An 8700 year paleoclimate reconstruction from the southern Maya lowlands

An 8700 year paleoclimate reconstruction from the southern Maya lowlands

Quaternary Science Reviews 103 (2014) 19e25 Contents lists available at ScienceDirect Quaternary Science Reviews journal homepage: www.elsevier.com/...

1001KB Sizes 3 Downloads 24 Views

Quaternary Science Reviews 103 (2014) 19e25

Contents lists available at ScienceDirect

Quaternary Science Reviews journal homepage: www.elsevier.com/locate/quascirev

An 8700 year paleoclimate reconstruction from the southern Maya lowlands David Wahl a, b, *, Roger Byrne b, Lysanna Anderson a a b

U.S. Geological Survey, 345 Middlefield Rd. MSe975, Menlo Park, CA, 94025, USA Department of Geography, University of California, Berkeley, CA, 94720, USA

a r t i c l e i n f o

a b s t r a c t

Article history: Received 11 February 2014 Received in revised form 29 July 2014 Accepted 5 August 2014 Available online

Analysis of a sediment core from Lago Puerto Arturo, a closed basin lake in northern Peten, Guatemala, has provided an ~8700 cal year record of climate change and human activity in the southern Maya lowlands. Stable isotope, magnetic susceptibility, and pollen analyses were used to reconstruct environmental change in the region. Results indicate a relatively wet early to middle Holocene followed by a drier late Holocene, which we interpret as reflecting long-term changes in insolation (precession). Higher frequency variability is more likely attributable to changes in ocean/atmosphere circulation in both the North Atlantic and the Pacific Oceans. Pollen and isotope data show that most of the period of prehispanic agricultural settlement, i.e. ~5000e1000 cal yr BP, was characterized by drier conditions than previous or subsequent periods. The presence of Zea (corn) pollen through peak aridity during the Terminal Classic period (~1250e1130 cal yr BP) suggests that drought may not have had as negative an impact as previously proposed. A dramatic negative shift in isotope values indicates an increase in precipitation after ~950 cal yr BP (hereafter BP). Published by Elsevier Ltd.

Keywords: Paleoclimate Holocene Maya lowlands Oxygen isotopes Pollen Zea

1. Introduction Our understanding of Holocene climate change in the Maya lowlands has improved significantly over the last several decades. Studies have shown that precessional forcing of insolation caused reorganization of global atmospheric dynamics on millennial time scales (Haug et al., 2001; Fleitmann et al., 2003; Broccoli et al., 2006; Koutavas et al., 2006). Less clear, however, are the causes of higher frequency (centennial to decadal) changes recognized in many Holocene climate reconstructions (Mayewski et al., 2004; Thompson et al., 2006; Wanner and Beer, 2008). The drivers of climate dynamics in the Yucatan Peninsula have been difficult to identify in part because ocean-atmosphere systems from both the Atlantic and Pacific oceans are involved (Malmgren et al., 1998; Giannini et al., 2000; Czaja et al., 2002). Impacts of human populations in the region after ~5000 BP further complicate interpretation of paleoenvironmental records (Leyden, 1987; Islebe et al., 1996; Pohl et al., 1996; Wahl et al., 2006, 2007). Elucidating spatial and temporal patterns of past climate variability is essential

to assessing the role of interacting forcing mechanisms and improving the precision of global climate modeling predictions. Reconstructions of Holocene climate change in the Yucatan peninsula are particularly useful in testing hypotheses that extreme climate events, specifically droughts, negatively impacted the Maya civilization (Hodell et al., 1995; Gill, 2000; Haug et al., 2003; Hodell et al., 2005). Several paleoclimate datasets indicate that the Yucatan peninsula experienced persistent, multi-decadal droughts centered around 1000 BP (Curtis et al., 1996; Hodell et al., 2001, 2005; Medina-Elizalde et al., 2010; Kennett et al., 2012). Increased spatial coverage of well dated, high temporal resolution climate reconstructions is needed to resolve environmental changes with the precision necessary to correlate with Maya prehistory. Here we present a new high-resolution stable isotope record that reflects changes in moisture balance along with previously reported pollen and magnetic susceptibility data (Wahl et al., 2006, 2007). Our research provides an ~8700 year record of climate change and anthropogenic activity from the southern Maya lowlands, in what is now northern Peten, Guatemala (Fig. 1). 2. Physical setting

* Corresponding author. 345 Middlefield Rd. MSe975, Menlo Park, CA, 94025, USA. Tel.: þ1 650 329 4533; fax: þ1 650 329 4936. E-mail address: [email protected] (D. Wahl). http://dx.doi.org/10.1016/j.quascirev.2014.08.004 0277-3791/Published by Elsevier Ltd.

The southern Maya lowlands are well known as the area in which the Maya civilization developed. Archaeological evidence

20

D. Wahl et al. / Quaternary Science Reviews 103 (2014) 19e25

3. Methods

Fig. 1. Map of the Yucatan peninsula showing the locations of Lago Puerto Arturo and other sites mentioned in the text. Hatched area indicates the Mirador Basin archaeological zone. LPA ¼ Lago Puerto Arturo, LPI ¼ Lago Peten-Itza, LY ¼ Laguna Yaloch, PL ¼ Punta Laguna, LC ¼ Lake Chichancanab. Dashed lines indicate the boundary of the Maya lowlands.

shows that the earliest settlement occurred ~3350 BP in the Mirador Basin, an archaeological zone in the southern Maya lowlands contained within a poorly defined drainage basin generally flowing to the west/northwest (Fig. 1; Hansen, 1991). From ~2350 to 1800 BP, numerous urban centers that supported extensive populations developed here (e.g., El Mirador, Nakbe, Tintal, and Xulnal). The area was largely abandoned at the end of the Preclassic Period, around 1800 BP. Smaller populations persisted during the Classic period (1650e1050 BP) until the area was permanently abandoned between 1150 and 1050 BP. The modern climate of Peten is strongly seasonal. More than 90% of the annual precipitation (1500e2000 mm) falls during the rainy season between April and November. Seasonal variability is controlled by the strength and positions of the North Atlantic High and the inter-tropical convergence zone (ITCZ). Interannual variability of climate appears to be driven by a combination of the ~ o/Southern OscilNorth Atlantic Oscillation (NAO) and the El Nin lation (ENSO) (Malmgren et al., 1998; George and Saunders, 2001; Giannini et al., 2001). The dominant geomorphic feature of northern Peten is the extensive system of poorly drained depressions locally known as bajos. These seasonally inundated areas account for over half the surface area of the region. In the Mirador Basin, approximately 60% of the surface is flooded annually with up to 1 m of water, some portion of which appears to drain into the Rio Candelaria to the northwest. Upland areas between the bajos have a subdued relief, rarely rising more than 20 m above bajo surfaces. The vegetation of Peten is broadly characterized as a semi-deciduous closed canopy tropical forest, although the bajo forests are more open and have a more herbaceous understory than upland forests (Lundell, 1934, 1937).

In 2001, we recovered a 7.28 m sediment core from Lago Puerto Arturo (17 32’ N, 9011’ W; Fig. 1), a small closed basin lake (~120 ha) near the western margin of the Mirador Basin. Cores, including replicates, were retrieved using a Livingstone piston corer lined with butyrate tubing. Core lithology, sediment properties, magnetic susceptibility, and pollen methods and results were reported previously (Wahl et al., 2006, 2007). Extended scans of pollen slides at 125  magnification were conducted to establish the presence or absence of Zea pollen. Up to three slides were examined from each pollen sample; once Zea pollen was encountered no additional slides were scanned. If Zea pollen was not encountered after scanning three slides it was considered not present. Zea mays (corn) was differentiated from other Poaceae pollen by size, long axis/pore ratio and phase contrast light microscopy (Irwin and Barghoorn, 1965; Whitehead and Langham, 1965). The minimum long axis measurement used for determining Zea was 60 mm (range: 60e100 mm, x ¼ 68 mm); the long axis/pore ratios fell between 5 and 9. Nomarski phase contrast microscopy was used on several Zea grains to verify the spacing of intertectile columella, a characteristic that distinguishes Zea from other large Poaceae pollen. It is unlikely that these grains represent teosinte given that the study area is well outside its known range (Doebley, 1990). A revised core chronology, based on previous published accelerator mass spectrometry (AMS) radiocarbon determinations, was developed using CLAM 2.2 (Wahl et al., 2006; Blaauw, 2010). The mean difference in ages between models is ~35 years, and the revised age model does not change previous interpretations of Lago Puerto Arturo data. A third order polynomial was fitted to the uncalibrated 14C dates (Fig. 2). The model uses 10,000 iterations based on the probability curves of the calibrated dates (IntCal13.14C; Reimer et al., 2013). Dates are assigned to specific core depths based on the weighted average of the model estimates of all iterations falling with in the 95% confidence interval. Three dates exhibiting age reversals were not included in the age model.

Fig. 2. Age-depth model based on seven radiocarbon determinations and produced using CLAM 2.2 (Blaauw, 2010).

D. Wahl et al. / Quaternary Science Reviews 103 (2014) 19e25

Changes in18O/16O of biogenic carbonate in closed basin Yucatecan lakes have been used to identify changes in precipitation and evaporation (Covich and Stuiver, 1974; Hodell et al., 1995; Curtis et al., 1996). Lake water becomes enriched in 18O (relative to source water) as 16O is selectively evaporated during periods of high evaporation/low precipitation; changes in lake water d18O values are recorded in carbonate shells as they precipitate. Oxygen isotope ratios were measured on a monogeneric set of gastropod shells (Pyrgophorus sp.; Curtis et al., 1996, 1999). A small sub-set of shells lacked spinose features and, although they may have been a smooth variety of Pyrgophorus coronatus (Covich, 1976), the authors limited identification to the generic level. Two hundred and sixty three samples were analyzed using an Micromass Optima isotope ratio mass spectrometer. Multiple adult individuals were selected to create an aggregate for each sample. Each sample spans 1 cm of deposition and represents, on average, ~15 years. Average sampling interval is ~30 years, with the exception of three sample gaps of 247, 162, and 239 years, centered on 5380 BP, 4640 BP, and 560 BP, respectively. A regime shift index (RSI) algorithm was used to identify the timing and magnitude of regime shifts in the d18O (Rodionov, 2004). Regime shift is defined here as a rapid change in the data from one period characterized by a specific mean and variance to a period with a statistically significant difference in mean and variance. A sequential Students t-test was applied to determine the magnitude of the change required for statistically significant shifts. The data were analyzed using a ten-year minimum window width (cutoff ¼ 10) and a 0.10 confidence interval. Outliers were identified conservatively as residuals greater than one standard deviation from the mean. A weighted mean was calculated using the Huber's weight function (Huber, 1964). The weighted mean was used in the significance test for regime shift. 4. Results and discussion 4.1. Holocene climate variability Relatively heavy isotopic values at the base of the core likely reflect larger surface area to volume associated with initial infilling of the lake, followed by persistently lighter values/wetter conditions through the early Holocene (Fig. 3). Mean values increase after ~6250 BP, continuing to rise until around 4500 BP. The RSI reflects this transition to a drier regime. Values are variable from ~4500 to 950 BP, yet indicate relatively dry conditions compared to before or after. Values decrease rapidly after 950 BP, marking a transition to the wettest conditions since 6250 BP. Changes in vegetation appear to have strongly influenced lake n Itza and water isotope ratios in two nearby Peten lakes, Pete n (Rosenmeier et al., 2002). Results from these sites suggest Salpete lighter d18O values reflect enhanced surface and groundwater input associated with forest clearance. The Lago Puerto Arturo results do not indicate such a response to local opening of the forest, which had occurred by 3000 BP. Heavier values during the period of anthropogenic disturbance in the watershed, however, may reflect, in part, a localized climatic response to widespread forest clearance (Lean and Warrilow, 1989; Shaw, 2003; Manoharan et al., 2009). The aquatic pollen evidence supports our interpretation of the isotope evidence (Fig. 4). The Cyperaceae (sedge)/Nymphaea (water-lily) pollen index is a normalized difference index calcutated as (CeN)/(C þ N), where C is percent Cyperaceae and N is percent Nymphaea (Rouse et al., 1974; Mensing et al., 2006). The center of Lago Puerto Arturo is currently relatively shallow (<1e2 m), and supports a dense growth of sedges. During drier periods with lower water levels, sedges expand into water-lily habitat. Thus, the Cyperaceae/Nymphaea index is interpreted to reflect the expansion

(A)

21

(B)

Fig. 3. A) Lago Puerto Arturo oxygen isotope data. Dashed line indicates raw data and the black line represents a 5 point running mean; B) Weighted mean values of regimes identified using the Regime Shift Index (RSI) algorithm.

and contraction of sedges into water lily habitat. Higher percentages of Cyperaceae pollen relative to Nymphaea coincide with periods of increased d18O values. The broader trends in the isotope data are, in many ways, similar to other Holocene paleoclimate records from the Yucatan. For example, several pollen and geochemical studies suggest that the climate of the area became more humid by 10,350 BP (Leyden, 2002; Hillesheim et al., 2005). Many small lakes began to fill in between 9000 and 8000 BP (Hodell et al., 1995; Leyden, 2002; Wahl et al., submitted for publication). Increased summer insolation in the early to middle Holocene would have warmed tropical North Atlantic SSTs, increased the cross-equatorial temperature gradient, slackened the NE tradewinds, and drawn the ITCZ to the north. The resulting increase in latent heat flux and rise in the height of the tradewind inversion would have resulted in increased convective precipitation. A more northern mean position of the ITCZ and wetter conditions are reported for the circum-Caribbean region in the early to middle Holocene (Hodell et al., 1991; Haug et al., 2001). Shin et al. (2006) suggested that orbitally forced increased northern hemisphere seasonality around 6000 BP amplified cross equatorial Atlantic SST differences and moved the tropical Atlantic ITCZ northward. In the modern climate system, this climatology, ~ a-like conditions in the tropical Pacoupled with persistent La Nin cific Ocean (i.e., more northern mean position of the ITCZ) results in increased precipitation in Central America. A wet early to middle Holocene followed by a relatively dry late Holocene is reported from the Caribbean, Atlantic, and Central America (Hodell et al., 1991; Haug et al., 2001; Poore et al., 2003; Dull, 2004; Gischler and Storz, 2009; Lane et al., 2009). Desertification in North Africa marked the end of the African Humid Period around 5500 BP (Claussen et al., 1999; deMenocal et al., 2000). The timing of transition to drier conditions on both sides of the North Atlantic, while variable, is centered around 5500 BP. Temporal differences between sites are possibly due to error inherent in radiocarbon based age-models, though temporal lags in climate response were also a factor given that the maximum seasonal insolation disparity at 18 N occurred around 9000 BP.

22

D. Wahl et al. / Quaternary Science Reviews 103 (2014) 19e25

Fig. 4. Oxygen isotope, selected pollen, and magnetic susceptibility data from Lago Puerto Arturo. The Cyper/Nymph index is calculated (CeN)/(C þ N), where C ¼ percent Cyperaceae and N ¼ percent Nymphaea pollen. Weedy taxa represent pollen from the Asteraceae and Poaceae families. Stars indicate pollen samples with Zea pollen present; diamond indicates earliest Zea pollen found in the Maya lowlands to date (Pohl et al., 1996). Zones follow Wahl et al. (2006) and mark significant changes in proxy data.

Comparison of the Lago Puerto Arturo stable isotope record with Holocene length SST reconstructions from the Atlantic and Gulf of Mexico suggests that longer-term changes in Yucatan precipitation are related to changes in subtropical SSTs (Fig. 5). The shift to less moisture in the late Holocene likely resulted from decreased northern hemisphere summer insolation and a reversal of the ocean-atmosphere dynamics described above, leading to cooler SSTs, a more southern position of pressure systems (ITCZ and North Atlantic High), and lowering of the trade wind inversion (Lane and Horn, 2013). Increased ENSO activity after 3500 BP, also associated with a more southern ITCZ, likely began contributing to variability on the Yucatan peninsula in the late Holocene. Warmer dry season temperatures caused by increased winter insolation would have enhanced evaporative stress, further decreasing annual moisture availability. A positive excursion in isotopic values at Lago Puerto Arturo and a marked shift in the RSI centered around 3000 BP appear to reflect the driest period since 8250 BP (Fig. 3). Anomalously dry conditions around this time have also been reported from other Peten lakes as well. Coring at Laguna Yaloch (Fig. 1) was terminated by stiff (possibly terrestrial) deposits that date to around 3400 BP (Wahl et al., 2013). Grass pollen percentages in the basal sediment of the Laguna Yaloch core are similar to that of local savannas (Bhattacharya et al., 2011), suggesting that the low-lying floodplain around the lake may have been drier at this time. Mueller et al. (2009) reported a drier climate beginning around 4500 and peaking around 3500 BP. Their evidence is based on CaCO3 deposition and the ratio of Ca to elements assumed to represent clastic input to the lake (Ti, Fe, and Al). This model assumes that drier conditions (higher evaporation/precipitation (E/P)) will reduce erosion and increase the precipitation rate of authigenic carbonate. They conclude that wetter conditions abruptly returned after 3000 BP based on a dramatic drop in CaCO3 and the ratio of Ca to the sum of terrestrial elements (Ti, Fe, Al). An alternative interpretation of

these data is that they reflect a dilution of CaCO3 by increased deposition of terrestrial material (i.e. Unit U2, the “Maya Clay”) due to human activity and associated erosion, as opposed to an increase in precipitation. Mueller et al. (2009) noted evidence for widespread drying around 3500 BP on both sides of the Atlantic and suggested a southward shift of the ITCZ may have been responsible. A more southerly ITCZ in the Atlantic drives a wind-evaporation-SST mode characterized by increased NE tradewinds, increased surface heat flux, and decreased SSTs. This climatology, reminiscent of a positive phase of the North Atlantic Oscillation (þNAO), results in decreased Caribbean/Central American precipitation and is often associated ~ o/Southern Oscillation (þENSO) with a positive phase of the El Nin in the Pacific (Giannini et al., 2001). Positive ENSO activity, though not the primary driver of interannual variability, can lead to decreased precipitation in the Caribbean. Giannini et al. (2001) suggested that impacts of ENSO in the region are largely determined by the phase of the NAO in such a way that þ NAO conditions coupled with þENSO result in decreased precipitation. Reconstructions of ENSO activity from the Atlantic and Pacific basins show increased frequency and intensity of þENSO events between 3500 and 3000 BP (Haug et al., 2001; Moy et al., 2002), which have been interpreted to reflect a more southern mean position of the ITCZ. Dry conditions near Lago Puerto Arturo at 3000 BP coincide with Ti decreases in the Cariaco Basin core at 3000 and 2800 BP. These signals may be associated with anomalously strong þENSO. Evidence for cooler SSTs around 3000e3500 BP comes from the Gulf of Mexico and may indicate a widespread cooling event (Poore et al., 2003). In the Gulf of Mexico, high concentrations of Globorotalia crassaformis (a planktonic foraminifer interpreted as reflecting cooler water) at ~3200 BP led the authors to conclude the presence of a “regional scale environmental event” (Poore et al., 2003). That this cooling event was widespread is corroborated by a tree ring based terrestrial temperature reconstruction from

D. Wahl et al. / Quaternary Science Reviews 103 (2014) 19e25

23

Fig. 5. Comparison of Lago Puerto Arturo isotope data (red line) with A) Loop Current strength reconstruction from Gulf of Mexico core GOM-MD2553 (Poore et al., 2003) GOM ¼ Gulf of Mexico; B) SST reconstruction from ODP Hole 658C off the coast of west Africa; *values decline to 18.3  C. (Zhao et al., 1995; deMenocal et al., 2000); C) Latitudinal shifts in the mean position of the ITCZ inferred from Ti concentrations in Cariaco Basin sediment (Haug et al., 2001), and D) reconstructions of ENSO activity from Laguna Pallcacocha (solid black line represents number of events per 100 years; Moy et al., 2002) and El Junco Lake (dashed gray line indicates percent sand; Conroy et al., 2008). SML¼ Southern Maya lowlands, E/P ¼ evaporation/precipitation. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)

northern Finland indicating ~3400 BP was the coolest period of the last 7600 years (Helama et al., 2010). Their results also show a positive correlation between solar activity and regional temperature at bi-millennial timescales during the middle Holocene, suggesting external drivers of climate were in play at this time. Inferred cooler terrestrial and marine temperatures across the North Atlantic basin may be related to Bond event 2, which is characterized by excursions in the GISP2 ice core and increased concentrations of ice rafted debris (Bond et al., 2001). These studies point to solar activity as the primary driver of millennial-centennial climate variability in the North Atlantic region, though it is unclear how direct the solar-climate link is. The annually resolved reconstruction of Helama et al. (2010) shows little to no time lag between changes in solar activity and shifts in northern European temperatures, leading to the conclusion that solar activity may have driven climate directly through atmospheric processes. Coupled general circulation models, however, show a more southern position of the Atlantic ITCZ can be associated with increased north Atlantic sea ice and attenuated thermohaline circulation (Manabe and Stouffer, 1988; Chiang, 2005), leaving the question of internal vs. external forcing of tropical Atlantic climate open. The Lago Puerto Arturo isotope record shows drier conditions around 1750 BP, coincident with the end of the Preclassic period. Archaeological evidence indicates that many large Preclassic centers experienced decreased population, and/or were abandoned around this time (Dunning et al., 2013). Paleoclimate

reconstructions from Yucatan and the circum-Caribbean suggest drier conditions from ~1800 to 1700 BP, leading to proposals that climate may have had an effect on cultural processes in the Maya lowlands at the end of the Preclassic period (Dahlin, 1983; Hansen, 1990; Gunn et al., 1995; Gill, 2000; Hodell et al., 2001; Haug et al., 2003; Kennett et al., 2012; Dunning et al., 2013). Proxy evidence for agricultural disturbance from Lago Puerto Arturo suggests a local abandonment at this time (Wahl et al., 2007); the isotope data show a dry excursion coinciding with the abandonment (Fig. 4). While there is little evidence for significant variability in the north Atlantic ~1750 BP, paleoclimate and SST reconstructions from the eastern tropical Pacific suggest a strong þ ENSO state (Moy et al., 2002; Woodroffe et al., 2003; Conroy et al., 2008). Several geochemical studies from the Maya lowlands indicate a persistent drier climate associated with the Terminal Classic (Hodell et al., 1995; 2001, Curtis et al., 1996; Webster et al., 2007; Medina-Elizalde et al., 2010; Kennett et al., 2012; Wahl et al., 2013). Relatively heavy d18O values at Lago Puerto Arturo provide corroborating evidence of dry conditions from ~1250 to 900 BP, peaking between ~1250 and 1100 BP, though the isotope signal does not indicate anomalous drought at this time. 4.2. Climate and agriculture The pollen record from Lago Puerto Arturo shows evidence of agricultural activity from ~4600 to 1100 BP (Fig. 4). Zea mays pollen

24

D. Wahl et al. / Quaternary Science Reviews 103 (2014) 19e25

first appears ~4600 BP, after which it is absent until ~3500 BP. Sporadic occurrences of very early Zea pollen are not uncommon in the Maya region (Rue, 1987; Pohl et al., 1996); current research aimed at concentrating rare Zea pollen from several sediment cores dating this time period, using the techniques of Whitney et al. (2012), will help clarify the chronology of early agriculture in the region. After 3500 BP, Zea pollen is commonly present until ~1100 BP, after which it disappears permanently. Pollen from disturbance taxa (Poaceae and Asteraceae) show an abrupt and sustained increase during this period. Although these increases may in part reflect the drier climate discussed above, the presence of Zea mays during this period indicates anthropogenic activity. Magnetic susceptibility (a proxy for watershed erosion) shows peak values corresponding to peaks in agricultural pollen, further tying environmental change at this time to human agency. The pattern of decreased forest/increased disturbance taxa, together with the presence of Zea pollen, has been reported from several sites across the southern Maya lowlands and can be considered an “ecological signature” of prehispanic land use practices. The highest values of pollen types indicating disturbance occur during the Late Preclassic (2250e1700 BP), when the area was most heavily occupied (Hansen, 1990, 1998; Wahl et al., 2007). The presence of Zea and weedy pollen types through peak aridity during the Late Classic indicates nearby settlement and agriculture, and suggests that drought may not have caused abandonment throughout the southern Maya lowlands. Similar evidence for persistence of settlement and agricultural activity through the Late Classic dry period has been reported from eastern Peten (Wahl et al., 2013). Agricultural activity around Lago Puerto Arturo ultimately ended ~1100 cal yr BP (Fig. 4). Within 40 years weedy taxa drop to pre-agricultural values and afforestation is underway, suggesting complete abandonment of the area (Wahl et al., 2007). One important implication of these results is that the relatively dry climate of the late Holocene may have offered favorable conditions for agriculture in the southern Maya lowlands. The relationship between drier conditions and agriculture has been noted previously (Curtis et al., 1996). The earliest evidence for agriculture in lowland Mesoamerica comes from fossil maize pollen recovered from riverine sites (Pohl et al., 1996; Pope et al., 2001), providing evidence in support of hypotheses that early maize agriculture in the Maya lowlands may have been based on dry season floodplain margin strategies (Puleston and Puleston, 1971). Floodplain strategies could have easily been applied to the edges of extensive wetlands in the interior of Yucatan, particularly the ubiquitous bajo systems currently found in northern Peten. A drier late Holocene climate may have created a more favorable environment for these strategies relative to more humid conditions in the middle Holocene. Further investigation into the temporal and spatial details of prehispanic agricultural strategies employed in the area is needed to fully understand the implication of these results. 5. Conclusions The Lago Puerto Arturo d18O based climate reconstruction provides new insight into Holocene climate change on the Yucatan Peninsula. Lighter d18O values in the early to middle Holocene reflect a moisture balance characterized by a relatively low E/P. After ~6250 BP, climate appears to have become drier. Millennial variability reflects orbitally forced changes in insolation, resulting in a southward shift of the ITCZ and associated decreased SSTs. Comparison of our results with paleoclimate and paleoceonographic reconstructions from the north Atlantic, Gulf of Mexico, and eastern Pacific suggests Atlantic variability was the primary driver of Yucatan climate during the middle Holocene. In contrast, variability during the late Holocene was driven by regional ocean-

atmosphere dynamics in both the Atlantic and Pacific oceans, reflecting increased influence of ENSO activity. Proxy evidence of widespread agricultural activity in the southern Maya lowlands is constrained to a period of overall drier conditions during the late Holocene. This suggests that a drier climate may have favored the development and expansion of sedentary agriculture. A prominent dry phase around 1750 BP correlates with evidence of local abandonment, raising the possibility that drought may have played a role in regional demographic shifts associated with the Preclassic/Classic period transition. The Terminal Classic droughts recorded elsewhere in the Yucatan and circum-Caribbean are weakly evidenced in the Lago Puerto Arturo isotope data. The presence of Zea pollen and pollen evidence of ecological disturbance during the Terminal Classic droughts, however, indicate that the area around the lake was not permanently abandoned until after peak aridity. Acknowledgments This research was funded by the National Science Foundation (DDIG #0327305), the Foundation for Anthropological Research and Environmental Studies (FARES), and the Mirador Basin/Cuenca Mirador project. We are grateful to Richard Hansen for facilitating field logistics and providing thoughtful discussion, Tom Schreiner and Oscar Tun for assisting ably with fieldwork, and the Instituto de Antropología e Historia for cooperative support. We thank John Barron and Jenn Kusler for valuable comments on earlier versions of this manuscript. References Bhattacharya, T., Beach, T., Wahl, D., 2011. An analysis of modern pollen rain from the Maya lowlands of northern Belize. Rev. Palaeobot. Palynol. 164 (1e2), 109e120. Blaauw, M., 2010. Methods and code for 'classical' age-modelling of radiocarbon sequences. Quat. Geochronol. 5 (5), 512e518. Bond, G., Kromer, B., Beer, J., Muscheler, R., Evans, M.N., Showers, W., Hoffman, S., Lotti-Bond, R., Hajdas, I., Bonani, G., 2001. Persistent solar influence in North Atlantic climate during the Holocene. Science 294, 2130e2136. Broccoli, A., Dahl, K., Stouffer, R., 2006. Response of the ITCZ to Northern Hemisphere cooling. Geophys. Res. Lett. 33 (1). Chiang, J.C.H., 2005. Present-day climate variability in the tropical Atlantic: a model for paleoclimate changes? In: Diaz, H.F., Bradley, R.S. (Eds.), The Hadley Circulation: Present, Past and Future. Kluwer Academic Publishers, Dordrecht/Boston/London, pp. 465e488. Claussen, M., Kubatzki, C., Brovkin, V., Ganopolski, A., 1999. Simulation of an abrupt change in Saharan vegetation in the Mid-Holocene. Geophys. Res. Lett. 26 (14), 2037e2040. Conroy, J.L., Overpeck, J.T., Cole, J.E., Shanahan, T.M., Steinitz-Kannan, M., 2008. pagos Holocene changes in eastern tropical Pacific climate inferred from a Gala lake sediment record. Quat. Sci. Rev. 27 (11), 1166e1180. Covich, A., Stuiver, M., 1974. Changes in oxygen 18 as a measure of long-term fluctuations in tropical lake levels and Molluscan populations. Limnol. Oceanogr. 19 (4), 682e691. Covich, A., 1976. Recent changes in Molluscan species diversity of a large tropical lake (Lago de Peten, Guatemala). Limnol. Oceanogr. 21, 51e59. Curtis, J.H., Hodell, D.A., Brenner, M., 1996. Climate variability on the Yucatan Peninsula (Mexico) during the past 3500 years, and implications for Maya cultural evolution. Quat. Res. 46, 37e47. Curtis, J.H., Brenner, M., Hodell, D.A., 1999. Climate change in the Lake Valencia Basin, Venezuela, ~12600 yr BP to present. Holocene 9, 609e619. Czaja, A., Van Der Vaart, P., Marshall, J., 2002. A diagnostic study of the role of remote forcing in tropical Atlantic variability. J. Clim. 15, 3280e3290. Dahlin, B.H., 1983. Climate and prehistory on the Yucatan Peninsula. Clim. Change 5, 245e263. deMenocal, P., Ortiz, J., Guilderson, T., Adkins, J., Sarnthein, M., Baker, L., Yarusinsky, M., 2000. Abrupt onset and termination of the African Humid Period: rapid climate responses to gradual insolation forcing. Quat. Sci. Rev. 19, 347e361. Doebley, J., 1990. Molecular evidence and the evolution of Maize. Econ. Bot. 44 (3 Suppl.), 6e27. Dull, R.A., 2004. An 8000-year record of vegetation, climate, and human disturbance from the Sierra de Apaneca, El Salvador. Quat. Res. 61, 159e167. Dunning, N., Wahl, D., Beach, T., Jones, J., Luzzadder-Beach, S., McCane, C., 2013. The end of the beginning: drought, environmental change, and the preclassic to classic transition in the East-central Maya lowlands. In: Iannone, G. (Ed.), The

D. Wahl et al. / Quaternary Science Reviews 103 (2014) 19e25 Great Maya Droughts in Cultural Context: Case Studies in Resilience and Vulnerability. University Press of Colorado, Boulder, CO, pp. 115e138. Fleitmann, D., Burns, S.J., Mudelsee, M., Neff, U., Kramers, J., Mangini, A., Matter, A., 2003. Holocene forcing of the Indian monsoon recorded in a Stalagmite from Southern Oman. Science 300 (5626), 1737e1739. George, S.E., Saunders, M.A., 2001. North Atlantic Oscillation impact on tropical north Atlantic winter atmospheric variability. Geophys. Res. Lett. 28 (6), 1015e1018. Giannini, A., Cane, M.A., Kushnir, Y., 2001. Interdecadal changes in the ENSO teleconnection to the Caribbean region and the North Atlantic Oscillation. J. Clim. 14, 2867e2879. Giannini, A., Kushnir, Y., Cane, M.A., 2000. Interannual variability of Caribbean rainfall, ENSO and the Atlantic Ocean. J. Clim. 13, 297e311. Gill, R.B., 2000. The Great Maya Droughts: Water, Life, and Death. University of New Mexico Press, Albuquerque, p. 484. Gischler, E., Storz, D., 2009. High-resolution windows into Holocene climate using proxy data from Belize corals (Central America). Palaeobiodivers. Palaeoenviron. 89 (3e4), 211e221. Gunn, J.D., Folan, W.J., Robichaux, H.R., 1995. A landscape analysis of the Candelaria watershed in Mexico: insights into paleoclimates affecting upland horticulture in the southern Yucatan peninsula semi-karst. Geoarchaeology 10 (1), 3e42. Hansen, R.D., 1990. Excavations in the Tigre Complex. Utah, Papers of the New World Archaeological Foundation, El Mirador Series, Part 3. El Mirador, Peten, Guatemala, Provo, p. 308. Hansen, R.D., 1991. The road to Nakbe. Nat. Hist. N5, 8e14. Hansen, R.D., 1998. Continuity and Disjunction: The Pre-Classic Antecedents of Classic Maya Architecture. In: Houston, S.D. (Ed.), Function and Meaning in Classic Maya Architecture. Dumbarton Oaks, Washington D.C., pp. 49e122. Haug, G.H., Gunther, D., Peterson, L.C., Sigman, D.M., Hughen, K.A., Aeschlimann, B., 2003. Climate and the collapse of Maya civilization. Science 299 (5613), 1731e1735. €hl, U., 2001. Southward Haug, G.H., Hughen, K.A., Sigman, D.M., Peterson, L.C., Ro migration of the intertropical convergence zone through the Holocene. Science 293, 1304e1308. Helama, S., Fauria, M.M., Mielik€ ainen, K., Timonen, M., Eronen, M., 2010. SubMilankovitch solar forcing of past climates: mid and late Holocene perspectives. Geol. Soc. Am. Bull. 122 (11e12), 1981e1988. Hillesheim, M.B., Hodell, D.A., Leyden, B.W., Brenner, M., Curtis, J.H., Anselmetti, F.S., Ariztegui, D., Buck, D.G., Guilderson, T.P., Rosenmeier, M.F., Schurrenberger, D.W., 2005. Climate change in lowland Central America during the late deglacial and early Holocene. J. Quat. Sci. 20 (4), 363e376. Hodell, D.A., Brenner, M., Curtis, J.H., Guilderson, T., 2001. Solar forcing of drought frequency in the Maya lowlands. Science 292 (5520), 1367e1370. Hodell, D.A., Curtis, J.H., Brenner, M., 1995. Possible role of climate in the collapse of Classic Maya civilization. Nature 375, 391e394. Hodell, D.A., Brenner, M., Curtis, J.H., 2005. Terminal Classic drought in the northern Maya lowlands inferred from multiple sediment cores in Lake Chichancanab (Mexico). Quat. Sci. Rev. 26, 1413e1427. Hodell, D.A., Curtis, J.H., Jones, G.A., Higuera-Gundy, A., Brenner, M., Binford, M.W., Dorsey, K.T., 1991. Reconstruction of Caribbean climate change over the past 10,500 years. Nature 352, 790e793. Huber, P.J., 1964. Robust estimation of a location parameter. Ann. Math. Stat. 35 (1), 73e101. Islebe, G.A., Hooghiemstra, H., Brenner, M., Curtis, J.H., Hodell, D.A., 1996. A Holocene vegetation history from lowland Guatemala. Holocene 6 (3), 265e271. Irwin, H., Barghoorn, E.S., 1965. Identification of the pollen of Maize, Teosinte and Tripsacum by phase contrast microscopy. Botanical Museum Leaflets, vol. 21 no. 2. Harvard University, pp. 37e47. Kennett, D.J., Breitenbach, S.F., Aquino, V.V., Asmerom, Y., Awe, J., Baldini, J.U., Bartlein, P., Culleton, B.J., Ebert, C., Jazwa, C., 2012. Development and disintegration of Maya political systems in response to climate change. Science 338 (6108), 788e791. Koutavas, A., deMenocal, P., Olive, G.C., Lynch-Steiglitz, J., 2006. Mid-Holocene El ~ o-Southern Oscillation (ENSO) attenuation revealed by individual foramiNin nifera in Eastern Tropical Pacific sediments. Geology 34 (12), 993e996. Lane, C.S., Horn, S.P., Mora, C.I., Orvis, K.H., 2009. Late-Holocene paleoenvironmental change at Mid-elevation on the Caribbean Slope of the Cordillera Central, Dominican Republic: a multi-site, multi-proxy analysis. Quat. Sci. Rev. 28, 2239e2260. Lane, C.S., Horn, S.P., 2013. Terrestrially derived n-alkane dD evidence of shifting Holocene paleohydrology in Highland Costa Rica. Arct. Antarct. Alp. Res. 45 (3), 342e349. Lean, J., Warrilow, D., 1989. Simulation of the regional climatic impact of Amazon deforestation. Nature 342 (6248), 411e413. Leyden, B.W., 1987. Man and climate in the Maya lowlands. Quat. Res. 28, 407e414. Leyden, B.W., 2002. Pollen evidence for climatic variability and cultural disturbance in the Maya lowlands. Anc. Mesoam. 13, 85e101. Lundell, C.L., 1934. Preliminary Sketch of the Phytogeography of the Yucatan Peninsula. Publication 436, Contribution No. 12. Carnegie Institution of Washington, Washington D.C. Lundell, C.L., 1937. The Vegetation of Peten. Publication 478. Carnegie Institution of Washington, Washington D.C, p. 244. ~ o-Southern Oscillation and North Malmgren, B.A., Winter, A., Chen, D., 1998. El Nin Atlantic Oscillation control of climate in Puerto Rico. J. Clim. 11 (10), 2713e2718. Manabe, S., Stouffer, R., 1988. Two stable equilibria of a coupled ocean-atmosphere model. J. Clim. 1 (9), 841e866.

25

Manoharan, V.S., Welch, R.M., Lawton, R.O., 2009. Impact of deforestation on regional surface temperatures and moisture in the Maya lowlands of Guatemala. Geophys. Res. Lett. 36 (21). n, W., Maasch, K.A., David Mayewski, P.A., Rohling, E.E., Curt Stager, J., Karle Meeker, L., Meyerson, E.A., Gasse, F., van Kreveld, S., Holmgren, K., 2004. Holocene climate variability. Quat. Res. 62 (3), 243e255. Medina-Elizalde, M., Burns, S.J., Lea, D.W., Asmerom, Y., von Gunten, L., Polyak, V., Vuille, M., Karmalkar, A., 2010. High resolution stalagmite climate record from n Peninsula spanning the Maya terminal classic period. Earth Planet. the Yucata Sci. Lett. 298 (1e2), 255e262. Mensing, S., Livingston, S., Barker, P., 2006. Long-term fire history in Great Basin sagebrush reconstructed from macroscopic charcoal in spring sediments, Newark Valley, Nevada. West. North Am. Nat. 66, 64e77. ~ o/Southern Moy, C., Seltzer, G., Rodbell, D., Anderson, D., 2002. Variability of El Nin Oscillation activity at millennial timescales during the Holocene epoch. Nature 420 (6912), 162e165. Mueller, A.D., Islebe, G.A., Hillesheim, M.B., Grzesik, D.A., Anselmetti, F.S., Ariztegui, D., Brenner, M., Curtis, J.H., Hodell, D.A., Venz, K.A., 2009. Climate drying and associated forest decline in the lowlands of northern Guatemala during the late Holocene. Quat. Res. 71, 133e141. Pohl, M.D., Pope, K.O., Jones, J.G., Jacob, J.S., Piperno, D.R., deFrance, S.D., Lentz, D.L., Gifford, J.A., Danforth, M.E., Josserand, J.K., 1996. Early agriculture in the Maya lowlands. Lat. Am. Antiq. 7 (4), 355e372. Poore, R., Dowsett, H., Verardo, S., Quinn, T., 2003. Millennial-to century-scale variability in Gulf of Mexico Holocene climate records. Paleoceanography 18 (2), 1048. Pope, K.O., Pohl, M.E., Jones, J.G., Lentz, D.L., von Nagy, C., Vega, F.J., Quitmyer, I.R., 2001. Origin and environmental setting of ancient agriculture in the lowlands of Mesoamerica. Science 292, 1370e1373. Puleston, D.E., Puleston, O.S., 1971. An ecological approach to the origins of Maya civilization. Archaeology 24, 330e337. Reimer, P.J., Bard, E., Bayliss, A., Beck, J.W., Blackwell, P.G., Bronk Ramsey, C., , C., Heaton, T.J., Grootes, P.M., Guilderson, T.P., Haflidason, H., Hajdas, I., Hatte Hoffmann, D.L., Hogg, A.G., Hughen, K.A., Kaiser, K.F., Kromer, B., Manning, S.W., Niu, M., Reimer, R.W., Richards, D.A., Scott, E.M., Southon, J.R., Staff, R.A., Turney, C.S.M., van der Plicht, J., 2013. IntCal13 and Marine13 radiocarbon age calibration curves 0e50,000 years cal BP. Radiocarbon 55, 1869e1887. Rodionov, S.N., 2004. A sequential algorithm for testing climate regime shifts. Geophys. Res. Lett. 31, L09204. Rosenmeier, M.F., Hodell, D.A., Brenner, M., Curtis, J.H., Martin, J.B., Anselmetti, F.S., Ariztegui, D., Guilderson, T.P., 2002. Influence of vegetation change on watershed hydrology: implications for paleoclimatic interpretation of lacustrine d18O records. J. Paleolimnol. 27, 117e131. Rouse, J.W., Haas, R.H., Schell, J.A., Deering, D.W., 1974. Monitoring the Vernal Advancement and Retrogradation (Green Wave Effect) of Natural Vegetation. Prog. Rep. RSC 1978-1. Remote Sensing Center, Texas A&M Univ., College Station, p. 164. Rue, David J., 1987. Early agriculture and early Postclassic Maya occupation in western Honduras. Nature 326, 285e286. Shaw, J.M., 2003. Climate change and deforestation: implications for the Maya collapse. Anc. Mesoam. 14 (01), 157e167. Shin, S.-I., Sardeshmukh, P.D., Webb, R.S., Oglesby, R.J., Barsugli, J.J., 2006. Understanding the mid-Holocene climate. J. Clim. 19 (12), 2801e2817. n, B., Les, D., Lin, P.Thompson, L.G., Mosley-Thompson, E., Brecher, H., Davis, M., Leo N., Mashiotta, T., Mountain, K., 2006. Abrupt tropical climate change: past and present. Proc. Natl. Acad. Sci. 103 (28), 10536e10543. Wahl, D., Byrne, R., Schreiner, T., Hansen, R., 2006. Holocene vegetation change in the northern Peten and its implications for Maya prehistory. Quat. Res. 65, 380e389. Wahl, D., Byrne, R., Schreiner, T., Hansen, R., 2007. A Late Holocene record of settlement and abandonment in the Mirador Basin, Peten, Guatemala. Holocene 17 (6), 813e820. Wahl, D., Estrada-Belli, F., Anderson, L., 2013. A 3,400 Year Paleolimnological record of prehispanic human-environment interactions in the Holmul region of the southern Maya lowlands. Palaeogeogr. Palaeoclimatol. Palaeoecol. 379e380, 17e31. Wahl, D., Hansen, R.D., Byrne, R., Anderson, L., 2014. Holocene Climate Variability and Anthropogenic Impacts From Lago Paixban, a Perennial Wetland in Peten, Guatemala (submitted for publication), pp. 1e43. Wanner, H., Beer, J.R., Buetikofer, J., Crowley, T.J., Cubasch, U., Flueckiger, J., Goosse, H., Grosjean, M., Joos, F., Kaplan, J.O., 2008. Mid-to Late Holocene climate change: an overview. Quat. Sci. Rev. 27 (19), 1791e1828. Webster, J.W., Brook, G.A., Railsback, L.B., Cheng, H., Edwards, R.L., Alexander, C., Reeder, P.P., 2007. Stalagmite evidence from Belize indicating significant droughts at the time of Preclassic Abandonment, the Maya Hiatus, and the Classic Maya collapse. Palaeogeogr. Palaeoclimatol. Palaeoecol. 250 (1e4), 1e17. Whitehead, D.R., Langham, E.J., 1965. Measurement as a means of identifying fossil maize pollen. Bull. Torrey Bot. Club 92, 7e20. Whitney, B.S., Rushton, E.A.C., Carson, J.F., Iriarte, J., Mayle, F.E., 2012. An improved methodology for the recovery of Zea mays and other large crop pollen, with implications for environmental archaeology in the Neotropics. Holocene doi: 0959683612441842. ~ o variWoodroffe, C.D., Beech, M.R., Gagan, M.K., 2003. Mid-late Holocene El Nin ability in the equatorial Pacific from coral microatolls. Geophys. Res. Lett. 30 (7). Zhao, M., Beveridge, N., Shackleton, N., Sarnthein, M., Eglinton, G., 1995. Molecular stratigraphy of cores off northwest Africa: sea surface temperature history over the last 80 ka. Paleoceanography 10 (3), 661e675.