Environmental Conservation 32 (2): 149–155 © 2005 Foundation for Environmental Conservation doi:10.1017/S0376892905002134

Conservation units: a new deforestation frontier in the Amazonian stateof Rondonia, Brazil

M.A. PEDLOWSKI 1 * , E .A.T. MATRICARDI 2 , D . SKOLE 2 , S .R . CAMERON 2 ,W. CHOMENTOWSKI 2 , C . F E R N A N D E S 3 AND A. LIS BO A 3

1Universidade Estadual do Norte Fluminense, Avenida Alberto Lamego, 2000, Laboratorio de Estudos do Espaco Antropico, Centro de Ciencias doHomem, Universidade Estadual do Estadual do Norte Fluminense,Campos dos Goytacazes, RJ, 28013-602, Brazil, 2Center for Global Change andEarth Observations, Michigan State University, 1405 S. Harrison Road, Room 101, East Lansing, MI 48823, USA, and 3Secretaria de Estado doDesenvolvimento Ambiental de Rondonia, Estrada de Santo Antonio, 900 Parque Cujubim, Porto Velho, RO, 78900-000, BrazilDate submitted: 17 August 2004 Date accepted: 18 May 2005

SUMMARY

Over the past several decades, the Brazilian Stateof Rondonia has been the destination of many ruralmigrants drawn from Brazil’s middle southern regionsby massive government colonization projects. Factorssuch as explosive population growth, logging, mining,small-scale farming and ranching have synergisticallyfuelled deforestation in the state. The total areadeforested in Rondonia in 1978 was 4200 km2. In1988, the area increased to 30 000 km2, in 1998 to53 300 km2 and by the year 2003, a total of 67 764 km2

of Rondonia was deforested. In response to the highrate of deforestation observed in Rondonia and otherAmazonian states, state and federal agencies workedto create a network of conservation units (CUs) inBrazil during the 1990s that was signed into law(Law 9985/00) in 2000. The ability of these CUs to reducethe rate of deforestation was analysed. Remotely-sensed data from Landsat and thematic coverageswere used to measure deforestation inside all CUs inRondonia. A more detailed analysis of CUs with thehighest levels of deforestation, including an analysisbetween soil types and deforestation and a forecastof potential future deforestation, was conducted. Thecreation of conservation units in Rondonia has beenuseful in curbing deforestation within their boundaries;however, many CUs face pressure from the combinedactivities of illegal loggers, cattle ranchers and small-scale farmers seeking new sources of timber andagricultural land. For example, an exponential increasein the amount of deforestation was observed inRondonia’s Bom Futuro National Forest between 1992and 2000. A regression model indicated a total of20 500 ha deforested by 2002, while measurements from2002 imagery showed an actual total deforestationof 20 720 ha. Should this trend persist, Bom FuturoNational Forest could be completely deforested by 2017.CUs in Rondonia must be developed and implemented

* Correspondence: Dr M.A. Pedlowski Telephone: + 55 22 27261583 e-mail: pedlowma@uenf.br

jointly by all stakeholders through the creationof partnerships between local communities, non-governmental organizations and government agencies.

Keywords: Brazilian Amazon, biodiversity, conservation,conservation units, ecology, remote sensing, tropicaldeforestation

INTRODUCTION

The conversion of tropical rainforests to pasture andcropland has become a serious environmental and social issuewith important global ramifications that include social andbiological impoverishment and an important contribution toglobal warming (Houghton 1991; Brown et al. 1992). At thesame time, the advance of deforestation in tropical countriesis being followed by the realization that the biodiversity thatexists in these forests must be preserved (Gallegos 1997).Many efforts have, however, been tainted by the lack of soundimplementation of legally-created unidades de conservacao orconservation units (CUs) and, as such, have become thesubject of harsh criticisms (Redford et al. 1998). Yet despiteconcern about the effectiveness of CUs in the tropics, growingevidence indicates that they have been surprisingly useful toolsin curbing deforestation (Bruner et al. 2001).

The Brazilian Amazon attracts special attention owingto widespread environmental degradation that arose in partfrom the implementation of public policies aimed at rapidlydeveloping the region (Hecht 1985; Browder 1988; Mahar1989). The region has had the highest detected rates ofdeforestation in the tropical world. Satellite data indicate thatbetween 1978 and 1988, the average rate of deforestation was1.6–2.2 × 106 ha yr−1 (INPE [Instituto Nacional de PesquisasEspaciais] 1992; Skole & Tucker 1993). Total deforestationin the Brazilian Amazon increased from approximately37.7 × 106 ha in 1988 to 58.7 × 106 ha by 2000, at an averageannual rate of 1.75 × 106 ha yr−1 (INPE 2004). Laurance et al.(2001) demonstrated that deforestation pressures have notdiminished in the Brazilian Amazon and note changes inthe spatial patterns of deforestation events that may causeaccelerated loss of pristine forest across the region.

150 M.A. Pedlowski et al.

In response to such widespread evidence of environmentaldegradation, there is an ongoing effort to protect the region’snatural ecosystems (Moran 1993) with the establishment ofprotected areas identified as a main strategy for conservation(Uhl 1988). Nepstad et al. (2002) have argued that a regionalconservation strategy based on the creation of CUs is urgentlyneeded to protect the Amazon. Indeed, within the BrazilianAmazon, the federal government began the creation of CUs in1959, with the creation of Araguaia National Park, in the stateof Tocantins (Rocha 1992). More recently, in conjunctionwith the World Bank, the Global Environmental Facility andWorld Wide Fund for Nature (WWF), Brazil committed totripling the amount of protected areas within the Amazon by2012 through the Amazon Region Protected Areas (ARPA)programme (World Bank 2002).

In July 2000, Brazil established its National System ofConservation Units through Federal Law 9985/2000 anddesignated two basic categories of CUs, namely IntegralProtection Conservation Units (IPCUs) and SustainableUse Conservation Units (SUCUs) (Rocco 2002). IPCUsare CUs in which the use of natural resources is strictlyforbidden, and include areas such as national parks, ecologicstations, biologic reserves and wildlife sanctuaries. SUCUsare CUs where the sustainable use and management ofnatural resources is allowed, and include CUs such as nationalforests, extractive reserves and sustainable developmentreserves. Given the existing pressure for the use ofAmazonian natural resources, Verıssimo et al. (2002) haveemphasized the potential contribution of national forests toestablishment of a more rational use of natural resources in theAmazon.

In the state of Rondonia, a consortium of state and federalagencies is working with funds from a World Bank loan toconsolidate a network of CUs in the hopes of curbing thedestruction of the state’s remaining rainforest. This effortwas a direct response to national and international pressuresto ease the alarming pace of deforestation associated withland-use and land-cover change resulting from colonizationprojects established in the region in the early 1970s (Coy 1987;Malingreau & Tucker 1988; Millikan 1988). Remote sensingdata from Landsat were used to assess land cover change insideCUs in Rondonia over a period of seven years. Though manyCUs have experienced more cryptic forms of degradation suchas those brought about by poaching (Peres & Lake 2003) orselective logging and fire (Nepstad et al. 1999), the currentresearch is directly concerned with deforestation. We focusour attention on the case of the Bom Futuro National Forest,where the highest rates of deforestation for the period 1996–1999 were found, to examine possible relationships betweendeforestation, soil types, and the presence of squatters andillegal loggers. Our ultimate goal is to determine whether thesocioeconomic and political reality of the Amazon requiresbroad programmes to strengthen the institutional capabilityof environmental agencies and to regulate the use of naturalresources by different stakeholders to ensure that successfollows the creation of CUs.

The state of Rondonia is located in the western portionof the Brazilian Amazon and contains 243 044 km2 (anarea slightly larger than Portugal). The first settlements inRondonia occurred during two rubber booms (the first in thenineteenth century and the second during the Second WorldWar), but neither had a lasting impact on the occupation ofthe state (Martine 1990). The modern population rush toRondonia, which began in the late 1960s, was caused by theabandonment of colonization projects along the TransamazonHighway and by changes in land tenure patterns in mid-southern Brazil (Martine 1990). More recently, Rondonia hasbeen characterized by a consolidation of land holdings and thespread of pasture at the expense of forested areas (Ferraz et al.2005).

In 1987, a World Bank mission went to Rondonia tobegin negotiations for the Projeto Agropecuario e Florestal deRondonia (PLANAFLORO), the Rondonia Natural ResourceManagement Programme. PLANAFLORO addressed majorproblems caused by previous development projects andincluded a series of goals related to environmental protection.The demarcation of CUs and Amerindian reserves was akey aspect of PLANAFLORO’s resource-use component(World Bank 1997). On many occasions mixed tenure patternsresulted in great difficulties for the legal establishment,demarcation and transfer of land ownership between distinctgovernmental spheres and private individuals (Umana 1998).During this process, ranchers and colonists encroached onareas known to be targeted for conservation, the end resultbeing that only 40.3% of the 5 355 000 ha originally intendedfor inclusion in CUs were actually demarcated (Millikan1998).

Most of Rondonia’s CU natural resources managementplans, which are legal documents defining the permitted usesof the natural resources within a given CU, were developedduring PLANAFLORO (Olmos et al. 1999). However, mostof the areas have no permanent staff and only sporadicprotective measures have been taken to enforce their limits.This situation is similar to that observed by Fearnside andFerreira (1984) during previous development programmes,where the creation of CUs was not followed by a clearcommitment to protection and management. In spite ofthese problems, PLANAFLORO and the funds provided bythe World Bank resulted in the consolidation of a networkincluding 51 CUs in both IPCU and SUCU categoriesthat include national, state and municipal entities (Table 1).In addition to the creation of CUs, the demarcation of 22Amerindian reserves has formally placed 35% of Rondonia’sterritory under protection (Fig. 1).

METHODOLOGY

Sixty-four Landsat scenes covering the state of Rondonia forthe years 1992, 1996 and 1999 were used to calculate thedeforested area within 51 CUs (Fig. 1) at the Center of GlobalChange and Earth Observations (CGCEO) at Michigan StateUniversity. Eight additional scenes (path/row 232/067 and

Land use and land cover change analysis 151

Table 1 Number of IPCUs andSUCUs in different categories ofconservation units in the Rondoniaconservation units network. 1Aportion (∼95 km2) of the JaruBiological Reserve overlaps part ofthe Igarape Lourdes IndigenousTerritory. 2The total area (687476 km2) of the Pacaas NovosNational Park overlaps part of theUru-Eu-Wau-Wau IndigenousTerritory. Source: SEPLAN(1999).

Category IPCU SUCU

Quantity Area (ha) Quantity Area (ha)

Ecological station 4 195 321 – –Biological reserve1 3 550 787 – –Municipal park 4 1329 – –State park 3 672 035 – –National park2 1 687 476 – –Extractive reserves – – 25 1 333 262State forest – – 9 267 251National forest – – 2 483 397Total 15 2 106 948 36 2 083 910

Figure 1 Diagrammatic map ofconservation units and Landsat tilesystem for the state of Rondonia in theBrazilian Amazon.

232/066), acquired for 1998, 1999, 2000 and 2002 were usedto perform a more detailed analysis of Bom Futuro NationalForest. Landsat 5 Thematic Mapper satellite images wereacquired for 1992, 1996 and 1998 and Landsat 7 EnhancedThematic Mapper+ images were acquired for 1999, 2000 and2002.

All images were geometrically corrected by using nearest-neighbour resampling with the four points derived fromephemeris data supplied with the images at the time ofordering. Additionally, the corrected images were validatedand tested using collected GPS ground points from severallocations in the Amazon. Image classification proceduresfor each year were carried out in ERDAS Imagine usingan unsupervised classification of Digital Numbers (DNs) tocreate clusters of distinct land uses. Once spatial clusters weregenerated, interpreters determined the nature of the clustersand provided a classification. The final deforestation layers or

grids contained seven classes, namely forest, deforestation,cerrado, secondary forest, water body, cloud and cloudshadow. Forest degradation that results from logging and firewas not included in this analysis as it cannot be accuratelydetected using unsupervised techniques (Nepstad et al. 1999;Asner et al. 2002).

A map containing the CUs’ limits at a scale of 1:250 000was used as a cartographic base for measurements. Inaddition, deforestation grids produced by CGCEO weretabulated according to the CUs map. Total deforestation inCUs was calculated for the period 1992 to 1999 using thedeforestation grid information. In areas where we observedhigher levels of deforestation from 1992 to 1999, we useda time-series for 1992, 1996 and 2000 to calculate thetemporal variation in deforestation. In the case of BomFuturo National Forest, deforestation was measured for1992, 1996, 1998, 1999, 2000 and 2002 and an additional

152 M.A. Pedlowski et al.

Table 2 Deforestation in sixconservation units in the state ofRondonia (1992, 1996 and 1999).

Conservation unit Total CU area (ha) Deforestation

1992 1996 1999

ha % ha % ha %

Bom Futuro National Forest 269 144 413 0.15 2230 0.83 6906 2.57Corumbiara State Park 433 157 899 0.21 899 0.21 1493 0.34Jaci-Parana State Extractive Reserve 200 635 267 0.13 290 0.14 480 0.24Jamari National Forest 216 101 5538 2.56 5660 2.62 6581 3.05Ouro Preto Federal Extractive Reserve 207 450 2849 1.37 7044 3.4 8964 4.32Pacaas Novos National Park 690 718 69 0.01 411 0.06 864 0.13

information layer, soil composition as derived from soil mapsproduced in 1999 by the Rondonia Planning Secretariat(SEPLAN [Secretaria de Estado do Planejamento eCoordenacao Geral] 1999) was used to investigate the possiblelinks between deforestation and endogenous environmentalvariables.

We also conducted fieldwork during June 2002 thatincluded field observations and GPS ground control pointmeasurements in Bom Futuro National Forest. We comparedthe areas mapped on the satellite imagery with the fieldobservations around the roads. Ground truthing of imageryclassification was used to reclassify areas that were originallymisclassified, such as several small areas of rock exposures thatwere provisionally classified as deforestation.

RESULTS

The percentage of the total area of CUs in Rondonia (roughly4 200 000 ha) classified as deforested increased to roughly0.7% from 1992 to 1999. This measure may underestimate theactual amount of deforestation because of the exclusion of areascovered by cloud and cloud shadow (around 0.1% per year)in the imagery. Approximately 19 000 ha of new deforestationwere observed between 1992 and 1999. Although the area ofdeforestation can be regarded as relatively low, the total areadeforested increased 350% over this period. A large portion ofthis deforestation, 64% in 1999, is concentrated in only six ofthe state’s 51 CUs (Table 2). Five of the 51 CUs investigatedshowed no increase in deforested area over the study period.

The highest absolute levels of deforestation were foundin Ouro Preto Federal Extractive Reserve, Jamarı NationalForest and Bom Futuro National Forest, with 4.32%,3.05% and 2.57% of total area deforested by 1999,respectively. These CUs also exhibited the greatest increasein deforestation over the same period. In Ouro Preto FederalExtractive Reserve and Jamarı National Forest, deforestationis attributed to agents operating legally within the confines ofeach CU’s natural resources management plan; rubber tappersin the case of the former and cassiterite miners in the latter.The process of deforestation in the Bom Futuro NationalForest is the result of activities by exogenous and illegal agents(i.e. squatters, cattle ranchers and loggers).

0

2000

4000

6000

8000

10000

Year

Def

ores

ted

area

(ha

)

1992 1994 1996 1998 2000

Figure 2 Plot of deforested area over time 1992–2000.

Bom Futuro National Forest

Approximately 9000 ha of Bom Futuro National Forest weredeforested between 1992 and 2000. The total deforested arearemains relatively low, but the rate of deforestation may begrowing at an exponential rate (Fig. 2). A curve fitted tothe data yielded the exponential equation y = 422.69e0.388x,where y is the amount of deforestation (ha) and x isyears since 1992 (R2 = 0.9922, p < 0.001). Imagery for 2002indicated a total area deforested of 20 720 ha, while the model(y = 422.69e0.388x) predicted a total of 20 500 ha, under-estimating total deforestation by only 1%.

One possible explanation for the situation observed inBom Futuro National Forest is that relatively fertile soilswith agricultural potential have made it a target of choiceamong those seeking to gain access to new land. In 1996,deforestation was disproportionately located on low fertilitydystrophic cambisols (Table 3). By 1998 and continuingto 2000, deforestation occurred predominantly on the morefertile yellow dystrophic latosols, yellow dystrophic podzolsand in particular red-yellow dystrophic latosols (Table 3).

DISCUSSION

Though experiencing relatively low levels of deforestationin comparison to the state as a whole, CUs in Rondoniaface increasing pressure from the incursion of illegal loggersand squatters. Only 10 of the 51 areas studied here had no

Land use and land cover change analysis 153

Table 3 Soil types anddeforestation in Bom FuturoNational Forest. Source:SEPLAN(1999).

Soil type Total area (ha) Deforestation in soil type subset

1996 1998 2000

ha % ha % ha %

Dystrophic cambisols 13 144.97 324.1 2.47 324.1 2.47 324.1 2.47Gley dystrophic 3400.47 0 0 0.2 0.00 5.4 0.16Yellow dystrophic latosols 124 574.95 816.5 0.66 893.9 0.72 2593.9 2.08Red-yellow dystrophic latosols 60 218.67 974.5 1.62 1994.5 3.31 5252.1 8.72Yellow dystrophic podzols 65 997.38 114.8 0.17 450.5 0.68 1269.4 1.92Sandy soils 832.14 0 0 0 0 0 0Total 268 168.58 2230.9 3663.2 9445.0

settlements detected within their borders. Where these agentsgain a foothold, such as in Bom Futuro National Forest, theiractions can lead to an exponential increase in deforestation, atrend often observed in newly opened frontiers. Owing to thecomplex process involved in the creation of CUs in Brazil, itcan take years or even decades for a targeted area to becomeeligible for full legal protection. During this time, incipientCUs are vulnerable to invasion. The reduction in the areaof CUs created under PLANAFLORO was largely causedby the exclusion of areas that had been invaded during thisprocess. Fearnside (2003) notes that after the formation ofBrazil’s national system of conservation units (SNUC), delaysin legislation provided loopholes that were exploited by manygroups.

Within Rondonia’s network of CUs, areas experiencing thelowest levels of disturbance have been those most isolatedowing to lack of access via the expanding road network.Proximity to previously deforested areas has been shownto be a good predictor of future deforestation (Alves et al.1999; Alves 2002). A possible explanation for the bulk ofdeforestation being present in a few CUs is their proximityto old settlements where land aggregation into large estateshas expelled small farmers from their properties. Landlessfarmers tend to form alliances with illegal loggers to generateincome needed to meet their material needs and gain accessto new land. The interaction between these social agentsresults in the construction of illegal roads for the extractionof timber resources, encroaching on the CUs with networksof access roads (Fig. 3). This also provides the opportunityfor squatters to establish themselves within CU boundaries.The results of this study confirm the notion that proximity toactively deforested areas increases the chance of a CU sufferingencroachment and consequently having its natural resourcesexploited by a multitude of social actors.

In Bom Futuro National Forest, visual analysis of satelliteimages indicates patterns of land-cover change that reflectthe presence of land-use systems adopted by small farmersand cattle ranchers inside the area. According to Brazil’sColonization and Agricultural Reform Agency (InstitutoNacional de Colonizacao e Reforma Agraria, INCRA), mostlandholders within Bom Futuro were not living permanentlyinside the area (INCRA 2002). However, 550 families claimed

Figure 3 Roads and conservation units in Rondonia.

land within the national forest and had cleared roughly 9900 hafor agriculture (INCRA 2002). Half of the families withagricultural plots were living in nearby urban areas, with15% of the families being only part-time dwellers and 35%permanent residents. Typically, illegal logging agents enterthe area and establish verbal contracts with the squatters toexploit the timber resources under their control, providingincome and capital to reinvest in further forest clearing.

Similar to Jones et al. (1995), we did not find a directrelationship between deforestation and soil types. However,given the relatively low fertility of the area’s soil types, wepredict that pasture will become the dominant use of landcleared for agriculture in Bom Futuro National Forest. Wenote that a growing demand for pasture already exits, and thatthere are many areas within the CU that are susceptible toillegal deforestation. Although Bom Futuro National Foresthas neither had its demarcation concluded, nor a naturalresources management plan in place, most squatters realizethe illegality of their presence inside the area. This senseof illegality creates a climate of uncertainty about the futurethat may serve as an additional incentive for the rapid, andconsequently unsustainable, exploitation of natural resources

154 M.A. Pedlowski et al.

Table 4 Evolution of timber processing companies in fourmunicipalities of Rondonia from 1992 to 2002. Source: IBAMA(2002).

Year Alto Paraıso Buritis Campo Novo Ariquemes

1992 2 1 1 471996 2 2 3 621999 6 26 7 852002 10 51 8 107

in Bom Futuro National Forest as illegal agents seek to ‘mine’all the resources that they can while the opportunity to do soexists.

We observed illegal selective logging activities inside BomFuturo National Forest, with loggers using heavy machinerysuch as skidders, bulldozers and large trucks to log and quicklymove out of the area. There was a steady growth in permitsissued by the Brazilian Institute for the Environment andRenewable Natural Resources (Instituto Brasileiro do MeioAmbiente e dos Recursos Naturais Renovaveis, IBAMA) forlogging and sawmill operations from 1992 to 2002 in fourmunicipalities in the vicinity of Bom Futuro National Forest(Table 4). Although our fieldwork did not establish a directlink between illegal logging and legally established timbercompanies, we believe that there is a synergy between illegalloggers, squatters and the wood sector in the surroundingmunicipalities. This belief is reinforced by the fact that Buritisand Alto Paraiso are respectively located at the southernand eastern limits of Bom Futuro and have become focalpoints for timber sawing in Rondonia. Additionally, weobserved sawmills operating within Bom Futuro NationalForest without proper authorization from IBAMA. We expectthat these clandestine operations are likely to expand theirinfluence to neighbouring Jamari Extractive Reserve in thenear future.

A final explanation for the deforestation observedwithin Rondonia’s CUs and, in particular, Bom FuturoNational Forest is the lack of institutional capacity toeffectively monitor and administer these areas. Agenciesresponsible for enforcing Brazil’s environmental laws areoften characterized as lacking the institutional strength orthe resources needed to achieve their goals (Rocha 1992;Peres & Terbough 1995; WWF 1999). Oliveira (2002)notes that the success of conservation programmes in theBrazilian state of Bahia hinged on the successful formationof alliances between federal, state and municipal agencies inconjunction with non-governmental organizations and localinterest groups. As the objectives of different stakeholdersare often at odds with each other, finding ways to forge newcoalitions is an important step in promoting more effectiveconservation within the region.

CONCLUSIONS

CUs have slowed the pace of deforestation in areas ofpristine forest in Rondonia; however, the borders between

CUs and settled areas are permeable, and increasing demandsfor pasture and new areas for logging are putting pressureon the protected resources. The creation of CUs alone isnot enough to stop land clearing for agriculture and otherforms of resource exploitation. Areas juxtaposed with previoussettlements are at the greatest risk, as the demands beingplaced on a shrinking resource base push further afield. Thissituation is compounded by the fact that CUs are normallylocated in areas rich in natural resources and are easy targetsfor illegal agents. In Rondonia, the bulk of land clearing isoccurring either in areas where old settlements previouslyexisted, or where unofficial new settlements have been createdrecently. Although we did not find that soil compositionhelped to explain deforestation in CUs, we argue that there isa gap in the knowledge about how such environmental factorsserve to enhance or preclude the occurrence of deforestationon protected lands. A fact that may further complicatethe future of conservation efforts in Rondonia is the lackof management plans for most of the newly-created CUs.Moreover, only three CUs have the required staff to conductthe routine tasks necessary to protect their territory.

Bom Futuro National Forest presents a worst-case scenarioof what can happen in the absence of sufficient institutionalcapacity to manage protected resources. As timber andagricultural land become scarce, greater pressure will bebrought to exploit protected resources. Consolidation ofRondonia’s CUs under PLANAFLORO is a first step towarda more effective management of these areas. Future effortsshould be aimed at better management and sustainable useof resources both within CUs and without. As a SUCU,Bom Futuro National Forest is open to sustainable resourceuse, including logging, provided that it is done under anapproved management plan. Involvement of the diverse groupof stakeholders including local, state and federal agencies,as well as private logging firms and local communities informulating management plans for all CUs should be a toppriority lest the case of Bom Futuro set a precedent for otherareas.

References

Alves, D.S., Pereira, J.L.G, de Sousa, C.L., Soares, J.V. &Yamaguchi, F. (1999) Characterizing landscape changes in centralRondonia using Landsat TM imagery. International Journal ofRemote Sensing 20(14): 2877–2882.

Alves, D.S. (2002) Space-time dynamics of deforestation in BrazilianAmazonia. International Journal of Remote Sensing 23(14):2903–2908.

Asner, G.P., Keller, M., Pereira, R. & Zweed, J.C. (2002) Remotesensing of selective logging in Amazonia: assessing limitationsbased on detailed field observations, Landsat ETM + , and texturalanalysis. Remote Sensing of Environment 80: 483–496.

Browder, J.O. (1988) Public policies and deforestation in theBrazilian Amazon. In: Public Policies and the Misuse of ForestResources, ed. M. Gillis & R. Repetto, pp. 247–298. London, UK:Cambridge University Press.

Land use and land cover change analysis 155

Brown, I.F., Nepstad, D.C., Pires, I.O., Luz, L.M. & Alexandre,A.S. (1992) Carbon storage and land-use in extractive reserves inAcre, Brazil. Environmental Conservation 19(4): 307–315.

Bruner, A.G., Gullison, R.E., Rice, R.E. & Fonseca, G.A.B. (2001)Effectiveness of parks in protecting tropical biodiversity. Science291: 125–128.

Coy, M. (1987) Rondonia: frente pioneira e o programa Polonoroeste.O processo de diferenciacao socio-economica na periferia e oslimites do planejamento publico. Tubinguen Geographische Studien95: 253–270.

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

Fearnside, P.M. & Ferreira, G.L. (1984) Roads in Rondonia: highwayconstruction and the farce of unprotected reserves in Brazil’sAmazonian forest. Environmental Conservation 11: 358–360.

Ferraz, S.F.B., Vettorazzi, C.A., Theobald, D.M. & Ballester,M.V.R. (2005) Landscape dynamics of Amazonian deforestationbetween 1984 and 2002 in central Rondonia, Brazil: assessment andfuture scenarios. Forest Ecology and Management 204(1): 69–85.

Gallegos, C.M. (1997) Madagascar: unrealized potential in naturalresources. Journal of Forestry 95(2): 10–15.

Hecht, S. (1985) Environment, development and politics: capitalaccumulation and the livestock sector in eastern Amazonia. WorldDevelopment 13(6): 663–684.

Houghton, R.A. (1991) Tropical deforestation and atmosphericcarbon dioxide. Climatic Change 19: 99–118.

IBAMA (2002) Controle de Emissao de Certificados. Departamentode Coordenacao Comercial. Unpublished report. Brasılia, DistritoFederal, Brazil.

INCRA (2002) Levantamento cadastral na floresta nacional do bomfuturo. Relatorio Tecnico. Unpublished report. Porto Velho,Rondonia, Brazil.

INPE (1992) Deforestation in Brazilian Amazonia. Unpublishedreport. Sao Jose dos Campos, Brazil.

INPE (2004) Monitoring of the Brazilian Amazonia by satellite,PRODES Project. Available on http://www.obt.inpe.br/prodes/prodes 1988 2003.htm. Sao Jose dos Campos, Sao Paulo, Brazil.

Jones, D.W., Dale, V.H., Beauchamp, J.J., Pedlowski, M.A. &O’Neill, R.V. (1995) Farming in Rondonia. Resource and EnergyEconomics 17: 155–188.

Laurance, W.F., Albernaz, A.K.M. & Costa, C. (2001) Isdeforestation accelerating in the Brazilian Amazon? EnvironmentalConservation 28(4): 305–311.

Mahar, D.J. (1989) Government policies and deforestation in Brazil’sAmazon Region. Washington, DC, USA: World Bank: 64 pp.

Malingreau, J.P. & Tucker, C.J. (1988) Large-scale deforestation inthe southeastern Amazon Basin of Brazil. Ambio 17(1): 49–55.

Martine, G. (1990) Rondonia and the fate of small producers. In:The Future of the Amazon: Destruction or Sustainable Development?,ed. D. Goodman & A. Hall, pp. 23–48. New York, USA: SaintMartin’s Press.

Millikan, B.H. (1988) The dialectics of devastation: tropicaldeforestation, land degradation, and society in Rondonia, Brazil.M.A. Thesis, University of California, Berkeley, USA: 186 pp.

Millikan, B.H. (1998) Zoneamento socio-economico-ecologicoe polıticas publicas no estado de Rondonia: oportunidades,limites e desafios para o desenvolvimento sustentavel. ProjetoBRA/94/007, Cooperacao Tecnica do PNUD ao Planafloro, PortoVelho, Brazil: 161 pp.

Moran, E.F. (1993) Deforestation and land use in the BrazilianAmazon. Human Ecology 21: 1–21.

Nepstad, D.C., Verıssimo, A., Alencar, A., Nobre, C., Lima, E.,Lefebvre, P., Schlesinger, P., Potter, C., Moutinho, P.,Mendoza, E., Cochrane, M. & Brooks, V. (1999) Large-scaleimpoverishment of Amazonian forests by logging and fire. Nature398: 505–508.

Nepstad, D., McGrath, D., Alencar, A., Barros, A.C., Carvalho, G.,Santilli, M. & Diaz, M.C.V. (2002) Frontier governance inAmazonia. Science 295: 629–630.

Oliveira, J.A.P. (2002) Implementing environmental policies indeveloping countries through decentralization: the case ofprotected areas in Bahia, Brazil. World Development 30(10): 1713–1736.

Olmos, F., Queiroz Filho, A.P. & Lisboa, C.A. (1999) As Unidadesde Conservacao de Rondonia. Unpublished report. CooperacaoTecnica do PNUD ao Planafloro. Porto Velho, Rondonia, Brazil:94 pp.

Peres, C.A. & Terborgh, J.W. (1995) Amazonian nature reserves: ananalysis of defensibility status of existing conservation units anddesign criteria for the future. Conservation Biology 9(1): 34–46.

Peres, C.A. & Lake, I.R. (2003) Extent of nontimber resourceextraction in tropical forests: accessibility to game vertebrates byhunters on the Amazon basin. Conservation Biology 7(2): 521–535.

Redford, K., Brandon, K. & Sanderson, S.E. (1998) Holdingground. In: Parks in Peril: People, Politics, and Protected Lands,ed. K. Brandon, K.H. Redford & S.E. Sanderson, pp. 455–464.Washington, DC, USA: Natural Conservancy/Island Press.

Rocco, R. (2002) Legislacao Brasileira do Meio Ambiente. Rio deJaneiro, Brazil: DP&A Editora.

Rocha, S.B. (1992) Diagnostico das unidades de conservacaoda Amazonia. In: Anais do Seminario Internacional sobre MeioAmbiente, Pobreza e Desenvolvimento da Amazonia, ed. U.H.Bezerra, pp. 347–360. Belem, PA, Brazil: Processamento de Dadosdo Estado do Para (PRODEPA).

SEPLAN (1999) Diagnostico Socio-Economico-Ecologico eAssistencia Tecnica para Formulacao da Segunda Aproximacao doZoneamento Socio-Economico-Ecologico do Estado de Rondonia:Mapa de Solos. Porto Velho, Rondonia, Brazil: 47 pp.

Skole, D.L. & Tucker, C.J. (1993) Tropical deforestation and habitatfragmentation in the Amazon: satellite data from 1978 to 1988.Science 260: 1905–1910.

Uhl, C. (1988) Restoration of degraded lands in the Amazon basin.In: Biodiversity, ed. E.O. Wilson, pp. 326–332. Washington, DC,USA: National Academy Press.

Umana, A. (1998) The World Bank Inspection Panel: The First FourYears 1994–1998. Washington, DC, USA: World Bank.

Verıssimo, A., Cochrane, M.A. & Souza Jr., C. (2002) NationalForests in the Amazon. Science 297: 1478.

World Bank (1997) Report on Progress Review of Implementation ofBrazil: Rondonia Natural Resources Management Project (LoanNo. 3444-BR). World Bank (International Bank of Reconstructionand Development), Washington, DC, USA.

World Bank (2002) Brazil to triple amount of protected Amazonrainforest over 10 years [www document]. URL http://web.worldbank.org / WBSITE/EXTERNAL / NEWS / 0, contentMDK:20066968∼menuPK:34460∼pagePK:64003015∼piPK:64003012∼theSitePK:4607,00.html.

WWF (1999) Areas Protegidas ou Espacos Ameacados? RelatorioWWW sobre o grau de implementacao e a vulnerabilidade dasunidades de conservacao federais brasileiras de uso indireto. SerieTecnica I, WWF, Brasılia, Brazil.