Deforestation in Brazilian Amazonia: History, Rates and Consequences

Philip M. Fearnside



Paper submitted for Special Section; “Brazilian Conservation: Challenges and Opportunities

Editors: Katrina Brandon, Gustavo A. B. da Fonseca, Anthony B. Rylands, José Maria Cardoso da Silva.


Address of first author:

Philip M. Fearnside

Instituto Nacional de Pesquisas da Amazônia (INPA)

Caixa Postal 478

Manaus 69083-000




E-mail of first author: <>

                        16 Nov. 2004                                     

                        13 Dec. 2004


Word count

Abstract to end of written text, excl. Literature cited:

Abstract to end of written text, incl. Literature cited:


1 Figure (graph)


Paper submitted: November 16, 2004; revised manuscript accepted December 18, 2004.


Deforestation in Brazilian Amazonia: History, Rates and Consequences





Instituto Nacional de Pesquisas da Amazônia (INPA), Caixa Postal 478, Manaus 69083-000, Amazonas, Brazil





Paper submitted: November 16, 2004; revised manuscript accepted December 18, 2004.



Abstract: Brazil’s Amazon forest remained largely intact until the “modern” era of deforestation began there with the inauguration of the Transamazon Highway in 1970.  Amazonian deforestation rates have trended upward since 1991, with clearing proceeding at a variable but always rapid pace.  Amazonian forests are cut for a number of reasons, but cattle ranching predominates.  The large and medium-sized ranches account for about 70% of clearing activity.  Profit from beef cattle sales is only one of the income sources that makes deforestation profitable.  Forest degradation results from logging, ground fires (made more likely by logging), and by the effects of fragmentation and edge formation.  Degradation contributes to forest loss.  The impacts of deforestation include loss of biodiversity, reduced water cycling (and rainfall), and contributions to global warming.  Strategies to slow deforestation include repression through licensing procedures, monitoring and fines.  The severity of penalties for deforestation needs to be sufficient to deter illegal clearing but not so great as to be inapplicable in practice.  Increasing the probability of detection and punishment is essential to for effective repression.  Policy reform is also needed to address root causes of deforestation, including the role of clearing in establishing land claims for both small and large actors.


Desmatamento na Amazônia Brasileira: História, Taxas e Conseqüências.







The major onslaught on the Amazon began only in the early 1970s, and,  while enormous tracts are still intact, the rate of forest loss is dramatic, most especially in the “arc of deforestation” along the southern and eastern edges.  Biodiversity loss and climatic impacts are both major concerns arising from the destruction of the region’s forests in the Amazon are the, and t.  The vastness size of the remaining forests remaining means that the potential impacts of continued clearing are far more serious than those, already severe, resulting from forest their loss to date. 


Combating deforestation in Brazil is a priority for government action and international assistance because of the major environmental impacts.  Monitoring and repression is currently the principal strategy.  Effective inspection and the levying of fines for those lacking the necessary permits of the Brazilian Institute for the Environment and Renewable Natural Resources (IBAMA), however, must be accompanied by an understanding of the social, economic and political aspects necessary to address the problem and the motives through changes in policy.


Extent and Rate of Deforestation


LANDSAT satellite data interpreted at Brazil’s National Institute for Space Research (INPE) (Fig. 1) indicate that, by 2003, forest cleared in Brazilian Amazonia had reached 648.5 × 10³ km² (16.2% of the 4 × 106 km² originally forested portion of Brazil's 5 × 106 km² Legal Amazon Region), including approximately 100 × 10³ km² of "old" (pre-1970) deforestation in Pará and Maranhão (Brazil, INPE 2004).  Both the current rate and the cumulative extent of deforestation represent vast areas.  The rate is often discussed in Brazil in terms of  Belgiums” – annual loss approaching the country’s area (30.5 × 10³ km²) – while the cumulative amount is compared to France (547.0 × 103 km2).  Clearing since construction of the Transamazon Highway in 1970 now surpasses the area of France.  Almost five centuries of European presence before 1970 deforested an area slightly larger than Portugal.  The original extent of Brazil’s Amazon forest was approximately the area of Western Europe.  Current values for deforestation can be obtained from INPE's web site:  Note, however, that the official explanations as to why deforestation rates fluctuate (decrees affecting incentives and programs for inspection and levying fines) are unlikely to be correct, as will be explained below.  In addition, a variety of technical questions remain open regarding the numbers themselves (Fearnside & Barbosa 2004).


[Figure 1 here]


Causes of Deforestation


In Brazilian Amazonia, the relative weight of small farmers versus large landholders changes continually with economic and demographic pressures.  Large landholders are most sensitive to economic changes such as interest rates offered by money markets and other financial investments, government subsidies for agricultural credit, the rate of general inflation, and changes in land prices.  Tax incentives were a strong motive in the 1970s and 1980s, and although a decree in June 1991 suspended new incentives,. the old incentives continue to the present day, contrary to the popular impression fostered by statements by government officials that all had been ended.  Other incentives, such as government-subsidized credit at rates well below inflation, became much scarcer after 1984..


Hyperinflation dominated the Brazilian economy for decades preceding Brazil's "Plano Real" economic reform (July 1994).  Land was at a premium, and prices reached levels way higher than could be justified as an input to agricultural and ranching production.  Deforestation was necessary to hold claim to land, and cutting the forest for cattle pasture was the cheapest and most effective in this sense, although the extent to which this activity was merely land speculation has been a matter of debate (Hecht et al. 1988; Faminow 1998; Fearnside 1987, 2002a).


Tax incentives and subsidized credit for large cattle ranches were important motives driving deforestation in the 1970s (e.g., Mahar 1979), but have played a lesser role since 1984.  Till around 1987, Lland speculation was also important,  until around 1987, but there was a subsequent increase in the role of pasture profit from beef production (Mattos et al. 1994; Margulis 2003), although this was limited to regions where soil and rainfall are suitable for pasture.  Rainfall is too high overIn large areas, especially in the west, rainfall is too high (Schneider et al. 2000).


Brazil's deepening economic recession best explains the decline in deforestation rates from 1987 through 1991.  Ranchers were unable to expand their clearings as quickly, and the government also lacked funds for highway construction and settlement projects.  The impact of repressive measures (helicopter patrols, confiscation of chainsaws, fines, etc.) was probably minor.  Most people, while complaining, continued to clear anyway.  Likewise, policy changes on fiscal incentives were also relatively ineffectual.  The decree suspending incentives (No. 153) was dated 25 June 1991 – subsequent to most of the observed decline in the rate (Fig. 21).  Even for the last year (1991) the effect would have been minimal, as the average date for the LANDSAT images for the 1991 data set was August.  At the low point in 1991 many ranchers were unable to use their funds for investment in clearing because then-president Fernando Collor de Melo had seized bank accounts in March of 1990, with funds subsequently released in small installments over a period of years.



      The 1995 peak is probably largely, in large part, a reflection of economic recovery under the Plano Real.    The reforms increased the availability of funds for investment (, which includinged cattle ranches), and municipal elections in 1994 also lead to an increase in agricultural credit which, in providing cash to landholders, is much more effective in spurring deforestation than economic changes which affect the value of less liquid assets such as land.  The subsequent fall in deforestation rates in 1996 and 1997 is a logical consequence of the Plano Real having sharply cut the rate of inflation.  Land values peaked in 1995 and fell by about 50% by the end of 1997.  Falling land values make land speculation unattractive to investors.  Deforestation rates then climbed to a level of 17-18 × 10³ km² per year, which remained constant for the next four years, followed by a jump in 2002 to a new plateau at 23 × 10³ km² per year (Fig. 1).


The association of major swings in deforestation rate with macroeconomic factors such as money availability and inflation rate is one indication that much of the clearing is done by those who invest in medium and to large cattle ranches, rather than by small farmers using family labor.  The predominant role of larger landholders is also indicated by the location of clearing activity.  The distribution among the region's nine states of in 1991 clearing among the region's nine states indicates that most is in those dominated by ranchers: Mato Grosso alone accounted for 26% of the 11.1 × 10³ km² total.  Mato Grosso has the highest percentage of its privately held land in ranches of 1000 ha or more: 84% at the time of the 1985 agricultural census.  By contrast, Rondônia – famous for its deforestation by small farmers – accounted for only 10% of the 1991 deforestation total, and Acre 3%.  The rise to a rate of 23 × 10³ km²/year in 2002, even with a lackluster domestic economy, can be partly attributed to an increasing globalization of the forces of deforestation.  In addition to the old forces, there was a sharp growth in the international market for soybeans and especially for beef, the latter previously restricted to the domestic market because of foot-and-mouth disease (Alencar et al. 2004;  Kaimowitz et al. 2004).


      The number of properties censused in each size class explains 74% of the variation in deforestation rates among the nine Amazonian states.  Multiple regressions indicate that 30% of the clearing in both 1990 and 1991 can be attributed to small farmers (properties ≤ 100 ha in area), and the remaining 70% to medium or large ranchers (Fearnside 1993a).  Surveys carried out in 1998 by tThe Institute for Environmental Research in Amazonia (IPAM) in 202 properties in the "arc of deforestation" from Paragominas to Rio Branco in the "arc of deforestation" indicated only 25% of the clearing in properties of 100 ha or less (Nepstad et al. 1999a).  The social cost of substantially reducing deforestation rates would therefore be much less than is implied by frequent pronouncements that blame "poverty" for environmental problems in the region.  Understanding who is to blame for deforestation is vital for any program attempting to reduce it, and a strategy to this end which promotes agroforestry among small farmers is doomed to fail when large when cattle ranchers with large estates are the principal villains.


Money laundering is also involved.  "Dirty" money from drugs, corruption and many other illegal sources can be converted into "clean" money by investing in Amazonian business ventures, such as gold mining dredges and cattle ranches, even if they are unprofitable based on the face value of return on investment.  The rapidly increasing drug traffic in Amazonia is likely to be reflected in greater investments in ranching and other activities for money-laundering purposes.  The incipient advance of soybean plantations in Amazonia, currently poses the greatest potential threat with its provision and stimulus for massive government investment in infrastructure such as waterways, railways and highways.  This infrastructure development unleashes an insidious chain of investment and profiteering that can be expected to destroy more forest than the plantations as such (Fearnside 2001a).  Logging roads, especially for mahogany extraction, precede and accompany highways, opening up frontiers for investing timber profits in soybean plantations and cattle ranches.  Timber extraction and edge effects of forest fragments result in drier understorey microclimates that greatly increase the flammability of the forest, leading to understory fires on the forest floor that set in motion a vicious cycle of tree mortality, increasing fuel loads, and re-entry of fire, until the forest is completely destroyed.  Reaching this endpoint can increase the role of fire from that of “cryptic deforestation” to a level of damage that would appear on LANDSAT imagery as deforestation (Cochrane et al. 1999; Nepstad et al. 1999b). 



One of the most basic causes of deforestation is the effect of transportation infrastructure in accelerating migration to remote areas and in increasing the rate of clearing in already-established properties.  The Avança Brasil program, a development package for the period 2000-2007, includeed US$20 billion in infrastructure in the Amazon region (Laurance et al. 2001; Fearnside 2002b; Nepstad et al. 2001).  Particularly damaging are the BR-163 (Santarém-Cuiabá) and BR-319 (Manaus-Porto Velho) highways, which will provide access to large blocks of forest currently little disturbed.  The “Pluriannual Plan” (PPA) that supplanted Avança Brasil for the 2004-2007 period is virtually identical to its predecessor.  The prominence of transporting soybeans as the economic justification of much of the planned infrastructure makes them especially controversial, as mechanized soybean cultivation brings little in the way of employment and other social benefit for local populations.


The Role of Forest Degradation in Forest Loss



Logging greatly increases the susceptibility of forest to fire, and once fire enters it kills trees and increases fuel loads and understorey drying, thereby increasing the risk of more-damaging future fires and the complete degradation of the forest.  Because only some of the tree species in Amazonia are commercially valuable, harvested wood volume is small relative to the total timber stock.  This is highly misleading as an indication of the damage logging causes.  Lacking special procedures to minimize damage, the additional volume severely damaged or killed is almost double the volume of the the harvested trees (Veríssimo et al. 1992).  Because many smaller trees are killed, the effect on individuals is even greater: One one study near Paragominas, Pará counted 27 trees killed by collateral damage for every tree harvested (Veríssimo et al. 1992)].  Gaps in the canopy allow sun and wind to reach the forest floor, resulting in drier microclimates.  The number of rainless days needed for the understorey to reach flammable condition is much less for a forest that has been logged than for one that has not (Nepstad et al. 2004).  The larger fuel load in the logged forest results in greater fire damage.



Rather than the canopy fires typical of coniferous forests, fires in Amazonian forests take the form of a slowly moving line of flame in the understorey.  The bases of many trees are burned as the fire lingers.  Amazonian forest trees are not adapted to fire and lack thick bark and other forms of protection, and .  mMortality from fire damage from a first burn provides the fuel and dryness needed to make the second and subsequent fires much more damaging.  The temperatures reached and the height of flames in the second fire are significantly greater than in the first fire, thereby killing many additional trees (Cochrane 2003).  After several fires the area is cleared to the point where it appears as deforestation in LANDSAT imagery.


During the 1997-1998 El Niño event, the “Great Roraima Fire” burned 11,394-13,928 km² of intact primary forest (Barbosa & Fearnside 1999), and fires in the “arc of deforestation” were estimated to have totalled a further 15 × 10³ km² (Cochrane 2003; Nepstad et al. 1999b).  Substantial burning also occurred in logging areas near Tailândia, southern Pará, while additional areas of standing forest burned in the state of Amazonas.  In southern Pará, the damage from El Niño events is magnified by the combination of the dry season being longer than in other parts of Amazonia, the concentration of logging activity in this part of the region, and the concentration of deforestation and associated burning for agriculture and ranching in this major part of the “arc of fire.”


Impacts of Deforestation


Loss of productivity

Loss of productivity through soil erosion, nutrient depletion, and soil compaction are among the most obvious impacts of deforestation.  Agricultural productivity declines as soil quality degrades, although a lower plateau of productivity can be maintained by systems such as shifting cultivation. Continuous inputs of lime, manure, and nutrients can counter degradation, but economic and physical resource limitations render this unviable for large areas distant from urban markets (Fearnside 1997a).  Deforestation removes options for sustainable forest management for both timber and, presently little-valued, genetic and pharmacological resources.  Also lost. are environmental services (e.g., biodiversity maintenance, water cycling and avoidance of global warming), which presently fall outside of most economic accounting systems.


Changes in hydrological regime

Watershed functions are lost when forest is converted to uses such as pasture.  Precipitation in deforested areas quickly runs off, creating flash floods followed by periods of greatly reduced or no stream flow.  Regular flooding patterns are important both for natural ecosystem functioning in and near the river as well as for floodplain agriculture.


The percentage of water recycled within the Amazon Basin is now believed to be 20-30% (Lean et al. 1996), rather than the traditional figure of 50% (Salati et al. 1979; Salati & Vose 1984).  Although indicating that the hydrological impact of deforestation would be less than otherwise thought, in reality the opposite is true.  That almost exactly 50% of the rain falling in the Basin basin flows out through the Amazon River implies that the other 50% has been recycled, assuming that water vapor stays within the Basinbasin.  In fact, some water vapor escapes to the Pacific, especially in the northwest corner of the Basin basin in Colombia.  More importantly, a substantial amount is transported to south and south-central Brazil, Paraguay, Uruguay and Argentina, and some continues across the Atlantic to southern Africa.  This gives Amazonian deforestation a level of impact to date little appreciated at the policy level (Fearnside 2004).  Rio de Janeiro and São Paulo, were subject to repeated blackouts and electricity rationing in 2001, as a result of low water levels in hydroelectric reservoirs in the non-Amazonian portion of the country.  The role of Amazonian water vapor in providing rainfall to this region should make the importance of Amazonian forest conservation readily apparent.  Water is supplied to central-south Brazil by air currents (low-level jets) coming from Bolivia and from the western part of Brazilian Amazonia (western Rondônia, Acre and western Amazonas).  Water vapor supply to the central-south has different magnitudes and differing importance depending on the season.  During the dry-to-wet transition period (September-October) in southwest Amazonia, water vapor supply is particularly important to avoid a lengthening of the dry season in São Paulo, critical Brazil’s most productive agricultural regions.  Hydroelectric generation capacity, on the other hand, is particularly dependent on rain in the austral summer (December), corresponding to the rainy season in southwest Amazonia when the difference between the hydrological behavior of forested and deforested areas is least.  According to preliminary estimates by Pedro Silva Dias of the University of São Paulo, roughly 70% of the rainfall in the state of São Paulo comes from Amazonian water vapor during this period.


In addition to maintenance of basin-wide precipitation and long-range water transport, deforestation also produces meso-scale effects.  Recent observations of a slight (approximately 5%) increase in rainfall in the heavily deforested Ji-Paraná area of Rondônia, together with satellite observations showing cloud formation occurring preferentially over clearings as small as 5 km in diameter, corroborate preliminary theoretical results on meso-scale effects of deforestation.  The potential of deforestation to increase local precipitation by providing convective updrafts of air that trigger cloud formation might mislead the unwary to conclude that deforestation is not so bad.  It could provide a deceptive temporary improvement as deforestation advances, only to be followed by a precipitous decline in rainfall as deforestation passes a threshold and falls off quickly.  In addition, the increase of rainfall over a clearing means that rain has been, in effect “stolen” from somewhere else.  This includes both the distant destinations of water vapor transport and the nearby forest edges.  Forest edges would suffer because the convection cells formed over clearings will not only take wet air aloft to provoke rain, but will also create a down draft over the nearby forest, bringing dry air down that will inhibit rainfall and dry the forest near the edge of the clearing (perhaps in a band about 20-km wide, provided prevailing winds are not blowing).  This drying from edges adds an additional feedback reinforcing the degradation of forest edges through fire and water stress.


Biodiversity loss

Biodiversity maintenance is a function to which many attribute value beyond the commercial sale of products (Fearnside 1999).  The loss of major portions of Brazil’s tropical forests is impoverishing, above all, the earth’s biodiversity (Capobianco et al. 2001; Mittermeier et al. 2002).  The biodiversity impact of continued deforestation is much greater in areas with little remaining forest and high levels of endemism, such as the Atlantic Forest (Mittermeier et al. 1999); Brooks & Rylands 2003).  If Amazonian deforestation is allowed to continue to near complete destruction, the same levels of risk to biodiversity would also apply there.


Net Emissions of Greenhouse Gases

Forest fires emit greenhouse gases.  The “Great Roraima Fire” during the 1997-1998 El Niño event emitted released a 17.9-18.3 × 106 t CO2-equivalent C through combustion, of the which 67% (12.0-12.3 × 106 t CO2-equivalent C) was from fires in primary forest, or 12.0-12.3 × 106 t CO2-equivalent C (Barbosa & Fearnside 1999).  CO2 carbon equivalent converts the various greenhouse gases to CO2-equivalent C considering the global warming potential of each gas over a 100-year time horizon using the conversions adopted under the Kyoto Protocol.  Clearing at the rate prevailing in 2003 implies approximately 429 × 106 tons of CO2-equivalent carbon emission, while for the 1988-1994 period (the base period used by Brazil for its initial greenhouse gas inventory under the climate convention) released 275 × 106 tons including all components (updated from Fearnside [2000a], including corrections in Fearnside & Laurance [2004] and Nogueira et al. [ndin press]), or 252 × 106 tons if only the emissions and wood density considered in the National Inventory are used.  This figure is slightly more than double the official value of 116.9 × 106 tons (Brazil, MCT 2004,, p.149).  The difference is explained by a series of omitted components in the official estimate (including roots and necromass) and by a high estimate for carbon uptake by secondary forests that does not reflect the slow rate at which these grow in degraded Amazonian pastures.


What most distinguishes the global warming implications of Amazonian deforestation from those of other tropical forests is the huge potential for future emissions.  In 1990, net committed emissions from Brazilian deforestation represented 5% of the global total from all sources, including both land-use change and fossil fuels, at that time (Fearnside 1997b), while the carbon stock in biomass in Brazilian Amazonia represented 38% of the tropical total (Fearnside 2000b, p.129).  “Net committed emissions” refers to the net result of emissions and uptakes as an area of forest is replaced by a patchwork of other land uses (in the proportions that would be reached at equilibrium if current land-use patterns continue).


Strategies to Slow Deforestation



In Brazil, deforestation control is mainly by repression through clearing licenses, inspections and fines.  Campaigns are often announced simultaneously with the annual conclusions of INPE’s monitoring program.  The first major effort to repress deforestation was in 1989 under the “Our Nature” (Nossa Natureza) program.  Since then a series of crackdowns has failed to halt its progress.  Clearing rates in the region seem to rise and fall independent of these programs.  Repression, while undoubtedly necessary, needs to be done differentlyrethought and underlying causes also need to be addressed.


A strong indication that Brazil is capable of controlling deforestation was provided by the dramatic drop in the number of fires on 1 July 2000, when a prohibition on burning came into effect.  Imagery from the AVHRR sensor interpreted at INPE indicates that this drop was more than 80%.  Deforestation also fell because of a deforestation and licensing program that was active in Mato Grosso over the 1999-2001 period, despite subsequent changes in the state government having halted the program as an effective impediment to deforestation (Fearnside 2003a; Fearnside & Barbosa 2003).


The reduction of burning in Mato Grosso was achieved by a combination of measures.  A system of permits was instituted by the state’s environmental agency (FEEMA) that included a printout of a satellite image showing the property boundaries and the existing deforestation.  Fines were issued with a satellite image printed on them, thus discouraging argument and attempts to misrepresent the area actually cleared.  Portions of Mato Grosso with the greatest decreases in burning were those subjected to special community training and education programs in fire management by the Amazonian Working Group (GTA) and the Friends of the Earth -Brazilian Amazonia, with support from FEEMA and from the Prevention and Control Programme for Forest and Savannah Fires in the Legal Amazon (PROARCO) project of the Pilot Program to Conserve the Brazilian Rainforest (PP-G7).  Plans have been announced to extend the system to select municipalities in Pará and Rondônia.


Policy Reform on Taxes, Credit and Subsidies

A major problem in controlling deforestation is that much of what needs to be done is outside of the purview of agencies such as IBAMA that are charged with environmental issues.  Authority to change tax laws and credit policies rest with other government agencies, as do resettlement policies and road-building and development priorities.  Tax subsidies for cattle ranches approved by the Superintendency for Development of Amazonia (SUDAM) were an important force motivating deforestation in the 1970s and 1980s; the ending of approval of new subsidized projects in June 1991 did not revoke the projects for which tax subsidies had already been granted.  SUDAM-approved projects not only gave tax-exemption on income they generated by the projects, but also allowed the owners to invest in their ranches part of the tax they owed on earnings from operations elsewhere.  The exclusion of ranches in 1991 did not affect other damaging activities, such as sawmills and pig-iron smelters fueled by charcoal.  The remaining tax subsidies need to be removed.


Another motive for deforestation, more prominent in the 1970s and 1980s than now, is land speculation.  The capital gain from selling a property after holding it for a few years was a major source of profit for ranchers as long as land values increased faster than inflation.  While average land values are no longer increasing at the rates prior to the abrupt slowing of inflation with the 1994 Real Plan, individual properties can still produce speculative profits, particularly when near a newly-built or improved road.  Heavy taxes should be applied to take the profit out of land speculation, both to remove the remaining speculative force in areas favored by infrastructure and to provide protection should there someday be a return to the astronomical inflation rates prevailing in Brazil for most of the past century.


Tax evasion can be a significant source of investment funds in Amazonian ranches.  Some of the ranchers who deforest the most are medical doctors and other professionals from urban areas.  People in such professions often have large incomes that they fail to declare.  Investing their money in the stock market or urban real estate they are likely draw the attention of tax authorities, but most of the investment in Amazonian ranches is of types that authorities have little basis for evaluation.  Even if the soil and rainfall regimes are unfavorable for pasture, resulting in some loss on the investment, money from beef sales from an Amazonian ranch will be “clean.”  As mentioned above, profits from drug trafficking, corruption and other illegal activities also provide a powerful motivation for investment in Amazonian ranches.  The government must invest in law-enforcement and in tightening the tracking of financial movements to eliminate this important driver of deforestation.


Deforestation also receives a strong impetus from subsidized agricultural credit.  The provision of agricultural credit in a given municipality and year is one of the factors with the strongest correlation to deforestation.  The government subsidy goes beyond low interest rates and generous grace periods.  There are also frequent “amnesties”, either forgiving debts or converting them to virtually token payments over long periods at low interest.  Amnesties are granted when production is reduced by droughts or other “acts of God.”  While usually viewed as “one-time” interventions, in fact they are a regular feature and represent a large additional subsidy to deforestation.


      A variety of other subsidies also increase the profitability of agriculture and ranching.  These include price supports for many agricultural products, with government guarantees of the price paid to the farmer irrespective of how distant from markets the farm may be.  Many special programs supply inputs such as fertilizer or lime to specific crops, and the vast network of transportation infrastructure, at government expense.


Land-Tenure and Settlement Policy Reform

The nature of settlements established by the National Institute of Colonization and Agrarian Reform (INCRA) has changed markedly over the years.  In the 1970s and 1980s most were placed in areas deliberately chosen by INCRA.,  Since the mid-1990s INCRA has claimed that new settlements are only sited in areas already deforested so as and in order to minimize their impact on deforestation. , since the mid-1990s INCRA has claimed that new settlements are only sited in areas already deforested.  Despite numerous official statements that such a policy was in effect, however, new settlements continued to be placed in forested areas, such as the Rio Acarí and Rio Juma settlements in the state of Amazonas, established in 1996.  More recently, INCRA has essentially ceded its role of determining settlement sites to squatter organizations such as the Landless Rural Workers’ Movement (MST); squatters invade either public land or the “legal reserves” (areas required to be kept forested) of large ranches, and INCRA subsequently “legalizes” the settlements when they are faits accomplis and compensates the ranchers for the lost land.  Because the compensation has generally been higher than the market price for land, some ranchers quietly encourage squatters to invade their land.  Bankrupt ranches undergoing foreclosure by the Banco do Brasil have been particularly prone to invasion – a situation that both assures squatters a resistance-free invasion and solves the financial problem of the Banco do Brasil when the compensation is paid by INCRA.  The areas chosen by squatters for invasion are invariably under primary tropical forest rather than pasture, agriculture or secondary forest.  The timber provides capital for the squatters and the soils are considerably better than could be expected in a degraded cattle pasture.  The de facto shift of INCRA activity to following in the path of initiatives by landless peasant organizations creates an additional barrier to effective control of this form of deforestation (Fearnside 2001b).


Although small farmers account for only about 30% of deforestation (Fearnside 1993a), the intensity of deforestation within the area they occupy is greater than for the medium and large ranchers holding 89% of the Legal Amazon's private land.  Deforestation intensity (the impact per km² of land) declines with increasing property size.  Deforestation would, therefore, increase if forest areas held by large ranches were redistributed into small holdings.  This emphasizes the importance of using already cleared areas for agrarian reform, rather than following the politically easier path of distributing areas still in forest.  Large as the area already cleared is, it falls far short of the potential demand for settlement.  Indeed, the Legal Amazon as a whole falls short of this demand.  Recognizing the existence of carrying capacity limits, and then maintaining population levels within these, is fundamental to any long-term plan for the sustainable use occupation of Amazonia (Fearnside 1997c).


Deforestation for cattle pasture is considered to be an "improvement" (benfeitoria) for the purpose of establishing and maintaining land title.  As long as this situation remains one can expect landholders will clear cut down their forest despite prohibitions.  A change in land-titling procedures to cease recognizing pasture as an improvement has yet to take place.


Agrarian reform is a particularly difficult area of policy affecting deforestation.  It is needed both in Amazonia and in the source areas of migrants outside of the region in order to stem the flow of people to new areas in search of land.  The availability of alternative employment both in rural and in urban areas is also related to these population flows.  At the same time, an “industry of settlement” has grown up where people who are granted land in one settlement area sell their claims (often informally if definitive land title has not yet been granted) and move on to try to obtain a parcel of land in a new settlement.  INCRA’s often-unsuccessful efforts to detect and disqualify those who have received land before only result in the creation of a permanently landless class that also contributes to deforestation.  In the end, the goal of provision of employment opportunities for all Brazilians will have to be met in ways that are less environmentally destructive than granting plots in Amazonian settlement areas (Fearnside 2001b).


Environmental Services

Economic activities in Amazonia almost exclusively involve material commodities – timber, minerals, the products of agriculture and ranching, and non-timber products such as natural rubber and Brazil nuts.  The potential is much greater, in terms of both monetary value and sustainability, for pursuing a radically different strategy for long-term support: finding ways to tap the environmental services of the forest as a means of both sustaining the human population and maintaining the forest (Fearnside 1997d, 2003b).


At least three classes of environmental services are provided by Amazonian forests: biodiversity maintenance, carbon storage, and water cycling.  The magnitude and value of these services are poorly quantified, and the diplomatic and other steps through which they might be compensated are also in their infancy – facts which do not diminish their importance nor the pressing need to focus effort on providing both the information and the political will required to integrate them into the economy in such a way that maintains rather than destroys the forest.  The role of tropical forests in averting global warming is much closer to serving as a basis for international financial flows than are other environmental services such as biodiversity maintenance.  This is because the United Nations Convention on Climate Change (UN-FCCC) has advanced further than the Convention on Biodiversity, even though both were signed simultaneously at the United Nations Conference on Environment and Development (UNCED) in Rio de Janeiro in 1992.  The UN-FCCC has been supplemented with the December 1997 Kyoto Protocol. 


Investment interest in carbon with a view to short-term returns is likely to be limited, given the fact that the agreement over the Kyoto Protocol reached in Bonn in July 2001 excludes credit for forest maintenance in the Clean Development Mechanism (CDM) during the Protocol’s first commitment period (2008-2012).  Should this be allowed in future commitment periods, Brazil could potentially gain substantially from CDM projects to reduce deforestation.  A proposal is also under discussion for the creation of a means whereby deforestation avoided now could generate credits after 2012 (Santilli et al. 2003).  The political struggles underlying this Bonn decision on the first commitment period can be expected to shift in the future because the “assigned amount” (national emissions quota) of each party is renegotiated for each successive commitment period, thereby removing the advantage to key actors (especially in Europe) of forcing parties (specifically the United States) to satisfy the commitments they made in Kyoto almost entirely through relatively expensive domestic measures (Fearnside 2001c).  The negotiations since the 1997 Kyoto conference have been unique because industrialized countries had agreed to specific assigned amounts (quotas) for the first commitment period before the rules were defined on such questions as inclusion of avoided deforestation in the Clean Development Mechanism.  For future commitment periods, inclusion of avoided deforestation would help induce countries to agree to larger commitments than they would accept in the absence of such a provision, and would therefore have a net benefit for climate.


The break with past inaction represented by the Bonn agreement could convince major investors, such as pension funds, to initiate or increase investment in long-term carbon ventures.  As global warming worsens and efforts to combat it become stronger and more universal, the carbon value of tropical forests can be expected to increase dramatically.  This is likely to happen, for example, by the end of a 30-year forest management cycle initiated now.


Although not currently favored by Brazil’s ministry Ministry of Eexternal Aaffairs, the country always has the option of accepting national limits on emissions that would allow it to earn much more by emissions trading under Article 17 of the Protocol, rather than through the CDM of Article 12 (Fearnside 2001d).  Emissions trading has substantially larger potential for carbon credit because the Kyoto Protocol does not require that the reductions be causally linked to a specific project.  It is also not required that the changes be “additional” to what would have occurred in a no-project scenario, the baseline for calculation being the country’s first national inventory (i.e., emissions in 1988-1994 in the case of Brazil).


Regardless of the future of decisions on the CDM under the Kyoto Protocol, global warming represents a long-term problem that is likely to gain urgency in the international policy arena as impacts become increasingly apparent to the public and to political leaders.  Sooner or later the major role played by tropical deforestation will be recognized, and appropriate measures, Brazilian and international, taken to finance combating deforestation, and to provide the basis for an alternative to destructive development.


One of the greatest impediments to effective action is fatalism.  Many believe that the forest will be cut down no matter what, and consequently argue that we should worry about other problems.  Fatalism acts as a deterrent to taking action that involves commitment of substantial financial resources and the acceptance of perceived or real political risks.  Many of the key determinants of the future path of development are in the hands of decision makers, who need to make their decisions based on the responsibility that this entails. While the future depends on human decisions, limits also exist.  We cannot go on destroying forests without dire and long-lasting consequences.




The author’s work is supported by the Conselho Nacional de Desenvolvimento Científico e Tecnológico National Council for Scientific and Technological Development (CNPq: Proc. 470765/01-1) and the National Institute for Research in the AmazonInstituto Nacional de Pesquisas da Amazônia (INPA: PPI 1 3620).


Literature Cited

Alencar, A., D. C. Nepstad, D. McGrath, P. Moutinho, P. Pacheco, M. del C. V. Diaz, and B. Soares-Filho. 2004. Desmatamento na Amazônia: iIndo aAlém da eEmergência cCrônica. Instituto de Pesquisa Ambiental da Amazônia (IPAM), Belém, Pará, Brazil.. 87pp.

Barbosa, R. I., and P. M. Fearnside. 1999. Incêndios na Amazônia brasileira: eEstimativa da emissão de gases do efeito estufa pela queima de diferentes ecossistemas de Roraima na passagem do evento "El Niño" (1997/98). Acta Amazonica 29(4): 513-534.

Brazil, INPE. 2004. Monitoramento da fFloresta aAmazônica bBrasileira por sSatélite/Monitoring of the Brazilian Amazon fForest by sSatellite: 2002-2003. Instituto Nacional de Pesquisas Espaciais (INPE), São José dos Campos, São Paulo, Brazil.

Brazil, MCT. 2004. Brazil’s National Communication to the United Nations Framework Convention on Climate Change. General Coordination on Global Climate Change, Ministry of Science and Technology. (MCT), Brasília, Brazil. 271pp.

Brooks, T. and A. B. Rylands. 2003. Species on the brink: Critically endangered terrestrial vertebrates. In: The Atlantic Forest of South America: Biodiversity Status, Threats and Outlook, C. Galindo-Leal & I. de G. Câmara (eds.), pp.360-371. Island Press, Washington, DC.

Capobianco, J. P. R., A. Veríssimo, A. Moreira, I. dos Santos, L. P. Pinto, and D. Sawyer (eds.).  2001. Biodiversidade na Amazônia Amazônia bBrasileira. Editora Estação Liberdade & and Instituto Socioambiental, São Paulo. 544pp. Available from ( (accessed November, 2004)).

Cochrane, M. A. 2003. Fire science for rainforests. Nature, London 421:913-919.

Cochrane, M. A., A. Alencar, M. D. Schulze, C. M. Souza Jr., D. C. Nepstad, P. Lefebvre, and E. A. Davidson. 1999. Positive feedbacks in the fire dynamic of closed canopy tropical forests. Science 284: 1832-1835.

Faminow, M. D. 1998. Cattle, Deforestation and dDevelopment inn the Amazon: aAn eEconomic and eEnvironmental pPerspective. CAB International, New York. 253pp.

Fearnside, P. M. 1987.  Causes of dDeforestation in the Brazilian Amazon. pp.Pages 37-61 iIn: R. F. Dickinson,  (editor. .) The gGeophysiology of Amazonia: vVegetation and Climate climate iInteractions. John Wiley & Sons, New York. 526pp.

Fearnside, P. M. 1993a. Deforestation in Brazilian Amazonia: The effect of population and land tenure. Ambio 22(8): 537-545.

Fearnside, P. M. 1993b. Desmatamento na Amazônia: qQuem tem razão nos cálculos – o INPE ou a NASA? Ciência Hoje 16(96): 6-8.

Fearnside, P. M. 1997a. Limiting factors for development of agriculture and ranching in Brazilian Amazonia. Revista Brasileira de Biologia 57(4): 531-549.

Fearnside, P. M. 1997b. Greenhouse gases from deforestation in Brazilian Amazonia: nNet committed emissions. Climatic Change 35(3): 321-360.

Fearnside, P. M. 1997c. Human carrying capacity estimation in Brazilian Amazonia as a basis for sustainable development. Environmental Conservation 24(3): 271-282.

Fearnside, P. M. 1997d. Environmental services as a strategy for sustainable development in rural Amazonia. Ecological Economics 20(1): 53-70.

Fearnside, P. M. 1999. Biodiversity as an environmental service in Brazil's Amazonian forests: rRisks, value and conservation. Environmental Conservation 26(4): 305-321.

Fearnside, P. M. 2000a.  Greenhouse gas emissions from land-use change in Brazil's Amazon region. pp.Pages 231-249 iIn : R. Lal, J. M. Kimble and& B. A. Stewart, (editors.s.) Global cClimate cChange and Tropical tropical eEcosystems. Advances in Soil Science. CRC Press, Boca Raton, Florida. 438pp.

Fearnside, P. M. 2000b. Global warming and tropical land-use change: gGreenhouse gas emissions from biomass burning, decomposition and soils in forest conversion, shifting cultivation and secondary vegetation. Climatic Change 46 (1/2): 115-158.

Fearnside, P. M. 2001a. Soybean cultivation as a threat to the environment in Brazil. Environmental Conservation 28(1): 23-38.

Fearnside, P .M. 2001b. Land-tenure issues as factors in environmental destruction in Brazilian Amazonia: The case of southern Pará. World Development 29 (8): 1361-1372.

Fearnside, P. M. 2001c. Saving tropical forests as a global warming countermeasure: aAn issue that divides the environmental movement. Ecological Economics 39(2): 167-184.

Fearnside, P. M. 2001d. The potential of Brazil's forest sector for mitigating global warming under the Kyoto Protocol. Mitigation and Adaptation Strategies for Global Change 6(3-4): 355-372.

Fearnside, P. M. 2002a. Can pasture intensification discourage deforestation in the Amazon and Pantanal regions of Brazil?  pp.Pages 283-364 iIn: C. H. Wood and R. Porro,  (editors.s.) Deforestation and lLand uUse in the Amazon. University Press of Florida, Gainesville, Florida.


Fearnside, P. M. 2002b. Avança Brasil: eEnvironmental and social consequences of Brazil’s planned infrastructure in Amazonia. Environmental Management 30(6): 748-763.

Fearnside, P. M. 2003a. Deforestation control in Mato Grosso: aA new model for slowing the loss of Brazil’s Amazon forest. Ambio 32(5): 343-345.

Fearnside, P. M. 2003b. Conservation policy in Brazilian Amazonia: understanding the dilemmas. World Development  31(5): 757-779.

Fearnside, P. M. 2004. A água de São Paulo e a floresta amazônica. Ciência Hoje 34(203): 63-65.

Fearnside, P. M., & and R. I. Barbosa. 2003. Avoided deforestation in Amazonia as a global warming mitigation measure: tThe case of Mato Grosso. World Resource Review 15(3): 352-361.

Fearnside, P. M., and R. I. Barbosa. 2004. Accelerating deforestation in Brazilian Amazonia: tTowards answering open questions. Environmental Conservation 31(1): 7-10.

Fearnside, P. M., and W. F. Laurance. 2004. Tropical deforestation and greenhouse gas emissions. Ecological Applications 14(4): 982-986.

Hecht, S. B., R. B. Norgaard, and C. Possio. 1988. The economics of cattle ranching in eastern Amazonia. Interciencia 13(5): 233-240.

Kaimowitz, D.,  B. Mertens, S. Wunder,  and P. Pacheco. 2004.  Hamburger connection fuels Amazon destruction. Technical Report, Center for International Forest Research (CIFOR), Bogor, Indonesia . 10 pp. Available from (accessed November 2004).

Laurance, W.F., M. A. Cochrane, S. Bergen, P. M. Fearnside, P. Delamônica, C. Barber, S. D’Angelo, and T. Fernandes. 2001. The Future future of the Brazilian Amazon. Science 291: 438-439.

Lean, J., C. B. Bunton, C. A. Nobre, and P. R. Rowntree. 1996. The simulated impact of Amazonian deforestation on climate using measured ABRACOS vegetation characteristics. Pages  pp. 549-576 iIn: J. H. C. Gash,  C. A. Nobre, J. M. Roberts and R. L. Victoria,  (editors.s.)  Amazonian dDeforestation and cClimate. Wiley, Chichester, United Kingdom.K. 611pp.

Mahar, D. J. 1979. Frontier Development development pPolicy in Brazil: aA sStudy of Amazonia. Praeger, New York., 182pp.

Margulis, S. 2003. Causas do desmatamento na Amazônia brasileira. Banco MundialThe World Bank, Brasília, Brazil.. 100pp.Available from http://www/ (accessed November 2004).

Mattos, M. M., and C. Uhl. 1994. Economic and ecological perspectives on ranching in the eEastern Amazon. World Development 22(2): 145-158.

Mittermeier, R. A., G. A. B. da Fonseca, A. B. Rylands, and& C. G. Mittermeier. 1999. Atlantic Forest. Pages  136-147 inIn: R. A. Mittermeier, N. Myers, P. Robles Gil and C. G. Mittermeier, editors. Hotspots: Tthe Earth's earth's Biologically biologically rRichest and mMost eEndangered tTerrestrial rEcoregions, R. A. Mittermeier, N. Myers, P. Robles Gil & C. G. Mittermeier (eds.), pp.136-147. CEMEX, Agrupación Serra Madre, SC, Mexico.

Mittermeier, R. A., C. G. Mittermeier, P. Robles Gil, J. Pilgrim, G. A. B. da Fonseca, T. Brooks, and W. R. Konstant ((eds.) 2002. Wilderness: Earth’s earth’s lLast wWild plPlaces. CEMEX, Agrupación Serra Madre, SC, Mexico.

Nepstad, D., G. Carvalho, A. C. Barros, A. Alencar, J. P. Capobianco, J. Bishop, P. Moutinho, P. Lefebvre, U. L. Silva, Jr., and E. Prins. 2001. Road paving, fire regime feedbacks, and the future of Amazon forests. Forest Ecology and Management 154(3): 395-407.

Nepstad, D. C., P. Lefebre, U. L. da Silva, J. Tomasella,, P. Schlesinger, L. Solórzano, P. Moutinho, D. Ray, and J. G. Benito. 2004. Amazon drought and its implications for forest flammability and tree growth: A a basin-wide analysis. Global Change Biology 10(5): 704-717.

Nepstad, D. C., A. G. Moreira, and A. A. Alencar. 1999a. Flames in the Rain rain fForest: oOrigins, Impacts and aAlternatives to Amazonian fFires., The World Bank, Brasília, Brazil..

Nepstad, D. C., A. Veríssimo, A. Alencar, C. Nobre, E. Lima, P. Lefebvre, P. Schlesinger, C. Potter, P. Moutinho, E. Mendoza, M. Cochrane, and V. Brooks. 1999b. Large-scale impoverishment of Amazonian forests by logging and fire. Nature, London 398: 505-508.

Nogueira, E. M., B. W. Nelson, and P. M. Fearnside. nd.In press. Wood density in dense forest in central Amazonia, Brazil. Forest Ecology and Management:   (in press).

Salati, E., A. Dall'Olio, E. Matusi, and J. R. Gat. 1979.  Recycling of water in the Brazilian Amazon Basin: An isotopic study. Water Resources Research 15: 1250-1258.

Salati, E., and P. B. Vose. 1984.  Amazon Basin: A a system in equilibrium. Science 225: 129-138.

Santilli, M., P. Moutinho, S. Schwartzman, D. Nepstad, L. Curran and C. Nobre,. 2003. Tropical Deforestation and the Kyoto Protocol: a new proposal, Paper presented at COP-9, 1st – 12th December 1-12, 2003, Milan, Italy Italy. 6pp.Available at  ( (accessed November 12, 2004).).

Schneider, R. R., E. Arima, A. Veríssimo, P. Barreto, and C. Souza Junior. 2000. Amazônia sSustentável: lLimitantes e oOportunidades para o dDesenvolvimento rRural. Banco MundialThe World Bank, Brasília,  DF and Instituto para o Homem e o Meio Ambiente na Amazônia (IMAZON), Belém, Brazil.. 58pp.

Veríssimo, A., P. Barreto, M. Mattos, R. Tarifa,  and C.  Uhl. 1992. Logging impacts and prospects for sustainable forest management in an old Amazonian Frontier: tThe case of Paragominas. Forest Ecology and Management 55: 169-199.
Figure Legend


Figure 1.  Cumulative deforested area and annual deforestation rate in Brazilian Amazonia.  Data from Brazil, INPE (, 2004)  (except for 1978 (: see Fearnside 1993b).