WTP5l 5 2002 A, _ 70 I .j }R R DA 711 I CUINAIA: G IANA 0 250 RAINFALL ZONES KLBoa0T EGUYANA\ ' ALARSNC -C s o f RORAI/MA . ,r.- /4 A,A,Pii OCcEAN ,-,-, %-e -t s>u Man]aus' Af TMcao * ;M1MA . A A// / A R ./: _ / (r Zu J m s ; MARANHAOt ¸'it:j __-'/ t } t,',_(- *i '('*f\ (i .#0 A C RE ~\t|Xriii *- &'V S -\ _:K. )_} ; ~- '- T VeIhOZ-Y Z4 \...7\ g~ Bron%COF N r / ¢' I ) trOAImlNs ~~- I" r~~~OCEATNSf- ___________________ RONDONIA ! \t 0 COLI ti 1A.-\ /f MATO - VIDIaof mop | B I I-lJ GOIAS - \KL1 B R A Z I L |RSII /| fUII ^1suif /l { gIr'i'f> t | t____ HUID ZONE NATIOAL CAMA h&t$3c 41 ____ TI1ANSIT1ONAL ZONE k EGICN flOUM0AY I° 2,200 mm/year). Statistical frontiers of Sinop (north-central Mato Gros- analysis revealed that, keeping other factors so) and the Vilhena - Ji Parana corridor constant, higher levels of rainfall in the (Rond6nia). The reasons is that, in these dryer Amazon reduce land conversion rates for areas, it has been possible to develop an alter- agriculture as the well as pasture productivity. native economy based on agriculture, particu- Moreover, the the wettest areas experience an larly grains. 2 Sustainable Amazon: Limitations and Opportunities for Rural Development Third, in the majority of land in the Ama- Finally, local economic forces make sus- zon, especially in the humid zone, forest man- tainable development in the Amazon frontier agement could provide a more stable difficult, because local political interests are economy-income, employment, and taxes - served by rapid economic growth, even if it than that produced by agriculture. leads to transient communities based on un- Fourth, it is necessary to expand and con- sustainable economies. Interest in the benefits solidate the national and state forest system as of sustainable development is generally na- part of a broader strategy to promote sustain- tional and global. able land-use and biodiversity protection. The Hence, it is essential that the federal creation of national forests could (1) help pro- government help state and local governments tect large areas by creating a mosaic of man- to guarantee sustainable development in the aged National Forests and national parks, the Amazon. The federal government could former serving as buffer zones for the latter; help stabilize the local economy through (2) prevent rural colonization in areas without both economic and regulatory instruments. agricultural potential; and (3) separate agri- These instruments include taxing wood cultural and logging frontiers, thereby reduc- derived from predatory logging, creating ing the economic benefits that unsustainable National Forests, improving the monitoring agriculture receives from being associated and control system, and providing selective with predatory logging. assistance to the implementation of forest A study by Verissimo and others (2000) management. revealed that there are 1.15 million km2 of However, a more proactive government forest in the Amazon with the potential to es- role is not without risk, which will have to be tablish National Forests. Of this total, ap- carefully managed from the outset. These risks proximately 38 percent coincides with areas include that of high importance for conservation of biodi- 1. Ill-conceived policies or poor implementa- versity (1999 Macapa Consultancy). In cases tion may do more harm than good (imple- of overlap, the authors recommend the desig- mentation risk). nation of these areas for absolute protection. 2. A strengthened and visible commitment of Even so, 0.7 million km2 would remain for the government to sustainable development in creation of National Forests, an area more the Amazon might raise expectations and than capable of sustainably supplying the cur-, increase government's exposure to criticism rent and expected near-term demands on the for a less-than-complete resolution of the Amazonian wood sector. problem (reputational risk). To ensure the success of National Forests, The authors believe that both implemen- it is crucial that an environmentally responsi- tation and reputational risk can be controlled ble and administratively efficient concession through careful attention to project phasing, system be established. In addition, to elimi- participation of stakeholders, and information nate incentives for unsustainable logging, it is dissemination. Based on the analysis presented crucial that a tax be imposed on wood de- in this report, the authors also believe that the rived from predatory harvests outside Nation- benefits of a more proactive government role al Forests. are well worth any possible risk. Introduction T his report comes at an opportune mo- opportunities derive from a heightened gov- ment in the Amazonian debate. After ernment commitment to confront its almost two decades without substantial environmental responsibilities, clearly articu- investment by the federal government in Ama- lated public support for more rational zonian transport infrastructure, the Brazilian land-use patterns in the Amazon, and a the government is planning actions that will pro- wealth of accumulated experience through the foundly alter the regional landscape. Pilot Program to Conserve the Brazilian Rain- First, a signficant expansion of the trans- forests (PPG7). portation system is planned under the pro- In any case, the government cannot re- gram, National Axes of Integration and main passive. Either it will lead the redirection Development. This initiative, the most signifi- of economic forces that are degrading the Am- cant since the pavement of BR 364 (Cuiaba - azon, or the national heritage will be appro- Porto Velho interstate) at the beginning of the priated for private interests. 1980s, will dramatically increase access to the The general objective of this report is to Amazon's natural resources. assist decisionmakers, civil society, and the pri- Second, the Ministry of the Environment vate sector in guaranteeing that Amazonian is implementing a new forest policy based on natural resources will be maintained in the use the expansion and consolidation of national that maximizes their social and environmental and state forests in the legal Amazon (referred value. In the first chapter, the authors review to in the present report as the Amazon). The the effects of rainfall on the agricultural pro- government's goal is to allocate 500,000 km2 ductivity of Amazon. In the second, the au- (10 percent of the Brazilian Amazon) for the thors analyze the economic performance of the creation of National Forests. principal land-use activities. For this analysis, Finally, the federal government has made the authors consider a typical Amazonian an international commitment to protect biodi- county in the humid zone whose economy is versity in the Amazon through expanding na- based on extensive ranching and predatory tional reserves (areas of complete protection) timber harvesting. In the third chapter, the au- to cover a minimum of a representative 10 thors examine the strategic role that govern- percent of the territory. ment can play in ensuring economically These government initiatives offer both sustainable development in the Amazon. Final- opportunities and risks. The risks stem from ly, the authors analyze the government's plan the investments in infrastructure and from to expand and consolidate national and state the difficulty of organizing the economic forests (public areas of sustainable use) within forces that improved access will unleash. The this context of sustainable development. 3 I Effect of Rainfall on the Performance of Agriculture and Ranching in the Amazon In this chapter the authors summarize the the gains obtained in soil fertility are largely information available in the literature on lost due to excessive rainfall. Cochrane and agricultural performance in the tropics, in Sanchez 1992) concluded that excessive rain- particular the Amazon. For the purpose of fall and saturated soils, especially in the cen- subsequent analysis, they then conceptually tral Amazon, impose a natural barrier to divide the Amazon into three rainfall zones. agricultural development. Gallup and Sachs Finally, the authors employ data from the (2000) observed that climate is one of the key 1995-96 agricultural census (Brazilian Insti- factors for the relative failure of agriculture in tute of Geography and Statistics, or IBGE) to the tropics. These authors stated that, despite evaluate the effect of rainfall on the productiv- research efforts, the humid tropics continue to ity and economic return of agriculture. display low productivity for their primary crops, including corn, rice, tubers, vegetables, Literature and cattle and pig ranching. The exceptions are perennial crops such as banana, cocpnut There is an ample literature noting the gener- and oil palm. ally low agricultural potential of the humid A study of the state of the art of agricul- Amazon (Goodland and Irwin 1975, Moran ture in the humid tropics, commissioned by 1981, Smith 1982, Hecht and others 1998, the North American Council of Research on Cochrane and Sanchez 1992, Mattos and Uhl Sustainable Agriculture and the Environment (1994). Schubart (1999), for example, con- in the Humid Tropics (a commission of the cluded that approximately 90 percent of Ama- National Research Council) sunmmanzed the zonian soils are acidic, chemically poor, and biological limits of agriculture in the tropics: excessively humid. The last characteristic fa- vors the development of insects and plant dis- The hot and humid climate provides ease. Goodland and Irwin (1975) affirmed ideal conditions for pests and diseas- that the hot, humid climate of the Amazon es. The growing season is essentially frequently is associated with high biotic pres- continuous and facilitates the devel- sure and acidic, infertile soils. Smith (1982) opment of persistent pests. Losses of revealed that excessive rainfall can make slash crops to pests in the humid tropics burning (to prepare fields) unviable in many are great. Preharvest losses are esti- humid areas. Even when burns are successful, mated to be 36 percent of yield, and 6 Sustoinabre Amazon: LUmitations and Opportunities for Rural Deveiopment postharvest losses are estimated to be marked dry season exists, losses are greater in 14 percent. The impacts of fungi, and 'drying grains due to pests and diseases. viral, and bacterial pathogens in de- Human and animal health. Health is se- veloping countries have been studied verely affected in areas without a pronounced less than those for insects, but the dry season. The dry season is a positive factor, most comprehensive studies suggest because it restricts the multiplication of en- that losses caused by pathogens are demic diseases and their vectors. about equal to those caused by in- Agriculture: In areas without a defined sects. Weed growth is often so prolif- dry season, burning recently cleared areas is ic and hard to control that it is generally incomplete. Crops such as rice, thought to be the most important beans, and corn require a dry period for ma- cause of yield depression. (National turing and drying, as well as to prevent rot- Research Council 1993). ting. Soy, in particular, requires a dry season d'oe to its vulnerability to pest and disease at- tRack while in the vegetative stage, especially if Recenl Evidence. i the humidity of the air near the soil surface re- Recent scientific discussions have also empha- mains high during a large part of the day. sized the effect of climate on agricultural pro- The use of heavy machinery on a corn- duction in the Brazilian Amazon. First, the mercial scale is viable only where the soil sur- analysis conducted by Win Sombroek (in face is relatively dry during the planting and press) stresses the necessity of a marked dry harvest periods. Hence, in the Amazon, mech- period for agricultural success (particularly for anized planting and harvesting would be lImit- grains). Second, a study recently concluded by ed to those zones that have at least one month Kenneth Chomitz and Timothy Thomas with rainfall below 10 mm (figure 1). (2001) statistically verifies the negative effect In summary, the effect of rainfall is most of high rainfall on agricultural productivity in significant for grains and, especially, soybeans, Amazon. Finally, a recent symposium spon- sored by Embrapa on the production potential of soy beans in the Amazon yielded similar Figure 1. Consecutive months with rainfall less conclusions to those of Sombroek and than 10 mm Chomitz and Thomas. These studies are sum- ( - marized below. -r+ >c a & t Sombroek (2001) Sombroek emphasizes the necessity of a pro- nounced dry season for the establishment of - agriculture. Using a minimum of two consecu- tive months with rainfall lower than 100 mm V as the criterion for defining a dry season, he w concluded: Roads. The construction and mainte- | nance of roads are problematic where there is I m nanih no pronounced dry season. 2 =nth Storage. The construction and mainte- M,1 -= nance of warehouses and silos are more expen- Source. Chomitz and Thomas calculations based sive in humid regions. In these areas, where no on CAMREX data. Effect of Rainfall on the Performance of Agnculture and Ranching in the Amazon 7 and slightly less so for ranching. In the case of Table 2 shows the effect of rainfall on the perennial crops, the effect of rainfall is signifi- stocking rate on a 500-hectare ranch with the cant for crops such as black pepper, but has same characteristics as above.' not been a limiting factor for crops such as oil Consistent with the observations of Som- palm, banana, and coconut. broek, Chomitz and Thomas note that soy represents a large proportion of agricultural Chomitz and Thomas (200 1) production in areas with the following charac- Chomitz and Thomas conducted a statistical teristics: analysis of the 1995-96 agricultural census to * Annual rainfall reaches 1,600 to 2,000 mm. relate agricultural land use (table 1) and pas- * There are 3 to 4 consecutive dry months. ture stocking rate (number of animals per * Limiting soil factors are a low level of hectare of pasture) (table 2) to soil character- phosphorus, nitrogen, and retention of or- istics, precipitation, market access, and histor- ganic matter. ical deforestation. The advantage of this * The basic vegetation is cerrado (savanna multivariate analysis is that it makes it possi- and scrub forest). ble to separate the effect of rainfall from that Curiously, dairy ranching also appears to of other factors, for example, roads, distance be facilitated by a dry climate. Dairy farming to market, and soil characteristics. These au- occurs almost exclusively in areas with annual thors confirmed that, controlling for all of rainfall below 2,200 mm. these factors, precipitation has a significant negative effect on the intensity and type of Embrapa: Conference on Soy (1999) land use. In December 1999, Embrapa-Cpatu sponsored The indpendent effect of rainfall can be a symposium in Belem to discuss the potential observed by considering an example of a typi- and technical limitations of soybean cultiva- cal property in the Amazon with the following tion in the humid Amazon. In his presenta- characteristics: tion, Nelson Ferreira Sampaio, executive * Location: western Para director of Embrapa-Rond6nia, made the fol- * Soils: oxisols lowing observations, which are highly consis- * Distance to primary road: 25 km tent with those of Sombroek and Chomitz and * Distance from the nearest area deforested Thomas: in 1996: between 100 and 200 km 1. Climatic, edaphic, and agronomic factors * Distance from the closest city: 200 km. are fundamental to soy cultivation. A large Table 1 shows the prediction of the effect part of the Amazon is covered by forests of rainfall on land use, holding all other vari- under an intense rainfall regime, with a ables constant. reduced dry season. These conditions Table 1. Effect of rainfall on land in agricultural use Table 2. Effect of rainfall on stocking rate Rainfall (mm) Land in agricultural Rainfall (mm) Stocking use (%) (animallha of pasture) 1,600 22 1,600 .38 2,000 8 2,000 .31 2,300 -0 2,300 .27 Source Chomitz and Thomas 2001. Source: Chomitz and Thomas 2001. 8 Sustainable Amazon: Limitations and Opportunities for Rural Development eliminate the opportunity for large-scale biophysical conditions and agricultural poten- grain production in most Amazonian terri- tial in these zones. The Radam classification tory, simply due to either the presence of of agricultural potential was based on infor- forest or the inability to intensively mecha- mation on climate, soils, relief, geology, and nize operations. vegetation. The Radam Brasil project covered 2. The potential for cultivating grains is an area of approximately 3.7 million km2, found in the natural savannas and grass- equivalent to 74 percent of Amazon. The cer- lands of the Amazon, which occur primari- rado and pantanal regions of Mato Grosso ly in peripheral areas (south of the legal were not included. Amazon and Roraima), where there is a pronounced dry season. Dry Amazon (rainfall below 1,800 mm/year) 3. The forests that cover the majority of the The dry Amazon, receiving less than 1,800 Amazon represent the natural vocation of mm rainfall per year, comprises approximately the region. There is a great need to define 17 percent of the territory. This area is con- areas with economic potential for occupa- centrated in the south of the Amazon basin tion by current and future populations. and in isolated areas of natural grassland lo- cated primarily in the north of Roraima. In this region, climatic conditions are relatively favorable for agriculture. Although soils are The analysis of Chomitz and Thomas (2001) predominantly poor, sparse stretches of fertile based on the 1995-96 Agricultural Census soils exist in Rond6oia, Para, and Mato Gros- data (IBGE) enabled the authors to identify so. Soils are generally well drained, and the re- three rainfall zones in the Amazon with dis- lief relatively favorable for mechanized tinct differences in agricultural performance agriculture. The vegetation is largely savanna (figure 2). The authors used data from the with some sparse areas of open and semi-de- Radam Brasil project (1973-1978) to describe ciduous forests. These forests contain low vol- umes of commercially valuable timber species. Transitional Amazon (rainfall between 1,800 Figure 2. Principal rainfall zones In the Amazon and 2,200 mm/year) +- $, $ ",,,;,,,>^ i ly 38 percent of the Amazon and is located largely in the transition belt between the cen- tral region (humid zone) and the arc of the l R ,Amazon deforestation in the south (dry zone). This transitional region is generally covered . w - _ - < by dense terra firma forest, with areas of open forest in Mato Grosso and southern Para. In general, soils are chemically poor (although patches of fertile soil do exist) and relatively well drained. The topography is largely rolling with significant elevation in Roraima and the Tmnsmal m1,8OO-2,2oo0,m/yce north of Para. In addition, isolated higher ele- _ Hum.d >2,200 ../y-R HNod /,a mmIyOC vation areas are found in the center (Carajas) Source: Chomitz and Thomas calculations based and south (Cachimbo) of Para and the center on CAMREX data. of Mato Grosso (Parecis). Effect of Rainfall on the Performance of Agriculture and Ranching in the Amazon 9 Excess rainfall and a short dry season cre- markets. For example, perennial crops (pri- ate severe agronomic and economic difficulties marily black pepper, malva, oil palm, passion for grain production in this zone. Perennial fruit, oranges, papaya) are cultivated in the crops have had somewhat better agricultural outskirts of Bel6m, where good infrastructure success, although diseases such as leaf plague and market conditions predominate (Serrao (Microcyclus ulei), which attacks rubber trees; and Homma 1993). However even these initi- witches broom (Crinipellis perniciosa); infes- atives face a difficult battle with disease and tations of Cacao fusarium (Fusarium solani f. pests. sp. Piperis) affecting black pepper; and the Combining the three rainfall zones with deadly yellowing (cause unknown) that dam- the data from Radam Brasil (study area 3.7 ages oil palm have greatly restricted their eco- million kin2), the authors found that 84 per- nomic viability. However, small landowners cent of the area possessed high or medium po- who employ diversified agricultural systems tential for timber extraction. In contrast, have achieved a reasonable increase in stan- Radam Brasil concluded that only 7 percent dard of living (Schneider 1994, Moran 1989, (approximately 0.25 million kni2) shows agri- Oz6rio de Almeida 1992, Jones and others cultural promise, whereas 93 percent of the 1992, and Toniolo and Uhl 1994). land presents either low or no agricultural In the case of ranching, Mattos and Uhl potential. (1994) documented the relative success of in- tensive cattle ranching in Paragominas, east- ern Para. Calculations made by the authors Agricultural Performance (appendix 1) demonstrate that reasonable in Relation to Rainfall economic returns to cattle ranching occur The humid Amazon is perhaps the area of the only under relatively advanced technological globe with the highest probability of shelter- conditions. ing a natural predator for any crop introduced by human beings. This implies that modifica- Humid Amazon (annual rainfall greater than tions introduced to control one pest have a 2,200 mm) high probability of rendering the crop vulner- In this zone, annual rainfall exceeds 2,200 able to a different predator. The history of ag- mm, with some areas receiving levels as high ricultural failures in the Amazon is instructive as 4,000-4,500 mm. In general, soils are infer- (box 1). tile and poorly drained. In areas of relatively There is a certain risk in generalizing high relief, intense rains increase the risk of from successes or failures of agricultural ex- erosion. This zone, comprising 45 percent of periments without either a long period or a the Amazon, is located primarily in the cen- wide range of experience to observe the effect tral region, occupying a large part of Amazo- of varying production factors. Nevertheless, nas and Amapa states; the northeast of the results of Chomitz and Thomas (2001), re- Rond6nia; and the southeast, northwest, and viewed above, which control statistically for northeast (the island of Maraj6 and the many of these factors, enable the authors to Bragantine region) of Para. The majority of generalize with some confidence about the this area is covered by dense forest. The ad- negative effects of high levels of rainfall on ag- verse natural environmental conditions (ex- ricultural productivity. A review of the most cess rainfall and poorly drained soils) render recent census data provides an additional test virtually all forms of agriculture economically of the validity of these generalizations. uncompetitive. Profitable activities occur only The most recent information pertaining in areas well endowed with infrastructure and to land use in the Amazon comes from the 10 Sustainable Amazon: Limitations and Opportunities for Rurol Development Box 1. Agricultural failures in the Amazon Bragantina (Par6). Attempts to transform the grains ended in failure. Only perennial crops (par- Amazon into a vast area of agricultural produc- ticularly fruit trees) show economic potential. tion began at the beginning of the twentieth Ranching. In Acre, approximately 550,000 hectares century. During this period, the federal govern- (ha) of pasture are in an advanced state of degra- ment supported the agricultural occupation of dation. A large majority of pasture is formed by the Bragantine Region, northeastern Par6. In Brachiario bizontha, a gross that is intolerant of over 100 years of agricultural experimentation, acidic and poorly drained soils. This type of soil is almost all crops failed. Excessive rainfall (>2,200 distributed over vast areas of the Amazon (ap- mm) and a short dry season made vegetable proximately 20 percent of the territory), including and grain cultivation unviable Perennial crops, Acre and southern Amazonas. such as black pepper, also failed due to dis- ease (fusariose). Today, the regional landscape Soybeans. At the end of the 1990s, the govern- is dominated by degraded and abandoned ar- ment of Amazonas stimulated planting of soy in eas, extensive cattle ranching, slash-and-burn Humaita, in the south of the state. Despite financial agriculture, and isolated crops (passion fruit, incentives the initiative failed. The drenched soils papaya. acerola, black pepper, oil palm) and high rainfall rendered cultivation unprofitable. Similar problems occurred in Santarem (Para) with Perimetral norte (Amapd). The recent occupation an experimental planting by the Quinico group. of an extensive area of northeast Amapd for Due to excessive rains, it was impossible to harvest agricultural reform resulted in failure Poor soils and a third of the 600 ha that had been planted. excessive rainfall made grain cultivation unprofit- able. Despite infrastructural support (paved roads, Long-cycle perennial crops. Expenments with ho- electricity, brick houses), the majority of lots are faled In the cse ofrnd brazil-nut plantations abandoned. Mlcrocyclus uley, facilitated by the high humidity, Transamazonica (Para). Excessive rainfall results in is a limiting factor that remains insurmountable. The prohibitively high costs for road building and main- productivity of Brazil nut in open field conditions tenance. In over three decades of occupation, (Itacoatiara, Amazonas) was significantly lower farmers have faced enormous natural (rainfall and than in the experiments conducted by Embrapa. humidity) and infrastructural (roads) challenges to This failure discouraged Brazil nut planting in other agricultural development. All attempts to cultivate parts of the Amazon (Dean 1989). 1995-96 Agricultural Census (IBGE). This in- data to analyze the relationship between rain- formation can be superimposed on rainfall fall patterns and land use in the Amazon. Figure 3. Areas in agricultural use (percent) Figure 4. Areas in agricultural use (percent) In relation to rainfall: Within 25 km of a main road In relation to rainfall: More than 25 km from a main road 50. 40 5, 30 _ e . 1l ~~~~~~~~~~~~~~~40 1 30 30 20 ...... 5_|L; .....20 10 0~~~~~~~~~~~~~~~~~~~~~~~~1 1200 1400 1600 1800 2000 2200 2400 2600 2800 3000 3200 3400 3600 0 1200 1400 1600 1800 2000 2200 2400 2600 2800 3000 3200 3400 3600 Annual rainfall (mm) Annual rainfall (mm) -Agricultural use - -Ranching Abandoned -Agricultural use - -Ranching - Abandoned Source: Authors' calculations based on IBGE and CAMREX Source: Authors' calculations based on IBGE and CAMREX data provided in Chomitz and Thomas 2001. data provided in Chomitz and Thomas 2001. Effect of Rainfall on the Performance of Agriculture and Ranching in the Amazon 11 Figures 3 and 4 are based on land-use data at The low agricultural productivity in the level of census unit (the finest scale of the the Amazon frequently is blamed on poor census analysis) and on rainfall data interpo- infrastructure and absence of markets, rather lated between rainfall measurements of than on climatic factors. The spike of the curve ANEEL (National Electric Energy Agency) of agricultural use at 2,900 mm of rainfall and those of the U. S. National Aeronautics shows the limitation of good transport and and Space Administration (NASA) Earth- market systems in overcoming the negative ef- Observing System (EOS) Amazon project of fect of excessive rainfall. This spike represents the University of Washington. As would be largely the humid zone microregions of Belem- predicted by Radam Brasil, these data demon- Braganca (Para) and Macapai (Amapa), situat- strate a severe reduction in agricultural land ed 25 km from a main road with easy access to area as rainfall levels increase. large urban markets. Nevertheless, the zone To isolate the effect of distance from the displays low land use and a high proportion of property to the primary road, the authors abandoned land.2 To better understand the dy- considered two scenarios, represented by fig- namics of land use and rainfall, the authors ures 3 and 4. Figure 3 includes only census ar- classify the census data into the three rainfall eas that were less than 25 km from a primary categories discussed above. These zones corre- road. Figure 4 displays results for census ar- spond to 17 percent (dry), 38 percent (transi- eas more than 25 km from roads. Interesting- tional), and 45 percent (humid) of the Amazon. ly, both figures show the same abrupt drop in Table 3 presents the proportions of the the percentage of census area in agricultural census areas formally designated as being in use in areas with rainfall greater than 1,800 farms (agricultural establishments) and the mm per year. Below 1,800 mm, both figures percentage actually converted to agricultural show that 30 percent to 40 percent of the use, including abandoned lands.3 The effect of land is designated agricultural. This number rainfall is evident: 55.6 percent of land in the falls abruptly to 5 percent as annual rainfall dry zone is in agricultural establishments, and rises to 2,200 mm. In areas farther than 25 38.2 percent is in agricultural use, while, in km from roads, the agricultural land covers the humid zone, these proportions reach only no more than 5 percent. In the case of land 7.5 percent and 3.2 percent, respectively.4 closer to roads, agricultural use increases to Table 4 displays the types of agricultural approximately 23 percent at 2,800-3,000 mm land use in each rainfall zone. Approximately rainfall, then drops rapidly to almost 83 percent of land in agricultural use in the dry 0 percent. zone is under pasture, and approximately Table 3. Land use in the Amazon by rainfall zone Rainfall z,,ne A! A4rea Dkn"I o t ,;fa l i % of zone ifn jreo oj lu s hents agriculturol use Drya 3 572jf ~ I76 38.2 Transiional o- 1;8 1 640 3 28 13.0 Humicd 1 , A- 45 3.2 Total if 44840 , 100 13.0 a. The rainfall categories correspond to less than 1,800 mm (dry); 1,800-2,200 mm (transitional); and more than 2,200 mm (humid). b. Agricultural establishment is a category used in the Brazilian census. It refers to privately held land in both agricultural and forest use. Source: Authors' calculations based on Chomitz and Thomas 2001. 12 Sustainable Amazon. Limitations and Opportunities for Rural Development Table 4. Use of agricultural areas In the Amazon (percent) Rainfacll V nriG Perennial zone r u Pasture crops Others Dry° 83.3 - 0.5 2.6b Transitional 77.7 _9 P7.9 3.6 Humid 56.8 E '7j2 4.4 10.7 Humid with old colonizationc h® - 54.4 4.6 6.7 a. The rainfall categories corresponds to less than 1,800 mm (dry); 1,800-2,200 mm (transitional); and more than 2,200 mm (humid). b. Includes plantation forests and settlements. c Corresponds to Bel6m-Bragantmna (Para) and Macapa-Mazagao (Amapa). This area is humid but possesses relatively good infra- structure and market conditions. Note: Area in agricultural use as a percentage of total land in rainfall area is as follows: dry: 38%; transitional: 13%: humid: 3%. Source: Authors' calculations based on Chomitz and Thomas 2001. 8 percent is abandoned. Pasture falls to consistent with the modest success found in roughly 60 percent in the humid zone, with a Schneider's review of colonization studies in concomitant increase in abandoned land to ap- the dry and transitional areas (1994), based proximately 20 percent. Relatively good infra- largely on work of Oz6rio de Almeida structure and access to markets do not appear (1992), Emilio Moran (1989), FAO/UNDP/ to improve the economic sustainability MARA (1992), and Jones and others (1992). of agriculture in this zone (box 1, Bragantine However, data discussed above demonstrate Region). that even this moderate success experienced The data presented above highlight the in dry and transitional areas cannot be ex- dangers in generalizing the modest success pected for the 45 percent of the Amazon with observed to date in Amazonian agriculture to rainfall above 2,200 mm. For these areas, the future perspectives. The relatively high con- most probable scenario is low agricultural version rates observed for the dry zone in the productivity, weak economic performance, 1995-96 Agricultural Census (IBGE) are and eventual land abandonment. 2 "Boom-and-Bust" or Economic Sustainability: The Community's Dilemma B ased on the land-use data and literature Internal Rate of Return (IRR) of forest man- values assembled for this report (appen- agement was estimated at 71 percent,' while dix 1), it is possible to predict the eco- for the predatory logging system, IRR reaches nomic future of a typical Amazonian county 122 percent (appendix 1). in the humid zone (45 percent of the region). Ranching If market forces operate freely in the region, land use will be based largely on predatory Cattle ranching is the dominant land-use ac- logging associated with extensive ranching. In tivity in deforested areas, representing 77 per- this case, the local economies will tend to fol- cent of converted areas in economic use low a "boom-and-bust" cycle. Rapid growth (Chomitz and Thomas 2001). The current in the first years (boom) will be followed by Amazonian herd is estimated at 32 million a severe decline in profit and employment head. The average stocking rate is only 0.7 an- (collapse). imals per ha. Ranching generates roughly 118,000 permanent jobs. In general, ranching presents a very low IRR (4.2 percent), in iso- lated cases (ranching in reformed pastures) The major uses of land in the Amazon are tim- achieving rates up to 13 percent (appendix 1). ber harvesting and ranching. rimber Harvesting The Community's Dilemma The legal Amazon produces approximately 90 The authors will consider a county containing percent of the native wood in Brazil. The 1 million ha of dense forest.7 Migrant loggers wood industry is the primary economic activi- begin to arrive in the area in search of new tim- ty in the Amazon, representing roughly 15 ber stocks. The community discusses whether percent of gross domestic product (GDP) of to try to control the influx of loggers and insti- the states of Para, Mato Grosso, and tute a sustainable management system, or to Rond6nia. In 1998, the gross receipts of the permit predatory logging of the forest with sector were estimated at US$2.5 billion.S subsequent market-driven conversion to pasture. Moreover, the wood industry generates ap- If market forces act freely in the region, proximately 500,000 direct and indirect jobs the community can anticipate rapid growth (Verissimo and others unpublished). The followed by a severe decline, as illustrated in 13 14 Sustainable Amazon: Limitations and Opportunities for Rural Development figure 5. Economic activity grows relatively Implications for employment are equally rapidly in the first eight years as trees are ex- dramatic. Both models employ roughly the tracted and pastures formed. However, after same number of workers during the first eight eight years, when the supp]y of high-value years. After this period, the predatory rnodel trees has been depleted and a second round of peaks at 4,500 jobs in timber harvesting and logging (focused on low-value species) begins, ranching combined, while the sustainable the economy starts to decline. By approxi- model, based only on forest management, mately the twentieth year, marketable wood is would remain stable, at 3,500 jobs (figure 6). completely exhausted, and the local economy However with depletion of marketable wood enters a crisis. in the twenty-third year in the predatory mod- Economic activity, measured by the gross el, the employment base migrates to another revenue from timber harvesting (extraction county, leaving behind fewer than 500 work- and processing) and ranching in this boom- ers involved in ranching. If the community's and-bust cycle, reaches a maximum of timber resource were managed sustainably, the US$100 million in the eighth year and falls to 3,500 jobs would be maintained indefinitely. below US$5 million in the twenty-third year. In areas of extremely good infrastructure During this period, loggers will abandon the and markets, the long-run employment ad- county, leaving behind only low-productivity vantage of managed forestry compared to ranching. If the communLty were to compel business as usual is less pronounced, because the logging companies to adopt sustainable the market-led conversion of land to ranching forest management, the gross revenue would is considerably greater. Nevertheless, even in reach US$70 million in the eighth year, instead these favorable areas, authors' calculations of the US$100 million obtained by the preda- show that direct employment drops by more tory model. However, revenue would be sus- than half (from year 21 to year 24), and the tainable indefinitely at this level, instead of economic base collapses. dropping drastically with the exhaustion of The above analysis indicates clearly that timber supplies in the twenty-third year, as oc- from the long-run perspective of stable growth, curs in the predatory model. Figure 6. Jobs: Managed forest vs. predatory Figure 5. Gross revenue: Managed forest vs. logging and ranching with poor infrastructure predatory logging and ranching with poor in the humid Amazon Infrastructure in the humid Amazon .0 3>500 = 30 I N 1 3 5 7 9 11 13 15 17 19 21 23 25 27 29 I 3 5 7 9 11 13 15 17 19 21 23 25 27 29 Years since initiation of colonization Years since initiation of colonization Forest management - -Predatory - Forest Management - -Predatory Source: Authors' calculaticns, appendix 1. Source: Authors' calculations., appendix 1. "Boom-and-Bust" or Economic Sustainobility: The Community's Dilemma 15 building community, and investing in people, the expansion of the agricultural frontier. the managed forest offers a better alternative Although there are several successful forest than uncontrolled market exploitation. management initiatives, such ventures still Figures 5 and 6 compare the effect of represent a small fraction of logging activity managed forest and market-led exploitation (less than 5 percent of the volume extracted). on economic activity in the community. The dominant pattern continues to be preda- Figure 7 illustrates the relationship between tory logging, characterized by severe damage the private and social benefits of an unsus- to the forest, excessive pressure on popula- tainable "boom-bust" pattern of develop- tions of high-value species, and increased sus- ment. Figure 7 displays a boom period, as the ceptibility of harvested areas to fire (Uhl and new activity attracts migrants and provides others 1997). the economic base for growth of the private Predatory logging has already exhausted and public sectors. The collapse begins when forest resources in old logging centers. The cir- the natural resource base that sustains this cled areas in figure 8 display the old logging activity (in this case, timber) is depleted and frontiers: Paragommas (Par6), Sinop (Mato the economy attains a new equilibrium based Grosso), Vilhena - Ji Parand - Ariquemes on extensive ranching. Although in this stage, (Rond6nia); frontiers of intermediate age, economic benefits still exist, costs are elevated such as the north of Mato Grosso and Tailan- due to unemployment and migration of dia - Maraba (Para); and the new frontiers: companies and people, and reductions in Novo Progresso (Para), Novo Aripuana-Apui public services resulting from loss of the (Amazonas) and Senador Jose Porfirio - Portel economic base of the community. In addition, (Para). The authors estimate that the scarcity the environmental costs, in the form of of wood in the old frontiers will require log- biodiversity loss and carbon emissions, are ging companies to either migrate or close increased. within five years. In the intermediate frontiers, natural timber stocks are sufficient for 10 to Logging Dynamic 20 additional years, whereas in the new areas, Most timber harvesting in the Amazon has oc- curred as a complement to agriculture. As a result, the logging frontier has accompanied Figure 8. Migration of logging In the Amazon Figure 7. Social costs and private benefits of the "boom-and-bust" economy Ph-, I Pha.. 11 Ph..b III Croih D.d- ColhP.. 0 Sawm,iI centers Porential National Forest .o /, iligration of sawidlis Source' Authors. Source. Verissimo and others 2000. 16 Sustainable Amazon: Limitations and Opportunities for Rural Development timber supplies will be depleted in 30 to 40 In Sinop, the number of sawmills fell years. from approximately 400 at the end of the Loggers in Rond6nia are relocating in the 1980s to fewer than 100 at the end of the direction of Bolivia and the state of Amazonas, 1990s. However, in this county, the decline in while companies from the old frontiers in Para the timber sector has been largely compensat- and Mato Grosso are migrating to the most ed by the rapid growth of agriculture, princi- recent frontiers (west of Para and southeast pally ranching and soybean cultivation. The of Amazonas), frequently illegally logging in- greater agricultural potential of the dry Ama- digenous and protected areas (figure 8). zon makes this transition possible. On the other hand, Paragominas, situat- Migration of Logging Companies ed on the border between the transitional and and Its Effect on Local Economies humid zones, has displayed a slower rate of The current logging model has a strong im- forest decline due to its denser forests, which pact on the economies of communities in the provide greater commercial volume per ha. Amazon. Following the expansion period, the Nevertheless, despite a relatively long history consequent exhaustion of resources results in of land-use experimentation, a consistently an inevitable economic recession in the local lucrative form of agriculture that is capable economy. The gravity of this recession de- of maintaining the vitality of the local econo- pends on the local agricultural potential, that my has not emerged.8 is, on the extent to which the emerging agri- cultural economy can replace the loss of the timber economy. For example, Paragominas, Summary the oldest logging frontier in Amazon, estab- As the frontier moves from the dry zone into lished at the end of the 1970s, is confronting to the transitional zone, and especially to the a grave shortage of primary material due to interior of humid areas, agricultural perfor- the exhaustion of its forests. In the past 5 mance will decline. If market forces are not years, approximately 50 sawmills closed or restrained, communities constructed in the migrated, and the volume of wood processed latter areas during the logging "boom" will fell approximately 30 percent. A similar phe- become increasingly depressed during the nomenon can be witnessed in the county of subsequent "bust" to an ever weaker agricul- Sinop (Mato Grosso), one of the largest log- tural base. Unfortunately, local decisionmak- ging centers in the 1980s, and in Reden,co ers often have little incentive to adopt a (south of Para). However, due to the fact that sustainable development model. Therefore, Sinop and Redencao are located in areas of below the authors analyze the potential for a open forest (characterized by low density of national policy that prevents this decay of marketable timber) situated in the dry zone, communities and contributes to sustainable the decline in timber harvesting has been development with a better distribution of more rapid than in Paragominas. benefits. 3 Role of Government T he Brazilian government has a crucial long-term interests of the community they role in determining the quality of dev- are representing. elopment in the Amazon and in pro- 3. Most important, even if a community were tecting the interests of the larger Brazilian so- to opt for the sustainable model, it would ciety. Government policy should reconcile (1) have great difficulty in attracting loggers the short- and long-term interests of society; under conditions of good forest manage- and (2) the interests of the various segments of ment practices. The Internal Rate of Re- society, which involve local, state, national turn for conventional harvesting and and global levels. processing of timber (predatory model) is 122 percent, while the sustainable system (forest management) achieves a maximum Stabilize Local and Regional Economies of 71 percent (appendix 1). When faced In frontier areas, land-use instability is caused with the choice between a community that primarily by economic forces. Therefore, the enforces forest management and one that creation of a political coalition capable of pro- permits conventional logging, loggers natu- moting more orderly development is a difficult rally tend to opt to work in areas that per- task. Local and regional interests support rap- mit the more lucrative system. id development (in general, unsustainable), while national and global interests support the benefits of a slower, sustainable economic Avoid Shortsightedness growth (Schneider 1994). of Local Governments The case of "the commlnity's dilemma" The problem of the tendency for local govern- (chapter 2) is instructive. In the absence of ments to be shortsighted and "captured" by government intervention, a community must local economic interests has been recognized choose between the boom-and-bust model worldwide. In the literature, the phenomenon (predatory) and sustainable development. A whereby local governments mortgage their fu- community has at least three reasons to tures for short-term benefits is called "regula- choose the boom-and-bust model. tory competition." This competition involves 1. The short period of municipal mandates a range of unhealthy practices, from negligent does not allow political leaders to adopt environmental and social regulations to subsi- long-term perspectives with the objective of dies and tax breaks offered by local govern- stabilizing and improving quality of life. ments to attract industries and corporations. 2. Many political leaders are personally in- These practices, guided by short-term political volved in the predatory natural resource and economic benefits, frequently threaten the "mining" economy and do not consider the long-run financial health of the community. In 17 18 Sustainable Amazon: Limitations and Opportunities for Rural Development the long term, such practices are unsustainable percent of private property in the Amazon be for all levels of society. maintained as a legal reserve; and, if it is logged, forest management must be used. The regulations of the Environmental Crimes Law (Lei de crimes ambientais) offer an opportuni- Various alternatives exist to stimulate sustain- ty to make fully effective the existing forest able forest resource use and thereby stabilize legislation. It is important that a division of the local economy. These alternatives are sum- responsibilities between local and federal gov- marized below. ernments be worked out to ensure an efficient and rigorous monitoring and control system. 1. Increase Profitability of Forest Management 4. Organize Regional Occupation Revenue from forest management would be The federal government should take the lead augmented if, for example, a market existed in altering the dynamic of disorganized terri- for the environmental services provided by torial occupation in the Amazon. This pattern forests. One of these services is the retention of colonization catalyzes deforestation, preda- of carbon. Carbon retention contributes to the tory logging, and fragmentation. One promis- equilibrium of the global climate. Payment for ing means of regulating this occupation is to this service has been the subject of interna- expand and consolidate a network of public tional debate, but no decision has yet been forests (national and state forests) in the Ama- made (box 2). The authors calculate that a re- zon (chapter 4). muneration of between US$2 and US$3 per 5 A ton of additional carbon sequestered in forest . dopt Compensatory Measures management, compared with that sequestered The federal government could adopt compen- under predatory logging, would be sufficient satory measures to (1) increase political sup- to induce loggers to adopt forest management port for initiatives that reduce unsustainable practices (appendix 3).9 short-term growth of the local economy and (2) avoid regulatory competition among coun- 2. Reduce Com rive ties. These initiatives might include improving of Predatory Logging local public services such as sanitation, educa- Refining the command and control system tion, and healthcare. Companies often prefer combined with adopting a tax on the value of to install their operations in counties endowed timber derived from predatory operations with good healthcare systems, quality schools, could reduce the advantage of predatory log- cultural and recreation options, low crime ging. The authors estimate that if a tax on rates, and healthy environments (low pollu- wood from predatory harvests remained be- tion, green spaces, clean water), instead of in tween US$1 and US$4 per m3 (depending of districts whose only attractions are financial the discount rate), logging companies would incentives and negligent regulations. have no incentive to migrate from the county to avoid management restrictions (appendix 3). Distribute Benefits from Use 3. Engender Respect for Forest Law of National Patrimony It is critical that respect for forest legislation, The government should guarantee that the in particular, the national Forest Code (C6di- benefits from both logging and the conversion ga Florestal), be ensured. Among other re- to agriculture of Brazilian forests are properly quirements, this legislation mandates that 80 distributed within Brazilian society. The Role of Government 19 Box 2. Clean development mechanism In the logging sector also, benefits are dis- tributed inequitably. Logging companies pay Significant alterations in world climate are being less than 20 percent of the income tax (ICMS) observed. The accumulation of "greenhouse gases," especially carbon dioxide emitted that is due, and most are exempt from pay- through fossil fuel burning since the industrial ment of income tax due to regional incentive revolution, provokes a warming of the earth's programs. In addition, loggers do not pay for atmosphere and, secondarily, a rise in sea level, In 1990, governments initiated discussions con- their use of federal lands. For example, in Bra- cerning what measures should be taken to zil, no fee is charged for standing timber ex- combat global warming. tracted from public lands. Charging such fees Brazil has had an important role in these discus- is a common procedure in national forests in sions. For example, during the 1992 Rio Earth most other countries. Summit (Rio-92), a Convention on Climate Change was approved, and its first signatory For economic, social, and environmental was Brazil's then-President Fernando Collor. In reasons, the Brazilian government's coloniza- addition, in 1997, the official Brazilian delega- tion policy-applicable in areas with agricul- tion played an important role in making the Kyoto Protocol viable. Brazil has continued to tural potential-favors small landholders. be one of the key Parties in the negotiations to However, it does not discourage the acqulsi- implement the Protocol, in particular, the clean tion of large properties. The policy for lands development mechanism (CDM). The CDM en- ables industrialized countries to achieve part of with agricultural potential should be changed their Kyoto Protocol emissions reduction targets to discourage speculation because: by financing projects that reduce or sequester 1. The acquisition of such large areas encour- carbon In developing countries. carbon in eveloping ountries.ages predatory activities conducted merely The Sixth Conference of the Parties of the Con- arator tivities oneds mrel vention on Climate Change in July 2001 agreed to guarantee the right of ownership. on a compromise that makes afforestation and 2. Large properties that are unproductive, or reforestation eligible for the CDM. In November abandoned, encourage land speculation 2001, the Seventh Conference of the Parties agreed on a package of rules and modalities and subsequent ownership conflcts. for implementing the CDM Other types of 3. The distribution of quality land to small projects that result in carbon sequestration holders improves the quality of life for through alterations in land use and forest man- low-income populations and contributes to agement may be considered by the Parties for subsequent commitment periods. the equitable distribution of national re- It has been estimated that the global volume of sources. of trade in CO2 (without the United States) could reach US$5 billion to $10 billion annually from 2008 to 2012, a part of which may go to Preserve Options for Future Economic Use developing countries to invest in CDM projects. of the Amazon Thus, the CDM would make viable a series of af- forestation and reforestation projects. Brazil In the future, technological changes may gen- could expect to attract 5 percent tolO percent erate new economic benefits for the humid of the market. tropics, such as genetic engineering based on existing biodiversity, non-timber forest prod- current situation reveals a series of social ucts, ecotourism, sustainable forest manage- disparities. On one hand, the 1 percent of ment, and ecological agriculture, that exceed largest landowners (who own more than today's values. By preventing forest degrada- 2,000 ha) hold 47 percent of the agricultural tion, the government would be preserving this land. At the other extreme, the 54 percent of ecosystem for possible future economic use. smallest landowners (who own fewer than 20 This argument justifies the designation of ha) represent only 1.1 percent of the agricul- conservation and sustainable use areas, for tural land. example, national forests and extractive 20 Sustainable Amazon: Limitations and Opportunities for Rural Development reserves, as a mechanism to maintain the op- role of government in this process, with 41 tion of employing the highest-value land uses percent of logging companies preferring a sys- in the future. tem in which the government merely granted concessions and companies were responsible Protect Blodiversity for management, and 56 percent preferring that the government be responsible for man- The Brazilian Amazon supports the richest agement. Similar percentages were obtained biodiversity and source of genetic information among nongovernmental organizations in the world. The Brazilian government is (NGOs), academic institutions, and profes- committed to do its part to help preserve this sionals outside the forestry sector. heritage for future generations. National For- The migration of companies in the Ama- ests could perform a key role within a mosaic zon has already begun (figure 8). Neither the of Conservation Areas, acting as a buffer zone loggers nor the receiving communities are between protected areas and private lands. comfortable with the current process. There- fore, the window of opportunity to initiate the transition to an industry based on forest man- Help Stabilize the Wood Sectoragenisow agement iS now. The depletion of timber resources in the oldest processing centers has led to the migration of sawmills to frontier areas. This migration will Organize Regional Occupation intensify in the next five years. Disorganized To ensure conservation and sustainable re- migration results in irregular occupation of source use, government intervention to con- untitled lands, conflicts with indigenous trol market forces is necessary. As "The groups, predatory logging, deforestation, and Community's Dilemma" (chapter 2) illus- extensive ranching. In this scenario, the wood trates, in the absence of payments for envi- industry is the catalyst of a boom-and-bust oc- ronmental services, the federal government cupation process. should prevent regulatory competition. In However, the majority of logging compa- other words, the government should prevent nies would prefer to function within a system states and counties from competing to attract of greater stability and certainty that would predatory logging and ranching industries include defined regulations, secure land own- with the goal of obtaining the economic ership, and sustainable timber stocks. Recent "boom," despite the inevitability of the long- research with 96 timber companies (Barreto term collapse. Preventing regulatory competi- and Arima 2000) revealed that a large majori- tion entails both good forestry legislation and ty (80 percent) of business owners want to ex- effective enforcement. The federal govern- change the current disorganized process for ment should discourage states and counties managed harvesting based on a forest conces- from using tenuous rules for monitoring and sion system (and the creation of National For- control to attract unsustainable and transito- ests) (chapter 4).1o Opinions differed on the ry investments. 4 Role of National Forests I nvolving the federal government in pre- standards, such as those of the international vention of regulatory competition is in Forest Stewardship Council (FSC). the interests of all parties. First, by insist- In forests on private lands, the govern- ing that long-term interests be considered, ment should encourage the adoption of forest the government fortifies the authority of management, including through the develop- communities. Second, responsible logging ment of a system to effectively monitor the ex- companies would be protected by law from ecution of forest management projects. If the unjust competition from other companies monitoring system is not federal, the that harvest wood illegally. Finally, greater government should periodically evaluate the formality in the industry would guarantee integrity of the state or local system. To stim- a greater contribution to state and local ulate forest management, it is essential that a revenue. tax be imposed on wood deriving from defor- The stabilization of the wood sector will estation permits. This tax must be at least require the adoption of forest management equal to the difference in cost between sus- in both public and private areas. The mining tainably and unsustainably produced timber. of resources from forests on private land in Such a tax would eliminate the unfair compet- old logging centers (Sinop, Paragominas, Ji- itive advantage of wood derived from preda- Parana) has impelled the migration of tory operations. sawmills to unoccupied lands in the west of Para and southeast of Amazonas. In these re- gions, the government can act now to avoid National Forests continued predatory resource use and the National and state forests today represent a alienation of public lands. The most promis- modest portion (83,000 km2, or 1.6 percent) ing mechanism to do so is the creation of of the Amazon region. This area would be suf- National Forests. ficient to sustainably supply only 10 percent of National, state, or municipal forests are the current demand for unsawn timber in the sustainable-use conservation units whose pur- region. To satisfy the present and near-future pose is to produce goods (timber, non-timber demand in a sustainable fashion, the govern- forest products) and maintain environmental ment would need to designate approximately services. The federal government can directly 700,000 km2, or 14 percent of the Amazon, manage these forests or temporarily concede for the creation of National Forests." forest-use rights to private companies (Verissi- Verissimo and others (2000) developed mo and others 2000). In either case, forest criteria to identify areas with potential for des- management in these forests should be certi- ignation as National Forests. The principal re- fied in accord with recognized international sults from their study follow. 21 22 Sustainable Amazon: Limitations and Opportunities for Rural Development Absence of Competitive Use Figure 10. Areas with potential for National Forests To reduce possible land-use conflicts, the new ..- National Forests should be established in ar- I > r ' ®,a l eas that have minimal competitive uses, avoid- ing zones of agricultural occupation and protected areas. Protected areas represent approximately mu 1.4 million km2, or 28 percent of the Amazon, of which 1 million km2, or more than 20 per- cent, are indigenous lands (Verissimo and oth- ers 2000). Settled areas were mapped based . on "hot pixels" (from the thermal channel of the NOAA AVHRR sensor),'2 government agrarian reform settlements, and county seats. _ Ipotenttal The superimposition of these maps reveals ornopotJtital sm.(M that approximately 9 percent of unprotected Source: Verissimo and others 2000. forested areas possess detectable human occu- pation (figure 9). (2) forest cover and marketable timber, (3) Economic Potential low human occupation, and (4) location with- Figure 10 displays the results of an analysis of in the radius of economic accessibility. forestry potential (vegetation map) and eco- nomic accessibility. The result is an area of Biodiversity 1.15 million km2 (23 percent of the Amazon) To protect areas of high biological signifi- that could be designated for forest manage- cance, it is desirable for the federal govern- ment, possessing the following combination ment to create a mosaic of conservation areas of characteristics: (1) no official protection, Figure 11. Priority areas for conservation Figure 9. Occupation of the Amozonian forest of blodiversity __ ~ \ i g , 0= Ytotal good infrastructure and market access I k t (Bragantine region, Para), this value increased to 13 percent (table A1-2). Figure Al -2. Aging processing capacity The authors considered that pasture of logging center under forest management planting would occur shortly after logging medium-value species (between years 4 and .0 - a__ ___ 1 Table Al -2. Pasture planted per rainfall zone and average stocking rate of pastures o _4 /* Percent of totol --si.s Rainfall zone pasture area o 0 Dry 31.82 > Transitional 10.10 l .. ]1 ~~~~ ~ ~ ~~~Humid 1.82 i 2 4 6 8 10 12 14 16 18 20 22 24 26 28 30 Humid Years since InitHatlon of colonization (good infrastructure) 13.00 . Source: Authors' calculations above. Source: Chomitz and Thomas 2001. 34 Sustainable Amazon: Limitations and Opportunities for Rural Development 22). At each period, an area proportional to the herd of the pasture area according to the the percentage of pasture in each region was sum of the stocking rate in each cell. Stocking converted following logging. In this fashion, pastures in each cell varied according to the at the end of the twentieth year, the pasture- age of the pasture. The average stocking rate land in the dry region corresponded to during the period was 0.67 animals per ha (ta- 31.8 percent of the total land area. The same ble A1-3). For example, pasture with 2 years procedure was adopted to calculate pasture of use supported a density of 2 animals per area in the transitional and humid zones ha, while pastures with 12 years of use sup- (figure A1-3). ported only 0.2 animals per ha. In the humid zone, a constant pasture-stocking rate of 0.4 AP,, = M, 'r animals per ha was used, consistent with P Chomitz and Thomas (2001). After 12 years t = 4 .. 22 of use, the present authors assumed that pas- tures would be replanted and the cycle would where recommence. AP = area converted to pasture in year t Herd, = YL,,,C, in region r Mt= area logged for timber of medium where value in year t y= proportion of pasture in region r Herd = number of animals in year t p = proportion of the total area pos- N = total number of cells in pasture in sessing medium-value timber in year t the entire period (0.788). L = stocking of pasture of age a in cell i (animals/kmi2) The number of cattle was determined as C = area in cell i (km 2) follows: for each year, the authors calculated t = 4 .. 30. Figure Al -3. Aging of area In pasture by rainfall zone Table Al -3. Stocking pastures 3500 . Pasture age &teekin§ 30007 ( .s years) X E 2500 . . M E 2 @2O -/30 ,, 1500 4-- -- o - 1000 .. / o' o D _/o°_ -s6 Sg O5_0 7 8 ). 1 3 5 7 9 11 13 IS 17 19 21 23 25 27 29 31 9 Years since initlation of colonizaflon 10 ( Dry - -Humid (good infrastructuure) 11 - Transitional - - - Humid 12 Source: Authors' calculations (see text above). Source: Hecht and others 1988. Appendix 1. Methods Used in the Economic Anolysis of Logging and Ranching Activity 35 Table Al -4. Assumed zootechnical Indices Table Al -6. Sale value of animals Assumed zootechnicol indices Animal Birth rate Bull Death rate of heifers Cow Death rate of yearlings Cow Death rate of adults Fattened bull Source: Arima and Uhl 1997. Source: Arima and Uhl 1997. After the authors had obtained the treatments and harvest planning, In a man- total number of animals in each year aged system, 258 m3 of wood are necessary to (herd), the authors simulated the herd compo- support one job. The number of jobs was cal- sition according to zootechnical indices de- culated using technical coefficients taken scribed in table A1-4. The simulation of the from the literature (Verissimo and others herd also provided the number of animals 1992, Barreto and others 1998). sold. Ranching generates one job per group The authors assumed that the whole of 39 animals, including permanent and herd is meant for beef production-calf temporary employment. Employment coeffi- production, steer raising, fattening-the cients were calculated based on the works of most common case in the Amazon (table Mattos and Uhl (1994) and Arima and Uhl A1-5). (1997). The gross revenue from ranching was calculated by multiplying the number of ani- Net Present Value mals sold by the prices (table A1-6). The Net Present Value (NPV) and Internal Rate of Return (IRR) of logging activity, including timber extraction and wood pro- Predatory logging employs workers in the cessing, were obtained through the methodol- harvest, transport, and processing phases. To ogy described for items A and B (table A1-7). generate one job, 283 m3 of unprocessed The value of equipment, investments, and wood are required. On the other hand, in ad- useful life were obtained from Verissimo dition to the above phases, forest manage- and others (1992) and Barreto and others ment employs people in silvicutural (1998). Table Al -5. Example of herd composition Herd composition: Cow-calf operation Adult 3-4 years 2-3 year's 1-2 years 0- 1 year Bulls Males = 2Males Males M E ales - m Total Existing 2,402 17, 814 18, 177 19.337 - - 173,112 Purchased 721 - - - _ - - - _ - 42,038 Births - - - - 21,019 2 34,434 Deaths - 356 3 3 64 1,160 2 1,682 173,112 Sales 721 9 17.457 7 _ - - - - _ Balance 2,402 1.2 17,,814 18, 177 19,337 9 Source: Authors' calculations. 36 Sustainable Amazon: Limitations and Opportunities for Rural Development Table Al -7. Calculated Internal Rate of Return (IRR) and Net Present Value (NPV) of ranching and of extracting and processing wood Activity NPV 6% _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _( (U S$) Extracting and processing wood 138.615,463 Ranching (dry zones; 31.82% of pasture area) -9,672,789 Source: Authors' calculations. Appendix 2 Financial Returns from Timber Harvesting and Ranching in the Amazon he objective of this brief literature re- predatory logging and forest management in T view is to compare the financial re- tropical forests (summarized in table A2-1). turns of logging and ranching presented Note that, unless specified, the analyses in in this study with data in other published stud- these studies did not include log processing. ies on the subject. The comparison is rendered In general, this review points out that for- difficult by the different perspectives taken in est management is not economically competi- other studies. For example, no article on finan- tive with predatory logging. In virtually all cial returns from logging addresses the harvest cases, management generates positive Net and processing phases while considering the Present Values (NPV) (with discount rates increase in transport costs that result from varying from 5 percent to 20 percent). The wood scarcity in the proximity of sawmills. only exception is the study on Peru by South- gate and Elgegren (1995), in which the NPV was negative. Logging Activity Practically all the NPVs of conventional Recently, Pearce and others (1999) conducted systems cited in this review are superior to a literature review on financial returns from those encountered in the present study (table Table A2-1. Review of literature Study NPV (US$/ho) s Bann 1997 i &imi iSb"2 PL =1,697 / FM = 408 Haltia and Kelpi 1997 -F : Forest management better than ranching - Howard and others 1996 PL= 334-449 / FM= 204-263 Kishor and Constantino 1993 W - R PL =1292 / FM 854 Kumari 1996 K lid PL = 860-1380 / FM = 322-944 Mendoza and Ayemou 1992 qcq Forest management + processing = 160 18 Richards 1991 0-5 14%-15% annual return on capital, - including processing Southgate and Elgegren 1995 WPg NPV negative Note: PL = predatory logging; FM = forest management. Source: Pearce and others 1999, 16-18. 37 38 Sustainable Amazon Limitations and Opportunities for Rural Development A2-2). This difference occurs due to the scale wood up to 100 km from the mill obtained of the analysis. These studies (cited in table IRR of 14 percent to 62 percent,. A2-1) analyzed the financial return at the Oz6rio de Almeida and Uhl (1995) used scale of the property. The present authors' the data of Verissimo and others (1992) to cal- study was done at the municipal scale, at culate an IRR for logging activity on the mu- which transport costs have a large influence nicipal scale. In the predatory system, on financial returns. extraction and processing generated IRR of Only two articles incorporated both the 108 percent, whereas in the managed system, extraction and processing phases. Mendoza IRR was 103 percent. Considering only the and Ayemou (1992), in C6te d'Ivoire, ob- harvesting phase, the returns were 33 percent served a NPV/ha of US$160, using a discount and 29 percent with and without manage- rate of 10 percent. At this same discount rate, ment, respectively. The differences in the re- the present authors' study found a value of turns obtained in Oz6rio de Almeida and Uhl approximately US$100/ha (table A2-2). Rich- and in the present study (122 percent without ards and others (1991), in Mexico, noted an management, 73 percent with) occurred prin- annual return rate on capital of 14 percent to cipally due to the fact that Oz6rio de Almeida 15 percent. Unfortunately, it is not possible to and Uhl did not explicitly incorporate differ- compare these results with the present au- ences in transport costs or the "extraction thors' study, because the latter used a horizon waves" described in appendix 1. The differ- of 30 years. ence in area harvested under managed forestry Studies of timber harvesting in the Ama- and traditional predatory logging is shown in zon consist primarily of those conducted by, table A2-3. Imazon. Studies by Uhl and others (1991), Barreto and others (1998) analyzed the Verissimo and others (1992, 1995), Barros costs and benefits of forest management on an and Uhl (1995), and Johns and others (1996) experimental scale (100 ha). The authors ob- analyzed the financial returns of predatory served a NPV per ha of US$430 with a dis- harvesting and processing in different regions count rate of 20 percent. In addition, they of the Amazon. The scale of the analyses was concluded that management is more profitable that of a typical sawmill. In general, the annu- than predatory logging due to greater efficien- al profit margins (profit/gross revenue) were cy of machine use and better use of logs. greater than 25 percent. Barros and Uhl ob- Another study, for the USDA Forest Ser- served Internal Rates of Return (IRR) of 124 vice, conducted recently by Holmes and others percent for small sawmills that used fluvial in the Paragominas region, observed results transport in the estuary of the lower Amazon. similar to those of Barreto and others (1998). Larger sawmills on terra firma harvesting Table A2-3. Area harvested Table A2-2. Net Present Values (NPV) per ha _ ,R X f ~~~~~~~~~~~~~~Area harvested (ha) Discountrate 1imne-'-rn& preciai./rr (%/year) *r | iD?.gr,a trip% /nal Wood value doss ii Predatory 6 i5.'4 1 o w iQ f otXr,,0 go j 1 C,. 6 . 5 High, medium, low 4 E4l i 492.200 15 '2 C- 3 High, medium 38,®~ 303,600 20 -.36-32,~~, c i ` Medium12 80 30 jZP 4.q 2' 3 Totall 921,600 Soturce: Authors. Source' Authors' calculations. Appendix 2. Financial Retums from Timber Harvesting and Ranching in the Amazon 39 Holmes and others compared an industrial- 4 There was a high ratio between cattle pric- scale managed harvest (500 ha) to the preda- es and input prices. tory system and found that management was Browder (1988) and Fearnside (1980) more lucrative: US$11.6/mi3 versus US$9.84/ reached similar conclusions. m3, or US$294/ha versus US$250/ha. In the 1990s, in general, the ranching Stone (1998a) employed GIS and an opti- studies supported the conclusions from work mization model of industry profit (harvesting in the previous decades, but they also demon- and processing) to project logging in the state strated the economic viability of specific of Para for the period 1996-2000. In the sce- ranching models, such as small-scale milk pro- nario in which timber prices increase 3 per- duction. For example, Mattos and Uhl (1994) cent per year and processing capacity grows at analyzed ranches in Paragominas and ob- an annual rate of 16 percent, an area of 22 served that ranching practiced in an extensive million ha would be logged during this 4-year form generated IRR lower than 5 percent. In period, generating gross revenue of US$1,677 contrast, dairy ranching on a small scale pro- per ha in today's values (discount rate of 5 duced returns of 12 percent, and beef ranch- percent). In the present authors' study, the ing in reformed pastures obtained returns of present value of gross revenue was US$940 12 percent to 21 percent.'6 per ha for the predatory system compared to Muchagata and others (1999) completed US$965 per ha for managed forests, using the a detailed survey of 20 small properties in the same discount rate (5 percent). Maraba region (PA) during one year. The au- thors noted annual incomes (net of variable costs and depreciation) that varied from nega- Ranching tive R$39/ha for very small ranches (12 ha of The literature on ranching on terra firma in pasture) to positive values of R$42/ha for the Brazilian Amazon can be divided into larger ranches (85 ha of pasture). These prop- three phases. In the first phase-1960 to erties sold milk and animals and, in some cas- 1970-the studies had an agricultural and es, rented pastures. zootechnical focus and demonstrated that the Faminow and others (1998) demonstrated Amazon region was appropriate for cattle that the predominance of pasture and cattle on production, because the grasses grew vigor- small properties was due to the lower risk of ously and the animals achieved good weight this activity in relation to agroforestry systems. gains (Falesi 1976). In addition, they showed that the risks of vari- The articles in the second phase of the ation in prices and production limited the literature, the 1980s, showed that ranching adoption of more intensive technologies. The did not have a satisfactory financial perfor- average profits per ranch were R$6,000, much mance. Hecht and others (1988) observed higher than the Brazilian per capita GDP. that, using traditional technology, the IRR The data used in the present study were was negative. The return rates were positive obtained from Arima and Uhl (1996). Ranch- (5 percent to 31 percent) only when there was ing on terra firma (south of Pari state) in the a combination of two or more of the follow- traditional extensive system generated IRR of ing factors: 3 percent to 5 percent. Small ranches special- 1 Ranches received financial incentives and izing in dairy production obtained higher IRR, subsidized credit. around 9 percent. These numbers are consis- 2 Land prices increased (speculation). tent with a study conducted by the Rural Syn- 3 Overstocking took place in the initial dicate of Araguaina, Tocantins (Nehmi Filho period. 1999) in which the IRR was 5 percent in 40 Sustainable Amazon: Limitations and Opportunities for Rural Development Redencao (south of Para), 7 percent in Ara- returns (-10 percent). Despite these findings, guaina (TO), and 5 percent in Guapore (MT). the increases in the Amazonian cattle herd In summary, the studies conducted to and in extensive ranching continue without date demonstrate that the extensive ranching apparent economic justification. Various practiced by the majority of ranchers gener- explanatory hypotheses, such as capital ates a very low return. Small dairy ranches gains with land valuation, await empirical located close to highways obtain satisfactory verification. Appendix 3 Calculating Carbon Compensation and Taxation Values under Predatory Logging his appendix explains how the authors the benefits were discounted in time (gross T obtained the carbon compensation val- revenue). The values obtained are listed in ta- ues and the taxes necessary to equalize bles A3-2a and A3-2b. revenues between predatory logging and forest 23 Bt 31 Bm management [1] t =B , C [11 ~ t=O ( + r)t 1( ) Analysis of Gross Revenue where Table A3-1 shows the Gross Present Values (GPV) without compensation for carbon or Bt = gross revenue of the predatory sys- taxation. tem (extraction + processing + The calculation of carbon compensation ranching) values is simple. The authors assumed the Bm = gross revenue of management payments for additional carbon retained in the (extraction + processing) system under management were made in year CP = carbon payment t=O, following equation [1]. Only the values of r = discount rate. Table A3-1. Gross Present Value (GPV) In the current situation: Humid zone and transitional areas ~PV t 10 ~U~'$~ GPV at 4.92% (US$) t.%US$ Predatory logging + ranching 525,713,391 876,524,943 (humid zone)a Predatory logging + ranching 50, . 4 - 1 l A1B (transitional zone) Managed system 4 876,524,943 F Ul@@ Note: Under a discount rate of 4.92% (in humid areas), society would be indifferent between the GPV of predatory logging/ranch- ing and managed forestry. Under a discount rate of 3.86% (in transitional areas), society would be indifferent between the GPV of predatory logging/ranching and managed forestry. a. 1.83% pasture, without carbon payments or taxes. Source: Authors' calculations. 41 42 Sustainable Amazon: Limitations and Opportunities for Rural Development Table A3-2a. Value of payments for carbon necessary to equalize the GPV of the predatory system and management (humid zone) under different discount rates Discount rate value of carbon (percent) credits 10 11 . 114.384. 17;2 _ 1 5 - 1g 111823.577 _ 20 96.709.188 Note: Assumes carbon payments in year t . a. For calculation of carbon saved, see D gelow. Source: Authors' calculations. Table A3-2b. Value of payments for carbon necessary to equalize the GPV of the predatory system and management (transitional zone) under different discount rates Discount rate V value of carbon (percent) fij - credits S$ 10 0 - 129,146,186 15 119,730,092 20 - 101,326,961 Note: Assumes carbon payments in year t a. For calculation of carbon saved, see D gelow. Source: Authors' calculations. Private Sector Analysis: where Compensation for Carbon The analysis for the private sector was Bt, Ct = costs and benefits of extraction similar to the calculation above. The differ- and processing, predatory system ence is in the use of liquid values {equation Bm, Cm = costs and benefits of [2]). The current values without compensation management are given in table A3-3, and the values CP = carbon payment with carbon compensation appear in table r = discount rate. A3-4. 23 Bt - Ct, 30 Bmt - Cm, Private Sector Analysis: Taxes [2] Bt t (1 + B=I (1 + +) The authors calculated the tax on convention- al logging necessary for equalization of finan- cial returns of predatory and managed logging (equation [311. Table A3-3. Economic performance of predatory logging and forest management ereEJetci~y Forest Parameter e management IRR 71% NPV at 10% - 89,468,031 Source: Authors' calculations. Appendix 3 Calculating Carbon Compensation and Taxotion Values under Predatory Logging 43 Table A3-4. Value of carbon payments necessary to equalize NPV under different discount rates Value of carbon credits (US$) . Equalize IRR (122%) - 1,749,196 NPV at 10% ,9 18,311.019 , NPV at 15% 66348923,750,838 NPV at 20% 46E73,9223,049,966 -- Note: Assumes carbon payments in year to. a. See calculation of reduction in carbon emissions in D below: 9,100,000 tons. Source Authors' calculations. 23 (Bt, x ir) - Ct 30 Bm, - Cmt Carbon Emissions [3] (1+r)t t=o (1+ r) The difference in the amount of carbon re- leased by predatory logging and forest man- where agement was obtained following equation [5]. The quantities of carbon emitted at each har- t < 1 is the tax charged. vest intensity are given in table A3-6. The value of the tax per cubic meter of [5] AC,, = EC,N, - EC.N. wood harvested (X) was obtained by dividing =' r= the total value of the tax by the total volume extracted (table A3-5). where 23 C = amount of carbon released under 14] XBet x~ (1- 7r) = Xpredatory logging under harvest intensity i (tons/ha) =X Cm= amount of carbon released by forest Vol management at harvest intensity m (tons/ha) where N = area extracted at intensity i, m (ha). = value of tax (US$/m3) Predatory logging releases approximately Vol= ota vlum hrvete 9 million additional tons of carbon when Vol= total volume harvested (m3 of logs). compared to forest management in an area of Table A3-5. Value of taxes on wood derived from predatory logging to equalize NPV under different discount rates Value of tax (US$/m3) Equalize I RR t71f.) _ .4 6.20 NPV at 10% i |8. 1.03 NPV at 15% j >2.41 NPV at 20% - 3 3.30 Source- Authors' calculations. 44 Sustainable Amazon: Limitations and Opportunities for Rural Development Table A3-6. Carbon emission per ha Harvest level 1 Managed forest , ... .. ...... . ............. High0 - 2.19 High and mediuma 0 7.13 High, medium and low 17.00 Pasture - Source: Gerwing and others. a. Assuming that management reduces carbon emissions proportionally to high harvest intensity (17/31 = 0.548). equal size (table A3-7). Pastures release an ad- this difference between ranching and forest ditional 4 million tons of carbon, although oc- management reaches 22.7 million tons. In curring in only 1.83 percent of the total area. sum, the predatory system releases 13 million In transitional areas (10.1 percent pasture), more tons of carbon than is released under forest management. Table A3-7. Carbon emissions In each activity Acttvity CareN Additionality Forest management M 04 Predatory logging 0D & 9,090,768 Pasture - humid area (1.83% of total) 4,095,616 Pasture - transitioncl areas (10.1% of total) 22,760,684 Note: Base = forest management Source: Authors' calculations. Notes 1 As would be expected, the size of the Embrapa, and the Tropical Forest Foun- property also has an important effect on dation (TFF). Basically, the system con- the stocking rate. sists of a selective harvest based on an 2 A longer histor-y of colonization (approxi- inventory of commercial trees; planned mately one century), which resulted in roads, patios, and skid trails; vine cut- more prolonged soil use, could be part of ting; directional felling; and planned the cause of this low level of land use and skidding. In addition, the management high abandonment rate. On the other plan should contain techniques to stimu- hand, this long history also implies that late regeneration and growth of commer- ample time has passed to identify adapted cial trees and an annual harvesting crops and technologies; yet none appears to have emerged. 7 The economic model underlying this 3 Agricultural uses include pasture, annual section is described in appendix 1. crops, perennial crops, extractivism, and 8 Initially, Paragominas was established as abandoned land. a ranching frontier, stimulated by finan- cial incentives. Mattos and Uhl (1994) 4 To interpret these data, it is important to interviewed 27 ranchers at the beginning observe that the majority of public of the 1990s and observed an Internal, protected areas are in the humid zone, Rate of Return for ranching of 13 per- thus reducing the percentage of census . . . areas in agricultural use. This high per- reveal a rate of return of 3.1 percent for centage in protected areas undoubtedly is traditional techniques and 10 percent to also partially influenced by the low 14 percent for more technologically agricultural potential of these areas. advanced operations (appendix 1). Field 5 The net revenue is estimated at US$500 evidence indicates that significant million, assuming a profit margin of 20 decapitalization has occurred in the percent. (Profit margins oscillate from a region. However, it remains to be seen minimum of 15 percent to a maximum whether the decapitalization will lead to of 25 percent. Verissimo and others, the abandonment of the rural area, or its unpublished. consolidation under advanced technology. 6 This forest management system has been 9 This value, US$2-3 per ton, equalizes the developed and described by Imazon, Net Present Value (NPV) of sustainable 45 46 Sustainable Amozon. Limitations and Opportunities for Rural Development and predatory harvesting techniques, eters (AVHRR) aboard the NOAA-7, -9, under discount rates varying from 10 -11, and -14 polar orbiting satellites. See percent to 20 percent. . 10 The results of these interviews are: 80 13 Zoning land for agricultural use is seen percent were in favor of National Forests, by some farmers as a possible way to 3 percent were opposed to their creation, escape the restrictions of the legal reserve, and 17 percent had no opinion. which require at least 80 percent of each 1 1 This section is based on the Minis.6rio do property to be under managed forests. Meio Ambiente (MMA) document, 14 See . the Creation of National Forests in the Legal Amazon" (Verissimo and others 15 The simulation area corresponds to half 2000). The data and methodology are of Paragominas county, the most impor- detailed in that publication. tant logging center in the Amazon. 12 The NOAA/ NASA AVHRR Oceans 16 Pasture reform consists of removing Pathfinder sea surface temperature (SST) invading vegetation, grading the area, data are derived from the 5-channel planting better-adapted grasses, and, in Advanced Very High Resolution Radiom- some cases, applying fertilizer. References Arima, E., and C. Uhl. 1996. "Pecuaria na Browder, J. 1988. "Public Policy and Defores- Amaz6nia Oriental: desempenho atual e tation in the Brazilian Amazon." In R. perspectivas futuras." S6rie Arnaz6nia Repetto and M. Gillis, eds. Public Policy N°1. Belem: Imazon. and the Misuse of Forest Resources. Arima, E., and C. Uhl. 1997. "Ranching in Cambridge: Cambridge University Press. the Brazilian Amazon in a National Chomitz, K. M., and T. S. Thomas. 2001. Context: Economics, Policy, and Practice." "Geographic Patterns of Land Research Society and Natural Resources 10: 433- Use and Land Intensity in the Brazilian 51. Amazon." Policy Research Working Paper Bann, C. 1997. "An Economic Assessment of Series 2687. World Bank, Washington, Tropical Forest Land Use Options, D.C. . Ratanakiri Province, Cambodia." EEPSEA Cochrane, T., and P. Sanchez. 1992. "Land Research Report Series. Economy and Resources, Soil and Their Management in Environment Program for Southeast Asia, the Amazon Region: A State of Knowledge Singapore. Report." In S. Hecht, ed., Amazon: Agri- Barreto, P., P. Amaral, E. Vidal, and C. Uhl. culture and Land Use. Cali, Col6mbia: 1998. "Costs and Benefits of Forest CIAT: 137-209. Management for Timber Production in the Dean, W. 1989. A Luta pela Borracha no Eastern Amazon. " Forest Ecology and Brasil: um estudo de hist6ria ecol6gica. Management 108: 9-26. Sao Paulo: Nobel. Barreto, P., and E. Arima. 2000. "As Florestas Embrapa-Cpatu. 1999. Seminirio Sobre Nacionais na Amaz6nia: uma consulta a P empresarios madeireiros e a atores afins ardo n tens d aledutrs , . . ~~~~~Graos na Amazomna: potencialidades e politica florestal." Minist6rio do Meio limitacoes tecnol6gicas e ambientais. Ambiente (MMA) e Organizacao das Bel6m, 15-16 de dezembro. Nac6es Unidas para Agricultura e Alimentacao (FAO). Brasilia. Falesi, I. 1976. "Ecossistema da pastagem Barros A., andC.hl195Logcultivada na Amaz6nia brasileira." Boletim Barros, A., and C. Uhl. 1995. "Logging along d eqia mrp-pr,Blm the Amazon River and Estuary: Patterns, Problems, and Potential." Forest Ecology Faminow, M., C. Dahl, S. Vosti, J. Witcover, and Management 77: 87-105. and S. Oliveira. 1998. "Smallholder Risk, 47 48 Sustainable Amazon: Limitations and Opportunities for Rural Development Cattle, and Deforestation in the Western Holmes, T. P., J. Blate, C. J. Zweede, R. Brazilian Amazon." Paper presented at Pereira, Jr., P. Barreto, and R. Bauch. FAO Expert Consultancy on Policies for 2001. "Financial and Ecological Indicators Animal Production and Natural Resource of Reduced Impact Logging Performance Management. Brasilia. May. in the Eastern Amazon." Forest Ecology and Management 5583: 118. FAO/UNDP/MARA. 1992. "Principais Indicadores Socioecon6micos dos Howard, A., R. Rice, and R. Gullison. 1996. Assentamentos de Reforma Agraria." "Simulated Financial Returns and Selected FAO/PNUD-Ministerio da Agricultura e Environmental Impacts from Four Alter- Reforma Agraria. Projeto BRA-87/022. native Silvicultural Prescriptions Applied Rio de Janeiro. to the Neo-Tropics: A Case Study of the Chimanes Forest, Bolivia." Forest Ecology Fearnside, P. M. 1980. "The Effects of Cattle and Management 89: 43-57. Pasture on Soil Fertility in the Brazilian Amazon: Consequences for Beef Produc- Instituto Brasileiro de Geografia e Estatistica tion Sustainability." Tropical Ecology 21: (IBGE). 1996. Censo Agropecudrio [Agri- 125-37. cultural Census] 1995-1996. Rio de Gallup, J. L., and J. D. Sachs. 2000. "Agricul- J ture, Climate, and Technology: Why Are Instituto Socioambiental (ISA); Instituto do the Tropics Falling Behind?" American Homem e Meio Ambiente da Amaz6nia Journal of Agricultural Economics 82: (Imazon); Instituto de Pesquisa Ambiental 731-37. da Amaz6nia (IPAM); Instituto Sociedade, Populacao e Natureza (ISPN); Grupo de Gerwing and others. Unpublished. "Land Trabalho Amaz6nico (GTA); e Conserva- Cover and Carbon Density Maps for the tion International. 1999. Seminario Brazilian Legal Amazon." Imazon, Belem. Consulta de Macapi 99: avaliacao e Manuscript prepared for the World Bank identificacao de ac6es prioritArias para a Global Overlays project. conservacao, utilizacao sustentavel e Goodland, R. J. A., and H. S. Irwin. 1975. reparticao dos beneficios da Amazon Jungle: Green Hell to Red Desert: biodiversidade na Amaz6nia. An Ecological Discussion of the Environ- Johns, J., P. Barreto, and C. Uhl. 1996. "Log- mental Impact of the Highway Construc- ging Damage in Planned and Unplanned tion Program in the Amazon Basin. New Logging Operations and Its Implications York: Elsevier Scientific Publishing. for Sustainable Timber Production in the Haltia, O., and K. Keipi. 1997. "Financing Eastern Amazon." Forest Ecology and Forest Investments in Latin America: The Management 89: 59-77. Issue of Incentives." Environment Divi- Jones, D. W., and others. 1992. "Farming in sion, Inter-American Development Bank, Rond6nia." Working Paper. Oak Ridge Washington, D.C. National Laboratory, Oak Ridge, Tenn. Hecht, S., R. Norgaard, and G. Posio. 1988. Kishor, J., and L. Constantino. 1993. "Forest "The Economics of Cattle Ranching in Management and Competing Land Uses: the Eastern Amazon." Interciencia 13: An Economic Analysis for Costa Rica." 233-40. LATEN Dissemination Note, World Bank, Washington D.C. References 49 Kumari, K. 1996. "Sustainable Forest Man- the Brazilian Amazon." World Develop- agement: Myth or Reality? Exploring the ment 23: 1745-64. Prospects for Malaysia." Ambio 25 (7): Pearce, D. W., F. Putz, and J. Vanclay. 1999. 459-67. "A Sustainable Forest Future." CSERGE Mattos, M., and C. Uhl. 1994. "Economic Working Paper GEC 99-15. London. and Ecological Perspectives on Ranching in the Eastern Amazon in the 1990s." Projeto Radam. 1973-1978. "Levantamento World Development 22 (2): 145-58. de Recursos Naturais." Minist6rio das Minas e Energia, Departamento Nacional Mendoza, G., and A. Ayemou. 1992. "Analy- de Producao Mineral. Brasilia. sis of Some Forest Management Strategies . inCt 'vir:ARgoalCs td. Richards, E. M. 1991. "The Forest Eiidos of in Cote d'lvoire: A Regional Case Study." South-East Mexico: A Case Study of Forest Ecology and Management 47: 149- Community Based Sustained Yield Man- 74. agement." Commonwealth Forestry Moran, E .T. 1981. Developing the Amazon. Review 70 (4): 20-311. Bloomington: Indiana University Press. Schneider, R. R. 1994. Government and the . 1989. "Adaptation and Maladaptation Economy on the Amazon Frontier. World in Newly Settled Areas." In D. A. Bank Environment Paper 11. Washington, Schumann and W. L. Partridge. eds., The D.C. Human Ecology of Tropical Land Settle- Schneider, R. R., A. Verissimo, and V. Viana. ment in Latin America. Boulder, Col.: Unpublished. "Logging and Tropical Westview Press. Forest Conservation." World Bank, Muchagata, M., W. Rendeiro, and R. DECWD, Washington, D. C. Machado. 1999. "Sustentabilidade da . Atividade Pecuaria. Relat6rio do Encontro entre Agricultores e Pesquisadores." ODG/ Schubart, H. 1999. "Biodiversid e DEV, Norwich, U.K. March. territorio na Amazonia." Paper presented at the III Congresso Sul-Americano de National Research Council. 1993. Sustainable Alternativas de Desenvolvimento Agriculture and the Environment in the Resgatando a Amaz6nia: uma nova Humid Tropics. Washington, D.C.: Na- hist6ria. Manaus. tional Academy Press. Serrao, E. A. S., and A. K. 0. Homma. 1993. Nehmi Filho, V. A. 1999. "Relacao terra/gado "Brazil." In National Research Council, define o preco da propriedade." Sustainable Agriculture and the Environ- Informativo Rural Ano V (N° 38) ment in the Humid Tropics, 265-351. Sindicato Rural de Araguaina. Washington, D.C.: National Academy Oz6rio de Almeida, A. L. 1992. Deforestation Press. and Turnover in Amazon Colonization. Smith, N. J. H. 1982. Rainforest Corridors: World Bank Discussion Paper. Washing- The Transamazon Colonization Scheme. ton, D.C. Berkeley: University of California Press. Oz6rio de Almeida, A. L., and C. Uhl. 1995. Sombroek, Win. 2001. "Spatial and Temporal "Developing a Quantitative Framework Patterns of Amazon Rainfall: Conse- for Sustainable Resource Use Planning in quences for the Planning of Agricultural 50 Sustainable Amazon: Limitations and Opportunities for Rural Development Occupation and the Protection of Primary Verissimo, A., P. Barreto, M. Mattos, R. Forests." Ambio 30(7): 388-96. Tarifa, and C. Uhl. 1992. "Logging Im- Southgate, D., and J. Elgegren. 1995. "Devel- pacts and Prospects for Sustainable Forest opment of Tropical Timber Resourcs bManagement in an Old Amazonian Fron- opet of Trpia Tibe Resourcesb tier: The Case of Paragominas." Forest Local Communities: A Case Study from the Peruvian Amazon." Commonwealth Ecology and Management 55: 169-99. Forestry Review 74 (2): 142-46. Verissimo, A., P. Barreto, R. Tarifa, and C. Stone, S. 1998a. "Evolution of the Timber Uhl. 1995. "Extraction of a High-Value Sne,stry.along 8an" tiong Frofnther Timber Natural Resource from the Amazon: The Industry along an Aging Frontier: The Caeo,aogn.,oet clg n Case of Paragominas (1990-95)." World Case of Mahogany. Forest Ecology and Development 26 (3): 433-48. Stone, S. 1998b. "Using a Geographic Infor- Verissimo, A., C. Souza Jr., S. Stone, and C. mation System for Applied Policy Analysis: ihe Brzla Aaon ." Conservation The Case of Logging in the Eastern Ama- Biology 12 (1): 1-10. zon." Ecological Economics 27: 43-61. Toniolo, A., and C. Uhl. 1994. "Economic Verissimo, A., C. Souza, Jr., R. Salomao, and P. Barreto. 2000. "Identificacao de Areas and Ecological Perspectives on Agriculture cornto. 2000. paaentiaqo de Freas in the Eastern Amazon." World Develop- bca den pra na Amaz6nia loe P6blicas de Producao na Arnazomia Le- ment 23: 95 9-73. gal." Minist6rio do Meio Ambiente Uhl, C., A. Verissimo, A. Mattos, Z. (MMA), Organizacao das Nac6es Unidas Brandino, and I. Vieira. 1991. "Social, para a Agricultura e Alimentacao (FAO). Economic, and Ecological Consequences Rio de Janeiro. of Logging in the Amazon Frontier: The Verissimo, A., C. Souza, Jr., and E. Lima. Case of Tailnndia." Forest Ecology and Unpublished. "Bases para o zoneamento Management 46: 243-73. da atividade madeireira na Amaz6nia Uhl, C., P. Barreto, A. Verissimo, A. C. Barros, Legal." Imazon, Bel6m. P. Amaral, J. Gerwing, J. Johns, and E. Verissimo, A., and E. Lima. Unpublished. Vidal. 1997. "An Integrated Research "Caracterizaq5o dos p6los madeireiros na Approach to Address Natural Resource Amaz6nia Legal." Imazon, Bel,m. Problems in the Brazilian Amazon." Bioscience 47 (3): 160-68. Recent World Bank Technical Papers (continued) No 463 Stephen Foster, John Chilton, Marcus Moench, Franklin Cardy, and Manuel Schiffler, Groundwater in Rural Developmncnt. Facinig the Challenges of Supply and Resource Sustaznabzilty No. 465 Csaba Csaki and Zvi Lerman, eds., Structural Change in the Farming Sectors in Central and Eastern Europe Lessonsfor EU Accession -Second World Bank! FAO Workshop, June 27-29, 1999 No. 466 Barbara Nunberg, Readyfor Europe: Public Adunnistration Reformi and Europeatn Unizon Accession in Central and Eastern Europe No. 467 Quentin T Wodon with contributions from Robert Ayres, Matias Barenstein, Norman Hicks, Kihoon Lee, William Maloney, Pia Peeters, Corinne Siaens, and Shlomo Yitzhaki, Poverty and Policy in Latin America and the Caribbean No. 469 Laurian Unnevehr and Nancy Hirschhorn, Food Safety Issues in the Developing World No. 470 Alberto Valdes, ed., Agricultural Support Policies in Transition Economies No. 471 Brian Pinto, Vladimir Drebentsov, and Alexander Morozov, Dismantling Russia's Nonpayments System: Creating Conditionsfor Growth No 472 Jtt B. S. Gill, A Diagnostic Frameworkfor Revenue Admniiistration No 473 Esen Ulgenerk and Leila Zlaoui, From Transition to Accession Developing Stable and Coinpetitive Financial Markets in Bulgaria No. 474 loannis N Kessides, ed., Hungary. A Regulatory and Structural Review of Selected Infrastructuire Sectors No 475 Csaba Csaki, Zvi Lerman, and Sergey Sotnikov, Farm Sector Restructuring in Belarus: Progress and Constraints No. 476 Katherine Terrell, Czech Republic: Labor Market Report No. 481 Csaba Csaki, John Nash, Achim Fock, and Holger Kray, Food and Agriculture in Bulgaria The Challenge of Preparing for EU Accession No 482 Peter Havlik, Trade and Cost Competitiveness in the Czech Republic, Hungary, Poland, and Slovenia No. 483 Mojmir Mrak, Communal Infrastructuire in Slovenia: Survey of Investmient Needs and Policies Aimed at Encouraging Private Sector Participation No. 484 Csaba Csaki and Laura Tuck, Rural Development Strategy Eastern Europe and Central Asia No. 488 Nina Bubnova, Governance Imtipact on Private Investment No. 489 Tim Schwarz and David Satola, Telecommunilcations Legislation in Transitional and Developing Econoniies No. 490 Jesko Hentschel and Radha Seshagiri, The City Poverty Assessmenit. A Priner No 491 Daniel Mtiller-Jentsch, The Devlopmenit of Electricity Markets in the Euro-Mediterranean Area No. 492 Tuntivate Voravate, Douglas F. Barnes, and V Susan Bogach, Assessimig Markets for Renewable Energy in Rural Areas of Northwestern China No. 496 Jerry Lebo and Dieter Schelling, Design and Appraisal of Rural Transport Infrastructure: Ensuring Basic Access for Rural Communinties No. 497 Julian A. Lampietti, Anthony A. Kolb, Sumila Gulyani, and Vahram Avenesyan, Utility Pricing and Poor: Lessonsfrom Armiienia No. 498 Gillian Perkins and Ruslan Yemtsov, Armenia Restructuring to Sustain Universal Gemieral Education No. 499 Rogrigo A. Chaves, Susana Sanchez, Saul Schor, and Emil Teshuc, Financial Markets, Credit Constraints, and Investment in Rural Romania No. 500 Zvi Lerman and Karen Brooks, Turkmenistan An Assessment of Leasehold-Based Farm Restructuring No. 501 Aldo Baietti, Private Infrastructure in East Asia Lessons Learned in the Aftermath of the Crisis No. 505 Ali Hashim and Bill Allan, Treasury Reference Model No 506 Omer Gokcekus, Nick Manning, Ranjana Mukherjee, and Raj Nallari, Institutional Environment and Public Officials'Perforniance in Guyania No. 507 Ranjana Mukherjee, Omer Gokcekus, Nick Manning, and Pierre Landell-Mills, Bangladeshi The Experiemice and Perceptiois of Public Officials No. 509 World Bank, Kosovo Economic and Social Reformsfor Peace and Reconciliation No 510 Anatoly Vinokur, Joana Godinho, Christopher Dye, Nico Nagelkerke, The TB and HIV/AIDS Epidemics in the Russian Federation No 512 Geremia Palomba, Milan Vodopivec, Financing, Efficiency, and Equity in Albanian Eduction No. 513 Thomas O'Brien, Christian Filipov, The Current Regulatory Framework Governimig Business in Bulgaria ;fHf12ON Uf -wfirdszIlnumMuBu1 1 5031 Jeannette Marie Smith 00803 ISN I MC C3-301 WASHINGTON DC 9702,537