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
Department of Anthropology, University of Cincinnati, Cincinnati, OH 45221, USA
b
Department of Geography, University of Cincinnati, Cincinnati, OH 45221, USA
c
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: Powder X-ray diffraction and petrographic analyses of reservoir sediments from Tikal, Guatemala have
Received 21 March 2011 identified significant quantities of decomposed volcanic ash in the form of smectite and euhedral
Received in revised form bipyramidal quartz crystals. X-ray fluorescence trace element content analysis was used to eliminate
30 May 2011
distant Sahara-Sahel and Antilles sources. The Zr/Y and Ni/Cr ratios of reservoir sediment from Tikal are
Accepted 31 May 2011
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
Keywords:
fell during the Preclassic, Classic, and Postclassic Maya cultural periods. It may now be possible to
Powder X-ray diffraction
X-ray fluorescence
develop an effective chronology of ash fall at Tikal and the greater Peten.
Maya Lowlands Published by Elsevier Ltd.
Tikal
Tephra

1. Introduction capacity of more than 570,000 m3 (Scarborough and Gallopin,

2003:661). Surface water drained into the reservoirs and cultur-
While volcanic ash has been well documented as temper in ally modified aguadas (i.e., natural depressions) and bajos (i.e., huge
ancient Maya ceramics, its sources remain in question (Ford, 1991; solutional dolines) (Scarborough, 1993; 1994:116; 2003:51). Unlike
Ford and Fedick, 1992; Ford and Glicken, 1987; West, 2002). The deeply inundated deposits from lake basins and ocean floors,
plethora of ash temper in ceramics recovered from the limestone abandoned and in-filling reservoir sediments can be easily sampled
lowlands of Guatemala has led some investigators to suggest that with solid sediment drill cores.
ash fell during the pre-Hispanic period of Maya occupation (Ford While it is possible for ash to survive in deep lakes, such as Yojoa
and Rose, 1995:149). This theory has significant implications for in Honduras, this is not the case in smaller and shallower reservoirs.
understanding the prehistoric exploitation of volcanic resources, In these settings, a significant problem in sourcing ash from the
landscape modification, and sustainability in the Maya Lowlands. Maya Lowlands is the fact that volcanic glass quickly weathers (i.e.,
If significant quantities of volcanic ash fell on the limestone chemically decomposes) into smectite clay in moist, tropical and
lowlands of Guatemala during the pre-Hispanic occupation of the alkaline environments, all of which are characteristic to the region.
region, then we should expect to find direct positive evidence in the This phenomenon is exemplified by an ash fall from the El Chichón
numerous large reservoirs constructed in the Maya city of Tikal Volcano, which lasted from 28 March to 4 April 1982 (Robock,
(Fig. 1). The Maya constructed reservoirs at Tikal to conserve water 2002). Although Tikal was blanketed by several centimeters of
during the annual dry season and to control and contain floodwa- ash during this event, there is no visible evidence of the event in the
ters during the rainy months. Six major reservoir catchment areas soils today, and local residents report that much of the ash had
drained the elevated precincts of Tikal (Fig. 2), which covered an already been incorporated into the soil within a few months.
area of approximately 300 ha with a total maximum reservoir Smectite is a group of expanding-lattice clay minerals that
include beidellite, hectorite, montmorillonite, nantronite, saponite,
* Corresponding author. Department of Anthropology, University of Cincinnati,
and sauconite, and is the principal component of bentonite clay
Cincinnati, OH 45221, USA. Tel.: þ1 513 556 2772; fax: þ1 513 556 2778. deposits (Laird et al., 1991). Smectite originates from the decom-
E-mail address: [email protected] (K.B. Tankersley). position of eruptive igneous rocks (e.g., tuff) and volcanic ash (i.e.,

0305-4403/$ e see front matter Published by Elsevier Ltd.

doi:10.1016/j.jas.2011.05.025
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 forma- dramatically in the 9th century and by AD 900 was largely aban-
tion of smectite include magnesium-rich environments with poor doned except for a small residual population that persisted for
drainage, which are characteristic of the reservoirs of Tikal. The another 200 or so years.
expandable nature of clays of this type is revealed by x-ray
diffraction analysis of the separated clay fraction (Moore and 3. Corriental reservoir
Reynolds, 1997). Likely sources for the weathered ash can be
identified using bulk X-ray fluorescence (XRF) and electron Corriental is one of the largest reservoirs at Tikal (Fig. 3). It is
microprobe data as documented by Huff (2008) and Huff et al. strategically positioned to collect most of the surface water runoff
(1999, 2000) for multiple Ordovician bentonites. from the southeastern margins of the central city promontory
(Scarborough and Gallopin, 2003:251 and 252). The reservoir was
2. Archaeological context likely used as a water source for drinking, cooking, and probably
bathing as indicated by water jar sherds. Although there are no
Pre-Hispanic human occupation of the Maya Lowlands is often prepared steps or walkways associated with the reservoir, the
divided into the following chronological periods: PaleoIndian (pre- scarcity of ancient debris in the sediments and the apparent former
7000 BC), Archaic (7000e2200 BC), Preclassic (2200 BCe250 AD), existence of carbonate sand filters suggest that it was constructed
Early Preclassic (2200e1000 BC), Middle Preclassic (1000e400 BC), for public drinking water. The inferred presence of ancient sand
Late Preclassic (400 BCeAD100), Terminal Preclassic (AD 100e250), filtration berms positioned at the ingress of the reservoirdno
Classic (AD 250e900), Early Classic (AD 250e600), Late Classic (AD doubt occasionally “blown out” by capricious seasonal flooding
600e770), Terminal Classic (AD 770e900), Postclassic (AD eventsdis suggested by the repeated stratigraphic sand lensing
900e1500), Early Postclassic (AD 900e1250), and Late Postclassic reported in most excavation profiles; carbonate sand sources have
(AD 1250e1500). While PaleoIndian, Archaic, and Early Preclassic not been identified within 10 km of Tikal, suggesting an anthro-
materials and occupations have been found widely distributed pogenic origin. The catchment area for the Corriental Reservoir is
within the Maya Lowlands, no materials from these periods have about 40 ha, with a surface area of more than 15,000 m2 and an
thus far been recovered at Tikal. Sometime prior to about 700 BC, estimated capacity of more than 57,000 m3 of water (Gallopin,
small populations began to reside at Tikal, and by 600 BC the first 1990:60). It was constructed in a pre-existing localized depres-
monumental architectural constructions appeared (Laporte, 2003). sion at the junction of several small seasonal streams. The reservoir
By the Late Preclassic period (ca. 350 BC), Tikal had developed into featured two separate ingress gates at different elevations,
a significant “player” in the emerging political landscape of the approximately 205 m amsl on the south side and 208 m amsl on the
Maya Lowlands. Unlike larger centers in the nearby Mirador Basin, northwestern side (Gallopin, 1990:32e38).
Tikal survived the turmoil of the 2nd century AD and emerged as In 2009, ten pits and trenches were hand excavated from the
a major center of the Early Classic period (Dunning et al., in press). center and margins of the Corriental Reservoir, the northwestern
Tikal enjoyed variable prosperity during the Classic era, including and southern ingress gates, and the eastern egress (see Fig. 3).
a notable downturn in its fortune or “hiatus” in the 6th century AD, Additionally, a hand-operated Environmental Subsoil Probe (ESP)
but emerged as a paramount center in the Late Classic period with was used to extract twenty-six 2-cm diameter cores from
a population estimated to have been around 60,000 in the 8th Corrientald16 from the center of the reservoir, 7 from the earthen
century (Martin and Grube, 2008). Tikal’s affluence declined 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 and soil organic matter (SOM) (Table 2)dthe former abundantly
in the field by Dunning during excavations on the basis of color, associated with airborne wood ash from the many cooking fires and
texture, structure, and pedogenic features. Tankersley confirmed the like from within the low-density urban setting. This work
the reservoir-wide stratigraphy in the lab during analysis and established that the stratigraphy (i.e., soil horizons) extended from
correlation of the cores with particle size analysis, magnetic the early Holocene, 8960  60 14C yr BP (Beta-270566) to at least the
susceptibility, and Munsell soil color. Early Postclassic 990  40 14C yr BP (Beta-258720). It also suggested
that the Corriental Reservoir was in use by the Early Classic,
3.1. Stratigraphy and geochronology 1560  40 14C yr BP (Beta-274990), but may have been initially
constructed toward the end of the preceding Late Preclassic period.
The stratigraphy for the interior sediments and buried soils The 1560  40 14C yr BP (Beta-274990) AMS radiocarbon date
within the Corriental Reservoir, as defined in Operation 1C, are was obtained from core 17 on charred plant remains beneath the
outlined in Table 1. Operation 1C was a 1  1.5 m trench excavated earthen berm (i.e., wall) 3 m above the northwest gate (Fig. 4). This
near the center point of the Corriental Reservoir. The pit reached date demonstrates that this feature was built during or subsequent
a final depth of 3.15 m, at which point weathered limestone to the Early Classic. Stable carbon isotope values (d 13C 19.1, 19.5,
bedrock was encountered. Dr. Pat Culbert analyzed ceramics from and 20.3&) on the AMS radiocarbon dated soil organic matter
this operation at Tikal. suggests that C4 photosynthetic pathway plants such as maize
Radiocarbon dating at Corriental was performed in a series of were present in the Corriental catchment area during the Late
AMS radiocarbon measurements made on carbonized plant remains 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.

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
Table 1 (Dunning and Beach, 2010). The Early Classic berm additions may
Age and relative percent composition of sediments identified in the Corriental have been built to strengthen the northwestern ingress against
reservoir based on XRD. heavy runoff during strong seasonal rains and hurricanes, which
Depth Horizon Measured 2s Calcite Smectite Quartz Void are powerful geomorphic agents at any time throughout the region
14
C yr BP Calibrated (%) (%) (%) (%) (Dunning and Houston, 2011), as well as to increase the carrying
Age capacity of the reservoir. The lowermost horizons exposed in
10 A1 12.84 37.91 35.86 12.84 Operation 1C (i.e., 3Ab and 3AC horizons) represented a highly
20 A2 88.72 1.93 4.32 6.97
compacted, skeletal soil overlying limestone bedrock. A radio-
30 ACss 80.51 0.00 0.88 18.61
40 C2ss 70.75 14.70 5.41 12.40 carbon date on soil organic matter from the analogous soil horizon
50 C2ss 67.51 13.13 2.87 16.49 in Core 8 produced a date of 8960  60 14C yr B.P or early Holocene
60 C2ss 51.77 8.89 3.48 38.23 age (see below, Fig. 4 and Table 2). Nevertheless, based on simi-
70 C3 990  40 AD 1010e1170 73.72 9.30 3.12 15.53 larities to other deeply buried soils in small depressions elsewhere
80 C4ss 71.10 13.98 7.02 12.96
90 C4ss 47.01 37.91 35.86 12.84
in the northeast Peten and northwest Belize, this soil may date as
100 C4ss 88.72 1.93 4.32 6.97 far back as the late Pleistocene (ca. 11,000e13,000 BP) (Dunning
110 C4ss 80.51 0.00 0.88 18.61 et al., 2006; Beach et al., 2008).
120 C4ss 70.75 14.70 5.41 12.40 With the onset of wetter conditions in the mid-Holocene period,
130 C5 67.51 13.13 2.87 16.49
the small depression filled with sediment eroded from upslope
140 C6ss 51.77 8.89 3.48 38.23
150 C6ss 2010  40 340e30 BC 73.72 9.30 3.12 15.53 areas (2C1 and 2C2 horizons) on which a new soil surface (2Ab and
160 C6ss 71.10 13.98 7.02 12.96 2AC horizons) gradually developed. Radiocarbon dating of organic
170 C7 19.56 73.59 2.68 5.40 matter within the 2C1 horizon place the age of this soil surface at
180 C7 2110  40 190e80 BC 17.73 77.77 1.46 3.88 approximately 2340  40 14C yr BP (Middle Preclassic). However,
190 C7 2120  40 380e170 BC 52.31 39.91 6.32 5.48
200 C8 78.70 11.59 9.81 6.31
this date is based on soil organic matter (organic matter which
210 C9 68.33 21.62 0.00 10.05 accumulated over hundreds of years, given the lack of significant
220 C10 82.41 4.92 3.46 11.30 occupation and associated cooking fires at this time), hence the soil
230 C11ss 80.27 10.02 4.68 8.22 surface was probably last exposed sometime in the Late Preclassic.
240 C12 69.76 18.11 11.14 9.78
This soil is a Vertisol, typical of small seasonally wet/dry depres-
250 2Abss 84.16 19.86 12.30 13.21
260 2ACss 83.61 4.43 4.43 31.56 sions in the Peten. Four weathered sherds were recovered from this
270 2C1bss 2340  40 760e400 BC 79.43 9.65 2.02 33.01 level, but were chronologically non-diagnostic.
280 2C1bss 92.20 6.56 4.76 11.29 At some point within the next 500 years, drainage within the
290 2C2bss 76.03 9.86 11.48 10.69 depression appears to have been substantially modified and sedi-
300 2C2bss 63.70 18.84 12.04 13.78
310 3Abss 8960  60 8290e7970 BC 90.48 15.74 3.97 19.78
ment 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 14

C yr BP 2 s Calibrated Age Cultural Period

Beta-258720 Charcoal 26.9 C3 990  40 AD 1010e1170 Early Postclassic

Beta-274990 Charcoal 23.7 Anthrosol 1560  40 AD 400e570 Early Classic
Beta-280839 SOM 19.5 C6 2010  40 340e30 BC Late Preclassic
Beta-266124 SOM 20.2 C7 2110  40 190e80 BC Late Preclassic
Beta-280837 SOM 20.3 C7 2120  40 380e170 BC Late Preclassic
Beta-258721 SOM 19.1 2C1 2340  40 760e400 BC Middle Preclassic
Beta-270566 SOM 23.1 3Ab 8960  60 8290e7970 BC Pre-habitation?

Reservoir, probably accomplished by a combination of quarrying the reservoir. One possible explanation is that sand was used to
(widening the natural depression), building up of the encircling filter water entering the reservoir and that sand from the filters was
berm, and diversion of inflowing seasonal stream runoff. Subse- occasionally flushed into the reservoir proper by storm-related
quently, the floor of the reservoir began to aggrade (fill with sedi- flooding. On the other hand, the organic clay layers are typical of
ment over time). Alternating strata of organic clay and laminated still water deposits and likely accumulated slowly while the
carbonate sand and small, rounded gravel are found between 65 reservoir was in active use. There are no obvious signs of dredging
and 253 cm depth within Operation 1C (C3 to C12 horizons). The within the sediments revealed in Operation 1C or other excavations
laminated or stratified nature of the deposits indicates that the and cores in the reservoir.
sands were deposited by running water (i.e., a fluvial process). Three radiocarbon dates were obtained from organic matter in the
Notably, there are no known natural sand sources upstream from 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.
2930 K.B. Tankersley et al. / Journal of Archaeological Science 38 (2011) 2925e2938

- 3.06Å CALCITE
SMECTITE, ZEOLITES, QUARTZ
SMECTITE, ZEOLITES

- 2.29Å CALCITE
SMECTITE, ZEOLITES

CALCITE
CALCITE

CALCITE
- 2.5Å CALCITE
- 3.37Å QUARTZ
- 3.88Å CALCITE

- 1.88Å
- 2.1Å

- 1.92Å
- 15.63Å

- 1.61Å

- 1.53Å
- 5.98Å

- 4.5Å

- 2.58Å

- 2.35Å

- 1.5Å
15.3
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).
 K.B. Tankersley et al. / Journal of Archaeological Science 38 (2011) 2925e2938 2931

CALCITE
- 3.05Å
- 4.48Å SMECTITE, ZEOLITES, QUARTZ
- 14 -17Å SMECTITE, ZEOLITES

SMECTITE, ZEOLITES

- 2.29Å CALCITE

CALCITE
CALCITE
- 2.5Å CALCITE
- 3.88Å CALCITE

QUARTZ

- 1.61Å CALCITE
- 2.03Å
- 2.1Å
- 2 35Å

- 1.88Å
- 1.92Å
- 3.37Å

- 2.86Å
- 5.82Å

- 1.63Å
- 2.58Å
80.4

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

6.0 TK2 10-20

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 of this organic matter also indicates that it is derived from C3
Preclassic period and are appreciably older than the ceramic sherds terrestrial plants and not aquatics such as algae. Notably, soil surfaces
found in the underlying C8 through C12 horizons. This inversion dating from the Late Preclassic period are typical of the Maya
strongly suggests that the organic matter being dated in the C6 and C7 Lowlands, but they were subject to pulses of erosion resulting in the
horizons is derived from older soil eroded from surfaces in the deposition of sediments derived from these soils in countless local
watershed above the reservoir. Analysis of the d 13C and d 15N content depressions (Anselmetti et al., 2007; Dunning and Beach, 2000).
2932 K.B. Tankersley et al. / Journal of Archaeological Science 38 (2011) 2925e2938

CALCITE
- 3.05Å
SMECTITE, ZEOLITES, QUARTZ
SMECTITE, ZEOLITES

CALCITE
SMECTITE, ZEOLITES

CALCITE
CALCITE
CALCITE
- 3.88Å CALCITE

QUARTZ

CALCITE
- 2.29Å

- 1.88Å
- 14 -17Å

- 2.5Å

- 1.92Å
- 2.1Å
- 4.53Å

- 3.36Å

- 1.61Å
- 1.63Å

- 1.53Å
11.9
- 6Å

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 have continued to be in use to some extent as late as the Early
C12 horizons principally in the sandy strata. Small and large water Postclassic period.
jar forms predominate in all strata. The large majority of sherds Subsequently, there is no evidence of reservoir use, though it
were too weathered to be chronologically diagnostic. C12, the has naturally continued to seasonally collect water. The modern soil
deepest alluvial stratum, and C9 included identifiable Early Classic that has developed within the reservoir (Oi through C2 horizons) is
types. C10 had no diagnostic sherds. C8 contained a mix of Early and a Terric Fibrist, an organic soil with mineral subsoil typical of
Late Classic types. C5eC7 included only Late Classic types. C3 had regional depressions that remain partially moist year-round.
no diagnostic sherds. Charcoal within the C3 horizon (65 cm) In summary, the Corriental Reservoir, on the southern flank of
produced an AMS radiocarbon date of 990  40 14C yr BP (2 sigma central Tikal, was likely constructed sometime toward the end of
correlated range: AD 1010e1170), suggesting that the reservoir may the Late Preclassic or very early in the Early Classic period. The
 K.B. Tankersley et al. / Journal of Archaeological Science 38 (2011) 2925e2938 2933

reservoir was constructed by widening a small pre-existing natural 4. Powder X-ray diffraction
depression by quarrying and mounding earth to form an encircling
berm and filled by diverting water flow from a local seasonal Powder XRD was used to identify the relative percent mineral
stream. Thick sediment deposits within the reservoir included composition of the reservoir sediments of Tikal. For this study, solid
alternating deposits of stratified carbonate sands and organic clays. sediment core samples were extracted from the center of Corriental
The clays indicate periods of stability during which clay and organic to obtain the deepest and most complete stratigraphic samples
matters gradually settled onto the reservoir floor. The sandy strata possible. Initially, the cores were cut into 10 cm subsamples. Soil
are indicative of running water, perhaps deposited during higher- strata were defined on the basis of particle size analysis, sedi-
energy storm runoff events. The origin of the sand is uncertain, mentary boundaries, and changes in Munsell soil color, then
but it may have been used to filter water as it entered the reservoir, correlated with the horizons established in Operation 1C.
and then was occasionally flushed into the reservoir proper during
flooding. Ceramics recovered from within the reservoir sediments 4.1. XRD methods
were generally very weathered, but contained a mixture of Early
Classic and Late Classic types. Most notable were the quantities of Following the methodology of Tankersley and Ballantyne (2010),
huge jar fragments. XRD samples were taken at 10-cm intervals and sieved through
- 4.41Å SMECTITE, ZEOLITES, QUARTZ
SMECTITE, ZEOLITES

CALCITE

CALCITE
- 1.87Å CALCITE
- 3.31Å QUARTZ

- 1.6Å CALCITE
- 3.83Å CALCITE

- 3.01Å

- 2.27Å CALCITE
- 2.49Å CALCITE

- 1.54Å
- 2.03Å

- 1.62Å
- 2.08Å
- 14 -20Å

- 4.22Å

- 2.55Å

- 1.9Å
- 2.34Å
- 2.82Å

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).
2934 K.B. Tankersley et al. / Journal of Archaeological Science 38 (2011) 2925e2938

a 2 mm mesh. Approximately 20 g of clay from each stratum was

146
131
123
110
126
211
214
214
202
198
158

310
107
153
132
108
154
mixed with deionized water to make slurry in 100 ml beakers. Clay

na
Zr
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

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

47
na
Y
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

88.3
173
134
136
168
211
prepared from each sample and equilibrated overnight with

46
41
35
32
39
77
82
87
83
79
46
na
V
ethylene glycol vapor.
XRD patterns were obtained for both the air-dried and glyco-

160

179
383
662
485
420
568
40
41
42
41
42
22
22
24
27
25

na
Sr
lated 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

61.4
27.9
60.8
41.5
31.2
Rb
17
17
19
16
16
32
32
31
30
30
55

21
na
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

62.3
23.3
12.1
9.74
peak position and peak intensity as described by Chen (1977).

41
38
35
28
32
69
74
76
61
67
17

23

37
na
Ni
Glycolated samples were prepared to test for the presence of
expandable clay minerals. Relative mineral percentages were

5.13
12.7

2.32
7.3
Nb
10
10

10
15
16
15
14
15

30

25
na
calculated from the total counts per second (cps), which were

9
9

4
totaled for each 10-cm sample. The sum of all peaks cps for each

197
mineral was divided by the total cps for each 10-cm sample. The

38
37
33
33
31
79
78
87
76
77
12

71

43
19
12
56
na
Cr
relative percent of each mineral was calculated from the total cps
per mineral divided by the total cps in the 10-cm sample.

217
350
340
132
265
289
264
290
193
217
813

449
246
600
654
695
502
na
Ba
4.2. XRD results

P2O5
0.12
0.14
0.16
0.16
0.13
0.01
0.01
0.01
0.01
0.01
0.11

0.24
0.12
0.24
0.30
0.21
1.36
na
The minerals calcite, smectite, and quartz were found in the
Corriental Reservoir sediments (Figures 5e8). Calcite was the most
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
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
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

remainder of the sediments consisted of smectite (27%), charac-

terized 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
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
unaffected by glycolation.
The abundance of calcite is undoubtedly related to the Creta-
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
ceous and Tertiary limestone bedrock, which determines the terrain
CaO

in the Maya Lowlands (Dunning et al., 1998). Because quartz is

known to occur in carbonate rock (Chafetz and Zhang, 1998),
Trace element content of Tikal reservoir sediments and comparative volcanic sources (wt% or ppm).

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
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
MnO
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
quartz and smectite were completely absent from the bedrock 9.63
samples. This finding suggests that smectite and quartz entered the
reservoirs as aeolian minerals rather than dissolution of the
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

surrounding bedrock and subsequent water transport (contra

Cowgill and Hutchinson, 1963:41). Grim and Güven (1978:128)
Al2O3
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
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
SiO2
180e190
190e200
200e210
210e220
220e230
180e190
190e200
200e210
210e220
220e230

average
average
average
average
average
average
average
Depth
Corriental
Corriental
Corriental
Corriental
Corriental
Reservoir

Perdido
Perdido
Perdido
Perdido
Perdido

TephraSaharan Dust
Tierra Blanca Joven

Moreno et al., 2006.

El Salvador Tephra
Guatemala Tephra

Honduras Teprha

na ¼ not analyzed.
Ilopango Tephra

Antilles Tephra

Carr et al., 2007.

Mexico Tephra

Dull et al., 2001

Location
Table 3

Tikal
Tikal
Tikal
Tikal
Tikal
Tikal
Tikal
Tikal
Tikal
Tikal

Fig. 9. Photomicrograph of a representative euhedral bipyramidal quartz crystal from

the C7 horizon of Corriental reservoir sediment (see Table 1).
b
a

c
 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 sedi-
ments 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 Core samples, which contained high relative percentages of
varying in composition from rhyolitic (quartz-rich) to basaltic smectite were selected for XRF analysis from the C7 and C8 soil
(quartz-poor). However, most bentonites have formed from ash horizons of Corriental (i.e., Preclassic, Classic, and Postclassic
ranging from rhyolitic to dacitic in composition. In other words, the strata). As a control, samples from the pre-Maya strata of the
association of quartz and smectite in altered volcanic ashes is quite nearby Perdido reservoir (see Fig. 2), AMS radiocarbon dated
common. 15,110  60 14C yr BP (Beta-289286), 15,480  60 14C yr BP (Beta-
Grains isolated from all soil horizons were identified as quartz 289285), and 15,310  60 14C yr BP (Beta-289284), were also
bearing with XRD (see Table 1). Under high magnification (up to 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 given the difficulties associated with waterborne contaminants
elements; Mo, Ba, Co, Cr, Cu, Nb, Pb, Rb, Sr, Th, U, V, Y, and Zn. affecting any tropical setting. Secondly, the timing of the initial Late
Following the methodology of Gerke et al. (2006:140), sample Preclassic construction and harvesting of water from the reservoir
powders were pressed into briquettes at 2000 psi. A separate coincides with the posited drought-like conditions recently argued
aliquot was heated to 1000  C for 1 h to measure volatile content. for this period (Dunning et al., in press). And the Classic period build
Intensity data were converted to parts per million (ppm) using up of the northwestern gate berm may suggest the return of a wetter
bivariate and multiple variable regressions applied to United States period, with the necessity of securing the principal ingress into the
Geological Survey, National Institute of Standards, and Japan reservoir to prevent excess waters from eroding the gate. At this
Geological Survey rock standards (Table 3). point, our understanding of regional paleoclimate does not allow us
to determine if any of the volcanic eruptions that dumped ash on
5.2. XRF results Tikal also affected local, regional, or global climate.
Nevertheless, the most revealing data sets come from the XRD
Significant differences in the trace element content of sediment and XRF assessments of weathered volcanic ash over Tikal and by
from the reservoirs of Tikal and Sahara-Sahel dust were found in extension over much of the Maya Lowlands through time. The
the ratios of Ni, Cr, Zr, and Y. These discriminations can be illus- impact of volcanism went far beyond the immediacy of the Guate-
trated in a plot of the ratio of Cr to Ni against the ratio of Zr to Y malan or Chiapan highlands into the limestone rich soils of the Maya
(Fig. 10). Ratios are used rather than absolute amounts to eliminate Lowlands. The study indicates the wealth of information sealed and
the effect of the large amounts of locally derived calcite. The trace potentially extractable from any abandoned reservoir setting. While
element content of Sahara-Sahel dust has a significantly lower reservoir sediments can be compromised by human activities such
range of Cr/Ni ratio and a higher Zr/Y ratio than do the samples as dredging, this was not the case at Corriental. Such reservoirs can
from Tikal. Using the same comparison for volcanic materials from provide significant paleoenvironmental data because (1) we can
the Lesser Antilles, Tikal has comparable ratios of Zr/Y, but a greater obtain many more controlled cores, and (2) the diminutive size of
ratio of Ni/Cr (see Fig. 10). The Ni/Cr trace element content of Tikal a tank allows a much more meaningfuldand accuratedsample of an
sediment is, however, comparable to volcanic rocks and tephra entire basin than a lake. Reservoirs can act as a corrective for the
from Guatemalan or Salvadoran volcanoes (Fig. 11). “randomness” of lake core extraction. Ideally, lake and ocean floor
The Ilopango TBJ eruption is the largest and best-documented cores should be used in combination with reservoir sediment data.
Holocene volcanic event in Central America (Hart and Steen-
McIntyre, 1983; Sheets, 2002). It occurred during the Early Classic 7. Conclusion
Period, sometime between A.D. 408 and 536 and its ecological and
cultural impact would have been felt throughout the Maya region Corriental is an ancient Maya reservoir at Tikal, Guatemala. It
(Dull et al., 2001). Professor Payson Sheets of the University of provides the first evidence of volcanic ash fall in the Maya Lowlands
Colorado kindly provided our team with samples of the TBJ tephra during pre-Hispanic times. Corriental is a unique test location to
for comparison. Data for other volcanic components are available in examine the chronology and stratigraphy in the reservoir, the
the database developed by Carr (Carr et al., 2007). XRD analysis aeolian minerals, and their potential source regions. Unlike large
demonstrated that the TBJ tephra is composed of plagioclase feld- lakes, Maya reservoirs are restricted areas with a known constric-
spar with lesser amounts of quartz and a large amount of glass. In tion point over which volcanic minerals are deposited.
other words, there are no clays in the TBJ tephra, because the glass Powder XRD and petrography show that volcanic-derived
is still intact, unlike the Tikal sediments where all the glass has minerals, smectite and euhedral bipyramidal quartz, are present
converted to smectite. It is possible that the abundance of calcite in in significant quantities in the sediments of Corriental. AMS
the Tikal reservoirs compared to the slopes of the Ilopango volcano radiocarbon dating of the Corriental reservoir demonstrates that
accounts for this difference (Cowgill and Hutchinson, 1963). glass (now smectite) and euhedral quartz were deposited during
Weathering, however, should not affect the trace element ratios the Preclassic, Classic, and Postclassic cultural periods. Although
for high-field strength elements like Cr, Nb, Ti, Y, and Zr (Winchester smectite is known to occur in Sahara-Sahel dust, XRF trace element
and Floyd,1977; Floyd and Winchester,1978; Maynard,1992). Nickel analysis indicates a different origin, one comparable to tephra from
is more mobile than Cr in acidic soils under tropical conditions Guatemalan or Salvadoran volcanoes (Cabadas-Baez et al., 2010).
(Maynard, 1983), and accumulates lower in the profile, which is the Fingerprinting the exact sources of these minerals will, however,
mechanism of genesis for lateritic nickel deposits. However, it require a comprehensive survey of tephra deposits from the Maya
should remain fixed in the high carbonate environment of these Highlands, Mexico, and the greater Caribbean region.
deposits. The Ni/Cr ratios in the individual cores at Tikal are constant The presence of airborne volcanic minerals in the reservoir
with depth, indicating that vertical migration of Ni has been sediments of Tikal supports a model of centuries-long periods of
minimal. volcanism during the prehistoric Maya occupation of the limestone
Our XRF analysis of the TBJ tephra sample found a higher ratio of lowlands (Ford and Glicken, 1987). As suggested by Ford and Rose
Zr/Y than the sediments from the reservoirs of Tikal. Similar results (1995:159), the region likely experienced regular and periodic
were obtained using the analyses in the Carr database. Note that the tephra eruptions and ash falls into the lowlands. Ash temper and
position for Ilopango in Fig. 11 is an average of our sample and the assemblage of crystals in Maya pottery include biotite, hornblende,
Carr data. hypersthene, and zircon, which all are minerals consistent with
Guatemala Highland tephra (Drexler et al., 1980; Rose et al., 1981).
6. Implications The presence of biotite suggests El Chichon, Tajamulco, Acatenango,
and Atitlán as possible volcanic sources, which were active in the
Several insights can be gleaned from this study, somewhat period of AD 600e900 and Cerro Quemado, which was active at
independently of the surprising amounts of weathered volcanic ash about AD 800 and for some time there after (Conway et al., 1992;
now revealed from Tikal. First, the pedological and stratigraphic Ford and Rose, 1995). Because each reservoir stratum has its own
controls suggest the presence of a sand filtration system presumably unique trace element composition, different volcanoes were likely
to purify the water source in the Corriental Reservoir. Such a tech- erupting at different moments in time. Montmorillonite (i.e.,
nology was previously unknown, but it is not altogether surprising smectite) in Lake Péten Itzá sediments radiocarbon dated A.D. 880
 K.B. Tankersley et al. / Journal of Archaeological Science 38 (2011) 2925e2938 2937

to 1140 and El Bajo de Santa Fé may also correlate with these Lane, Benjamin Thomas, and Eric Weaver accomplished the coring
volcanic events (Cowgill and Hutchinson, 1963:41; Mueller et al., of Corriental Reservoir. Payson Sheets provided a tephra sample
2010:525). from Ilopango Volcano.
The presence of airborne volcanic minerals also supports the
position that the Maya living in the limestone lowland had References
a dependable source of volcanic ash for the production of ceramics.
Indeed, tephra is superior to the temper used as early as the Pre- Anselmetti, F.S., Hodell, D.A., Ariztegui, D., Brenner, M., Rosenmeier, M.F., 2007.
classic and Early Classic as a result of its angularity (West, 2002). Quantification of soil erosion rates related to ancient Maya deforestation.
Geology 35, 915e918.
Also given the reverence of the Maya for sacred volcanoes, the Beach, T., Luzzadder-Beach, S., Dunning, N.P., Cook, D., 2008. Human and natural
tephra likely had religious implications. It may have been collected impacts on fluvial and Karst systems in the Maya Lowlands. Geomorphology
at Tikal and the greater Peten from air-fall deposits and acquired by 101, 301e331.
Cabadas-Báez, H., Solleiro-Rebolledo, E., Sedov, S., Pi-Puig, T., Gama-Castro, J., 2010.
trade from sources in the Maya Highlands.
Pedosediments of karstic sinkholes in the eolianites of NE Yucatán: a record of
Given the accounts of the 1982 ash fall from El Chichón in the Late Quaternary soil development, geomorphic processes and landscape
Peten it is clear that it would have been very easy for the Maya to stability. Geomorphology 122, 323e337.
collect and stockpile large quantities of ash. Although it is beyond 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
the scope of this paper, if we are able to develop an effective Hazards. Taylor and Francis, London, pp. 565e590. Chapter 22.
chronology of ash fall at Tikal and the greater Peten, then it might Chafetz, H.S., Zhang, J., 1998. Authigenic euhedral megaquartz crystals in a Quater-
be possible to compare this record to chronological patterns in nary dolomite. Journal of Sedimentary Research 68 (5), 994e1000.
Chen, P.Y., 1977. Table of key lines in X-ray powder diffraction patterns of minerals
temperingdto say nothing about chronological control of strati- in clays and associated rocks. Indiana Geological Survey Occasional Paper 21,
graphic deposits. In addition to the use of ash for temper, smectite is 1e67.
an important raw material for ceramic production because of its Conway, F.M., Vallance, J.W., Rose, W.I., Johns, G.W., Paniagua, S., 1992. Cerro Que-
mado, Guatemala: the volcanic history and hazards of an exogenous volcanic
plasticity. dome complex. Journal of Volcanology and Geologic Research 52, 303e323.
It has long been assumed that tropical soils in the Peten and Cowgill, U.M., Hutchinson, G.E., 1963. El Bajo de Santa Fe. Transactions of the
surrounding areas were the result of weathering of the local American Philosophical Society 53, 3e51.
Dull, R.A., Southon, J.R., Sheets, P., 2001. Volcanism, ecology, and culture: a reas-
carbonate bedrock. Ash enrichment of the soils would have also sessment of the Volcan Ilopango TBJ eruption in the southern Maya Realm. Latin
enhanced soil fertility by increasing the porosity, permeability, and American Antiquity 12, 25e44.
nutrients needed for maize agriculture, thus helping support pop- Dunning, N.P., Beach, T., 2000. Stability and instability in pre-Hispanic Maya
Landscapes. In: Lentz, D. (Ed.), Imperfect Balace: Precolumbian New World
ulation increases at Tikal and elsewhere (Ford and Rose, 1995). It is
Ecosystems. Columbia University Press, New York.
likewise possible that a short-term problem generated by heavy fall Dunning, N.P., Beach, T., 2010. Farms and forests: spatial and temporal perspective
would have been the clogging of water catchment and storage on ancient Maya landscapes. In: Martini, P., Chesworth, W. (Eds.), Landscapes
systems. Additionally, ash falls would have directly affected local and Societies. Springer, London.
Dunning, N., Beach, T., Farrell, P., Luzzadder-Beach, S., 1998. Prehispanic agricultural
flora and fauna proportional to depth. For example, it would have and adaptive regions in the Maya Lowlands. Culture and Agriculture 20,
clogged the spiracles of insects and suffocated them, thus, insect 87e101.
pollinated plants would have suffered. Additionally, volcanic ash Dunning, N.P., Beach, T., Luzzadder-Beach, S., 2006. Environmental Variability
Among Bajos in the Southern Maya Lowlands and Its Implications for Ancient
would have been deleterious to aquatic food resources such as Maya Civilization and Archaeology. In Precolumbian Water Management,
shellfish, fish, and manatees. Tempe. University of Arizona Press, Tempe.
Although difficult to quantify, several of our stratigraphic 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.
columns examined in 10 cm arbitrary levels yielded volcanic ash In: Lentz, D.L., Grazioso Sierra, L., Dunning, N.P., Scarborough, V.L. (Eds.),
and quartz ejecta in amounts as high as 70% by volume (see Proyecto de Silvacultura y Manejo de Aguas de los Antiguos Mayas de Tikal:
Table 1). It is important to further note that preliminary analysis of Temporada de 2009. Dirección Patrimonio Cultural y Natural de Guatemala,
Guatemala City.
other soils and reservoir sediments from Tikal suggest that smectite
Dunning, N.P., Houston, S., 2011. Chan Ik’: hurricanes as a disruptive force in the
clay derived from weathered volcanic ash is a major component of Maya Lowlands. In: Persson, B.L., Isendahl, C. (Eds.), Ecology, Power, and Reli-
the inorganic fraction of these soils. Future investigation may reveal gion in Maya Landscapes. Verlag Anton Saurwein, Berlin.
Dunning, N.P., Wahl, D., Beach, T., Jones, J.G., Luzzadder-Beach, S., McCormick, C.
that aeolian volcanic material is, indeed, a dominant component of
Environmental instability and human response in the late Preclassic east-
the mineral portion of the soils across wide areas of the Maya central Yucatan Peninsula. In. Iannone, G. (Ed.), The Great Maya Droughts in
Lowlands. 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
Acknowledgments basins. Quaternary Research 13, 327e345.
Floyd, P.A., Winchester, J.A., 1978. Identification and discrimination of altered and
This study was made possible with funding from the Court metamorphosed volcanic rocks using immobile elements. Chemical Geology 21,
291e306.
Family Foundation, the Charles Phelps Taft Foundation, the Ford, A., 1991. Economic variation of ancient Maya residential settlement in the
Alphawood Foundation, the Wenner-Gren Foundation, and a grant upper Belize River area. Ancient Mesoamerica 2, 35e46.
from the National Science Foundation (BCS0810118) to David Lentz, 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
Vernon Scarborough, and Nicholas Dunning. Fieldwork was of Field Archaeology 19, 35e49.
undertaken in 2009 as part of the University of CincinnatidIDAEH Ford, A., Glicken, H., 1987. The significance of volcanic ash tempering in the
Tikal Project co-directed by David Lentz and Liwy Grazioso Sierra. 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
The laboratory assistance of Meredith Coates, Nathan Marshall, Jon Series, 345 Oxford.
Paul McCool, Jim Milawski, Andras Nagy, Leslie Neal, and Tony Ford, A., Rose, W.I., 1995. Volcanic ash in ancient Maya ceramics of the limestone
Tamberino were invaluable. Chris Carr led digitization and rectifi- lowlands: implications for prehistoric volcanic activity in the Guatemala
highlands. Journal of Volcanology and Geothermal Research 66, 149e162.
cation of the original University of Pennsylvania map of central
Gallopin, A.L., 1990. Water Storage at Tikal, Guatemala. Unpublished M.A. Thesis,
Tikal and produced Fig. 3. Eric Weaver led field mapping of the Department of Anthropology, University of Cincinnati, Cincinnati.
Corriental Reservoir in 2009. Liwy Grazioso Sierra directed many of Gerke, T.L., Stocker, S.R., Davis, J.L., Maynard, J.B., Dietsch, C., 2006. Sourcing volcanic
the excavations in Corriental Reservoir assisted by Raquel Macario, Millstones from Greco-Roman Sites in Albania source. Journal of Field Archae-
ology 31 (2), 137e146.
Sheryl Carcruz, Ana Arriola, and Marielos Corado. Pat Culbert Grim, R.E., Güven, N., 1978. BentonitesdGeology, Mineralogy, Properties and Uses:
analyzed the ceramics recovered from these excavations. Brian 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 the forest ecosystem in the tropical lowlands of northern Guatemala after
eruption of the Ilopango Caldera. In: Sheets, P.D. (Ed.), The Zapotitan Valley of El disintegration of Classic Maya polities. Geology 38, 523e526.
Salvador. University of Texas Press, Austin. Muhs, D.R., Budahn, J.R., Prospero, J.M., Carey, S.N., 2007. Geochemical evidence for
Huff, W.D., 2008. Ordovician K-bentonites: issues in interpreting and correlating African dust inputs to soils of western Atlantic islands: Barbados, the Bahamas,
ancient tephras. Quaternary International 178, 276e287. and Florida. Journal of Geophysical Research 112, FO2009. doi:10.1029/
Huff, W.D., Bergström, S.M., Kolata, D.R., Muftuoglu, E., 1999. K-bentonite bed 2005JF000445.
preservation and its event stratigraphic significance. Acta Universitatis Caro- Robock, A., 2002. Volcanic eruption, El Chicho  cn. In: MacCracken, M.C., Perry, J.S.
linae Geologica 43, 491e493. (Eds.), Encyclopedia of Global Environmental Change. John Wiley & Sons, Ltd,
Huff, W.D., Bergström, S.M., Kolata, D.R., 2000. Silurian K-bentonites of the Chichester.
Dnestr basin, Podolia, Ukraine. Journal of the Geological Society, London 157, Rose, W.I., Hahn, G.A., Drexler, J.W., Malinconico, M.L., Peterson, P.S.,
493e504. Wunderman, R.L., 1981. Quaternary tephra of northern Central America. In:
Laird, D.A., Barak, P.E., Nater, A., Dowdy, R.H., 1991. Chemistry of smectitic and illitic Self, S., Sparks, R.S.J. (Eds.), Tephra Studies. Reidel, Dordrecht.
phases in interstratified soil smectite. Soil Science Society of America Journal Scarborough, V.L., 1993. Water management in the southern Maya Lowlands: an
55, 1499e1504. accretive model for the engineered landscape. In: Scarborough, V.L., Isaac, B.L.
Laporte, Juan Pedro, 2003. Architectural aspects of interaction between Tikal and (Eds.), Economic Aspects of Water Management in the Prehispanic New World,
Teotihuacan during the Early Classic Period. In: Geoffrey Braswell (Ed.), The Research in Economic Anthropology, (Sup.7). JAI Press, Greenwich.
Maya and Teotihuacan: Reinterpreting Early Classic Interaction. University of Scarborough, V.L., 1994. Water management as a function of locational and
Texas Press, Austin, TX, pp. 199e216. appropriational movements and the case of the Classic Maya of Tikal. In:
Martin, S., Grube, N., 2008. Chronicle of the Maya Kings and Queens: Deciphering Colin, A.M., Tandy, D.W. (Eds.), From Political Economy to Anthropology: Situ-
the Dynasties of the Ancient Maya, second ed. Thames & Hudson, London and ating Economic Life in Past Societies. Black Rose Books, New York.
New York. Scarborough, V.L., 2003. The Flow of Power: Ancient Water Systems and Land-
Maynard, J.B., 1983. Geochemistry of Sedimentary Ore Deposits. Springer Press, scapes. School of American Research Press, Santa Fe.
Verlag, New York. Scarborough, V.L., Gallopin, A.L., 2003. A water storage adaptation in the Maya
Maynard, J.B., 1992. Chemistry of modern soils as a guide to interpreting Precam- Lowland. Science 251, 658e662.
brian paleosols. Journal of Geology 100, 279e289. Sheets, P., 2002. Before the Volcano Erupted. University of Texas Press, Austin.
Moore, D., Reynolds, R.C., 1997. X-Ray Diffraction and the Identification and Analysis Tankersley, K.B., Ballantyne, 2010. X-ray power diffraction analysis of late Holocene
of Clay Minerals, second ed.. Oxford University Press, New York. reservoir sediments. Journal of Archaeological Science 37, 133e138.
Moreno, T., Querol, X., Castillo, S., Alastuey, A., Cuevas, E., Herrmann, L., West, G., 2002. Ceramic analysis in the late classic and post classic Maya Lowlands:
Mounkaila, M., Elvira, J., Gibbons, W., 2006. Geochemical variations in aeolian a diachronic approach. In: Masson, M.A., Freidel, D.A. (Eds.), Ancient Maya
mineral particles from the SaharaeSahel dust corridor. Chemosphere 65, Economies, pp. 140e196. Oxford.
261e270. Winchester, J.A., Floyd, P.A., 1977. Geochemical discrimination of different magma
Mueller, A.D., Flavio, Islebe G.A., Anselmetti, S., Ariztegui, D., Brenner, M., series and their differentiation products using immobile elements. Chemical
Hodell, D.A., Hajdas, I., Hamann, Y., Haug, G.H., Kennett, D.J., 2010. Recovery of Geology 20, 325e343.