Acta Amazonica

=20

services

Print version ISSN 0044-5967

Acta Amaz. vol.39 no.4 Manaus 2009

doi: 10.1590/S0044-59672009000400030

Global warming=20 in Amazonia: Impacts and Mitigation

Aquecimento=20 Global na Amaz=F4nia: Impactos e Mitiga=E7=E3o

Philip Martin=20 Fearnside

Instituto Nacional=20 de Pesquisas da Amaz=F4nia - INPAC. P. 478. CEP 69.011-970 = Manaus-Amazonas.=20 E-mail: pmfearn@inpa.gov.br

ABSTRACT

Global = warming has=20 potentially catastrophic impacts in Amazonia, while at the same time = maintenance=20 of the Amazon forest offers one of the most valuable and cost-effective = options=20 for mitigating climate change. We know that the El Ni=F1o phenomenon, = caused by=20 temperature oscillations of surface water in the Pacific, has serious = impacts in=20 Amazonia, causing droughts and forest fires (as in 1997-1998). = Temperature=20 oscillations in the Atlantic also provoke severe droughts (as in 2005). = We also=20 know that Amazonian trees die both from fires and from water stress = under hot,=20 dry conditions. In addition, water recycled through the forest provides = rainfall=20 that maintains climatic conditions appropriate for tropical forest, = especially=20 in the dry season. What we need to know quickly, through intensified = research,=20 includes progress in representing El Ni=F1o and the Atlantic = oscillations in=20 climatic models, representation of biotic feedbacks in models used for=20 decision-making about global warming, and narrowing the range of = estimating=20 climate sensitivity to reduce uncertainty about the probability of very = severe=20 impacts. Items that need to be negotiated include the definition of = "dangerous"=20 climate change, with the corresponding maximum levels of greenhouse = gases in the=20 atmosphere. Mitigation of global warming must include maintaining the = Amazon=20 forest, which has benefits for combating global warming from two = separate roles:=20 cutting the flow the emissions of carbon each year from the rapid = pace of=20 deforestation, and avoiding emission of the stock of carbon in = the=20 remaining forest that can be released by various ways, including climate = change=20 itself. Barriers to rewarding forest maintenance include the need for = financial=20 rewards for both of these roles. Other needs are for continued reduction = of=20 uncertainty regarding emissions and deforestation processes, as well as=20 agreement on the basis of carbon accounting. As one of the countries = most=20 subject to impacts of climate change, Brazil must assume the leadership = in=20 fighting global warming.

Keywords:=20 Carbon, Climate change, Deforestation, Environmental services, Global = warming,=20 Greenhouse effect

RESUMO

O = aquecimento=20 global tem impactos potencialmente catastr=F3ficos na Amaz=F4nia, e, ao = mesmo tempo,=20 a manuten=E7=E3o da floresta amaz=F4nica oferece uma das op=E7=F5es mais = valiosas e=20 baratas para mitigar as mudan=E7as clim=E1ticas. N=F3s sabemos que o = fen=F4meno de El=20 Ni=F1o, causado por uma oscila=E7=E3o da temperatura da superf=EDcie da = =E1gua no=20 Pac=EDfico, tem impactos s=E9rios na Amaz=F4nia, causando secas e = inc=EAndios=20 florestais, como aconteceram em 1997-1998. Oscila=E7=F5es de temperatura = no=20 Atl=E2ntico tamb=E9m provocam secas severas, como em 2005. N=F3s = tamb=E9m sabemos que=20 =E1rvores amaz=F4nicas morrem, tanto do fogo como do estresse h=EDdrico = sob condi=E7=F5es=20 quentes e secas. Al=E9m disso, a =E1gua reciclada pela floresta fornece = chuva que=20 mant=E9m as condi=E7=F5es clim=E1ticas apropriadas para floresta = tropical, especialmente=20 durante a esta=E7=E3o seca. O que n=F3s precisamos saber com urg=EAncia, = por meio de=20 pesquisa intensificada, inclui como representar melhor o El Ni=F1o e as = oscila=E7=F5es=20 no Atl=E2ntico, nos modelos clim=E1ticos, como representar as = retroalimenta=E7=F5es=20 bi=F3ticas nos modelos usados para tomada de decis=E3o sobre o efeito = estufa, e um=20 estreitamento da gama das estimativas da sensitividade clim=E1tica (para = reduzir a=20 incerteza sobre a probabilidade de impactos muito severos). Assuntos que = precisam ser negociados incluem a defini=E7=E3o de mudan=E7a de clima = "perigosa", com=20 os correspondentes n=EDveis m=E1ximos das concentra=E7=F5es de gases de = estufa na=20 atmosfera. Mitiga=E7=E3o do efeito estufa tem que incluir a = manuten=E7=E3o da floresta=20 amaz=F4nica, o que traz benef=EDcios para o combate ao efeito estufa por = meio de=20 dois pap=E9is separados: diminuir o fluxo de emiss=F5es de = carbono que=20 acontece em cada ano devido ao ritmo r=E1pido do desmatamento, e evitar = a emiss=E3o=20 do estoque de carbono na floresta restante que pode ser liberada = de=20 v=E1rias maneiras, inclusive por causa da pr=F3pria mudan=E7a = clim=E1tica. Barreiras=20 impedindo a recompensa=E7=E3o da manuten=E7=E3o de floresta incluem a = necessidade por=20 recompensas financeiras para ambos estes pap=E9is. Outras necessidades = s=E3o=20 continuar a redu=E7=E3o da incerteza relativo =E0s emiss=F5es e ao = processo de=20 desmatamento, assim como tamb=E9m um acordo sobre a base da = contabilidade do=20 carbono. Por ser um dos pa=EDses mais sujeito aos impactos da mudan=E7a = clim=E1tica, o=20 Brasil tem que assumir a lideran=E7a na luta contra o aquecimento = global.=20

Palavras-chave: = Aquecimento=20 global, Carbono, Desmatamento, Efeito estufa, Mudan=E7a clim=E1tica, = Servi=E7os=20 Ambientais

I.) = THE ROLE OF=20 AMAZON FOREST IN GLOBAL WARMING

Climate = change is a=20 threat to the Amazon forest and loss of the forest is a threat to the = climate.=20 Land-use change in Amazonia produces globally significant emissions of=20 greenhouse gases such as carbon dioxide, methane and nitrous oxide = (Fearnside,=20 2007a, 2008a, 2008b). Because each hectare of deforestation has a = significant=20 net emission, reducing deforestation rates represents an avoided impact. = The=20 potential value of this avoided impact is much larger than the value = that can be=20 earned through traditional land uses such as the sale of timber and = cattle,=20 which are based on destruction of the forest. Avoided deforestation is a = means=20 of transforming the value of the forest's environmental services into a = cash=20 flow that can be used to maintain both the forest and the human = population that=20 defends it. Avoiding emissions of greenhouse gases represents the = environmental=20 service that it is closest to becoming a significant source of income in = Amazonia, while maintaining biodiversity and water cycling are potential = sources=20 of value over the long term (Fearnside, 2008c). Use of the value of the = forest's=20 role in maintaining the global carbon balance depends on a reliable=20 quantification of the emissions caused by deforestation.

Global = warming is a=20 grave threat to the world as a whole, and Brazil, including Amazonia, is = one of=20 the places expected to suffer the most severe impacts if emissions of = greenhouse=20 gases continue unchecked. It is therefore necessary to reduce the total = global=20 emission from all sources, whether or not the sources count as "direct=20 human-induced emissions" that are the responsibility of any particular = country=20 (Allen et al., 2009; Meinshausen et al., 2009).

II.) = IMPACTS OF=20 GLOBAL WARMING IN AMAZONIA

A.) = WHAT WE=20 KNOW

1.) El = Ni=F1o impacts=20 in Amazonia

The key = question=20 for Amazonia is the future of El Ni=F1o. The recent Fourth Assessment = Report of=20 the Intergovernmental Panel on Climate Change (IPCC) finds agreement = among=20 models on an increase in "El Ni=F1o-like conditions" with continued = global warming=20 (Meehl et al., 2007, pp. 779-780). This refers to warm water in = the=20 Pacific Ocean, which triggers El Ni=F1o. However, the report finds no = agreement=20 among the models on El Ni=F1o itself, meaning the droughts and floods at = different=20 locations around the world. El Ni=F1o is very difficult to model and = should=20 obviously be a top priority. It is hard to get all the symptoms = happening=20 simultaneously in a model, with rain on the coast of Peru, drought in = Amazonia,=20 floods in southern Brazil, drought in Ethiopia and Borneo, etc. Our = problem in=20 Amazonia is that we know from direct experience that whenever there is = warm=20 water in the Pacific we have droughts and forest fires in Amazonia. If a = climate=20 model shows the Pacific warming up and nothing happening in Amazonia, it = means=20 that there is something missing from the model, not that we are safer = here. The=20 link between warm water in the Pacific Ocean and droughts and forest = fires in=20 Amazonia is demonstrated by the complete coincidence of these events in = Amazonia=20 with past "El Ni=F1o-like conditions," for example in 2003, 1997 and = 1982. El Ni=F1o=20 occurrence and impact is therefore an anchor in reality - it derives = from direct=20 observation and does not depend on the results of climate = models.

2.) = Atlantic=20 oscillation

A = previously=20 unappreciated climatic threat became apparent in 2005 when a devastating = drought=20 struck Amazonia. Streamflows in tributaries on the south side of the = Amazon=20 River were so low that boats could not navigate the rivers and riverside = communities were isolated from hospitals and other essential services. = Forest=20 fires burned in Acre and in neighboring parts of the state of Amazonas, = an=20 unprecedented event (Brown et al., 2006; Vasconcelos and Brown, = 2007).=20 The forest lost biomass through stunted growth and increased tree = mortality=20 (Phillips et al., 2009). The year 2005 was not an El Ni=F1o year; = rather=20 than warm water in the Pacific there was warm water in the southern part = of the=20 North Atlantic and colder-than-usual water in the northern part of the = South=20 Atlantic. The warm water in the North Atlantic gave energy to Hurricane = Katrina=20 that struck New Orleans that year. It also contributed to the Amazon = drought by=20 causing more air to rise when the Intertropical Convergence Zone (ITCZ) = reached=20 the area of warm water. The air at the ITCZ rises to an altitude of=20 approximately 1800 m, divides into northward and southward flows, and = then moves=20 towards the poles by approximately 30=BA of latitude before descending = to ground=20 level and returning towards the equator at low altitude, thus forming = the Hadley=20 cell. When the air rises, its moisture content condenses and falls as = rain, and=20 when the air later descends to ground level it is dry and desiccates the = area=20 where it comes to earth. With more hot, humid air rising at the ITCZ in = 2005,=20 there was also more cold, dry air coming down 30=BA farther south. With = the=20 temperature gradient between warm water in the North Atlantic and cold = in the=20 South Atlantic, the ITCZ was drawn further north than usual, and at the = time of=20 year when the seasonal movement of the ITCZ was near its northern = extreme, dry=20 air was descending over the headwaters of the Amazon tributaries on the = southern=20 side of the basin (e.g,, Fearnside, 2006a; Marengo et al., = 2008).=20

The = warm water in=20 the North Atlantic in 2005 was the combined result of several factors. = One was=20 the Atlantic Multi-decadal Oscillation (AMO), which produces = warmer-than-average=20 water in this area at approximately 40-year intervals. However, this = oscillation=20 explains no more than 11% of the anomalous temperature in 2005, whereas = 50% of=20 the temperature increase could be directly attributed to global warming=20 (Trenberth and Shea, 2006). There was also a 22% indirect contribution = from=20 global warming from the remains of an El Ni=F1o period in the preceding = years. In=20 addition, the reduction of industrial air pollution in Europe and North = America=20 resulted in less shielding of this area of ocean by aerosols (Cox et = al.,=20 2008).

The = north-south=20 temperature gradient in the Atlantic is significantly correlated with = rainfall=20 in the southwestern portion of Amazonia, and in 2005 both the Atlantic = gradient=20 and drought in southwestern Amazonia were at extreme levels (Cox et = al.,=20 2008). Results from the Hadley Centre model indicate a tremendous = increase in=20 the magnitude of the Atlantic temperature gradient and associated Amazon = droughts if greenhouse-gas emissions continue their current path (Cox = et=20 al., 2008). The simulation results indicate that the probability of = a=20 drought as severe as that of 2005 was 5% (1 year in 20) in 2005, but = this would=20 rise to 50% (1 year in 2) by 2025 and 90% (9 years in 10) by 2060. The=20 probability of these droughts explodes if atmospheric CO₂ = exceeds 400=20 ppmv, a level only slightly higher than the 2009 level of 387 ppmv. = Atmospheric=20 CO₂ concentrations are rising at approximately 2.6=20 ppmv/year.

3.) = Forest=20 dieback

The = Amazon forest=20 is quite vulnerable to climate changes in the direction predicted to = result from=20 continued global warming: higher temperatures combined with less = rainfall,=20 including longer dry seasons. When temperatures are higher, trees need = more=20 water just to survive. Trees in standing forest can be killed due to = climate=20 variability in today's climate, as occurred in the 1997-1998 El Ni=F1o = drought=20 (caused by warm water in the Pacific Ocean) and in the severe drought in = 2005=20 (caused by warm water in the Atlantic). In addition to trees essentially = dying=20 of thirst during droughts, the greater chance of forest fires starting = and=20 spreading greatly increases the risk to the forest were the climate to = change as=20 projected (Nepstad et al., 1999, 2001). Tropical forests are more = susceptible to fire than are other vegetation types because forest fires = have=20 been so rare over the last millennia that the tropical-forest trees did = not need=20 to evolve defenses against fire. Amazonian trees have thin bark, and = when fires=20 occur the cambium is heated under the bark at the base of the trunk, and = the=20 trees die. Trees in other biomes, such as the cerrado, have thick = bark=20 and resist fire.

Various = lines of=20 evidence indicate that Amazonian forest can succumb to these changes. In = the=20 INPA/STRI Biological Dynamics of Forest Fragments Project north of = Manaus, over=20 65,000 trees have been tagged and monitored; the study shows clearly = that trees=20 die with much greater frequency near forest edges, where microclimatic=20 conditions are hotter and drier than in the interior of a continuous = forest=20 (Nascimento and Laurance, 2004; Laurance et al., 2006). The same = result=20 has been confirmed near Santar=E9m by the "Seca Floresta" Project, which = is a part=20 of the Large-Scale Atmosphere Biosphere Experiment in Amazonia (LBA). = Plastic=20 sheeting was installed to intercept 60% of the throughfall in the forest = over a=20 one-hectare plot, and tree mortality greatly increased (Nepstad et = al.,=20 2007). Both studies found that large trees are the first to die, thus = both=20 increasing the release of carbon and allowing canopy opening to further = dry the=20 forest microclimate.

Unfortunately,=20 Brazil's official positions in this area have been to deny the problem: = when the=20 IPCC released its report on impacts in April 2007, the Folha de S=E3o = Paulo=20 reported that "Yesterday [5 April 2007] the Folha de S=E3o Paulo=20 [newspaper] found that the Brazilian delegation [to the = Intergovernmental Panel=20 on Climate Change meeting in Brussels] objected to the final text of the = summary=20 mentioning savannization of Amazonia caused by global warming" (Folha = de S=E3o=20 Paulo, 2007). However, no less than four chapters of the IPCC report = highlight the danger of savannization in Amazonia and the summary for = policy=20 makers was approved with the statement that "By mid century, increases = in=20 temperature and associated decreases in soil water are projected to lead = to=20 gradual replacement of tropical forest by savanna in eastern Amazonia" = (IPCC,=20 2007a). The Brazilian delegation's position is symptomatic. Just as US = President=20 George W. Bush denied the very existence of global warming, thereby = freeing=20 himself of any onus to do something about the problem, the Brazilian=20 delegation's efforts to avert official recognition of the danger of=20 savannization in Amazonia has the effect of forestalling the need for = the=20 Brazilian government to confront the problem, including containing = deforestation=20 and reducing greenhouse-gas emissions.

4.)=20 Evapotranspiration and rainfall

Simulations=20 indicate that if over 40% of the original forest in Amazonia were cut = and=20 converted to cattle pasture or soybeans, the rainfall in eastern = Amazonia would=20 decline precipitously, especially during the dry season (Sampaio et = al.,=20 2007). These changes in precipitation could lead to degradation of the = remainder=20 of the forest in this area. Whether or not 40% deforestation is the = "tipping=20 point," or threshold, the existence of such a threshold has long been = suspected=20 and constitutes a strong argument for stopping deforestation (Fearnside, = 1985,=20 1997). As of 2008 a total of 18% of the originally forested area in the = Legal=20 Amazon region had been cleared (Brazil, INPE, 2009). However, no safe = level of=20 deforestation exists below which there is no risk. Rather than a single = "tipping=20 point," every tree that falls increases slightly the probability that=20 irreversible feedbacks will be set in motion and destroy the remaining = forest.=20 This is a risk that Brazil and the world can ill afford.

B.) = WHAT WE NEED TO=20 KNOW

1.) = Modeling El=20 Ni=F1o

Although our=20 knowledge contains various gaps, it is important to emphasize that this = fact=20 does not represent a valid reason for any delay in actions to fight = global=20 warming. On the contrary, the existence of uncertainties means that we = should be=20 reducing emissions of greenhouse gases even more in order to insure that = the=20 worst-case scenarios do not become reality. This is the well-known=20 "precautionary principle," which should be applied to policies on global = warming=20 but obviously isn't.

There = are several=20 things we need to know fast. The various global climate models show = widely=20 different results for some parts of the world, especially Amazonia. The = Hadley=20 Centre model, from the UK Meteorological Office, shows a complete = catastrophe,=20 with the Amazon forest being essentially wiped out by the year 2080 as a = result=20 of heat and drought (Cox et al., 2000, 2004). Several other = models show a=20 more modest but still serious drying in the Amazon, including the = National=20 Center for Atmospheric Research (NCAR) model from the USA and the Max = Planck=20 Institute model from Germany. Some others show little change, and one = (the=20 Geophysical Fluid Dynamics Laboratory model from the USA) even showed = (as of the=20 time of the 2007 IPCC report) an increase in Amazon rainfall (Kundzewicz = et=20 al., 2007, p. 183), but subsequent improvements in the model have = eliminated=20 this unrealistic behavior (Stephen M. Griffies, personal communication, = 2009).=20 Given the importance of the question of what happens to the Amazon = rainforest,=20 it is amazing that discrepancies of the magnitude still existing among = models=20 have not been resolved. The Hadley Centre results were first published = in 2000,=20 and subsequent refinements of their model have only reinforced the = catastrophic=20 conclusion for Amazonia. Of the 21 models submitted to Phase 2 of the = Coupled=20 Model Intercomparison Project (CMIP2) (http://www.pcmdi.llnl.gov/CMIP); Covey et al. = (2003)=20 the HadCM3LC Hadley Centre model produced the best representation of the = link=20 between "El Ni=F1o-like conditions" and Amazon droughts (see Cox et = al.,=20 2004, p. 153). The Hadley Centre model is not perfect: it produces a = present-day=20 climate in Amazonia that is hotter (C=E2ndido et al., 2007) and = dryer=20 (Huntingford et al., 2004) than the real climate. Establishment = of a=20 "permanent El Ni=F1o" indicated by the Hadley model is the key feature = leading to=20 the catastrophic result in Amazonia. However, the danger of = savannization does=20 not depend on the Hadley Centre model proving to be the best = representation of=20 future climate: an average of the results of 15 models indicates = savannization=20 of eastern Amazonia as a consequence of continued global warming = (Salazar et=20 al., 2007).

Linking = El Ni=F1o to=20 global warming has tremendous moral and political implications, in = addition to=20 the scientific ones. When El Ni=F1o killed 200,000 people in Ethiopia in = 1982=20 there was an enormous outpouring of charity and sympathy, with rock = stars=20 singing "We are the World" to attract donations for the victims. But the = drought=20 was presented as an act of God that was nobody's fault in particular. It = wasn't=20 your car or your deforestation that caused that disaster. But if the = frequency=20 of El Ni=F1o is linked to global warming, then everything changes = because=20 greenhouse-gas emissions have identifiable culprits. Contributions to = global=20 warming can be quantified and allotted to countries and = individuals.

2.) = Biotic=20 feedbacks

There = are several=20 important features that were left out of the models used by the IPCC = that need=20 to be analyzed and included. One is the "biotic feedbacks," especially = the=20 carbon that would be released from the biomass and from the soil if the = Amazon=20 forest were to die and be replaced by grassland with scattered trees or, = alternatively, by a low-biomass woody second growth. Only half a dozen = of the=20 over 20 models analyzed by the IPCC had the capability of simulating the = biotic=20 feedbacks, but these features were switched off in order to make the = simulations=20 of all models comparable. If these feedbacks are included the = temperature in=20 2100 is 38% higher (Cox et al., 2004, p. 138). Amazonia is an = important=20 part of this effect.

Biotic = feedbacks=20 involve not only the carbon stock in forest biomass but also the stock = in the=20 soil under the forest. When forest is replaced by grass, whether through = deforestation or climate-induced mortality, much of this soil stock is = released=20 (Fearnside and Barbosa, 1998). In fact, even without loss of the forest = cover=20 the increase in soil temperature from global warming has the potential = to=20 release massive amounts of carbon to the atmosphere, forming an = additional=20 positive feedback relationship with climate change (Fearnside,=20 2009a).

3.) = Climate=20 sensitivity

Another = important=20 topic is the matter of "climate sensitivity." This is the amount by = which the=20 average global temperature at equilibrium would increase over the = pre-industrial=20 average with the concentration of atmospheric CO₂ at double = the=20 pre-industrial level, a benchmark that would be crossed in approximately = 2050=20 without mitigation. The "most probable" value for this is approximately = 2.8=BAC=20 and the assumption of climate sensitivity close to this value is the = basis of=20 all of the model runs that generated the graphs and maps in the IPCC = report's=20 summary for policy makers. The problem is that this "most probable" = value for=20 climate sensitivity implies that there is a 50% chance that the real = value is=20 higher than this, and it could be a lot higher. The point at which there = is 95%=20 assurance that the real value is included is now believed to be about = 6.2=BAC=20 according to a reassessment in 2006, down from a previous assessment = that was=20 substantially higher (Hegerl et al., 2006). There needs to be = much more=20 scientific work on constraining the estimates of climate sensitivity, = and much=20 more policy work on the implications of changes at the high end of the=20 range.

The = relevance of=20 the more devastating results at the high end of the climate-sensitivity = range=20 can be visualized by an example. In 1998 the Palace II apartment = building in Rio=20 de Janeiro collapsed. For someone living in a building, the possibility = that it=20 might collapse represents a catastrophe. If someone who lives in an = apartment=20 building were to ask an engineer if the building will collapse and were = to=20 receive the answer that "it is most likely that the building will stay=20 standing," would the apartment dweller be satisfied? This means that = there may=20 be a 51% chance that the building will stay standing, but there could be = a 49%=20 chance that it will collapse like the Palace II. The apartment dweller = will=20 therefore not be satisfied with a 50% probability of the building = remaining=20 standing, and will want an assurance of 99% or more. The same applies to = the=20 climate change scenarios based on a climate sensitivity value that = represents=20 the "most likely" outcome.

C.) = WHAT WE NEED TO=20 NEGOTIATE

The = United Nations=20 Framework Convention on Climate Change (UN-FCCC), better known as the = "climate=20 convention," specifies the purpose of the Convention as preventing the=20 atmospheric concentrations of greenhouse gases from reaching levels that = cause=20 "dangerous interference with the global climate system" (UN-FCCC, 1992, = Article=20 2). Negotiation of the definition of "dangerous interference" is = currently=20 underway.

The = significance of=20 defining a maximum increase in global temperature, and the corresponding = permissible concentrations of greenhouse gases, is that the nations of = the world=20 must achieve this global target, and it no longer matters where the = emissions=20 are coming from. This gives Amazonia a substantial additional role in = mitigating=20 climate change. Not only does the forest have a role from the large = annual=20 emission provoked by deforestation, but the stock of carbon in the = remaining=20 forest has a value beyond its potential for future deforestation. The = stock of=20 carbon in the remaining forest could be released by climate change = itself, and=20 not only global climate change from the greenhouse effect but also = partly from=20 regional climate change resulting from loss of evapotranspiration as=20 deforestation proceeds.

Due to = the=20 importance of the Amazonian forest to Brazil, it is very important that = the=20 country use its diplomatic weight to press for a very low value as the=20 definition of this number. The definition of dangerous climatic change = should be=20 at the most 2=C2=B0C above the pre-industrial level. This is the value = chosen by the=20 European Union as their definition of "dangerous" and is now endorsed by = over=20 100 countries. Substantial uncertainty surrounds the equilibrium = CO₂₌₂₀concentration to which this amount of warming corresponds, but it = is in=20 the neighborhood of 400 ppmv of CO₂ equivalent, including the = effects=20 of the trace gases such as methane and nitrous oxide. Since the = concentration of=20 CO₂ alone is already at 387 ppmv as of 2009, and the trace = gases=20 elevate the total effect to over 400 ppmv, the cuts in the global = emissions will=20 have to be very large and fast (Hare and Meinshausen, 2006). Therefore, = all=20 mitigation options need to be used, including both reduction in burning = fossil=20 fuels and decrease in deforestation.

The = forest is=20 already vulnerable, especially in the dryer areas near the eastern and = southern=20 edges of the forest (Hutyra et al., 2005; Salazar et al., = 2007).=20 Any further change in the direction expected from global warming = increases the=20 risk of savannization. Because of delays within the climate system, = warming=20 would continue for 20-30 years even if anthropogenic emissions were to = halt=20 immediately. After that period, however, the difference in simulated = global=20 temperatures is very great depending on what course emissions take over = the=20 coming years (e.g., IPCC, 2007b). The course of future emissions = depends=20 on international negotiations currently underway to define "dangerous=20 interference with the global climate system."

The = delays in the=20 climate system mean that actions must be taken well before critical = thresholds=20 are passed. Because of the tremendous mass of the oceans (and because = water=20 vapor that evaporates is a greenhouse gas) the water continues to warm = and=20 affect climate even after emissions are reduced: the result of a kind = "inertia."=20 Controlling climate change is analogous to steering a large ship, which = can take=20 kilometers to change course after the helm is turned. In the case of the = Titanic, the crew saw the iceberg before the ship hit it, but they = couldn't get=20 the ship to turn fast enough. In the case of climate, we may foresee a=20 catastrophe occurring in model results after 2050, with the Amazon = forest=20 drying, but if we wait until the last minute to reduce emissions it will = be too=20 late.

III.) MITIGATION=20 BY AMAZON FOREST MAINTENANCE

A.) = WAYS TO FIGHT=20 GLOBAL WARMING IN AMAZONIA

A wide = variety of=20 mitigation options has been proposed in Amazonia, including plantation=20 silviculture (including charcoal production), agroforestry, forest = management=20 for timber, reduced-impact logging, soil carbon enhancement (including = charcoal=20 additions, no-till agriculture and pasture fertilization), and, most = recently,=20 biofuels (including oil palm). However, tapping the native forest's role = in=20 mitigation is by far the best both in terms of carbon benefits = (Fearnside, 1995)=20 and social and environmental concerns (Fearnside, 1996, 2006b). This = discussion=20 will be limited to the maintenance of standing forest.

Deforestation can=20 be reduced by a variety of means, including command-and-control programs = using=20 inspections and fines, creation of protected areas of a variety of types = (including "sustainable use" areas), various kinds of integrated = development=20 projects aimed at channeling labor and capital resources to sustainable = land=20 uses in deforested areas instead of clearing forest, direct payment for=20 environmental services, and policy measures affecting infrastructure=20 construction, agricultural credit, land-tenure and taxation. = Environmental=20 agencies such as the Brazilian Institute for the Environment and = Renewable=20 Natural Resources (IBAMA) only have access to some of these options, = such as=20 command-and-control measures, reserve creation and payment for = environmental=20 services.

It = cannot be=20 overemphasized that infrastructure decisions must be an integral part of = efforts=20 to control deforestation, rather than being separate as at present. = Today major=20 highways and other developments are, in practice, decided by decree = based on=20 political considerations, and the role of environmental agencies is = restricted=20 (in practice) to legalizing the construction process and suggesting = small=20 changes to minimize impacts. This must change such that a full = evaluation of=20 impacts, including resulting deforestation and climate impacts, are = considered=20 before decisions are made on building the infrastructure = (e.g.,=20 Fearnside and Gra=E7a, 2006; Fearnside, 2007b). In order for the = deforestation and=20 emissions impacts of individual infrastructure decisions to be = quantified (and,=20 conversely, the benefits of the "no project" option), one must have=20 deforestation models with realistic provisions for the stimulation = effect of=20 roads on deforestation and the percentage of "leakage" when = deforestation is=20 inhibited by reserves and other measures. "Leakage" refers to effects a=20 mitigation project provokes outside of its boundaries that cancel out = all or=20 part of the climatic benefits of the project itself, for example if = people who=20 would have deforested in a protected areas simply move somewhere else in = the=20 forest and continue clearing. Existing models for regional deforestation = (e.g., Soares-Filho et al., 2006) do not yet have this = capability,=20 but progress has been made towards enhancing these models to simulate = the=20 effects of individual decisions (Fearnside et al., 2009). This = opens the=20 possibility of much more effective proposals for capturing carbon credit = through=20 avoided deforestation, as well as providing a tool for improved = decision-making=20 on infrastructure.

B.) = BARRIERS TO=20 REWARDING FOREST MAINTENANCE

If = keeping=20 Amazonian forest standing is to assume a role in strategies to mitigate = global=20 warming one must look at what has so far prevented this from happening = and then=20 address these issues. The forest really has two separate roles with = respect to=20 global warming: the flow of carbon to the atmosphere represented = by=20 annual emissions from clearing (which can be reduced through avoided=20 deforestation), and the stock of carbon in the standing forest.=20 Deforestation activity in Amazonia is spatially highly concentrated, = over 80% of=20 it occurring in the "arc of deforestation" along the eastern and = southern edges=20 of the forest. In this area a system of rewarding avoided deforestation = could be=20 established where emissions are compared to a baseline (presumably based = on the=20 recent history of clearing). This is what is implied by the Kyoto = Protocol under=20 what is commonly known as "additionality," or what is additional as = compared to=20 what would have happened in the absence of a mitigation project. In = parts of the=20 region where little or no deforestation has taken place, a different = system of=20 rewarding environmental services is needed based on stocks. Clearly the = role of=20 the forest in mitigating global warming will only be tapped if the way = that the=20 reward is calculated for each location results in a reasonable=20 return.

Brazil's diplomatic=20 community has traditionally resisted the idea of rewarding the Amazon = forest's=20 climatic role under the UN-FCCC (e.g., CFR Independent Task = Force. 2001;=20 Fearnside, 2001). This has its root in a lack of confidence that = Brazil's=20 government is capable of controlling deforestation, the implication = being that=20 were Brazil to agree to reduce emissions by a given amount and then find = that=20 deforestation could not be reduced as promised, the country would (in = the view=20 of these people) be exposed to pressures that would interfere with = Brazil's=20 sovereignty in Amazonia. The solution to this is an unambiguous = demonstration=20 that the government can control deforestation if it is serious about = doing so.=20 Several lines of evidence indicate that the government has this ability=20 (Fearnside, 2003; Fearnside and Barbosa, 2003).

A = different set of=20 concerns has been raised by various national governments and other = entities that=20 are committed to designing a system of mitigating global warming that is = sufficiently reliable to be used as credit against emissions from fossil = fuels.=20 These include the level of uncertainty associated with each hectare of=20 deforestation that is avoided and uncertainty as to how many hectares = have been=20 avoided (see Fearnside, 2000). Both require improvements of data and of=20 monitoring capabilities. Much has improved in both respects. We at INPA = have=20 made significant progress in better quantifying the carbon stocks in the = forests=20 being cleared, in accounting for the emissions and uptakes that occur = after=20 clearing, and in modeling the process and distribution of deforestation. = In=20 addition, the large amounts of emission that can be avoided at = relatively low=20 cost mean that uncertainty can always be more than compensated for by = granting=20 less credit than the amount of physical emission believed to be avoided. = In=20 addition to these concerns over the data used in carbon accounting, = there are=20 also disagreements over the theoretical basis of the accounting itself,=20 particularly with regard to "permanence" (the time the carbon remains = out of the=20 atmosphere), or more generally the value attributed to time (Fearnside, = 2002a,=20 2002b; Fearnside et al., 2000) and leakage (Fearnside, 2009b). = Several=20 proposals exist to address these problems, including proposals generated = at=20 INPA. Lastly, negotiations for rewarding environmental services must = address the=20 question of what is to be done with the money such that it ensures = attainment of=20 both objectives: maintaining the forest with its environmental services = and=20 maintaining the wellbeing of the human population in the Amazonian=20 interior.

C.) = BRAZIL MUST=20 ASSUME LEADERSHIP

Brazil = is one of=20 the countries of the world with the greatest expected impacts from = global=20 warming (Dias et al., 2007; Fearnside, 2008d; Magrin et = al.,=20 2007). In addition to massive mortality of the Amazonian forest, other = parts of=20 the country would also be affected. In the Northeast global warming = would cause=20 drying in an area that already suffers constantly from lack of rain. In = the=20 South it would increase torrential rain events and typhoons. Sea-level = rise=20 affects the entire coast where much of the country's population lives.=20

Brazil = must make a=20 quantitative commitment to reduce deforestation. It is important that = this be=20 under the climate convention, as opposed to internal commitments that = might be=20 rewarded through voluntary funds outside of the Kyoto process, as = proposed by=20 Brazil at the conferences of the parties in Nairobi, Bali and Poznan in = 2006,=20 2007 and 2008 (e.g., Brazil, 2006; Brazil, Comit=EA = Interministerial sobre=20 Mudan=E7a do Clima, 2008). If the industrialized countries decide to = fight global=20 warming in a serious way, agreeing in large cuts in their emissions = under the=20 climate convention, all of their resources will be used to meet their = negotiated=20 targets and nothing will remain for voluntary funds.

At the = United=20 Nations meeting on climate change, held in Vienna in August 2007, the = Brazilian=20 representative repeated the country's position since the 1992 "Earth = Summit"=20 (ECO-92) of avoiding any quantitative commitment regarding emissions = (O=20 Estado de S=E3o Paulo, 2007). Undoubtedly, Brazil can choose to = continue to=20 resist taking on a commitment for a few more years, but sooner or later = it will=20 be forced to make a commitment if global warming is to be controlled. It = is in=20 Brazil's national interest that such a change of position be made now = rather=20 than later. Brazil must assume the leadership in the fight against = global=20 warming.

ACKNOWLEDGEMENTS

The = Conselho=20 Nacional do Desenvolvimento Cient=EDfico e Tecnol=F3gico (CNPq: Proc. = 306031/2004-3,=20 557152/2005-4, 420199/2005-5, 474548/2006-6), Rede GEOMA and Instituto = Nacional=20 de Pesquisas da Amaz=F4nia (INPA: PRJ05.57) contributed financial = support. The=20 portion of this text on mitigation was presented as a discussion paper = for=20 INPA's Grupo de Estudos Estrat=E9gicos Amaz=F4nicos (GEEA).

LITERATURE=20 CITED

Allen, = M. R.;=20 Frame; D. J.; Huntingford, C.; Jones, C. D.; Lowe, J. A.; Meinshausen, = M.;=20 Meinshausen, N. 2009. Warming caused by cumulative carbon emissions = towards the=20 trillionth tonne. Nature, 458: 1163-1166.=20 [ Links ]

Brazil. = 2006. The=20 12^th Conference of the Parties of the UNFCCC, Nairobi, Kenya. = Positive incentives for voluntary action in developing countries to = address=20 climate change: Brazilian perspective on reducing emissions from = deforestation.=20 United Nations Framework Convention on Climate Change (UNFCCC), Bonn, = Germany. 4=20 pp. (Available at: http://unfccc.int/files/meetings/dialogue/application/pdf= /wp_21_braz.pdf).=20 [ Links ]

Brazil, = Comit=EA=20 Interministerial sobre Mudan=E7a do Clima. 2008. Plano Nacional sobre = Mudan=E7a=20 do Clima - PNMC -- Brasil. Minist=E9rio do Meio Ambiente, = Bras=EDlia, DF,=20 Brazil. 129 pp. (Available at: http://www.mma.gov.br/estruturas/imprensa/_arquivos/96_01= 122008060233.pdf) &nb= sp;[ Links ]

Brazil, = INPE=20 (Instituto Nacional de Pesquisas Espaciais). 2009. Projeto PRODES: = Monitoramento=20 da Floresta Amaz=F4nica Brasileira por Sat=E9lite. INPE, S=E3o Jos=E9 = dos Campos, S=E3o=20 Paulo, Brazil. (Available at: http://www.obt.inpe.br/prodes/).=20 [ Links ]

Brown, = I. F.;=20 Schroeder, W.; Setzer, A.; Maldonado, M. J. R.; Pantoja, N.; Duarte, A. = F.;=20 Marengo, J. 2006. Monitoring fires in Southwestern Amazonia rain forest. = EOS,=20 Transactions of the American Geophysical Union, 87(26): 253-264.=20 [ Links ] =

C=E2ndido, L. A.;=20 Manzi, A. O.; Tota, J.; da Silva, P. R. T.; da Silva, F. S. M.; dos = Santos, R.=20 N. N.; Correia, F. W. S. 2007. O Clima atual e futuro da Amaz=F4nia nos = cen=E2rios=20 do IPCC: A quest=E3o da savaniza=E7=E3o. Ci=EAncia e Cultura, = 59(3): 44-47.=20 [ Links ]

CFR = Independent=20 Task Force. 2001. A letter to the President and a memorandum on U. S. = policy=20 toward Brazil. Council on Foreign Relations (CFR), New York, U.S.A. 13 = pp.=20 (Available at: http://www.cfr.org/).=20 [ Links ]

Covey, = C.;=20 AchutaRao, K. M.; Cubasch, U.; Jones, P.; Lambert, S. J.; Mann, M. E.; = Phillips,=20 T. J. Taylor, K. E. 2003. An overview of results from the Coupled Model=20 Intercomparison Project. Global and Planetary Change, 37: = 103-133.=20 [ Links ]

Cox, P. = M.; Betts,=20 R. A.; Jones, C. D.; Spall, S. A.; Totterdell, I. J. 2000. Acceleration = of=20 global warming due to carbon-cycle feedbacks in a coupled climate model. = Nature, 408: 184-187.=20 [ Links ]

Cox, P. = M.; Betts,=20 R. A.; Collins, M.; Harris, P. P.; Huntingford, C.; Jones, C. D. 2004. = Amazonian=20 forest dieback under climate-carbon cycle projections for the = 21^st=20 century. Theoretical and Applied Climatology, 78: 137-156,=20 doi:10.1007/s00704-004-0049-4.=20 [ Links ]

Cox, P. = M.; Harris,=20 P. P.; Huntingford, C.; Betts, R. A.; Collins, M.; Jones, C. D.; Jupp, = T. E;.=20 Marengo, J. A.; Nobre, C. A. 2008. Increasing risk of Amazonian drought = due to=20 decreasing aerosol pollution. Nature, 453: 212-215.=20 [ Links ]

Dias, = P. L. S.;=20 Ribeiro, W. C.; Nunes, L. H. (Eds.). 2007. A Contribution to = Understanding=20 the Regional Impacts of Global Change in South America. Instituto de = Estudos=20 Avan=E7ados, Universidade de S=E3o Paulo, S=E3o Paulo, Brazil. 418 pp.=20 [ Links ]

O = Estado de S=E3o=20 Paulo. 2007. "Brasil diz que combate efeito estufa, mas n=E3o aceita = metas" 27=20 August 2007. (Available at: http://www.estadao.com.br/vidae/not_vid41671,0.htm). = [ Links ]

Fearnside, P. M.=20 1985. Environmental change and deforestation in the Brazilian Amazon. = In:=20 Hemming, J. (Ed.) Change in the Amazon Basin: Man's Impact on Forests = and=20 Rivers. Manchester University Press, Manchester, U. K. pp. 70-89.=20 [ Links ]

Fearnside, P. M.=20 1995. Global warming response options in Brazil's forest sector: = Comparison of=20 project-level costs and benefits. Biomass and Bioenergy, 8: = 309-322.=20 [ Links ]

Fearnside, P. M.=20 1996. Socio-economic factors in the management of tropical forests for = carbon.=20 In: Apps, M. J.; Price, D. T. (Eds.). Forest Ecosystems, Forest = Management=20 and the Global Carbon Cycle. NATO ASI Series, Subseries I "Global=20 Environmental Change," Vol. 40. Springer-Verlag, Heidelberg, Germany. = pp.=20 349-361. [ Links ]

Fearnside, P. M.=20 1997. Limiting factors for development of agriculture and ranching in = Brazilian=20 Amazonia. Revista Brasileira de Biologia, 57(4): 531-549.=20 [ Links ]

Fearnside, P. M.=20 2000. Uncertainty in land-use change and forestry sector mitigation = options for=20 global warming: Plantation silviculture versus avoided deforestation. = Biomass=20 and Bioenergy, 18: 457-468.=20 [ Links ]

Fearnside, P. M.=20 2001. Saving tropical forests as a global warming countermeasure: An = issue that=20 divides the environmental movement. Ecological Economics, 39: = 167-184.=20 [ Links ]

Fearnside, P. M.=20 2002a. Why a 100-year time horizon should be used for global warming = mitigation=20 calculations. Mitigation and Adaptation Strategies for Global = Change, 7:=20 19-30. [ Links ]

Fearnside, P. M.=20 2002b. Time preference in global warming calculations: A proposal for a = unified=20 index. Ecological Economics, 41: 21-31.=20 [ Links ]

Fearnside, P. M.=20 2003. Deforestation control in Mato Grosso: A new model for slowing the = loss of=20 Brazil's Amazon forest. Ambio, 32: 343-345.=20 [ Links ]

Fearnside, P. M.=20 2006a. A vazante na Amaz=F4nia e o aquecimento global. Ci=EAncia = Hoje,=20 38(231): 76-78. = [ Links ]

Fearnside, P. M.=20 2006b. Mitigation of climatic change in the Amazon. In: Laurance, W. F.; = Peres,=20 C. A. (Eds.) Emerging Threats to Tropical Forests. University of = Chicago=20 Press, Chicago, Illinois, U.S.A., pp. 353-375.=20 [ Links ]

Fearnside, P. M.=20 2007a. Uso da terra na Amaz=F4nia e as mudan=E7as clim=E2ticas globais. = Brazilian=20 Journal of Ecology, 10: 83-100.=20 [ Links ]

Fearnside, P. M.=20 2007b. Brazil's Cuiab=E2-Santar=E9m (BR-163) Highway: The environmental = cost of=20 paving a soybean corridor through the Amazon. Environmental = Management,=20 39: 601-614. [ Links ]

Fearnside, P. M.=20 2008a. Quantifica=E7=E3o do servi=E7o ambiental do carbono nas florestas = amaz=F4nicas=20 brasileiras. Oecologia Brasiliensis, 12: 743-756.=20 [ Links ] =

Fearnside, P. M.=20 2008b. Deforestation in Brazilian Amazonia and global warming. Annals = of Arid=20 Zone, 47(3-4): 1-20. (in press).=20 [ Links ]

Fearnside, P. M.=20 2008c. Amazon forest maintenance as a source of environmental services. = Anais=20 da Academia Brasileira de Ci=EAncia, 80: 101-114.=20 [ Links ]

Fearnside, P. M.=20 2008d. Mudan=E7as clim=E2ticas globais e a floresta amaz=F4nica. In: = Buckeridge, M. S.=20 (Ed.) Biologia e Mudan=E7as Clim=E2ticas Globais no Brasil. RiMa = Editora, S=E3o=20 Paulo, Brazil, pp. 131-150.=20 [ Links ]

Fearnside, P. M.=20 2009a. Estoque e estabilidade do carbono nos solos na Amaz=F4nia = brasileira.=20 In: Teixeira, W. G.; Madari, B. E.; Benites, V. M.; Kern, D. C.; = Falc=E3o,=20 N. P. S. (Eds.) As Terras Pretas de =CDndio: Caracteriza=E7=E3o e = manejo para=20 forma=E7=E3o de novas =E1reas. Empresa Brasileira de Pesquisa = Agropecu=E1ria=20 (EMBRAPA) Amaz=F4nia Ocidental, Bel=E9m, Par=E1, Brazil. pp. 259-262.=20 [ Links ]

Fearnside, P. M.=20 2009b. Carbon benefits from Amazonian forest reserves: Leakage = accounting and=20 the value of time. Mitigation and Adaptation Strategies for Global=20 Change. 14(6): 557-567.=20 [ Links ]

Fearnside, P. M.;=20 Barbosa, R. I. 1998. Soil carbon changes from conversion of forest to = pasture in=20 Brazilian Amazonia. Forest Ecology and Management, 108: 147-166.=20 [ Links ]

Fearnside, P. M.;=20 Barbosa, R. I. 2003. Avoided deforestation in Amazonia as a global = warming=20 mitigation measure: The case of Mato Grosso. World Resource = Review, 15:=20 352-361. [ Links ]

Fearnside, P. M.;=20 Gra=E7a, P. M. L. A. 2006. BR-319: Brazil's Manaus-Porto Velho Highway = and the=20 potential impact of linking the arc of deforestation to central = Amazonia.=20 Environmental Management, 38: 705-716.=20 [ Links ]

Fearnside, P. M.;=20 Gra=E7a, P. M. L. A.; Keizer, E. W. H.; Maldonado, F. D.; Barbosa, R. = I.;=20 Nogueira, E. M. 2009. Modelagem de desmatamento e emiss=F5es de gases de = efeito=20 estufa na regi=E3o sob influ=EAncia da Rodovia Manaus-Porto Velho = (BR-319).=20 Revista Brasileira de Meteorologia, 24(2): 208-233.=20 [ Links ]

Fearnside, P. M.;=20 Lashof, D. A.; Moura-Costa, P. 2000. Accounting for time in mitigating = global=20 warming through land-use change and forestry. Mitigation and = Adaptation=20 Strategies for Global Change, 5: 239-270.=20 [ Links ]

Folha de S=E3o=20 Paulo. 2007. "Conclus=E3o de texto involve debate intenso", 6 April = 2007, p.=20 A-14. [ Links ]

Hare, = B.;=20 Meinshausen, M. 2006. How much warming are we committed to and how much = can be=20 avoided? Climatic Change, 75: 111-149.=20 [ Links ]

Hegerl, = G. C.;=20 Crowley, T. J.; Hyde, W. T.; Frame, D. J. 2006. Climate sensitivity = constrained=20 by temperature reconstructions over the past seven centuries. = Nature,=20 440: 1029-1032. = [ Links ]

Huntingford, C.;=20 Harris, P. O.; Gedney, N.; Cox, P. M.; Betts, R. A.; Marengo, J. A.; = Gash, J. H.=20 C. 2004. Using a GCM analogue model to investigate the potential for = Amazonian=20 forest dieback. Theoretical and Applied Climatology, 78: 177-185. = [ Links ]

Hutyra, = L. R.;=20 Munger, J. W.; Nobre, C. A.; Saleska, S. R.; Vieira, S. A.; Wofsy, S. C. = 2005.=20 Climatic variability and vegetation vulnerability in Amazonia. = Geophysical=20 Research Letters, 32: L24712, doi:10.1029/ 2005GL024981.=20 [ Links ]

IPCC=20 (Intergovernmental Panel on Climate Change). 2007a. Summary for = policymakers.=20 In: Parry, M. L.; Canziani, O. F.; Palutikof, J. P.; van der = Linden, P.=20 J.; Hanson, C. E. (Eds.). Climate Change 2007: Impacts, Adaptation = and=20 Vulnerability. Contribution of Working Group II to the Fourth Assessment = Report=20 of the Intergovernmental Panel on Climate Change, Cambridge = University=20 Press, Cambridge, U. K., pp. 7-22.=20 [ Links ]

IPCC=20 (Intergovernmental Panel on Climate Change). 2007b. Summary for = policymakers.=20 In: Solomon, S.; Qin, D.; Manning, M.; Chen, Z.; Marquis, M.; Averyt, K. = B.;=20 Tignor, M. Miller, H. L. (Eds.). Climate Change 2007: The Physical = Science=20 Basis. Contribution of Working Group I to the Fourth Assessment Report = of the=20 Intergovernmental Panel on Climate Change, Cambridge University = Press,=20 Cambridge, U. K., pp 1-18.=20 [ Links ] =

Kundzewicz, Z. W.;=20 Mata, L. J.; Arnell, N. W.; D=F6ll, P.; Kabat, P.; Jim=E9nez, B.; = Miller, K. A;.=20 Oki, T.; Sen, Z.; Shiklomanov, I. A. 2007: Freshwater resources and = their=20 management. In: Parry, M. L.; Canziani, O. F.; Palutikof, J. P.; = van der=20 Linden, P. J.; Hanson, C. E. (Eds.). Climate Change 2007: Impacts, = Adaptation=20 and Vulnerability. Contribution of Working Group II to the Fourth = Assessment=20 Report of the Intergovernmental Panel on Climate Change, Cambridge=20 University Press, Cambridge, U. K., pp. 173-210.=20 [ Links ]

Laurance, W. F.;=20 Nascimento, H. E. M.; Laurance, S. G.; Andrade, A.; Ribeiro, J. E. L. = S.;=20 Giraldo, J. P.; Lovejoy, T. E.; Condit, R.; Chave, J.; Harms, K. E.; = D'Angelo,=20 S. 2006. Rapid decay of tree-community composition in Amazonian forest=20 fragments. Proceedings of the National Academy of Sciences, = 103(50):=20 19010-19014. [ Links ]

Magrin, = G.; Gay=20 Garc=EDa, C.; Cruz Choque, D.; Gim=E9nez, J. C.; Moreno, A. R.; Nagy, G. = J.; Nobre,=20 C.; Villamizar, A. 2007. Latin America. In: Parry, M.L.; = Canziani, O. F.;=20 Palutikof, J. P.; van der Linden, P. J.; Hanson, C. E. (Eds.) Climate = Change=20 2007: Impacts, Adaptation and Vulnerability. Contribution of Working = Group II to=20 the Fourth Assessment Report of the Intergovernmental Panel on Climate=20 Change, Cambridge University Press, Cambridge, U.K., pp. 581-615.=20 [ Links ]

Marengo, J. A.;.=20 Nobre, C. A.; Tomasella, J.;. Oyama, M. D.; Sampaio de Oliveira, G.; de=20 Oliveira, R.; Camargo, H.; Alves, L. M.; Brown, I. F.. 2008. The drought = of=20 Amazonia in 2005. Journal of Climate, 21: 495-516.=20 [ Links ]

Meehl, = G. A.;=20 Stocker, T F.; Collins, W. D.; Friedlingstein, P.; Gaye, A. T.; Gregory, = J M.;=20 Kitoh, A.; Knutti, R.; Murphy J M.; Noda, A.; Raper, S.C B.; Watterson, = I. G. ;=20 Weaver, A. J.; Zhao, Z-C. 2007. Global Climate Projections. In: Solomon, = S.;=20 Qin, D.; Manning, M.; Chen, Z.; Marquis, M.; Averyt, K. B.; Tignor, M.; = Miller,=20 H. L. (Eds.). Climate Change 2007: The Physical Science Basis. = Contribution=20 of Working Group I to the Fourth Assessment Report of the = Intergovernmental=20 Panel on Climate Change, Cambridge University Press, Cambridge, U. = K., pp.=20 247-845. [ Links ]

Meinshausen, M.;=20 Meinshausen, N.; Hare, W.; Raper, S. C. B.; Frieler, K.; Knutti, R.; = Frame, D.=20 J.; Allen, M. R. 2009. Greenhouse-gas emission targets for limiting = global=20 warming to 2=BAC. Nature, 458: 1158-1162.=20 [ Links ]

Nascimento, H. E.=20 M.; Laurance, W. F. 2004. Biomass dynamics in Amazonian forest = fragments.=20 Ecological Applications, 14(4) Supplement: S127-S138.=20 [ Links ]

Nepstad, D. C.;=20 Alencar, A; Nobre, C.; Lima, E; Lefebvre, P.; Schlesinger, P.; Potter, = C.;=20 Moutinho, P.; Mendoza, E.; Cochrane, M.; Brooks, V. 1999. Large-scale=20 impoverishment of Amazonian forests by logging and fire. Nature, = 398:=20 505-508. [ Links ]

Nepstad, D. C.;=20 Carvalho, G.; Barros, A. C.; Alencar, A.; Capobianco, J. P.; Bishop, J.; = Moutinho, P.; Lefebvre, P.; Silva, Jr., U. L.; Prins, E. 2001. Road = paving, fire=20 regime feedbacks, and the future of Amazon forests. Forest Ecology = and=20 Management, 154: 395-407.=20 [ Links ]

Nepstad, D. C.;=20 Tohver, I. M.; Ray, D.; Moutinho, P.; Cardinot, G. 2007. Mortality of = large=20 trees and lianas following experimental drought in an Amazon forest.=20 Ecology, 88: 2259-2269.=20 [ Links ]

Phillips, O. L. and=20 65 others. 2009. Drought sensitivity of the Amazon rainforest. = Science,=20 323: 1344-1347. = [ Links ]

Salazar, L. F.;=20 Nobre, C. A.; Oyama, M.D. 2007. Climate change consequences on the biome = distribution in tropical South America. Geophysical Research = Letters, 34:=20 L09708, doi:10.1029/2007GL029695.=20 [ Links ]

Sampaio, G.; Nobre,=20 C.; Costa, M. H.; Satyamurty, P.; Soares-Filho, B. S.; Cardoso, M. 2007. = Regional climate change over eastern Amazonia caused by pasture and = soybean=20 cropland expansion. Geophysical Research Letters, 34: L17709,=20 doi:10.1029/2007GL030612.=20 [ Links ]

Soares-Filho B. S.;=20 Nepstad, D. C.; Curran, L. M.; Cerqueira, G. C.; Garcia, R. A.; Ramos, = C. A.;=20 Voll, E.; McDonald, A.; Lefebvre, P.; Schlesinger, P. 2006. Modelling=20 conservation in the Amazon Basin. Nature, 440: 520-523.=20 [ Links ]

Trenberth, K. E.;=20 Shea, D. J. 2006. Atlantic hurricanes and natural variability in 2005.=20 Geophysical Research Letters, 33: L12704, = doi:10.1029/2006GL026894.=20 [ Links ]

UN-FCCC = (United=20 Nations Framework Convention on Climate Change). 1992. United Nations = Framework=20 Convention on Climate Change. (Available in English at: http://www.unfccc.de/ = and in=20 Portuguese at: http://www.mct.gov.br/).=20 [ Links ]

Vasconcelos, S. S.;=20 Brown, I. F. 2007. The use of hot pixels as an indicator of fires in the = MAP=20 region: Tendencies in recent years in Acre, Brazil. In: = Epiphanio, J. C.=20 N.; Galv=E3o, L. S.; Fonseca, L. M. G. (Eds.) Anais XIII Simp=F3sio = Brasileiro de=20 Sensoriamento Remoto, Florian=F3polis, Brasil 21-26 abril 2007. = Instituto=20 Nacional de Pesquisas Espaciais (INPE), S=E3o Jos=E9 dos Campos, S=E3o = Paulo, Brazil.=20 pp. 4549-4556. [ Links ]

Recebido em=20 26/11/2007
Aceito em 16/06/2009

All the = content of=20 the journal, except where otherwise noted, is licensed under a Creative=20 Commons License

Instituto Nacional de Pesquisas da = Amaz=F4nia

Av.=20 Andr=E9 Araujo, 2936 Aleixo
CEP 69011-970 Manaus AM Brasil
Caixa = Postal=20 478
Tel.: +55 92 642-3438
Fax: +55 92 643-3223

acta@inpa.gov.br