Get 20M+ Full-Text Papers For Less Than $1.50/day. Start a 14-Day Trial for You or Your Team.

Learn More →

Do we need to devalue Amazonia and other large tropical forests?

Do we need to devalue Amazonia and other large tropical forests? Introduction Concern over loss of global habitat and biotic diversity coupled with the widely held assertion that there is not enough funding to save more than a fraction of the remaining habitat in many important regions has lead to a number of laudable attempts to prioritize global hotspots and establish conservation funding priorities ( Olson & Dinerstein, 1998 ; Myers , 2000 ). A primary appeal of these approaches is that they offer a package to politicians and funding agencies both elegant in simplicity and realizable in scope. A negative consequence of such approaches is that, intentionally or not, they devalue areas that do not make the short lists. In a similar vein, one approach taken to highlight the value of some regions has been to employ comparative values such as numbers of endemic species, in order to establish the relative worth of certain areas ( Mares, 1992a ; Fjeldså & Rahbek, 1998 ). Depending on how these data are presented, such approaches also devalue regions with lower comparative values. We question, on a global or continental scale, whether it is wise to take approaches such as these that pit areas against one another. We are not attempting here to review the growing field of prioritization. However, while we agree that prioritization is helpful in bringing attention to the plight of biologically important regions, we are concerned that the structure of these approaches may needlessly lead to devaluation of other areas of high biotic diversity. As a particular example of this trend, we discuss recent treatments of an area that we are familiar with, Amazonia. Although recent publications predict a disturbing future for Amazonia ( Laurance , 2001 ), there has been a trend in the conservation literature that has suggested Amazonia and other large tropical forests (the Congo Basin and New Guinea) can be lower priorities than other regions. Thus, we focus on several technical pitfalls and consequences of various approaches that have evaluated Amazonia and emphasize the need for better data on the regional heterogeneity that may exist in these large tropical forests. When tropical deforestation first began receiving global conservation attention, habitat loss in parts of Amazonia, in particular dramatic images from the Brazilian State of Rondônia, were highlighted in the international press. Thus, despite Amazonia’s tremendous size, it became a poster child for habitat loss. As global concern about habitat loss increased, scientists studying declining biodiversity in other regions wanted to bring these disturbing trends to international attention. In many cases, the plight of these other ecoregions (or ecosystems) was and continues to be worse than that of Amazonia in terms of relative percentage of habitat destroyed ( Myers , 2000 ); however, in highlighting these other areas, several authors have resorted to approaches that directly attempt to devalue the biotic riches of Amazonia. AMAZONIA VS. THE ‘NEOTROPICAL DRYLANDS’ In a paper published in Science , entitled ‘Neotropical mammals and the myth of Amazonian diversity’, Mares (1992a) highlighted the biodiversity value of the Neotropical drylands in general and the cerrado region in particular. He assembled species lists of mammals and purported to show that mammalian diversity was higher in the South American drylands than in Amazonia. This approach has been criticized by various authors (e.g. Chesser & Hackett, 1992 ; Voss & Emmons, 1996 ) because of bias in how species numbers were totalled (for example, Mares added together endemic species from all South American drylands and compared total diversity and number of endemic taxa to that of Amazonian forest, including the Chocó and Pacific lowlands, but not Atlantic forest). These critics argued that Mares’ comparisons were constructed to bias results in favour of the drylands to highlight habitat loss in these regions. Although the South American drylands have not received adequate attention from the conservation community, the very title of the paper demonstrates the adversarial approach that was taken (also see Mares, 1992b ). AMAZONIA VS. THE ANDES Unfortunately, others also have followed this path to some extent. Fjeldså and colleagues have used species‐level comparisons between Amazonia and the Andes to document the need to conserve habitat in the tropical Andes. They also have employed the rhetoric that Amazonian diversity is a ‘myth’ ( Fjeldså & Rahbek, 1998 ). Using endemic avian species and the presumed evolutionary age of these taxa, they suggest that the Andean regions of South America are a more important ‘species pump’ than Amazonia, and conclude that the tropical Andes should have a higher global conservation priority. These authors appreciate the diversity that exists in Amazonia, and there is no doubt that more forested area exists in Amazonia than in the Andes, but they and others have made little effort to examine the complexities of the comparisons they have made. For example, the number of endemic biological species of birds in the Andes has been calculated to be twice that of Amazonia ( Stotz , 1996 , Table 3.12). This is a striking figure, yet we would argue that it does not speak to a myth of Amazonian diversity. For example, one family (the hummingbirds, Trochilidae) is responsible for 94 of the 318 endemic Andean species (30%, data from Stotz , 1997 ). Conservation of hummingbirds clearly requires a focus on the Andes; however, other avian families have their centres of diversity and endemism in Amazonia ( Fjeldså & Rahbek, 1998 ). Such patterns are well known in the conservation literature (e.g. Prendergast , 1993 ). The analyses of Fjeldså & Rahbek (1998) also suggest that Amazonia is a region dominated by widely distributed homogenous species, and therefore preserving existing protected areas will protect the region’s biotic diversity. In short, Amazonia is already well protected. We believe that reliance on the Biological Species Concept ( Mayr, 1963 ) in current avian classifications greatly underestimates the diversity of Amazonia, and therefore affects analyses of both levels and patterns of diversity such as the GAP analyses they present. Although still few in number and restricted in geographical scope, studies of genetic differentiation in Amazonian birds (e.g. Capparella, 1988 ; Bates , 1999 ; Bates, 2000 ), mammals ( Patton , 1994 ; Jacobs , 1995 ; Patton , 2000 ) and amphibians ( Gaston , 1998 ) show that substantial levels of genetic differentiation within Amazonian biological species are common. These levels are equal to levels of genetic divergence found in congeneric species from other regions ( Johns & Avise, 1998 ). Analyses of vocal characters in biological species of Amazonian birds with presumably innate songs also show substantial levels of regional differentiation ( Isler , 1997, 1999 ). There is no doubt that unrecognized genetic differentiation also exists in the Andes, but for historical reasons related to the relative ages of the two regions (uplift of parts of the Andes has been relatively recent), levels of divergence in the Andes may often be lower than those in Amazonia (e.g. Bates & Zink, 1994 ; Garcia‐Moreno , 1999 ). We can examine this possible taxonomic artifact in another way using described subspecies instead of species in calculating numbers of endemic taxa for the two regions. Although subspecies have been considered problematic units in ornithology, such comparisons can still be informative ( Bates , 1998 ). We created a data set to determine the number of endemic subspecies of passerines or perching birds (Order Passeriformes) from both the Andes and Amazonia (Demos and Bates unpubl. data). We assume that this taxonomic level is roughly equivalent to the number of phylogenetic species ( sensu Zink & McKittrick, 1995 ) that would likely be recognized, and we believe that this approach provides a more meaningful indicator of the relative amounts of avian diversity in Amazonia (also see Peterson & Navarro‐Següenza, 1999 ; Peterson , 2000 ). At the level of biological species, the ratio of endemic species of passerines in the Andes vs. Amazonia is 1.7 : 1. If we calculate this ratio incorporating subspecies of polytypic species for the two regions, the ratio is lowered to 1.4 : 1. Thus, although it still may be true that the Andes harbour more endemic taxa than Amazonia using either concept, what seems most important is that both regions harbour high diversity. An additional contention, that Amazonian lineages are older than Andean lineages, also has been used to suggest that the diversity of Andean birds is of higher value ( Fjeldså, 1995 ; Fjeldså & Lovett, 1997 ). This hypothesis, that the Andes are a ‘species pump’ for younger lineages, also rests on the assumption that biological species in Amazonia do not have high levels of genetic structure. As mentioned earlier, genetic and vocal differences in some Amazonian taxa suggest that this may be an artifact of current taxonomy. Because we believe this is the case, we are not convinced that the impressive frequency of adaptation to local ecological conditions in the Andes ( Fjeldså, 1997, 1999 ) is necessarily more critical to preserve than the genetically structured populations found across Amazonia. Once again, this approach devalues Amazonian diversity in an effort to highlight the high levels of endemism in the tropical Andes. AMAZONIA AND GLOBAL PRIORITIZATION FOR CONSERVATION We also believe that Amazonia has been devalued unnecessarily in recent prioritization projects at the global scale. Here, Amazonia has been either not included in lists because of its large size ( Myers , 2000 ), or broken up into subregions, some of which were not considered priorities ( Olson & Dinerstein, 1998 ). Along with the Congo Basin and New Guinea, Myers (2000) excluded Amazonia from their list of global ‘hotspots’ because, despite high diversity in these areas, substantial habitat remains intact. These authors know the value of these regions, but they believe that habitat loss in other regions and the relative levels of endemic biotic diversity these areas harbour supersedes the need to prioritize habitat conservation in these ‘larger’ regions. We agree that the global hotspots identified by Myers . (2000) are greatly in need of conservation, but we question whether we want to devalue larger areas such as Amazonia. We are especially concerned about global priority‐setting projects that persist in assuming that Amazonia is a single biogeographic unit (e.g. Beissinger , 1996 ; Myers , 2000 ) when studies of many groups have shown that it has distinct biogeographic subdivisions (e.g. Müller, 1973 ; Haffer, 1978 ; Cracraft, 1985 ; Silva & Oren, 1996 ; Bates , 1998 ; Ron, 2000 ). To their credit, Myers . (2000) mention that substantial portions of the Guianan region and western Amazonia are intact, but they fail to mention that this is not true for south‐eastern Amazonia. The ‘Global 200 Ecoregions’ analysis ( Olson & Dinerstein, 1998 ) has divided Amazonia into regions. Initially, Amazonian upland forest was split into four regions, the Guianan region, north‐western Amazonia (their Rio Negro‐Juruá moist forests), south‐western Amazonia and south‐eastern Amazonia. Of the four, south‐eastern Amazonia was left off the original priority list. Recently, the list has been expanded to 233 regions and includes Rondônia (a distinctive region on the south bank on the Amazon between the Rios Madeira and Tapajos, http://www.wwf.org/global200/ , Dinerstein, personal communication). This is laudable, but their approach still leaves the Belém/Pará region, which lies south of the Amazon to the east of Rondônia (east of the Rio Tapajos), off the list. This is unfortunate for several reasons. First, we believe that the endemism (a primary criterion in the Global 200 approach) of this region is currently under‐estimated. For several vertebrate groups, the apparent lack of endemic taxa is largely an artifact of using biological species. Parsimony Analyses of Endemism on birds, primates, anurans and lizards ( Silva & Oren, 1996 ; Bates , 1998 ; Ron, 2000 ) have demonstrated the distinctiveness of this region in analyses that have used species and/or subspecies. Again, levels of genetic differentiation between taxa from these regions and others in Amazonia can be quite high (e.g. Patton , 2000 ). We know too little about the distributions of other groups such as plants, but there is evidence that their distributions also are not homogeneous across Amazonia ( Henderson , 1995 ; Terborgh & Andresen, 1998 ). Second, forest in the Belém/Pará region has been reduced to an extent that unfortunately rivals regions included in the Myers et al . global hotspot list and deforestation continues ( Instituto Nacional de Pesquisas Espaciais, 2000 ; Laurance , 2001 ). Level of habitat loss is not a criterion of the Global 200 approach, but the degree of habitat loss is well documented in various assessments (also see Bryant , 1997 ). Finally, there are currently few protected areas within the Belem/Pará regions ( Peres & Terborgh, 1995 ). We suspect that New Guinea and the Congo Basin forests, which were also considered by Myers et al . to be large enough to leave off their list, possess similarly distinct and valuable subregions. Birds of New Guinea have a strong tendency to be locally distributed or geographically variable ( Diamond, 1985 ; Beehler , 1986 ; Cracraft, 1992 ). Few data exist for the Congo Basin, but distinct populations have been documented here as well ( Beresford & Cracraft, 1999; Roy , 2001 ). The categorization of the large tropical forests of the world, Amazonia, New Guinea, and possibly the Congo, as single entities under approaches such as the hotspots of Myers . (2000) combined with the undervaluation that has resulted from other treatments of Amazonia could be disastrous. Such approaches leave distinct and threatened subregions such as the Belém/Pará region in much greater risk of being further deforested, and in addition, lack of attention will only serve to promote development in other parts of the basin. On the heels of these global prioritization studies, the Brazilian government has launched an ambitious project called Avança Brasil ( Laurance , 2001 ; Peres, 2001 ), a massive project designed to bring $40 billion dollars worth of roads and other development to the southern Brazilian Amazon ( Peres, 2001 ). The global prioritization studies discussed above could be used to suggest that such large development projects are not affecting a unique biota. Major NGOs such as Conservation International and World Wildlife Fund do have less publicized initiatives underway to create additional protected areas across Amazonia. However, from an outside prospective, the major global prioritization projects still often treat Amazonia as a vast homogeneous region which means that areas such as the Belém/Pará region may become nonpriorities. We suggest that other distinct regions in the three largest tropical forests could receive similar treatment. We believe that no myths exist regarding Amazonian diversity, even if Amazonia does have more undisturbed area than other ecoregions. Amazonian diversity is divided into evolutionarily distinct subregions that even today are not understood for many organisms. Amazonia is an incredibly diverse region that should not be considered adequately protected and subsequently neglected ( Laurance , 2001 ). Our concerns lead us to make several suggestions. First, we urge the conservation community to look beyond simple comparisons of endemicity that pit ecoregions against one another on continental scales. We would advocate a more regional approach that also considers international borders. In the case of South America, it is probably true that the tropical forests of the Andes are a higher conservation priority than some Amazonian forests (based on habitat size and lack of people). However, for Brazil, a country with no Andean forests, the Belém/Pará region (which lies completely within the country) deserves a conservation priority that equals that given to the cerrado and Atlantic forests, two regions that are considered global hotspots by Myers . (2000) . We believe that south‐eastern Amazonia also warrants inclusion as both a global hotspot and a valuable ecoregion. Finally, we agree with others (e.g. Patton , 2000 ) who advocate gathering the data necessary to further investigate and document floral and faunal subdivision in the large tropical forests of Amazonia, New Guinea and the Congo Basin, so that the diversity in subregions is better documented for future prioritization projects. Not to do this will make leave subregions more vulnerable to development pressures and lead to the loss of globally unique biodiversity. Acknowledgments While the opinions expressed in this paper are our own, this manuscript was improved by the comments of J.M. Cardoso da Silva, W. Figueiredo, P.Z. Goldstein, S. J. Hackett, L. Heaney, W. Laurance, B. Patterson, J.G. Tello, and discussions with E. Dinerstein, R. Foster, D. Stotz and Joel Brown’s laboratory group. http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png Diversity and Distributions Wiley

Do we need to devalue Amazonia and other large tropical forests?

Download PDF
7 pages

Loading next page...
 
/lp/wiley/do-we-need-to-devalue-amazonia-and-other-large-tropical-forests-n0LoJYTgru

References (44)

Publisher
Wiley
Copyright
Copyright © 2001 Wiley Subscription Services, Inc., A Wiley Company
ISSN
1366-9516
eISSN
1472-4642
DOI
10.1046/j.1366-9516.2001.00112.x
Publisher site
See Article on Publisher Site

Abstract

Introduction Concern over loss of global habitat and biotic diversity coupled with the widely held assertion that there is not enough funding to save more than a fraction of the remaining habitat in many important regions has lead to a number of laudable attempts to prioritize global hotspots and establish conservation funding priorities ( Olson & Dinerstein, 1998 ; Myers , 2000 ). A primary appeal of these approaches is that they offer a package to politicians and funding agencies both elegant in simplicity and realizable in scope. A negative consequence of such approaches is that, intentionally or not, they devalue areas that do not make the short lists. In a similar vein, one approach taken to highlight the value of some regions has been to employ comparative values such as numbers of endemic species, in order to establish the relative worth of certain areas ( Mares, 1992a ; Fjeldså & Rahbek, 1998 ). Depending on how these data are presented, such approaches also devalue regions with lower comparative values. We question, on a global or continental scale, whether it is wise to take approaches such as these that pit areas against one another. We are not attempting here to review the growing field of prioritization. However, while we agree that prioritization is helpful in bringing attention to the plight of biologically important regions, we are concerned that the structure of these approaches may needlessly lead to devaluation of other areas of high biotic diversity. As a particular example of this trend, we discuss recent treatments of an area that we are familiar with, Amazonia. Although recent publications predict a disturbing future for Amazonia ( Laurance , 2001 ), there has been a trend in the conservation literature that has suggested Amazonia and other large tropical forests (the Congo Basin and New Guinea) can be lower priorities than other regions. Thus, we focus on several technical pitfalls and consequences of various approaches that have evaluated Amazonia and emphasize the need for better data on the regional heterogeneity that may exist in these large tropical forests. When tropical deforestation first began receiving global conservation attention, habitat loss in parts of Amazonia, in particular dramatic images from the Brazilian State of Rondônia, were highlighted in the international press. Thus, despite Amazonia’s tremendous size, it became a poster child for habitat loss. As global concern about habitat loss increased, scientists studying declining biodiversity in other regions wanted to bring these disturbing trends to international attention. In many cases, the plight of these other ecoregions (or ecosystems) was and continues to be worse than that of Amazonia in terms of relative percentage of habitat destroyed ( Myers , 2000 ); however, in highlighting these other areas, several authors have resorted to approaches that directly attempt to devalue the biotic riches of Amazonia. AMAZONIA VS. THE ‘NEOTROPICAL DRYLANDS’ In a paper published in Science , entitled ‘Neotropical mammals and the myth of Amazonian diversity’, Mares (1992a) highlighted the biodiversity value of the Neotropical drylands in general and the cerrado region in particular. He assembled species lists of mammals and purported to show that mammalian diversity was higher in the South American drylands than in Amazonia. This approach has been criticized by various authors (e.g. Chesser & Hackett, 1992 ; Voss & Emmons, 1996 ) because of bias in how species numbers were totalled (for example, Mares added together endemic species from all South American drylands and compared total diversity and number of endemic taxa to that of Amazonian forest, including the Chocó and Pacific lowlands, but not Atlantic forest). These critics argued that Mares’ comparisons were constructed to bias results in favour of the drylands to highlight habitat loss in these regions. Although the South American drylands have not received adequate attention from the conservation community, the very title of the paper demonstrates the adversarial approach that was taken (also see Mares, 1992b ). AMAZONIA VS. THE ANDES Unfortunately, others also have followed this path to some extent. Fjeldså and colleagues have used species‐level comparisons between Amazonia and the Andes to document the need to conserve habitat in the tropical Andes. They also have employed the rhetoric that Amazonian diversity is a ‘myth’ ( Fjeldså & Rahbek, 1998 ). Using endemic avian species and the presumed evolutionary age of these taxa, they suggest that the Andean regions of South America are a more important ‘species pump’ than Amazonia, and conclude that the tropical Andes should have a higher global conservation priority. These authors appreciate the diversity that exists in Amazonia, and there is no doubt that more forested area exists in Amazonia than in the Andes, but they and others have made little effort to examine the complexities of the comparisons they have made. For example, the number of endemic biological species of birds in the Andes has been calculated to be twice that of Amazonia ( Stotz , 1996 , Table 3.12). This is a striking figure, yet we would argue that it does not speak to a myth of Amazonian diversity. For example, one family (the hummingbirds, Trochilidae) is responsible for 94 of the 318 endemic Andean species (30%, data from Stotz , 1997 ). Conservation of hummingbirds clearly requires a focus on the Andes; however, other avian families have their centres of diversity and endemism in Amazonia ( Fjeldså & Rahbek, 1998 ). Such patterns are well known in the conservation literature (e.g. Prendergast , 1993 ). The analyses of Fjeldså & Rahbek (1998) also suggest that Amazonia is a region dominated by widely distributed homogenous species, and therefore preserving existing protected areas will protect the region’s biotic diversity. In short, Amazonia is already well protected. We believe that reliance on the Biological Species Concept ( Mayr, 1963 ) in current avian classifications greatly underestimates the diversity of Amazonia, and therefore affects analyses of both levels and patterns of diversity such as the GAP analyses they present. Although still few in number and restricted in geographical scope, studies of genetic differentiation in Amazonian birds (e.g. Capparella, 1988 ; Bates , 1999 ; Bates, 2000 ), mammals ( Patton , 1994 ; Jacobs , 1995 ; Patton , 2000 ) and amphibians ( Gaston , 1998 ) show that substantial levels of genetic differentiation within Amazonian biological species are common. These levels are equal to levels of genetic divergence found in congeneric species from other regions ( Johns & Avise, 1998 ). Analyses of vocal characters in biological species of Amazonian birds with presumably innate songs also show substantial levels of regional differentiation ( Isler , 1997, 1999 ). There is no doubt that unrecognized genetic differentiation also exists in the Andes, but for historical reasons related to the relative ages of the two regions (uplift of parts of the Andes has been relatively recent), levels of divergence in the Andes may often be lower than those in Amazonia (e.g. Bates & Zink, 1994 ; Garcia‐Moreno , 1999 ). We can examine this possible taxonomic artifact in another way using described subspecies instead of species in calculating numbers of endemic taxa for the two regions. Although subspecies have been considered problematic units in ornithology, such comparisons can still be informative ( Bates , 1998 ). We created a data set to determine the number of endemic subspecies of passerines or perching birds (Order Passeriformes) from both the Andes and Amazonia (Demos and Bates unpubl. data). We assume that this taxonomic level is roughly equivalent to the number of phylogenetic species ( sensu Zink & McKittrick, 1995 ) that would likely be recognized, and we believe that this approach provides a more meaningful indicator of the relative amounts of avian diversity in Amazonia (also see Peterson & Navarro‐Següenza, 1999 ; Peterson , 2000 ). At the level of biological species, the ratio of endemic species of passerines in the Andes vs. Amazonia is 1.7 : 1. If we calculate this ratio incorporating subspecies of polytypic species for the two regions, the ratio is lowered to 1.4 : 1. Thus, although it still may be true that the Andes harbour more endemic taxa than Amazonia using either concept, what seems most important is that both regions harbour high diversity. An additional contention, that Amazonian lineages are older than Andean lineages, also has been used to suggest that the diversity of Andean birds is of higher value ( Fjeldså, 1995 ; Fjeldså & Lovett, 1997 ). This hypothesis, that the Andes are a ‘species pump’ for younger lineages, also rests on the assumption that biological species in Amazonia do not have high levels of genetic structure. As mentioned earlier, genetic and vocal differences in some Amazonian taxa suggest that this may be an artifact of current taxonomy. Because we believe this is the case, we are not convinced that the impressive frequency of adaptation to local ecological conditions in the Andes ( Fjeldså, 1997, 1999 ) is necessarily more critical to preserve than the genetically structured populations found across Amazonia. Once again, this approach devalues Amazonian diversity in an effort to highlight the high levels of endemism in the tropical Andes. AMAZONIA AND GLOBAL PRIORITIZATION FOR CONSERVATION We also believe that Amazonia has been devalued unnecessarily in recent prioritization projects at the global scale. Here, Amazonia has been either not included in lists because of its large size ( Myers , 2000 ), or broken up into subregions, some of which were not considered priorities ( Olson & Dinerstein, 1998 ). Along with the Congo Basin and New Guinea, Myers (2000) excluded Amazonia from their list of global ‘hotspots’ because, despite high diversity in these areas, substantial habitat remains intact. These authors know the value of these regions, but they believe that habitat loss in other regions and the relative levels of endemic biotic diversity these areas harbour supersedes the need to prioritize habitat conservation in these ‘larger’ regions. We agree that the global hotspots identified by Myers . (2000) are greatly in need of conservation, but we question whether we want to devalue larger areas such as Amazonia. We are especially concerned about global priority‐setting projects that persist in assuming that Amazonia is a single biogeographic unit (e.g. Beissinger , 1996 ; Myers , 2000 ) when studies of many groups have shown that it has distinct biogeographic subdivisions (e.g. Müller, 1973 ; Haffer, 1978 ; Cracraft, 1985 ; Silva & Oren, 1996 ; Bates , 1998 ; Ron, 2000 ). To their credit, Myers . (2000) mention that substantial portions of the Guianan region and western Amazonia are intact, but they fail to mention that this is not true for south‐eastern Amazonia. The ‘Global 200 Ecoregions’ analysis ( Olson & Dinerstein, 1998 ) has divided Amazonia into regions. Initially, Amazonian upland forest was split into four regions, the Guianan region, north‐western Amazonia (their Rio Negro‐Juruá moist forests), south‐western Amazonia and south‐eastern Amazonia. Of the four, south‐eastern Amazonia was left off the original priority list. Recently, the list has been expanded to 233 regions and includes Rondônia (a distinctive region on the south bank on the Amazon between the Rios Madeira and Tapajos, http://www.wwf.org/global200/ , Dinerstein, personal communication). This is laudable, but their approach still leaves the Belém/Pará region, which lies south of the Amazon to the east of Rondônia (east of the Rio Tapajos), off the list. This is unfortunate for several reasons. First, we believe that the endemism (a primary criterion in the Global 200 approach) of this region is currently under‐estimated. For several vertebrate groups, the apparent lack of endemic taxa is largely an artifact of using biological species. Parsimony Analyses of Endemism on birds, primates, anurans and lizards ( Silva & Oren, 1996 ; Bates , 1998 ; Ron, 2000 ) have demonstrated the distinctiveness of this region in analyses that have used species and/or subspecies. Again, levels of genetic differentiation between taxa from these regions and others in Amazonia can be quite high (e.g. Patton , 2000 ). We know too little about the distributions of other groups such as plants, but there is evidence that their distributions also are not homogeneous across Amazonia ( Henderson , 1995 ; Terborgh & Andresen, 1998 ). Second, forest in the Belém/Pará region has been reduced to an extent that unfortunately rivals regions included in the Myers et al . global hotspot list and deforestation continues ( Instituto Nacional de Pesquisas Espaciais, 2000 ; Laurance , 2001 ). Level of habitat loss is not a criterion of the Global 200 approach, but the degree of habitat loss is well documented in various assessments (also see Bryant , 1997 ). Finally, there are currently few protected areas within the Belem/Pará regions ( Peres & Terborgh, 1995 ). We suspect that New Guinea and the Congo Basin forests, which were also considered by Myers et al . to be large enough to leave off their list, possess similarly distinct and valuable subregions. Birds of New Guinea have a strong tendency to be locally distributed or geographically variable ( Diamond, 1985 ; Beehler , 1986 ; Cracraft, 1992 ). Few data exist for the Congo Basin, but distinct populations have been documented here as well ( Beresford & Cracraft, 1999; Roy , 2001 ). The categorization of the large tropical forests of the world, Amazonia, New Guinea, and possibly the Congo, as single entities under approaches such as the hotspots of Myers . (2000) combined with the undervaluation that has resulted from other treatments of Amazonia could be disastrous. Such approaches leave distinct and threatened subregions such as the Belém/Pará region in much greater risk of being further deforested, and in addition, lack of attention will only serve to promote development in other parts of the basin. On the heels of these global prioritization studies, the Brazilian government has launched an ambitious project called Avança Brasil ( Laurance , 2001 ; Peres, 2001 ), a massive project designed to bring $40 billion dollars worth of roads and other development to the southern Brazilian Amazon ( Peres, 2001 ). The global prioritization studies discussed above could be used to suggest that such large development projects are not affecting a unique biota. Major NGOs such as Conservation International and World Wildlife Fund do have less publicized initiatives underway to create additional protected areas across Amazonia. However, from an outside prospective, the major global prioritization projects still often treat Amazonia as a vast homogeneous region which means that areas such as the Belém/Pará region may become nonpriorities. We suggest that other distinct regions in the three largest tropical forests could receive similar treatment. We believe that no myths exist regarding Amazonian diversity, even if Amazonia does have more undisturbed area than other ecoregions. Amazonian diversity is divided into evolutionarily distinct subregions that even today are not understood for many organisms. Amazonia is an incredibly diverse region that should not be considered adequately protected and subsequently neglected ( Laurance , 2001 ). Our concerns lead us to make several suggestions. First, we urge the conservation community to look beyond simple comparisons of endemicity that pit ecoregions against one another on continental scales. We would advocate a more regional approach that also considers international borders. In the case of South America, it is probably true that the tropical forests of the Andes are a higher conservation priority than some Amazonian forests (based on habitat size and lack of people). However, for Brazil, a country with no Andean forests, the Belém/Pará region (which lies completely within the country) deserves a conservation priority that equals that given to the cerrado and Atlantic forests, two regions that are considered global hotspots by Myers . (2000) . We believe that south‐eastern Amazonia also warrants inclusion as both a global hotspot and a valuable ecoregion. Finally, we agree with others (e.g. Patton , 2000 ) who advocate gathering the data necessary to further investigate and document floral and faunal subdivision in the large tropical forests of Amazonia, New Guinea and the Congo Basin, so that the diversity in subregions is better documented for future prioritization projects. Not to do this will make leave subregions more vulnerable to development pressures and lead to the loss of globally unique biodiversity. Acknowledgments While the opinions expressed in this paper are our own, this manuscript was improved by the comments of J.M. Cardoso da Silva, W. Figueiredo, P.Z. Goldstein, S. J. Hackett, L. Heaney, W. Laurance, B. Patterson, J.G. Tello, and discussions with E. Dinerstein, R. Foster, D. Stotz and Joel Brown’s laboratory group.

Journal

Diversity and Distributions – Wiley

Published: Sep 1, 2001

There are no references for this article.