Collectively and individually, IIRSA projects represent an enormous threat to the conservation of the biodiversity of the South American continent. All but one of the ten IIRSA corridors intersects with a Biodiversity Hotspot (Andes, Cerrado, Atlantic Coastal Rainforest) or Wilderness Area (Amazon, Pantanal, Gran Chaco, Caatinga). The IIRSA highway corridors planned for Bolivia, Ecuador, and Peru, the Arco Norte in the Guayana Shield region, and the PPA investments in the Brazilian Amazon are particularly worrisome because they will radically increase access to wilderness areas with very high levels of biological endemism. IIRSA will expose the western Amazon and the Andean foothills to potent global and regional economic forces, and the region's interconnected ecosystems will be inalterably changed. Climate change and geological history have left their mark on today's ecosystems and their species. The distribution of biodiversity is radically different in each of the region's major biomes due to the distinct physical attributes of mountains versus lowlands, and terrestrial versus aquatic ecosystems. Consequently, IIRSA and other development phenomena will have variable impacts across the Amazon, Andes, and the Cerrado biomes. Designing mitigation programs and conservation strategies must be predicated on a thorough understanding of the regional nature of biodiversity.
Montane forests are the most biologically diverse habitats within the Andes Mountains, occupying the terrain between the grasslands of the Andean highlands and the humid forests of the Amazonian lowlands. Landscapes include broad valleys, narrow canyons, slopes of varying steepness, cliff faces, and ridge tops situated between 500 and 3,500 m in elevation. Tropical montane forests are characterized by strong gradients related to topography and manifest as differences in elevation, precipitation, humidity, soil type, slope, aspect, and radiation. Species adapt to these gradients in often contrasting ways, and their distribution depends on the characteristics that define their reproduction and survival (Kessler et al. 2001, Young et al. 2002). For example, the diversity of epiphytic bromeliads is correlated with elevation, as this functional group is adapted to cool and humid cloud forest habitats, whereas aroids are more abundant in warm and humid forest communities. Terrestrial bromeliads and cacti are most abundant in dry valleys where high levels of incident radiation reach the forest floor; fern richness is correlated with moss cover, because the fern gametophytes depend on the water captured and held by moss (Kessler 2000, 2001, 2002). Trees, the most important functional group in the ecosystem, are most diverse at lower elevations, but species composition changes over multiple gradients. Birds and bats decrease in species richness with elevation, but the diversity of rodents is unrelated to elevation (Patterson et al. 1998). Forest communities are subject to periodic disturbances from landslides due to high precipitation and mountainous terrain (Veblen et al. 1981). Because of disturbances and multiple gradients, montane forests are extraordinarily complex spatially; thus, habitat diversity and species turnover are prominent attributes of this ecosystem.
Lower montane forest communities are similar to lowland rain forests; however, with increasing altitude, montane species become more abundant and lowland species more rare (Gentry 1988, 1992a, 1992b). At the top of the montane forest sequence are “cloud forests,” where ridge tops in the foothills and the flanks of the eastern cordillera intersect with the cloud layer that forms over the adjacent lowlands. Cloud forests have unique environmental conditions characterized by mist, low temperatures, and limited solar radiation. Because they are surrounded by forest types that are radically different in structure, function, and composition, patches of cloud forest are effectively geographically isolated (Killeen et al. 2005). One of the central tenets of conservation biology is that island-like habitats are important in establishing the reproductive isolation that drives allopatric speciation (Stebbins 1950, MacArthur & Wilson 1967). The effective isolation of cloud forests is manifest in the negative relationship between mean range size and mean elevation; essentially, species with the smallest range sizes occur at higher altitudes (Kessler 2002).
Environmental differences among patches of cloud forest are caused by elevation gradients and the frequency of cloud formation (Figure 3.1). The Andean foothills are composed of a series of parallel ridges of increasing altitude situated between the lowland plain and the eastern cordillera. Evaporation over the valleys leads to cloud formation over the adjacent ridges in a diurnal cycle that exaggerates both the humidity on ridges and the evapotranspiration in the valleys (Troll 1968, Kessler et al. 2001, Killeen et al. 2007a); consequently, cloud-impacted communities occur at elevations as low as 1,000 m and as high as 3,500m. High levels of endemism are particularly pronounced for higher taxa and functional groups such as amphibians (Köhler 2000, Kattan et al. 2004) epiphytic orchids (Vasquez et al. 2003), aroids (Vargas et al. 2004), and mosses (Churchill et al. 1995). Taxa that are well represented in other biomes also have experienced a radiation of species in montane forests: Ericaceae (Luteyn 2002), Inga (Pennington 1997), Solanaceae (Knapp 2002), and the Podocarpaceae (Killeen et al. 1993). Animal species such as hummingbirds have coevolved with plant taxa (Ericaceae and Bromeliaceae) that are both abundant and speciose in montane forests (Stotz et al. 1996). Many species are known to exist at a single locality, making them extremely vulnerable to extinction (Figure 3.2).
Mid-level and lower montane forest communities provide additional complexity. Some of the valleys of the eastern Andes experience annual precipitation greater than 6,000 mm (Hijmans et al. 2004), but rain shadow and the diurnal cycles in deep canyons create semiarid habitats with precipitation less than 1,000 mm (Troll 1968, Killeen et al. 2007a). Montane dry forests occur as isolated habitats from Argentina to Venezuela. Although they share a common biogeographic history, they have experienced thousands of years of divergent evolution. Andean dry forests contain numerous endemic species or infraspecific regional variants (Pennington et al. 2005); many have been heavily affected by human populations.
The magnitude of IIRSA-related impacts on montane forests is impossible to map with any level of precision because the environmental gradients and the resulting habitat mosaic of these forests is complex. Nonetheless, the consequences of road improvement and expansion in the humid montane regions of Bolivia, Ecuador, and Peru are very predictable. Due to the extremely high levels of endemism associated with cloud forests, there is a high probability that any highway construction will directly lead to species extinction (Ricketts et al. 2005). Deforestation on the adjacent piedmont may reduce cloud cover and raise the height of the cloud base during the dry season (Lawton et al. 2001, Nair et al. In press), possibly altering the environmental conditions of the cloud forest during a critical time of the year. If such a trend develops, then cloud forest species will have to shift their distribution upward or be exposed to environmental conditions different than those to which they are adapted. Many will not be able to migrate upward in response to a rapidly changing altitudinal gradient or adapt to the changing conditions at their current elevation distribution. If so, these species will become extinct.
In lower montane forests the largest impacts will be indirect because they are associated with increased deforestation and forest fragmentation (see discussion on lowland tropical rainforests below). Fortunately, montane forests of the central Andes Mountains are still largely intact; a dozen highway corridors connect the Andean highlands and the Amazon lowlands, but large forest blocks remain unsettled and wild. IIRSA threatens to change this rare wilderness area, permanently degrading a global biodiversity hotspot. Several of the planned transportation corridors will transect areas that have been inhabited for decades (Yungas and Chapare in Bolivia, Huallaga Valley in Peru, Napo in Ecuador), but road improvement will accelerate the expansion of secondary roads into the intact forest blocks, degrading forest remnants. The construction of the Interoceanic Corridor in southern Peru53 will produce even greater impacts because it will improve a road that is closed during several months of the year and that is sparsely settled where it transects montane forests. The situation is even more critical in Colombia, Ecuador, and Venezuela, where montane forest ecosystems have been occupied by human civilizations for centuries. The montane forests of the northern Andes are already highly fragmented, and the cloud forest habitats have been encroached upon from above and below the cloud zone.54 The potential impact from IIRSA investments in the northern Amazon basin will take place on landscapes that have already been substantially altered by human activities. Consequently, special care needs to be taken so that the few remaining patches of native habitat are identified and provided maximum protection.
LOWLAND TROPICAL RAINFOREST
The Amazon region is sometimes described as a monotonous expanse of tall forest that covers more than half of the continent; however, this apparent uniformity is deceiving: the variability in species composition is manifest at the basin, the regional, and the local scale. Botanical research has shown that the floras have strong regional differentiation (Prance 1972, 1989, Mori & Prance 1990, Baker et al. 2004) and that plant communities vary over latitudinal and longitudinal gradients (ter Steege et al. 2000). The distribution of vertebrates is constrained by rivers, leading to many endemic species and subspecies (Wallace 1852, Emmons 1997, Patton & da Silva 1998, da Silva et al. 2005). Scientists with different taxonomic and disciplinary perspectives have fiercely debated the origin and evolution of biodiversity in the Amazon (Haffer 1969, Colinvaux 1993, Nelson et al. 1993, Irion et al. 1995, Marriog & Cerqueira 1997, Lovejoy et al. 1998, Burnham & Graham 1999, Maslin 2005, Mayle & Bush 2005). There is no argument, however, over the distinctiveness of the regional biota and that IIRSA projects will impact some regions more than others.
Studies of vertebrate taxa have identified eight subregions of the Amazon separated by rivers and differentiated according to the distribution of endemic frog, lizard, bird, and primate species (see review in da Silva et al. 2005). The most commonly accepted hypothesis is that the largest Amazonia tributaries and the main trunk of the Amazon River are situated in very old valleys that have acted as barriers to the distribution of forest species over millions of years. Given this geographic isolation, a characteristic suite of endemic species or subspecies has emerged within each subregion. Each one of these biogeographic subregions (Figure 3.3) represents a basic unit of conservation planning because the species in it cannot be affected by conservation measures in the other subregions. One of the central tenets of conservation planning is the need to protect large blocks of forest (Tarabelli & Gascon 2005); however, if a large block is composed of two or more separate subregions that are differentiated by endemism, then the effective size of that forest block is proportionally reduced. Consequently, the large forest remnants in the southern and eastern Amazon are not as effective conservation units as they may seem. Regional endemism should be the primary consideration in designing conservation corridors (da Silva et al. 2005). Unfortunately, IIRSA corridors have been designed without consideration of their potential impact on regional endemics. Regional endemism is less important in defining geographic stratification on the Andean piedmont, where landscapes are younger and there is little evidence that rivers have acted as barriers to the distribution of species (Aleixo 2004); nonetheless, there are still regions of high endemism that must be considered in development planning. Botanical studies show a gradual change in forest composition between Bolivia and Ecuador (Smith & Killeen 1998, Terborgh & Andresen 1998, Pitman et al. 2001). However, the latitudinal gradient does not correspond simply with the precipitation gradient that is responsible for a concomitant reduction in biodiversity. There are three regions that experience anomalously high rainfall at some distance from the equator (Hijmans at al. 2004); each of these regions is floristically distinct and contains high levels of species richness and endemism (Killeen et al. 2007a). They have been hypothesized to be climatically stable and probably remained wet during the Pleistocene, when the distribution of humid forest species was reduced.55 Habitat conversion in these regions is particularly unfortunate, especially because these areas are prime candidates for biological reserves that are resistant to future climate change (Killeen et al. 2007a).
The structure and composition of Amazon forest communities make them particularly susceptible to deforestation and fragmentation because they are characterized by a small number of oligodominant tree species (usually ten to twenty) that are more or less abundant, constituting up to 50 percent of all stems (Pitman et al. 2001, 2002). Oligodominants tend to have very wide distributions and can be viewed as successful species in terms of their ability to reproduce, disperse seeds, and recruit adult individuals into the forest canopy. However, the rest of the tree community is composed of hundreds of species that are represented by only one to a few individuals per hectare. Rare species represent the bulk of tree diversity in tropical forest communities (Pitman et al. 2001, 2002, Condit et al. 2002). Detailed studies on the range distribution of rare species have not been undertaken, but it is reasonable to hypothesize that most are regional endemics. Because of their low population densities, they are particularly susceptible to forest fragmentation and deforestation, which will tend to homogenize the Amazon forest as these rare regional endemics are eliminated, leaving the widespread oligodominants.
The deleterious effects of forest fragmentation are well documented (Laurance et al. 2002, Tabarelli & Gascon 2005). Most deforested landscapes retain patches of natural forest surrounded by a matrix of human modified landscapes that include crops, pastures, and secondary forest. Forest fragments are relatively small and have a large edge-to-interior ratio that exposes the fragment to further degradation. Edge effects include greater penetration of light and wind, which leads to drier conditions in the understory, impedes the regeneration of native species, and exposes adult trees to increased mortality from wind sheer (Laurance & Wiliamson 2001). In addition, fire is an increased threat because forest fragments are often surrounded by pasture that is periodically burned for weed control; fire invades forest remnants, increasing adult mortality and further degrading the forest structure (Cochrane & Laurance 2002). Because forest remnants are located in areas with relatively high population density and reduced regional fores cover, logging is also more intensive to meet the local population's ongoing need for timber and firewood. Similarly, intensive hunting quickly leads to the local extinction of the vertebrate fauna, which limits the dispersal of seeds and genetically degrades tree populations. Edge effect has been shown to penetrate up to 300 m into forests, and because remnants are often quite small, there is often no place free from some sort of edge effect (Laurance et al. 2002). Disturbance from adult mortality, logging, and fire promote the regeneration of pioneer species that are both invasive and widely distributed; all of these processes lead to further homogenization of the forest flora (Tabarelli & Gascon 2005).
Lowland forests will be degraded over an extensive area due largely to the economic and social forces unleashed by the IIRSA projects (Fearnside & Graça 2006). Highway construction in remote areas stimulates human migration, which leads to accelerated land use change and increased habitat fragmentation. Because the land on the piedmont and adjacent plains is essentially flat, secondary roads will proliferate, bringing extensive deforestation, forest fragmentation, and forest degradation (Laurance et al. 2001). It is not unreasonable to predict that at least 250,000 km2 will be deforested in the Amazon over the next decade, with a total of at least 50,000 km2 in Bolivia, Colombia, Ecuador, and Peru combined.56 If the proposed highways are constructed, forest habitats on the eastern slope of the Andes and the adjacent piedmont will be fragmented into at least eight distinct blocks with several highways transecting proposed biodiversity conservation corridors. Even larger areas will be subject to forest degradation caused by logging and fire.
GRASSLANDS, CERRADOS, AND DRY FORESTS
The regions adjacent to the Amazon support a variety of ecosystems adapted to the seasonal climates of the dry tropics (Daly & Mitchell 2000). The largest and the most biologically diverse of them is the Cerrado Biodiversity Hotspot, a complex of savannas and shrublands that stretches across the southern Amazon from Maranhão to eastern Bolivia (Machado et al. 2007). Structurally similar but floristically distinct are the grasslands in Colombia and Venezuela (Llanos de Orinoco); Bolivia (Llanos del Moxos); and on the border of Bolivia, Brazil, and Paraguay (Gran Pantanal), which are recognized as wilderness areas because the greatest part of their surface area remains essentially intact. Natural grasslands also occur in the Guayana Shield region of Venezuela (Gran Sabana), Guayana (Rupinini), and Brazil (Roraima), as well as on isolated patches associated with poor soils or seasonally inundated landscapes in the Brazilian Amazon, such as the Araguaia River (Ilha do Bananal) and white sand caatingas of the central Amazon. Savannas and scrublands predominate on these landscapes due to environmental factors such as seasonal drought, poor soils, and impeded drainage, with fire almost always playing an important role in modulating the density of the woody cover. Savannas almost always occur within a landscape mosaic with a seasonal forest habitat where the edaphic constraints that limit forest formation are less severe. Dry or seasonal forest formations exist on landscapes with relatively fertile soils, such as in eastern Bolivia (Chiquitano, Gran Chaco), northeastern Brazil (Caatinga), and Venezuela (Pennington et al. 2005).
The Cerrado biome is renowned for its high levels of diversity and endemism, with an estimated 40 percent of its woody plants and 38 percent of its reptiles occurring only in this biogeographic region (Colli 2005, Ratter et al. 2006). One of the key ecological attributes of the Cerrado ecosystem is its habitat diversity, which ranges from open grasslands to dense scrubland and gallery forests along water courses; specialized species are often most abundant or even restricted to a specific habitat type (Figure 3.4). Habitat diversity leads to complex landscape mosaics with multiple ecotones. Gallery forests are particularly important for providing cover for wildlife, and most savanna fauna depend on this habitat type within the savanna ecosystem. The biodiversity of the Cerrado Hotspot has not been adequately mapped, but the distribution of woody plant species reveals that there is strong regional differentiation (Ratter et al. 2006), and a gap analysis performed for 244 threatened and endemic amphibians, reptiles, birds, mammals, and plants demonstrated that almost 30 percent of the Cerrado biodiversity is not represented in the existing protected areas (Machado et al. 2007).
Modern development in the Cerrado region began in the 1950s and 1960s with the expansion of cattle ranching on the Planalto de Mato Grosso. However, because most native grasses are coarse and unpalatable, landowners cleared them to plant cultivated grasses that allowed them to increase their livestock herds and improve animal management. Beginning in the late 1980s, agronomists learned that the application of lime (CaCO2) to Cerrado soils could dramatically improve soil fertility. Landholders then began to convert ranches into cropland, and the Cerrado has since become the largest soybean producing region, making Brazil the largest exporter of soy in the world. The original extent of the Cerrado habitat has been estimated at 2.2 million km2, but approximately 55 percent has been converted to either pasture or cropland. The rate of habitat conversion was estimated at 1.36 to 2.2 million ha per year between 1985 and 2002, and some models predict that what remains of the Cerrado will disappear by 2030 (Machado et al. 2004, 2007). Unfortunately, only 2.2 percent of the original Cerrado area has been set aside as protected areas (Klink & Machado 2005, Rylands et al. 2005).
Rapid land use change and habitat degradation are similarly occurring in seasonally dry forest formations, where climates and soils have long favored human settlement and agriculture (Olson & Dinerstein 1998). These geographically separate forests share a common biogeographic history (Prado & Gibbs 1993, Pennington et al. 2005). Apparently, they coalesced near the equator during the last glacial maximum surrounding a much reduced rainforest ecosystem in the central and western Amazon (Mayle et al. 2004); nonetheless, today, each subregion of seasonally dry forest has its own group of endemic species or subspecies (Pennington et al. 2005). Nonsustainable logging practices are particularly damaging in these ecosystems because of the inherently slow growth rates of species adapted to these climatic regimes (Dauber 2003). The Caatinga region is the most degraded of the seasonal forest regions due to the long history of human settlement and colonization in northeastern Brazil. Recent deforestation has impacted dry forest regions in Bolivia, Venezuela, and southern Brazil.
Fortunately, the savannas of the Llanos de Orinoco, Moxos, and the Gran Pantanal are still relatively intact due largely to the seasonally inundated landscapes, which inherently limit cultivation. Nonetheless, cattle ranchers do graze livestock on the native grasses of these regions, which are more palatable than the native grasses on the upland savannas, and some areas are locally impacted by overgrazing. In addition, there are recent experiments to convert seasonally inundated habitats to paddy rice cultivation in Bolivia, and there is clearing of woody plants in the Gran Pantanal.
IIRSA and PPA investments will accelerate habitat degradation in most, if not all, of the extra-Amazonian ecosystems. The Cerrado Biodiversity Hotspot is the most endangered due largely to its suitability for mechanized agriculture. Although the Federal Government of Brazil makes repeated commitments to conserve the Amazon, similar actions to conserve in the Cerrado region are weighed against national priorities to expand agricultural production (Figure 3.5). Thus, although the Amazon ecosystem faces extensive degradation and fragmentation due to IIRSA investments over the medium term, the Cerrado faces virtual annihilation over the next half century (Machado et al. 2007). Considering the advanced state of the destruction of this biome, protected area creation should be part of any strategy that seeks to mitigate the effects of the PPA's investments in modern highways. The Brazilian forest code requires that 20 percent of private properties be left as native habitat within the Cerrado biome. These efforts to conserve gallery forests within private lands could significantly reduce the effect of habitat conversion within savanna ecosystems.57
The Amazon River system is the world's largest freshwater ecosystem, with nearly 20 percent of the Earth's freshwater discharge (Goulding 1980). The aquatic biodiversity of the Amazon is the consequence of geological history, the sheer size of the basin, the contrasting nature of the constituent sub-basins, and the extraordinary habitat heterogeneity of each individual floodplain. The Amazon's sub-basins have been traditionally classed into three broad categories on the basis of turbidity, color, and pH of the water: 1) white water rivers originating in the Andes are characterized by high sediment loads and approximately neutral pH; 2) black water rivers drain lowland landscapes with white sand soils dominated by plants with very high tannin58 levels, so their runoff is very dark and acidic; 3) clear water rivers arise in moderate terrain, particularly the Brazilian and Guayana Shield, and have relatively clear water of nearly neutral pH (Sioli 1968, Junk 1983). Recent studies have emphasized the diversity of aquatic habitats associated with the morphological variability of channels, lakes, and wetlands, as well as distinct hydrological regimes in upstream, midstream, and downstream sections of the river (Mertes et al. 1996, Goulding et al. 2003). Like their terrestrial counterparts, aquatic ecosystems are characterized by high levels of biodiversity and species endemism (Figure 3.6). Estimates of the number of Amazonian fish species vary between 1,300 and 3,000, but the true number may be much higher as systematists revise the status of headwater populations of widespread but poorly studied taxonomic groups (Ruffino 2001).
Migration is a behavioral trait characteristic of many Amazonian fish. Some of the most economically important commercial species such as the piramutaba (Brachyplatystoma vaillantii) and the dourada (B. flavicans) migrate long distances between the estuary and upstream sections of the basin. Other species are less pelagic in their migration, moving within certain sectors of the basin, such as the tambaqui (Colossoma macropomum), pacú (Mylossoma spp.), jaraqui (Semaprochilodus spp.), and curimatã (Prochilodus nigricans), among others (Barthem & Goulding 1997, Ruffino 2001). One of the most salient features of the Amazon River system is the importance of the floodplain in providing multiple niches for aquatic organisms, and many fish species migrate locally between the channel and flood plain habitats in accord with annual fluctuations in river levels (Goulding 1980, Goulding & Ferreira 1996). Along white water rivers, floodplain forests known as varzea are particularly productive because the sediments washed down from the Andes bring essential chemical nutrients. Frugivorous fish migrate into varzea during periods of high water to spawn and feed on a rich assortment of fruits, then return to the river channels during low-water periods (Barthem & Goulding 1997).
Modifications to the Amazon's aquatic systems through the construction of dams and reservoirs will bring obvious direct impacts in the immediate area of each project. The most serious are those relating to fish populations. At the Tucuruí dam on the Tocantins River in Brazil, a program to monitor fish populations before and after the flooding of the reservoir showed declines in species diversity upstream (25 percent) and downstream (19 percent); in addition, species diversity was down 27 percent compared with the preexisting riverine habitat (LaRovere & Mendes, 2000). Another recent study of mercury contamination in the Caroni River in Venezuela revealed that mercury levels in fish populations were several-fold higher within the reservoir than in populations below the reservoir. Apparently the high organic content of lake bed sediments (the result of flooding tens of thousands of kilometers of forest vegetation) creates anoxic conditions that enable mercury-methylate to form, a type of mercury that is rapidly absorbed by biological organisms (Veiga 1997, Fearnside 2001a, 2005a).
Aquatic systems are particularly susceptible to secondary long-term impacts: upstream effluents and sediments are conveyed downstream, while downstream obstacles impede the migration of species to headwater regions. Consequently, the impacts of IIRSA investments on aquatic systems will be apparent both locally and regionally. Deforestation of upland forest in lowland landscape will bring about a pulse of nutrients associated with ash and increased sedimentation, as well as long-term changes in temperature and chemistry (Bojsen & Barrigo 2002). Muddy streams that meander through cattle pasture bear little resemblance to the cool, shaded, clear water habitats that existed prior to deforestation. Paddy rice cultivation is not common in the Amazon, but recent experiments in Bolivia are encouraging new migrants to adopt this production system in the accessible areas of the Llanos de Moxos of the Beni. The conversion of savanna wetlands for paddy rice cultivation will also have important, if poorly understood, impacts on fish populations.
Savanna fishes are known to be diverse, although they represent the most poorly studied segment of freshwater fishes in tropical America (Schaefer 2000). Deforestation of inundated forests is particularly devastating because it directly affects the prime feeding and spawning ground of the most economically important fish species. Road-building in mountainous regions will have severe short-term impacts. Because roads typically are built on slopes above rivers, earth-moving equipment can dump tens of thousands of tons of soil and rock into watercourses during construction. Highways and rivers usually run parallel for dozens of kilometers; thus, the riparian environment will be permanently altered.
A major emphasis of IIRSA is the renovation and improvement of hydrovias. This promotion of river transport will have fewer negative impacts than highway construction, and the creation of river ports will pose moderate impacts. Nonetheless, a revitalized Amazonian hydrovia will potentially lead to increased population density along major and minor rivers. This would almost certainly lead to increased fishing pressure, and may also foster increased deforestation in both upland and flooded forests. IIRSA investments to construct dams for energy production would constitute a long-term, permanent impact on dozens if not hundreds of migratory fish species. Damming the major tributaries of the Amazon River may have catastrophic effects on fish migration and fish populations and bring about huge economic consequences.
AndersenL. E. 1997A Cost-benefit Analysis of Deforestation in the Brazilian AmazonTexto para Discussão, no. 455Río de JanieroIPEA, Instituto de Pesquisas Econoimicas AplicadaGoogle Scholar
Banco do Brasil 2007PROEX – Programa de Financiamento às Exportações.Online. Available: http://www.bb.com.br/appbb/portal/gov/ep/srv/fed/AdmRecPROEXFin.jspGoogle Scholar
BarthemR. B. M.Goulding 1997The Catfish Connection: Ecology, Migration, and Conservation of Amazon PredatorsNew YorkColumbia University PressGoogle Scholar
BerryM. C. 1975The Alaska Pipeline: The Politics of Oil and Native Land ClaimsBloomington, INIndiana University PressGoogle Scholar
(BOA) Board on Agriculture, Committee on Sustainable Agriculture & the Environment in the Humid Tropics, National Research Council 1993Sustainable Agriculture and the Environment in the Humid TropicsWashington, DCNational Academy PressGoogle Scholar
Bolivia Forestal 2007Preliminar: Exportaciones forestales del 2006 superan los 170 Millones de $US.Cámara Forestal81Online. Available: http://www.cfb.org.bo/NoticiasBF/8.01/boletin.notaBF03.htm. June 1, 2007Google Scholar
Brito-CarreirasJ. M. J. M.Cardoso-Pereira M. L.Campagnolo Y. E.Shimabukuro 2005A land cover map for the Brazilian Legal Amazon using SPOT-4 VEGETATION data and machine learning algorithmsAnais XII Simpósio Brasileiro de Sensoriamento Remoto. April 16–21. Goiânia, Brasil. INPE. pp. 457–464. Online. Available: http://marte.dpi.inpe.br/col/ltid.inpe.br/sbsr/2004/11.19.14.07/doc/457.pdf. May 1, 2007Google Scholar
CadmanJ. D. 2000The Environmental Aspects of Six Hydro Reservoirs in the Amazon Basin.Submission to the World Commission on Dams, no. ENV061. Online. Available: http://www.dams.org/kbase/submissions/showsub.php?rec=ENV061. January, 13, 2007Google Scholar
CamposM. M.Francis F.Merry 2005Stronger by Association: Improving the Understanding of How Forest-Resource Based SME Associations can Benefit the PoorLondonInstituto de Pesquiza Ambiental da Amazônia & The International Institute for Environment and DevelopmentGoogle Scholar
ChinaView 2006CNPC to purchase EnCana's oil business in Ecudaor.Online. Available: http://news.xinhuanet.com/english/2005-09/15/content_3497826.htm. March 14, 2007Google Scholar
ChurchillS. P. D.Griffin M.LewisIII 1995Moss diversity of the tropical Andes.In ChurchillS. P. H.Balslev E.Forero J. L.Luteyn (Eds.)Biodiversity and Conservation of Neotropical Montane Forestspp335346 Bronx, NYNew York Botanical GardenGoogle Scholar
CochraneT. A. T. J.Killeen O.Rosale 2007Agua, Gas y Agroindustria: La Gestion Sostenible de la Riego Agrícola en Santa Cruz, BoliviaLa Paz, BoliviaConservation InternationalGoogle Scholar
ColinvauxP. A. 1993Pleistocene biogeography and diversity in tropical forests of South America.In GoldblattP. (Ed.)Biological Relationships between Africa and South Americapp473499New Haven, CTYale University PressGoogle Scholar
ColliG. R. 2005As origens e a diversificação da herpetofauna do Cerrado.In ScariotA. J. C.Souza-Silva J. M.Felfili (Eds.)Cerrado: Ecologia, Biodiversidade e Conservaçãopp247264BrasíliaMinistério do Meio AmbienteGoogle Scholar
CowellA. 1990The Killing of Chico MendesEpisode 4. The Decade of Destruction: A Unique Chronicle of the Destruction of the Amazonian Rainforest. PBS Frontline Documentary Series. VideotapeGoogle Scholar
DalyD. C. D. J.Mitchell 2000Lowland vegetation of tropical South America: An overview.In LentzD. (Ed.)Imperfect Balance: Landscape Transformations in the pre-Columbian Americaspp391454New YorkColumbia University PressGoogle Scholar
DauberE. 2003Modelo de Simulación para Evaluar las Posibilidades de Cosecha en el Primer y Segundo Ciclo de Corta en Bosques Tropicales de BoliviaDocumento Técnico 128/2003Santa Cruz, BoliviaProyecto BOLFORGoogle Scholar
EmmonsL. H. 1997Neotropical Rainforest Mammals: A Field Guide2d edChicagoChicago University PressGoogle Scholar
EspinozaG. B.Richards 2002Fundamentals of Environmental Impact AssessmentWashington, DCInter-American Development Bank (IDB) & Inter-American Association of Sanitary and Environmental Engineering (AIDIS)Google Scholar
FabeyM. 1997Free-Trade-Zone Status Turns Amazon Port into Boom Town.Global Logistics & Supply Chain StrategiesOnline. Available: http://www.glscs.com/archives/2.97.FTZ.htm?adcode=90. October 13, 2006Google Scholar
FearnsideP. M. 2005bIndigenous peoples as providers of environmental services in Amazonia: Warning signs from Mato Grosso.In HallA. (Ed.)Global Impact, Local Action: New Environmental Policy in Latin Americapp187198LondonUniversity of London School of Advanced Studies, Institute for the Study of the AmericasGoogle Scholar
FearnsideP. M. Graça 2006BR-319: Brazil's Manaus-Porto Velho highway and the potential impact of a migration corridor to Central Amazonia, Instituto Nacional de Pesquisas da Amazônia-INPA, Manaus, Amazonas, Brazil.Ecological Society of AmericaMérida mexicoGoogle Scholar
FoglemanV. M. 1990Guide to the National Environmental Policy Act. Interpretations, Applications, and ComplianceNew YorkQuorum BooksGoogle Scholar
GlaserB. W. I.Woods (Eds.)2004Amazonian Dark Earths: Explorations in Space and TimeBerlinSpringer-VerlagGoogle Scholar
Global Mapping International 2006World Language Mapping SystemCDROM, Colorado Springs, COGlobal Mapping InternationalGoogle Scholar
GoeschlT. D. C.Igliori 2004Property Rights, Conservation and Development: An Analysis of Extractive Reserves in the Brazilian Amazon, Natural Resources Management(FEEM) Fondazione Eni Enrico Mattei. Working Paper no. 60.04. Online. Available: http://www.feem.it/Feem/Pub/Publications/WPapers/default.htmGoogle Scholar
Gomez-RomeroE. T.Tamariz-Ortiz 1998Uso de la tierra y patrones de deforestacion en la zona de Iquitos.In KalliolaR. S.Flores-Paitan (Eds.)Geoecologia y Desarrollo AmazonicoSulkavaFinnreklama OyGoogle Scholar
GouldingM. 1980The Fishes and the Forest: Explorations in the Amazonian Natural HistoryBerkeleyUniversity of California PressGoogle Scholar
GouldingM. E. G.Ferreira 1996Pescarias Amazônicas, Porteção de Habitas e Fazendas nas Várzeas:Uma Visão Ecológica e EconômicaRelatório Banco Mundial. Brasília: BIRDGoogle Scholar
GouldingM. R.Barthem E.Ferreira 2003The Smithsonian Atlas of the AmazonWashington, DCSmithsonian Institution PressGoogle Scholar
GroganJ. E. P.Barreto A.Veríssimo 2002Mahogany in the Brazilian Amazon: Ecology and perspectives on managementBelém, Brazil(IMAZON) Amazon Institute of People and the EnvironmentGoogle Scholar
HaggettP. A. D.Cliff A.Frey 1977Locational Analysis in Human GeographyNew YorkWileyGoogle Scholar
HallA. 2004Extractive Reserves: Building Natural Assets in the Brazilian Amazon.Working Paper Series, no. 74Amherst, MA(PERI) Political Economy Research InstituteGoogle Scholar
HanaiM. 1998Formal and garimpo mining and the environment in Brazil.In WarhurstA. (Ed.)Mining and the Environment: Case Studies from the Americaspp181197OttowaInternational Development Research CenterOnline. Available: http://reseau.crdi.ca/en/ev-31006-201-1-DO_TOPIC.htmlGoogle Scholar
HarperG. J. M. K.Steininger Y.Talero M.Sanabria T. J.Killeen L. A.Solorzano 2007Deforestation Assessments Across the Andes.Online. Available: http://science.conservation.org/portal/server.pt?open=512&objID=755&&PageID=128505&mode=2&in_hi_userid=124186&cached=true. May 1, 2007Google Scholar
HechtS. B. A.Cockburn 1989The Fate of the Forest: Developers, Destroyers, and Defenders of the AmazonLondonVersoGoogle Scholar
HezelF. X. 2001The New Shape of Old Island CulturesHonoluluUniversity of Hawaii PressGoogle Scholar
(IBGE) Instituto Brasileiro de Geografia e Estadística 2006Síntese de Indicadores Sociais 2006.Online. Available: http://www.ibge.gov.br/home/estatistica/populacao/condicaodevida/indicadoresminimos/sinteseindicsociais2006/default.shtm. May 5, 2007Google Scholar
(IDB) Inter-American Development Bank 2006Building a New Continent: A Regional Approach to Strengthening South American InfrastructureWashingtonIDBOnline. Available: http://www.iadb.org/publications/Reports.cfm?language=en&parid=4. May, 15, 2007Google Scholar
(IPCC) Intergovernmental Panel on Climate Change 2007Climate Change 2007: The Physical Science Basis, Summary for PolicymakersContribution of Working Group I to the Fourth Assessment Report of the Intergovenmental Panel on Climate ChangeGenevaIPCCOnline. Available: http://www.ipcc.ch/SPM2feb07.pdf. April 1, 2007Google Scholar
IUCN Commission on National Parks & Protected Areas & World Conservation Monitoring Centre 1994Guidelines for Protected Area Management CategoriesGland, SwitzerlandIUCNGoogle Scholar
KabatP. M.Claussen P. A.Dirmeyer J. H. C.Gash L.Bravo de Guenni M.Meybeck R. A.PielkeSr. C. J.Vorosmarty R. W. A.Hutjes S.Lutkemeier (Eds.)2004Vegetation, Water, Humans and the Climate: A New Perspective on an Interactive SystemBerlinSpringer VerlagGoogle Scholar
KaimowitzD. A.Angelsen 1998Economic Models of Tropical Deforestation: A ReviewBogorCenter for International Forestry ResearchGoogle Scholar
KalliolaR. S.Flores-Paitan 1998Geoecologia y desarolla Amazonico: Estudio integrado en la zona de Iquitos, PeruAnnales Universitatis Turkuensis, Ser A IITurku, FinlandTurku UniversityGoogle Scholar
KaltnerF. J. G. F. P.Azevedo I. A.Campos A. O. F.Mundim 2005Liquid Biofuels for Transportation in Brazil: Potential and Implications for Sustainable Agriculture and Energy in the 21st CenturySubmitted report by Fundação Brasileira para o Desenvolvimento Sustentável. Commissioned by The German Technical Cooperation. (GTZ) Online. Available: http://www.fbds.org.br/IMG/pdf/doc-116.pdf. April 2007Google Scholar
KilleenT. J. S. G.Beck E.Garcia 1993Guía de Arboles de BoliviaLa Paz, BoliviaHerbario Nacional de Bolivia & Missouri Botanical GardenGoogle Scholar
KilleenT. J. M.Douglas T.Consiglio P. M.Jørgensen 2007aWet spots and dry spots in the Andean Hotspot, the link between regional climate variability and biodiversity.Journal of BiogeopgraphyIn pressGoogle Scholar
KilleenT. J. V.Calderon L.Soria B.Quezada M. K.Steininger G.Harper L. A.Solórzano C. J.Tucker 2007bThirty Years of Land-Cover Change in Bolivia.AMBIOIn pressGoogle Scholar
LehmannJ. D. C.Kern B.Glaser W. I.Woods 2003Amazonian Dark Earths: Origin, Properties, ManagementDordrecht, The NetherlandsKluwerGoogle Scholar
LugoA. E. 2002Homoegeocene in Puerto Rico.In ZarinD. J. J. R R.Alavalapati F. E.Putz M.Schmink (Eds.)Working Forests in the Neotropics: Conservation through Sustainable Managementpp266276New YorkColumbia University PressGoogle Scholar
MacArthurR. H. E. O.Wilson 1967The Theory of Island BiogeographyPrinceton, NJPrinceton University PressGoogle Scholar
MachadoR. M. B.Ramos-Neto M. B.Harris R.Lourival L. M. S.Aguiar 2004Análise de lacunas de proteção da biodiversidade no Cerrado.InAnais IV Congresso Brasileiro de Unidades de Conservaçãopp2938Curitiba, BrasilBrasil Fundação O Boticário de Proteção à NaturezaGoogle Scholar
MachadoR. B. M. B. R.Neto J. M. C.Silva R. B.Cavalcanti 2007Cerrado deforestation and its effects on biodiversity conservation.In KlinkC. A. R. B.Cavalcanti R.Defries (Eds.)Cerrado Land-Use and Conservation: Balancing Human and Ecological NeedsApplied Advances in Biodiversity Science, no. 8Washington, DCCenter for Applied Biodiversity Science, Conservation International (CI)(In press)Google Scholar
MalhiY. J.Wright 2005Late twentieth-century patterns and trends in the climate of tropical forest regions.In MalhiY. O. L.Phillips (Eds.)Tropical Forests & Global Atmospheric Changepp316OxfordOxford University PressGoogle Scholar
MannC. 20051491: New Revelations of the Americas before ColumbusNew YorkKnopfGoogle Scholar
MargulisS. 2004Causes of Deforestation in the Brazilian AmazonBrasiliaWorld BankGoogle Scholar
MaslinM. 2005The longevity and resilience of the Amazon rainforest.In MahliY. O. L.Phillips (Eds.)Tropical Forests & Global Atmospheric Changepp167183OxfordOxford University PressGoogle Scholar
MayleF. E. M. E.Bush 2005Amazonian ecosystems and atmospheric change since the last gl;acial maximum.In MalhiY. O. L.Phillips (Eds.)Tropical Forests & Global Atmospheric Changepp183191OxfordOxford University PressGoogle Scholar
MertesL. A. K. E. M. L.Novo D. L.Daniel Y. E.Shimabukuro J. E.Richey T.Krug 1996Classification of Rios Solimoes-Amazonas wetlands through application of spectral mixture analysis to landsat thematic mapper data.VIII Simposio Brasileiro de Sensoriamento RemotoSalvador, BrazilGoogle Scholar
MoriS. A. G. T.Prance 1990Lecythidaceae - part II: The zygomorphic-flowered New World genera (Couroupita, Corythophora, Bertholletia, Couratari, Eschweilera, & Lecythis).Flora Neotropica Monographno. 21Bronx, NYNew York Botanical GardenGoogle Scholar
NairU. S. D. K.Ray R. O.Lawton R. M.Welch R. A.PielkeSr. J.Calvo The impact of deforestation on orographic cloud formation in a complex tropical environment.In BruijnzeelL. A. J.Juvik F. N.Scatena L. S.Hamilton P.Bubb (Eds.)Mountains in the Mist: Science for Conserving and Managing Tropical Montane Cloud ForestsHonoluluUniversity of Hawaii PressIn PressGoogle Scholar
OrtizE. 2005Conservation Biology of Brazil-nut Rich ForestsWashingtonSmithsonian InstitutionGoogle Scholar
PachecoP. 1998Estilos de Desarrollo, deforestación y Degradación de Los Bosques en Las Tierras Bajas de BoliviaLa PazCIFOR, CEDLA, Fundacion TIERRAGoogle Scholar
PartidárioM. R. 1999Strategic environmental assessment: Principles and potential.In PettsJ. Handbook on Environmental Impact Assessmentpp6073LondonBlackwellGoogle Scholar
PartidárioM. R. R.Clark 2000Perspectives on Strategic Environmental AssessmentBoca Raton, FLCRC PressGoogle Scholar
PattonJ. L. M. N. F.da Silva 1998Rivers, refuges, and ridges: The geography of speciation of Amazonian mammals.In HowardD. J. S. H.Berlocher Endless Forms: Species and Speciationpp202213OxfordOxford University PressGoogle Scholar
PearceD. W. 1994Economic Value BiodiversityLondonJames & Jame, EarthscanGoogle Scholar
PenningtonT. 1997The Genus Inga – BotanyLondonRoyal Botanic Gardens, KewGoogle Scholar
PenningtonR. T. M.Lavin D. E.Prado C. A.Pendry S. K.Pell 2005Climate change and speciation in Neotropical seasonally forest plants.In MalhiY. O. L.Phillips Tropical Forests & Global Atmospheric Changepp191198OxfordOxford University PressGoogle Scholar
PowersM. 2002Illegal loggers invade primordial indigenous natives.Environment News ServiceOnline. Available: http://www.ensnewswire.com/ens/aug2002/2002-08-09-01.asp. August 9, 2002Google Scholar
PranceG. T. 1972ChrysobalanaceaeFlora Neotropica Monograph, no. 9New YorkPublished for Organization for Flora Neotropica by HafnerGoogle Scholar
PranceG. T. 1989Chrysobalanaceae: SupplementFlora Neotropica. Monograph, no. 9SNew YorkOrganizaiton for Flora NeotropicaGoogle Scholar
(PROMPEX) Peruvian Export Promotion Agency 2006Boletines Sectoriales de Exportación: Enero – Marzo 2006.Online.Available: http://www.prompex.gob.pe/Prompex/Portal/Sector/DefaultSector.aspx?.menuId=3Google Scholar
PutzF. E. M. A.Pinard T. S.Fredericksen M.Peña-Claros 2004Forest science and the BOLFOR experience: Lessons learned about natural forest management in Bolivia.In ZarinD. J. J. R. R.Alavalapati F. E.Putz M.Schmink Working Forests in the Neotropics: Conservation through Sustainable Managementpp6496New YorkColumbia University PressGoogle Scholar
RatterJ. A. S.Bridgewater J. F.Ribeiro 2006Biodiversity patterns of the woody vegetation of the Brazilian Cerrados.In PenningtonR. T. G.Lewis J. A.Ratter Neotropical Savannas and Dry Forests: Plant Diversity, Biogeography and ConservationBoca Raton, FLCRC PressGoogle Scholar
RedwoodIII.J. 2002World Bank Approaches to the Brazilian Amazon: The Bumpy Road toward Sustainable Development.Latin America and Caribbean Region Sustainable Development Working Paper, no. 13. Washington: The World BankOnline. Available:http://wbln0018.worldbank.org/.../b8234d558447e77e85256ccd005dbbc5/$FILE/redwood%Google Scholar
ReidW. V. S. A.Laird R.Gamez A.Sittenfeld D. H.Janzen M. A.Gollin C.Juma 1993A new lease on life.In ReidW. V. S. A.Laird C. A.Meyer R.Gamez A.Sittenfeld D. H.Janzen M. A.Gollin C.Juma Biodiversity Prospecting: Guidelines for Using Genetic and Biochemical Resources Sustainably and Equitablypp152WashingtonWorld Resources InstituteGoogle Scholar
Reuters 2007South American Heads Meet in Brazil.January 7. Online. Available at http://www.reuters.com/news/video/videoStory?videoId=30147Google Scholar
RicardoF. A.Rolla 2006Mineração em Unidades de Conservação na Amazônia BrasileiraSão PauloInstituto SocioambientalGoogle Scholar
RiceD. C. A.Sugal S. M.Ratay G. A. B.da Fonseca 2001Sustainable Forest Management: A Review of Conventional WisdomAdvances in Applied Biodiversity Science, no. 3Washington DCCenter for Applied Biodiversity Science at Conservation InternationalGoogle Scholar
RosenfeldA. B. D. L.Gordon M.Guerin-McManus 1997Reinventing the Well Approaches to Minimizing the Environmental and Social Impact of Oil Development in the TropicsWashington, DCConservation InternationalGoogle Scholar
RosenthalJ. P. 1997Equitable sharing of biodiversity benefits: Agreements on genetic resources.InInvesting In Biological Diversity: Proceedings of the Cairns Conferencepp253274ParisOrganisation for Economic Cooperation and Development (OECD)Google Scholar
RuffinoM. L. 2001Strategies for Managing Biodiversity in Amazonian FisheriesManaus, BrazilThe Brazilian Environmental and Renewable Natural Resources Institute (IBAMA)Online. Available: http://www.unep.org/bpsp/HTML%20files/TS-Fisheries2.htmlGoogle Scholar
RylandsA. B. M.Fonseca R.Machado R.Cavalcanti 2005Brazil.In SpaldingM. S.Chape M.Jenkins The State of the World's Protected AreasCambridgeUnited Nations Environment Programme (UNEP) and World Conservation Monitoring Centre (WCMC)Google Scholar
SchaeferS. 2000Fishes of Inundated Tropical Savannas: Diversity and Endemism in the Serrania Huanchaca of Eastern BoliviaFinal report sponsored by The American Museum Center for Biodiversity and Conservation. Online. Available: http://220.127.116.11/scholar?hl=en&lr=&q=cache:h-ivoaIKpAJ:research.amnh.org/ichthyology/bolivia.pdf+Schaefer+Fishes+Tropical+inundatedGoogle Scholar
SchwartzmanS. 1985Banking on disaster.Multinational Monitor67Online. Available: http://www.multinationalmonitor.org/hyper/issues/1985/0615/schwartzman.htmlGoogle Scholar
SmithD. N. T. J.Killeen 1998A comparison of the structure and composition of montane and lowland tropical forest in the Serranía Pilón Lajas, Beni, Bolivia.In DallmeierF. J. A.Comiskey Forest Biodiversity in North, Central and South America and the Caribbean: Research and MonitoringMan and the Biosphere Series, no. 22pp681700Carnforth, UKUNESCO, The Parthenon Publishing GroupGoogle Scholar
StebbinsG. L. 1950Variation and evolution in plantsNew YorkColumbia University PressGoogle Scholar
StewardJ. H. 1948Handbook of South American Indians. Vol. 3. The Forest TribesWashington, DCBureau of American Ethnography & The Smithsonian InstitutionGoogle Scholar
StotzD. F. J. W.Fitzpatrick T. A.ParkerIII D. K.Moskovits 1996Neotropical Birds: Ecology and ConservationChicagoUniversity of Chicago PressGoogle Scholar
TierneyP. 2000Darkness in El Dorado: How Scientists and Journalists Devastated the AmazonNew YorkWW Norton and CompanyGoogle Scholar
TreeceD. 1988Brutality and Brazil: The Human Cost of Cheap Steel.Multinational Monitor92Online. Available: http://multinationalmonitor.org/hyper/issues/1988/02/mm0288_08.html#nameGoogle Scholar
TrollC. 1968The Cordilleras of the Tropical Americas: Aspects of Climatic, Phytogeographical and Agrarian EcologyBonnFerd DümmlersGoogle Scholar
(UNFCCC) United Nations Framework Convention on Climate Change 2006Background Paper for the Workshop on Reducing Emissions from Deforestation in Developing Countries30 August – 1 September 2006. Rome, Italy. Online. Available: http://unfccc.int/methods_and_science/lulucf/items/3757.phpGoogle Scholar
VeigaM. M. 1997Mercury in Artisanal Gold Mining in Latin America: Facts, Fantasies and Solutions.UNIDO -Expert Group Meeting: Introducing new technologies for abatement of global mercury pollution deriving from artisanal gold miningVienna. July 1–3. Online. Available: http://www.facome.uqam.ca/. November 5, 2006Google Scholar
WanderlyI. F. R. L.Fonseca P. G.Pereira P.do A. C.Prado A.de A. B.Oliveira F. P.Barbosa F.Panciera 2007Implicações da Iniciativa de Integração da Infraestrutura Regional Sulamericana e projetos correlacionados na política de conservação no Brasil.InPolítica Ambiental, no. 3BrasíliaConservation InternationalOnline. Available: http://www.conservacao.org/publicacoes/index.php?t=5Google Scholar
WarhurstA. 1998Mining and the Environment: Case Studies from the AmericasOttowaInternational Development Research CenterGoogle Scholar
World Bank 1991Environmental Assessment SourcebookVol. 1Policies, Procedures, and Cross-sectoral Issues. World Bank Technical Paper Number 139Washington, DCWorld BankGoogle Scholar
World Bank 2003aA Common Framework: Converging Requirements of Multilateral Financial Institutions No. 1, Environmental Impact Assessment (EIA)WashingtonWorld BankOnline. Available: http://www1.worldbank.org/harmonization/romehlf/Background/MFI%20Final%20Jan17%202003-Eng.pdf. May 15, 2007Google Scholar
World Bank 2006Finding Sustainable Ways to Extract Forest Products in the Amazon.Pilot Program Extractive Reserves ProjectOnline. Available: http://web.worldbank.org/WBSITE/EXTERNAL/COUNTRIES/LACEXT/BRAZILEXTN/0,,contentMDK:20754543~pagePK:141137~piPK:141127~theSitePK:322341,00.html. June 1, 2007Google Scholar
Tables A.1 through A.4 provide simple models that estimate the value of the carbon stored in Amazonian forests (Table A.1), the value of the carbon released each year via deforestation (Table A.2), the potential value of a 5 percent reduction in deforestation in the eight countries of the greater Amazon Wilderness Area compared against the documented baseline deforestation rates (Table A.3), and the potential value of a 5 percent reduction in deforestation for four Andean countries when compared to a Business as Usual Scenario (Table A.4). Tables A.5 through A.7 provide statistics on protected areas and indigenous lands.
 53 The Interoceanic Corridor is a project in the Peru–Brazil–Bolivia Hub rather than a component of the Central Interocenic Hub that spans Bolivia and Brazil to the south (see Figure 1.1 and Figure A.2).
 54 In Colombia, potatoes are cultivated near the ecotone between cloud forest and high altitude paramo grasslands.
 55 The presence of humid forest refugia in the Pleistocene does not imply that Amazonian speciation occurred during that period (e.g., Refugium Hypothesis), which is generally believed to be a Tertiary phenomenon (see Maslin 2005), but only that several disjunctive or loosely connected areas on the Andean piedmont would have provided appropriate conditions for humid forest species (Killeen et al. 2007a).
 56 Annual land use change in the Brazilian Amazon is approximately 20,000 km2 yr−1 (Laurence et al. 2004); in Bolivia it is 2,400 km2 yr−1 (Killeen et al. 2007b). Studies are underway in Colombia, Ecuador, Peru, Venezuela, and the Guayanas, with a sum of 3,000 km2 yr−1 being a conservative estimate for these countries (see Table A.2).
 57 Conservation International–Brazil is working with farmers and ranchers to restore areas cleared beyond the legally allowed amount; voluntary certification schemes are being promoted to reward farmers in compliance by providing secured access to European markets.