the pathogenic diseases are dissemina ted by infected seeds, including anthracnose ( Colletotrichum dematium var. truncata ), Phomopsis spp., purple seed stain ( Cercospora kikuchii ), Cercospora sojina brown spot ( Septoria glycines ) and Diaporthe phaseolorum f.sp meridionalis The use of fungicides on seeds, besides a voiding the spread of pathogenic diseases, protects the seeds from fungi present in the soil, which can cause damage to seedlings (EMBRAPA, 2004). The seed treatment with insecticides can be done together with fungici des. The insecticides protect the seed in the soil until its germination, from pests th at are detrimental to plant health (Fundao Mato Grosso, 2006). The fungicides and in secticides used for seed treatment by the interviewed farmers are displayed in Table 3-1.

PAGE 73

73 During the process of applying seed treatments, applications of micr o-nutrients such as molybdenum and cobalt, or the addition of inocul ants to enhance nodulation, may be included as well. For the latter purpose, seeds may be either sprayed with inoculants or mixed with a peatbased material containing Bradyrhizobium which enhances the plants ability to fix nitrogen. To optimize effectiveness, fungicides and insecticides are applied be fore the micro-nutrients, while seed inoculation is conducted la st (Fundao Mato Grosso, 2006). Herbicide Use Weeds can interfere in different ways duri ng soybean production. They can cause a serious decrease in soybean yield, re duce the quality of the soybean grain, cause difficulty during harvest, and can harbor diseases or pests that may spread to adjacent crops (Gazziero, 2006). Weeds are a pervasive challenge in soybean plan tations and require high levels of herbicide application, especially when thes e areas are managed under a zero ti llage regime as mentioned in Chapter 1. Herbicides can be soil-applied or foliageapplied, and are app lied pre-planting, preemergence, and post-emergence to manage weed s. Nonselective herbicid es or desiccants are applied before planting soybean to kill cover cr ops and any weeds present. Desiccants are also applied prior to soybean harvest to kill and dry the crop and weeds, to improve efficiency during mechanical harvest (Powers and McSorley, 2000). The herbicides used by the producers interviewed are displayed in Table 3-2. Insecticide and Fungicide Use It is estimated that up to 15% of the worlds food production is lost to insect pests each year (Fundao Mato Grosso, 2006). Therefore, in soybean plantations as with most other intensive crop cultivation, chemi cal and biological methods are extensively used to manage

PAGE 74

74 insects. Biological insecticides are an alternativ e to the use of chemical insecticides; however, chemical insecticides provide the most rapid me thod for responding to an emergency situation. It is also estimated that up to 25% of the worlds agricultural produc tion is lost due to fungi, viruses, and bacteria (Fundao Mato Gr osso, 2006). Fungicides are used to prevent the development of plant disease unde r favorable environmental conditi ons, rather than to cure or reduce active epidemics (Powers and McSorley, 2000) Therefore, some fungicides may be used on a routine preventive basis for this purpose. In secticides may be used to manage the insect vectors that carry plant viruses. The insecticid es and fungicides used by the interviewed soybean farmers are shown in Tables 3-3 and 3-4, respectively. Methods Data Analyses Specific information on pesticide use obtaine d from 40 grower interviews and surveys conducted during June and July 2006 in northern Ma to Grosso were compared in four different ways. The group comparisons and the number of farms in each sub-group are structured as follows: Financial groups: between 20 farms pre-financed, a nd 20 farms not pre-financed by Amaggi Group; Municipal groups: among 10 farms located in Sinop, 20 farms in Sorriso, and 10 farms in Tapurah; Forest-1 groups: between 20 farms with less than 30% of natural forest on farm, and 20 farms with greater than or equal to 30% of natural forest on farm; and Forest-2 groups: between 10 farms with less than or equal to 20% of natural forest, and 10 farms with more than 40% of natural fore st, excluding the 20 farms that follow in the middle. In order to test the hypothesis that farms surr ounded by natural forest or in close proximity to forest use fewer types of pesticides, forest -1 groups and forest-2 groups were added to the

PAGE 75

75 analyses. Although no data were collected regardi ng neighboring natural fore sts for these farms, the percentage of natural forest on farm reported by the farmers was used as a proxy. Descriptive Analyses Descriptive analyses of the percent of fa rmers using herbicides, insecticides, and fungicides were conducted with in the four group comparisons previously defined. Group comparisons for pesticide use (by common pesticid e name) are displayed in Figures 1 to 14 for each stage of pesticide application: seed tr eatment; herbicide use as pre-planting and preharvesting desiccants, and post-emergence; and po st-planting insecticide and fungicide use. As shown in Tables 1 to 4, farmers use a variety of pesticides; therefore more than one pesticide common name can by used by a farm. Some pesticide uses were grouped according to their class or group in the charts. The fungicides were displayed in the figures by chemical groups. The fomosafen and lactofen (common names) post-emergent herbicides for broadleaf weeds are from the same nitrophenyl ether herbicide group. The haloxyfop and fluazifopp-butyl post-emergent herbicides for narrowleaf weeds are from the same aryloxyphenoxypropioni c herbicide class; although the name of the chemical haloxyfop was kept in the graphs for better visualization. The insecticides diflubenzuron, triflumuron, and novaluron are ins ect growth regulators, and were grouped as chitin synthesis inhibitors. The cypermethrin, lambda-cyhalothrin, and pe rmethrin insecticides are pyrethroid ester insecticides and were grouped as pyrethroids. Statistical Analyses Comparisons of independent sample means we re conducted to test for differences in timing of fungicide applications and percentage of soybean area defoliated at a 95% confidential interval between financial groups, municipal gr oups, forest-1 groups, or forest-2 groups. The statistical analysis was conducted as follows: first, a pre-test (F-test) concerning two-population

PAGE 76

76 variances at the 5% level of signi ficance was carried out to test fo r equality of variances between groups (Ott and Longnecker, 2004). If the varian ces were equal, a tw o-tailed t-test for comparison of means from samples with equal va riances was conducted; if the F-test revealed inequality between variances, a two-tailed t-test for unequal variances was conducted. Both types of tests (equal or unequal variances) were carried out at 95% confidence intervals. Results Seed Treatment Data on financial groups reveal that 75% of pre-financed farm ers use insecticides for seed treatment, 95% use fungicide, 80% inoculate seeds, and 5% use micronutrients for seed treatment. On the other hand, 50% of farmers not pre-financed use insecticide, 95% use fungicides, and 55% of farmers i noculate seeds. Data for the sa me variables between municipal groups demonstrate that in Sinop, 60% of the farm ers use insecticide and 90% use fungicide and inoculate seeds. In Sorriso, 50% of farmers use insecticide and inoculate seeds, 95% use fungicide, and 5% use micronutrients for seed tr eatment. All farmers in Tapurah use fungicide, 90% use insecticide, and 80% inoculate seeds (Figure 3-1). Data on forest-1 groups reveal that 95% of farmers from both groups reported using fungicide for seed treatment. For farmers with less than 30% of natural forest on farm, 65% use insecticides, 65% inoculate seeds, and 5% use mi cronutrients for seed treatment. Among farmers with greater than or equal to 30% of natural forest on farm 60% use insecticide, and 70% inoculate seeds. Data for forest-2 groups reveal that for farmers with less th an or equal to 20% of natural forest on farm, 50% use insecticide, 90% use fungicide, 10% use micronutrients for seed treatment, and 60% inoculate seeds. For farmers with more than 40% of natural forest on farm, all of them use fungicide, and 70% use inse cticide and inoculate seeds (Figure 3-2).

PAGE 77

77 Fungicide use for seed treatment Although not all farmers reported which fungici de they used, data for financial groups reveal that 40% of pre-financed farmers use the combination carboxin + thiram, 15% use carbendazim + thiram, 25% use fluodioxonil + metalaxyl-M, and 5% do not use fungicide for seed treatment. Some 45% of farmers not pre-fi nanced use the combinations carboxin + thiram and carbendazim + thiram, 10% use fluodioxonil + metalaxyl-M, and 5% do not use fungicide for seed treatment (Figure 3-3). Data based on municipal groups reveal th at in Sinop, 70% of the farmers use the combination carboxin + thiram, 10% use fluodi oxonil + metalaxyl-M, and 10% do not use fungicide for seed treatment. In Sorriso, 35% of farmers use the mixture carboxin + thiram, 40% use carbendazim + thiram, 25% use fluodioxonil + metalaxyl-M, and 5% do not use fungicide for seed treatment. Finally, in Tapurah, 30% of farmers use the combination carboxin + thiram, 40% use carbendazim + thiram, and 10% use fl uodioxonil + metalaxyl-M for seed treatment. Data from forest-1 groups show that for farm ers with less than 30% of natural forest on farm, 45% use the combination carboxin + thiram, 35% use carbendazim + thiram, 15% use fluodioxonil + metalaxyl-M, and 10% do not use fungici de for seed treatment. For farmers with greater than or equal to 30% of natural forest on farm, 40% use carboxin + thiram, 25% use carbendazim + thiram, and 20% use fluodioxonil + me talaxyl-M for seed treatment (Figure 3-4). Data from forest-2 groups reveal that for farm ers with less than or e qual to 20% of natural forest on farm, 20% use the combination car boxin + thiram, 20% use fluodioxonil + metalaxylM, 50% use carbendazim + thiram, and 10% do not us e fungicide for seed treatment. For farmers with more than 40% of natural forest on farm 30% use carboxin + thiram, 30% use fluodioxonil + metalaxyl-M, and 20% use carbendazim + thiram for seed treatment.

PAGE 78

78 Herbicide Use Desiccants before soybean planting All farmers use the chemical glyphosate before soybean planting. Data for financial groups demonstrate that 25% of farmers from both groups mix the chemical 2,4-D with glyphosate. Data based on municipal groups reveal that 40% of farmers in Tapurah, 30% of farmers in Sinop, and 15% of farmers in Sorriso mix gl yphosate with 2,4-D (Figure 3-5). Data for forest-1 and forest-2 groups show that 20% of farmers with less than 30% of natural forest on farm and that 10% of farmers with less than or equal to 20% of natural forest on farm mix 2,4-D with glyphosate, and 30% of farmers with greater than or equal to 30% of natural forest on farm, and another 30% of farmers with more than 40% of natural forest on farm use 2,4-D with glyphosate before soybean planting (Figure 3-6). Post-emergent herbicides Data based on financial groups demonstrate th at 55% of pre-financed farmers use the chemical imazethapir, 75% use chlorimuron ethyl, 85% use nitrophenyl ether, 85% use haloxyfop, 35% use other post-emergent chemicals, and 5% reported that they do not use postemergent herbicides. While 74% of not pre-financed farmers use imazethapir, 79% use chlorimuron ethyl, 79% use haloxyfop, 68% use nitrophenyl ether a nd 21% use other postemergent herbicides (Figure 3-7). According to data based on municipal groups, 50% of farmers in Sinop use the chemical imazethapir, 90% use chlorimuron ethyl, 90% use nitrophenyl ether, 80% use haloxyfop, and 30% use others post-emergent herbicides. In So rriso, 58% of farmers use imazethapir, 63% use chlorimuron ethyl, 68% use nitr ophenyl ether, 79% use haloxyf op, 26% use other chemicals and 5% do not use post-emergent herbicides. In Tapurah, 90% use imazethapir, chlorimuron ethyl and haloxyfop, 80% use nitrophe nyl ether and 30% use other post-emergent herbicides.

PAGE 79

79 Data from forest-1 groups indi cate that for farmers with less than 30% of natural forest on farm, 74% use imazethapir, chlorimuron ethyl a nd haloxyfop; 68% use n itrophenyl ether; and 32% use other post-emergent herbicides. For farmers with greater than or equal to 30% of natural forest on farm, 55% use imazethapir, 80% use chlorimuron ethyl, 85% use haloxyfop, 90% use nitrophenyl ether, 25% use ot her chemicals, and 5% of fa rmers do not use post-emergent herbicides (Figure 3-8). Data for forest-2 groups show that for farmer s with less than or equal to 20% of natural forest on farm, 67% use imazetaphir, 56% use chlo rimuron ethyl and nitrophenyl ether, and 44% use haloxyfop and other post-emergent herbicides. For farmers with more than 40% of natural forest on farm, 70% use imazethapir, 80% use ch lorimuron ethyl, 90% use nitrophenyl ether and haloxyfop, 20% use other chemicals, and 10% of farmers do not use post-emergent herbicides. Desiccants prior to soybean harvest Data from the two financial groups reveal th at 80% of pre-financed farmers use diquat before soybean harvest, 50% use paraquat, and 10% use glyphosate. In comparison, 85% of not pre-financed farmers use diquat, 35% use pa raquat, 5% glyphosate, and 10% do not use desiccants before soybean harvest. Data separa ted among municipal groups indicate that 70% of farmers from Sinop use diquat, 60% use para quat, 10% use glyphosate, and 10% do not use desiccants before soybean harvest. In Sorris o, 85% of farmers use di quat, 30% use paraquat, 10% use glyphosate, and 5% do not use desiccants before soybean harvest. In Tapurah, 90% of farmers use diquat and 50% use paraquat before soybean harvest (Figure 3-9). When data are analyzed for forest-1 groups, of the farmers with less than 30% of natural forest on farm, 95% use diquat, 30% use para quat, 5% use glyphosate, and 5% do not use desiccants before soybean harvest. For farmers with greater than or e qual to 30% of natural forest on farm, 70% use diquat, 20% use para quat, 10% use glyphosate, and 5% do not use

PAGE 80

80 desiccants before soybean harvest. Data from fo rest-2 groups reveal that all farmers with less than or equal to 20% of natural forest on farm use diquat, and 30% use paraquat. For farmers with more than 40% of natural forest on fa rm, 70% use diquat and paraquat, and 20% use glyphosate before soybean harvest (Figure 3-10). Comparisons of means with regards to the percentage of soybean area in which the reported desiccants were applied (Table 3-5) reve al that there are no st atistically significant differences between the financia l groups. However, there is a significant difference (P<0.01) between the farms in Sorriso and Tapurah. Farm ers in Tapurah sprayed desiccant herbicides prior to soybean harvest in 82% of their s oybean area while farmers in Sorriso sprayed desiccants in 50% of their soybean area. Insecticides and Biological Control Data from financial groups demonstrate that 25% of pre-financed farmers manage insects with biological control, 70% with chitin synt hesis inhibitors, 75% use methamidophos, 40% use endosulfan, 50% use pyrethroids, and 25% use othe r insecticides. Twenty percent of not prefinanced farmers use biological control, 70% use chitin synthesis inhibitors, 90% use methamidophos, 25% use endosulfan, 85% use py rethroids, and 10% use other insecticides (Figure 3-11). Data from municipal groups i ndicate that in Sinop, 20% of farmers manage insects with biological control, 70% with chitin synthesis inhibitors, 70% use methamidophos, 40% use endosulfan, 60% use pyrethroids, and 30% use other insecticides. In Sorriso, 25% of farmers use biological control, 65% use ch itin synthesis inhibitors, 80% use methamidophos, 20% use endosulfan, 70% use pyrethroids, and 15% use other insecticides. In Tapurah, 20% use biological control, 80% use ch itin synthesis inhibitors, 100% use methamidophos, 50% use endosulfan, 70% use pyrethroids, an d 10% use other insecticides.

PAGE 81

81 Results from forest-1 groups reveal that for fa rmers with less than 30% of natural forest on farm, 30% manage insects with biological control, 70% with chitin synthesis inhibitors, 90% use methamidophos, 30% use endosulfan, 70% use pyre throids, and 15% use other insecticides. In the case of farmers with greater than or equal to 30% of natura l forest on farm, 15% of farmers use biological control, 70% use chitin synt hesis inhibitors, 75% use methamidophos, 35% use endosulfan, 65% use pyrethroids, and 20% use other insecticides (Figure 3-12). Data from forest-2 groups show that for farm ers with less than or e qual to 20% of natural forest on farm, 40% manage insects with biological control, 60% with chitin synthesis inhibitors, 80% use methamidophos, 50% use pyrethroids, a nd 10% use endosulfan and other insecticides. Considering farmers with more than 40% of natura l forests on farm, 10% use biological control, 60% use chitin synthesis inhibitors, 90% use methamidophos, 70% use pyrethroids, and 20% use endosulfan and other insecticides. Fungicide Use All farmers from all groups use triazole fungici des. Data from financ ial groups reveal that all pre-financed farmers use strobilurin fungi cides and 20% use benzamidazole fungicides. Similarly, 90% of not pre-financed farmers us e strobilurin, and 25% use benzamidazole. Data from municipal groups show that all farmers fr om Sorriso and Tapurah and 80% of the farmers in Sinop use strobilurin fungicides. The benzimid azole fungicides are used by 15% of farmers in Sorriso, 20% in Sinop, and 40% in Tapurah (Figure 3-13). Data from forest-1 groups and forest-2 groups (Figure 3-14) indicate that the strobilurin fungicides are used by all farmers with less percen tages of natural forest on farm, and by 90% of farmer with more percentages of natural fo rest on farm in both groups. The benzimidazole fungicides are used by 20% of farmers with less than 30% of natural forest on farms, and by 25% of farmers with greater than or equal to 30% of natural forest on farm in forest-1 groups, and by

PAGE 82

82 30% of farmers with less than or equal to 20% of natural fore st on farm, and by 40% of farmers with more than 40% of natural fo rest on farm in forest-2 groups. Statistical comparisons of mean s with regards to the number of fungicide applications in the soybean plantation (Table 3-5) do not reveal any significant differenc e between groups in any of the four group comparisons: financial groups, municipal groups, forest-1 groups, and forest-2 groups. Discussion Seed Treatment Although 25% more pre-financed farmers reported using insecticide for seed treatment, the product used by both groups was Standak insecticid e. Standak contains the active ingredient fipronil and is consider the leadi ng insecticide for seed treatment in Brazil. For protection against seed-borne and soil-borne fungi which cause deca y, damping-off, and seed ling blight, most of the farmers used fungicide for seed treatment. Inoculation with Bradyrhizobium bacteria provides the soybean plant with nitrogen (N). Symbiotic N2 fixation is the main source of N for soybean plants (EMBRAPA, 2004). Inoculation usually increases crop yield, the %N in plant tissues, and post-harvest levels of N in the soil (Powers and McSorley, 2000). The Am aggi Group recommends the use of inoculation for seed treatment, however it did not have a si gnificant impact in the result between financial groups. With regard to municipal groups, farmer s reported using the combination carboxin + thiram. According to the Vitavax-Thiram product la bel, the fungicides that combine the systemic action of carboxin with the surface action of thiram control various seed and seedling diseases. It is particularly effective against foliar blight (Rhizoctonia so lani) and anthracnose (Colletotrichum truncatum). The effectiveness of this fungicide may explain why farmers in

PAGE 83

83 Sinop did not report any foliar blight diseas e during the 2005/2006 soybean harvest period as mentioned in Chapter 2. Independent of the percentage of natural forest on farm, almost all farmers reported using fungicide for seed treatment. According to the Vitavax-Thiram label, the use of fungicide for seed treatment often results in increased and more uniform stands of seedlings with a higher yield potential. Therefore, the percentages of natural forest on farm does not appear to reduce pesticide use for seed treatment, at least in terms of types of fungi cide and insecticide used, since amounts of pesticide applic ation in seeds were not statistically analyzed. Herbicide Use Desiccants before soybean planting The desiccants currently available in the market are glyphosate (the brand name RoundUp in particular), paraquat, and paraquat + diuron. These products may be applied alone or mixed with 2,4-D, particularly in areas with high density of broadleaf weeds. Some weeds such as Ipomoea sp., Conyza bonariensis and C. canad ensis, Richardia brasiliensis and Commelina benghalensis are tolerant to gl yphosate and require th e addition of associated herbicides for effective weed control (Vargas et al., 2006). Farmers that practice the no-tillage system need to rely on herbicide use before planting soybeans to create the dried cover crop resi dues on the soil surface. The surface residues suppress weeds, moderate soil temperature and mo isture, and can have allelopathic effects on weeds (Vargas et al., 2006). As mentioned in Ch apter 2, all the farmers surveyed employed the no-tillage system, and consequently relied on th e herbicide glyphosate to prepare the field for soybean plantation. Results also demonstrated that there was no significant differen ce in the percent of farmers mixing the herbicide 2,4-D with glyphosate betw een the financial groups, among the municipal

PAGE 84

84 groups, and within forest-1 and forest-2 groups. Therefore, the hypotheses related to desiccants are rejected: there is no difference in desiccant use before soybean plantin g between the financial groups; among farm locations, or between farms with different amounts of forest cover. Post-emergent herbicides Results did not show significant differences in the percentage of fa rmers using specific post-emergent herbicides between the financia l groups, among the municipal groups and within the forest-1 and forest-2 groups. For example, over 50% of farmers, regardless of financial status, location or different amount of forest cover, applied post-emergent herbicides such as imazethapir, chlorimuron ethyl, nitrophenyl ether and haloxyfop, and only 5% of pre-financed farmers from Sorriso in areas with higher percen tage of forest cover do not use post-emergent herbicide. The only farmer that reported not us ing post-emergent herbicides explained that his area was new and it did not need herbicide use. The amounts that each farmer used was not analyzed. The reason why the interviewed farmers use so many different post-emergent herbicides is that most farmers do not plant genetically m odified soybeans (GM). Genetically modified herbicide-resistant R oundup Ready soybean plants (RR soybeans) require only one general purpose herbicide such as glyphosate. The pe rcentage of farmers using glyphosate for RR soybeans in the study region is ve ry small and is included with other post-emergent herbicides. To avoid weed resistance to glyphosate, it is recommended to rotate GM soybeans with conventional soybean and/or rotate the herb icides active ingredients (Gazziero, 2006). The constant use of the same herbicide cla ss can cause resistance problems. According to Gazziero (2006), repeated use of he rbicides that present the same mode of action such as plant hormones, photosynthesis inhibitors cellular division inhi bitors, or specific enzyme inhibitors, are partially responsible for newl y resistant weed species An invasive plant can have resistance

PAGE 85

85 to an herbicide product line, in which mode of ac tion is similar; or the plant can have multiple resistances to differe nt modes of action. The results indicate that, contrary to the hypothe ses, post-emergent herbicides are not used by fewer pre-financed farmers, nor are they us ed by fewer farmers in Sinop, since no significant differences were observed between financial gr oups, and among municipali ties. Moreover, based on the observations, areas surrounded by forests do not seem to use fewer kinds of post-emergent herbicides. Desiccants prior to soybean harvest Based on the descriptive statisti cs, there is no significant diff erence in the percentage of farmers using the desiccants di quat, paraquat and gl yphosate prior to soybean harvest between financial groups, among municipa l groups and within forest groups. The desiccants diquat and paraquat are nonselective herbicides, for which the active ingredients only affect the green parts of plants sprayed, destroying the energy produci ng cells (chloroplasts) a nd rapidly desiccating the tissue. According to the Brazilian agricultu re research and extension agency (EMBRAPA, 2004), paraquat should be used in areas where narrow-leaf weeds such as Cardiospermum halicacabum are predominant, while the product diquat should be used in areas where broadleaf weeds such as Ipomoea grandifolia predominate. With regard to applications of desiccants prior to soybean harvest, statistical analyses revealed that farmers in Tapurah applied desiccants in a greater percentage of soybean area than the farmers located in Sorriso (P<0.01). This sugg ests that soybean farmers in Tapurah had more weed infestation in their crop or that the soybean plants were not dried enough to be harvested due to environmental conditions such as rainfall and moisture. In addition, as reported in Chapter 2, farmers in Tapurah had stinkbug infestation wh ich causes foliar retenti on at the end of the

PAGE 86

86 soybean plant cycle. This might have contribut ed to the need for more extensive desiccant application. Insecticide Use Although Amaggi Group recommen ds that their pre-financed farmers adopt biological control strategies or insect gr owth regulators when possible, the percent of farmers adopting biological control methods is still small and th ere is no difference in the level of adoption between financial groups. The level of adoption of insect growth re gulator insecticides is greater than the biological insecticides and equal betw een the financial groups. Therefore, the hypothesis that pre-financed farmers would use more biological and growth regulator insecticides than the not pre-financed farmers is rejected. Among the municipal groups, there are no large differences in percent of farmers adopting biological control and insect growth regulators. Biological controls, such as Baculovirus anticarsia and insect growth regulators, used to control foliar feeding insects, are more effec tive for small caterpillars (Degrande and Vivan, 2006). Some farmers reported that they did not use biological c ontrol methods because of their ineffectiveness against stinkbugs, an d that other chemicals applied to control other insects kill caterpillars as well. According to Gassen (2002), stinkbugs are controlled by insecticides such as monocrotophos, methamidophos, endosulfan, fen itrothion, and trichlor fon. He argues that pyrethroid insecticides are not efficient agai nst this type of pest. However, pyrethroid insecticides are generally recommended fo r some species of stinkbug. For example, cypermethrin, permethrin, and lambda-c yhalothrin are recommended for controlling Piezodorus guildinii Nezara viridula and Euchistus heros by Brazilian Minist ry of Agriculture In some regions of Brazil, some accounts report that Euschistus heros is resistant to insecticides such as endosulfan (Degrande and Vivan, 2006).

PAGE 87

87 During the interviews, some farmers repor ted using methamidophos and endosulfan to control stinkbugs and pyrethroid to control caterpillars at the end of the soybean cycle. However, based on the observations, there are no significant differences in percent of farmers using chemical insecticides between financ ial groups and among municipal groups. With regards to forest groups, it was expected that farms with a higher forest cover would use less insecticides. Accordi ng to Powers and McSorley (2 000), maintenance of vegetation heterogeneity and diversity with regard to cr ops and proximate natural vegetation within a region, is an effective integrated pest manageme nt strategy for some pests. Non-crop vegetation may serve as a reservoir for predators and parasite s of the pests; pests may also be attracted to the natural vegetation rather than the planted crop. However, results showed that farmers, regardless of different amounts of forest cover, applied biol ogical and growth regulator insecticides, as well as different chemical in secticides. Therefore, there are no significant differences in percent of farmers using insecticides with regard to percentage of forest cover on farm. Fungicide Use Results indicate that most of the farmers between financial groups and among municipal groups used strobilurin and triazole fungicides and fewer farmers used benzamidazole fungicide. The former fungicides are recommended for cont rolling pathogenic diseases such as Asian Soybean Rust (Phakopsora pachyrhizi). As show n in Chapter 2, all farmers reported having Asian Soybean Rust in their soybean crops. Fungicide application as a pr eventive treatment for Asian so ybean rust is recommended prior to the detection of symp toms of the disease. Triazole fungicides combined with a strobilurin or a benzamidozole s hould be applied. The objective is to protect the soybean against rust and other diseases that o ccur during the flowering stage, such as anthracnose, leaf spot

PAGE 88

88 (Corynespora cassiicola), foliar blight, and powdery mildew (Microsphaera diffusa). However, if disease symptoms are observed, then other fungici des that are effective in controlling the rust should be applied (Fundao Mato Grosso, 2006). Spores of Asian soybean rust can travel hundreds of miles on a windy day and infection can occur in favorable weather conditions such as high air moisture and with temperatures between 18C and 26C (Wyse, 2005). Bearing this in mind, and the fact that the disease started in the south of Brazil, one woul d expect that farms located in the northernmost regions would require less fungicide ap plications. However, statistical an alysis rejected the hypothesis that farms in Sinop apply fungicides fewer times. Another hypothesis was that farmers with more forest cover would use less fungicide. It was expected that the forest would function as a windbreak, helping prot ect the crop from the spores of Asian soybean rust. However, results di d not show significant diffe rences in percent of farmers using insecticide within each forest groups. Moreover, statistical an alyses with regard to the number of times that fungicide was applied did not reveal significan t differences within forest-1 and forest-2 groups. Therefore, the hypoth esis that farmers with more forest on farm apply fungicide fewer times is rejected. The foliar blight disease (Rhizo ctonia solani), also mentione d by farmers in Chapter 2, is efficiently controlled when adopting an inte grated pest management strategy. There are no effective fungicides registered by the Ministry of Agriculture a nd Agrarian Reform (MARA) to control foliar blight disease. Nonetheless, experiments have shown that fungicides such as azoxystrobin, metconazole, pyraclo strobin + epoxiconazole, and tr ifloxystrobin + cyproconazole can effectively control the dis ease (EMBRAPA, 2004) and as shown in Table 3-4, some of these pesticides were used by farmers to control pathogenic diseases.

PAGE 89

89 Conclusion To sum up, most explanatory va riables had no effect on pestic ide use patterns, with the exception of desiccant use between the farms in Ta purah and Sorriso. The similarities in the data reject the hypotheses that pre-fi nanced farms use fewer types of pesticide than not pre-financed farms, that the producers located in Sinop use le ss fungicide and apply it less often than the producers located in Sorriso and Tapurah, and that farms surrounded by natural forest or in close proximity to forest need less pesticide use. Ho wever, there are some interesting findings among the farmers. For example, all farmers, regardless of financial status, locat ion, or amount of forest cover, rely on the herbicide glyphosate before so ybean planting, since they adopted the no-tillage system for soil conservation. Also, farmers use different types of post-emergence herbicides, because most of their seeds are non-GM soybeans.

PAGE 90

90 Table 3-1. Pesticide used for seed treatment as reported by farmers in Sinop, Sorriso, and Tapurah. Common name Trade name Fungicide Carbedanzin + Thiram Derosal Plus / ProTreat Carboxin + Thiram Vitavax-Thiram PM / Vitavax-Thiram 200 SC Fluodioxonil + Metalaxyl M Maxin XL Insecticide Fipronil Standak 250 FS Table 3-2. Herbicide used in soybean planta tion as reported by farmers in Sinop, Sorriso, and Tapurah. Common name Trade name Desiccants before soybean plantating Glyphosate Glifosato* / Roundup* / Trop 2,4-D not reported Post-emergent Carfentrazone Aurora 400 CE Chlorimuron Ethyl Classic / Clorimuron Master Nortox Clethodim Select 240 EC Diclosulam Spider 840 WG Fenoxaprop-P-Ethyl + Clethodim Podium S Fluazifop-P-Butyl Fusilade 250 EW Fomosafen Flex Glyphosate Glifosato* / Roundup* Haloxyfop-Methyl Verdict R Imazethapir Pivot Lactofen Cobra / Naja Trifluralin Trifluralina* Desiccants before soybean harvest Diquat Reglone Glyphosate Glifosato* / Roundup* Paraquat Gramoxone 200 *Trade names with many variations and not specified by farmers.

PAGE 91

91 Table 3-3. Insecticide used in soybean planta tion as reported by farmers in Sinop, Sorriso, and Tapurah. Common name Trade name Cypermethrin Cipermetrina Nortox 250 CE Diflubenzuron Dimilin 250 WP Dimethoate not reported Endosulfan Thiodan* / Endosulfan* Lambda-Cyhalothrin Karate Zeon 250 CS Methamidophos Metamidofs Fersol 600 / Tamaron BR Methomyl Lannate BR Methyl Parathion Folidol CS Novaluron Gallaxy 100 CE Permethrin Permetrina / Pounce / Talcord / Valon Teflubenzuron Nomolt 150 SC Thiamethoxam + Lambda-cyalothrin Engee Pleno Thiodicarb Larvin 800 WG Triflumuron Certero 480 CS *Trade names with many variations and not specified by farmers. Table 3-4. Fungicide used in soybean planta tion as reported by farmers in Sinop, Sorriso, and Tapurah. Common name1 Trade name Azoxystrobin (S)+ Cyproconazole (T) Priori Xtra Carbendazim (B) Bendazol / Derosal Carbendazim (B) + Thiram (D) Derosal Plus Epoxiconazole (T) + Pyraclostrobin (S) Opera Flutriafol (T) Impact 125 SC Flutriafol (T) + Thiophanate-methyl (B) Impact Duo Myclobutanil (T) Systane CE Propiconazole (T) + Cyproconazole (T) Artea Tebuconazole (T) Folicur 200 CE / Orius 250 CE Thiophanate-Methyl (B) Cercobi n 500 SC / Cercobin 700 PM Trifloxystrobin (S) + Cyproconazole (T) Sphere 1. (T) = Triazole; (S) = Strobilurin; (B ) = Benzimidazole; (D) = Dithiocarbamate

PAGE 92

92 Table 3-5. Descriptive statistics and comparis on of means of pesticid e use. Comparisons between financial groups, am ong municipal groups, between farms with <30% forest and farms with 30% forest, and between farms with less 20% forest and farms with >40% forest. Table shows number of farms (N) in each group; minimum, maximum, mean, and standard deviation of values reported within each group; and t-test statistics for comparisons between groups. Comparisons N Min. Max. Mean Std. Dev. t-value P Desiccant use Pre-financed 14 10% 100% 58% 34% -1.037 0.308 before Not financed 18 0% 100% 70% 33% harvest (% 1-Sinop 6 0% 100% 73% 43% 1-2 1.434 0.167 of soybean 2-Sorriso 16 10% 100% 50% 31% 2-3 -2.989 0.006** area) 3-Tapurah 10 50% 100% 82% 19% 1-3 -0.482 0.646 <30% forest 17 10% 100% 65% 34% 0.186 0.854 30% forest 15 0% 100% 63% 34% 20% forest 8 14% 100% 59% 32% 1.188 0.253 >40% forest 9 35% 100% 76% 27% Number of Pre-financed 20 1 3 2.1 0.456 -0.080 0.936 times Not financed 20 2 3 2.2 0.319 fungicide 1 Sinop 10 2 3 2.2 0.334 1-2 0.856 0.399 was 2 Sorriso 20 1 3 2.1 0.353 2-3 -1.425 0.176 applied 3 Tapurah 10 2 3 2.3 0.483 1-3 -0.700 0.493 <30% forest 20 1 3 2.2 0.492. 0.977 0.337 30% forest 20 2 3 2.1 0.246 20% forest 10 1 3 2.2 0.576 0.433 0.670 >40% forest 10 2 3 2.1 0.316 **indicates t-value significant at P<0.01.

PAGE 93

93 0% 20% 40% 60% 80% 100% Pre-financed Not pre-financed Sinop Sorriso TapurahPercent of farmers using pesticide for seed treatment Insecticide Fungicide Inoculation Micro Nutrients Figure 3-1. Pesticide and inocul ant use for seed treatment for financial groups and municipal groups in 2006. 0% 20% 40% 60% 80% 100% F1 <30%F1 30%F2 20%F2 >40%Percentage of natural forest on farmPercent of farmers using pesticide for seed treatment Insecticide Fungicide Inoculation Micro Nutrients Figure 3-2. Pesticide and inoculan t use for seed treatment for fore st-1 groups (F1) and forest-2 groups (F2) in 2006. 0% 10% 20% 30% 40% 50% 60% 70% Pre-financed Not pre-financedSinopSorrisoTapurahPercent of farmers using fungicides for seed treatment Carboxin + Thiram Carbendazim + Thiram Fludioxonil + Metalaxyl-M none Figure 3-3. Fungicide use for s eed treatment for financial groups and municipal groups in 2006.

PAGE 94

94 0% 10% 20% 30% 40% 50% 60% F1 <30%F1 30%F2 20%F2 >40%Percentage of natural forest on farmPercent of farmers using fungicides for seed treatment Carboxin + Thiram Carbendazim + Thiram Fludioxonil + Metalaxyl-M none Figure 3-4. Fungicide use for seed treatment for fore st-1 groups (F1) and forest-2 groups (F2) in 2006. 0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100% Pre-financed Not prefinanced Sinop Sorriso TapurahPercent of farmers using desiccant herbicides before soybean plantation Gyphosate 2,4-D Figure 3-5. Desiccant use before soybean plantin g for financial groups and municipal groups in 2006. 0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100% F1 <30%F1 30%F2 20%F2 >40%Percentage of natural forest on farmPercent of farmers using desiccants befores soybean plantation Gyphosate 2,4-D Figure 3-6. Desiccant use before soybean planting for forest-1 groups (F1) and forest-2 groups (F2) in 2006.

PAGE 95

95 0% 20% 40% 60% 80% 100% Pre-financed Not pre-financedSinopSorrisoTapurahPercent of farmers using post-emergent herbicides Imazethapir Chlorimuron Ethyl Nitrophenyl Ether Haloxyfop others none Figure 3-7. Post-emergent herbicide use for financial groups and municipal groups in 2006. 0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100% F1 <30%F1 30%F2 20%F2 >40%Percentage of natural forest on farmPercent of farmers using post-emergent herbicides Imazethapir Chlorimuron Ethyl Nitrophenyl Ether Haloxyfop others none Figure 3-8. Post-emergent herbicide use for fore st-1 groups (F1) and forest-2 groups (F2) in 2006. 0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100% Pre-financed Not pre-financed Sinop Sorriso TapurahPercent of farmers using desiccants prior to soybean harvest Diquat Paraquat Glyphosate none Figure 3-9. Desiccant use prior to soybean harves t for financial groups and municipal groups in 2006.

PAGE 96

96 0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100% F1 <30%F1 30%F2 20%F2 >40%Percentage of natural forest on farmPercent of farmers using desiccants prior to soybean harvest Diquat Paraquat Glyphosate none Figure 3-10. Desiccant use prior to soybean harves t for forest-1 groups (F1) and forest-2 groups (F2) in 2006. 0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100% Pre-financed Not pre-financedSinopSorrisoTapurahPercent of farmers using insecticides Biological Control Chitin Synthesis Inhibitors Methamidophos Endosulfan Pyrethroid other s Figure 3-11. Insecticide use for financ ial groups and municipal groups in 2006. 0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100% F1 <30%F1 30%F2 20%F2 >40%Percentage of natural forest on farmPercent of farmers using insecticides Biological Control Chitin Synthesis Inhibitors Methamidophos Endosulfan Pyrethroid others Figure 3-12. Insecticide use for forest-1 gr oups (F1) and forest-2 groups (F2) in 2006.

PAGE 97

97 0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100% Pre-financed Not pre-financed Sinop Sorriso TapurahPercent of farmers using fungicides Strobilurin Triazole Benzimidazole Figure 3-13. Fungicide use for financ ial groups and municipal groups in 2006. 0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100% F1 <30%F1 30%F2 20%F2 >40%Percentage of natural forest on farmPercent of farmers using fungicides Strobilurin Triazole Benzimidazole Figure 3-14. Fungicide use for forest-1 gr oups (F1) and forest-2 groups (F2) in 2006.

PAGE 98

98 CHAPTER 4 SOYBEAN PROFITABILITY AND RISK MODELING Introduction The first objective of this chapter is to anal yze if there are differences in soybean yield among soybean producers in the study region. Congr uent with this objective, the following specific hypotheses will be tested : (1) The pre-financed group, who use management practices recommended by Amaggi Group, has greater yiel ds than those who are not; and (2) the producers located in Sinop, where soybeans have been cultivated for fewer years, have lower yields than those located in Sorriso and Tapura h. The second objective is to present a case study comparing net revenues for the 2005/2006 soybean production year of a smaller and a larger farm. Reasons for the differences in this variable are explored. Finally, a risk modeling exercise, using Pallisades @Risk for Microsoft Excel, is conducted to evaluate the sensi tivity of the net revenue of case study farms to fluctuations in th e soybean price and the exchange rate with the US dollar. To accomplish these tasks, first, the excha nge rate, world soybean price and soybean production costs will be discusse d. Then, statistical analyses of soybean yield in the 2005/2006 soybean harvest are undertaken and compared between the financial groups and among the municipal groups. Finally, net revenue per hectare fo r a larger and a smaller farm, both located in Sinop, are compared to discuss the difference in this variable and how it is affected by changes in exchange rate (Brazilian real related to the US dollar) and US dollar soybean price. Exchange Rate In early 1999, Brazil adopted a floating exchan ge rate causing the real to depreciate considerably from R$1.21/US$ to an aver age of R$1.52/US$ in January, R$1.91/US$ in February, and R$1.90/US$ in March (Marques, 2004) The devaluation of the Brazilian currency

PAGE 99

99 benefited exporters, while reducing the competitiveness of imports. Devaluation of the real related to the US dollar increases the producer pr ice of internationally traded commodities such as soybeans while the agricultural inputs measured in foreign currency became more expensive. The devaluation of the real rais ed expected returns to soybeans, which in turn led to a 20% expansion in the area planted to soybean s in the 2000/01 crop year (Valdes, 2006). Not only is the export price of soybean dete rmined in U.S. dollars, but most of the operating costs such as fertilizer, pesticides, and fuel are also in U.S. dollars. Therefore, these costs increase when there is a local currency deva luation or an exchange rate increase. However when the real appreciates, these costs do not decrease at the same rate as the exchange rate (J. Y. Shimada, personal communication, November 21, 2006). Since September 2004, the real started a new pe riod of appreciation (Figure 4-1), affecting Brazils competitive pricing and the profitability of its agricultural exports. By July 2006, the real had appreciated 32% against the U.S. dolla r, making Brazilian prod ucts about one-third more expensive in importing coun tries (Valdes, 2006), which also resulted in a lower soybean producer price for Brazilian farmers. Reduced export competitiveness resulting from a less favorable exchange rate has caused protests in Br azil, with attempts to block deliveries and force the price up (Baer, 2006). According to the Mato Grossos Alliance for Agriculture and Cattle Ranching (FAMATO) and Mato Grossos Association of Soybean Producers (APROSOJA), soybean producers operating costs are closely linke d to the exchange rate. In the 2006/2007 crop harvest, producers bought inputs at an exchange rate of R$2.30, and sold the production at an exchange rate of less than R$2.00. This is the third consecutive ha rvest period where producers buy inputs at a less

PAGE 100

100 favorable exchange rate than the exchange rate th ey receive when selling their product (Dirio de Cuiab, 2007). Soybean Price (US$) The soybean price is determined in the Chi cago Board of Trade (CBOT), and current and future soybean demands are established. The Braz ilian free-on-board (FOB) price is based on the CBOT price and the Rotterdam cost of insu rance and freight (CIF) price (Machado and Margarido, 2000). The differential be tween producer price and the FOB price is 20% less in the case of Mato Grossos producers. This differentia l is due to marketing margins of the trading companies such as port costs, transportation to port and taxation cost s (Roessing, 2005; Valdes, 2006). The history of international soybean pr ices has impacted soybean expansion and production in Brazil, as discu ssed in detail by Brum (2004) : During the 1970s, high world soybean prices combined with government incenti ves resulted in an incr ease in Brazils soybean production. In 1972, the soybean quotation in Chi cago reached US$ 10.00/bus hel (equivalent to 27.216 kilos), with direct impacts on the Brazilian producers soybean price. From 1970 to 1977, Brazilian soybean production increased from 1.5 million tons on an area of 1.3 million hectares to 12.5 million tons on an area of 8.3 million hectar es. In the last two years of the 1970s, soybean production decreased to 9.6 million tons due to weather-related losses. During the 1980s the high growth rates in soybean production were diminishing due to uncertainties and risks related to this activity. In Chicago, the av erage annual soybean price was varying between US$6.00 and US $7.25/bushel during the first half of the decade. In 1985 until 1987, the average annual soybean price was varying between US$5.00 and US$5.50/bushel. In 1988, the average monthly soybean prices incr eased again reaching US$9.00/bushel due to a

PAGE 101

101 soybean supply crisis in the United States. Ho wever, the average annua l soybean price was US$ 7.60 in 1988 and US$6.75/bushel in 1989 (Brum, 2004). In the 1990s, the average soybean price in Chicago was US$6.20/bushel. Even though the international soybean prices were not high from an historical perspec tive, Brazilian soybean production increased significantly during the de cade due to technological advances. Soybean production increased from 20.4 million tons in 1990 to 31.4 million tons and an area of 13 million hectares in 1999. The increase in area pl anted and higher production translated into a 35% growth in soybean export vo lume. This export expansion in Brazil led to changes in world prices, such as a 2 percent decline in world soybean prices in 2001 (Valdes, 2006). From 1999 to 2002, the average annual soybean price did not exceed US$5.00/bushel; and in 2003, the soybean price increased to US$6.34/bushel (Brum, 2004). In 2004 there was an increase in the soybean pr ices due to the bad harvest period in the United States and to production losses in Braz il. In the beginning of March 2004, the soybean price reached US$9.60/bushel, the highest pr ice since 1988. In 2005, the price in Chicago dropped 25% with regard to the previous year and producer prices in Mato Grosso decreased 28% in the same period (Dirio de Cuiab, 2007). Between 2006 and 2007, stimulated by the increased corn demand in the United States for ethanol production, the so ybean price increased 25% in a year, increasing the Br azilian exports despite the low exchange rate (Riveras, 2007). Production Costs According to EMBRAPA, around 20% of total s oybean production costs are for pesticides (Bickel and Dros, 2003). Fertiliz ers account for 30% or more of soybean cost of production (Goldsmith and Hirsch, 2006). The average transpor tation cost, when exporting soybeans, is 83% higher than in the United States, the largest soybe an producer, and 94% higher than in Argentina, the third largest soybean producer (Valdes, 2006). Fu el is another expensive input in Brazil; from

PAGE 102

102 2001 to 2007, the price of fuel increased 97%, while the price of diesel in creased 213% (Dirio de Cuiab, 2007). According to Roessing (2005) the soybean pr oduction cost for non-GM soybeans with two fungicide applications for As ian soybean rust was around R$1,303 per hectare in 2005 in central west and southeastern regions of Braz il. The FOB price was US$13.00 per 60kg bag (2.205 bushel) and the exchange rate was R$2.27 per U. S. dollar, equivalent to R$29.51 per bag. Since the producer price for farmers in Mato Grosso is 20% less due to marketing margins, the estimated producer price was approximately R$2 4.00 per bag. Roessing (2005) explains that for producers to break even, they would need to produce more than 54.31 bags (3.26 tons) per hectare in 2005, and as such, soybean farming wa s a very high risk endeavor. In the 2005/2006 harvest, the average soybean yield in Mato Gr osso was 44 bags (2.64 tons) per hectare (L. M. Ribeiro, personal communication, June 13, 2007). The president of FAMATO states that produc ers income has accumulated a drop of 46%; in two consecutive soybean harv est periods, the loss accumulated was more than R$2.07 billion (Dirio de Cuiab, 2007). The average production cost in the 2007/2008 harv est period is already 25% higher in Mato Grosso: the producer w ho paid around R$950.00 to plant one hectare of soybean, will spend R$1.187, an increase of R$237.5 compared with the previous harvest period. This is a reflection of the 50% increase in production costs, on average. Therefore, the producers debt and the high agri cultural production cost given cu rrent price and exchange rates obstruct the expansion of cultivated ar ea in Mato Grosso (A Gazeta, 2007). Since from the individual producers perspectiv e, little can be done with regard to the soybean price and exchange rates, the alternativ e is to reduce costs or increase yield through better technology. Some options are available su ch as acquiring specific seeds recommended for

PAGE 103

103 the site, amendment corrections to the soil based on laboratory analysis of soil samples, avoiding unnecessary pesticide applica tions, minimizing mechanical damages to the product during harvesting, and perhaps, planting GM soybeans de pending on weed infestation levels and seed royalties (Roessing, 2005). To reduce fuel costs, some farmers are producing their own biodiesel with oilseeds such as soybeans a nd sunflowers, as verified by th e researcher in Tapurah (Figure 4-2). Methods Statistical Analyses Comparisons of independent sample means we re conducted to test for differences in soybean yield at a 95% level of confidence between financial groups and among municipal groups. The statistical analysis was conducted as follows: first, a pre-test (F-test) concerning two-population variances was carried out to test for equality of variances between groups (Ott and Longnecker, 2004). If the varian ces were equal, a two-tailed ttest for comparison of means from samples with equal varian ces was conducted; If the F-test revealed inequality between variances, a two-tailed t-test for unequal variances was conduc ted, both at a 95% level of confidence. Net Revenue Analyses In order to determine the prof itability of soybean farms for one year on a smaller and a larger farm, total net revenue and net revenue per hectare is calculat ed by deducting soybean production costs from revenues for the 2005/2006 ha rvest. The soybean production costs for both farms were collected in June 2006 while conducting field work in the city of Sinop. The size of the larger farm is 14,000 hectares; 6,151 hectares are planted with soybeans and the farm is located 100 km from Sinop. The si ze of the smaller farm is 1,391 hectares; 650 hectares are planted with soybeans and it is located 22 km from Sinop.

PAGE 104

104 In comparing net revenue analysis for both fa rms, the costs consider ed are: fertilizers, pesticides, seed, machinery operation, freight, le ased land, salary, and machinery depreciation. Costs with salary and depreciation for the smalle r farm are incorporated with machinery costs since it is leased. Financing aspects such as bank loans were not considered since they were not revealed by the landowners. Income tax (5.5%) is discounted from the revenue because it is common to register the property as a family enterprise rather than a corporation. It is known that a part of the larger farm is leased and another part was bought in 1996; and that the smaller farm was bought in 2002. However, for the purpose of this analysis, the area planted to soybean is considered 40% leased, 60% owned. Land rental ra tes are paid in bags (60kg) of soybean. For land near Sinop, the rate is approximately five to six bags of soybean per hectare, depending how far the property is from the city, more precisely from the BR-163 (H. C. Ribeiro, personal communication, June 15, 2007). Hen ce, it is assumed that the smaller farm leases land at 6 bags per hectare si nce it is closer to the city, and the larger farm pays 5 bags per hectare since it is farther from the city. Risk Analyses Since soybean farmers are concerned with th e uncertainty surrounding future values for exchange rates and world prices of soybean, a risk analysis was incorporated in the net revenue analyses using the @Risk software application for Micr osoft Excel. The @Risk routine performs a Monte Carlo simula tion, which randomly generates values for uncertain variables according to a user-specified probability distribution and iteration limit. In the case of the present analysis, uncertain variables considered were the exchange rate and the world soybean price (Campbell and Brown, 2003). For the purposes of this analysis, it is assumed that a triangular distribution would represent a reasonable descripti on of the variables uncertaint y. A triangular distribution shows

PAGE 105

105 the range of possible values the variables uncer tainty could take and shows the probability of variables lying within any particular range of possible va lues. The function used is TRIGEN, which estimates the minimum, best guess a nd maximum values for each variable. This function estimates the bottom and top percentile va lues and makes the distribution inclusive of the maximum and minimum values (Palisade Corporation, 2006). The most likely values were based on the aver age exchange rate and soybean price in the study region during the year 2006. The maximum values and the minimum values for the exchange rate were based on the percentage that it varied over the last 3 years. The maximum and the minimum values for the soybean price were defined as the percentage that it varied in the year of 2006. Therefore, it is assumed that the ex change rate and the s oybean price could vary 10% around the values of R$2.18 and US$ 9.60 resp ectively. The bottom pe rcentile and the top percentile chosen for this analysis follow th e procedure in Campbell and Brown (2003), where the extreme values for exchange rates occur at th e 10% and 90% percentiles, and for share prices at the 5% and 95% percentiles. The RiskTrigen functions are shown as: Exchange rate (Figure 4-1) = RiskTrigen(1.96,2.18,2.4,10,90); Soybean price (US$) = RiskTrigen(8.64,9.6,10.56,5,95). The following three scenarios were modeled for both farms: (1) the impact of uncertainty in the exchange rate on net revenue per hectare while holding the world soybean price constant; (2) the impact of uncertainty in the exchange rate on the ne t revenue per hectare where the exchange rate affects pesticide and fertilizer costs, while holding the world soybean price constant; and (3) the impact of uncertainty in the soybean price on the net revenue per hectare while holding the exchange rate constant.

PAGE 106

106 Results Statistical Analyses Comparisons of means with regard to soybean yield (measured in 60kg bags per hectare) in the 2005/2006 soybean harvest (Table 4-1) rev eal that there are no si gnificant differences between the financial groups. However, there is a significant difference (P <0.05) between the farms in Sinop and Sorriso. Farmers in Sinop ha d an average soybean yield of 51.7 bags (3.1 tons) per hectare, while farmers in Sorriso had an average soybean yield of 56 bags (3.36 tons) per hectare. Net Revenue Analyses The analyses reveal that the larger fa rm had a net revenue of R$848,782 or R$138 per hectare. The production cost was R$930 per hect are of which 36.6% was for fertilizers, 20.9% for pesticides, 17.4% for machinery, 9.3% for salary and commission, 6.4% for soybean seed, 4.5% for leasing land, 3.8% for depreciation, an d 1.0% for freight. The farmers revenue was R$1,130 per hectare, since the yield was 54 bags ( 60 kg/bag) per hectare w ith an exchange rate of R$2.18 and soybean price of US$9.60 per bag, from which income tax (5.5%) was deducted and then production costs subtracted. The smaller farm had a negative net reve nue of R$-43,785 or R$-67 per hectare; a difference of R$205 compared with the larger farms net revenue. The production cost was R$1,036 per hectare (R$106 greater than the larg er farm) of which 35.7% was for pesticides, 34.1% for fertilizers, 14.8% for machinery, 6.1 % for seed, 4.9% for leasing land, and 4.5% for freight. The farmers revenue was R$1,025 per hect are, given the soybean yield of 49 bags per hectare with an exchange rate of R$2.18 and soybean price of US$ 9.60 per bag, from which income tax (5.5%) was deducted and then production costs subtracted.

PAGE 107

107 Risk Analyses Despite all the benefits and costs in produci ng soybean, there are some risks that farmers take when they enter this market. The soybean pr ice is fixed in US dollars, and when converted to the real, it is subject to fl uctuating exchange rates. The risk modeling incorporated in the net revenue analyses showed different results fo r the three different scenarios (Table 4-2). In the first scenario (Figur e 4-4), with the exchange ra te varying 10% around R$2.18 and holding the soybean price at US $9.60 per bag, there is a 90% chance of the net revenue per hectare for the larger farm falling between R$9.73 and R$266. The minimum value for net revenue in this scenario is R$-45.64 and the maximum value is R$321.78. For the smaller farm, there is a 90% chance of the net revenue per hectare falli ng between R$-182.20 and R$47.28; the minimum value is R$-231.79 and the maximum value is R$97.23. In the second scenario (Figure 4-5), with the exchange rate varying 10% around R$2.18 and with the exchange rate linked to the price of fertilizers and pestic ides, while keeping the soybean price fixed at US$9.60 per bag, there is a 90% chance of the net revenue per hectare for the larger farm falling between R$76.75 and R $199.22. The minimum value for net revenue in this scenario is R$50.53 and the maximum valu e is R$226.63. For the smaller farm, there is a 90% chance of the net revenue per hectare falling between R$ -91.76 and R$-42.92; the minimum value is R$-102.21 and the maximum value is R$-32.00. In the third scenario (Figure 4-6), with the soybean price varying 10% around US$ 9.60 per bag while keeping the exchange rate at R $2.18, there is a 90% chance of the net revenue per hectare for the larger farm falling between R$35.25 and R$240.59; the minimum value is RS$8.92 and the maximum value is RS$285.45. For the smaller farm, there is a 90% chance of the net revenue per hectare falling between R$ -159.35 and R$24.52; the minimum value is R$198.90 and the maximum value is R$64.69.

PAGE 108

108 Discussion Statistical Analyses With regard to soybean yield, it was expected that pre-financed fa rms would have higher yields than not pre-financed farms due to the Amaggi Groups recommendations on best management practices and respons ible pesticide use as mentioned in the previous chapters. However, this hypothesis was rejected because th ere was no statistically significant difference in soybean yields between the financial groups. This result may be explained by the fact that most farmers interviewed have adopted the same farm ing practices, such as the no-till system, crop rotation, and integrated pest management. It was also hypothesized that farms in Si nop would have a lower average soybean yield than farms in Sorriso and Tapurah. In the recen t past, the main economic activity in Sinop was timber production; agricultural production began af ter 1995 (Pichinin, no date). Since farm land in Sinop is relatively young, it was expected that it would have lo wer yields. The results showed that there is a significant di fference (P<0.05) in soybean yield between Sinop and Sorriso, but not between Sinop and Tapurah. The soybean yield in Sinop is lower than in Sorriso, confirmi ng the initial hypothesis that areas in Sinop have lower yields. According to IBGE data (discussed in Chapter 1), the municipality of Sinop had a soybean productiv ity of 2.88 tons per he ctare in the 2005/2006 soybean harvest, compared with 3.12 tons per hectare in Sorriso, and 3.06 tons per hectare in Tapurah. Although it is unknown if these differen ces are statistically different, they do lend support to the presen t studys findings. Tapurah is also a relatively recently establis hed municipality; land cl earing began in the 1980s to make way for agriculture and cattle ranchi ng. Nonetheless, this fact was not reflected in the results of this analysis, since there were no statistical differences between farms in Tapurah

PAGE 109

109 and Sorriso and between farms in Tapurah and Sinop. Mato Grossos average soybean yield over the last few years has been approximately 50 ba gs (3 tons) per hectar e (Folha do Estado, 2006). According to this analysis, the average soybean yield for the financial and municipal groups is above the state average. Net Revenue Analyses The results revealed that the smaller farm with 650 hectares of soybean area is losing money in the soybean business. It s net revenue per hectare was R$67, while the larger farm with 6,150 hectares of soybean had a net revenue of R$138 per hectare; a difference of R$205 per hectare, of which 51.0% is due to soybean yi eld, and 48.2% is due to production cost (not including income tax). According to Kaimowitz and Smith (2001), soyb ean in the Cerrado region is characterized by economies of scale. The large and modern proce ssing and storage facilitie s, low cost access to transportation, infrastructure and financial, t echnological and marketing systems required to produce soybean competitively implies economies of scale at the sector level. Moreover, mechanized soybean production also exhibits economies of scale at the farm level. According to Conte (2006), soybean producers in Mato Gosso benefit from economies of scale up to a threshold of 7,900 hectares. This help s explain why the larger soybean farm is more profitable than the smaller farm. Conte states that increasing returns to scale enables more efficient use of land, labor and machinery, and ma rket advantages for the purchase of inputs and the sale of outputs. However, this study did not result in enough data to make any conclusions regarding economies of scale from these two farms. There is a large litera ture that hypothesizes that small farms are more profitable (on a per hectare basis) than large farms, because of re duced labor costs (salar ies/benefits) and labor supervision costs (Kuma, 1980). Although the above may be accurate when considering family

PAGE 110

110 farms that produce goods for home consumption and sell goods to local markets, the present analysis demonstrates that the si ze of the smaller commercial farm considered here is not more profitable due to reduced labor costs. Another factor explaining the difference in net revenue per hect are between these two farms is that yield in areas that have been recently cleared is ge nerally lower than in areas that have been cultivated for longer periods of time (A. L. M. Pissollo, personal communication, November 28, 2006). This is evident in the yiel d comparison between the smaller farm and the larger farm: 54 bags per hectare in the case of the larger farm and 49 bags per hectare in the smaller farm. In the previous year (2004/2005 so ybean harvest), when the smaller farmer first planted soybeans in his farm, the difference in yi eld was even more exaggerated: in the larger farm where 4,950 hectares of soybean were plante d, the yield was 53 bags per hectare; in the case of the smaller farm, 250 hectares of soybean yielded 36 bags per hectare. Moreover, farmers entering the industry on land that was previously used for other purposes face other challenges as well. Initial investment costs required to farm soybeans are high. According to the smaller farm owner, he is indebted as a conseque nce of high investment costs such as land purchase, la nd clearing, soil prepara tion, and interest to pay on loans for land purchase, soil conditioning, and other farm operati ons; none of the above is included in this analysis. Furthermore, in the case of the smalle r farm owner, he was new to the business which, although not quantified, undoubtedly imposed additional costs in the way of inefficiencies due to a lack of experience. Although the smaller farmer is losing money, as the farm becomes more productive as the soil conditions improve and the fa rmer gains experience in the industry, it is expected to make positive net revenues. The break-even point for the smaller farm in this case would be

PAGE 111

111 approximately 52.5 bags per hectare with an exch ange rate of R$2.18 and the soybean price of US$9.60 per bag. Risk Analyses Risk modeling was conducted to determine how th e exchange rate and soybean price affect the viability of the farms in this case study and how the smaller farm is particularly susceptible to these fluctuations: a variation of 10% in the exchange rate wit hout directly impacting pesticide and fertilizer costs (Scenario 1) showed that th ere was less than 5% risk of the larger farmer losing money with soybeans. For the smaller farm er that was already having a loss, there was more than 80% risk of losing money. Variation of 10% in the exchange rate direc tly impacting soybean pr ice and fertilizer and pesticide costs (Scenario 2) did not represent a risk of losing money for the larger farmer. For the smaller farmer, who was already losing money, it did not represent a chance of having positive net revenue but it did increase his losses. The fre quency (y values on the triangular distribution) randomly chosen by @Risk for net revenues were higher for Scenario 2 than for Scenario 1, which led to a narrower dispersion of the net re venue values around the mean net revenue. This means that the chances of losing money and making profit were smaller for the farmers when variations in exchange ra te directly impact fertilizer and pesticide costs. Variation of 10% in the US do llar soybean price (Scenario 3) revealed a small risk of the larger farmer losing money, and more than an 85 % risk of the smaller farmer losing money; the risks taken by both farmers are smaller in Scenar io 1 and higher in Scenario 3. The smaller farm had a higher risk or higher probabi lities of losing money than the larger farm in all scenarios. Fluctuations in exchange rate and soybean pri ce (US$) strongly influences the structure of the soybean industry. Small farmers have a smalle r profit which renders th em more susceptible

PAGE 112

112 to uncertainties in the exchange rate and soybean price. This helps expl ain why the industry is dominated by larger farms. Conclusion This chapter showed that there was no signi ficant difference in soybean yield between the financial groups, since farmers, regardless of whether they were pre-financed by Amaggi Group or not, adopt the same agricultural practices su ch as the no tillage system. However, soybean farms in Sinop had lower yield than the farms locat ed in Sorriso. This resu lt lends support to the hypothesis that soybean yield in ar eas that were recently deforested is lower than in areas that have been cultivated for longer periods. The case study demonstrated that the smaller farmer is more susceptible to uncertainties in soybean pr ice and the exchange rate. Moreover, the smaller farm was less profitable; this ma y be explained in part by the s horter length of time that the smaller farm was cultivated affecting soybean yi eld, and the farmers lack of experience in soybean farming.

PAGE 113

113 Table 4-1. Descriptive statisti cs and independent samples test for soybean yield in 2005/2006 harvest. Comparisons are between fina ncial groups and among municipal groups. This table shows number of farms (N) in each group; minimum, maximum, mean, and standard deviation of values reported wi thin each group; and t-test statistics for comparisons between groups. Comparisons N Min. Max. Mean Std. Dev. t-value P Soybean yield Pre-financed 17 45 61 53.88 5.023 -0.862 0.395 (60 kg bag Not financed 20 45 63 55.23 4.456 per hectare) 1-Sinop 10 45 59 51.70 5.376 1-2 -2.474 0.020* 2-Sorriso 20 48 63 56.05 4.084 2-3 0.790 0.437 3-Tapurah 7 48 59.5 54.64 3.966 1-3 -1.298 0.238 indicates t-value significant at P<0.05. Table 4-2. Summary statistics fo r risk analyses, showing net s oybean revenue per hectare for a larger farm (6150 ha) and a smaller farm (650 ha) in different scenarios: (1) exchange rate varying 10%, (2) exchange rate varyi ng 10% and directly in fluencing pesticide and fertilizer costs, and (3) soybean price (US$) varying 10%. Min. Max. Mean Std. Dev. 5%tile 95%tile Scenario 1 Larger farm R$-45.64 R$321.78 R$138.01 R$76.49 R$9.73 R$266.00 Smaller farm R$-231 R$97.23 R$-67.33 R$68.49 R$-182.20 R$47.28 Scenario 2 Larger farm R$50.53 R$226.63 R$138.01 R$36.57 R$76.75 R$199.22 Smaller farm R$-102.21 R$-32.00 R$-67.33 R$14.58 R$-91.76 R$-42.92 Scenario 3 Larger farm R$-8.92 R$285.45 R$138.01 R$61.27 R$35.25 R$240.59 Smaller farm R$-198.90 R$64.69 R$-67.33 R$54.87 R$-159.35 R$24.52

PAGE 114

114 0 0.5 1 1.5 2 2.5 3 3.5 4 4.54/1/2002 7/1/2002 10/1/2002 1/1/2003 4/1/2003 7/1/2003 10/1/2003 1/1/2004 4/1/2004 7/1/2004 10/1/2004 1/1/2005 4/1/2005 7/1/2005 10/1/2005 1/1/2006 4/1/2006 7/1/2006 10/1/2006 1/1/2007 4/1/2007 Exchange rate (R$/US$) Figure 4-1. Trend line for the Brazilian real:US Dollar exchange rate from April 2002 to April 2007. [Source: FXHistory: historical currency exchange rates, 2007. OANDA corporation. Available from: http://www.oanda.com/convert/fxhistory (accessed May 2007).] Figure 4-2. Biodiesel production w ith soybeans in a farm in Tapur ah. Pictures taken by author June 23rd, 2006.

PAGE 115

115 Figure 4-3. Screen capture of the Microsoft Ex cel sheet showing the RiskTrigen formula for exchange rate and dependent variables in scenario 1.

PAGE 116

116 A B Figure 4-4. Triangular dist ribution for net soybean revenue in s cenario 1: exchange rate varying 10%. A) Probability distribu tion for the larger farm. B) Probability distribution for the smaller farm. A B Figure 4-5. Triangular dist ribution for net soybean revenue in s cenario 2: exchange rate varying 10% and directly influencing pe sticide and fertilizer costs. A) Probability distribution for the larger farm. B) Probability distribution for the smaller farm. Distribution for Net Revenue/ha-#2/C34 0.000 0.005 0.010 0.015 0.020 0.025 0.030 Mean=-67.33094 -110 -90 -70 -50 -30@RISK Trial VersionFor Evaluation Purposes Only -110 -90 -70 -50 -30 5% 90% 5% -91.7609 -42.9248 Mean=-67.33094 Distribution for Net Revenue/ha-#1/C22 0.000 0.002 0.004 0.007 0.009 0.011 0.013 0.016 0.018 0.020 Mean=138.014 40 90 140 190 240@RISK Trial VersionFor Evaluation Purposes Only 40 90 140 190 240 5% 90% 5% 76.7484 199.2198 Mean=138.014 Distribution for Net Revenue/ha-#2/C34 Values in 10^ -3 0 1 2 3 4 5 6 Mean=-67.33234 -250 -200 -150 -100 -50 0 50 100@RISK Trial VersionFor Evaluation Purposes Only -250 -200 -150 -100 -50 0 50 100 5% 90% 5% -182.2036 47.2754 Mean=-67.33234 Distribution for Net Revenue/ha-#1/C22 Values in 10^ -3 0 1 2 3 4 5 6 Mean=138.0121 -50 0 50 100 150 200 250 300 350@RISK Trial VersionFor Evaluation Purposes Only -50 0 50 100 150 200 250 300 350 5% 90% 5% 9.7319 265.9979 Mean=138.0121

PAGE 117

117 A B Figure 4-6. Triangular distributi on for net soybean revenue in s cenario 3: soybean price (US$) varying 10%. A) Probability distributi on for the larger farm. B) Probability distribution for the smaller farm. Distribution for Net Revenue/ha-#2/C34 Values in 10^ -3 0 1 2 3 4 5 6 7 8 Mean=-67.33131 -200 -150 -100 -50 0 50 100@RISK Trial VersionFor Evaluation Purposes Only -200 -150 -100 -50 0 50 100 5% 90% 5% -159.3504 24.5226 Mean=-67.33131 Distribution for Net Revenue/ha-#1/C22 Values in 10^ -3 0 1 2 3 4 5 6 7 Mean=138.0132 -50 0 50 100 150 200 250 300@RISK Trial VersionFor Evaluation Purposes Only -50 0 50 100 150 200 250 300 5% 90% 5% 35.2528 240.5892 Mean=138.0132

PAGE 118

118 CHAPTER 5 SUMMARY AND CONCLUSIONS This study provided a comprehensive evaluati on of the management practices adopted by soybean farmers in northern Mato Grosso, Brazil The soybean farmers in Mato Grosso claim that they are adopting sustaina ble agricultural practices to meet the growing demand for food while minimizing their impact on the environment and natural ecosystems. They seem to be aware that intensive land use is unsustainable b ecause of contamination of soil and water, and the importance of landscape conservation for the prot ection of biodiversity and recreation values. Given that the Amaggi Group is well known fo r its environmental work with their prefinanced producers, the management practi ces adopted by soybean farmers who were prefinanced by Amaggi Group in the year of 2006 we re compared to farmers who were not prefinanced by the Group. Moreover, based on the fa rms location and the different biomes in northern Mato Grosso, differences in farming pr actices among the study farmers located in the municipalities of Sinop, Sorriso, a nd Tapurah were also evaluated. Based on Amaggi Groups pre fi nancing requirements such as having a registered Legal Reserve and not deforesting illegally, it was hypothes ized that the pre-financed farmers are more likely to preserve forested area than farmer s that were not. However, this hypothesis was rejected. This result may be explained by the f act that soybean buyers other than Amaggi Group may have similar pre-financing policies. Comparisons among municipalities of farm forest cover, however, revealed that farms in Sorriso have a smaller percentage of natural forest on farm than farms in Sinop and Tapurah. This result lent support to the hypot hesis that the producers located in Sorriso are less likely to preserve forested areas than those located in Si nop and Tapurah. Farms in Sorriso are situated in the Cerrado biome, while farms in Sinop are in the Amazon forest biome. According to the New

PAGE 119

119 Forestry Code, 80% of the properties in the Amazon cannot be cleared, but only 35% of the properties must be conserved if located in the Cerrado. Although farms in Tapurah are also in the Cerrado biome, the municipality of Tapurah is not situated along the BR-163 highway, which could have contributed to a greater percentage of natural forest on farms compared to those located in Sorriso. It is generally accepted that in the Brazilian Amazon, deforestation occurs in proximity to roads (Pfaff et al., 2007). With regard to soybean management practi ces, Amaggi Group recommends the adoption of the no-tillage system and that the area where no -tillage is implemented is increased. However, statistical analyses re jected the hypothesis that the pre-financed fa rmers have a greater percentage of soybean area in a no tillage system. It was also hypot hesized that the pre-financed farmers have a greater cover crop diversity. Alth ough this hypothesis was not tested statistically, survey data suggested that prefinanced farmers are interested in diversifying their production, since more pre-financed farmers reported planting different cover crops, besides millet and corn. Among the municipalities, it was expected th at farmers in Sorriso would use a higher percentage of their soybean area to plant corn as a cover crop, since the municipality of Sorriso is the fourth largest corn producer in Brazil. Howe ver, this hypothesis wa s rejected. One possible explanation is that high levels of precipitation in the region of Sorriso could have delayed the soybean harvest and consequently the pl anting of corn in the study year (2006). Related to soybean planting practices, one important finding was that although GM soybeans are legally permitted to be planted in Mato Grosso, the percentages of area planted among the study farmers were low. Results indicate that farmers are not very interested in planting GM soybeans. This is likely due to the fact that in the case of pre-financed farmers, they were receiving better prices for non-GM s oybean. Amaggi Group exports non-GM soybean

PAGE 120

120 through a private port thus reducing transportatio n costs. In addition, farmers did not perceive advantages in planting GM soybeans, since ther e is no difference in production costs between GM soybeans and non-GM soybeans (Fundao Mato Grosso, 2006). Survey data did not show any major differen ces between financial gr oups with regard to pathogenic diseases; all farmers reported havi ng Asian rust and some farmers reported having anthracnose and foliar blight disease in their soybean cr ops. Stinkbugs, whiteflies, and caterpillars were the insect pest s most often reported in the fa rmers soybean crop as well as nematodes. Among the municipalities, high levels of precipitation in the regions contributed to the appearance of Asian rust and foliar blight, however, farmers located in Sinop were the only ones who did not report foliar blight in their crop, maybe as a resu lt of the fungicide used for seed treatment. With regard to pesticide use, it was hypothesized that the pr e-financed farmers use fewer types of pesticides than those who are not, due to the fact that th e Amaggi Group recommends that pre-financed farms practice integrated pest management and as such, use inoculations in seed treatment, adopt biological control or insect growth regulators when possible, and vary the types of pesticides used to avoid insect and w eed resistance. However, the survey data did not support this hypothesis since all fa rmers appeared to have adopted similar integrated pest management strategies. The assumption that farms surrounded by natural fore st or in close proximity to forest need to use less pesticides, led to the hypotheses that the producers located in Sinop use less fungicide and apply it fewer times than the producers located in Sorriso and Tapurah, and that farms with a higher percentage of forest cover use fewer types of pesticides. However, statistical analyses did not reveal differences in the number of times fungicides were applied among the municipal

PAGE 121

121 groups; and based on survey data, proximity to fore st or amount of forest cover on farm did not appear to affect pesticide use pa tterns. However, farmers in Tapur ah applied desiccants prior to soybean harvest in a greater percentage of soybean area than the farmers located in Sorriso. This result can be partially explained by the fact that the soybean plants were not dried enough to be harvested due to heavy rainfall and high levels of moisture in the region. Farm proximity to forested area likel y was not a factor. There were some interesting tr ends in pesticide use among the farmers. Farmers, regardless of financial status, location, or amount of forest cover, repo rted using fungicide for seed treatment. Farmers also rely on the herbicide glyphosate since no-tillag e requires a pre-harvest herbicide application to dry th e soybean plants. Farmers also use different types of postemergence herbicides because most of their seeds are non-GM soybeans. The percentage of farmers adopting biological contro l methods is small due to its in effectiveness against stinkbugs. Finally, an average of two fungicide applica tions per crop was observed among the farmers. With regard to soybean yields, the hypothesis th at pre-financed farmers have greater yields than those who are not was rejected. Although the Amaggi Group recommends that the prefinanced farmers adopt good farmi ng practices, there were no di fferences in soybean yields between pre-financed and not pre-financed farmers. This result may be explained by the fact that most of the interviewed farmers have adopted the same farming practi ces, such as the no-till system, crop rotations, and integrated pest management. Among the municipalities, the hypothesis that the producers located in Sinop have lower yields than those located in Sorriso and Tapurah was partially rejected. Soybean farms in Sinop had lower yield than the farms lo cated in Sorriso but similar yield when compared to farms in Tapurah. Since farm land in Sinop is relatively younger than farm land in Sorriso, the result

PAGE 122

122 lends supports to the assertion that soybean yield in areas that were recently deforested is lower than in areas that have been cultivated for longer periods. The case study presented in Chapter 4 also supports this finding. In a study of two case farms, the smaller farm was less profitable compared to the larger farm due to dissimilarities in soybean yield and production costs. The great difference in net revenue per hectare may be explained in pa rt by economies of scale in soybean farming (although this was not tested in th e study), the shorter le ngth of time that the smaller farm was cultivated, and the farmers lack of experience in soybean farm ing. Moreover, the case study also showed that the smaller farmer was more susceptib le to risks and uncertain ties in the exchange rate and soybean price. Variations in these va riables increased the risk of the smaller farm increasing his losses. This study provides feedback to the Amaggi Group about the farming practices adopted by their pre-financed farmers, not pre-financed fa rmers and whatever differences that may be the result of their specific location in northern Mato Grosso. This info rmation is important in that it can serve to direct Amaggi Gr oups extension programs to enable farmers to overcome barriers to the adoption of good farming practices and pr otect environmental values. In addition, the analytical framework presented here can be applied to other agricultural regions, showing whether or not adoption of similar management practices can be su stainable, produce less environmental impact, and be economically viable. This study identifies a number of areas for fu rther research. In the case of smaller farms and given the current soybean price and the exch ange rate, benefit-cost analysis of farm diversification, biodiesel producti on and conservation tillage, for ex ample, could be particularly fruitful with environmental and energy concerns Increasing the productiv ity of soybean farming

PAGE 123

123 is also critical. New varieties of soybeans that are better suited to th e physical and biological conditions of the region can increase supply fr om the existing agricultural land-base. How producer cooperatives can improve farm-gate prices is also an interesting area to be pursued. These innovations may have a significant impact on reducing deforestatio n and the expansion of soybean production toward the Amazon forest.

PAGE 124

124 APPENDIX SEMI-STRUCTURED INTERVIEW Tpico 01: Sobre a fazenda (ltimos 2 anos, ano atual, estimativa para o prximo ano) Fazenda: ( )prpria ( )arrendada Ano de compra:___________ Situao em que se encontrava: ( )pastagem degradada ( )cultivo de soja ( )mata fechada ( )juquira ( )esterada ( )semi-aberta ( )outro:______ Distncia da cidade(km): _____________ ltimo ano de desmate:______________ rea (ha): ________________________ rea total da propriedade (ha): Planta safrinha? ( )no ( )sim rea de soja plantada: Safra 2003/2004 (ha): saca/ha: Safra 2004/2205 (ha): saca/ha: Safra 2005/2206 (ha): saca/ha: Safra 2006/2007 (ha): saca/ha: Safrinha: ( )milho ( )milheto ( )sorgo ( )outro:___ Safra 2003/2004 (ha): Safra 2004/2205 (ha): Safra 2005/2206 (ha): Safra 2006/2007 (ha): Planta outro tipo de cultura? ( )no ( )sim Outra cultura: ( )feijo ( )arroz ( )algodo ( )outro:____ Safra 2003/2004 (ha): Safra 2004/2205 (ha): Safra 2005/2206 (ha): Safra 2006/2007 (ha): Possui rea de pastagem? ( )no ( )sim rea de pastagem (ha): Cabea de gado: Ano 2004: total: /ha: Ano 2005: total: /ha: Ano 2006: total: /ha: Ano 2007: total: /ha: GMO: ( )no ( )sim, safra:___________ ha:__________ Presena de rvore no pasto: ( )nativa ( )reflorestada ( )n.a. Tpico 02: Boas Prticas Agrcolas O que o senhor entende por Boas Prticas Agrcolas? ( )aumento de custo ( )reduo nos custos ( )preservao do meio ambiente ( )conservao do solo ( )aumento no rendimento ( )aumento na produtividade ( )outro:_____________________ Boas Prticas agrcolas adotadas e desde quando: ( )plantio direto, ano:_____ ( )rotao de cultura, ano:_____ ( )controle biolgico, ano: _____ ( )outro:__________________________________ Por que essas prticas foram adotadas? ( )exigncia do financiador ( )conscientizao prpria ( )exigncia dos clientes ( ) exigncia do fornecedor ( )outro:__________________ Essas prticas tm mostrado algum resultado? ( )sim ( )no Que tipo de resultado essas prticas tem mostrado? ( )resultados econmicos ( )resultados na produo ( )no meio ambiente ( )na conservao do solo ( )melhor visto no mercado ( )captao de mais clientes ( )outros:__________________ Que outros resultados gostaria de ter: ( )econmicos ( )resultados na produo ( )no meio ambiente ( )na conservao do solo ( ) no mercado ( )captao de mais clientes ( )outros:________________________________ Plantio Direto Controle Biolgico Desde quando: Hectares de soja plantados: poca de plantio: Vantagens: Desvantagens: Resultado na produo de soja: Desde quando: Hectares de soja plantados: poca de plantio: Vantagens: Desvantagens: Resultado na produo de soja:

PAGE 125

125 Resultados econmicos: nos custos: na rentabilidade: Resultados econmicos: nos custos: na rentabilidade: Rotao de soja com pasto Desde quando: Freqncia: Hectares de pasto: Hectares de soja: Quantas cabeas de gado/ha.: poca de plantio: poca de criao de gado: Vantagens: Desvantagens: Resultados na produo (kg/carne/h ou @/h): Resultados econmicos: nos custos: na rentabilidade: Tpico 03: Reserva Legal e rea de Preservao Permanente Possui rea de Reserva Legal? ( )no ( )sim rea (ha): ___________ ou % ______________ Se sim, RL : ( ) nativa ( ) reflorestada rea averbada ( )sim ( )no Possui rea reflorestada ( )no ( )sim rea:_____________ espcie:_______________ Se no, pretende reflorestar? ( )no ( )sim Se sim, que espcie de rvore pretende plantar? Possui rea de preservao Permanente? ( )no ( )sim. Se sim, qual a largura? ____________ So reas ( )nativa ( )reflorestada: espcie:___________ Se no, pretende reflorestar? ( )no ( )sim Se sim, que espcie de rvore pretende plantar? Tpico 04: Conservao do Solo H reas mais propcias a eroso do solo? ( )no ( )sim Na sua propriedade tem eroso de solo? ( )no ( )sim. Se sim, quantos ha.?________________ Tem conhecimento por que a eroso ocorre? ( )no ( )sim, por que? Como o senhor controla a eroso do solo? ( )boas prticas agrcolas ( )no controla Tpico 05: Pestes Agrcolas Que tipo de doenas agrcolas tem na sua propriedade? ( )erva daninha: ( )patognicos: ( )ferrugem ( )outros:___________ ( )nematide: ( )cisto ( )galha ( )insetos: ( )lagarta ( )percevejo ( )mosca branca Classifique de 1 a 6 p/ a mais severa a menos severa: ( )erva daninha ( )patognicos/ferrugem ( )nematide ( )mosca branca ( )percevejo ( )lagarta Etapas Agro-qumicos Combater: Dessecao plantio ( )Roundup original ( )Rondoup Transorb ( )Roundup W.G. ( )24D ( )Glifosato ( )Paraquat ( )outros:_______________ Tratamento semente Inceticida: ( )Standak ( )Fungicida: ( )Vitavax Thiran Inoculante: ( )no ( )sim:____________ Ps emergente Folha larga Folha estreita ( )Pivot ( )Verdict ( )Classic ( )outro:______________ ( )Cobra ( )outro: ____________ Inseticida Biolgico: ( )no ( )Baclovirus ( )Larvin ( )outro:

PAGE 126

126 Fisiolgico: ( )Nomolt ( )outro:______________ Veneno: ( )Metamidofs ( )Talcord ( )Endosulfan ( )Folidol ( )outro:___________________ Fungicidas ( )pera ( )Folicur ( )Cercobin ( )Priori ( )Impact ( )Priori xtra ( )Impact Duo ( )Stratego ( )Sphere ( )outro:___________________ Dessecao colheita ( )Round-up ( )24D ( )Gramoxone ( )Reglone ( )Smash Quantos hectares plantados com cu lturas resistentes a nematides? Qual o destino das embalagens vazias? ( )Trp lice lavagem ( )reciclagem ( )devoluo das embalagens ( )jogado no lixo ( )aterro privado ( )outro:________________

PAGE 127

127 LIST OF REFERENCES A Gazeta. 2007. Custo mdio da produo de so ja est 25% maior pa ra safra 07/08. Noticias Agrometeorolgicos. INPE/CPTEC. Ministrio da Cincia e Tecnologia. Available from: http://www.cptec.inpe.br/cgi-bin/we bpub/noticia.cgi?6538 (in Portuguese). Agnol, A. D. 2006. Perspectivas da Soja Brasilei ra. Embrapa Soja. Centro de Inteligncia da Soja. Available from: http://www.ciso ja.com.br/index.php?p=artigo&idA=9 (in Portuguese). Altmann, N. 2006. Rotao, sucesso e consorcio de espcies para ag ricultura sustentvel. Pages 236-240 in S. Suzuki, M. M. Yuyama, and S. A. Camacho, editors. Boletim de Pesquisa de Soja 2006. Fundao Mato Grosso, Rondonpolis, MT (in Portuguese). ASSECOM (Assessoria de Comunicao da Prefe itura de Sinop). 2006. Identificando nosso municpio. Sinop, MT. Available from: http://www.conheca.sinop.mt.gov.br/ (accessed June 2007) (in Portuguese). Baer, J. 2006. Agricultural commodity facts. Private Banking Investment Research. Bank Julius Baer & Co. Ltd. Available from: http://www.juliusbaer.com/runappl.cfm/Agr icultural_EN?wm=a(302)&dp=a(dl)d(2557)&e xt=. Bickel, U. and J. M. Dros, 2003. The Impacts of Soybean Cultivation on Brazilian Ecosystems, Three case studies. WWF, the global cons ervation organization. Available from: http://assets.panda.org/downloads/impactsofsoybean.pdf. Bortolini, C. G. 2006. Integrao Lavoura Pecur ia: A gerao da terceira safra do ano. Pages 242-248 in S. Suzuki, M. M. Yuyama, and S. A. Camacho, editors. Boletim de Pesquisa de Soja 2006. Fundao Mato Grosso, Rondonpolis, MT (in Portuguese). Brazil, Lei n. 4.771, de 15 de setembro de 1965. Dirio Oficial da Repblica Federativa do Brasil, Braslia (DF), 16 de set. 1965. Brazil, Medida Provisria n. 1.956/50, de 27 de maio de 2000. Dirio Oficial da Repblica Federativa do Brasil, Braslia (DF), 28 de maio de 2000. Brown J. C., M. Koeppe, B. Coles, and K. P. Price. 2005. Soybean Production and Conversion of Tropical Forest in the Brazilian Amazon: The Case of Vilhena, Rondnia. AMBIO: A Journal of the Human Environment 34(6):462. Brum, A. L. 2004. Economia da Soja: Historia e Futuro. Fazendas MT, Cuiab, MT. Available from: http://www.fazendasmt.com.br/ar tigos/ler_art.php?id= 3 (in Portuguese). Buchholz, D. D., E. Palm, G. Thomas, and D. L. Pfost. 1993. No-Till Planting Systems. MU Extension, University of Miss ouri-Columbia. Available from: http://extension.missouri.edu/e xplore/agguides/crops/g04080.htm.

PAGE 128

128 Campbell, H., and R. Brown. 2003. Benefit-cost analysis. Cambridge University Press, 345pp. Conte, L. 2006. Economia de escala e substituio de fatores na produo de soja no Brasil. Escola Superior de Agricultura Luiz de Quei roz. Universidade de So Paulo 1:115. Tese de doutorado (in Portuguese). Degrand, P. E., and L. M. Vivan. 2006. Pragas da Soja. Pages 153-179 in S. Suzuki, M. M. Yuyama, and S. A. Camacho, editors. Boletim de Pesquisa de Soja 2006. Fundao Mato Grosso, Rondonpolis, MT (in Portuguese). Dirio de Cuiab. 2007. Preo da soja regi strou queda de 28% em 2005 no MT. Avicultura Industrial. Available from: http://www.aviculturaindustrial.com.br/sit e/dinamica.asp?id=27249& tipo_tabela=negocios &categoria=insumos (accessed June 2007) (in Portuguese). Dias, W. P., J. F. V Silva, and A. Garcia. 2006. Ne matides de Importncia para a soja no Brasil. Pages 139-151 in S. Suzuki, M. M. Yuyama, and S. A. Camacho, editors. Boletim de Pesquisa de Soja 2006. Fundao Mato Grosso, Rondonpolis, MT (in Portuguese). EMBRAPA (Empresa Brasileira de Pesquisa Ag ropecuria). 2004. Tecnologias de Produo de Soja Regio Central do Brasil 2004. Sistema de produo N0 1. Embrapa soja. Available from: http://www.cnpso.embrapa.br/pr oducaosoja/index.htm (in Portuguese). FAO (Food and Agriculture Organization) and UNDP (United Nations Development Programme). 2001. Zero tillage: Brazil. Shari ng innovative experiences Examples of successful initiatives in agriculture and ru ral development in the South. FAO and UNDP. Available from: http://tcdc.undp.or g/sie/experiences/vol5/zero.pdf. Fearside, P. M. 2001. Soybean as a threat to the environment in Brazil. Environmental Conservation 28(1):23-38. Folha do Estado, 2006. Doena aumenta custos de produo da soja. Agrolink O Portal do Contedo Agropecurio. Available from: http://www.agrolink.com.br/agrolinkfito/pg_ detalhe_noticia.asp?cod=36640 (in Portuguese). Gassen, D. N. 2002. Percevejos em soja. Agro link O Portal do C ontedo Agropecurio. Available from: http://www.agrolink.com.br/coluni stas/pg_detalhe_co luna.asp?Cod=344 (in Portuguese). Gazziero, D. L. P, 2006. Manejo de espcies infe stantes de reas cultivadas com Soja. Pages 181-185 in S. Suzuki, M. M. Yuyama, and S. A. Camacho, editors. Boletim de Pesquisa de Soja 2006. Fundao Mato Grosso, Rondonpolis, MT (in Portuguese). Goldsmith, P. and R. Hirsch. 2006. The Brazili an Soybean Complex. American Agricultural Economics Association 21(2):97-104.

PAGE 129

129 Grupo Andr Maggi, 2006. Recomendaes Ambienta is. Internal document (in Portuguese). Grupo Andre Maggi. 2007. Empresas do Grupo. Rondonpolis, MT. Available from: http://www.grupomaggi.com.br/pt-br/index.jsp (accessed June 2007). Fasulo, T. R. 2006. USDA Whitefly Knowle dgebase. Department of Entomology and Nematology. University of Florida. Available from: http://whiteflies. ifas.ufl.edu/wfly0002 .htm. IBGE (Instituto Brasileiro de Geografia e Esta tstica). 2005. O municpio de Sorriso (MT) se destaca na produo de soja. Produo Agr cola Municipal Cer eais, Leguminosas e Oleoginosas 2004. Available from: http://www.ibge.gov.br/home/presidencia/noticias/ noticia_visualiza.php ?id_noticia=395 (in Portuguese). IBGE (Instituto Brasileiro de Geografia e Es tatstica). 2006. Produo agrcola municipal. Ministrio do Planejamento, Oramento e Gesto. Available from: http://www.sidra.ibge.gov.br/bda/acer vo/acervo2.asp?e=v&p=PA&z= t&o=10 (accessed June 2006) (in Portuguese). ISA (Instituto Socioambiental). 2005. BR-163 Sustentvel. ISA. Available from: http://www.socioambiental.org/esp/BR163/ (accessed June 2007) (in Portuguese). Joels, L. M. 2002. Reserva Legal e Gesto Am biental da Propriedade Rural: um estudo comparativo da atitude e comportamento de agricultores orgnicos e convencionais do distrito federal. Planeta Orgnico. Available from: http://www.planetaorganico.com.br/trabjoels2.htm (in Portuguese). Kaimowitz, D., and J. Smith. 2001. Soybean Tec hnology and the Loss of Natural Vegetation in Brazil and Bolivia. Pages 195-211 in Agricultur al Technologies and Tropical Deforestation of A. Angelsen and D. Kaimowitz, editors. CAB International Pub lishing. Available from: http://www.cifor.cgiar.org/publicatio ns/pdf_files/Books /CKaimowitz0101E0.pdf. Kuma, N. N. 1980. Reviewed Work: Agrarian Structure and Produc tivity in Developing Countries by R. Albert Berry; William R. Cline. Southern Economic Journal 47(1):255257. Lal, R. 2000. Physical Management of Soils of the Tropics: Priorities for the 21st Century. Soil Science 165(3):191-207. Machado, E. L., and M. A. Margarido. 2000. Seasonal Price Transmission in Soybean International Market: the case of Brazil a nd Argentine. IAMA Agribusiness Forum in Chicago, Illinois. Available from: http://www.ifama.org/conferences/2000Congress/ 2000_forum_papers.htm.

PAGE 130

130 Marques, A. B. F. A. 2004. Poltica Cambial Br asileira. Economia & Energia. MAK Editorao Eletrnica. Available from: http://ecen.com/eee15/cambio.htm (accessed July 2007) (in Portuguese). Mittermeier, R. A., P. R. Gil, M. Hoffmann, J. Pilgrim, T. Brooks, C. G. Mittermeier, J. Lamoreux, and G. A. B. da Fonseca. 2004. Hotspot s Revisited: Earth's Biologically Richest and Most Endangered Terrestrial Ecoregions Cemex Books on Nature. Available from: http://www.biodiversityscience.org/ publications/hotspots/cover.html Moreno, G. 2005. A Colonizao no sculo XX. A Po ltica estadual e federal de colonizao em Mato Grosso 1900/1990. Pages 52-71 in G. More no and T. C. S Higa, editors. Geografia de Mato Grosso. Entrelinhas, Cuiab, MT (in Portuguese). Moreno, G. 2005. Polticas pblica de infra-estrut ura e de desenvolvimento regional. Pages 172205 in G. Moreno and T. C. S Higa, editors. Geografia de Mato Grosso. Entrelinhas, Cuiab, MT (in Portuguese). Nair, P. K. R. 1990. Classification of agrofore stry systems. Pages 31-57 in Agroforestry: classification and management. New York: John Wiley & Sons. Ondro, W., L. Couto, and D. R. Betters. 1995. The stat us and practice of forestry in Brazil in the early 1990s. The Forestry Chronicle 71:106-119. Ott, R. L., and M. T. Longnecker. 2004. A Firs t Course in Statistical Methods. Brooks/Code Thomson Learning, Belmont, CA. Palisade Corporation. 2006. @Risk 4.5 for Excel Product Support Center. Available from: http://www.palisade-eu rope.com/Default.asp. Pfaff, A., J. Robalino, R. T. Walker, S. Aldrich, M. Caldas, E. J. Reis, S. Perz, W. Laurance, and K. Kirby. 2007. Road Investments, Spatial Spil lovers, and Deforestation in the Brazilian Amazon. Regional Science 47(1):109-123. Pichinin, E. S. no date. As se rrarias no contexto da Amaznia Mato-grossense. Universidade Estadual Paulista-FCT Unesp. Available from: http://www2.prudente.unesp.br/eventos/ semana_geo/ericapichinin.pdf (in Portuguese). Powers, L. E., and R. McSorley. 2000. Ecologica l Principles of Agriculture. Delmar Thomson Learning, Albany, NY. Riveras, I. 2007. Preo alto da soja mantm fluxo de dlares. Reuters. Agrolink O Portal do Contedo Agropecurio. Available from: http://www.agrolink.com.br/noticias/pg_detalh e_noticia.asp?cod= 55962 (accessed June 2007) (in Portuguese).

PAGE 131

131 Roessing, A. C. 2005. Soja Safra 2005/2006: e a gora? Embrapa Soja. Agrolink O Portal do Contedo Agropecurio. Available from: http://www.agrolink.com.br/ferrugem/ artigos_pg_ detalhe_noticia.asp?cod=33715 (acce ssed June, 2007) (in Portuguese). Russo, R. O. 1996. Agrosilvopastoral systems: a practical approach toward sustainable agriculture. J. Sust. Agric. 7(4):5-17. Schwenk, L. M. 2005. Domnios Biogeogrficos Pages 250-271 in G. Moreno and T. C. S. Higa, editors. Geografia de Mato Grosso. Entrelinhas, Cuiab, MT. Shimada, J. Y. 2006. Gesto Ambiental: Como produzir sem destruir. Pages 250-259 in S. Suzuki, M. M. Yuyama, and S. A. Camacho, editors. Boletim de Pesquisa de Soja 2006. Fundao Mato Grosso, Rondonpolis, MT. Sorriso City Hall. 2005. Histria Sorriso, MT. Available from: http://www.sorriso.mt.gov.br/inde x.php?pg= institu cional&tipo=1 (accessed June 2007) (in Portuguese). Souza, O. B. 2006. Deforestation in the Am azon Region and agribusiness. ISA (Instituto Socioambiental). Available from: http://www.socioambiental.org/esp/desmatamento/en/ agrobusiness Suzuki, S., M. M. Yuyama, and S. A. Camacho, editors. 2006. Boletim de Pesquisa de Soja 2006. Fundao Mato Grosso, MT (in Portuguese). Tapurah City Hall, 2006. Portiflio do municpio de Tapurah-MT. Tapurah, MT. USDA (United States Department of Ag riculture). 2004. World Soybean Production 2004. World Statistics. Available from: http://www.soystats.com/2005/page_30.htm (accessed November 2006). USDA. (United States Department of Agricult ure). 2006. Economic Research Service. Briefing Rooms. Brazil. Available from: http://www.ers.usda.gov/Briefing/Brazil/ (accessed June 2007). USDA. (United States Department of Agri culture). 2007. Record 2006/2007 Soybean Crop in Brazil. Commodity Intelligence Report. Fore ign Agricultural Serv ice. Available from: http://www.pecad.fas.usda.gov/highlights/200 7/03/brazil_soybean_30mar2007/ (accessed June 2007). Valdes, C. 2006. Brazils Booming Agriculture Faces Obstacles. AMBER WAVES 4(5):28-35. Vargas, L., Bianchi, M. and M. Rizzardi, 2006. Dessecao: manejo das invasoras antes da semeadura da soja. Embrapa Trigo. Agroli nk O Portal do Contedo Agropecurio. Available from: http://www.agrolink.com.br/ferr ugem/artigos_pg_detalhe_noticia.asp? cod=47151 (in Portuguese).

PAGE 132

132 WHRC (Woods Hole Research Center). 2007. Ag ricultural Frontier Explosion in Brazil. Available from: http://www.whrc.org/southamerica/agric_expans.htm (accessed June 2007). World Agroforestry Centre. 2006. Informati on ResourcesGlossary. Available from: http://www. worldagroforestry.org/Informati onResources /A-B.asp (accessed February 2006). Wyse, J. 2005. Soybean Asian Rust; Research upda te. North Central Soybean Research Program. Available from: http://www.planthealth.info/pdf _docs/AsianRust_screen.pdf Yorinoni, J. T., J. N. Junior, and J. J. Lazzar oto. 2004. Ferrugem asitica da soja no Brasil: evoluo importncia econmica e controle Documento 247. Embrapa Soja. Londrina. Available from: http://www.cnpso.embrapa.br/download/alerta/documento 247.pdf (in Portuguese).

PAGE 133

133 BIOGRAPHICAL SKETCH Carolina Maggi Ribeiro was bor n in So Miguel do Iguau, Pa ran state, Brazil and was raised in Rondonpolis, Mato Grosso where she graduated from high school in 1998. She earned a bachelors degree in business administration from The Centro Universitrio Franciscano do Paran in Curitiba in December 2004. During these studies, she also studied accounting for three years and participated in an extr acurricular course in management consulting. In her last year as an undergrad student she worked in the Finance Department of Fertipar Fertilizers do Paran. Before attending the University of Florida, sh e undertook some course work in environmental issues.