Holocene fire and occupation in amazonia records from two lake districts


Published on

Published in: Technology
  • Be the first to comment

  • Be the first to like this

No Downloads
Total views
On SlideShare
From Embeds
Number of Embeds
Embeds 0
No embeds

No notes for slide

Holocene fire and occupation in amazonia records from two lake districts

  1. 1. Holocene Fire and Occupation in Amazonia: Records from Two Lake DistrictsAuthor(s): Mark B. Bush, Miles R. Silman, Mauro B. de Toledo, Claudia Listopad, William D.Gosling, Christopher Williams, Paulo E. de Oliveira, Carolyn KriselSource: Philosophical Transactions: Biological Sciences, Vol. 362, No. 1478, BiodiversityHotspots through Time: Using the Past to Manage the Future (Feb. 28, 2007), pp. 209-218Published by: The Royal SocietyStable URL: http://www.jstor.org/stable/20209833 .Accessed: 22/03/2011 13:08Your use of the JSTOR archive indicates your acceptance of JSTORs Terms and Conditions of Use, available at .http://www.jstor.org/page/info/about/policies/terms.jsp. JSTORs Terms and Conditions of Use provides, in part, that unlessyou have obtained prior permission, you may not download an entire issue of a journal or multiple copies of articles, and youmay use content in the JSTOR archive only for your personal, non-commercial use.Please contact the publisher regarding any further use of this work. Publisher contact information may be obtained at .http://www.jstor.org/action/showPublisher?publisherCode=rsl. .Each copy of any part of a JSTOR transmission must contain the same copyright notice that appears on the screen or printedpage of such transmission.JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range ofcontent in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new formsof scholarship. For more information about JSTOR, please contact support@jstor.org.The Royal Society is collaborating with JSTOR to digitize, preserve and extend access to PhilosophicalTransactions: Biological Sciences.http://www.jstor.org
  2. 2. PHILOSOPHICALTRANSACTIONSpM ^^ R^ ? (200?)^^ 209_21gTHEROYAL[4ndoi:10.1098/rstb.2006.1980SOCIETY jSJ) Published online9 January2007Holocene fire and occupation inAmazonia:records from two lake districtsMark B. Bush1*, Miles R. Suman2, Mauro B. de Toledo14,Claudia Listopad1, William D. Gosling1*, Christopher Williams1,Paulo E. de Oliveira and Carolyn Krisel1Department ofBiological Sciences, Florida Institute ofTechnology,Melbourne, FL 32901, USA2Department ofBiology, Wake Forest University,Winston Salem, NC 27106, USA3Department ofBiological Sciences, Universidade do Guarulhos, Sao Paulo 07023-070, BrazilMarine Geology and Geophysics Department, Universidade Federal Fluminense,Niter?i RJ 24.210-340, BrazilWhile large-scale pre-Columbian human occupation and ecological disturbance have beendemonstrated close tomajor Amazonian waterways, less is known of sites in terra firme settings.Palaeoecological analyses of two lake districts in central and western Amazonia reveal long histories ofoccupation and land use. At both locations, human activitywas centred on one of the lakes,while theothers were either lightlyused or unused. These analyses indicate that the scale of human impacts inthese terrafirme settings is localized and probably strongly influenced by thepresence of a permanentopen-water body. Evidence is found of forest clearance and cultivation ofmaize and manioc. Thesedata are directly relevant to the resilience of Amazonian conservation, as they do not support thecontention that all ofAmazonia is a built landscape and therefore a product of past human land use.Keywords: agriculture; charcoal; fossil pollen; Peru; Brazil; pre-Columbian1. INTRODUCTIONThe role of pre-Columbian human activity in shapingAmazonian ecosystems has received considerableattention in the light of occupational records from theXingu River basin, Marajos Island, and near themodern cities of Santar?m and Manaus (Roosevelt1991; Roosevelt etal 1991; Heckenberger etal. 1999).Advocates ofwidespread human influence inAmazoniapoint to what appear to be urban centres (Roosevelt1991; Roosevelt etal 1991; Heckenberger et al. 1999,2003), large earthworks inBolivian savannahs borderingAmazonia (Mann 2000; Erickson 2001), and thewidespread occurrence of soils enriched with carbon(terra pr?ta) within Amazonia (Glaser etal. 2001; Limaet al. 2002). IfAmazonia was a managed landscapeprior to European contact, the biodiversity of theregion has withstood substantial fragmentation, hunting and human interaction, and such manifest resilience would influence conservation decision making.While Erickson (2001) has described the Boliviansavannah landscape as having been manufactured byhumans, other authors are generally more circumspectwhen dealing with the densely forested regions.Major uncertainties exist in the size of the preColumbian human population of Amazonia, withestimates ranging from 1 to 11 million inhabitants.Similarly, the large areal extents of some archaeologicalsettlements, e.g. 5 km2 of terra pr?ta underlying themodern city of Santar?m, have been variously interpreted as representing continuous occupation by alarge population (Roosevelt 1987) or repeated resettlement and abandonment by smaller groups (Meggers1995). Heckenberger et al (2003) have argued fordense settlement and societies with strong socialhierarchies with the capability to transform landscapes.The alternate view is that human endeavours inAmazonia were constrained by poor soils and thatoccupation was sparse, lacking the societal complexityormonuments ofAndean or Central American systems(Meggers 1954, 2003a,?). Denevan (1996) suggestedthat occupation was centred on sandy bluffs overlooking navigable channels and floodplains. Thisinterpretation of existing data suggests local alterationof landscapes extending 5-10 km around settlements,but with vast interfluvial areas where humans had littleimpact on the biota (Denevan 1996).From an ecological standpoint, humans are likely todeplete animal populations and transform landscapesthrough burning. While it isvery difficult to documentthe past population size of people, let alone animals,evidence of burning is well preserved in palaeoecological records. Indeed, we suggest that fire is thecharacterizing disturbance of human activity inmuchofwestern Amazonia and we will take itspresence to bean indicator of human activity.(a) Fire in Amazonian landscapesHuman settlers modified the landscape through fire.Some parts of Amazonia are more susceptible to fire*Author for correspondence (mbush@fit.edu)."Present address: Department of Earth Sciences, CEPSAR, TheOpen University, Walton Hall, Milton Keynes MK7 6AA, UK.One contribution of 14 to a Theme Issue Biodiversity hotspotsthrough time: using the past tomanage the future.209 This journal is? 2007 The Royal Society
  3. 3. 210 M. B. Bush et al. Holocene fire and occupation inAmazoniathan others (Nepstad et al 2004) and considerableuncertainty exists over the natural occurrence ofAmazonian forest fires. Some researchers have foundevidence of natural fire (e.g. Sanford et al 1985;Saldarriaga & West 1986; Piperno & Becker 1996),whereas other data point to fire as very rare in theabsence of humans (e.g. Hammond & terSteege 1998;Turcq et al 1998; Behling & Da Costa 2000; Behling2001; Behling etal 2001). In a 50 000-year palaeoecological record from theHill of Six Lakes innorthwesternAmazonia, only one sedimentary layer at 5600 calendaryears before the present (yrBP) was found to containsignificant amounts of charcoal, suggesting the rarityoffire in that landscape (Bush etal 2004).Piperno & Becker (1986) recorded charcoal in soilsaround Manaus with a concentration of ages between ca900 and 1260 yrBP, and attributed these fires tonaturalclimatic change rather than agriculture. From a nearbysite, Santos et al (2000) also recovered charcoal fromsoils with peaks between 1000 and 1300 yr BP. Theperiod ca 1100-1300 yrBP was a phase of intensifiedElNi?o activity (Moy etal 2002), making itplausible thatextreme drought caused fire at this time.However, somecaution isneeded in interpreting these data. Rather thanslash and burn agriculture, less radical landmanagementinvolving firemay have prevailed, e.g. that practised forBrazil nut harvesting in Peru (Phillips 1993) andCaryocar edule fruitsby indigenous peoples in theXingu(P. E. de Oliveira 2005, personal observation).Nepstad et al (2004) document that the forests ofcentral Amazonia are only susceptible to natural fireduring the late stages of the dry season. Under modernconditions in strongly seasonal forests, there is about a10-day window each year when natural fire ispossible.Near Manaus, the probability of fire decreases exponentially with distance from sites of human occupation(Laurance & Williamson 2001). In Peruvian Amazonia, indigenous people find itdifficult to burn the forestat any time.Evidence of the rarity of fire inwestern Amazoniacomes from historical and palaeoecological observations. Cocha Cashu inManu National Park lies lessthan 200 km from the western Lake District and hasbeen continuously monitored for the past 40 years. Inthat time, no fire has been documented within theapproximately 50 000 km2 reserve. The only charcoalrecovered from the Cocha Cashu lake sediments werefragments derived fromwood burned since 1968 by thelake-shore field station (M. R. Silman, unpublishedcore data). Sediment cores raised from three oxbowlakes along theMadre de Dios, all ofwhich span thepast 220-700 years contain no charcoal (Listopad2001). Similarly, the ca 3300-year record from LakeWerth (below) contains no evidence of local forest fires.We conclude from these observations that undermodern conditions, fire is not a natural phenomenonwithin this forest area.Three factors combine to provide the probabilitythat a given site is susceptible to fire: the amount ofprecipitation; the strength of the dry season; and theamount of human activity. Once an area has burned,however, successive burning is promoted by increaseddead biomass, reduced canopy cover and a drier litterlayer (Nepstad etal 2004). Such fires are documentedin lake records in the absolute concentration ofcharcoal in lake sediment. However, as lake size andtaphonomic processes influence rates of charcoaldeposition (Whitlock et al 1997; Whitlock & Larsen2001), comparisons of concentration to infer intensityof land use aremost robust within a lake record, ratherthan between lake records.Although landscape alteration is clearly going tohave occurred around known archaeological sites, theextent towhich those sites are representative of otherportions ofAmazonia is unknown. Consequently, thespatial and temporal scale of landscape alteration inpreviously undocumented settings may provide analternative image of Amazonian occupation. Wepresent data from two Amazonian lake districts thatdocument human occupation, and use the charcoal andpollen records from those sites to gauge the spatial andtemporal extent of landscape alteration at thoselocations. In particular, we ask did human occupationresult inwidespread landscape modification at thesetwo widely separated Amazonian sites? or, ashypothesized by Denevan (1996), is human disturbance centred on particular features of the landscape?2. MATERIAL AND METHODS(a) Study areasA paucity of long Holocene palaeoecological records exists inAmazonia, but here we report on two suites of cores raised toinvestigate climate change in ecologically sensitive areas ofAmazonia (Listopad 2001; M. B. de Toledo 2004, unpublished Ph.D. thesis). Sites were selected to provide lakes thatwere not directly influenced by rivers, lay within terra firmeforest and where replication could be achieved within a smallgeographical area. A western lake district (hereafter westernlakes) near Puerto Maldonado, Peru, that comprised lakesGentry, Vargas, Parker and Werth, and an eastern lake district(hereafter eastern lakes) near Prainha, Brazil, that comprisedlakes Geral, Santa Maria and Saracuri, were selected for study.The western lakes lie within a lake district containingapproximately 50 lakes and swampy depressions (approx.12.5? S, 69.0?W; figure la). The lake district lies atapproximately 230 m elevation and all of the lakes arerelatively small, approximately from 0.3 to 2 km in diameterand are approximately 1-3 m deep. The origin of the lakes isunknown, but they lie at the highest points on the rollinglandscape, outside fluvial influence, and are underlain bysands and clays.The lakes lack inflowing streams and their permanence isindicated by the presence of fishes. The lakes appear to bemesotrophic and are fringed by beds of Sagittaria, Pontederiaand Eichornia. The forests in the region are upper-Amazonianrainforest and have been extensively documented, e.g. Pitmanet al. (2001) and are rich in Calycophyllum,Calophyllum,Cecropia, Cedrela, Ceiba, Dipteryx, Ficus (35 species) andPoulsenia (Foster 1990). The palms Iriartea deltoidea andAstrocaryum murumuru are the commonest stems in the forest(Pitman et al 2001). Of note is the local abundance ofLecythidaceae (particularly Bertholletia excelsa and Couratarispp.) in the forests. Temperatures are relatively constant witha mean of 25?C, but the lake district lies across a region ofsharply changing precipitation ranging from approximately2000 mm yr_lin the south to approximately 1700 mmyr~lin the north. Similarly, there is some variability in the dryseason which ranges from two to four months in duration.A modern road passes closest to Lake Gentry, and on a knollPhil. Trans. R. Soc. B (2007)
  4. 4. Holocene fire and occupation inAmazonia M. B. Bush etal 211beside a stream within 1 km of Gentry is a modernfarmhouse. The occupant of the farm pointed out that theknoll was rich in stone tools and pottery, and he showed us hiscollection of axe heads and other stone tools that he hadfound close to his house. No formal description wasattempted for this site; we report it to emphasize some prioruse and occupation of the site. No archaeological evidence ofoccupation was noted for the other lakes.The eastern lakes lie between 20 and 30 m in elevation on anorth-south transect perpendicular to the Amazon River(figure lb). The three lakes occupy old river valleys andprobably filled as sea-level rose in the early Holocene. Theseelongate lakes range in size from approximately 400 X100-3000 X 1000 m. Mauritia and Mauritiella were commonon shorelines and in swamps. Mean annual temperature is 27?Cand rainfall is approximately 2200 mm per annum over an eightand nine month wet season (Instituto Brasiliero de Geograf?a eEstat?stica 1990). Though thisclimate could support tropicalsemideciduous forest, the local vegetation is a savannah-forestmosaic. The origin of the savannah isunknown and could be theresult of edaphic factors or a long history of human landscapealteration (Prance& Schubart 1977).Saracuri is completely enclosed by forest, with just somerecent deforestation on the northern shoreline. Parts of theSanta Maria catchment are recently deforested, but a patch oftall dry forest lies within its catchment. Geral is an elongatedlake running approximately east-west. Along its northernshore is forest, with savannah along its southern shore. Today,Geral is extensively used for leisure and fishing, receivingvisitors from Prainha (28 km away). The other lakes are usedby cattle and for fishing by the owners, but show no otherevidence of human use. The fossil pollen and phytolithhistory ofGeral have previously been described (Piperno &Pearsall 1998; Bush etal. 2000).(b) Field and laboratory techniquesBetween 1990 and 2001, cores were raised from the centre ofeach lake using a Colinvaux-Vohnout coring rig from a raft ofrubber boats. The core from Geral was initially analysed at theSmithsonian Tropical Research Institute, Panama, and later,like the others, at the Florida Institute of Technology. Pollenanalysisfollowedstandardprotocols (Stockmarr 1971;Faegri&Iversen 1989), and is reported inBush etal. (2000; inpress)andM. B. de Toledo (2004, unpublished Ph.D. thesis).AllPoaceae pollen grains identified as maize had a distinctivesurface pattern and a diameter greater than 90 urn (note thatthis is a more exacting definition than often used). The pollensum is based on a minimum of 300 terrestrial pollen, with theexception of Geral, which is based on 200 terrestrial pollen.Aquatic taxa (e.g. Cyperaceae, Ludwigia, Alismataceae andPontederiaceae) were excluded from the pollen sum, but areexpressed as a percentage of the terrestrial pollen.Charcoal samples were disaggregated in 10% KOH andsieved with a 180 urn screen. Particles retained on the screenwere recorded digitally and their area calculated throughvideo-capture and analysis using National Institutes ofHealth-image (Clark & Hussey 1996; Clark & Patterson1997). Diagrams were plotted using C2 (Steve Juggins,University of Newcastle).3. RESULTS AND DISCUSSION(a) ChronologyWith the exception ofGeral, the chronologies of all thelake records are supported by at least five 14Caccelerator mass spectrometry (AMS) dates (seetable 1). The Geral chronology is based on two bulkcarbon dates. All dates were calibrated using Calib 5.1(Stuiver & Reimer 1993) using the Southern Hemisphere correction and are reported as yrBP.(b) Western lakesClimatic fluctuations of theHolocene are evident inthese records. Although three of the lakes (Vargas,Parker and Gentry) formed between 8000 and 6000 yrBP, this was not a uniformly wet period. Consistentwith other regional data (Bush et al in press), a midHolocene dry event influenced the region, but wasinterrupted by wet episodes, presumably when thelakes filled (Paduano et al 2003). The lakes showdifferent levels of sensitivity to drought, with Vargasmost sensitive, then Parker, and Gentry the leastsensitive. However, even Gentry was susceptible todrying and prior to ca 4800 yrBP exhibited at least twobrief phases when pollen was oxidized from thesediments. Another period of apparently slow sedimentation between ca 1270 and 2800 yrBP may pointto intermittent accumulation, although high pollenconcentrations suggest that thismay simply have been atime of little sediment input; otherwise the Gentryrecord appears to have a continuous history of pollenaccrual. The last lake to formwas Werth, which startedto accumulate sediment ca 3400 yrBP, and deepenedat ca 900 yrBP.Once the lakes had formed fully, theyprovide pollenrecords with greater than 70% arboreal pollen throughout their history (figure 2). Only Gentry showedevidence of agriculture, with Zea pollen found between3700 and 500 yr BP and two grains ofManihot pollenat ca 2400 yr BP. Charcoal was most abundant inGentry, ranging between 1 and 34 mm2 cm-3 throughout much of the record. Notably, charcoal wasabsent during the second of the oxidizing phases at ca5000 yrBP.Lake Parker also contained charcoal through muchof itshistory, including the time immediately prior tothe formation of Gentry. Charcoal concentrations inParker were generally about half those of Gentry,between approximately 0.5 and 17mm2 cm-3(figure 2). Lake Vargas contains no charcoal formostof its history, except a strong spike of charcoal withapproximately 17mm2 cm-3 at ca 1000 yr BP. TheLake Werth record contains no charcoal except forthree tiny fragments at ca 800 yrBP.(c) Eastern lakesThe eastern lakes, Saracuri, Santa Maria and Geral allformed between 8200 and 8400 yrBP. The previouslyreported history forGeral (2000; Piperno & Pearsall1998) is substantiated and supplemented by the newanalyses.The small basin of Santa Maria provides a sensitiveproxy for lake level change. High modern aquatic pollenabundances are due to a 40 m wide floating mat ofmarshland species. Poaceae pollen rises and fallssynchronously with the aquatics, indicating that at thissitePoaceae pollen are probably derived from themarshfringe.Notably, Poaceae percentages range as high as85% at Santa Maria, whereas they seldom rise above 12%at the other two lakes. Saracuri provides a remarkablystable pollen record throughout itshistory (figure 3) andPhil Trans. R. Soc. B (2007)
  5. 5. 212 M. B. Bush et al Holocene fire and occupation inAmazoniaFigure 1. Satellite images showing the location of the (a) Maldonado lakes, southeastern Peru and (?>) the Prainha lakes relativeto major rivers and South America. Forest is shown as darker areas, with brown areas being savannah or cerrado. Charcoal isexpressed as mm2 per cm3. Latitude and longitude of Gentry 12?1038.3l" S; 69?05/51.54//W and Geral 1?38;48.85" S;53?3543.9" W.contrasts with the post- 4000 yrBP portion of theGeralrecord, lacking high proportions of Poaceae, highcharcoal content and the pollen from known cultivars.The divergence of thePoaceae curvefromGeral with thatof Saracuri is consistent with the onset of agriculture andprobably reflectshuman activity around Geral.Geral is the only lake of this trio inwhich Zea mayspollen was found. Pollen and phytoliths ofZea are foundconsistently after ca 4030 yrBP and prior to ca 850 yrBP.A phase ofmore active forestburning is evident inboththe local charcoal reported here (figure3) and a doublingof the concentration of charred Heliconia phytoliths atthis level (Piperno & Pearsall 1998). Clearly, whilecultivation ofmaize may have taken place on the shorelineand in seasonally exposed shallows (Roosevelt 1980;Matheny & Gurr 1983), there was also clearance ofadjacent forest. Santa Maria and Saracuri both containedcharcoal, including samples predating the first evidenceof charcoal atGeral but at lower concentrations than inGeral. Within the period of known crop cultivation,charcoal concentrations at both Santa Maria andSaracuri fall from theirmid-Holocene highs, whereasthose ofGeral show a much lesser decline. Indeed, theearlyHolocene peaks of concentration are slightlyhigherat Santa Maria (as befits a small basin) than atGeral, butfall to about one-fifth those ofGeral post-4000 yr BP,suggesting much lower firefrequency than atGeral.(d) The charcoal recordA regional drought influenced western and southernAmazonia between ca 6200 and 3400 yr BP (Servantetal 1981; Mayle etal 2000; Listopad 2001; Mayle &Beerling 2004), though its occurrence in centralAmazonia ismuch less clearly defined (Bush et al2000; Behling et al 2001; Listopad 2001; M. B. deToledo 2004, unpublished Ph.D. thesis). During thisperiod, fire frequency in the western lakes does notappear to be markedly different to other periods of theHolocene even though drought causes lake level to fall,exposing the mud-water interface. While organicmaterial fails to accumulate and pollen is oxidized,charcoal continues to accumulate as it oxidizes veryslowly. In none of the lakes is there evidence of a charcoalspike associated with a time of lowpollen concentrations.Fires in these forests probably accompanied humanactivities and dry episodes caused site abandonment. AtGentry, the few samples that contained no pollen alsocontained no charcoal. Similarly, the record from Vargasoverlaps with thatofParker between ca 7500 and 7000 yrBP and shows no peak of charcoal associated with thehighest proportion ofPoaceae. We inferthatfirewas rareor absent from this setting in the absence of humandisturbance.Vargas was thefirst lake to form and appears tohavebeen briefly occupied and then abandoned. Theabandonment is approximately coincidental with thecolonization of Lake Parker and the first charcoalevident in that record. As Gentry filled, itwas quicklyoccupied and the fire intensity and frequency becamegreater than that of any of the other sites.In the eastern lakes, all three lakes contain somecharcoal, but the lake that yielded crop pollen had themost consistent occurrence of relatively high concentrations of charcoal. Although Geral shows a moreconsistent firehistory than the other two lakes, even theforested sites of Santa Maria and Saracuri were proneto fire, especially in the early Holocene. The highestcharcoal peaks in those records are between 6900 and7500 yrBP. The Geral record isnot aswell dated, but apeak of charcoal 6700 and 7200 yrBP is seen to overlapwith the fire peak at other sites.At Saracuri and Santa Maria, diminishing charcoalinputs indicate that fire becomes relatively rare ca6800 yr BP. Between 6200 and 4500 yr BP, anoscillation in the aquatic and Poaceae pollen signatures, and a spike of charcoal at Santa Maria, suggestPhil. Trans. R. Soc. B (2007)
  6. 6. Holocenefire and occupation inAmazonia M. B. Bush et al 213Table 1. Radiocarbon data and calibrated ages for lakes Vargas, Parker, Gentry, Werth, Geral, Saracuri and Santa Maria.lab number depth (cm)?13C%0 4C age (yr)calibratedage (yrBP)median probableage (yr)GentryNSRL-11997NSRL-11998NSRL-11999NSRL-12001NSRL-12000NSRL-12002ParkerOS-38415CAMS 109894OS-38416CAMS 109895OS-38417OS-35829VargasOS-38418OS-38419OS-39955OS-39954OS-35343WerthCAMS -74839CAMS -75227NSRL-11994CAMS -75228CAMS -74982NSRL-11994NSRL-11995NSRL-11996Geral?-41654?-39702Santa MariaOS 24122OS 24123OS 24124OS 24125OS 24126CURL-5386SaracuriOS-38383OS-38384OS-38385OS-38386OS-38387CURL-5385334349517710650110130167178216558088110166373750575790100140375-387542-55117224040752857080670263355462666859-22.35-25.06-23.48-24.03-24.33-24.50-25-25-22.61-25-26.07-27.11-27.2-29.86-28.74-28.53-24.43-25-27-14.94-27-25-25-18.93-23.19-28.9-28.2-25-25-25-25-25-25-25-25-25-25-25-25> modern940 ?402250?302610?504070?355440 + 40525?2528151353530135585013561401456410145> modern9451301390130630014570601605801701020150107013514701401850140320014513013512 75016557601907500110029601453450135466014061801406770145674014578512541301303780130439013559801457690175743-9022149-23062504-27494429-45226033-6279507-5282789-29183693-38276546-66586809-71447181-7413763-9021189-12987029-72547762-7932506-628803-930914-9641294-13451633-18133274-34420-25014 934-15 1756403-66308179-83692960-31413585-36905093-54486938-71567513-76207506-7592662-7164448-47843989-41464850-49606674-67928378-85360770227026204490623052028303750660069507290084012807200783054084094013201750337014015 1006540827030503640527070507560755069046104070491067308450the possibility of a somewhat drier ormore fire-pronesystem, although the lake does not dry out. At Geral,increased charcoal and Cecropia abundance characterize thisperiod, suggesting increased disturbance andgap formation in the local forests, probably attributableto human activity.At Saracuri, a very slight increase incharcoal abundance is coincident with a modestincrease inPoaceae pollen. Taken together, these datasuggest a period of increased human activity andpossibly a more flammable forest under drier conditions. These data are consistent with a rather weakmanifestation of the mid-Holocene dry event documented in western Amazonia and the High Andes(Bush etal 2005).The charcoal record of Geral iswholly consistentwith that of the phytolith record (Piperno & Pearsall1998). Heliconia is a genus generally associated withforest gaps and clearings. Most (70%) of theHeliconiaphytoliths recovered at ca 6500 yrBP were charred, aswere a substantial proportion of Poaceae and leafphytoliths from arboreal species. These data stronglysuggest that firewas used to clear forests near the lake.While this evidence clearly indicates human activity,the first occurrence of landscape alteration probablydates to at least 7700 yrBP, i.e. the first occurrence ofcharcoal in the record. One possibility is that land useintensified around 6000 years ago, coinciding with theincrease in Cecropia pollen and the nearlyPhil. Trans. R. Soc. B (2007)
  7. 7. 214 M. B. Bush et al. Holocene fire and occupation inAmazonia(a) Gentry0 40 80 0 0 40 80 0 80 0 20 40(b)Parker?>0m ,pq<1?6^ &0 40 80 0 0 0 40 0 20 40(c)VargasOicl, 2cdo4Tsedimentary hiatus0 40 80 0 0 40J*is~WF40 80 0 20(d)Werth40 80 0 0Figure 2. Fossil pollen summary percentage data andcharcoal for the Maldonado lakes. Summary groups are:arboreal (sum of more than 200 types), Poaceae, aquatics,crops (plus, Zea; open circle, Manihot) and charcoal, plottedagainst time. Asterisk, change of scale for Gentry aquatics(double other percentage scales); dashed line, trace presenceof charcoal (Werth). Charcoal is expressed as mm2 per cm3.contemporaneous occurrence of burned phytoliths andPoaceae pollen. However, it can also be argued thatprior to the formation of an open body ofwater at thecoring site, the deposition of microfossils may havereflected so local a source that land use beyond theimmediate vicinity of the sample site cannot bedetected. Hence, we cannot specify a time for theonset of disturbance atGeral.O40 80 040 0 4080040 800 20(c) SaracuriFigure 3. Fossil pollen summary percentage data andcharcoal for the Prainha lakes. Summary groups are: arboreal(sum ofmore than 150 types), Poaceae, aquatics, crops (plus,Zed) and charcoal, plotted against time.(e) The scale of impactPrior towhat may have been a single local fire event atVargas, the scale of occupation appears to be consistentwith modern anthropological observations that indigenous people manipulate relatively small territories. Theactivelymanaged area ismuch smaller than the huntingrange. Studies of the Secoya-Siona inEcuador revealedan occasional hunting range as large as 2500 km2, with590 km2 being heavily used (Vickers 1988). TheSecoya-Siona hunters live in the black-water systems ofthe Cuyabeno River, and hunting ranges in the fertilewhite-water systems of Peruvian Amazonia might besmaller. The Tsimane of northern Bolivia concentratetheir farming and hunting activitywithin 3 km of theirsettlement (Apaza et al. 2002). Carneiro (1970) hassuggested that a radius of approximately 5 km couldsupport a sedentary population of 500 through swiddenagriculture indefinitely.We have no way to estimatePhil Trans. R. Soc. B (2007)
  8. 8. Holocenefire and occupation inAmazonia M. B. Bush et al 215population size around Geral or Gentry. Ifwe assumethat these lakeswere the centre of occupation, itisevidentthat the strength of human influence on the landscapedecreased markedly with increasing distance from them.After several thousand years of apparent use, the firstpalynologically identifiable crop isfound inthe sedimentsof Gentry. Only Gentry provided direct evidence ofcultivation with pollen of Zea being found regularlybetween 3700 and 500 yrBP and manioc at ca 2400 yrBP. There isno clear spike incharcoal associated with theonset ofmaize cultivation; therefore, the field systemswere already cleared or the cultivation took place onexposed mud when lake levels were low.The onset ofmaize cultivation coincides with a period of very slowsedimentation at Gentry, consistent with intermittentsediment accumulation and lowered lake levels, offeringthe opportunity to cultivate exposed wetlands. Detectionofmaize pollen at this timemay not be a coincidence, assuch large pollen grains (greater than 90 um) are poorlydispersed, and cultivation closer to the sampling pointimproves the probability of recovering the fossil grain.Taking the charcoal and crop data together,we inferthathumans have occupied this region ofAmazonia formorethan 8000 years, and through the use of fire they havealtered the landscape at least at a local level. From ourdata, we cannot determine whether the occupationwasseasonal or permanent, though the abundance ofpotteryshards on the ridge beside Gentry suggests a degree ofpermanence.The sequential occupation of Vargas (and itsabandonment), Parker and Gentry suggests that inhabitantswere continually taking advantage of better homesites andmay have abandoned these sitesduring thepeakdrought event at ca 5000 yr BP. Ifwe are correct thatGentry was the centre of the occupation,we can comparethe relative impacts toParker, Vargas andWerth in termsof distance from the primary settlement.The distance between Gentry and Parker is 9.5 kmand between Gentry and Vargas is 17.5 km. Parkerappears tohave been used fairlyextensively, but perhapsless intensively thanGentry, as no evidence was found ofcrop cultivation. Only in the last millennium wereGentry, Parker and Vargas simultaneously occupied.This expansion is coincident with the major Xingulandscape transformation (Heckenberger et al. 2003),but we note that it did not extend to the more remotesetting of Lake Werth, 50 km fromGentry, which wasnever impacted by human occupation.A similar image emerges from the eastern sites. If it isassumed that the long history of agriculture at Geralindicates that this lake formed the centre of localoccupation, the distance to the adjacent lakes providesa scale of local impact. Saracuri at 5 km distance andSanta Maria at 6.7 km distance were both lightlyused.No evidence exists to suggest that sitesmore than 5 kmfrom the centre of occupation were heavily modified,though undoubtedly theywould have been the site ofhunting activity. Although the eastern sites liewithin150 km of the major pre-Columbian settlementsaround Santar?m, that pattern of large landscapetransformation isnot seen in these records.Our data are entirely consistent with the observationby Smith (1980) that the average size of terrafirme terrapr?ta deposits was approximately 1.4 ha, whereasriverside sites averaged 21.2 ha. This observation supports a history of patchy disturbance, with riversidesettings not being typical of settings remote from rivers.Smith (1980) specifically noted that sites near theconfluence of the Rio Negro and the Solimoes riversand on theXingu were particularly large. Consequently,we conclude that while disturbance was probablyprofound at a landscape level around major settlements,such localities should not be used to generalize about thestate of Amazonia as a whole.4. CONCLUSIONSWe offer the first landscape-scale view of disturbance attwo sites in Amazonia and see long histories ofdisturbance at what appears to be a central location,with relatively small impacts on nearby systems and noapparent impacts on a system 50 km away.While thesedata do not support the contention of widespreadlandscape alteration, we note a considerable increase inhuman alteration of the landscape of thewestern lakesin the millennium prior to abandonment. Clearly,palaeoecological evidence exists for a populationcollapse atmany sites across Amazonia and in isthmianregions of Central America (e.g. Bush et al 1989;Piperno 1990; Bush & Colinvaux 1994).In our experience of studying more than 30 lakerecords from Amazonia, a general rule of thumb hasemerged: ifthe lake has modern use, has a road to itorhas a settlement beside it, there is a very goodprobability of finding a long record (thousands ofyears) of disturbance. If, on the other hand, the lake isremote frommodern society, itprobably has a historylacking the characteristic signature of human activity.Denevans (1996) bluff model of occupation oftencoincides with the sites of modern occupation andhence our observations align well with his model.Although exceptions can always be found, modern useof an area is a powerful predictor ofpast human activity.We observe that the extent of deforestation aroundthe lakes that we discuss in this paper, and also theminimally seasonal locations of Kumpaka (Liu &Colinvaux 1988) and Ayauch1 (Bush et al 1989) inEcuador, was probably considerably less than aroundlakes inCentral America with comparable histories ofoccupation, e.g. La Yeguada (Piperno et al 1991) andLake Wodehouse (Bush & Colinvaux 1994). Itmay bethat the sites with very strong seasonally, whichincludes the Xingu, were more extensively impactedthan less seasonal settings. We suggest that extrapolating from known areas of intense occupation to inferthatmost of Amazonia was a parkland in 1492 AD(458 yr BP) would be unwise. Rather we suggest thatthe highly heterogeneous history of land use, exploitation levels and population densities in Amazoniashould be recognized, as these probably varied widelyacross thisvast region in response to local soils, climateand potential protein sources. We now know thattropical forests have been subject to considerableclimate change throughout theQuaternary. Forests ofthe Holocene are different in relative species abundance and productivity than theywere during much ofthe Pleistocene. And even within theHolocene, naturalPhil Trans. R. Soc. B (2007)
  9. 9. 216 M. B. Bush et al. Holocene fire and occupation inAmazoniadisturbance cycles such as droughts and disease may haveplayed an important role in reshaping communities.Hence, pre-Columbian Amazonian landscapes thatdiffer from those of today could have been due to bothdirect and indirect human influence, natural causes, ormost likely, some combination of all of these. Directhuman actions include land conversions such as fellingtrees and setting fires. The temporal and spatial scalesof these disturbances need to be resolved across therange of Amazonian landscape types. Indirect actionsinclude hunting, introduction of exotic species thatthenmodify a landscape or a disease that crosses fromone host to another. Megafauna have large and clearlandscape-level effects on forest structure and composition in systems where they still exist.What was theeffect of their extinction on Amazonian forests? If, ashas been suggested, the loss of megafauna wasattributable to humans, then the case for a humanmodification ofAmazonian landscapes is strengthened,although not on the time- or spatial-scale of the currentarchaeological debate.Natural disturbance in the Holocene of Amazoniawas probably also profound. While there is increasingdocumentation ofmillennial scale droughts (see ?3d),contemporary studies indicate the potential importance of unexplained changes in animal abundance. Inthe 1980s, white-lipped peccaries disappeared, apparently stochastically, from an area of more than70 000 km2 for 12 years. The absence of thismajorseed predator resulted in a surge in recruitment of thedominant rainforest tree species (Silman et al 2003,Wyatt & Silman 2004), an imprint that will last fordecades, perhaps centuries, on this region.Two potentially crucial differences exist betweenpre-Columbian human and natural disturbance inmany areas of Amazonia. The first is the scale ofdisturbance. Modern studies of fragmentation showthat duration and areal extent are critically importantto post-abandonment succession (Laurance et al2002). Were the landscapes created by pre-Columbianoccupants ofAmazonia on the scale of a large tree-fallgap, or orders of magnitude larger? How long diddisturbance persist on the same location? The second isthe introduction of fire. Repeated burning transformsrainforest communities, whereas a single fire has lessereffects (Cochrane & Schulze 1998; Haugaasen et al2003). Determining whether the same patch of forestwas burned repeatedly or whether there were longperiods of recovery between events is an importantquestion that needs to be resolved before makingpredictions of human impacts in any locale.The conservation message that emerges from this andother palaeoecological studies is thatmany forestsmay beonly one or two tree generations away from a period ofland use and that in such areas populations are unlikely tobe equilibrial (e.g. Foster etal 2002). However, the typeofmixed land use of the past was utterly different to thedeforestation that accompanies modern ranching andsoyabean farming.Away from the largest settlements, itisprobable that gaps created by pre-Columbian farmerswere closer to the scale of natural blowdowns, i.e.1-10 ha, rather than the thousands of hectares createdby agro-industry. Consequently, seed sources remainedlocally available and biodiversity was not reduced.Modern biodiversity and forest composition are nota product of land management, but have persisteddespite it.We are grateful toPaul Colinvaux forhis role indirecting thecoring ofLake Gentry and toChengyuWeng, Emilio Ancayaand Matthew Scripter for coring lakes Parker and Vargas.Carol Mitchell and Rolando Soto are thanked for logisticalsupport in Peru. Hermann Behling and an anonymousreviewer provided thoughtful comments that strengthenedthis paper. This work was funded by NSF grants DEB9732951, 0237573 (M.B.B) & 0237684 (M.R.S) and CNPq(doctoral scholarship 200065/99-8) (M.B.T).REFERENCESApaza, L., Wilkie, D., Byron, E., Huanca, T., Leonard, W. &P?rez, E. 2002 Meat prices influence the consumption ofwildlife by theTsimane Amerindians ofBolivia. Oryx 36,382-388. (doi:10.1017/S003060530200073X)Behling, H. 2001 Late Quaternary environmental changes inthe Lagoa da Curuca region (eastern Amazonia, Brazil) andevidence of Podocarpus in the Amazon lowland. Veg. Hist.Archaeobot. 10, 175-183. (doi:10.1007/PL00006929)Behling, H. & Da Costa, M. L. 2000 Holocene environmental changes from the Rio Curu? record in theCaxiuan? region, eastern Amazon Basin. Quat. Res. 53,369-377. (doi:10.1006/qres.l999.2117)Behling, H., Keim, G., Irion, G., Junk, W. & Nunes de Mello,J. 2001 Holocene environmental changes in the centralAmazon Basin inferred from Lago Calado (Brazil).Palaeogeogr. Palaeoclim. Palaeoecol. 173, 87-101. (doi: 10.1016/S0031-0182(01)00321-2)Bush, M. B. & Colinvaux, P. A. 1994 A paleoecologicalperspective of tropical forest disturbance: records fromDarien, Panama. Ecology 75, 1761-1768. (doi: 10.2307/1939635)Bush, M. B., Piperno, D. R. & Colinvaux, P. A. 1989 A 6000year history of Amazonian maize cultivation. Nature 340,303-305. (doi:10.1038/340303a0)Bush, M. B., Miller, M. C, De Oliveira, P. E. & Colinvaux, P. A.2000 Two histories of environmental change and humandisturbance in eastern lowland Amazonia. The Holocene 10,543-554. (doi:10.1191/095968300672647521)Bush, M. B., De Oliveira, P. E., Miller, M. C, Moreno, E. &Colinvaux, P. A. 2004 Amazonian paleoecologicalhistories: one hill, 3 watersheds. Palaeogeogr. Palaeoclim.Palaeoecol. 214, 359-393.Bush, M. B., Hansen, B. C. S., Rodbell, D, Seltzer, G. O.,Young, K. R., L?on, B., Silman, M. R, Abbott, M. B. &Gosling, W. D. 2005 A 17,000 year history of Andeanclimatic and vegetation change from Laguna de Chochos,Peru. J. Quat. Sei. 20, 703-714. (doi:10.1002/jqs.983)Bush, M. B., Listopad, M. C. S. & Silman, M. R. In press.Climate change and human occupation in PeruvianAmazonia. J. Biogeogr.Carneiro, R. 1970 Hunting and hunting magic among theAmahuaca of the Peruvian Monta?a. Ethnology 9,331-341.Clark, J.S. & Hussey, T. C. 1996 Estimating themass flux ofcharcoal from sedimentary records: effects of particle size,morphology, and orientation. The Holocene 6, 129-144.Clark, J. S. & Patterson, W. A. I. 1997 Background and localcharcoal in sediments: scales of fire evidence in thepaleorecord. In Sediment records of biomass burning andglobal change, vol. 51 (eds J. S. Clark, H. Cachier, J. G.Goldammer & B. Stocks). NATO ASI series 1: globalenvironmental change, pp. 23-48. Berlin, Germany: Springer.Cochrane, M. A. & Schulze, M. D. 1998 Forest fires in theBrazilian Amazon. Conserv. Biol. 12, 948-950.Phil Trans. R. Soc. B (2007)
  10. 10. Holocenefire and occupation inAmazonia M. B. Bush et al 217Denevan, W. M. 1996 A bluff model of riverine settlementin prehistoric Amazonia. Ann. Assoc. Am. Geogr. 86,654-681. (doi:10.1111/j.l467-8306.1996.tb01771.x)Erickson, C. L. 2001 Pre-Columbian roads of the Amazon.Expedition 43, 21-30.Faegri, K. & Iversen, J. 1989 Textbook of pollen analysis.Chichester, UK: Wiley.Foster, R. B. 1990The floristiccomposition of theRioManufloodplain forest. In Four neotropical rainforests, vol. 627(ed. A. H. Gentry), pp. 99-111. New Haven, CT: YaleUniversity Press.Foster, D. R, Clayden, S., Orwig, D. A., Hall, B. & Barry, S.2002 Oak, chestnut and fire: climatic and cultural controlsof long-term forest dynamics in New England, USA.J. Biogoeogr. 29, 1359-1379. (doi: 10.1046/j. 1365-2699.2002.00760.x)Glaser, B., Haumaier, L., Guggenberger, G. & Zech, W. 2001The Terra Pr?ta phenomenon: a model for sustainableagriculture in the humid tropics. Naturwissenschaften 88,37-41. (doi: 10.1007/s001140000193)Hammond, D. S. & terSteege, H. 1998 Propensity offire inGuianan rainforests. Conserv. Biol. 12, 944-947.Haugaasen, T., Barlow, J. & Peres, C. A. 2003 Surfacewildfires in central Amazonia: short-term impact on foreststructure and carbon loss. For. Ecol. Manag. 179,321-331.(doi:10.1016/S0378-l 127(02)00548-0)Heckenberger, M. J., Kuikuro, A., Kuikuro, U. T., Russell,J. C, Schmidt, M., Fausto, C. & Franchetto, B. 2003Amazonia 1492: pristine forest or cultural parkland?Science^!, 1710-1714. (doi:10.1126/science.l086112)Heckenberger, M. J., Peterson, J. B. & Neves, E. G. 1999Village size and permanence in Amazonia: two archaeological examples from Brazil. Latin Am. Antiquity 10,353-376. (doi:10.2307/971962)Laurance, W. F. &Williamson, G. B. 2001 Positive feedbacksamong forest fragmentation, drought, and climate changein the Amazon. Conserv. Biol. 15, 1529-1535.Laurance, W. F., Lo ve joy, T., Vasconcelos, H. L., Bruna,E. M., Didham, R. K., Stouffer, P. C, Gascon, C,Bierregaard, R. O., Laurance, S. G. & Sampaio, E. 2002Ecosystem decay of Amazonian forest fragments: a22-year investigation. Conserv. Biol. 16, 605-618.(doi:10.1046/j.l523-1739.2002.01025.x)Lima, H. N, Schaefer, C. E. R, Mello, J.W. V, Gilkes, R. J.& Ker, J.C. 2002 Pedogenesis and pre-Colombian landuse of "terra pr?ta anthrosols" ("Indian black earth") ofWestern Amazonia. Geoderma 110, 1-17. (doi:10.1016/S0016-7061(02)00141-6)Listopad, C. 2001 Vegetational changes, fire history and humanimpact during the last 6000 years: a paleoecological study of theMadre de Dios Province, Lowland Peru. Melbourne, FL:Florida InstituteofTechnology.Liu, K.-b. & Colinvaux, P. A. 1988 A 5200-year history ofAmazon rain forest. J. Biogeogr. 15, 231-248. (doi: 10.2307/2845412)Mann, C. C. 2000 Earthmovers of the Amazon. Science 287,786-789. (doi:10.1126/science.287.5454.786)Matheny, R. T. & Gurr, D. L. 1983 Variation inprehistoricagricultural systems of the New World. Annu. Rev.Anthropol 12, 79-103. (doi:10.1146/annurev.an.l2.100183.000455)Mayle, F. E. & Beerling, D. J. 2004 Late Quaternary changesinAmazonian ecosystems and their implications for globalcarbon cycling. Palaeogeogr. Palaeoclim. Palaeoecol. 214,11-25. (doi:10.1016/j.palaeo.2004.06.016)Mayle, F. E., Burbridge, R. & Killeen, T. J. 2000 Millennialscale dynamics of southern Amazonian rain forests. Science290, 2291-2294. (doi:10.1126/science.290.5500.2291)Meggers, B. J. 1954 Environmental limitation on thedevelopment of culture. Am. Anthropol. 56, 801-824.(doi:10.1525/aa.l954.56.5.02a00060)Meggers, B. J. 1995 Judgingthe futureby thepast: the impactof environmental instability on prehistoric Amazonianpopulations. In Indigenous peoples and the future ofAmazonia: an ecological anthropology of an endangeredworld, vol. 162 (ed. L. E. Sponsel), pp. 15-43. Tucson,AZ: University of Arizona Press.Meggers, B. J. 2003a Natural versus anthropogenic sources ofAmazonian biodiversity: the continuing quest for ElDorado. In How landscapes change (eds G. A. Bradshaw& P. A. Marquet), pp. 89-107. Berlin, Germany: Springer.Meggers, B. J.20036 RevisitingAmazonia circa 1492. Science301, 2067. (doi:10.1126/science.302.5653.2067b)Moy, C. M., Seltzer, G. O., Rodbell, D. T & Anderson,D. M. 2002 Variability of El Ni?o/Southern Oscillationactivity at millennial timescales during the Holoceneepoch.Nature 420, 162-164. (doi:10.1038/nature01194)Nepstad, D., Lefebvre, P., Lopes da Silva, U., Tomasella, J.,Schlesinger, P., Sol?rzano, L., Moutinho, P., Ray, D. &Guerreira, B. J.2004 Amazon drought and itsimplicationsfor forest flammability and tree growth: a basin-wideanalysis. Global Change Biol. 10, 704-717. (doi:10.1111/j.l529-8817.2003.00772.x)Paduano, G. M., Bush, M. B., Baker, P. A., Fritz, S. C. &Seltzer, G. O. 2003 A vegetation and fire history of LakeTiticaca since the Last Glacial Maximum. Palaeogeogr.Palaeoclim. Palaeoecol. 194,259-279. (doi:10.1016/S00310182(03)00281-5)Phillips,O. 1993The potential forharvesting fruitsintropicalrainforests: new data from Amazonian Peru. BiodiversityConserv. 2, 18-38.Piperno, D. R. 1990Aboriginal agricultureand landusage inthe Amazon Basin, Ecuador. J. Archaeol. Sei. 17, 665-677.Piperno, D. R. & Becker, P. 1996Vegetational historyof a sitein theCentral Amazon basin derived fromphytolithandcharcoal records from natural soils. Quatern. Res. 45,202-209. (doi:10.1006/qres.l996.0020)Piperno, D. R. & Pearsall, D. M. 1998 The origins of agriculturein the lowland neotropics. San Diego, CA: Academic Press.Piperno, D. R, Bush, M. B. & Colinvaux, P. A. 1991Paleoecological perspectives on human adaptation inPanama. II: the Holocene. Geoarchaeology 6, 227-250.Pitman, N. C. A., Terborgh, J.W., Silman, M. R, Nunez,P. V, Neill, D. A., Cer?n, C. E., Palacios, W. E. & Aulestia,M. 2001 Dominance and distribution of tree species inupper Amazonian terra firme forests. Ecology 82,2101-2117. (doi: 10.2307/2680219)Prance, G. T & Schubart, H. O. R. 1977 Nota preliminarsobre a origem das campi?as abertas de areia branca dobaixo Rio Negro. Acta Amaz?nica 7, 567-570.Roosevelt, A. C. 1980 Parmana: prehistoric maize and maniocsubsistence along theAmazon and Orinoco. New York, NY:Academic Press.Roosevelt, A. C. 1987 Chiefdoms in the Amazon andOrinoco. In Chiefdoms in the Americas (eds R. C.Drennen & C. A. Uribe), pp. 153-185. Lanham, MD:University Press of America.Roosevelt, A. C. 1991 Moundbuilders of the Amazon:geophysical archaeology on Maraj? Island, Brazil SanDiego, CA: Academic Press.Roosevelt, A. C, Housley, R. A., Imazio da Silveira, M.,Maranca, S. & Johnson,R 1991 Eighth millenniumpottery from a prehistoric shellmidden in the BrazilianAmazon. Science 254, 1621-1624. (doi:10.1126/science.254.5038.1621)Saldarriaga, J.G. & West, D. C. 1986 Holocene fires in thenorthern Amazon basin. Quatern. Res. 26, 358-366.(doi: 10.1016/0033-5894(86)90095-5)Phil Trans. R. Soc. B (2007)
  11. 11. 218 M. B. Bush et al. Holocene fire and occupation inAmazoniaSanford, R. L., Saldarriaga, J.,Clark, K. E., Uhl, C. & Herrera,R. 1985 Amazon rainforest fires. Science 227, 53-55.Santos, G. M., Gomes, P. R. S., Anjos, R. M., Cordeiro,R. C, Turcq, B. J., Sifeddine, A., di Tada, M. L.,Cresswell, R. G. & Fifield,L. K. 2000 14CAMS dating offires in the central Amazon rain forest. Nucl. Instrum.Methods Phys. Res. B 172, 761-766.Servant, M., Fontes, J.-C, Rieu, M. & Sali?ge, X. 1981Phases climatiques arides holoc?nes dans le sud-ouest delAmazonie (Bolivie). C. R. Acad. Sei. Paris, Ser. II 292,1295-1297.Silman, M. R., Terborgh, J.W. & Kiltie, R. A. 2003Population regulation of a dominant rain forest tree by amajor seed predator. Ecology 84, 431-438.Smith, N. J.H. 1980 Anthrosols and human carrying capacityin Amazonia. Ann. Assoc. Am. Geogr. 50, 553-566.(doi:10.1111/j.l467-8306.1980.tb01332.x)Stockmarr, J. 1971 Tablets with spores used in absolutepollen analysis. Pollen Spores 13, 615-621.Stuiver,M. & Reimer, P. J. 1993 Extended 14C databaseand revised CALIB radiocarbon calibration program.Radiocarbon 35, 215-230.Turcq, B., Sifeddine, A., Martin, L., Absy, M. L., Soubles, R,Sugio, K. & Volkmer-Ribeiro, C. 1998 Amazoniarainforest fires: a lacustrine record of 7000 years. Ambio27, 139-142.Vickers, W. T. 1988 Game depletion hypothesis of Amazonian adaptation: data from a native community. Science 239,1521-1522. (doi: 10.1126/science.3353699)Whitlock, C, Bradbury, J. P. & Millspaugh, S. H. 1997Controls on charcoal distribution in lake sediments:case studies from Yellowstone National Park and northwestern Minnesota. In Sediment records of biomass burningand global change (eds J. S. Clark, H. Cachier, J. G.Goldammer & B. Stocks), pp. 367-386. Berlin, Germany:Springer.Whitlock, C. & Larsen, C. 2001 Charcoal as a fire proxy. InTracking environmental change using lake sediments, vol. 3(eds J. P. Smol, H. J. B. Birks & W M. Last), Terrestrial,algal and siliceous indicators, pp. 75-98. Dordrecht, TheNetherlands: Kluwer Academic Publishers.Wyatt, J. L. & Silman, M. R. 2004 Distance-dependence intwo Amazonian palms: effects of spatial and temporalvariation in seed predator communities. Oecologia 140,26-35.Phil. Trans. R. Soc. B (2007)