Evidence for volcanic ash fall in the Maya Lowlands from a reservoir at Tikal, Guatemala

Evidence for volcanic ash fall in the Maya Lowlands from a reservoir at Tikal, Guatemala

Journal of Archaeological Science 38 (2011) 2925e2938 Contents lists available at ScienceDirect Journal of Archaeological Science journal homepage: ...

2MB Sizes 0 Downloads 18 Views

Journal of Archaeological Science 38 (2011) 2925e2938

Contents lists available at ScienceDirect

Journal of Archaeological Science journal homepage: http://www.elsevier.com/locate/jas

Evidence for volcanic ash fall in the Maya Lowlands from a reservoir at Tikal, Guatemala Kenneth B. Tankersley a, c, *, Vernon L. Scarborough a, Nicholas Dunning b, Warren Huff c, Barry Maynard c, Tammie L. Gerke c a b c

Department of Anthropology, University of Cincinnati, Cincinnati, OH 45221, USA Department of Geography, University of Cincinnati, Cincinnati, OH 45221, USA Department of Geology, University of Cincinnati, Cincinnati, OH 45221, USA

a r t i c l e i n f o

a b s t r a c t

Article history: Received 21 March 2011 Received in revised form 30 May 2011 Accepted 31 May 2011

Powder X-ray diffraction and petrographic analyses of reservoir sediments from Tikal, Guatemala have identified significant quantities of decomposed volcanic ash in the form of smectite and euhedral bipyramidal quartz crystals. X-ray fluorescence trace element content analysis was used to eliminate distant Sahara-Sahel and Antilles sources. The Zr/Y and Ni/Cr ratios of reservoir sediment from Tikal are consistent with a source from Central American volcanism (e.g., Guatemalan and Salvadoran). AMS radiocarbon dating of the smectite and crystalline quartz-rich reservoir sediments show that volcanic ash fell during the Preclassic, Classic, and Postclassic Maya cultural periods. It may now be possible to develop an effective chronology of ash fall at Tikal and the greater Peten. Published by Elsevier Ltd.

Keywords: Powder X-ray diffraction X-ray fluorescence Maya Lowlands Tikal Tephra

1. Introduction While volcanic ash has been well documented as temper in ancient Maya ceramics, its sources remain in question (Ford, 1991; Ford and Fedick, 1992; Ford and Glicken, 1987; West, 2002). The plethora of ash temper in ceramics recovered from the limestone lowlands of Guatemala has led some investigators to suggest that ash fell during the pre-Hispanic period of Maya occupation (Ford and Rose, 1995:149). This theory has significant implications for understanding the prehistoric exploitation of volcanic resources, landscape modification, and sustainability in the Maya Lowlands. If significant quantities of volcanic ash fell on the limestone lowlands of Guatemala during the pre-Hispanic occupation of the region, then we should expect to find direct positive evidence in the numerous large reservoirs constructed in the Maya city of Tikal (Fig. 1). The Maya constructed reservoirs at Tikal to conserve water during the annual dry season and to control and contain floodwaters during the rainy months. Six major reservoir catchment areas drained the elevated precincts of Tikal (Fig. 2), which covered an area of approximately 300 ha with a total maximum reservoir * Corresponding author. Department of Anthropology, University of Cincinnati, Cincinnati, OH 45221, USA. Tel.: þ1 513 556 2772; fax: þ1 513 556 2778. E-mail address: [email protected] (K.B. Tankersley). 0305-4403/$ e see front matter Published by Elsevier Ltd. doi:10.1016/j.jas.2011.05.025

capacity of more than 570,000 m3 (Scarborough and Gallopin, 2003:661). Surface water drained into the reservoirs and culturally modified aguadas (i.e., natural depressions) and bajos (i.e., huge solutional dolines) (Scarborough, 1993; 1994:116; 2003:51). Unlike deeply inundated deposits from lake basins and ocean floors, abandoned and in-filling reservoir sediments can be easily sampled with solid sediment drill cores. While it is possible for ash to survive in deep lakes, such as Yojoa in Honduras, this is not the case in smaller and shallower reservoirs. In these settings, a significant problem in sourcing ash from the Maya Lowlands is the fact that volcanic glass quickly weathers (i.e., chemically decomposes) into smectite clay in moist, tropical and alkaline environments, all of which are characteristic to the region. This phenomenon is exemplified by an ash fall from the El Chichón Volcano, which lasted from 28 March to 4 April 1982 (Robock, 2002). Although Tikal was blanketed by several centimeters of ash during this event, there is no visible evidence of the event in the soils today, and local residents report that much of the ash had already been incorporated into the soil within a few months. Smectite is a group of expanding-lattice clay minerals that include beidellite, hectorite, montmorillonite, nantronite, saponite, and sauconite, and is the principal component of bentonite clay deposits (Laird et al., 1991). Smectite originates from the decomposition of eruptive igneous rocks (e.g., tuff) and volcanic ash (i.e.,

2926

K.B. Tankersley et al. / Journal of Archaeological Science 38 (2011) 2925e2938

Fig. 1. Regional setting of Tikal and Central American volcanoes.

glass). Favorable physical and chemical conditions for the formation of smectite include magnesium-rich environments with poor drainage, which are characteristic of the reservoirs of Tikal. The expandable nature of clays of this type is revealed by x-ray diffraction analysis of the separated clay fraction (Moore and Reynolds, 1997). Likely sources for the weathered ash can be identified using bulk X-ray fluorescence (XRF) and electron microprobe data as documented by Huff (2008) and Huff et al. (1999, 2000) for multiple Ordovician bentonites. 2. Archaeological context Pre-Hispanic human occupation of the Maya Lowlands is often divided into the following chronological periods: PaleoIndian (pre7000 BC), Archaic (7000e2200 BC), Preclassic (2200 BCe250 AD), Early Preclassic (2200e1000 BC), Middle Preclassic (1000e400 BC), Late Preclassic (400 BCeAD100), Terminal Preclassic (AD 100e250), Classic (AD 250e900), Early Classic (AD 250e600), Late Classic (AD 600e770), Terminal Classic (AD 770e900), Postclassic (AD 900e1500), Early Postclassic (AD 900e1250), and Late Postclassic (AD 1250e1500). While PaleoIndian, Archaic, and Early Preclassic materials and occupations have been found widely distributed within the Maya Lowlands, no materials from these periods have thus far been recovered at Tikal. Sometime prior to about 700 BC, small populations began to reside at Tikal, and by 600 BC the first monumental architectural constructions appeared (Laporte, 2003). By the Late Preclassic period (ca. 350 BC), Tikal had developed into a significant “player” in the emerging political landscape of the Maya Lowlands. Unlike larger centers in the nearby Mirador Basin, Tikal survived the turmoil of the 2nd century AD and emerged as a major center of the Early Classic period (Dunning et al., in press). Tikal enjoyed variable prosperity during the Classic era, including a notable downturn in its fortune or “hiatus” in the 6th century AD, but emerged as a paramount center in the Late Classic period with a population estimated to have been around 60,000 in the 8th century (Martin and Grube, 2008). Tikal’s affluence declined

dramatically in the 9th century and by AD 900 was largely abandoned except for a small residual population that persisted for another 200 or so years. 3. Corriental reservoir Corriental is one of the largest reservoirs at Tikal (Fig. 3). It is strategically positioned to collect most of the surface water runoff from the southeastern margins of the central city promontory (Scarborough and Gallopin, 2003:251 and 252). The reservoir was likely used as a water source for drinking, cooking, and probably bathing as indicated by water jar sherds. Although there are no prepared steps or walkways associated with the reservoir, the scarcity of ancient debris in the sediments and the apparent former existence of carbonate sand filters suggest that it was constructed for public drinking water. The inferred presence of ancient sand filtration berms positioned at the ingress of the reservoirdno doubt occasionally “blown out” by capricious seasonal flooding eventsdis suggested by the repeated stratigraphic sand lensing reported in most excavation profiles; carbonate sand sources have not been identified within 10 km of Tikal, suggesting an anthropogenic origin. The catchment area for the Corriental Reservoir is about 40 ha, with a surface area of more than 15,000 m2 and an estimated capacity of more than 57,000 m3 of water (Gallopin, 1990:60). It was constructed in a pre-existing localized depression at the junction of several small seasonal streams. The reservoir featured two separate ingress gates at different elevations, approximately 205 m amsl on the south side and 208 m amsl on the northwestern side (Gallopin, 1990:32e38). In 2009, ten pits and trenches were hand excavated from the center and margins of the Corriental Reservoir, the northwestern and southern ingress gates, and the eastern egress (see Fig. 3). Additionally, a hand-operated Environmental Subsoil Probe (ESP) was used to extract twenty-six 2-cm diameter cores from Corrientald16 from the center of the reservoir, 7 from the earthen berms, 2 from the northwestern ingress gate, and 1 from the

K.B. Tankersley et al. / Journal of Archaeological Science 38 (2011) 2925e2938

2927

Fig. 2. Main catchment areas and reservoirs of Tikal.

eastern egress (see Fig. 3). Twenty-three soil horizons were defined in the field by Dunning during excavations on the basis of color, texture, structure, and pedogenic features. Tankersley confirmed the reservoir-wide stratigraphy in the lab during analysis and correlation of the cores with particle size analysis, magnetic susceptibility, and Munsell soil color. 3.1. Stratigraphy and geochronology The stratigraphy for the interior sediments and buried soils within the Corriental Reservoir, as defined in Operation 1C, are outlined in Table 1. Operation 1C was a 1  1.5 m trench excavated near the center point of the Corriental Reservoir. The pit reached a final depth of 3.15 m, at which point weathered limestone bedrock was encountered. Dr. Pat Culbert analyzed ceramics from this operation at Tikal. Radiocarbon dating at Corriental was performed in a series of AMS radiocarbon measurements made on carbonized plant remains

and soil organic matter (SOM) (Table 2)dthe former abundantly associated with airborne wood ash from the many cooking fires and the like from within the low-density urban setting. This work established that the stratigraphy (i.e., soil horizons) extended from the early Holocene, 8960  60 14C yr BP (Beta-270566) to at least the Early Postclassic 990  40 14C yr BP (Beta-258720). It also suggested that the Corriental Reservoir was in use by the Early Classic, 1560  40 14C yr BP (Beta-274990), but may have been initially constructed toward the end of the preceding Late Preclassic period. The 1560  40 14C yr BP (Beta-274990) AMS radiocarbon date was obtained from core 17 on charred plant remains beneath the earthen berm (i.e., wall) 3 m above the northwest gate (Fig. 4). This date demonstrates that this feature was built during or subsequent to the Early Classic. Stable carbon isotope values (d 13C 19.1, 19.5, and 20.3&) on the AMS radiocarbon dated soil organic matter suggests that C4 photosynthetic pathway plants such as maize were present in the Corriental catchment area during the Late Preclassic and Middle Preclassic cultural periods. Consequently,

2928

K.B. Tankersley et al. / Journal of Archaeological Science 38 (2011) 2925e2938

Fig. 3. Location of Corriental Reservoir core samples and excavation units.

Table 1 Age and relative percent composition of sediments identified in the Corriental reservoir based on XRD. Depth Horizon Measured 14 C yr BP

2s Calibrated Age

10 20 30 40 50 60 70 80 90 100 110 120 130 140 150 160 170 180 190 200 210 220 230 240 250 260 270 280 290 300 310

12.84 88.72 80.51 70.75 67.51 51.77 AD 1010e1170 73.72 71.10 47.01 88.72 80.51 70.75 67.51 51.77 340e30 BC 73.72 71.10 19.56 190e80 BC 17.73 380e170 BC 52.31 78.70 68.33 82.41 80.27 69.76 84.16 83.61 760e400 BC 79.43 92.20 76.03 63.70 8290e7970 BC 90.48

A1 A2 ACss C2ss C2ss C2ss C3 C4ss C4ss C4ss C4ss C4ss C5 C6ss C6ss C6ss C7 C7 C7 C8 C9 C10 C11ss C12 2Abss 2ACss 2C1bss 2C1bss 2C2bss 2C2bss 3Abss

990  40

2010  40

2110  40 2120  40

2340  40

8960  60

Calcite Smectite Quartz Void (%) (%) (%) (%) 37.91 1.93 0.00 14.70 13.13 8.89 9.30 13.98 37.91 1.93 0.00 14.70 13.13 8.89 9.30 13.98 73.59 77.77 39.91 11.59 21.62 4.92 10.02 18.11 19.86 4.43 9.65 6.56 9.86 18.84 15.74

35.86 4.32 0.88 5.41 2.87 3.48 3.12 7.02 35.86 4.32 0.88 5.41 2.87 3.48 3.12 7.02 2.68 1.46 6.32 9.81 0.00 3.46 4.68 11.14 12.30 4.43 2.02 4.76 11.48 12.04 3.97

12.84 6.97 18.61 12.40 16.49 38.23 15.53 12.96 12.84 6.97 18.61 12.40 16.49 38.23 15.53 12.96 5.40 3.88 5.48 6.31 10.05 11.30 8.22 9.78 13.21 31.56 33.01 11.29 10.69 13.78 19.78

open environments may have led to increased erosion rates in the catchment area (Anselmetti et al., 2007), the latter influenced by increased precipitation with the onset of the Classic period as suggested by some climatic models for the greater Maya Lowlands (Dunning and Beach, 2010). The Early Classic berm additions may have been built to strengthen the northwestern ingress against heavy runoff during strong seasonal rains and hurricanes, which are powerful geomorphic agents at any time throughout the region (Dunning and Houston, 2011), as well as to increase the carrying capacity of the reservoir. The lowermost horizons exposed in Operation 1C (i.e., 3Ab and 3AC horizons) represented a highly compacted, skeletal soil overlying limestone bedrock. A radiocarbon date on soil organic matter from the analogous soil horizon in Core 8 produced a date of 8960  60 14C yr B.P or early Holocene age (see below, Fig. 4 and Table 2). Nevertheless, based on similarities to other deeply buried soils in small depressions elsewhere in the northeast Peten and northwest Belize, this soil may date as far back as the late Pleistocene (ca. 11,000e13,000 BP) (Dunning et al., 2006; Beach et al., 2008). With the onset of wetter conditions in the mid-Holocene period, the small depression filled with sediment eroded from upslope areas (2C1 and 2C2 horizons) on which a new soil surface (2Ab and 2AC horizons) gradually developed. Radiocarbon dating of organic matter within the 2C1 horizon place the age of this soil surface at approximately 2340  40 14C yr BP (Middle Preclassic). However, this date is based on soil organic matter (organic matter which accumulated over hundreds of years, given the lack of significant occupation and associated cooking fires at this time), hence the soil surface was probably last exposed sometime in the Late Preclassic. This soil is a Vertisol, typical of small seasonally wet/dry depressions in the Peten. Four weathered sherds were recovered from this level, but were chronologically non-diagnostic. At some point within the next 500 years, drainage within the depression appears to have been substantially modified and sediment began to accumulate much more rapidly. This modification likely corresponds to the initial construction of the Corriental

K.B. Tankersley et al. / Journal of Archaeological Science 38 (2011) 2925e2938

2929

Table 2 Radiocarbon dates for the Corriental reservoir. Lab Number

Sample

d 13C&

Horizon

Measured

Beta-258720 Beta-274990 Beta-280839 Beta-266124 Beta-280837 Beta-258721 Beta-270566

Charcoal Charcoal SOM SOM SOM SOM SOM

26.9 23.7 19.5 20.2 20.3 19.1 23.1

C3 Anthrosol C6 C7 C7 2C1 3Ab

990  40 1560  40 2010  40 2110  40 2120  40 2340  40 8960  60

Reservoir, probably accomplished by a combination of quarrying (widening the natural depression), building up of the encircling berm, and diversion of inflowing seasonal stream runoff. Subsequently, the floor of the reservoir began to aggrade (fill with sediment over time). Alternating strata of organic clay and laminated carbonate sand and small, rounded gravel are found between 65 and 253 cm depth within Operation 1C (C3 to C12 horizons). The laminated or stratified nature of the deposits indicates that the sands were deposited by running water (i.e., a fluvial process). Notably, there are no known natural sand sources upstream from

14

C yr BP

2 s Calibrated Age

Cultural Period

AD 1010e1170 AD 400e570 340e30 BC 190e80 BC 380e170 BC 760e400 BC 8290e7970 BC

Early Postclassic Early Classic Late Preclassic Late Preclassic Late Preclassic Middle Preclassic Pre-habitation?

the reservoir. One possible explanation is that sand was used to filter water entering the reservoir and that sand from the filters was occasionally flushed into the reservoir proper by storm-related flooding. On the other hand, the organic clay layers are typical of still water deposits and likely accumulated slowly while the reservoir was in active use. There are no obvious signs of dredging within the sediments revealed in Operation 1C or other excavations and cores in the reservoir. Three radiocarbon dates were obtained from organic matter in the C6 and C7 horizons: 2010  40 14C yr BP for C6, and 2110  40 and

Fig. 4. Profile drawing of the east wall of Op. 1C in the floor of Corriental Reservoir (after Dunning et al., 2009). Radiocarbon dates to the left of the profile are in measured 14C years. See Table 2 for additional chronological information. Dates in brackets were obtained from analogous soil horizons in Core 8.

K.B. Tankersley et al. / Journal of Archaeological Science 38 (2011) 2925e2938

- 1.53Å - 1.5Å

- 1.61Å

CALCITE

CALCITE

- 1.92Å - 1.88Å

- 2.1Å

CALCITE

- 2.29Å CALCITE - 2.35Å

- 3.37Å QUARTZ

- 3.88Å CALCITE

SMECTITE, ZEOLITES, QUARTZ

- 2.58Å - 2.5Å CALCITE

15.3

- 4.5Å

- 5.98Å

- 15.63Å

SMECTITE, ZEOLITES

SMECTITE, ZEOLITES

- 3.06Å CALCITE

2930

TK1 70-80

14.3

TK1 60-70

10.8 TK1 50-60

11.2 TK1 40-50

2.6 TK1 30-40

17.2 TK1 20-30 90.8 TK1 10-20

39.4 TK1 0-10 0

10

20

30

40

50

60

70

2Theta

Fig. 5. X-ray diffractograms of sediments extracted from core 8 located at the center of the Corriental Reservoir (first core extracted from 0 to 80 cm below the surface).

2931

- 1.63Å - 1.61Å CALCITE

CALCITE CALCITE - 1.92Å - 1.88Å

- 2.1Å

- 2.03Å

- 2 35Å - 2.29Å CALCITE

- 2.5Å CALCITE - 2.58Å

- 2.86Å

QUARTZ

- 3.88Å CALCITE

- 3.37Å

80.4

- 4.48Å SMECTITE, ZEOLITES, QUARTZ

- 5.82Å

SMECTITE, ZEOLITES

- 14 -17Å SMECTITE, ZEOLITES

- 3.05Å

CALCITE

K.B. Tankersley et al. / Journal of Archaeological Science 38 (2011) 2925e2938

TK2 70-78

2.3 TK2 60-70

20.8 TK2 50-60

76.0 TK2 40-50

78.5 TK2 30-40

3.1

TK2 20-30

TK2 10-20

6.0

31.9 TK2 0-10 0

10

20

30

40

50

60

70

2Theta

Fig. 6. X-ray diffractograms of sediments extracted from core 8 located at the center of the Corriental Reservoir (second core extracted from 80 to 160 cm below the surface).

2120  40 14C yr BP for C7. The dates correspond to the Maya Late Preclassic period and are appreciably older than the ceramic sherds found in the underlying C8 through C12 horizons. This inversion strongly suggests that the organic matter being dated in the C6 and C7 horizons is derived from older soil eroded from surfaces in the watershed above the reservoir. Analysis of the d 13C and d 15N content

of this organic matter also indicates that it is derived from C3 terrestrial plants and not aquatics such as algae. Notably, soil surfaces dating from the Late Preclassic period are typical of the Maya Lowlands, but they were subject to pulses of erosion resulting in the deposition of sediments derived from these soils in countless local depressions (Anselmetti et al., 2007; Dunning and Beach, 2000).

K.B. Tankersley et al. / Journal of Archaeological Science 38 (2011) 2925e2938

- 1.53Å

- 1.63Å - 1.61Å

- 1.92Å - 1.88Å

CALCITE

CALCITE CALCITE

CALCITE - 2.1Å

- 2.5Å

- 2.29Å

CALCITE

- 3.05Å QUARTZ

SMECTITE, ZEOLITES, QUARTZ

- 3.88Å CALCITE

- 3.36Å

- 6Å

11.9

- 4.53Å

- 14 -17Å

SMECTITE, ZEOLITES

SMECTITE, ZEOLITES

CALCITE

2932

TK3 70-80

39.4 TK3 60-70

18.2

-

TK3 50-60

39.2 TK3 40-50

37.3 TK3 30-40 353.5

TK3 20-30 2116.2

TK3 10-20

903.0 TK3 0-10 0

10

20

30

40

50

60

70

2Theta

Fig. 7. X-ray diffractograms of sediments extracted from core 8 located at the center of the Corriental Reservoir (third core extracted from 160 to 240 cm below the surface).

Ceramic sherds were recovered in the C3, C4, C5, C7, C8, C10, and C12 horizons principally in the sandy strata. Small and large water jar forms predominate in all strata. The large majority of sherds were too weathered to be chronologically diagnostic. C12, the deepest alluvial stratum, and C9 included identifiable Early Classic types. C10 had no diagnostic sherds. C8 contained a mix of Early and Late Classic types. C5eC7 included only Late Classic types. C3 had no diagnostic sherds. Charcoal within the C3 horizon (65 cm) produced an AMS radiocarbon date of 990  40 14C yr BP (2 sigma correlated range: AD 1010e1170), suggesting that the reservoir may

have continued to be in use to some extent as late as the Early Postclassic period. Subsequently, there is no evidence of reservoir use, though it has naturally continued to seasonally collect water. The modern soil that has developed within the reservoir (Oi through C2 horizons) is a Terric Fibrist, an organic soil with mineral subsoil typical of regional depressions that remain partially moist year-round. In summary, the Corriental Reservoir, on the southern flank of central Tikal, was likely constructed sometime toward the end of the Late Preclassic or very early in the Early Classic period. The

K.B. Tankersley et al. / Journal of Archaeological Science 38 (2011) 2925e2938

4. Powder X-ray diffraction Powder XRD was used to identify the relative percent mineral composition of the reservoir sediments of Tikal. For this study, solid sediment core samples were extracted from the center of Corriental to obtain the deepest and most complete stratigraphic samples possible. Initially, the cores were cut into 10 cm subsamples. Soil strata were defined on the basis of particle size analysis, sedimentary boundaries, and changes in Munsell soil color, then correlated with the horizons established in Operation 1C. 4.1. XRD methods Following the methodology of Tankersley and Ballantyne (2010), XRD samples were taken at 10-cm intervals and sieved through

- 1.54Å

- 1.62Å - 1.6Å CALCITE

- 1.9Å CALCITE - 1.87Å CALCITE

- 2.08Å - 2.03Å

- 2.34Å - 2.27Å CALCITE

- 2.55Å - 2.49Å CALCITE

CALCITE - 3.01Å - 2.82Å

- 3.31Å QUARTZ

- 3.83Å CALCITE

- 14 -20Å

SMECTITE, ZEOLITES

- 4.41Å SMECTITE, ZEOLITES, QUARTZ - 4.22Å

reservoir was constructed by widening a small pre-existing natural depression by quarrying and mounding earth to form an encircling berm and filled by diverting water flow from a local seasonal stream. Thick sediment deposits within the reservoir included alternating deposits of stratified carbonate sands and organic clays. The clays indicate periods of stability during which clay and organic matters gradually settled onto the reservoir floor. The sandy strata are indicative of running water, perhaps deposited during higherenergy storm runoff events. The origin of the sand is uncertain, but it may have been used to filter water as it entered the reservoir, and then was occasionally flushed into the reservoir proper during flooding. Ceramics recovered from within the reservoir sediments were generally very weathered, but contained a mixture of Early Classic and Late Classic types. Most notable were the quantities of huge jar fragments.

2933

9.5

TK4 60-70

50.3 TK4 50-60

14.9 TK4 40-50

22.1

TK4 30-40

10.2

TK4 20-30

0.0

TK4 10-20

38.7 TK4 0-10 0

10

20

30

40

50

60

70

2Theta

Fig. 8. X-ray diffractograms of sediments extracted from core 8 located at the center of the Corriental Reservoir (third core extracted from 240 to 320 cm below the surface).

38 35 33 30 31 33 36 38 36 37 20 na 25.1 24.1 23.9 17.9 21.3 47 b

a

Dull et al., 2001 Moreno et al., 2006. c Carr et al., 2007. na ¼ not analyzed.

average average average average average average average

Zr Y V

46 41 35 32 39 77 82 87 83 79 46 na 88.3 173 134 136 168 211 40 41 42 41 42 22 22 24 27 25 160 na 179 383 662 485 420 568

Sr Rb

17 17 19 16 16 32 32 31 30 30 55 na 61.4 27.9 60.8 41.5 31.2 21 41 38 35 28 32 69 74 76 61 67 17 na 23 62.3 23.3 12.1 9.74 37

Ni Nb

10 10 9 9 10 15 16 15 14 15 4 na 30 5.13 12.7 7.3 2.32 25 38 37 33 33 31 79 78 87 76 77 12 na 71 197 43 19 12 56

Cr Ba

217 350 340 132 265 289 264 290 193 217 813 na 449 246 600 654 695 502 0.12 0.14 0.16 0.16 0.13 0.01 0.01 0.01 0.01 0.01 0.11 na 0.24 0.12 0.24 0.30 0.21 1.36

P2O5 TiO2

0.20 0.18 0.40 0.27 0.33 0.45 0.45 0.43 0.40 0.40 0.46 0.18 1.03 0.71 0.68 0.68 0.79 1.40 0.09 0.08 0.09 0.01 0.02 0.17 0.17 0.17 0.18 0.18 2.16 2.77 1.76 0.94 2.02 2.07 1.54 3.56

K2O Na2O

1.29 1.12 6.23 2.01 2.83 0.04 0.05 0.04 0.05 0.05 3.19 3.99 0.94 3.10 3.96 3.91 3.60 8.61 17.35 23.10 7.07 17.39 9.19 9.68 9.26 8.73 7.63 7.86 3.01 1.15 5.78 8.44 6.58 6.03 7.15 4.32

CaO MgO

0.15 0.15 0.15 0.14 0.13 1.53 1.51 1.54 1.58 1.56 1.32 0.20 1.56 3.90 2.70 2.45 2.90 0.17 0.24 0.25 0.16 0.06 0.09 0.34 0.34 0.28 0.31 0.38 0.13 0.07 0.07 0.16 0.15 0.18 0.16 9.63

MnO Fe2O3t

4.20 3.55 4.99 2.41 3.94 6.62 6.70 7.21 7.34 7.09 3.37 1.25 4.90 7.74 6.48 5.97 7.98 8.75 14.9 12.3 13.0 10.8 12.6 17.8 17.9 18.5 18.6 18.6 15.3 13.0 11.0 17.3 17.8 17.5 17.3 18.3

Al2O3 SiO2 Depth Reservoir

Corriental Corriental Corriental Corriental Corriental Perdido Perdido Perdido Perdido Perdido

Location

Fig. 9. Photomicrograph of a representative euhedral bipyramidal quartz crystal from the C7 horizon of Corriental reservoir sediment (see Table 1).

Table 3 Trace element content of Tikal reservoir sediments and comparative volcanic sources (wt% or ppm).

The minerals calcite, smectite, and quartz were found in the Corriental Reservoir sediments (Figures 5e8). Calcite was the most abundant mineral (66%) and characterized by glycolated XRD peaks at 3.88 Å, 3.38 Å, 3.06 Å, 2.50 Å, 2.25 Å, 2.29 Å, 1.92 Å, and 1.88 Å. The remainder of the sediments consisted of smectite (27%), characterized by XRD peaks at 15.63 Å, 5.98 Å, and 4.50 Å, and quartz (6%), characterized by an XRD peak at 3.37 Å. Quartz and calcite were unaffected by glycolation. The abundance of calcite is undoubtedly related to the Cretaceous and Tertiary limestone bedrock, which determines the terrain in the Maya Lowlands (Dunning et al., 1998). Because quartz is known to occur in carbonate rock (Chafetz and Zhang, 1998), samples of local limestone bedrock were subjected to powder XRD analysis. XRD peaks at 3.85 Å, 3.03 Å, 2.84 Å, 2.49 Å, 2.28 Å, 2.09 Å, 1.91 Å, 1.62 Å, 1.60 Å, and 1.52 Å characterized calcite. Peaks typical of quartz and smectite were completely absent from the bedrock samples. This finding suggests that smectite and quartz entered the reservoirs as aeolian minerals rather than dissolution of the surrounding bedrock and subsequent water transport (contra Cowgill and Hutchinson, 1963:41). Grim and Güven (1978:128)

32.0 28.4 40.1 36.2 39.4 42.4 42.0 42.9 43.9 43.6 68.5 77.5 60.7 57.4 58.4 59.5 57.4 52.2

4.2. XRD results

180e190 190e200 200e210 210e220 220e230 180e190 190e200 200e210 210e220 220e230

a 2 mm mesh. Approximately 20 g of clay from each stratum was mixed with deionized water to make slurry in 100 ml beakers. Clay was dispersed using a high-speed stirrer and gravity settling was used to obtain a fraction of <2 mm. A 5 ml pipette was used to obtain a clay sample from the top of the slurry and transferred to a glass slide and air-dried. A second oriented glass slide was prepared from each sample and equilibrated overnight with ethylene glycol vapor. XRD patterns were obtained for both the air-dried and glycolated samples. All slides were initially scanned from 2 to 32 2q at 0.5 increments and then broadened to 60 2q on a Siemens D-500 X-ray diffractometer using a Cu-Ka radiation source. The intensity threshold was set at 1.6 and minerals were identified on the basis of peak position and peak intensity as described by Chen (1977). Glycolated samples were prepared to test for the presence of expandable clay minerals. Relative mineral percentages were calculated from the total counts per second (cps), which were totaled for each 10-cm sample. The sum of all peaks cps for each mineral was divided by the total cps for each 10-cm sample. The relative percent of each mineral was calculated from the total cps per mineral divided by the total cps in the 10-cm sample.

146 131 123 110 126 211 214 214 202 198 158 na 310 107 153 132 108 154

K.B. Tankersley et al. / Journal of Archaeological Science 38 (2011) 2925e2938

Tikal Tikal Tikal Tikal Tikal Tikal Tikal Tikal Tikal Tikal Ilopango Tephra Tierra Blanca Joven TephraSaharan Dust Antilles Tephra Mexico Tephra Guatemala Tephra El Salvador Tephra Honduras Teprha

2934

K.B. Tankersley et al. / Journal of Archaeological Science 38 (2011) 2925e2938

2935

640, Leica MZ12 stereomicroscope), the quartz grains appear as dispersed slivers, fragments of euhedral crystals, and complete bipyramidal crystals with scarce inclusions (Fig. 9). Together, these characteristics suggest they represent “first quartz” or an extrusive igneous or volcanic origin. While smectite, like the euhedral bipyramidal quartz crystals, suggests a volcanic ash source, Tikal is located in the Sahara-Sahel Dust Corridor (Moreno et al., 2006) and African dust has been identified as a major component of soils overlying other carbonate land masses in the Caribbean Basin including Florida, the Bahamas, and Barbados (Muhs et al., 2007). To determine if the smectite from the reservoirs of Tikal originated as airborne Sahara-Sahel dust or volcanic ash, the trace element composition of the reservoir sediments from Tikal was analyzed using XRF. 5. X-ray fluoresence XRF was used to determine the trace element concentration of smectite-bearing reservoir sediments of Tikal and compare it with the trace element content of airborne Sahara-Sahel dust (Moreno et al., 2006). Fig. 10. Trace element Ni/Cr and Zr/Y ratios of Tikal reservoir sediments, Sahara-Shahel dust, Antilles volcanoes, and Central American volcanoes.

5.1. XRF methods

demonstrate that smectite forms from weathering volcanic ash varying in composition from rhyolitic (quartz-rich) to basaltic (quartz-poor). However, most bentonites have formed from ash ranging from rhyolitic to dacitic in composition. In other words, the association of quartz and smectite in altered volcanic ashes is quite common. Grains isolated from all soil horizons were identified as quartz bearing with XRD (see Table 1). Under high magnification (up to

Core samples, which contained high relative percentages of smectite were selected for XRF analysis from the C7 and C8 soil horizons of Corriental (i.e., Preclassic, Classic, and Postclassic strata). As a control, samples from the pre-Maya strata of the nearby Perdido reservoir (see Fig. 2), AMS radiocarbon dated 15,110  60 14C yr BP (Beta-289286), 15,480  60 14C yr BP (Beta289285), and 15,310  60 14C yr BP (Beta-289284), were also analyzed and compared. A Rigaku 3070 X-ray Fluorescence

Fig. 11. Trace element Ni/Cr and Zr/Y ratios of Tikal reservoir sediments and Central American volcanoes (CR1 Arenal, ES1 Santa Ana, ES2 Cerro Verde, ES3 Apaneca, ES4 Meanguera, ES5 Izalco, ES6 Conchagua Peninsula, ES7 San Miguel, G1 Santa Maria, G2 Tacana, G3 Agua, G4 Pacaya, G5 Tecuamburro, G6 Moyuta, H1 Cerro de Hule, H2 Yojoa, M1 Venustaubi Carranza, M2 Santoton).

2936

K.B. Tankersley et al. / Journal of Archaeological Science 38 (2011) 2925e2938

spectrometer was used to determine the intensity of the trace elements; Mo, Ba, Co, Cr, Cu, Nb, Pb, Rb, Sr, Th, U, V, Y, and Zn. Following the methodology of Gerke et al. (2006:140), sample powders were pressed into briquettes at 2000 psi. A separate aliquot was heated to 1000  C for 1 h to measure volatile content. Intensity data were converted to parts per million (ppm) using bivariate and multiple variable regressions applied to United States Geological Survey, National Institute of Standards, and Japan Geological Survey rock standards (Table 3). 5.2. XRF results Significant differences in the trace element content of sediment from the reservoirs of Tikal and Sahara-Sahel dust were found in the ratios of Ni, Cr, Zr, and Y. These discriminations can be illustrated in a plot of the ratio of Cr to Ni against the ratio of Zr to Y (Fig. 10). Ratios are used rather than absolute amounts to eliminate the effect of the large amounts of locally derived calcite. The trace element content of Sahara-Sahel dust has a significantly lower range of Cr/Ni ratio and a higher Zr/Y ratio than do the samples from Tikal. Using the same comparison for volcanic materials from the Lesser Antilles, Tikal has comparable ratios of Zr/Y, but a greater ratio of Ni/Cr (see Fig. 10). The Ni/Cr trace element content of Tikal sediment is, however, comparable to volcanic rocks and tephra from Guatemalan or Salvadoran volcanoes (Fig. 11). The Ilopango TBJ eruption is the largest and best-documented Holocene volcanic event in Central America (Hart and SteenMcIntyre, 1983; Sheets, 2002). It occurred during the Early Classic Period, sometime between A.D. 408 and 536 and its ecological and cultural impact would have been felt throughout the Maya region (Dull et al., 2001). Professor Payson Sheets of the University of Colorado kindly provided our team with samples of the TBJ tephra for comparison. Data for other volcanic components are available in the database developed by Carr (Carr et al., 2007). XRD analysis demonstrated that the TBJ tephra is composed of plagioclase feldspar with lesser amounts of quartz and a large amount of glass. In other words, there are no clays in the TBJ tephra, because the glass is still intact, unlike the Tikal sediments where all the glass has converted to smectite. It is possible that the abundance of calcite in the Tikal reservoirs compared to the slopes of the Ilopango volcano accounts for this difference (Cowgill and Hutchinson, 1963). Weathering, however, should not affect the trace element ratios for high-field strength elements like Cr, Nb, Ti, Y, and Zr (Winchester and Floyd,1977; Floyd and Winchester,1978; Maynard,1992). Nickel is more mobile than Cr in acidic soils under tropical conditions (Maynard, 1983), and accumulates lower in the profile, which is the mechanism of genesis for lateritic nickel deposits. However, it should remain fixed in the high carbonate environment of these deposits. The Ni/Cr ratios in the individual cores at Tikal are constant with depth, indicating that vertical migration of Ni has been minimal. Our XRF analysis of the TBJ tephra sample found a higher ratio of Zr/Y than the sediments from the reservoirs of Tikal. Similar results were obtained using the analyses in the Carr database. Note that the position for Ilopango in Fig. 11 is an average of our sample and the Carr data. 6. Implications Several insights can be gleaned from this study, somewhat independently of the surprising amounts of weathered volcanic ash now revealed from Tikal. First, the pedological and stratigraphic controls suggest the presence of a sand filtration system presumably to purify the water source in the Corriental Reservoir. Such a technology was previously unknown, but it is not altogether surprising

given the difficulties associated with waterborne contaminants affecting any tropical setting. Secondly, the timing of the initial Late Preclassic construction and harvesting of water from the reservoir coincides with the posited drought-like conditions recently argued for this period (Dunning et al., in press). And the Classic period build up of the northwestern gate berm may suggest the return of a wetter period, with the necessity of securing the principal ingress into the reservoir to prevent excess waters from eroding the gate. At this point, our understanding of regional paleoclimate does not allow us to determine if any of the volcanic eruptions that dumped ash on Tikal also affected local, regional, or global climate. Nevertheless, the most revealing data sets come from the XRD and XRF assessments of weathered volcanic ash over Tikal and by extension over much of the Maya Lowlands through time. The impact of volcanism went far beyond the immediacy of the Guatemalan or Chiapan highlands into the limestone rich soils of the Maya Lowlands. The study indicates the wealth of information sealed and potentially extractable from any abandoned reservoir setting. While reservoir sediments can be compromised by human activities such as dredging, this was not the case at Corriental. Such reservoirs can provide significant paleoenvironmental data because (1) we can obtain many more controlled cores, and (2) the diminutive size of a tank allows a much more meaningfuldand accuratedsample of an entire basin than a lake. Reservoirs can act as a corrective for the “randomness” of lake core extraction. Ideally, lake and ocean floor cores should be used in combination with reservoir sediment data. 7. Conclusion Corriental is an ancient Maya reservoir at Tikal, Guatemala. It provides the first evidence of volcanic ash fall in the Maya Lowlands during pre-Hispanic times. Corriental is a unique test location to examine the chronology and stratigraphy in the reservoir, the aeolian minerals, and their potential source regions. Unlike large lakes, Maya reservoirs are restricted areas with a known constriction point over which volcanic minerals are deposited. Powder XRD and petrography show that volcanic-derived minerals, smectite and euhedral bipyramidal quartz, are present in significant quantities in the sediments of Corriental. AMS radiocarbon dating of the Corriental reservoir demonstrates that glass (now smectite) and euhedral quartz were deposited during the Preclassic, Classic, and Postclassic cultural periods. Although smectite is known to occur in Sahara-Sahel dust, XRF trace element analysis indicates a different origin, one comparable to tephra from Guatemalan or Salvadoran volcanoes (Cabadas-Baez et al., 2010). Fingerprinting the exact sources of these minerals will, however, require a comprehensive survey of tephra deposits from the Maya Highlands, Mexico, and the greater Caribbean region. The presence of airborne volcanic minerals in the reservoir sediments of Tikal supports a model of centuries-long periods of volcanism during the prehistoric Maya occupation of the limestone lowlands (Ford and Glicken, 1987). As suggested by Ford and Rose (1995:159), the region likely experienced regular and periodic tephra eruptions and ash falls into the lowlands. Ash temper and assemblage of crystals in Maya pottery include biotite, hornblende, hypersthene, and zircon, which all are minerals consistent with Guatemala Highland tephra (Drexler et al., 1980; Rose et al., 1981). The presence of biotite suggests El Chichon, Tajamulco, Acatenango, and Atitlán as possible volcanic sources, which were active in the period of AD 600e900 and Cerro Quemado, which was active at about AD 800 and for some time there after (Conway et al., 1992; Ford and Rose, 1995). Because each reservoir stratum has its own unique trace element composition, different volcanoes were likely erupting at different moments in time. Montmorillonite (i.e., smectite) in Lake Péten Itzá sediments radiocarbon dated A.D. 880

K.B. Tankersley et al. / Journal of Archaeological Science 38 (2011) 2925e2938

to 1140 and El Bajo de Santa Fé may also correlate with these volcanic events (Cowgill and Hutchinson, 1963:41; Mueller et al., 2010:525). The presence of airborne volcanic minerals also supports the position that the Maya living in the limestone lowland had a dependable source of volcanic ash for the production of ceramics. Indeed, tephra is superior to the temper used as early as the Preclassic and Early Classic as a result of its angularity (West, 2002). Also given the reverence of the Maya for sacred volcanoes, the tephra likely had religious implications. It may have been collected at Tikal and the greater Peten from air-fall deposits and acquired by trade from sources in the Maya Highlands. Given the accounts of the 1982 ash fall from El Chichón in the Peten it is clear that it would have been very easy for the Maya to collect and stockpile large quantities of ash. Although it is beyond the scope of this paper, if we are able to develop an effective chronology of ash fall at Tikal and the greater Peten, then it might be possible to compare this record to chronological patterns in temperingdto say nothing about chronological control of stratigraphic deposits. In addition to the use of ash for temper, smectite is an important raw material for ceramic production because of its plasticity. It has long been assumed that tropical soils in the Peten and surrounding areas were the result of weathering of the local carbonate bedrock. Ash enrichment of the soils would have also enhanced soil fertility by increasing the porosity, permeability, and nutrients needed for maize agriculture, thus helping support population increases at Tikal and elsewhere (Ford and Rose, 1995). It is likewise possible that a short-term problem generated by heavy fall would have been the clogging of water catchment and storage systems. Additionally, ash falls would have directly affected local flora and fauna proportional to depth. For example, it would have clogged the spiracles of insects and suffocated them, thus, insect pollinated plants would have suffered. Additionally, volcanic ash would have been deleterious to aquatic food resources such as shellfish, fish, and manatees. Although difficult to quantify, several of our stratigraphic columns examined in 10 cm arbitrary levels yielded volcanic ash and quartz ejecta in amounts as high as 70% by volume (see Table 1). It is important to further note that preliminary analysis of other soils and reservoir sediments from Tikal suggest that smectite clay derived from weathered volcanic ash is a major component of the inorganic fraction of these soils. Future investigation may reveal that aeolian volcanic material is, indeed, a dominant component of the mineral portion of the soils across wide areas of the Maya Lowlands. Acknowledgments This study was made possible with funding from the Court Family Foundation, the Charles Phelps Taft Foundation, the Alphawood Foundation, the Wenner-Gren Foundation, and a grant from the National Science Foundation (BCS0810118) to David Lentz, Vernon Scarborough, and Nicholas Dunning. Fieldwork was undertaken in 2009 as part of the University of CincinnatidIDAEH Tikal Project co-directed by David Lentz and Liwy Grazioso Sierra. The laboratory assistance of Meredith Coates, Nathan Marshall, Jon Paul McCool, Jim Milawski, Andras Nagy, Leslie Neal, and Tony Tamberino were invaluable. Chris Carr led digitization and rectification of the original University of Pennsylvania map of central Tikal and produced Fig. 3. Eric Weaver led field mapping of the Corriental Reservoir in 2009. Liwy Grazioso Sierra directed many of the excavations in Corriental Reservoir assisted by Raquel Macario, Sheryl Carcruz, Ana Arriola, and Marielos Corado. Pat Culbert analyzed the ceramics recovered from these excavations. Brian

2937

Lane, Benjamin Thomas, and Eric Weaver accomplished the coring of Corriental Reservoir. Payson Sheets provided a tephra sample from Ilopango Volcano. References Anselmetti, F.S., Hodell, D.A., Ariztegui, D., Brenner, M., Rosenmeier, M.F., 2007. Quantification of soil erosion rates related to ancient Maya deforestation. Geology 35, 915e918. Beach, T., Luzzadder-Beach, S., Dunning, N.P., Cook, D., 2008. Human and natural impacts on fluvial and Karst systems in the Maya Lowlands. Geomorphology 101, 301e331. Cabadas-Báez, H., Solleiro-Rebolledo, E., Sedov, S., Pi-Puig, T., Gama-Castro, J., 2010. Pedosediments of karstic sinkholes in the eolianites of NE Yucatán: a record of Late Quaternary soil development, geomorphic processes and landscape stability. Geomorphology 122, 323e337. Carr, M.J., Feigenson, M.D., Patino, L.C., 2007. Petrology and geochemistry of Lavas. In: Bundschuh, J., Alvarado, G. (Eds.), Central America: Geology, Resources and Hazards. Taylor and Francis, London, pp. 565e590. Chapter 22. Chafetz, H.S., Zhang, J., 1998. Authigenic euhedral megaquartz crystals in a Quaternary dolomite. Journal of Sedimentary Research 68 (5), 994e1000. Chen, P.Y., 1977. Table of key lines in X-ray powder diffraction patterns of minerals in clays and associated rocks. Indiana Geological Survey Occasional Paper 21, 1e67. Conway, F.M., Vallance, J.W., Rose, W.I., Johns, G.W., Paniagua, S., 1992. Cerro Quemado, Guatemala: the volcanic history and hazards of an exogenous volcanic dome complex. Journal of Volcanology and Geologic Research 52, 303e323. Cowgill, U.M., Hutchinson, G.E., 1963. El Bajo de Santa Fe. Transactions of the American Philosophical Society 53, 3e51. Dull, R.A., Southon, J.R., Sheets, P., 2001. Volcanism, ecology, and culture: a reassessment of the Volcan Ilopango TBJ eruption in the southern Maya Realm. Latin American Antiquity 12, 25e44. Dunning, N.P., Beach, T., 2000. Stability and instability in pre-Hispanic Maya Landscapes. In: Lentz, D. (Ed.), Imperfect Balace: Precolumbian New World Ecosystems. Columbia University Press, New York. Dunning, N.P., Beach, T., 2010. Farms and forests: spatial and temporal perspective on ancient Maya landscapes. In: Martini, P., Chesworth, W. (Eds.), Landscapes and Societies. Springer, London. Dunning, N., Beach, T., Farrell, P., Luzzadder-Beach, S., 1998. Prehispanic agricultural and adaptive regions in the Maya Lowlands. Culture and Agriculture 20, 87e101. Dunning, N.P., Beach, T., Luzzadder-Beach, S., 2006. Environmental Variability Among Bajos in the Southern Maya Lowlands and Its Implications for Ancient Maya Civilization and Archaeology. In Precolumbian Water Management, Tempe. University of Arizona Press, Tempe. Dunning, N.P., Griffin, R., Jones, J.G., Carr, C., Lane, B., Thomas, B., Weaver, E., 2009. Investigaciones de geoarqueologia y paleoambiente en la zona mayor de Tikal. In: Lentz, D.L., Grazioso Sierra, L., Dunning, N.P., Scarborough, V.L. (Eds.), Proyecto de Silvacultura y Manejo de Aguas de los Antiguos Mayas de Tikal: Temporada de 2009. Dirección Patrimonio Cultural y Natural de Guatemala, Guatemala City. Dunning, N.P., Houston, S., 2011. Chan Ik’: hurricanes as a disruptive force in the Maya Lowlands. In: Persson, B.L., Isendahl, C. (Eds.), Ecology, Power, and Religion in Maya Landscapes. Verlag Anton Saurwein, Berlin. Dunning, N.P., Wahl, D., Beach, T., Jones, J.G., Luzzadder-Beach, S., McCormick, C. Environmental instability and human response in the late Preclassic eastcentral Yucatan Peninsula. In. Iannone, G. (Ed.), The Great Maya Droughts in Cultural Context. University of Colorado Press, Boulder, in press. Drexler, J.W., Rose, W.I., Sparks, R.S.J., Ledbetter, M.T., 1980. The Los Chocoyos ash, Guatemala: a major stratigraphic Marker in Middle America and in three ocean basins. Quaternary Research 13, 327e345. Floyd, P.A., Winchester, J.A., 1978. Identification and discrimination of altered and metamorphosed volcanic rocks using immobile elements. Chemical Geology 21, 291e306. Ford, A., 1991. Economic variation of ancient Maya residential settlement in the upper Belize River area. Ancient Mesoamerica 2, 35e46. Ford, A., Fedick, S.L., 1992. Prehistoric settlement patterns in the upper Belize River area: initial results of the Belize River Archaeological Settlement Survey. Journal of Field Archaeology 19, 35e49. Ford, A., Glicken, H., 1987. The significance of volcanic ash tempering in the ceramics of the central Maya Lowlands. In: Rice, P.M., Sharer, R.J. (Eds.), Maya Ceramics. Papers from the 1985 Maya Ceramics Conference. BAR International Series, 345 Oxford. Ford, A., Rose, W.I., 1995. Volcanic ash in ancient Maya ceramics of the limestone lowlands: implications for prehistoric volcanic activity in the Guatemala highlands. Journal of Volcanology and Geothermal Research 66, 149e162. Gallopin, A.L., 1990. Water Storage at Tikal, Guatemala. Unpublished M.A. Thesis, Department of Anthropology, University of Cincinnati, Cincinnati. Gerke, T.L., Stocker, S.R., Davis, J.L., Maynard, J.B., Dietsch, C., 2006. Sourcing volcanic Millstones from Greco-Roman Sites in Albania source. Journal of Field Archaeology 31 (2), 137e146. Grim, R.E., Güven, N., 1978. BentonitesdGeology, Mineralogy, Properties and Uses: Developments in Sedimentology, vol. 24. Elsevier, New York.

2938

K.B. Tankersley et al. / Journal of Archaeological Science 38 (2011) 2925e2938

Hart, W.J.E., Steen-McIntyre, V., 1983. Tierra Blanca Joven tephra from the AD 260 eruption of the Ilopango Caldera. In: Sheets, P.D. (Ed.), The Zapotitan Valley of El Salvador. University of Texas Press, Austin. Huff, W.D., 2008. Ordovician K-bentonites: issues in interpreting and correlating ancient tephras. Quaternary International 178, 276e287. Huff, W.D., Bergström, S.M., Kolata, D.R., Muftuoglu, E., 1999. K-bentonite bed preservation and its event stratigraphic significance. Acta Universitatis Carolinae Geologica 43, 491e493. Huff, W.D., Bergström, S.M., Kolata, D.R., 2000. Silurian K-bentonites of the Dnestr basin, Podolia, Ukraine. Journal of the Geological Society, London 157, 493e504. Laird, D.A., Barak, P.E., Nater, A., Dowdy, R.H., 1991. Chemistry of smectitic and illitic phases in interstratified soil smectite. Soil Science Society of America Journal 55, 1499e1504. Laporte, Juan Pedro, 2003. Architectural aspects of interaction between Tikal and Teotihuacan during the Early Classic Period. In: Geoffrey Braswell (Ed.), The Maya and Teotihuacan: Reinterpreting Early Classic Interaction. University of Texas Press, Austin, TX, pp. 199e216. Martin, S., Grube, N., 2008. Chronicle of the Maya Kings and Queens: Deciphering the Dynasties of the Ancient Maya, second ed. Thames & Hudson, London and New York. Maynard, J.B., 1983. Geochemistry of Sedimentary Ore Deposits. Springer Press, Verlag, New York. Maynard, J.B., 1992. Chemistry of modern soils as a guide to interpreting Precambrian paleosols. Journal of Geology 100, 279e289. Moore, D., Reynolds, R.C., 1997. X-Ray Diffraction and the Identification and Analysis of Clay Minerals, second ed.. Oxford University Press, New York. Moreno, T., Querol, X., Castillo, S., Alastuey, A., Cuevas, E., Herrmann, L., Mounkaila, M., Elvira, J., Gibbons, W., 2006. Geochemical variations in aeolian mineral particles from the SaharaeSahel dust corridor. Chemosphere 65, 261e270. Mueller, A.D., Flavio, Islebe G.A., Anselmetti, S., Ariztegui, D., Brenner, M., Hodell, D.A., Hajdas, I., Hamann, Y., Haug, G.H., Kennett, D.J., 2010. Recovery of

the forest ecosystem in the tropical lowlands of northern Guatemala after disintegration of Classic Maya polities. Geology 38, 523e526. Muhs, D.R., Budahn, J.R., Prospero, J.M., Carey, S.N., 2007. Geochemical evidence for African dust inputs to soils of western Atlantic islands: Barbados, the Bahamas, and Florida. Journal of Geophysical Research 112, FO2009. doi:10.1029/ 2005JF000445. Robock, A., 2002. Volcanic eruption, El Chicho  cn. In: MacCracken, M.C., Perry, J.S. (Eds.), Encyclopedia of Global Environmental Change. John Wiley & Sons, Ltd, Chichester. Rose, W.I., Hahn, G.A., Drexler, J.W., Malinconico, M.L., Peterson, P.S., Wunderman, R.L., 1981. Quaternary tephra of northern Central America. In: Self, S., Sparks, R.S.J. (Eds.), Tephra Studies. Reidel, Dordrecht. Scarborough, V.L., 1993. Water management in the southern Maya Lowlands: an accretive model for the engineered landscape. In: Scarborough, V.L., Isaac, B.L. (Eds.), Economic Aspects of Water Management in the Prehispanic New World, Research in Economic Anthropology, (Sup.7). JAI Press, Greenwich. Scarborough, V.L., 1994. Water management as a function of locational and appropriational movements and the case of the Classic Maya of Tikal. In: Colin, A.M., Tandy, D.W. (Eds.), From Political Economy to Anthropology: Situating Economic Life in Past Societies. Black Rose Books, New York. Scarborough, V.L., 2003. The Flow of Power: Ancient Water Systems and Landscapes. School of American Research Press, Santa Fe. Scarborough, V.L., Gallopin, A.L., 2003. A water storage adaptation in the Maya Lowland. Science 251, 658e662. Sheets, P., 2002. Before the Volcano Erupted. University of Texas Press, Austin. Tankersley, K.B., Ballantyne, 2010. X-ray power diffraction analysis of late Holocene reservoir sediments. Journal of Archaeological Science 37, 133e138. West, G., 2002. Ceramic analysis in the late classic and post classic Maya Lowlands: a diachronic approach. In: Masson, M.A., Freidel, D.A. (Eds.), Ancient Maya Economies, pp. 140e196. Oxford. Winchester, J.A., Floyd, P.A., 1977. Geochemical discrimination of different magma series and their differentiation products using immobile elements. Chemical Geology 20, 325e343.