Forests, fields, and the edge of sustainability at the ancient Maya city of Tikal

Forests, fields, and the edge of sustainability at theancient Maya city of TikalDavid L. Lentza,1, Nicholas P. Dunningb, Vernon L. Scarboroughc, Kevin S. Mageeb, Kim M. Thompsona, Eric Weaverb,Christopher Carrb, Richard E. Terryd, Gerald Islebee, Kenneth B. Tankersleyc, Liwy Grazioso Sierraf, John G. Jonesg,Palma Buttlesh, Fred Valdezi, and Carmen E. Ramos Hernandezj

aDepartment of Biological Sciences, bDepartment of Geography, and cDepartment of Anthropology, University of Cincinnati, Cincinnati, OH 45221;dDepartment of Plant and Wildlife Sciences, Brigham Young University, Provo, UT 84602; eEl Colegio de la Frontera Sur, Unidad Chetumal Herbario, Chetumal,AP 424 Quintana Roo, Mexico; fLa Escuela de Historia, Universidad de San Carlos de Guatemala, Guatemala City, 01012 Guatemala; gArchaeologicalConsulting Services, Ltd., Tempe, AZ 85282; hSoftware Engineering Institute, Carnegie Mellon University, Pittsburgh, PA 15213; iDepartment of Anthropology,University of Texas, Austin, TX 78712; and jDepartamento de Monumentos Prehispanicos, Instituto de Antropología e Historia de Guatemala, Guatemala City,01001 Guatemala

Edited by B. L. Turner, Arizona State University, Tempe, AZ, and approved November 7, 2014 (received for review May 9, 2014)

Tikal has long been viewed as one of the leading polities of theancient Maya realm, yet how the city was able to maintain itssubstantial population in the midst of a tropical forest environ-ment has been a topic of unresolved debate among researchers fordecades. We present ecological, paleoethnobotanical, hydraulic,remote sensing, edaphic, and isotopic evidence that reveals howthe Late Classic Maya at Tikal practiced intensive forms of agriculture(including irrigation, terrace construction, arboriculture, householdgardens, and short fallow swidden) coupled with carefully con-trolled agroforestry and a complex system of water retention andredistribution. Empirical evidence is presented to demonstratethat this assiduously managed anthropogenic ecosystem of theClassic period Maya was a landscape optimized in a way thatprovided sustenance to a relatively large population in a pre-industrial, low-density urban community. This landscape produc-tivity optimization, however, came with a heavy cost of reducedenvironmental resiliency and a complete reliance on consistentannual rainfall. Recent speleothem data collected from regionalcaves showed that persistent episodes of unusually low rainfallwere prevalent in the mid-9th century A.D., a time period thatcoincides strikingly with the abandonment of Tikal and theerection of its last dated monument in A.D. 869. The intensifiedresource management strategy used at Tikal—already operatingat the landscape’s carrying capacity—ceased to provide adequatefood, fuel, and drinking water for the Late Classic populace in theface of extended periods of drought. As a result, social disorderand abandonment ensued.

paleoecology | Neotropics | paleoethnobotany | irrigation | root crops

The Late Classic period (A.D. 600–850) was a time of un-precedented architectural, astronomical, and artistic achieve-

ment at Tikal, one of the leading urban centers of the ancient Mayarealm. It was also a time of meteoric population growth at thisbustling cultural center. Notwithstanding its prominence as a majorMaya polity, how Tikal’s leaders and farmers managed to providefood, fuel, and other sustenance for its many occupants has neverbeen fully understood or quantified.To best assess resource potential at Tikal, we first defined an

extraction zone that was extrapolated from archaeological set-tlement data by creating a Voronoi diagram (1, 2) (Fig. 1). Es-sentially, this approach proscribes a proportional boundarybetween Tikal and its surrounding contemporaneous communi-ties: namely, Motul de San Jose, El Zotz, Uaxactún, Xultun, DosAguadas, Nakum, Yaxha, and Ixlu. Using this technique, in-cluding assigning greater economic clout to Tikal using a 2:1weighting scheme (see section on the Voronoi diagram in SIMaterials and Methods), we calculated that its Late Classic re-source extraction zone encompassed ∼1,100 km2. This is the areafrom which the residents of Tikal could obtain their necessaryfood, fuel, construction timbers, and other living essentials.

Superimposing the Voronoi Diagram over satellite images ofmodern Tikal (2, 3) (Fig. 1), which is mostly forested today,reveals that ∼850 km2 is upland tropical forest habitat and 250km2 is seasonal wetland or bajo (4).† Pollen data from LakePetén Itza (5), a deep lake that is downwind and less than 5 kmsouth of the extractive zone of Tikal, suggest that the range offorest clearance was from 60–70% during the Late Classicperiod (LCP). As with all calculations in this report, we usedthe most conservative number when a range of values is available.In this case, we use the 60% figure for the amount of uplandforest cleared during the LCP, leaving ∼340 km2 of forest in-tact. Pollen data from Aguada Vaca de Monte (Fig. S1), a smallpond located in the Bajo Santa Fe with a pollen content morereflective of the surrounding bajo, suggested that 37–32% ofthe bajo lands (80 km2 using the figure of 32%) were clearedfor agriculture and another 170 km2 remained as managedseasonal swamp forest (Fig. S2). Details of the pollen analysesare given in the section on palynological data in SI Materialsand Methods.

Significance

The rise of complex societies and sustainable land use associ-ated with urban centers has been a major focus for anthro-pologists, geographers, and ecologists. Here we present aquantitative assessment of the agricultural, agroforestry, andwater management strategies of the inhabitants of theprominent ancient Maya city of Tikal, and how their land usepractices effectively sustained a low-density urban populationfor many centuries. Our findings also reveal, however, that theproductive landscape surrounding Tikal, managed to the brinkof its carrying capacity during the Late Classic period, did nothave the resilience to withstand the droughts of the 9th cen-tury. These results offer essential insights that address thequestion of why some cities thrive while others decline.

Author contributions: D.L.L., N.P.D., and V.L.S. designed research; D.L.L., N.P.D., V.L.S.,K.S.M., K.M.T., E.W., C.C., R.E.T., L.G.S., J.G.J., P.B., F.V., and C.E.R.H. performed research;D.L.L., N.P.D., V.L.S., K.S.M., K.M.T., E.W., C.C., R.E.T., G.I., K.B.T., L.G.S., J.G.J., P.B., F.V.,and C.E.R.H. analyzed data; and D.L.L., N.P.D., V.L.S., and K.S.M. wrote the paper.

The authors declare no conflict of interest.

This article is a PNAS Direct Submission.1To whom correspondence should be addressed. Email: david.lentz@uc.edu.

This article contains supporting information online at www.pnas.org/lookup/suppl/doi:10.1073/pnas.1408631111/-/DCSupplemental.

†We are fully aware that the forest structure of the northern Petén is more complex thanthis simple dichotomy. Other studies of the area have subdivided the forest differently,for example Ford’s survey of the Tikal–Yaxha transect divided the landscape into threeclasses (4). For our purposes of calculating standing biomass, we surveyed the forestsextensively and bifurcated the forest cover into two classes (upland and bajo) that couldbe readily discerned on Landsat Enhanced Thematic Mapper Plus (ETM+) imagery.

www.pnas.org/cgi/doi/10.1073/pnas.1408631111 PNAS Early Edition | 1 of 6

AINABILITY

SCIENCE

AgroforestryBecause forests supplied essential resources, such as fuel (TableS1), construction material, habitat for game, wild plant foods,and a pharmacopoeia from medicinal species, the study of pastmethods of agroforestry is crucial to understanding the economicunderpinning of the ancient Maya. As part of this study, aboveground biomass (AGB) survey transects of modern forests werecombined with Landsat 7 multispectral imagery and statisticallymodeled to yield biomass estimates for the Tikal region. TheAGB of modern upland forest was calculated at 28.9 ± 2.6million kg·km−2 (SEM) and 18.2 ± 0.523 million kg·km−2 (SEM)for bajo forest. If we use these data to estimate the AGB of theLCP Tikal forest, with forest extent determined by pollen data,we find that 9.8 billion kg of wood would have been presentin the uplands (on 340 km2 of forested land) and 3.2 billion kgin the bajos (on 170 km2) during the LCP. It is reasonable to usethe modern Tikal forest as an analog for the ancient forest be-cause the archaeological wood assemblage (Table S2) showeda remarkable similarity to the oligarchic species of the modernforest (Table S3). See the sections on modern forest surveys,biomass calculations, and ancient wood use in SI Materials andMethods for expanded discussions.If the ancient forests were managed on a sustainable basis, and

there is ancient tree-ring evidence that the Maya attempted to doso (6), then the Tikal occupants could only have harvested theannual growth increment each year. To calculate the annualgrowth rate of Tikal forests, we extrapolated from the results ofa 10-y study (7) of a 50-ha plot in a similar type of CentralAmerican moist tropical forest on Barro Colorado Island, Pan-ama (BCI), where it was recorded that the biomass change ratewas 0.55 Mg·ha−1·yr−1. Because the modern Tikal forest hada larger basal area (39 m2·ha−1) than the BCI forest (28 m2·ha−1),a proportional adjustment to 0.76 Mg·ha−1·yr−1 was performed.Using these forest growth data, we calculated that the amount ofsustainably usable wood on an annual basis during the LCPwould have averaged 26 million kg·yr−1 in the uplands and 13million kg·yr−1 in the bajos for a total of 39 million kg·yr−1 (8)(Table 1). Justification for using forest growth data from BCI can

be found in the section on annual growth increment calculationsin SI Materials and Methods.By far the heaviest demand on the forest was firewood needed

for cooking. All of the major foods of the ancient Maya, espe-cially beans, root crops, and to a lesser degree, maize had to becooked before consumption. This requirement created a dailyand inexorable fuel need for all of the Tikal inhabitants. Thefiring of ceramics also required substantial amounts of fuel.Studies at well-preserved Maya archaeological sites (9, 10)revealed that each household, from the most humble to the elite,possessed from 70 to 80 ceramic vessels at any given time. Inaddition, the use life of a ceramic pot was only about 1 y onaverage (11, 12). To make matters worse, the kilns used by theTikal Maya were inefficient (13), even by preindustrial standards,and required about 5.2 kg of fuelwood per vessel (14). Thus,keeping the city supplied with pottery, not to mention the pos-sibility of export production, created a heavy demand for fuel.The production of lime (calcium oxide), an essential componentof plaster, also required considerable fuel input; it was made byburning crushed limestone and required 5 kg of wood to make1 kg of lime (15). All of the temples, plazas, causeways (sacbeob),reservoirs, and elite houses were covered with plaster and, al-though this was not a daily need, in the long run the processconsumed a substantial amount of fuel. Wood required forconstruction timbers and artifact manufacture created an es-sential, but less voluminous demand. Details of these calcu-lations are provided in the section on ancient wood use in SIMaterials and Methods and the results are summarized in Table 1and Table S1.The estimated wood quantities required annually for the main-

tenance of LCP populations at Tikal for fuel and construction was42 million kg·yr−1, approximately equal to the amount of woodavailable on a sustainable basis (39 million kg·yr−1) from the Tikalupland and bajo forests. The Maya could have compensated forany shortages in forest productivity through the importation ofpine wood (see section on ancient wood use in SI Materials andMethods) and intensive forestry techniques applied to a fixed-plot agroforestry system. [Evidence from burned wood retrieved

Fig. 1. Tikal extraction zones. The Voronoi cellequal to the extractive zone during the LCP is out-lined in cyan and in yellow for the Late Preclassicperiod. The built environment of the city, the por-tion that contained the majority of houses, temples,ball courts, and other structures during the LCP (inred) was determined by archaeological survey (3).The black square outlines the 9 km2 of the cere-monial core mapped by Carr and Hazard (2). Notethat of the eight neighboring Maya polities used tocreate the LCP Voronoi diagram, Xultun is not lis-ted. It is outside the borders of this map to thenortheast of the Voronoi diagram.

2 of 6 | www.pnas.org/cgi/doi/10.1073/pnas.1408631111 Lentz et al.

from middens and structures at Tikal indicate that forests werebeing managed as fixed plots or woodlots. Most (90%) of thearchaeological charcoal samples examined (n = 421) had parallelrays when viewed in transverse section, indicating that they werefrom mature trees, not saplings or branches, which have con-vergent rays, a condition that would be expected in young secondgrowth forest. Only 3% of archaeological wood samples had raysthat were clearly convergent (the other 7% were classified as in-determinate). Accordingly, the Tikal Maya appear to have beenharvesting mature trees for fuel and construction purposes.]

AgricultureThe ancient Maya agricultural practices at Tikal can be inter-preted with fresh and previously unavailable insights by using theextraordinarily well-preserved Late Classic village of Cerén (9)as a comparative model. Although Tikal operated on a muchlarger scale, the comparison is valid because both communitiesused essentially the same array of crops (Table S2), and theyoccupied land with similar native vegetation and elevation. As atCerén, the built environment of Tikal (160 km2) would havebeen cleared of its native forest cover and planted with dooryardgardens. Crops used at Tikal included maize (Zea mays L.),three species of beans (Phaseolus spp.), two species of squash(Cucurbita spp.), and several species of root crops, includingsweet potato (Ipomoea batatas [L.] Lam.) (16) (Fig. 2), achira(Canna cf. indica L.) (Fig. S3), and malanga (Xanthosoma sag-ittifolium [L.] Schott.). Manioc has not been identified at Tikal,but this important root crop has been discovered at other nearbysites (17, 18), so it seems quite likely that the farmers of Tikalhad access to it as well. Paleoethnobotanical remains of fruittrees from Tikal also parallel the Cerén model, with orchards ofcoyol (Acrocomia aculeata Lodd. ex. Mart), sapote (Pouteriasapota [Jacq.] H. E. Moore & Stearn), jocote (Spondias cf pur-purea L.), nance (Byrsonima crassifolia [L.] H.B.K.), avocado (Per-sea americana Mill.), and curiously, cacao (Theobroma cacao L.)(19). [Cacao was an important product to the ancient Maya andthe iconographic references to it at Tikal and elsewhere arenumerous. Yet it is believed (19) that cacao, which is intolerantof long periods of dryness, could not have survived the long, hotdry season at Tikal. Evidently, this supposition was untrue be-cause among the archaeological sediments at Tikal, we not onlyfound the seeds of cacao (which easily could have been impor-ted), but we also recovered the burned wood of cacao (Fig. S4),which would not have been imported. Two of the shade treesthat are often grown with cacao, Gliricidia sepium (Jacq.) Steud.,and Erythrina spp. were also found among the archaeologicalplant remains at Tikal. These trees are leguminous symbionts,traditionally associated with cacao production, which not onlyprovide cover for the shade-loving cacao, but bear nitrogen-fixing bacteria in their root nodules that help to fertilize the soilas well. Cacao and its symbionts were probably grown in specialareas, such as artificial sinkholes or rejolladas, where they wereprotected from the heat and could be watered during the dry

season.] As at Cerén, these orchards would have been planted ad-jacent to household compounds that were widely spaced at Tikal.Relying on Cerén as an interpretive model, intensively man-

aged fields of maize, other seed crops, and root crops (Table S2)that provided most of the calories for the city were found outsideof the built environment of Tikal (the zone in red in Fig. 1).Previous studies (20) of Petén farmers recorded that an averageof 0.18 ha was required to feed an individual using traditionalfarming techniques. If we apply this figure to the LCP Maya,then Tikal required ∼80 km2 of fertile land per year to feed itself(0.0018 km2 × 45,000 inhabitants) (8). Looking at the uplandareas only, and assuming an intensive, open field style of agri-culture, the Tikal Maya would have had enough land for 1 y ofplanting with about 270 km2 of vacant, cleared land in any givenyear, enough for about 3 y of fallow. A short fallow system asdescribed by Sanders (21) in a 1:3 ratio of planting to fallowyears or a ratio of 3:5 as described by Griffin (22) would havebeen feasible within the confines of available upland plantingsurfaces. Killion’s (23) observations of intensive, fixed-plot ag-riculture with short fallow periods in southern Veracruz seem toclosely match the projected conditions at LCP Tikal.Some scholars (24) have stated that this kind of short fallow

system would not have been sustainable. Others (25), however,have observed that alternative long fallow systems would nothave been possible in the Late Classic Maya Lowlands; there justwas not enough land. Short fallow systems could have been used

Table 1. Wood availability

Aboveground biomass Uplands Bajos

Modern AGB 28.9 ± 2.60 × 106 kg·km−2 18.2 ± 0.523 million kg·km−2

LCP Tikal total area 850 km2 250 km2

Forest cover (LCP) 340 km2 170 km2

Forest AGB (LCP) 9.8 × 109 kg 3.2 × 109 kgSustainably usable AGB (LCP) 26 × 106 kg·yr−1 13 × 106 kg·yr−1

Total annual usable AGB (LCP) 39 × 106 kg·yr−1

Total annual wood need (LCP) 42 × 106 kg·yr−1

Results are based on modern vegetation surveys coupled with pollen data to determine the forest extentduring the LCP. Wood use requirements for LCP Tikal are based on the estimate of 45,000 occupants (8).

Fig. 2. Electron micrograph of a burned sweet potato (Ipomoea batatas).Characteristically in sweet potato roots, the tissue organization is disruptedby the formation of radially oriented cavities (as seen in this image) whencarbonized (16). This sample from Middle Preclassic Tikal represents the firstevidence for sweet potato in the ancient Maya Lowlands. It was acceleratormass spectrometry radiocarbon dated to cal 640 ± 30 B.C. (SD).

Lentz et al. PNAS Early Edition | 3 of 6

AINABILITY

SCIENCE

if managed carefully with yields declining slowly over time (26).To facilitate production in a short fallow system, Tikal farmerscultivated at least three species of leguminous annual crops (TableS2). A common practice among modern Maya farmers is to plantfields with multiple crops, including nitrogen-fixing cultigens andmaize. Tikal farmers, because they had all a full array of nitrogen-fixing crops, likely used this same technique. These leguminousplants provide accessible nitrogen in soil and help prevent yielddeclines common in short fallow agricultural regimes.Phosphorus, however, is a critical element often in short

supply across the Maya Lowlands (27). When forest cover isremoved, the topsoil is gradually starved of this essential nu-trient (28), an outcome with potentially catastrophic conse-quences for ancient Maya farmers (29, 30). Fortunately for thepeople of Tikal, soils may have been naturally renewed via thecapture of windblown soot, dust, and volcanic ash. Analyses byX-ray diffraction and X-ray florescence of reservoir sediments

from Tikal demonstrate that there were sporadic but substantialinputs of aeolian volcanogenic materials deposited throughoutthe Classic period (31). This regular deposition of volcanic ashmay well have served to ameleorate exausted agricultural soilsat Tikal and helps to explain how intensive farming practicescould have succceeded for extended periods of time withoutrapidly declining yields.To be sure, the uplands were not the only areas cultivated;

there were also agricultural activities in the seasonal wetlands.Pollen data from Aguada de Terminos (Fig. S3) reveal thatmaize and achira (Canna cf. indica L.) were grown in that sectorof Bajo Santa Fe, demonstrating that portions of the bajos, es-pecially the margins, were active areas of intensive agriculture.Chert-lined terraces, created to prepare planting beds and pre-vent soil from eroding down the hillslope, were found near theadjacent Terminos plazuela group (Fig. S6) and signify a highlevel of labor investment in bajo agriculture.

Fig. 3. δ13C isotope enrichment. Map showing the location of the Perdido Reservoir and its relationship to the pocket bajo to the south/southwest. Sedi-mentary evidence in the Pocket Bajo (Fig. S5) and unusually high amounts of debris in the Perdido Reservoir indicate that the tank was used for irrigation.Episodes of δ13C enrichment in the pocket bajo are compatible with the interpretation of maize agriculture. The base map is from Carr and Hazard (2),courtesy of the University of Pennsylvania Museum of Archaeology and Anthropology.

In an area with a pronounced annual dry season and no per-manent rivers or lakes in the immediate vicinity, water was ascarce and highly valued resource. Tikal grew as a city arounda system of springs emanating from what is now the TempleReservoir, which sits at the head of a long ravine (32). Throughtime the ravine was blocked off in places with the purpose ofimpounding water that flowed from the springs. As Tikal con-tinued to grow, they built large plazas on either side of the ravineand canted the pitch of the pavements so that seasonal rainwaterwould flow into the reservoirs.This elaborate system designed for water retention gave rise to

another form of intensive agriculture south of the PerdidoReservoir (Fig. 3), where impounded water appears to have beenused to periodically flood a pocket of flat land located just belowthe tank and its egress. One of the smallest of the formal tanksidentified, Perdido, evidently functioned exclusively as a sourceof agricultural irrigation water, as inferred by the debris re-covered from our excavations. Compared with other Tikal res-ervoirs, several times the amount of artifactual waste was foundhaving washed into the Perdido tank. This was not a reservoir thatwas kept clean for potable water as seen elsewhere at the site.An ancient agricultural field locus 100–900 m south/southwest

of the Perdido Reservoir was identified lying stratigraphicallylower than the bottom of the plaster floor of the reservoir andthe bottom of the egress gate. Both the plaster floor of thereservoir and the stratified base of the agricultural field date tosometime between A.D. 350 and 550. The soil profile (Fig. S5)from the former agricultural field was especially telling. Bajo-likesoils were abruptly truncated around A.D. 485 (just prior to thebeginning of the LCP) and then were overlain by stratified al-luvial sediments. We interpret these laminated alluvial strata asevidence for repeated bouts of flooding from the reservoir, os-tensibly for agricultural purposes.Furthermore, δ13C values associated with Classic period land

surfaces and rooting zones in three soil pedons located between500 and 800 m south of the reservoir show clear signs of en-richment (Fig. 3). Soil samples for isotopic analysis were col-lected from Pedon 1 in the Pocket Bajo to a depth of 90 cm andresults revealed two peak shifts in δ13C values compared with thesurface reading: one of 4.3‰ at a depth of 30 cm and 4.7‰ ata depth of 75 cm. This pattern in the C isotope data suggest twoseparate periods of C4-dominated plant cover. Results fromPedon 2, only 45-cm deep, reveal a δ13C shift of 2.7‰, whichis inconclusive for the presence or absence of C4 plants. Pedon 3,however, which is located just to the east of Pedon 2, revealedstrong isotopic evidence of ancient C4 plants with a δ13C en-richment as high as 6.1‰. Pedon 4 was close to house structuresand the upward shift of 3.3‰ in δ13C provided some evidence ofancient C4 plants. Details of the radioisotopic methodology arepresented in the section on stable carbon isotope in SI Materialsand Methods.These data, considered together, indicated that C4 plants were

growing for long periods within the Perdido Pocket Bajo. Be-cause our surveys of the modern vegetation in the pocket bajorecorded no C4 or Crassulacean acid metabolism (CAM) plants,our analyses of archaeological plant remains revealed no C4cultigens (other than maize), and the δ13C results showed evi-dence of C4 enrichment, our interpretation of these multiplestrands of evidence is that the Pocket Bajo was an ancient ag-ricultural field and maize likely was one of those plants culti-vated. (In addition to maize, there are other C4 and CAM plantsin the Neotropics that could have caused the δ13C enrichment. Itis possible that the land was cleared in the past and some weedyC4 plants, such as wild amaranth, sedges, or wild grasses, invadedthe field and caused the δ13C enrichment. In either case, whetherit was maize or some other crop intermingled with weedy C4plants, all of the evidence considered together indicate that theland below Perdido Reservoir was periodically flooded for

agricultural purposes during the Classic period.) Using waterreleased from the Perdido Reservoir during the dry season, theTikal Maya could have double-cropped the area and obtaineda second harvest from this carefully tended pocket of land.Land below the Corriental Reservoir also appears to have

been a likely location for crop irrigation. A switching station,probably used to divert water downstream to arable land surfa-ces, was found at the low end of the reservoir (32). Other loca-tions where irrigation was feasable included areas below theBejucal and Tikal reservoirs, which lie just slightly higher in el-evation than broad flat areas of deep soil. There may have beenother reservoirs at Tikal that were used for irrigation but thesefour, at least, were well situated for it.

DiscussionOne of the conclusions that can be drawn from the evidencegenerated is that the Maya at Tikal were living quite near orperhaps beyond the sustainable carrying capacity of their highlyengineered landscape. Larger population estimates (33–35) thanthe one used in our calculations for LCP Tikal would not havebeen feasible without massive importation of food and fuel fromoutside the defined extraction zone. This importation seemshighly unlikely, given the absence of navigable waterways, draftanimals, or wheeled vehicles (36). Although there is abundantevidence for the long-distance movement of highly valued tradegoods, such as salt (37, 38) and cacao (39), the movement of low-value bulk goods probably was very localized (40).The reservoir system had the positive effect of maximizing and

carefully storing the rainwater that fell on the site core, but hadthe negative effect of cutting off recharge supplies to the springs,once a major attraction to early settlers. During most of the 7thand 8th centuries, precipitation was abundant enough to ac-commodate crops as well as reservoir recharge to provide for theneeds of a growing populace. Although these adaptations wereextremely effective in meeting the short-term demands of pop-ulation growth and increasing levels of social well-being, theunforeseen consequences of the extensive landscape alterationshad tragic results.By the early to mid-9th century (A.D. 820–870), speleothem

evidence indicates that an extremely dry period akin to episodicdrought (41–43) occurred in the central Maya Lowlands. Thismultidecadal drought coincides with the depopulation of Tikal(43). The last dated monument was erected in A.D. 869 (44),when the city was already in its death throes (45). Moreover, thedrought was likely anthropogenically influenced, as there isa growing body of evidence that indicates forest clearance, evenpartial forest clearance, will negatively impact the hydrologiccycle (46–48). In short, the construction of extensive pavementscombined with forest clearance likely exacerbated the effect ofthe drying trend, so by the mid-9th century there were in-adequate supplies of water and food with little resilience left inthe system to adapt to new conditions. As a consequence, thesocial structure of Tikal collapsed, leaving the site core aban-doned with only a tiny relict population huddled around the fewwater holes that did not dry up. Although the focus of our studyhas been the polity of Tikal, what we describe was not an isolatedevent; similar scenarios of human interaction with the environ-ment and climate change played out on a broader scale throughoutmuch of the Central Maya Lowlands at the end of the Classic pe-riod (30, 49).Although some may view this interpretation as environmental

determinism, we argue that the demise of LCP Tikal was aproduct of human agency where a carefully constructed nichewas designed to meet the immediate needs of a burgeoningpopulation. Ultimately, through the long arc of time as climaticpatterns changed, with influences from human activities, theintensified agricultural, hydraulic, and agroforestry systems thatmade the urban condition possible at Tikal reached a tipping

Lentz et al. PNAS Early Edition | 5 of 6

AINABILITY

SCIENCE

point and were unable to meet productivity demands in the faceof reduced precipitation.

Materials and MethodsDuring our 2009 and 2010 field seasons, we surveyedmodern forests at Tikal todetermine the number, size, and diversity of tree species. Forest surveys wereconducted, mostly in 500-m2 rectangular plots, and covered a total of 5.95 hain a variety of forest types. The diameter at breast height, tree height (assessedwith a hand-held clinometer), and species name for each tree over 6 cm di-ameter at breast height were recorded within each plot demarcated witha GPS unit. Voucher specimens were collected for tree species and brought tothe paleoethnobotanical laboratory at the University of Cincinnati for furtheridentification. Vouchers were compared with herbarium specimens at theMargaret H. Fulford Herbarium at the University of Cincinnati (CINC) and theMissouri Botanical Garden Herbarium. For specimens that were difficult toidentify, nuclear and chloroplast DNA was extracted from voucher leaves andsequenced for final identification (48). Sequences were compared with theBasic Alignment Search Tool (BLAST) from the National Center for Bio-technology Information database. Vouchers were housed at the University ofSan Carlos Herbarium and CINC as part of their permanent collections. Biomassof forest tracts was determined using satellite images in combination with

ground surveys. Environmental indices and data transforms were derived fromthe spectral data of a March 2003 Landsat 7 ETM+ image of the study areaobtained from the US Geological Survey Global Visualization Viewer (GLOVIS;glovis.usgs.gov). Excavations and coring procedures focused on the recovery ofpaleoethnobotanical remains and the collection of archaeological sedimentsthrough time. Archaeological wood and other plant remains were examinedwith a combination of light and environmental scanning electron microscopes(Philips XL30 ESEM). Soil samples were analyzed for pollen content, isotopicsignatures (δ13C and 14C), and chemical composition using powder- X-ray dif-fraction and X-ray florescence. Additional details of our research methods canbe found in the SI Materials and Methods.

ACKNOWLEDGMENTS. We thank administrators of the Guatemalan Ministryof Culture and Sports, the Institute of Anthropology and History ofGuatemala, and Tikal National Park for their logistic support; curators ofthe University of Pennsylvania Museum of Archaeology and Anthropologyfor sharing archaeobotanical specimens and other data; P. Sheets, M. Pohl,V. Slotten, L. Lentz, and S. Matter for editorial assistance; and V. Slotten forassisting with processing and imaging paleoethnobotanical remains. Thisstudy was supported in part by National Science Foundation Grant BCS-0810118, Wenner-Gren Foundation Grant 7799, the Alphawood Foundation,the University of Cincinnati, and Brigham Young University.

1. Montmollin OD (1988) Scales of settlement study for complex societies: Analyticalissues from the Classic Maya area. J Field Archaeol 15(2):151–168.

2. Carr RF, Hazard JE (1961) Map of the Ruins of Tikal, El Petén, Guatemala (UniversityMuseum, Univ of Pennsylvania, Philadelphia).

3. Fry RE (2003) The peripheries of Tikal. Tikal: Dynasties, Foreigners and Affairs of State,ed Sabloff JA (SAR Press, Santa Fe, NM), pp 143–170.

4. Ford A (1997) Obsidian procurement and distribution in the Tikal-Yaxha intersite areaof the Central Maya Lowlands. Anc Mesoam 8(1):101–110.

5. Islebe GA, et al. (1996) A Holocene vegetation history from lowland Guatemala.Holocene 6(3):265–271.

6. Lentz DL, Hockaday B (2009) Tikal timbers and temples: Ancient Maya agroforestryand the end of time. J Archaeol Sci 36(7):1342–1353.

7. Phillips OL, et al. (1998) Changes in the carbon balance of tropical forests: Evidencefrom long-term plots. Science 282(5388):439–442.

8. Haviland WA (1970) Tikal, Guatemala and Mesoamerican urbanism. World Archaeol2(2):186–198.

9. Sheets P, et al. (2012) Ancient manioc agriculture south of the Cerén village, El Sal-vador. Lat Am Antiq 23(3):259–281.

10. Inomata T, Triadan D (2010) Burned Palaces and Elite Residences of Aguateca: Ex-cavations and Ceramics (Univ of Utah Press, Salt Lake City).

11. Arnold PJ, III (1991) Domestic Ceramic Production and Spatial Organization: A Mex-ican Case Study in Ethnoarchaeology (Cambridge Univ Press, Cambridge, UK).

12. Nelson BA (1991) Ceramic frequency and use-life. Ceramic Ethnoarchaeology, edLongacre WA (Univ of Arizona Press, Tucson, AZ), pp 162–181.

13. Culbert TP (2003) The Ceramics of Tikal. Tikal: Dynasties, Foreigners, and Affairs ofState—Advancing Maya Archaeology, ed Sabloff JA (School of American ResearchPress, Santa Fe, NM), pp 47–82.

14. Arnold DE (2008) Social Change and the Evolution of Ceramic Production and Dis-tribution in a Maya Community (Colorado Univ Press, Boulder CO).

15. Wernecke DC (2008) A burning question: Maya lime technology and the Maya forest.J Ethnobiol 28(2):200–210.

16. Hather JG (1994) The identification of charred root and tuber crops from archaeo-logical sites in the Pacific. Tropical Archaeobotany: Applications and New Develop-ments, ed Hather JG (Routledge, London), pp 51–64.

17. Miksicek CH, Wing ES, Scudder SJ (1991) The ecology and economy of Cuello: Thenatural and cultural landscape of Preclassic Cuello. Cuello: An Early Maya Communityin Belize, ed Hammond N (Cambridge Univ Press, Cambridge, UK), pp 70–84.

18. Akpinar-Ferrand E, Dunning NP, Lentz DL, Jones JG (2012) Aguadas as water sourcesat southern Maya lowland sites. Anc Mesoam 23(1):85–101.

19. Gómez-Pompa A, Salvador-Flores J, Aliphat-Fernández M (1990) The sacred cacaogroves of the Maya. Lat Am Antiq 1(3):247–257.

20. Cheetham D (2010) Corn, calendars, and cooking: Early maize processing in the Mayalowlands and its implications. Pre-Columbian Foodways: Interdisciplinary Approachesto Food, Culture, and Markets in Ancient Mesoamerica, eds Staller JE, Carrasco M(Springer, New York), pp 345–368.

21. Sanders W (1957) Tierra y agua (Soil and Water): A study of the ecological factors inthe development of Meso-American civilizations. PhD thesis (Harvard University,Cambridge, MA).

22. Griffin RE (2012) The carrying capacity of Ancient Maya swidden maize cultivation:A case study in the region around San Bartolo, Petén, Guatemala. PhD thesis(Pennsylvania State University, University Park, PA).

23. Killion TW (1990) Cultivation intensity and residential site structure: An ethno-archaeological examination of peasant agriculture in the Sierra de los Tuxtlas,Veracruz, Mexico. Lat Am Antiq 1(3):191–215.

24. Cowgill UM (1962) An agricultural study of the southern Maya lowlands. Am An-thropol 64(2):273–286.

25. McAnany P (1995) Living with the Ancestors: Kinship and Kingship in Ancient MayaSociety (Univ of Texas Press, Austin, TX).

26. Mertz O, et al. (2008) A fresh look at shifting cultivation: Fallow length an uncertainindicator of productivity. Agric Syst 96(1-3):75–84.

27. Dunning NP, Beach TP, Farrell P, Luzzadder-Beach S (1998) Pre-Hispanic agricultureand adaptive systems in the Maya lowlands. Cult Agric 20(2-3):87–101.

28. Lawrence D, et al. (2007) Ecological feedbacks following deforestation create thepotential for a catastrophic ecosystem shift in tropical dry forest. Proc Natl Acad SciUSA 104(52):20696–20701.

29. Dunning NP, Beach TP, Luzzadder-Beach S (2012) Kax and kol: Collapse and resiliencein lowland Maya civilization. Proc Natl Acad Sci USA 109(10):3652–3657.

30. Turner BL, 2nd, Sabloff JA (2012) Classic Period collapse of the Central Maya Low-lands: Insights about human-environment relationships for sustainability. Proc NatlAcad Sci USA 109(35):13908–13914.

31. Tankersley KB, et al. (2011) Evidence for volcanic ash fall in the Maya lowlands froma reservoir at Tikal, Guatemala. J Archaeol Sci 38(11):2925–2938.

32. Scarborough VL, et al. (2012) Water and sustainable land use at the ancient tropicalcity of Tikal, Guatemala. Proc Natl Acad Sci USA 109(31):12408–12413.

33. Culbert TP, Levi LJ, Cruz L (1990) Lowland Maya wetland agriculture: The Rio Azulagronomy program. Vision and Revision in Maya Studies, eds Clancy FS, Harrison PD(Univ of New Mexico Press, Albuquerque, NM), pp 115–124.

34. Dickson BD (1980) Ancient agriculture and population at Tikal, Guatemala: An ap-plication of linear programming to the simulation of an archaeological problem. AmAntiq 45(4):697–712.

35. Harrison PD (1999) The Lords of Tikal: Rulers of an Ancient Maya City (Thames andHudson, New York).

36. Drennan RD (1984) Long-distance movement of goods in the Mesoamerican Forma-tive and Classic. Am Antiq 49(1):27–43.

37. Andrews AP (1983) Maya Salt Production and Trade (Univ of Arizona Press, Tucson, AZ).38. McKillop H (2002) Salt: White Gold of the Ancient Maya (Univ Press of Florida,

Gainesville, FL).39. Reents-Budet D, Ball JW (1994) Painting the Maya Universe: Royal Ceramics of the

Classic Period (Duke Univ Press, Durham, NC).40. Webster D (2002) The Fall of the Ancient Maya: Solving the Mystery of the Maya

Collapse (Thames and Hudson, New York).41. Kennett DJ, et al. (2012) Development and disintegration of Maya political systems in

response to climate change. Science 338(6108):788–791.42. Medina-Elizalde M, Rohling EJ (2012) Collapse of Classic Maya civilization related to

modest reduction in precipitation. Science 335(6071):956–959.43. Webster JW, et al. (2007) Stalagmite evidence from Belize indicating significant

droughts at the time of preclassic abandonment, the Maya hiatus, and the classicMaya collapse. Palaeogeogr Palaeoclimatol Palaeoecol 250(1-4):1–17.

44. Houston SD, Inomata T (2009) The Classic Maya (Cambridge Univ Press, Cambridge,UK).

45. Haviland WA (2003) Settlement, society, and demography at Tikal. Tikal: Dynasties,Foreigners & Affairs of State: Advancing Maya Archaeology, ed Sabloff JA (School ofAmerican Research Press, Santa Fe, NM), pp 111–142.

46. Cook BI, et al. (2012) Pre-Columbian deforestation as an amplifier of drought inMesoamerica. Geophys Res Lett 39(16):L16706.

47. Georgescu M, Lobell D, Field C, Mahalov A (2013) Simulated hydroclimatic impacts ofprojected Brazilian sugarcane expansion. Geophys Res Lett 40(5):972–977.

48. Oglesby RJ, et al. (2010) Collapse of the Maya: Could deforestation have contributed?J Geophys Res Atmos 115:D12106.

49. Demarest AA, Rice PM, Rice DS (2005) The terminal Classic in the Maya lowlands. TheTerminal Classic in the Maya Lowlands: Collapse, Transition, and Transformation, edsDemarest AA, Rice PM, Rice DS (Univ of Colorado Press, Boulder, CO), pp 545–572.

Supporting InformationLentz et al. 10.1073/pnas.1408631111SI Materials and MethodsVoronoi Diagram. Determining the territory controlled by an an-cient polity is a difficult problem for which only imperfect solutionscan be offered (1). Hieroglyphic inscriptions can sometimes offerinformation about interpolity relationships, including glimpses ofpolitical hierarchies, warfare, and royal marriages (e.g., ref. 2),but such texts, even where preserved, seldom contain informa-tion relevant to the ancient functional economy, nor the natureor extent of a polity’s effective resource extraction area. Thus, weare left to reconstruct ancient economic zones using archaeo-logical proxies.One approach to modeling ancient political and economic

spheres is through spatial analysis of archaeological settlementdata. This approach provides one way to model ancient politicaland economic spheres. For example, the relative size of ancientcommunities at various points in time offers information aboutthe labor and resources they controlled. Given the general sim-ilarity in residential density at most Maya sites, the areal extent ofsettlement can be used as a rough proxy for population size.Similarly, the volume of monumental architecture constructed ata site during a given time period can be used as a proxy for theamount of labor and resources controlled by the site’s rulers.For ancient Tikal, we attempted to reconstruct the size of its

territory or primary economic extraction zone for two points intime: about A.D. 100, the apogee of the Late Preclassic period,and about A.D. 700, the peak of Tikal’s regional influence in theLate Classic period (LCP). In both cases, we used available ar-chaeological and epigraphic data to determine Tikal’s nearestneighbors that were large enough to exert an appreciable degreeof economic autonomy. Because evidence indicates that none ofthese neighboring sites was either as large or regionally domi-nant as Tikal during the two times considered, we posited thattheir own zones of economic control did not extend as far as thatof Tikal, or some of their economic production went towardsupplying the needs of Tikal. Hence, for modeling purposes weweighted the extent of territories in favor of Tikal two-thirds toone-third.The amount of archaeological and epigraphic information

available for Tikal’s neighbors varies tremendously. Data havebeen available for many years for Uaxactun (3, 4), and Yaxha (5,6). Excavation data are now forthcoming for El Zotz and ElPalmar, Nakum (7), and Xultun (8). However, for most neigh-boring sites data are more preliminary, derived chiefly from earlyexpeditions (9, 10), the University of Pennsylvania Tikal Projectperipheral surveys (11–14), Instituto de Antropología e Historiade Guatemala site inventories (15, 16), and surveys of the RioHolmul drainage and intersite transects between Tikal, Yaxha,and Nakum (17–20). The Tikal Project surveys and those of Ford(17) and Fialko (19, 20) included a few test pit excavations toobtain chronological data, but the other projects examined onlysurface features, inscribed monuments, and sometimes looter’strenches.For the Late Preclassic zenith (ca. A.D. 100), Tikal’s nearest

competitive neighbors appear to have been Zocotzal, El Palmar,El Encanto, Jimbal, and Chalpate (also known as Ramonal). Atthe time, each of these centers included between two and fourarchitectural attributes associated with major centers: triadicgroups, pyramid complexes, ball courts, and intrasite sacbeob(causeways). For the Late Classic apogee (ca. A.D. 700), most ofthese Preclassic sites were either abandoned or subsumed assuburban or satellite centers of Tikal as it expanded its politicaland economic power. In this later period, Tikal’s likely nearest

competitive neighbors lay at greater distances: Motul de SanJose, El Zotz, Uaxactun, Xultun, Dos Aguadas, Nakum, Yaxha,and Ixlu. Hieroglyphic evidence suggests that several of thesecenters were either allied with or under the political sway ofTikal, but the large populations of these centers would likelyhave required the use of a sizeable portion of their own terri-torial resources.The Voronoi diagram, also known as a Thiessen polygon, has

been used as a tool in archaeological settlement analysis in manyparts of the world, including the Maya area (21). Although farfrom precise, this method has been widely used to approximatethe positions of intersite boundaries and zones of likely politicaleconomic jurisdiction. A primary limitation of earlier versions ofThiessen polygons has been that the method required eachgeographic center to have an equal amount of influence, whereasin fact the relative political power of individual centers was oftenunequal. Although solutions to this obstacle have been suggestedfor many years (22), the application of a weighted Voronoipolygon model, as presented here, has only recently begun to bepractical because of the increased processing speeds of computersystems and running geographic software programs. In our study,a Geographic Information System was used to produce a weightedVoronoi diagram to estimate the projected boundary of Tikal’seconomic extraction area. In a weighted Voronoi diagram, theweight of the Euclidean distance from each point in relationshipto all other points determines the boundary of each center. Tikalwas given a two-thirds to one-third weight over its neighbors,thus allowing Tikal to have a greater influence on positioningterritorial boundaries than any of its neighbors based on therationale noted earlier.

Palynological Data. Pollen data from Lake Petén Itza (23), a deeplake that is downwind and less than 5-km south of the extractivezone of Tikal, were evaluated to assess the amount of uplandforest clearance during the LCP. Pollen taxa characterizing highforest, such as Moraceae and Urticaceae, dropped to low levelsduring the LCP (Zone 3 of the Petén Itza core). The drop in highforest taxa is indicative of extensive forest clearance in the rangeof 60–70% in the Lake Peten Itza watershed based on palyno-logical and sediment data. Additionally, taxa indicating distur-bance, such as Ambrosia and other Asteraceae pollen, reach highpercentages in this zone. Pollen results were based on percen-tages of arboreal pollen versus nonarboreal pollen.Forest clearance co-occurred with erosion of sediments of low

organic content, indicating high siltation rates in the watershed.Increases in K and Fe suggest greater presence of clays, and anincrease in magnetic susceptibility indicates erosion of clasticmaterial into the lake during this time period (24). According tothese results at least 60% of the upland forest was cleared duringthe LCP, leaving ∼340 km2 of forest intact within the resourceextractive zone of Late Classic Tikal.Pollen data from Aguada Vaca de Monte (Fig. S1), a small

pond located in the Bajo Santa Fe with a pollen content morereflective of the surrounding bajo, indicate that 37–32% of thebajo lands (at least 80 km2) were cleared for agriculture and an-other 175 km2 remained as seasonal swamp forest (Fig. S2). Theamount of land clearance in the past was calculated by comparingthe modern percentage of arboreal pollen versus nonarboreal pol-len (at full forestation) to the percentages of arboreal pollen, bothmaximum and minimum, during the LCP.

Lentz et al. www.pnas.org/cgi/content/short/1408631111 1 of 11

Modern Forest Surveys. For the purposes of this study we havedivided the forests into upland and seasonal wetland (or bajo)even though the tropical forests of the northern Petén District aremuch more complex than this simplified dichotomy (25–27).Numerous transitional forest types exist between upland andbajo, many dominated by particular palm species, reflectingsubtleties in edaphic conditions and drainage patterns. In theLandsat 7 images we used in our biomass calculations we couldeasily distinguish between upland and bajo forest, but finer dis-tinctions were not possible. To compensate, our extensive sur-veys covered 5.95 ha of forests within the bounds of TikalNational Park and subsumed much of the upland and bajo forestvariability.

Biomass Calculations. Satellite images, remote sensing, and Geo-graphic Information System techniques were combined with datafrom forest surveys to calculate the modern biomass of the forestsaround Tikal and determine the extent of the various forest typeswithin the extraction zones of the Tikal community during theLate Preclassic and Late Classic periods. A March 27, 2003,Landsat 7 ETM+ image of the study area was acquired to aid inthe calculation of the above ground biomass (AGB) of modernTikal forests.The image was atmospherically corrected using a logarithmic

residuals technique (28) to mitigate as much as possible theundesired atmospheric component in all subsequent transforms.The use of this convenient, but less-exact, ad hoc atmosphericcalibration technique was preferred over a more intensive at-mospheric modeling approach (e.g., FLASH) because the at-mospheric data necessary for the computational models areabsent in this region. The use of reflectance data will be denotedby the symbol ρ (e.g., ρBand 5).We constructed (29) an object-based imagery analysis classi-

fication methodology for differentiating seasonal wetlands fromupland tropic forest and other forms of land cover and land useat a regional accuracy with a regional accuracy of 93.5% (κ =0.876). These classification data, which were constructed fromthe same Landsat ETM+ scene and a texturally transformedversion of the experimental global Advanced Spaceborne Ther-mal Emission and Reflection Radiometer digital elevation modelimagery, were used as a processing mask for each respective AGBmodel. The high accuracy of the classification reduced the potentialfor error because of AGB model use across the image. Once eachAGBmodel was derived, it was calculated on regions correspondingto their respective classes.Indices and data transforms were derived from atmospherically

corrected bands of the satellite dataset. By superimposing bio-mass parcels from vegetation surveys onto the remote-sensingdataset and experimentally regressing AGB values against theremotely sensed variables, we were able to calculate overall AGBwithin each forest type. A weighted mean value for pixel valueswas calculated for each parcel, with overlapping pixels beingproportionately weighted by their area of overlap to reduce anybias resulting from the spatial discontinuity between the parcelsand the coarse 30-m pixels of the satellite dataset. Ultimately, themean imagery dataset values were tabulated with AGB values andimported into a SPSS statistical package (v16) for regression.The relationship between predicted and measured biomass

derived from statistical bands, vegetation indices, and lineardecompositions (i.e., tasseled cap transform) can differ greatly inboth direction and strength between regions (30). Accordingly,separate models were experimentally determined for uplandtropical forest and the bajo vegetation parcels. To determinemean AGB error, both equations were entered into ENVI 4.7and compared back to the values for each vegetation parcel.Because vegetation parcels and pixels did not perfectly overlap,each parcel was related to an areally weighted sum of AGBvalues of the overlapping pixels to better assess the response of

each modeled pixel to their overlapping parcel. Each pixel thatwas modeled for AGB was summed and divided by the arealextent of its respective vegetation community. From these cal-culations the respective error bounds for each forest type wasdetermined.All possible combinations of relationships between bands and

variables were explored using multiple linear regression forbiomass calculations of bajo forests. Initial calculations indicateda strong linear relationship when data from ρBand 3 (pre-dominantly red visible light; 0.63–0.69 μm) and the Soil-AdjustedVegetation Index (SAVI) were combined, but the relationshipwas not a statistically significant one. There are inherent risks inassuming a global soil adjustment weight when attempting tomodel subtle local variations, which predicated dropping SAVI.A strong inverse relationship occurs between ρBand 3 and AGBoccurs because of the preferential absorption of red light.Chlorophyll absorbs red light for photosynthesis so lower ag-gregate red reflection corresponds to either denser or morevigorous healthy vegetation. This relationship was most stronglyindicated using a cubic curve (r2 = 0.758, P = 0.029, n = 8, SE =1.98) using the following equation: 937.0139 + (−30.906)* ρBand3+0* ρBand 32+0.005* ρBand 33. A cubic relationship was used inplace of a linear one not only because of the stronger statisticalrelationship but also because of the latter’s tendency to createanomalies (e.g., large negative or improbably high values forAGB) in a way that was less predictable than the cubic curve.This model was applied only to portions of the satellite datasetindicated to be seasonal wetlands by Magee (29).Upland forest biomass was calculated using all possible com-

binations of relationships between AGB, spectral bands, andderived satellite data. All were tested using multiple linear re-gression. Whereas previous studies have focused on the linearrelationship between forest biomass and vegetation indices (31),we found the strongest relationship of the biomass of uplandtropical forest to be the atmospherically corrected reflectance ofone of the satellite datasets ρBand 5 (sensitive to 1.55- to 1.75-μmwavelengths). Steinenger (32) observed this same relationshipfor AGB estimations in tropical forests in Bolivia and Brazil. Thestrongest correlation between ρBand 5 and each parcel’s AGB,again, was found using a cubic curve (r2 = 0.818, P = 0.006,SE = 12.02, n = 9) with the equation summarized as follows:7428.578 + (−175.784)* ρBand 5+1.041* ρBand 52+0* ρBand 53.ρBand 5 has a distinct inverse relationship to the internalmoisture of a plant’s chloroplasts, thus providing a strong ra-tionale for the model. Full details of the methodology of thebiomass portion of our study can be found in other publications(29, 30, 33). Our findings show that the AGB of modern uplandforest is 28.9 ± 2.6 million kg·km−2 and the AGB of the bajo is18.2 ± 0.523 million kg·km−2. The results of our surveys comparefavorably to data generated by biomass estimates from otherNeotropical forest surveys (34).

Annual Growth Increment Calculations. The study at Barro Colo-rado Island (BCI) was selected as a reference to estimate annualgrowth increment measurements in the ancient forest at Tikal fornumerous reasons: both areas are classified as moist tropicalforest, both have a pronounced wet and dry season, and both arelisted as having a tropical monsoon climate (Am) according to theKӧppen Climate Classification System (35, 36). Both forest systemsdeveloped out of the Neotropical flora (37), which evolved in SouthAmerica and then migrated into Central America when the twoland masses joined together ∼4 million y ago. Although the forestsof Tikal and BCI are not identical in their species make-up, theyshare many commonalities. For example, the dominant oligarchicspecies in terms of basal area at Tikal is Brosimum alicastrum Sw.This is also a common species at BCI (36) and a tree whose wood iswell represented among the paleoethnobotanical remains at Tikal(38, 39) (Table S2). Alternatively, the most common species at BCI

is Trichila cipo (A. Juss.) C. DC. Although this particular species isnot found in the Tikal forests, a closely related congeneric species,Trichila minutiflora Standl. is one of the dominant species there(40). Three other species of Trichilia are also found in the modernTikal forests: Trichila pallida Sw., Trichila moschata Sw., andTrichilia glabra L. Parenthetically, Trichilia wood was identifiedamong the archaeological plant remains at Tikal, evidence thattrees of this genus were present during ancient times and ofeconomic value to the Maya.Although it is true that the BCI forest is in the middle of a lake

and Tikal is not, that lake (Lake Gatun) is artificial and wascreated during the construction of the Panama Canal 100 y ago.The BCI forest was established as a result of natural processeslong before the lake appeared.Ideally speaking, the annual growth increment measurements

used in this study would have come from Tikal forests or frommoist tropical forests in the Petén region. Unfortunately, no long-term forest studies that generated the data that we required havebeen completed in the region. In short, the closest and most well-studied moist Neotropical forest in Central America with theappropriate measurements was the BCI forest.

Ancient Wood Use. Firewood for cooking undoubtedly representedthe major demand for fuel among the ancient Maya (41). Nu-merous ethnographic studies have recorded the daily firewoodrequirement of traditional Mesoamerican farmers, with the es-timates ranging from 2.3 kg per person per day (42) to 3.2 kg perperson per day (43). For our calculations, we used the minimumnumber of 2.3 kg per person per day. For 45,000 inhabitants (44),the need for cooking fuel would have been a staggering 38 mil-lion kg/yr (2.3 kg × 45,000 persons × 365 d).Judging from Maya archaeological sites that were catastro-

phically destroyed (45, 46), each household would use 70–80ceramic vessels at any given time and replace them on an annualbasis because of breakage (47, 48). The ancient Maya likely usedopen kilns (49), which were extremely inefficient and consumedabout 5.2 kg of wood per vessel (50), requiring around 73 kg perperson per year (assuming five persons per family). For the en-tire polity of Tikal during the LCP, the total need for wood tofire ceramics was ∼3.3 million kg/yr (73 kg per person per year ×45,000 persons).Lime, along with crushed limestone, weathered limestone, and

other fillers (all components of plaster) represented anothermajor demand on the forests for fuel. Lime is made by burningcrushed limestone and serves as the binder in traditional Mayaplaster (51). Most of the surfaces of the site core at Tikal (in-cluding plazas, temples, palaces, ball courts, reservoirs, andcauseways) were covered with plaster. This total area (620,000 m2)would have required 16 million kg to plaster the exterior surfacesof Late Classic Tikal. It took about 5 kg of wood to make 1 kg oflime using traditional open kiln technology (51). Therefore, itwould have taken around 80 million kg of wood to plaster all ofTikal. Unlike the need for hearth fuel and firewood for ceramics,however, the demand for plaster could be spaced out over manyyears. Studies at Copan (41) estimated that surfaces were plas-tered every 50 y. If this were true at LCP Tikal then only 1.6million kg of firewood would have been required for annualplaster manufacture and even less if the plasterers decided to“water down” the formula and use less lime in the mix.Similar to lime production, the need for construction wood

could be spread over many years, because, for example, tradi-tional Maya houses last an average of 25 y (52). The 1778 resi-dential structures within the central core of Tikal (53) required,according to our calculations, 60,000 kg·yr−1. If we double thisfigure to account for temples (comprised mostly of cut stone andrubble fill), scaffolding and outlying residences, then the amountof construction wood needed each year at Late Classic Tikal was

120,000 kg, a relatively insignificant figure compared with otherwood needs.The estimated annual need for fuel and construction wood for

a Tikal population of 45,000 slightly exceeds what our calculationsindicate was available (54) (Table S1). This shortage could haverepresented a challenge for the LCP Maya at Tikal. One way thisshortfall could have been accommodated was the importation ofpine charcoal or the cultivation of pine, as has been suggested forother Maya communities (55). Pine was widely used as a com-ponent of ceremonial activities among the ancient Maya (56, 57),but also served as an everyday fuel, especially in elite households(55). There is substantial evidence for this hypothesis at Tikalbecause pine charcoal was one of the most common plant re-mains in the paleoethnobotanical record, even though pine is nota common element of the upland or seasonal wetland forest.There is a small (180 ha) but ancient, stand (58) of pine about20 km to the northwest of the Tikal site core that evidently washeavily exploited by the ancient Maya, but never eliminated. Ifthe Maya were not carefully managing this resource, they easilycould have wiped it out after several centuries of populationgrowth and high demand. Conversely, it appears evident that theMaya were managing their pine and other forest resources to theoptimum productive capacity.The tree species represented in the archaeological wood as-

semblage are similar to the composition of tree species in themodern forests of Tikal. For example, 8 of the 10 most commonhardwood trees represented in the paleoethnobotanical record—that is, Manilkara zapota (L.) Royen, Haematoxylon campechianumL., Pouteria sp., Brosimum alicastrum Sw., Pseudolmedia glabrata(Liebm.) Berg., Nectandra sp., Protium copal (Schiltdl. & Cham.)Engl., and Lonchocarpus sp.—are among the modern forest ol-igarchic species (trees with high importance value based on basalarea and relative stem density). The other two most commonhardwood trees from the archaeological remains, Aspidospermasp. and Licaria sp., although not oligarchic species, were fre-quently encountered in survey plots of the modern forests.Further evidence that oligarchic tree species were prominentlyrepresented in the paleoethnobotanical record, is that they weresignificantly present in greater numbers than the overall mean(X1, 177 = 3.87, P = 0.049), whereas the mean abundance of non-oligarchic woods among archaeological samples was not signifi-cantly different from the overall mean (X2

1, 177 = 0.125, P = 0.72).When species found in both ancient and modern contexts onlywere compared, the relative species abundance in the archaeo-logical wood remains correlated positively with the importancevalue of woods found in the modern forest (Spearman’s ρ =0.292, P = 0.045, n = 46). These results indicate that the ancientTikal Maya were actively selecting long-lived forest dominantsfor use, but did so in a way that did not perceptibly alter theforest structure. Compare this approach to the historic practiceof mahogany exploitation in Belize, where a once common tree isnow a rare species (59).Although the pollen record indicates that the Late Classic

Maya of Tikal reduced their upland forests by about 60%, Mayaethnographic accounts suggest that the remaining woodlots couldhave been protected as ancestral forests. Possibly these forestswere managed using a harvesting technique, such as selective or“umbrella” felling that was used in the Middle Ages in northernEurope to preserve species diversity (60).The idea that theMaya weremanaging their forests is a concept

that has been presented elsewhere. Lentz and Hockaday (61)demonstrated that the monumental construction at Tikal re-quired specific highly valued deciduous hardwoods. Timbersfrom trees of considerable girth were used in the construction ofseveral of the major temples during the 8th century, long afterthe population boom of the LCP began. These trees could onlyhave come from old-growth forests that were somehow protectedfrom the heavy demand for prime agricultural land. The girth of

these temple timbers declined toward the end of the 8th century,however, indicating an attenuation of the conservation practicesthat protected the old-growth forests for many generations.Another study (62) concluded that the Maya in the coastal areaof Belize were managing their forests to maintain their majorfuel source for the salt production industry.

Agriculture at Tikal.To understand how the vegetated environmentmight have beenmanaged at Tikal, we turned to the ancientMayavillage of Cerén that was covered by volcanic ash near the be-ginning of the LCP in A.D. 650 (45). As a result of the excellentpreservation at Cerén, we know what plants the Maya weregrowing and also how the Maya inhabitants were growing them.Because of the overlap in cultigens used at Tikal and Cerén andthe cultural similarities between the two communities, it seemslikely that their agricultural practices would be analogous. Ac-cordingly, Cerén serves as an excellent model to help interpretland use activities at Tikal.In addition to farming activities in upland areas, which relied on

annual crops, orchard trees, and root crops, our studies show thatdeep, cumulic soils on foot and toe-slopes surrounding bajos werealso of agricultural importance. Excavations at the small settle-ment near Aguada de Terminos in Bajo de Santa Fe (Fig. S6)revealed that the ancient Maya occupants were indeed practicingmaize and root crop agriculture in the bajos and were constructingterraces to conserve soil and water resources. The small pondadjacent to the site provided a consistent supply of water andenabled the residents to survive through the drought and into thePostclassic period.

Stable Carbon Isotopes. Stable carbon isotopes in the soil organicmatter have been shown to hold a record of past vegetativehistories where C3 forest trees and vines have been replaced byC4 tropical grasses, such as maize. Woody plants use a C3 pho-tosynthesis pathway that is highly discriminatory toward heavy13CO2. Maize and other C4 grasses are much less discriminatorytoward the heavier 13CO2. Others (63–65) have demonstratedthat increases in 13C in the soils organic matter provides evi-dence that soil pedons once hosted maize and other C4 plantsassociated with forest clearance for agricultural use. Soil pedons1 through 4 (Fig. 2) were collected by bucket auger at locationsbetween 500- and 800-m south and west of the Perdido reservoir.Augured soil samples are highly disturbed and we were unable toobserve the soil laminations found in a pit located about 100-msouth of the reservoir outlet.The soil samples were collected at 15-cm depth intervals from

the surface to bedrock or to a maximum depth of 195 cm. Thesamples were crushed, sieved (<2 mm), and homogenized.Subsamples (2 g) were further ground and sieved to less than0.25 mm before acidification to remove carbonate and alkaline

pyrophosphate extraction to remove fulvic and humic acidfractions of the soil organic matter. Carbon isotope ratios of theresidual humin fraction were determined by isotope ratio massspectrometer coupled with an elemental analyzer (63–65).

Tikal Population. To provide an accurate representation of theresource requirements of Tikal during the LCP, we needed someidea of the number of its inhabitants. Fortunately, archaeologistshave addressed this subject and, using various datasets andinterpretations, have estimated the population of Tikal as 45,000(44), 62,000 (54), 80,000 (66), and 100,000–200,000 (67). Theresults of our research strongly indicate that the larger pop-ulation estimates cited by Culbert et al. (54), Dickson (66), andHarrison (67) could not have been supported in any sort ofsustainable land-use system without significant input of food andfuel from outside of the Tikal extractive zone. Note that thepopulation estimate of 62,000 offered by Culbert et al. (54) is forthe site core of 120 km2 only. The entire realm of Tikal, theystate, included 1,963 km2 and a population of 425,000. Thislarger population figure, in our viewpoint, would have beenimpossible given the agricultural potential and fuel producingtechnology of the time. Some scholars have proposed the ideathat major amounts of foodstuffs could have been moved fromone area in the ancient Maya realm to another (68), but others(69, 70) have discounted this possibility, mostly because of theenergetic constraints of moving supplies by human porter withoutthe aid of draft animals, wheeled vehicles, or accessible waterways.

Drought. Other, less-densely populated communities with moreconsistent water supplies near Tikal were able to survive throughthe 9th century drought. A small household near the Aguada deTerminos, several kilometers east of the site core of Tikal, per-sisted and continued its occupation into the Postclassic period,long after the city center was abandoned. El Zotz, Tikal’s nearneighbor to the west, with a huge oversized reservoir and smallpopulation also was able to endure the drought into the Post-classic period. Both of these communities were able to survivebecause they had reliable water sources and small populationsthat did not stretch the carrying capacity of their landscape be-yond its point of resilience to fluctuations in climate.

Anthropogenic Influences. Recent climate modeling studies inNeotropical areas (71–73) have concluded that deforestation willresult in a reduction of somewhere between 5% and 30% of latewet season precipitation. This reduction is caused by reducedevapotranspiration from less vegetation and increased surfaceheating that results in high pressure zones in the atmosphere,which in turn disrupt convection and, ultimately, rainfall cycles.Even partial forest removal as proposed herein can contribute tothis effect.

1. Garrison T, Dunning NP (2009) Settlement, environment, and politics in the SanBartolo – Xultun Territory, El Peten, Guatemala. Lat Am Antiq 20(4):525–552.

2. Martin S, Grube N (2008) Chronicle of the Maya Kings and Queens (Thames andHudson, New York).

3. Ricketson OG (1937) Uaxactun, Guatemala, Group E - 1926–1931. Part 1: The Exca-vations. Carnegie Institution of Washington Publication No. 477 (Carnegie Institution,Washington, DC).

4. Valdes JA (1986) Uaxactun: Recientes investigaciones. Mexicon 7(6):125–128.5. Rice DS, Rice PM (1981) Muralla de Leon: A lowland Maya fortification. J Field Archaeol

8(3):271–288.6. Cabrera T (2000) Resultados preliminares de las investigaciones realizadas en el Grupo

Maler, Yaxha, Petén. XIII Simposio de Investigaciones Arqueológicas en Guatemala,1999, eds Laporte JP, et al. (Museo Nacional de Arqueología y Etnología, GuatemalaCity, Guatemala), pp 354–373.

7. Hermes B, Calderón Z (2003) La secuencia de ocupación prehispánica en Nakum: Unavisión preliminar. XVI Simposio de Investigaciones Arqueológicas en Guatemala, 2002,eds Laporte JP, et al. (Museo Nacional de Arqueología y Etnología, Guatemala City,Guatemala), pp 305–319.

8. SaturnoWA, Stuart D, Aveni AF, Rossi F (2012) Ancient Maya astronomical tables fromXultun, Guatemala. Science 336(6082):714–717.

9. Morley SG (1937) Inscriptions of the Peten, Vol. 1. Publication No. 437. (CarnegieInstitution of Washington, Washington, DC).

10. Bullard W (1960) The Maya settlement pattern in northeastern Peten, Guatemala. AmAntiq 25(3):355–372.

11. Fry RE (2003) The peripheries of Tikal. Tikal: Dynasties, Foreigners and Affairs of State,ed Sabloff JA (SAR Press, Santa Fe, NM), pp 143–170.

12. Fry RE (1969) Ceramics and settlement in the periphery of Tikal. PhD thesis (Universityof Arizona, Tucson, AZ).

13. Puleston DE (1973) Ancient Maya settlement patterns and environment at Tikal,Guatemala: Implications for subsistencemodels. PhD. thesis (University of Pennsylvania,Philadelphia, PA).

14. Puleston DE (1983) The Settlement Survey of Tikal. Tikal Report No.13. (Univ MuseumPublications, Philadelphia, PA).

15. Quintana O, Noriega R (2006) Tikal y sus vecinos: Complejodad cultural en el triánguloYaxha-Nakum-Naranjo. XIX Simposio de Investigacioies Arqueológicas en Guatemala,2004, eds Laporte JP, et al. (Museo Nacional de Arqueología e Etnología, GuatemalaCity, Guatemala), pp 333–339.

16. Quintana O, Wurster W (2001) Ciudades Mayas del Noreste del Petén, Guatemala: Unestudio urbanístico comparativo. AVA –Materialien 59 (Philipp von Zabern, Mainz amRhein, Germany).

17. Ford A (1986) Population Growth and Social Complexity: An Examination of Settlementand Environment in the Central Maya Lowlands (Univ of Arizona Press, Tempe, AZ).

18. Lou B (1997) Chalpate: Análisis del asentamiento y orientación de una centro satellitede Tikal. X Simposio de Investigaciones Arqueológicas en Guatemala, 1996, edLaPorte EH (Museo Nacional de Arqueología y Etnología, Guatemala City, Guate-mala), pp 373–380.

19. Fialko V (2000) Recursos hidraulicos en Tikal y sus periferias. XIII Simposio de Inves-tigaciones Arqueológicas en Guatemala, 1999, eds Laporte JP, et al. (Museo Nacionalde Arqueología y Etnología, Guatemala City, Guatemala), pp 556–565.

20. Fialko V (2005) Diez años de investigaciones arqueológicas en la Cuenca del RíoHolmul, region norteste del Petén. XVIII Simposio de Investigacioies Arqueológicas enGuatemala, 2004, eds Laporte JP, et al. (Museo Nacional de Arqueología y Etnología,Guatemala City, Guatemala), pp 244–260.

21. Montmollin OD (1988) Scales of settlement study for complex societies: Analyticalissues from the Classic Maya area. J Field Archaeol 15(2):151–168.

22. Boots BN (1980) Weighting Thiessen polygons. Econ Geogr 56(3):248–259.23. Islebe GA, et al. (1996) A Holocene vegetation history from lowland Guatemala.

Holocene 6(3):265–271.24. Curtis JH, et al. (1998) A multi-proxy study of Holocene environmental change in the

Maya Lowlands of Petén, Guatemala. J Paleolimnol 19(2):139–159.25. Kunen JL, et al. (2000) Bajo communities: A case study from the central Petén. Cult

Agric 22(3):15–31.26. Schulze MD, Whitacre DF (1999) A classification and ordination of the tree community

of Tikal National Park, Petén, Guatemala. Bulletin of the Florida Museum of NaturalHistory 41(3):169–297.

27. Dunning N, Beach T, Luzzadder-Beach S (2006) Environmental variability among Bajosin the Southern Maya Lowlands and its implications for ancient Maya civilization andarchaeology. Pre-Columbian Water Management, eds Lucero L, Fash B (Univ ofArizona, Tempe, AZ), pp 111–133.

28. Green AA, Craig MD (1985) Analysis of aircraft spectrometer data, with logarithmicresiduals. Proceedings of the Airborne Imaging Spectrometer Data Analysis Work-shop, eds Vane G, Goetz A (Jet Propulsion Laboratory, Pasadena, CA), pp 111–119.

29. Magee KS (2011) Segmentation, object oriented applications for remote sensing landcover and land use classifications. PhD thesis (Univ of Cincinnati, Cincinnati, OH).

30. Foody GM, Boyd DS, Cutler MEJ (2003) Predictive relations of tropical forest biomassfrom Landsat TM data and their transferability between regions. Remote Sens Envi-ron 85(4):463–474.

31. Zheng D, et al. (2004) Estimating aboveground biomass using Landsat 7 ETM+ data acrossa managed landscape in northern Wisconsin, USA. Remote Sens Environ 93(3):402–411.

32. Steininger MK (2000) Satellite estimation of tropical secondary forest above-groundbiomass: Data from Brazil and Bolivia. Int J Remote Sens 21(6-7):1139–1157.

33. Lentz DL, Dunning NP, Scarborough VL, eds. (2014) Tikal: Paleoecology of an AncientMaya City (Cambridge Univ Press, Cambridge, UK), in press.

34. DeWalt SJ, Chave J (2004) Structure and biomass of four lowland Neotropical forests.Biotropica 36(1):7–19.

35. Akpinar-Ferrand E, Dunning NP, Lentz DL, Jones JG (2012) Aguadas as water sourcesat southern Maya lowland sites. Anc Mesoam 23(1):85–101.

36. Croat TB (1978) Flora of Barro Colorado Island (Stanford Univ Press, Stanford, CA).37. Greller AM (2000) Vegetation in the floristic regions of North and Central America.

Imperfect Balance, ed Lentz DL (Columbia Univ Press, New York), pp 39–87.38. Moholy-Nagy H (2003) The Artifacts of Tikal: Utilitarian Artifacts and Unworked

Material (University Museum, Univ of Pennsylvania, Philadelphia, PA).39. Pohl M, Pope K, Jones J (2000) Base agrícola de la civilización Maya de las Tierras Bajas. XIII

Simposio de Investigaciones Arqueológicas en Guatemala, eds Laporte JP, et al. (MuseoNacional de Arqueología y Etnología, Guatemala City, Guatemala), pp 258–267.

40. Thompson KM (2013) Biodiversity in forests of the ancient Maya lowlands and geneticvariation in a dominant tree, Manilkara zapota (Sapotaceae): Ecological and an-thropogenic implications. PhD thesis (University of Cincinnati, Cincinnati, OH).

41. Abrams EM, Rue DJ (1988) The causes and consequences of deforestation among theprehistoric Maya. Hum Ecol 16(4):377–395.

42. Romero JJ (2003) Leña y Carbón. MA thesis (Universidad de San Carlos de Guatemala,Guatemala City, Guatemala).

43. Breedlove DE, Laughlin RM (2000) The Flowering of Man: A Tzotzil Botany of Zina-cantan (Smithsonian Institution, Washington, DC).

44. Haviland WA (1970) Tikal, Guatemala and Mesoamerican urbanism. World Archaeol2(2):186–198.

45. Sheets P, et al. (2012) Ancient manioc agriculture south of the Cerén village, El Sal-vador. Lat Am Antiq 23(3):259–281.

46. Inomata T, Triadan D (2010) Burned Palaces and Elite Residences of Aguateca: Ex-cavations and Ceramics (Univ of Utah Press, Salt Lake City).

47. Arnold PJ, III (1991) Domestic Ceramic Production and Spatial Organization: A Mex-ican Case Study in Ethnoarchaeology (Cambridge Univ Press, Cambridge, UK).

48. Nelson BA (1991) Ceramic frequency and use-life. Ceramic Ethnoarchaeology edLongacre WA (Univ of Arizona Press, Tucson), pp 162–181.

49. Culbert TP (2003) The ceramics of Tikal. Tikal: Dynasties, Foreigners, and Affairs ofstate –Advancing Maya Archaeology, ed Sabloff JA (School of American Research,Santa Fe, NM), pp 47–82.

50. Arnold DE (2008) Social Change and the Evolution of Ceramic Production and Dis-tribution in a Maya Community (Colorado Univ Press, Boulder, CO).

51. Wernecke DC (2008) A burning question: Maya lime technology and the Maya forest.J Ethnobiol 28(2):200–210.

52. Vogt EZ (1969) Zinacantan: A Maya Community in the Maya Highlands of Chiapas(Belknap Press, Cambridge, MA).

53. Arnold JE, Ford A (1980) A statistical examination of settlement patterns at Tikal,Guatemala. Am Antiq 45(4):713–726.

54. Culbert TP, Levi LJ, Cruz L (1990) Lowland Maya Wetland Agriculture: the Rio AzulAgronomy Program. Vision and Revision in Maya Studies, eds Clancy FS, Harrison PD(Univ of New Mexico Press, Albuquerque, NM), pp 115–124.

55. Lentz DL, Yaeger J, Robin C, Ashmore W (2005) Pine, prestige and politics of the LateClassic Maya at Xunantunich, Belize. Antiquity 79(305):573–585.

56. Morehart CT, Lentz DL, Prufer KM (2005) Wood of the Gods: The ritual use of pine(Pinus spp.) by the ancient lowland Maya. Lat Am Antiq 16(3):255–274.

57. Morehart CT, Helmke CGB (2008) Situating power and locating knowledge: A pale-oethnobotanical perspective on Late Classic Maya gender and social relations. AmAnthropol 18(1):60–75.

58. Dvorak WS, Hamrick JL, Gutiérrez JL (2005) The origin of Caribbean pine in the sea-sonal swamps of the Yucatán. Int J Plant Sci 166(6):985–994.

59. Thomson P (2004) Belize: A Concise History (Macmillan, Oxford).60. Schulze ED, Beck E, Muller-Hohenstein K (2002) Plant Ecology (Springer, Heidelberg).61. Lentz DL, Hockaday B (2009) Tikal timbers and temples: Ancient Maya agroforestry

and the end of time. J Archaeol Sci 36(7):1342–1353.62. Robinson M, McKillop H (2014) Fuelling the ancient Maya salt industry. Econ Bot

68(1):96–108.63. Webb EA, Schwarcz HP, Healy PF (2004) Carbon isotope evidence for ancient maize

agriculture in the Maya lowlands. J Archaeol Sci 31(8):1039–1052.64. Burnett RL, et al. (2012) The ancient agricultural landscape of the satellite settlement

of Ramonal near Tikal, Guatemala. Quat Int 265(28):101–115.65. Burnett RL, et al. (2012) Upland and lowland soil resources of the ancient Maya at

Tikal, Guatemala. Soil Sci Soc Am J 76(6):2083–2096.66. Dickson BD (1980) Ancient agriculture and population at Tikal, Guatemala: An ap-

plication of linear programming to the simulation of an archaeological problem. AmAntiq 45(4):697–712.

67. Harrison PD (1999) The Lords of Tikal: Rulers of an Ancient Maya City (Thames andHudson, New York).

68. Pyburn KA (1998) The Albion island settlement pattern project: Domination and re-sistance in Early Classic northern Belize. J Field Archaeol 25(1):37–62.

69. Drennan RD (1984) Long-distance movement of goods in the Mesoamerican Forma-tive and Classic. Am Antiq 49(1):27–43.

70. Webster D (2002) The Fall of the Ancient Maya: Solving the Mystery of the MayaCollapse (Thames and Hudson, New York).

71. Kennett DJ, et al. (2012) Development and disintegration of Maya political systems inresponse to climate change. Science 338(6108):788–791.

72. Cook BI, et al. (2012) Pre-Columbian deforestation as an amplifier of drought inMesoamerica. Geophys Res Lett 39(16):L16706.

73. Demarest AA, Rice PM, Rice DS (2005) The terminal classic in the Maya lowlands.The Terminal Classic in the Maya Lowlands: Collapse, Transition, and Trans-formation, eds Demarest AA, Rice PM, Rice DS (Univ of Colorado Press, Boulder,CO), pp 545–572.

Fig. S1. Vaca de Monte pollen profile.

Fig. S2. Land use at Late Classic Tikal. Chart showing relative proportions of land use categories within the Tikal extractive zone.

Fig. S3. Pollen profile from Aguada de Terminos. Located in the Bajo Santa Fe, just east of the site core of Tikal. Note that pollen evidence for the root cropachira (Canna cf indica) can be found in the earlier levels and that maize pollen is evident well into the Postclassic period.

Fig. S4. Scanning electron microscopy of burned cacao (Theobroma cacao L.) wood. This micrograph provides evidence that cacao was cultivated at Tikal.

Fig. S5. Perdido Pocket Bajo soil profile. This is the south profile of an excavation in the Perdido Pocket Bajo (Op 8D) located just to the south of andstratigraphically below the Perdido Reservoir. The initiation of irrigated agriculture likely commenced with the radiocarbon-dated layer [cal A.D. 485 ± 85 (SD)]at the bottom of the stratified alluvial sediments. Because the size of the pit from which the profile was recorded was quite small (1 m × 1 m), the photographon the right side was taken from the ground surface, causing the lower levels of the pit to appear compressed in the image (i.e., the drawing is true to scale butthe photograph is not).

Fig. S6. Aguada de Terminos map. Note the agricultural terraces and the proximity to the bajo. The aguada, although small, is a consistent source of waterand undoubtedly enabled the occupants of the Grupo de Terminos to survive the 9th century dry period. The base map is from Puleston (14), courtesy of theUniversity of Pennsylvania Museum of Archaeology and Anthropology.

Table S1. Fuel and timber needs of Tikal during the LCP

Wood use

Quantity required (in millions of kg·yr−1)

LCP population 45,000 LCP population 62,000

Firewood for cooking 37.8 52.1Ceramic manufacture 3.3 4.6Plaster production 1.6 1.6Construction 0.1 0.1Total 42.8 58.4

The wood needs of 45,000 inhabitants at Tikal would have stressed thesystem, but was potentially manageable. When we compare the wood needsof a hypothesized LCP population of 62,000 (54), it becomes readily evidentthat the landscape could not have supported a population of that magni-tude. Total amount of wood available per year from upland and wetlandforest = 39 million kg·yr−1.

Table S2. Ancient plant remains identified at Tikal

Plant taxa Contextual data

Annual crops [common name]Cucurbita moschata (Lam.) Poir. [squash] C4;H3-4;N1;P2*,†

Cucurbita pepo L. [pumpkin] C3;H3-4;N3;P2*Gossypium hirsutum L. [cotton] C3;H3;N1;P2*Phaseolus coccineus L. [scarlet runner] C2-3;H3-4;N2;P2Phaseolus lunatus L. [lima bean] C3;H3-4;N2;P2*Phaseolus vulgaris L. [common bean] C3;H3-4;N2;P2*Zea mays L. [maize] C3-6;H3-4;N33;P2,6,7,8*

Tree crops [common name]Acrocomia aculeata Lodd. ex Mart. [coyol] C2-4;H3;N1;P4*Bactris major Jacq. [biscoyol] C3;H2,3;N2;P3,6Byrsonima crassifolia (L.) H.B.K. [nance] C2-6;H3;N2;P1,3*Persea americana Mill. [avocado] C5;H3;N1;P3*Pouteria sapota (Jacq.) Mre. & Strn. [sapote] C4;H1,3;N1;P1Spondias spp. [jocote] C2-5;H1,3;N8;P1,3*Theobroma cacao L. [cacao] C3-5;H3;N4;P1,2*

Root crops [common name]Canna cf indica L. [achira] C2;H2;N1;P8Ipomoea batatas (L.) Lam. [sweet potato] C1;H3,4;N1;P5Xanthosoma sagittifolium (L.) Schott. [malanga] C4,5;H3,4;N1;P5*,‡

Other useful plants [common name]Cyperus canus J. Presl & C. Presl. [tule] C4;H2;N1;P9*cf Morinda sp. [pinuela] C3;H3;N1;P1Piper sp. [cordoncillo] C3;H1;N2;P1Tecoma stans (L.) H.B.K. [flor amarilla] C3,4;H3,5;N1;P1Thevetia ahouai (L.) A. DC. [cocheton] C2;H1;N1;P2

Trees [common name]Acacia sp. [subín] C7;H1;N1;P1Acosmium panamense (Benth.) Yak. [billywebb] C3;H1;N2;P1Alvaradoa subovata Cronquist [cortacuero] C4,5;H5;N2;P1Ampelocera hottlei (Standl.) Standl. [bullhoof] C2-4;H1;N1;P1Aspidosperma spp. [white malady] C2-4;H1;N6;P1*Astronium graveolens Jacq. [glassywood] C2-4;H1;N1;P1Brosimum alicastrum Sw. [ramón] C2-5;H1;N8;P1,2Caesalpinia sp. [warree wood] C4;H1;N3;P1Cameraria latifolia L. [white poison Wood] C2;H2;N1;P1cf Carapa guianensis Aubl. [andiroba] C2,3;H1;N1;P1Casearia sp. [café de monte] C2;H1;N1;P1*Ceiba pentandra (L.) Gaertn. [ceiba] C5;H1;N1;P1Celtis iguanaea (Jacq.) Sarg. [wild cherry] C2-5;H1;N1;P3†

Clusia sp. [matapalo] C2;H1;N1;P1Croton sp. [hierba de jabali] C2,3;H2;N3;P1Chrysophyllum sp. [star apple] C2-4;H1;N2;P1Cupania sp. [grande betty] C3-5;H1,2;N1;P1Enterolobium cyclocarpum (Jacq.) Griesb. [guanacaste] C5;H1;N1;P2†

Erythrina spp. [tiger wood] C3,4;H3,5;N2;P1Eugenia spp. [guabillo] C2-4;H1;N3;P1Ficus sp. [fig] C3-4;H1;N2;P1*Garcinia cf intermedia (Pittier) Hamm. [jocomico] C5;H5;N1;P1Gliricidia sepium (Jacq.) Steud. [madre de cacao] C4;H2;N2;P1Guarea glabra Vahl [cedrillo] C2-4;H1;N1;P1cf Guettarda combsii Urb. [arepa] C7;H1;N2;P1Haematoxylum campechianum L. [logwood] C3-6;H2;N31;P1*Heliocarpus sp. [broadleaf moho] C2,3;H1;N2;P1Hirtella sp. [pigeon plum] C2-4;H1;N1;P1cf Lacmellea sp. [chicle dwarf] C7;H2;N1;P1Licaria spp. [laurelillo] C2-4;H1-2;N8;P1Lonchocarpus spp. [dogwood] C2-6;H1,2;N5;P1Manilkara zapota (L.) Royen [sapodilla] C2-5;H1-3,N34,P1,2*Metopium brownei (Jacq.) Urb. [poisonwood] C4;H2;N1;P1Nectandra spp. [timber sweet] C2-4;H1,2;N8;P1*Ocotea puberula (Rich.) Nees. [wakkowit] C2-4;H1;N2;P1Pimenta dioica (L.) Merr. [allspice] C5;H1;N1;P2

Table S2. Cont.

Plant taxa Contextual data

Trees [common name] cont.Pinus spp. [pine] C2-5;H5;N118;P1*Pouteria spp. [mamey] C2-5;H1;N20;P1Protium copal (Schiltdl. & Cham.) Engl. [copal] C2-4;H1;N6;P1,2Pseudolmedia glabrata (Liebm.) Berg. [cherry] C2-4;H1;N8;P1Psychotria sp. [night bloom] C4;H1;N1;P1Salix cf chilensis Molina [willow] C5;H2;N2;P1Sebastiana sp. [poison wood] C2-4;H1;N3;P1Sideroxylon sp. [silion] C2-4;H1;N6;P1Stemmadenia sp. [cojotón] C2-4;H1;N1;P1cf Tabebuia sp. [yellow mayflower] C4;H1;N2;P1Tapirira cf mexicana Marchand [tanto] C2-4;H1;N2;P1Terminalia buceras (L.) C. Wright [pukté] C3;H2;N3;P1Trichilia spp. [red cedar] C3;H1;N3;P1Trophis spp. [white ramón] C3-5;H1;N4;P1Zanthoxylum caribaeum Lam. [prickly yellow] C2-4;H1,2;N4;P1Zuelania guidonia (Sw.) Britt. & Millsp. [tamai] C2-4;H1;N1;P1

Chronological assessments (C) are numbered as follows: 1, Middle Preclassic (1000–300 B.C.); 2, Late Preclassic(300 B.C. to A.D. 250); 3, Early Classic (A.D. 250–600); 4, Late Classic (A.D. 600–850); 5, Terminal Classic (A.D. 850–950); 6, Postclassic (A.D. 950–1150); 7, unknown date. H represents the habitat where the plant species likelyoriginated, numbered as follows: 1, upland forest; 2, bajo; 3, kitchen garden; 4, field; 5, other. N represents thenumber of contexts from which the plant remains were recovered. P represents plant part, numbered as follows:1, wood; 2, seed; 3, pit; 4, endocarp; 5, tuber; 6, stem; 7, cob; 8, pollen; 9, leaf.*Indicates plants also found at Late Classic Cerén.†See Moholy-Nagy (38).‡See Pohl et al. (39).

Table S3. Modern Tikal forest survey results

Upland or Bajo Proportional BA Upland or Bajo Proportional density

Upland UplandBrosimum alicastrum 30.4 Cryosophila stauracantha 16.6Blomia prisca 8.2 Blomia prisca 10.8Ficus spp. 6.7 Trichilia minutiflora 10.1Clusia spp. 5.6 Pouteria reticulata 10.0Pouteria reticulata 5.4 Brosimum alicastrum 7.2Manilkara zapota 5.0 Pseudolmedia glabrata 5.0Spondias mombin 4.5 Manilkara zapota 2.2Cedrela odorata 2.8 Sabal mauritiiformis 2.1Forchhammeria trifoliata 2.7 Forchhammeria trifoliata 2.0Trichilia minutiflora 2.6 Protium copal 1.9

Total (of the top 10) 73.9 Total (of the top 10) 67.9

Bajo BajoHaematoxylum campechianum 20.3 Croton billbergianus 23.9Sabal mauritiiformis 12.3 Haematoxylum campechianum 11.6Croton billbergianus 9.8 Cryosophila stauracantha 9.9Cedrela odorata 8.1 Metopium brownei 6.6Metopium brownei 6.4 Gymnanthes lucida 6.4Cupania belizensis 5.0 Manilkara zapota 5.2Manilkara zapota 4.1 Sabal mauritiiformis 5.0Simira salvadorensis 2.9 Margaritaria nobilis 3.5Gymnanthes lucida 2.6 Lonchocarpus heptaphyllus 3.3Cryosophila stauracantha 2.3 Nectandra spp. 2.8

Total (of the top 10) 73.7 Total (of the top 10) 78.2

Most abundant species by stem density (number of trees per ha−1) and dominant species by basal area (BA = m2/ha−1) for eachhabitat (6 cm diameter at breast height threshold) in our survey plots (5.95 ha). These are the oligarchic or most dominant tree speciesof the modern forest (40).

Forests, fields, and the edge of sustainability at the ancient Maya city of Tikal

Documents

16.2 – Maya Kings & Cities. Maya Lands stretched from...

Tikal Rulers

Y EL LEGADO MAYA EN TIKAL Y COPÁN

BELIZE TO TIKAL - Mass Audubon...BELIZE TO TIKAL REEFS,...

Guatemala Corazón del Mundo Maya... Volcan Acatenango Tikal...

Guatemala Lugares Turísticos. Tikal La riqueza...

Tikal timbers and temples: ancient Maya agroforestry and the...

Research at El Zotz, Guatemala In the Shadow of a Giant ·....

Guatemala con Tikal · llegada, traslado hacia uno de los.....

Latin America Past & Present. The Maya Central America Tikal...

Opazität in Gesellschaftsspielen Tikal, Monopoly ... ·...

Ancient Maya City and Protected Tropical Forests of ...

MAYA MATH, MAYA INDIANS AND INDIAN MAYA · many ancient...

Kingdom of the Maya Trip highlights · 2018-11-15 · Hotel...

A MArvel of MAyA engineering - World History 7B · 2018. 9....

Informe del Taller de epigrafía maya En el centro del mundo...