Interactions among Amazon land use, forests and climate: prospects for a near-term forest tipping pointIntroductionThe growing profitability of deforestation-dependent land usesForest degradation through drought, fire, fragmentation and loggingThe drying of the AmazonInteractions among ecosystems, economics and climate: an Amazon forest tipping pointAvoiding the Amazon tipping pointConclusionWe gratefully acknowledge the contributions of Paul Lefebvre, Paulo Brando, Ane Alencar, Wendy Kingerlee and two anonymous reviewers. Funding was provided by the Worldwide Fund for Nature (WWF), the US National Science Foundation, the Large-Scale Biosp...ReferencesInteractions among Amazon land use, forestsand climate: prospects for a near-termforest tipping pointDaniel C. Nepstad1,2,*, Claudia M. Stickler1,2,3, Britaldo Soares-Filho2,4and Frank Merry1,21Woods Hole Research Center, 149 Woods Hole Road, Falmouth, MA 02540, USA2Instituto de Pesquisa Ambiental da Amazoˆnia, Avenida Nazare´, 669, CEP: 66035-170,Bele´m, Para´, Brazil3School of Natural Resources and Environment, University of Florida, Gainesville, FL 32611-6455, USA4Centro de Sensoriamento Remoto, Universidade Federal de Minas Gerais,CEP: 31270-901, Belo Horizonte, Minas Gerais, BrazilSome model experiments predict a large-scale substitution of Amazon forest by savannah-likevegetation by the end of the twenty-first century. Expanding global demands for biofuels and grains,positive feedbacks in the Amazon forest fire regime and drought may drive a faster process of forestdegradation that could lead to a near-term forest dieback. Rising worldwide demands for biofuel andmeat are creating powerful new incentives for agro-industrial expansion into Amazon forest regions.Forest fires, drought and logging increase susceptibility to further burning while deforestation andsmoke can inhibit rainfall, exacerbating fire risk. If sea surface temperature anomalies (such asEl Nin˜o episodes) and associated Amazon droughts of the last decade continue into the future,approximately 55% of the forests of the Amazon will be cleared, logged, damaged by drought orburned over the next 20 years, emitting 15–26 Pg of carbon to the atmosphere. Several importanttrends could prevent a near-term dieback. As fire-sensitive investments accumulate in the landscape,property holders use less fire and invest more in fire control. Commodity markets are demandinghigher environmental performance from farmers and cattle ranchers. Protected areas have beenestablished in the pathway of expanding agricultural frontiers. Finally, emerging carbon marketincentives for reductions in deforestation could support these trends.Keywords: fire; deforestation; biofuel; feedbacks; globalization; global warming1. INTRODUCTIONThe future of the Amazon Basin is a topic of greatconcern worldwide. The trees of the Amazon contain90–140 billion tons of carbon (Soares-Filho et al.2006), equivalent to approximately 9–14 decades ofcurrent global, annual, human-induced carbon emis-sions (Canadell et al. 2007). The prospect of reducingglobal warming and keeping global average tempera-tures from rising no more than 28C will be very difficultif emissions of carbon from tropical forests worldwide,and the Amazon in particular, are not curtailed sharplyin the coming years (Gullison et al. 2007). Beyond itsrole as a giant, somewhat leaky reservoir of carbon, theAmazon is home to one out of every five mammal, fish,bird and tree species in the world. Less recognized,perhaps, is the role of the Amazon in the global energyand water balance. Approximately eight trillion tons ofwater evaporate from Amazon forests each year, withimportant influences on global atmospheric circulation(IPCC 2007). The remainder of the rainfall enteringthis enormous basin flows into the Atlantic Ocean—15–20% of the worldwide continental freshwaterrun-off to the oceans.These ecological services may be threatened byglobal warming through a late-century, climate-drivensubstitution of forests by savannah and semi-aridvegetation in what has been called the Amazon forest‘die back’ ( Nobre et al. 1991; Cox et al. 2000, 2004;Botta & Foley 2002; Oyama & Nobre 2003). However,these climate–vegetation simulations do not includeland-use change, or the synergistic effects of land-usechange and near-term regional climate change on theAmazon fire regime. Could accelerating, forest-substituting and forest-damaging economic activitiesinteract with regional climate change to replace ordegrade a large portion of the Amazon forest over thenext two decades? And what counteracting trendscould prevent such a dieback? These questions are thefocus of this paper. We review current trends inAmazon economic, ecological and climatic processes,the growing evidence of positive feedbacks among theseprocesses and the potential for these interactions topush Amazon forests towards a tipping point. Weconclude with a brief review of some of the processesPhil. Trans. R. Soc. B (2008) 363, 1737–1746doi:10.1098/rstb.2007.0036Published online 11 February 2008One contribution of 27 to a Theme Issue ‘Climate change and thefate of the Amazon’.* Author and address for correspondence: Woods Hole ResearchCenter, 149 Wood Hole Road, Falmouth, MA 02540, USA([email protected]).1737 This journal is q 2008 The Royal Societythat could contribute to a strategy to reduce thechances of a major Amazon forest dieback. Thissynthesis draws on the published literature, someunpublished data, and presents two new maps ofthe Amazon.2. THE GROWING PROFITABILITY OFDEFORESTATION-DEPENDENT LAND USESThere are several trends underway that point togrowing, sustained economic pressure to convertAmazon forests to crop fields and cattle pasture.First, large areas of southern and eastern Amazoniahave eradicated foot-and-mouth disease, openingmuch of the region’s cattle and swine industries toexport out of the Amazon, often for higher prices(Kaimowitz et al. 2004; Arima et al. 2005; Nepstadet al. 2006a). Second, the rising international demandfor agro-industrial commodities, discussed below, iscolliding with the scarcity of appropriate land for agro-industrial expansion in the USA, Western Europe,China and many other agricultural countries (Brown2004). As a result, much of the recent surge in globalcropland area expansion is taking place in the BrazilianCerrado and Amazon regions (Brown 2004; Shean2004; Nepstad et al. 2006a). Third, the rising price ofoil has triggered new