141 Ecologr in the Long View:Settlement Histories, Agrosystemic Strategxes,and Ecological Performance KarlW. Butzer The University of Texas at Austin Austin, Texas TherecentColumbianpolemiccontrastedbeneficialNew Worldland usebefore1492 withdestructiveOld Worldland mana,gement. Sincearchaeolo,gists are uniquely equippedto documentand modellon,g-term settlementand land-usehistories,thereis bothchallen,ge and opportunityto empiricallyexaminetheecolo,gical impactofparticular a,grosystems withinlon,gtimefrawes. Thispaperexaminestherisk-minimization and ecolo,gicalfine-tunin,g of theMediterraneana,grosystem, and its lon,g-term ecolo,gicalperformgnce.TheMediterraneanecosystem is theproductof millenniaof co-evolution betweentheenvironmentand humanactivities,buttraditionalland usehasbeen conservative and ecolo,gically adaptive,despitesporadicdisequilibrium.Thedeviations from thenormposea numberof testablehypothesesforfurther examination,not onlyin theMediterraneanre,gionbutalsoin otherareassuchas theNew World. Introduction The recent polemic surroundingthe Columbian Encounter served to underscorethat the most interesting issues of today are those that cross disciplinarylines. Whereas theoreticaldebates within archaeologyremain ignored by other social scientists, bioarchaeologicalresearchin Mesoamericaor the Pacificregion is widelycited in the national or internationalliteraturedealing with human ecologicalimpact.Archaeologywas highly visible in earlierscientificexchangesabout agriculturalor human origins, and archaeologytoday not only has an opportunity but an obligationto becomeengagedin mattersof the human use and modificationof the environment.This paperjuxtaposesquestionsof Mesoamericanand Mediterraneanland-useand environmentaldegradation,centered on a regionalexplicationof the complexityand sophistication of adaptivestrategiesin the Mediterranean world.The concludingsection identifiesquestionsthat archaeologyis particularlysuitedto address. One of the unfortunate myths broadly disseminated before and during the Columbianpolemic was that of ecologically-invisibleNew Worldpeoples. Living in harmonywith nature,NativeAmericanssupposedlyemployed environmentally sensitiveand sustainableagricultural practices that preservedthe wildernessof the Americaswith "barelyperceptiblehuman disturbance."That claim, le- gitimized by the Smithsonianvolume, Seedsof Chan,ge (Shetler 1991: 226), flies in the face of generationsof archaeologicalresearchthat demonstratedlarge populations andintensifiedagriculturein Mesoamericaand other partsof the New World,includingurbanismand complex social organizationin some. Regardlessof an ethos of harmonywith nature,high populationdensitiesmust be providedwith food, and the recordfromNorth and South Americaprovidescountlessexamplesof ridgedor ditched fields, terracedhillsides, destroyed forests, or persisting roadwaysthat suppliedenergy for great mound sites or monumentalcities (Denevan1992; Doolittle 1992; Whitmore and Turner1992). To deny that New Worldpeoples humanizedand modifiedthe environment,or createdtheir own culturallandscapes,is to reinforcethe metaphorof the 17th-centurypreacher,Cotton Mather,who referred to them as primitiveaborigines,blendinginto the forest like beavers(see Zuckerman1987: 144). A moreappropriatequestionto raiseis whetherthe large populationsin centersof New Worldcivilizationwere able to grow and persistwithout ecologicaldamage (Williams 1989; Butzer 1992a). Demographicpressuresand finite resourceswill eventuallycurtail fallow cycles or require expansionof agricultureonto fragilesoils, and the accumulatingevidencerevealsthat there was no magicalformula to avoid the ecologicalconsequences.AroundTexcoco, in the Basin of Mexico, CarlosCordova(personal 142 Ecologyin theLongView/Butzer communication,1994) has recentlyidentifieda series of late Aztec sites within massivesheets of alluvium,derived from soil erosion in the foothills. Similar implications follow from the limnologicalwork of O'Hara,Street-Perrott, and Burt (1993) in the Tarascanheartlandof Lake Patzcuaro,where indigenousland use provokedepisodes of soil erosion as earlyas 1900 B.C., and soil destruction peaked during the three centuriesprior to the Spanish intrusion. In the Mixteca Alta of Oaxaca, severalprotractedepisodes of pre-Conquestsoil erosion have been identified.Large-scaledeforestationis evenverifiedin Late Formativeor Classic contexts in the Basin of Mexico (Gonzalez,Fuentes,and Fuentes 1980) and coastalVeracruz (Byrneand Horn 1989) . Suchevidenceraisesquestionswhetherindigenousagriculturewas indeed "sustainable,"a questionreconsidered below. It would also seem to urge cautionagainstwholeheartedlyembracingcontemporary,traditionalagriculture as a "system,"ratherthan as a sourceof informationto be drawnupon in selectingmore beneficialagriculturalpracticesandstrainsof crops(Butzer1992a,1993a). As repugnant as high-investmenttechnodevelopmentis to most of us, it would be irresponsibleto recommendtraditional land managementpracticessimplyon emotionalor ideological grounds,without pragmatictesting and comparatlve eva.uatlon. Prehispanicenvironmentalmodificationis quite pertinent for conservationargumentsor issuessuchas biodiversity. The early Spaniardsencountered deforested landscapes from Mexico to Panama (Sauer 1966: 265-88; Butzerand Butzer 1993), some of whichhad been reconstitutedby a centuryago, while earlierphasesof deforestation are amplyattestedbypalynology(Rue 1989; Abrams and Rue 1988; McDade 1994; Piperno,Bush, and Colinvaux 1991). Large swaths of the Peten rainforestwere completelycleared2000 yearsago, but have since recovered sufficientlyto be cited as reservoirsof primevalbiodiversity(Rice, Rice, and Deevey 1985; Vaughan,Deevey, and Garrett-Jones1985). That not only challengessome popularassumptionsbut, more constructively,could be used to estimateratesfor the reconstitutionof forestsand Butzer 1990). Biologistsare certainlyno better qualified with the human than archaeologiststo deal systematically side of that interrelationship.Like most largerproblems confronting human society today, ecological issues are sufficientlycomplex to demand a new and greatly exinteractionand colpanded medium of interdisciplinary laboration.Archaeologyshould, indeed must, be a major player.It has unique capabilities,not only to generatean indispensablearchaeologicaldatabase,but also to understandlong-termland use andits implications. The MediterraneanAgrosystem:Fine-tuning and Risk-minimization The Columbianpolemic servesto highlightsome critical issuesof ecology in the long view,specificallythe myth of the devastatedColoniallandscape(Butzer1992b). Did early European settlers indeed introduce agricultural methods, livestock, and technologies that promptlydegradedNew Worldecosystems?As a categoricalassertionit agrosystem,with its cattle,sheep,and is the Mediterranean plow technologythat continuesto be singled out as particularlydestructive (Klein 1920; Simpson 1952; Sale 1990; Melville 1990). In realitythere are two different issues here, one whetherthis agrosystemand its components are inherentlydestructive,the other whether the transferof this agrosystemto the New Worldautomatically had a negativeecologicalimpact.The firstof these questions will here be consideredin some detail, while the second is reviewedin the finaldiscussion. The basicMediterraneanagrosystemthat can be documented from Greek and Roman sources, or verifiedby sophistiethnographicand archivalwork, was remarkably cated (White 1970; Dimen and Friedl1976; Spurr1986; Delaigue1988; Butzer1993b,1994;Wells 1992: 87-101; also Burford 1993). It had four basic components: a) outfield cultivationof a selectionof grainsand legumes, suited for local soils and climate; b) infield tending of variousgreen vegetablesand condiments,in kitchen or marketgardens;c) orchardcrops,not only grapevinesand olive groves, but also an arrayof other fruit trees;and d) livestock,among which sheep, goats, or pigs were more commonly kept than cattle, with the exception of draft thelr specles dlverslty. These are a few examplesof archaeologyinterdigitating oxen. But specificplantsand animalsare less informativefor with problemsof ecology, in both the past and present. an agrosystemthantheirsystematicintegraMuch of the primarywork on environmentalchange, characterizing ecologicaldegradation,or sustainabilityhas been monop- tion in termsof ecology,schedulingand management,or olized by biologists or earth scientists,who admittedly cuisine (Butzer 1988b). Each of the components was carefullyadjustedto the local environmentalmosaic,and havethe specializedskillsrequisitefor certaincategoriesof includedwinteraswell as summercropsthatwereplanted, the analyticalresearch.But contemporaryecosystemsare product of millennia of co-evolution between environ- tended, and harvestedaccordingto an elaborateseasonal includedcomplicated mentalcomponentsand human activities,especiallysince schedule.Their micro-management instructionsfor plowing, spading,transplanting,grafting, lifeways(Birkset al.1988; the establishmentof agricultural . . . . . . 23, 1996 143 Journcllof FieldArchcleology/Vol. pruning, selection or applicationof fertilizers,watering, and pest control. Intercroppingwas the norm for fruit trees and outfieldcrops,with a growing appreciationthat grassfallow,nitrogen-bindinglegumes, and crop rotation all helpedto maintainor restorefertility. Stockraisingincludeda nucleusof householdanimals,as well as herdsof sheep and goats movedunderclose supervision within the municipalcommonageor in a seasonal, transhumantcycle between lowlands and adjacenthighlands (Vassberg1984; Halstead 1987; VVhittaker1988; Butzer 1988a; Boyazogluand Flamant1992). Goats and sheep provided dairy products or wool, with selective cullingof young male animalsand occasionalslaughterof an adultfemalein times of food shortage.Animalgrazing on stubbleafterthe harvestprovidedindispensablemanure for outfieldcrops.Herdsof pigs wererestrictedto environmentswith oak woodlandsthat allowedseasonalfattening on acorns.Finally,cattle were commonlyraisedon prime pasturesand bred to produce draft animals.Only at the peakof Romanurbanismwas cattle-raisingpracticedwith improvedbreeds, to supply meat and dairy products to urban markets (White 1970; Garnsey 1988; Butzer 1992b). This Mediterraneanagrosystememphasizedrisk-minimization, by a sequence of autumn, winter, spring, and earlysummerplantingsthat anticipatedsporadic,earlyor late, killing frosts. The mix of outfield agriculture,orchards, and infield gardens with micro-irrigationall reduced the risksof prolongedsummerdrought. Livestock, less vulnerableto climaticperturbationsby virtue of their mobility,offeredanotherbufferagainstsubsistencestress. Animalsalso supplied manureor fuel, fiber, and animal protein that furtherreduced subsistencerisk. Finally,the production of grains, fresh vegetables, preservedfruits, cheesesand hams,leatheror wool, and aboveall,wine and olive oil, representeda viablebut flexiblecommercialcomponent. That in turn favoredintra- and interregionalexchange networks(Renfrewand Wagstaff1982; Osborne 1983; Garnsey1988). Dur1987; Garnseyand VVhittaker ing periodsof populationgrowthand regionalintegration, such networksprovidedincentivesfor optimizing market opportunities,mitigatingthe dangersof periodic subsistence stressthroughinterregional"averaging." In view of the deep-seatedNorth Europeanand North Americanbiasesand misconceptionsabout the traditional Mediterraneanagrosystem,it is ironic that it was one of the world's most diversifiedruraleconomies, its stability predicatedon highlycomplexmanagement.The interlocking, risk-minimizingstrategieswere so effectivein anticipating periodic shortfallsthat disastrousfamineson the scaleexperiencedin NWEuropeduringmedievaland early modern times had no counterpartin the Mediterranean Basin. Furthermore,its commercialcomponents could readilybe scaledback duringperiodsof decliningurbanism, lowerpopulationlevels,andreducedregionalintegration, but expandingagain during times of economic and demographicgrowth.Thus ruralpopulationscould adjust theirallocationof laborandsurplusesto changingdemand acrossthe centuries,or compensatefor local crop failures in some yearsby greatermarketreturnsduringothers. Extensivepastoralismwas integral to the flexibilityof that agrosystem,as a meansto utilize extensivemountainous landscapesunsuitablefor agriculture,withoutcompeting for arableland. It also served as a "pioneer"form of land use in the resettlementof depopulatedareaswhich could only graduallybe convertedto agriculture.Recent experimentalstudiesof pasturemanagement,productivity, and nutrient cycles have yielded unexpected results (Bernaldez 1995) . Traditional management created a parklandvegetationof open treesthroughregularpruning and the controlleduse of fire (Trabaud1991). Wild and semi-domesticated legumes were naturally dispersed throughlivestockmovements,enrichingunimprovedpastureswith nitrogenas well as phosphates.The tree canopy favors microclimaticsheltering, and liveoak litter significantlyimproves soil nutrient levels. Even bushy or shrubbymatorral vegetationprotectsagainsterosion and enhancesinfiltration,while its leavesprovide browse and its woody partscan be convertedto charcoal. In other words, the ecology of supposedlyunimproved pasture land in the MediterraneanBasin is largely beneficial,its structureand dominantformsreflectingcoevolutionin responseto summerwaterstressand human manipulation(Bernaldez1991, 1995). Such managedenvironments represent a controlled subclimaxformation and are known by the Spanishterm dehesa.From this perspective,transhumancecan be seen as an effectivecomplement to agricultureas well as a sustainableform of utilizationfor marginallands.Mediterraneanagropastoral systemswerenot only tightlyintegratedbut carefullymanaged.The demiseof pastoralismduringthe last 50 years,in part under governmentpressure,has led to expansionof (Tomaselli unproductive, thorny and bushy matorral 1977; MontoyaOliver1983; Bernaldez1991). Much like the analogous chaparralof southern California,it is a potential tinderbox,waiting to explode in wildfiresthat also destroyinterfingeringwoodland(Le Houerou 1987). The Bottom Line:Long-termPerformance of the MediterraneanAgrosystem This ecological analysisof the Mediterraneanagrosystem servesto explicatethe complexweb of strategiesthat are designed to minimize vulnerabilityand risk. They reflectnot only the cumulativeexperienceof 10 millennia 144 Ecologyin theLongView/Butzer in a summer-dry,subtropical of agriculturaltrial-and-error climate?but also local and externalinnovationsthat were tested, adapted,and disseminated.This historicaltrajectory can be amplyreconstructedby archaeologyanddocumentarysources.It can also be independentlymonitored from geoarchaeologicalrecordsof alluviationin response to soil erosion, or from bioarchaeologicaldata on changing vegetation.There now are over three dozen, detailed and dated pollen cores from all around the Basin that recordlocal land use and ecological modificationover at leastthe last six to eight millennia. Withoutelaboratingon that investigationand its methodological underpinningshere, these palynologicaldata allow a) qualifieddistinctionof pastoraland agricultural "indicators",b) identificationof basicagriculturalmodes, and c) semiquantitativeassessmentof vegetation disturbance, deforestation,or secondaryregenerationof woodland (Bottema,Entjes-Nieborg,andvan Zeist 1990). Suchresultsdivergefrom charcoalor seed identification from site flotationstudies, becausethey representa composite picture for a larger area, much of it unutilized, ratherthan the disturbedenvironsof a site. Pollen spectra of the same age from a lake or swampcore and from an archaeologicalsite within the same watershed also are radicallydifferent(van Zeist and Bottema 1991: fig. 41a versus 41b). Long pollen cores tend to be limited in resolutionby samplingintervals,and dating must all too often be interpolatedbetween widely spacedradiocarbon dates. Yet the advantagesare obvious, with a relatively continuous trace of composite ecological change, sometimes punctuatedby strong regionalperturbations,comparedwith the discontinuousrecordsand much sparseror secondaryarborealrecord typical of charcoaland seeds from an archaeologicalsite (see Vernet,Badal,and Grau 1983; Vernet,Thiebault,and Heinz 1987; Miller 1991). This is primarilya matterof scale,but it is more than that, because survivalof more distant woodlands within the samewatershedfacilitatesecologicalrecoveryas the degree of disturbanceis reduced.In that sensethe multiplewatershedsof complexenvironmentssuch as those of the MediterraneanBasin also differ fundamentallyfrom mid-oceanic,islandecosystems,the ecology of which can be more rapidlyand perhapsirretrievablydegraded(Flenleyet al. 1991; Kirchet al. 1992). Equallyso thereareproblemsin reconcilingpollen core data with episodes of destructiveland use inferredfrom soil erosionas recordedin dated colluviumor alluvialfills. Such geomorphologicalcriterianormallyare discontinuous, documentingrupturesof soil and slope equilibrium (vanAndel, Zangger,and Demitrack1990; vanAndeland Runnels 1987). Yet the natureof cause-and-effectin the crossingof such a thresholdis open to interpretation,in part becausechannelfills,overbankdeposits,debrisflows, and colluvial spreads reflect different processes, and it remainsdifficultto exclude extremeweatherevents and their recurrenceintervalfrom a complementaryrole in erosionor deposition.Equallydifficultis exactdatingand accuratecorrelationwith a particularstageof a local settlement history (e.g., Wells, Runnels, and Zangger 1990; Bell and Boardman1992: Section C). For example,does initialclearance,long-termpressure,or the constructionor partial(versustotal) abandonmentof agriculturalterraces push disturbanceor degradation beyond that critical thresholdthat promotessoil erosion in a particulararea? Last but not least, a pollen core will sample a unique columnof sediment,whereassoil erosiondataarespatially distributedin a complexlandscape,or must be linked by difficultlateralcorrelations.Not surprisinglythe results may diverge,as evidentfrom comparisonof two interpretations of Lake Lerna in the GreekArgive Plain (Jahns 1993 versusZangger1991). Ideally,palynologicalrecords should be supplementedby measurementsof magnetic susceptibilitythat give a continuous,quantitativetraceof mineral sediment accretion,potentiallysensitive to soil erosionin the catchment. For our purposes,episodesof soil erosion give a signal for disequilibriumthat complements the palynological datainterpretedhere. Ideallyall these categoriesof informationrequirecomparisonanddiscussion,againsta background of regionalsettlementhistory(that is rarelyavailable), but that indeed would require book-length presentation.What follows is an outline that seeks to identify basic patterns and place certain archaeological . . ssues lnto > tocus. In general,vegetationwas relativelyundisturbeduntil about 5000 B.C. (calibrated)in the easternMediterranean region and until roughly2000 B.C. in the westernBasin. Nonetheless,therewere scatteredepisodesof moderateto intense disturbance,due to agropastoralintrusion into localcatchments,generallyfollowedby partialor complete woodland recovery.There also is some evidence of soil erosion, varyinggreatlyin intensityfrom place to place, and timing (van Andel, Zangger, and Demitrack1990; Bruckner1986; J. T. Abbott, personalcommunication, 1994). A combinationof agricultureandincipientarboriculture (chestnut,walnut)is firstevidentin Syriaby 5000 B.C. and in Greece about 3000 B.C., but olives or grapes, as a hallmarkof true Mediterraneanagriculture(Goor and Nurock 1968; Stager 1985), are only verifiedlater: in Palestineshortlyafter3500 B.C.; in Greeceabout2000 B.C.; and in the westernMediterraneanafter 500 B.C. (Baruch 23, 1996 145 Journglof FieldArchcleology/Vol. 1986; Jahns1993; Reille 1979, 1984; Runnelsand Hansen 1987; Reille and Lowe 1993; Menendez Amor and Florschutz1961). That E-W gradientis compatiblewith directarchaeologicalevidencefor olives or grapesor their products(Jashemski1979; Stager 1985; Heltzer and Eitam 1987; Zohary and Hopf 1993), as well as for plow agricultureand possiblyhillsideterracing(see Wells1992: 29-57, 123-130). In some places,but not in others, the apex of agriculturaldevelopmentled to modest soil erosion (van Andel, Zangger, and Demitrack1990; Butzer 1993c). Unexpectedis that this agrosystemsubsequently was regressive,as earlyas 300 B.C. in some areas,or as late as A.C. 1000 in others,with woodlandreconstitutionduring the early or late Middle Ages (Reille 1976, 1977; Lamb, Eicher, and Switsur 1989; Bottema 1984; Ikuniholm and Striker1987). LocalagriculturaldeclineduringClassicaltimes (Argive Plain: Jahns 1993; western Anatolia: van Zeist and Bottema 1991) reflectedprovincial?urbandecayin some areascontrolledby Rome and mayhave been accentuated by soil destruction.Agriculturalrecessionis also apparent during the Greek Dark Ages about 1000 B.C. (Bottema 1984). Distinctivein severalareasis strongpastoraldisturbance, in western Anatolia and northern Greece about 1200 to 800 B.C. (Bottema 1984; van Zeist and Bottema 1991: 82), in Iberiaaround 1000 to 500 B.C. (Stevenson and Harrison1992), or in Palestineand NorthwestAfrica from A.C. 900 to 1400 (Baruch1986; Reille 1976, 1977; Ritchie 1984). These representtimes when the areasin questionwere exposedto pastoralistinvasionsfrom either temperateEuropeor the adjacentdeserts,and such disturbancesometimesled to substantialsoil erosion.Ironically, however,the pollen recordsof northernGreeceshow that forest degradationincreasedspeciesdiversityby an order of magnitude(Bottema1984); biodiversityissuesin tropical and temperateforestsarenot identical. A combinationof charcoalstudiesand palynologysuggests that increasinghuman modificationof the environvegetation shift ment accentuateda climatically-induced during mid-Holocene times, favoringa change from deciduous to live oak woodland and an increasingprominence of matorral,beginningat some point between4000 and 2000 B.C. (Vernet, Badal, and Grau 1983; Vernet, Thiebault, and Heinz 1987). A recent, discriminating study in sw Spain (Stevensonand Harrison1992) argues managedgrazing,by meansof controlthatexperimentally led use of fire, was underwaysince 4000 B.C.; by 500 B.C. the local dehesasystemhad been perfected,with no basic changethroughRoman,Islamic,or Christiantimes.Alternatively, in the temperate woodlands of Anatolia and northern Greece, sylvopastoralpractices favored some treesoverothersafterabout 1700 B.C. (Bottema1984; van Zeist and Bottema1991). This synopsisgives a glimpseof the complexityof ecology in the long term. The history of land use and landscapeecology in the MediterraneanBasinwas a checkered one, with punctuatedchanges,long intervalsof stability, and shorterepisodes of mismanagement,periodicallyinterruptedby ecologicalrecovery.Despitea recentspateof archaeologicalor documentarystudies devoted to longterm settlementand land-use,particularlyin Greece(e.g. Renfrew and Wagstaff 1982; Keller and Rupp 1983; Snodgrass1987; van Andel and Runnels 1987; Cherry, Davis,andMantzourani1991; Wells1992), but alsoin the Levantand Spain,the ruralhistoriesof largeareasremain little explored.But what standsout is that this relatively modest environmenthas supportedagriculturalpopulations across 5-10 millennia, and urban societies for some 20004000 years. Whereasthe landscapehas indeed been humanized, modified, impaired, or altered, traditionalagropastoralsystemshave maintainedtheir basic productivityacross200 generationsor more. For all purposes,in the long test of history,Mediterraneanlanduse systemshave been sustainable,despite sporadicevidence for destructivehuman intervention in time and space. Implications for Mesoamerican Land-use Histories This Mediterraneanrecordappearsto be analogousto the more fragmentaryevidencecurrentlyemergingfrom Mexico and CentralAmerica.That should not be surprising because long ecological records will not reflect an idealizedrelationshipbetween people and their environment, but ratherthe harshrealitiesof events or circumstancesthat could not be controlled.Intervalsof destructive land use in PrehispanicMesoamericaare not an indictment against the conservationismof New World agrosystems,any more than they are of their Mediterranean counterpart.Both traditionsembodied millenniaof experiencethat was translatedinto a refined system to minimize subsistencerisk. Both, in the main part, were successfulin achieving that goal, although internal or external stress at times required a shift of prioritiesto short-termsurvival,with unfortunatebut not generally irreversibleecologicalrepercussions. Just what happenedwhen the Mediterraneanagrosystem was transferredto Mexico and "superimposed"on indigenous systems?The limited impact of Spanishlivestock and agriculture,as now documented(Butzer1992c, 1995; Butzer and Butzer 1993, 1995), is once again . counterlntultlve: . . 146 Ecolo,gyin the Lon,gView/Butser 1. The Spanish-Mediterranean agrosystemwas simplifiedin the New World,becauseof ecologicalconstraints, insufficientmanpoweror agronomicexpertiseamong the Europeansettlers,and limited economic competitiveness versusindigenousagriculturalproductivityandproducts. 2. The indigenousagrosystemexpandedits own repertoire of cultivars,especiallyby incorporatingMediterranean fruit trees in the highlands, or citrus fruits and bananasin the lowlands,the lastintroducedfromAfricavia the CanaryIslands.Outfieldandinfieldcropsandmanagement remainedlittle changed. 3. Giventhe almostunlimitedpastureslying beyondthe northernlimitsof indigenousagriculturein Mexico,stockraisingbecameless controlledbut farmore mobilethan in the Mediterranean world.Afterinitialconflictwith indigenous farmers,stockraisers were pushedout to the frontiers by governmentpolicy.Balancedagainstdamagessustained and lands lost, the indigenouspeoples did deriveimportant benefitsfromEuropeanlivestock:reducedsubsistence risk; improved nutrition with significantlymore animal protein;and manure-for theirfieldsor fuel for theirfires. 4. Probablyas a resultof high livestockmobility,thereis no evidenceforvegetationchangeor soil erosionduringat leastthe firstcenturyof colonizationin Mexico contrary to well-entrenched,popularassumption.But for reasons that are unclear,the Colonial governmentattemptedto interdictfiremanagementof rangelands,possiblybecause it could not be properlycontrolled,as a resultof the labor shortageand the inexperienceof hiredherders. 5. Finally,tangible evidence for ecological disbalance, including stream alluviationwith eroded soil products, only becomes apparentduring the course of the 18th century,a time of rapid demographicexpansion(C. D. Frederick,personal communication,1995; C. Cordova, personalcommunication,1995). But that comes on the eve of proto-industrialization and in the wake of fundamentaleconomic change, not only in Mexico but also in easternNorth America. regions there are a number of intervalswith very low archaeologicalvisibility,typicallylastingfrom 500 to 1000 years.Most of these fall between the earlyNeolithic and the earlyBronzeAge, and they areverifiedby local settlement surveys. C? lattiveexperiencesho?ldselectfor vnoreconservationiststrattegies. By BronzeAge times much largerpopulationshadless impacton forestor soil coverthandid their Neolithic predecessors.S?bsistence experienceis atbouthow to vnininnizeboth lon,g-tertnenvirontnentatl datvnat,ge gnd short-tertns?bsistencerisk. Diversificationof agronomic strategieswas the key to reducedsubsistencerisk,as suggested by continuing amplificationof the repertoireof orchard and garden crops, and tentative evidence that seasonal transhumancesystems were being operationalized. Nonetheless, increasingand more persistentpressureson resourcesreinforcedthe Holocene climatictrend to a more xericvegetation,thus creatingthe characteristic, subclimaxvegetationcommunitiesof the Mediterranean world.Fromthen on, the Mediterranean environmentwas more or less continuouslymanagedas an artificialecosystem that, on the whole, was sustainable. Expatnsionatndwestwardpro,gression of the Mediterraneatnat,grosystenn, with its diagnosticolive and grapeproduction, represents bothatn intensificattio n atndconntnerciatlizattionof prod?ction,linked to the developtnent of ?Wrbatn centersatndtheir vnarketdennatnd. Commercialization implies a redirectionand refinementof managementpractices,suchas the seasonalworkcalendar,in the directionof risk-reducingbut production-satisfyingstrategies that meet marketdemand.Thatin turnrequiresintensification and greatercapitalinvestment in technology,cultivars, livestock,and an appropriateinfrastructureof buildings, roads, and agriculturallandforms.Subsequentregression or reexpansionof suchcommercialagricultureshouldhave closelyparalleledregional,urbanhistories.At some point urbandemandand high ruralpopulationscould placethe ecosystemitself under stress,triggeringperiodsof degradation. ConcludingHypotheses Destructive pastoral land use was not endemic, but In retrospect,it is the temporaryexceptionsto agricul- region- and time-specific.In part it was linked to the turalstabilityin the long-term Mediterraneanrecordthat incursionof new peoplesfromother ecozones, suggesting suggestthe most interestinginferences. thattpriorenvirontnentatl experience catnnotbetratnsplatnted A,grosystenns evolvethro?,ghtriatlatnderror;earlyol,gric?l- ontonewecolo,gies witho?tinitiatldoltnat,ge. Only long-term t?rewatsexperinnentatl in natt?Wre, atndinitiatltechnostratte,giesenvironmentalexperiencewill encode reasonablyaccurate wereexploitattive, atndoftenephenneratl. informationas to how to use a particularenvironmental Substantialecological damagewas done by very small groups of Neolithic matrix,or as to the dispositionof micro-scaleactivities, people during fairlybrief periods (Kohler-Rollefsonand selection of crops and animals,and relatedmanagement Rollefson1990). Whethersucheventswereso maladaptive strategies.Equallyso, however,uncontrolledpastoralexas to requirelocal abandonmentand settlementdispersal ploitationwas facilitatedby politicaldevolution.This suginto new areas(Butzer 1990) is unclear,but acrosssome gests that politicollinstatbilityatnd r?Wratl insec?Writy will 23, 1996 147 Journgl of Field Arehgeolo,gvv/Vol. of revnatng,gennent incentivesfor conservattionist red?Wse sources,as long-term optimizationtakes second place to short-termsurvival. Finally,simplisticassumptionsabout the efficacyof differentland ethics in differentculturesshould be avoided. Latndethicsrepresentideatls,whetherexplicitlyincorporatted in reatltn,or tratnstnitted or costnolo,gicatl in thephilosophicgl sphere.The cognitivedimenatndseculatr prat,gvnattic a vnore sions of the Mediterraneanagrosystemare revealedin an explicit, didactic literature represented by agronomic worksthat span some 2500 years(Butzer 1993b, 1994). The concept of "good husbandry"pervadessuch GraecoRoman,medieval,and Islamicwritings,not as an ideology, but as a matter of economic rationaleand proper civic behavior.Judgementis best based on an assessmentof performance(Tuan 1968), ratherthan on simplistic,deductiveinferences. It requiresno emphasisthat these concludingobservations representtestable hypothesesthat can be reformulatedto matchotherregionaland diachroniccontexts,and fallwell withinthe interestsof manyarchaeologists.This is where I see the productiveintersectionbetweenarchaeology and ecology in the long run. Archaeologistsare uniquelyequippedto documentand model not only longterm settlement and land-usehistories,but also to place these in the broaderintellectualperspectivethey deserve. Acknowledgments I remaindeeplyindebtedto JuanF. Mateu(Universidad de Valencia)for the manyinsightshe providedinto Mediterraneanagricultureand cuisine over manyyearsof collaborationin the field. My interestsin long-termsustainabilitywere refinedas a resultof the symposiumArqueolomediterratneo, >ffz:Lathuellatdel hombreen el ecosistemat organized by BernardoMarti Oliver and myself, at the UniversidadInternacionalMenendez Pelayo in Valencia, July 1-5, 1991. I have also profitedfrom repeateddiscussions with my students CharlesD. Frederick,Carlos E. Cordova,and JamesT. Abbott, currentlycompletingdissertationson long-term ecologicalchange in Mexico and Italy.More indirectly,I havebeen affectedby the novelsof MarcelPagnol that so vividlycapturethe ethos of persistence in the rural Mediterraneanworld. An abbreviated version of this paper was given at the PlenarySession, Society for ArnericanArchaeologyMeeting in Anaheim, CA, April1994. ICatrlW Butzer (D.Sc., Universityof Bonn, Gertnatny, 1957) is R. C. DicksonCentenniatlProfessorof Liberatl Arts attthe Universityof Texgs,Austin. His currentre- atnd cult?ratlecolo,gy, interestsatre,geoatrchateolo,gy, seatrch atratndhistoricatl doc?vnentattion the interfatceof atrchivatl MatilMexico. coloniatl atnd Spatin vnedievatl in chateolo,gy Universityof of Geo,gratphy, Depatrtvnent in,gatddress: 78712. TX Austin, Texats, Abrams, Elliot M., and David J. Rue 1988 "The Causes and Consequences of Deforestation among 16: 377-395. the PrehistoricMaya,"HHmolnEcolo,gy Baruch, Uri 1986 "The Late Holocene VegetationalHistory of Lake Kinneret (Sea of Galilee), Israel,"Polleorient12 (2): 37-48. Bell, Martin, and John Boardman,eds. olnd Geo1992 Polst olnd Present Soil Erosion:Archoleolo,gicoll Oxford:Oxbow Monograph22. Perspectives. ,grolphicoll Bernaldez,Fernando Gonzalez 1991 "EcologicalConsequencesofthe Abandonmentof Traditional Land Use Systems in Central Spain," Options 15: 23-29. Mediterrolneennes 1995 "Western MediterraneanLand-use Systems as Antecedents for Semiarid America," in B. L. Turner and A. Gomez Sal, eds., GlobollLolnd Use Choln,ge.Madrid: Consejo Superior de Investigaciones Cientificas, 131149. Birks, H. H., H. J. B. Birks, P. E. Kaland,and D. Moe, eds. Polst,Present,olndFHtre. Cam1988 TheCHltroll Lolndscolpe: bridge:CambridgeUniversityPress. Bottema, Sytze 1984 "PalynologicalInvestigationsin Greece with SpecialReference to Pollen as an Indicator of Human Activity," 24 (1982): 257-289. Polloleohistoriol Bottema, Sytze, G. Enqes-Nieborg, and Willem van Zeist, eds. 1990 Moln'sRole in the Sholpin,gof the EasternMediterrolneoln Rotterdam:Balkema. Lolndscolpe. Boyazoglu, Jean, and Jean-ClaudeFlamant 1992 "MediterraneanSystemsof Animal Production,"in J. G. Galatyand D. L. Johnson, eds., The Worldof Polstorollism: Perspective.New York: Herdin,gSystemsin Compolroltive GuildfordPress, 353-393. Bruckner,Helmut 1986 "Man'sImpacton the Evolutionof the PhysicalEnvironment in the MediterraneanRegion in HistoricalTimes," GeoJornoll13: 7-17. Burford,Alison 1993 Lolnd olnd Lolborin the Greek World.Baltimore: Johns Hopkins UniversityPress. Butzer, KarlW. 1988a "Cattle and Sheep from Old to New Spain: Historical Antecedents," Annolls,Associoltionof AmericolnGeo,grol phers78: 29-56. 1988b "Diffusion, Adaptation, and Evolution of the Spanish Agrosystem,"in P. J. Hugill and D. B. Dickson, eds., The of IdeolsolndMolteriollCHlTrolnsferolnd Trolnsformoltion tre. College Station, TX: Texas A&M UniversityPress, 91-109. 148 Esolo,gyin the Lon,gView/Butser 1990 "A Human Ecosystem Frameworkfor Archaeology,"in Dimen, Muriel, and ErnestineFriedl, eds. E. F. Moran, ed., TheEcosystem Approolch in Anthropolo,gy. 1976 Re,gionollVolrioltion in ModernGreeceolndCyprHs:Towoltd Ann Arbor:Universityof Michigan Press,91-130. a Perspective on the Ethno,grolphy of Greece.Annolls, Nexv YorkAcoldemyof Sciences,268. New York. 1992a "Ethno-agricultureand CulturalEcology in Mexico: Historical Vistas and Modern Implications," Benchmolrk Doolittle, WilliamE. 1990, Conferenceof LoltinAmericolnistGeo,grolphers 171992 "Agriculturein North Americaon the Eve of Contact,"in 18: 139-151. K. W. Butzer, ed., TheAmericolsbeforeolnd olfter1492. CambridgeMA: Blackwell,386401. 1992b "The Americasbefore and after 1492: An Introductionto CurrentGeographyResearch,"in K. W. Butzer, ed., The Flenley,J. R., S. M. King, J. T. Teller, M. E. Prentice, J. Jackson, Americolsbeforeolndafter 1492. Cambridge,MA: Blackand C. Chew well, 345-368. 1991 "The Late QuaternaryVegetationaland ClimaticHistory of Easter Island,"Jornoll of QHolternolty Science6: 851992c "SpanishConquest Society in the New World:Ecological 115. Readaptation and Cultural Transformation,"in S. T. Wong, ed., Person,Plolce,Thin,g.Baton Rouge, LA:GeosGarnsey,Peter cience and Man, 211-242. 1988 Folmineolnd Food SHpplyin the Groleco-Romoln World. Cambridge: Cambridge University Press. 1993a "No Eden in the New World,"Noltre 362: 15-17. Garnsey,Peter, and C. R. Whittaker 1993b "The ClassicalTraditionof Agronomic Science: Perspec1983 Troldeolnd Folminein Clolssicoll AntiqHity.Cambridge: tives on CarolingianAgricultureand Agronomy,"in P. L. CambridgePhilologicalSociety. Butzer and D. Lohrmann, eds., Sciencein Westernolnd EolsternCivilizoltion.Basel:Birkhauser,539-596. Gonzalez Quintero, Luis Fuentes Mata, and MariaFuentes Mata 1980 "E1Holoceno de la porcion centralde la Cuencadel Valle 1993c "Arqueologiaurbana:Intervencion humana y respuesta de Mexico," in F. Sanchez, ed., III ColoqHiosobrePolambiental,"in F. Burillo, ed., Procesospostdeposicionolles. leobotoinicol y Pollinolo,gtol. Mexico City: Instituto Nacional Teruel:ArqueologiaEspacial16-17: 7-18. de Antropologiae Historia, 1 13-132. 1994 "The Islamic Traditions of Agroecology: Cross-cultural Goor, Asaph, and Max Nurock Experience,Ideas, and Innovation,"EcHmene: Jornoll of 1968 TheFrHitsof theHolyLolnd.Jerusalem:IsraelUniversities Environment,CHltre, Meolnin,g1: 7-50. Press. 1995 "BiologicalTransfer,AgriculturalChange, and EnvironHalstead, Paul mental Implications of 1492," in R. R. Duncan, ed., 1987 "Traditionaland Ancient Rural Economy in MediterraInternoltional GermplolsmTrolnsfer:Polst olnd Present. nean Europe,"Jornoll of HellenicStHdies107: 77-87. Madison,WI: Crop Science Society of America,3-29. Butzer, KarlW., and ElisabethK. Butzer 1993 "The Sixteenth-century Environment of the Central Mexican Bajio: Archival Reconstruction from Colonial Land Grants,"in K. Mathewson, ed., CHltre, Plolce,olnd Form.Baton Rouge, LA: Geoscience and Man, 89-124. 1995 "Transferof the MediterraneanLivestock Economy to New Spain:Adaptationand EcologicalConsequences,"in B. L. Turner and A. Gomez Sal, eds., GlobollLolnd Use Choln,ge.Madrid: Consejo Superior de Investigaciones Cientificas,151-193. Heltzer, M., and D. Eitam, eds. 1987 Olive Oil in AntiqHity.Haifa: Culture and Art Division, Ministryof Educationand Culture. Jahns, Susanne 1993 "On the Holocene Vegetation of the Argive Plain (Peloponnese,SouthernGreece)," Ve,getoltionHistoryolnd Archoleobotolny2: 187-203. Jashemski,WilhelminaF. 1979 TheGolvdens of Pompeii,HercHlolneHm, olndthe VillolsDestroyedby VesHviHs. New Rochelle, NY: Caratzas. Byrne, Roger, and Sally P. Horn 1989 "PrehistoricAgricultureand Forest Clearancein the Sierra de las Tuxtlas, Veracruz, Mexico," Pollynolo,gy 13: 181-193. Keller,Donald R., and David W. Rupp, eds. 1983 Archoleolo,gicoll Srvey in the Mediterrolneoln Areol.BAR Internoltionoll Series155. Oxford:B.A.R. Cherry,John F., Jack L. Davis, and Eleni Mantzourani 1991 Lolndscolpe Archoleolo,gy OlSLon,g-termHistory: Northern ICeosin the CycloldicIslolndsfFom the EolrliestSettlement HntilModernTimes.Los Angeles: Universityof California at Los Angeles, Institute of Archaeology. Kirch,PatrickV., J. R. Flenley,D. W. Steadman,F. Lamont, and S. Dawson 1992 "Ancient EnvironmentalDegradation: Prehistoric Human Impacts on an Island Ecosystem:Mangaia,Central Polynesia," Noltional Geo,grolphic Reseolrcholnd Exploroltion 8: 166-179. Delaigue, Marie-Christine 1988 Colpileirol,Villol,ge AndolloHs:Un holbitoltmontol,gnoltd a toitsplolts.BAR InternationalSeries466. Oxford:B.A.R. Klein, Julius 1920 The Mestol:A StHdyin SpolvishEconomicHistory, 12731836. Cambridge,MA: HarvardUniversityPress. Denevan, WilliamM. 1992 "The Pristine Myth: The Landscapeof the Americasin 1492," in K. W. Butzer,ed., TheAmericolsbeforeolndolfter 1492. Cambridge,MA: Blackwell,369-385. Kohler-Rollefson,Ilse, and Gary O. Rollefson 1990 "The Impact of Neolithic SubsistenceStrategieson the Environment: The Case of Ain Ghazal, Jordan," in S. Bottema, G. Entjes-Nieborg and W. van Zeist, eds., Jornal of Field Arshaeolo,gy/Vol. 23, 1996 149 Man's Role in the Shapin,gof the Eastern Mediterraneoln Lolndscolpe. Rotterdam:Balkema,3-14. Kuniholm, Peter I., and Cecil L. Striker 1987 "DendrochronologicalInvestigation in the Aegean and Neighboring Regions, 1983-1986," JoHrnollof Field Archoleolo,gy 14: 385-398. Lamb, H. F., U. Eicher, and V. R. Switsur 1989 "An 18,000 year Record of Vegetation, Lake-leveland Climatic Change from Tigalmamine,Middle Atlas, Morocco," JoHrnollof Bio,geo,grolphy 16: 65-74. Le Houerou, H. N. 1987 "VegetationWildfiresin the MediterraneanBasin:Evolution and Trends,"Ecolo,giol Mediterrolneol 13: 13-24. Reille, Maurice, and John J. Lowe 1993 "A Re-evaluationof the VegetationHistory of the Eastern Pyrenees (France)," QHolternolry ScienceReviews12: 4777. Renfrew,Colin, and Malcolm Wagstaff,eds. 1982 An IslolndPolity:TheArcholeolo,gy of Exploitoltionon Melos. Cambridge:CambridgeUniversityPress. Rice, Don S., Prudence M. Rice, and EdwardS. Deevey 1985 "ParadiseLost: ClassicMaya Impact on a LacustrineEnvironment," in M. Pohl, ed., PrehistoricLowlolndMolyo Environmentand SHbsistence Economy.Cambridge,MA: Harvard University, Peabody Museum of Archaeology and Ethnology,91-105. Menendez Amor, Josefa, and Franz Florschutz 1961 "Contribucional conocimiento de la historiade la vegetacion en Espana durante el Quaternario," EstHdios Geolo,gicos 17: 83-99. Ritchie, J. C. 1984 "Analysepollinique de sediments Holocenes superieures des hauts plateauxdu Maghreb Oriental,"Pollenet Spores 26: 489-496. McDade, Lucinda, ed. 1994 La Selvol:Ecolo,gyolnd NoltHrollHistory of a Neotropicoll Rolinforest.Chicago:Universityof Chicago Press. Rue, David J. 1989 "ArchaicMiddle American Agriculture and Settlement: Recent Pollen Data from Honduras," JoHrnollof Field Archoleolo,gy 16: 177-184. Melville, Elinor G. K. 1990 "Environmental and Social Change in the Valle del Mezquital, Mexico, 1521-1600," Compolroltive StHdiesin Societyand History32: 24-53. Runnels, Curtis N., and Julie Hansen 1987 "The Olive in the PrehistoricAegean: The Evidence for Domestication in the EarlyBronze Age," OxfordJoHrnal of Archoleolo,gy 5: 299-308. Miller, Naomi F. Sale, Kirkpatrick 1991 "The Near East," in W. van Zeist, K. Wasylikowa,and 1990 The ConqHestof Polroldise: ChristopherColHmbHs olnd the K. E. Behre, eds., Pro,gressin Old WorldPolloleoethnoboColHmbioln Le,golcy. New York:Knopf. tolny.Rotterdam:Balkema,133-160. Montoya Oliver, Jose Miguel 1983 Polstorollismomediterroineo. Madrid: Ministerio de Agricultura,Instituto Nacionalparala Conservacionde la Naturaleza25. O'Hara, SarahL., F. Alayne Street-Perrott,and Timothy P. Burt 1993 "AcceleratedSoil Erosion around a Mexican Highland Lake Caused by Prehistoric Agriculture," NoltHre362: 48-51. Sauer, Carl O. 1966 TheEolrlySpolnish Molin.Berkeley:Universityof California Press. Shetler, Sidney 1991 "Three Faces of Eden," in H. J. Viola and C. Margolis, eds., Seedsof Choln,ge.Washington:Smithsonian Institution Press,225-247. Osborne, Robin 1987 Clolssicoll Lolndscolpe with Fi,gHres: TheAncient GreekCity olndIts CoHntryside. London: George Philip. Simpson, Leslie B. 1952 Exploitoltionof Lolvd in CentrollMexico in the Sixteenth CentHry.Ibero-Americolnol 36. Berkeley: University of CaliforniaPress. Piperno, Dorothy R., Mark B. Bush, and Paul A. Colinvaux 1991 "Paleoecological Perspectiveson Human Adaptation in Central Panama, II. The Holocene," Geoolrcheolo,gy 6: 227-250. Snodgrass,Anthony M. 1987 An Archoleolo,gy of Greece:The PresentStolteolnd FHtHre Scopeof a Discipline. Berkeley:University of California Press. Reille, Maurice 1976 "Analyse pollinique de sediments postglaciaires des le Moyen Atlas et le Haut Atlas marocains,"Ecolo,giol Mediterraneol2: 153-170. Spurr,M. S. 1986 ArablecHltivationin RomolnItollyc. 200 B.C.-c. A.D. 100. London: Society for the Promotion of Roman Studies. 1977 "Contributionpollenanalytiquea l'histoireholocene de la vegetation des montagnes du Rif (Maroc Septentrional)," Bulletin, AssociationFranise Pour l'Etude du Quaternaire, Supplement50: 53-76. 1979 "Analyse pollinique du lac de Sidi Bou Rhaba, littoral atlantique(Maroc)," Ecolo,giaMediterranea4: 61-65. 1984 "Origine de la vegetation actuelle de la Corse sud-orientale," Pollenet Spores26: 43-60. Stager, LawrenceE. 1985 "The First Fruits of Civilization," in J. N. Tubb, ed., Palestinein theBronzeand IronA,ge.London: Instituteof ArchaeologyMonograph, 172-188. Stevenson, A. C., and R. J. Harrison 1992 "AncientForestsin Spain:A Model for Land-useand Dry Forest Management in South-west Spain from 4000 BC to 1900 AD," Proceedin,gsof the PrehistoricSociety58: 227-247. 150 Ecolo,gyin the Lon,gView/Butzer Tomaselli, Ruggero 1977 "The Degradationofthe MediterraneanMaquis,"Ambio 6: 356-362. Trabaud,L. 1991 "Is Fire an Agent FavouringPlant Invasions?"in R. H. of MediterrGroves and F. diCastri, eds., Bio,geo,graphy nean Invasions.Cambridge:CambridgeUniversityPress, 179-190. Williams,BarbaraJ. 1989 "Contact Period Rural Overpopulationin the Basin of Mexico: Carrying-capacityModels Tested with DocumentaryData," AmericanAntiqHit,v54: 715-732. Zangger, Eberhard 1991 "PrehistoricCoastal Environmentsin Greece: The Vanished Landscapesof Dimini Bayand LakeLerna,"JoHrnal 18: 1-15. of FieldArchaeolo,gy Tuan, Yi-Fu 1968 "Discrepanciesbetween EnvironmentalAttitude and Behavior: Examples from Europe and China," Canadian 12: 176-189. Geo,grapher Zohary, Daniel, and MariaHopf 1993 Domesticationof Plants in the Old World:TheOri,ginsand Spreadof CHltivatedPlants in WestAsia, EHropeand the Nile Valley.Oxford:ClarendonPress,2nd ed. van Andel, Tjeerd, and Curtis Runnels 1987 BeyondtheAcropolis:A RHral GreekPast. Stanford:Stanford UniversityPress. Zuckerman,Michael 1987 "Identity in British America: Unease in Eden," in N. Canny and A. Pagden, eds., Colonial Identit,vin the Atlantic World.Princeton: Princeton University Press, 1 15-158. van Andel, Tjeerd, EberhardZangger, and Anne Demitrack 1990 "LandUse and Soil Erosion in Prehistoricand Historical 17: 379-396. Greece,"Journal of FieldArchaeolo,gy van Zeist, Willem, and Sytze Bottema in the Near East.Wiesbaden: 1991 Late Qternary Ve,getation Beihefte zum TubingerAtlas des VorderenOrients,A18. Vassberg,David E. 1984 Land and Societyin GoldenA,geCastile.New York:Cambridge UniversityPress. Vaughan, Hague H., EdwardS.Deevey, and S. E. Garrett-Jones 1985 "Pollen Stratigraphyof Two Cores from the Peten Lake District,"in M. Pohl, ed., PrehistoricLowlandMayaEnviEconomy.CambridgeMA: Harronmentand SHbsistence ard University, Peabody Museum of Archaeology and Ethnology,73-89. Vernet, Jean-Louis,E. Badal Garcia,and E. GrauAlmero 1983 "La vegetation neolithique du sud-est de l'Espagne (Valencia, Alicante) d'apres l'analyse anthracologique," ComptesRendHes,AcademiedesSciencesde Paris III-296: 669-672. Vernet, Jean-Louis,StephanieThiebault, and Christine Heinz 1987 "Nouvelles donnees sur la vegetation prehistoriquepostglaciaire mediterraneenne d'apres l'analyse anthracologique," in PremieresCommHntes Paysannesen Mediterranee Occidentale.Paris:Centre National de la Recherche Scientifique,87-94. Wells, Berit, Curtis Runnels, and EberhardZangger 1990 "The Berbati-LimnesArchaeologicalSurvey: The 1988 Season," OpuscHlaAtheniensia18: 207-238. Wells, Berit, ed. 1992 A,gricultHrein Ancient Greece.Stockholm: SkrifterUtgivna av SvenskaInstitutet i Athen,4 Series,42. White, Kevin D. 1970 Roman Farmin,g.Ithaca,N.Y.: Cornell UniversityPress. Whitmore, Thomas M., and B. L. Turner, II 1992 "Landscapesof Cultivationin Mesoamericaon the Eve of the Conquest," in K. W. Butzer, ed., TheAmericasbefore and after 1492. Cambridge,MA: Blackwell,402-425. Whittaker,C. R., ed. 1988 Pastoral Economiesin Clgssical AntiqHity. Cambridge: CambridgePhilologicalSociety.