This article appeared in a journal published by Elsevier. The attached
copy is furnished to the author for internal non-commercial research
and education use, including for instruction at the authors institution
and sharing with colleagues.
Other uses, including reproduction and distribution, or selling or
licensing copies, or posting to personal, institutional or third party
websites are prohibited.
In most cases authors are permitted to post their version of the
article (e.g. in Word or Tex form) to their personal website or
institutional repository. Authors requiring further information
regarding Elsevier’s archiving and manuscript policies are
encouraged to visit:
http://www.elsevier.com/copyright
Author's personal copy
Journal of Archaeological Science 37 (2010) 3401e3411
Contents lists available at ScienceDirect
Journal of Archaeological Science
journal homepage: http://www.elsevier.com/locate/jas
Pre-pottery farmers on the Pacific coast of southern Mexico
Douglas J. Kennett a, *, Dolores R. Piperno b, c, John G. Jones d, Hector Neff e, Barbara Voorhies f,
Megan K. Walsh g, Brendan J. Culleton a
a
Department of Anthropology, University of Oregon, Eugene, OR 97403, United States
Archaeobiology Program, Department of Anthropology, National Museum of Natural History, Washington, DC 20560, United States
c
Smithsonian Tropical Research Institute, Balboa, Panama
d
Department of Anthropology, Washington State University, Pullman, WA 99164, United States
e
Department of Anthropology and Institute for Integrated Research in Materials, Environments, and Societies, California State University, Long Beach, CA 90840, United States
f
Department of Anthropology, University of California, Santa Barbara, CA 93106, United States
g
Department of Geography, University of Oregon, Eugene, OR 97403, United States
b
a r t i c l e i n f o
a b s t r a c t
Article history:
Received 17 March 2010
Received in revised form
19 July 2010
Accepted 20 July 2010
Large island-like shell mounds along the southern coast of Mexico are the earliest known archaeological
sites on the Pacific margin of Mesoamerica. These aceramic deposits date to between 7500 and
3800 cal BP and have been interpreted as locations where foragers, living elsewhere seasonally on the
coastal plain, harvested shellfish and other estuarine resources. Based on an accumulation of paleoecological data from elsewhere in the lowland Neotropics of Mesoamerica, southern Central America and
South America we pose and provide a first test of an alternative subsistence model: that the Archaic
Period populations in this area were slash and burn farmers. Burned maize phytoliths first appear in
these sedimentary records at 6500 cal BP in association with macroscopic charcoal and forest disturbance plant taxa. Periodic burning and forest disturbance, consistent with farming activities, are also
evident in the macroscopic charcoal record between 6500 and 4700 cal BP. Pollen, phytolith and charcoal
records all point to sustained burning, forest disturbance and the cultivation of maize between 4700 and
3800 cal BP. These data suggest that people were slash and burn farming during the Archaic Period prior
to the adoption of pottery and the proliferation of Early Formative Period villages and full-fledged
agriculture based on near or total reliance on crop plants after w3800 cal BP.
Ó 2010 Elsevier Ltd. All rights reserved.
Keywords:
Mesoamerica
Maize
Archaic period
Shell mounds
Paleoecology
1. Introduction
The Archaic Period in Mesoamerica is traditionally viewed as
a long transitional interval (w9000e3800 cal BP) between a poorly
defined Paleoindian big-game hunting tradition and the rise and
proliferation of agricultural villages across Mesoamerica starting
after 3800 cal BP. Archaeological sites dating to this w5000 year
pre-ceramic period are surprisingly rare. To a certain degree this
reflects the demographic and societal realities of the Archaic, but,
particularly in the lowlands, it is also a product of the burial
(natural and cultural) of older sites or their inundation due to postglacial sea-level rise in coastal settings. Mesoamerican archaeologists have also placed greater emphasis on later time periods.
Domestication, dispersal and experimentation of some key
Mesoamerican plant cultigens (maize and squash) are known to
have occurred during the Archaic (e. g., Smith, 1997a,b; Piperno and
* Corresponding author. Tel.: þ1 541 346 5106; fax: þ1 541 346 0668.
E-mail address: [email protected] (D.J. Kennett).
0305-4403/$ e see front matter Ó 2010 Elsevier Ltd. All rights reserved.
doi:10.1016/j.jas.2010.07.035
Flannery, 2001; Piperno et al., 2009; Pope et al., 2001; Pohl et al.,
2007; Ranere et al., 2009), but a substantial dietary role for these
and other cultigens has been generally thought to date to the later
Formative Period (after 3800 years ago) when settled village life
started becoming well-established (Flannery, 1982; Smith, 2001;
Blake et al., 1992a,b; Chisholm and Blake, 2006). Smalley and
Blake (2003) have suggested that the low-level use of maize
during the Archaic Period, and the reason for its early dispersal
from the Balsas region, was its sugary stalk to produce alcohol (and
fresh greens), rather than an important dietary staple (also see
Blake, 2006). This hypothesis, however, is not supported by the
available data on early maize use in its likely area of origin in the
Balsas River valley of Mexico (Piperno et al., 2009: 5023).
Voorhies and Kennett have argued previously that the people
living along the Pacific coast of southern Mexico during the Archaic
were foragers that may have used some domesticated plants at
a low level (Voorhies, 1976, 2004; Kennett and Voorhies, 1996;
Kennett et al., 2006). This inference is largely based on their work
at a series of large shellmounds along the coast (Voorhies, 2004:
32). Voorhies (1976) was drawn to Drucker’s (1948) original
Author's personal copy
3402
D.J. Kennett et al. / Journal of Archaeological Science 37 (2010) 3401e3411
description of these island-like pre-ceramic shell mounds in the
flatwater lagoon systems as evidence for early coastal forager lifeways. Six of these large shell mounds have now been identified and
radiocarbon work indicates that they accumulated between w7500
and 3500 years ago (Voorhies, 2004). Several lines of evidence (see
below) indicate that, despite their size, these shell mounds resulted
from periodic and repeated visitation to harvest and process marsh
clams and other estuarine resources (e.g., fish, shrimp), rather than
permanent settlements. More permanent settlements are hypothesized to have been located elsewhere on the coastal plain.
However, with one possible latest Archaic exception (Vuelta Limón;
see Voorhies, 2004: 100), high sedimentation rates on this flat
coastal plain have largely obscured these hypothetical settlements
and prevented their detection (Voorhies, 1989; Voorhies and
Kennett, 1995). Based on these fragmentary and imperfect data,
Voorhies and Kennett have argued that these people were foragers,
living in semi-permanent base camps and gathering wild resources
from a range of environments logistically, perhaps with some lowlevel use of domesticated plants (Voorhies, 2004: 396; Kennett
et al., 2006: 120). In this paper we present paleoecological data
as a first test of an alternative model; that the people living along
this stretch of Pacific coast were slash and burn farmers occupying
more interior habitats on the coastal plain and periodically visiting
the estuarine zone to harvest and process marsh clams and other
resources.
The basis for this alternative model is a series of paleoecological
studies carried out in the lowland Neotropics during the last twenty
years (Neff et al., 2006; Pearsall et al., 2004; Pohl et al., 1996;
Piperno and Pearsall, 1998; Pope et al., 2001; Piperno, 2006a;
Piperno et al., 2000, 2007, 2009; Ranere et al., 2009). These
studies demonstrate that the cultivation of maize (and other
domesticates) in slash and burn farming systems was in place well
before 3800 years ago in some parts of lowland tropical Mesoamerica, southern Central America and South America. Maize is
often the most detectable domesticate used within these slash and
burn farming systems, but other cultigens are also in evidence (see
Piperno and Pearsall, 1998; Piperno et al., 2000, 2007, 2009). The
earliest evidence for slash and burn maize farming in close proximity to the Pacific Coast of southern Mexico comes from the Gulf
Coast lowlands of Mexico (7300 cal BP; Pope et al., 2001; Pohl et al.,
2007); the central Balsas region of Mexico (7200 cal BP; Piperno
et al., 2007), and the Pacific coastal plain of Guatemala by
5500 cal BP (Neff et al., 2006). In this paper we present data from
sediment cores (pollen, phytoliths and charcoal) collected in the
vicinity of these Archaic Period shellmounds to test the hypothesis
that the prehistoric people that deposited them were farmers.
Before we present these paleoecological data we provide some
background on the archaeological record for the region.
2. Study area
The Pacific coast of Chiapas, Mexico and adjacent Guatemala
coast contain one of the longest known archaeological sequences in
the lowland Neotropics of Mesoamerica (Fig. 1). At European
contact (wAD 1520) the region, known as the Soconusco, was part
of the Aztec Empire and was well known for its agricultural
productivity and tropical forest products (Gasco and Voorhies,
1989). The deep and productive alluvial soils of this region,
coupled with resource rich wetland environments fostered early
settled village life and ceramic manufacture between 3800 and
3600 cal BP (Clark and Cheetham, 2002). A notable increase in
coastal agricultural communities occurs after this time and many of
these were positioned on the interior margins of the wetland
systems along the Pacific coast of Mesoamerica, from Guerrero
(Mexico) south to El Salvador (Neff et al., 2006; Kennett et al.,
2008). These early villages ultimately provided the social lattice
for the first appearance in the Maya region for institutionalized
social hierarchies, signaled by the development of larger and more
socially and politically important communities governed by higher
status individuals or chiefs (Clark and Blake, 1994). Some of these
communities grew in size and population density and started
monumental building campaigns, and it was in these early cities
that some of the first hieroglyphic writing was carved on freestanding stone monuments (Love, 1999, 2002).
Archaeological sites dating prior to 3800 cal BP are rare. The
earliest evidence for human occupation of the coastal plain comes
from six shell mound sites positioned on the landward edge of an
extensive mangrove-estuary system that formed along much of
coastal Chiapas, Mexico as sea-level started to stabilize around
7500 cal BP (Voorhies et al., 2002). These shellmounds form artificial islands in these wetlands and range in size from 0.2 to 1.17
hectares and are between 3 and 11 m above the watertable at their
centers (Voorhies, 2004). Shell deposits at all locations tested
extend well below the modern watertable. The earliest of these
sites, Cerro de las Conchas, has shell deposits dating to between
7500 and 5500 cal BP (Clark, 1986; Voorhies et al., 2002). It is situated on the inland margin of a large freshwater swamp, but the
dominant mollusk species (Polymesoda radiata) indicates that
a brackish-water estuary extended to this inland location at the
time of initial site formation. The five additional shell mounds are
located in the Acapetahua Estuary to the northwest of El Hueyate
and date to between w5500 and 3800 cal BP. Deposits of this age at
all sites are dominated by the marsh clam P. radiata.
These shell mounds are generally characterized by alternating
beds of well-preserved whole and burned fragmented marsh clam
shell that suggest periodic rather than continual use of these locations. This interpretation is bolstered by the general absence of house
floors, formal hearths and a low diversity of tools and fauna
(Voorhies, 2004). Oxygen isotope seasonality studies indicate that for
much of the Archaic Period shellfish were harvested throughout the
year with a greater emphasis on dry season exploitation (Kennett and
Voorhies, 1996; Voorhies et al., 2002). Voorhies (2004); also see
Kennett et al., 2006; hypothesized that residential base camps were
positioned elsewhere on the coastal plain in forest clearings.
Site visibility is a major problem confronting archaeologists
working in this region because rivers flowing out of the Sierra
Madre seasonally deposit large amounts of sediment across this
relatively flat coastal plain. In extreme cases, major tropical storms
produce large amounts of rain across the area, resulting in the
abrupt burial of modern towns. Locating living surfaces and
archaeological sites dating to the Archaic Period has been challenging because of these high sedimentation rates (Voorhies, 1989;
Voorhies and Kennett, 1995). The paleoecological data reported in
this paper therefore provide an important source of information
regarding prehistoric land use, subsistence and settlement during
the Archaic Period.
3. Paleoecological methods
Here we report incremental pollen, macroscopic charcoal, and
phytolith data from sediment cores taken in close proximity to the
five late Archaic Period shellmounds in the Acapetahua region. We
used a vibracorer (3.500 tubes) to obtain intact columns of sediment
that we split and sampled incrementally for radiocarbon, pollen,
phytolith and charcoal analysis. One core was driven in the pericoastal wetlands within the vicinity of Pijijiapan (SOC05-2; UTM
484865E, 1714546N) and the other was taken in very close proximity to the late Archaic Period shell mounds on the landward edge
of the Acapetahua Estuary (SOC05-3; UTM 520258, 1681451N) (see
Fig. 1 for core locations). Observations in the field indicated that
Author's personal copy
D.J. Kennett et al. / Journal of Archaeological Science 37 (2010) 3401e3411
3403
Fig. 1. Locations of sediment cores and Archaic Period sites mentioned in text. Inset shows study area on the southern Pacific coast of Chiapas, Mexico.
these low-energy wetland environments contained well-preserved
paleoecological records that were likely sensitive to anthropogenic
alterations of the landscape during the Archaic Period.
3.1. Chronology and age models
Age control for these cores was established with a series of 8
AMS 14C dates on organic-rich sediment and reported in calibrated
calendar years. These samples were analyzed at the NSF AMS
Facility at the University of Arizona or the National Ocean Sciences
Accelerator Mass Spectrometry Facility (NOSAMS; Table 1). At both
labs organic-rich sediment (peat) underwent a series of heated
acidebaseeacid leaches to remove inorganic carbon and mobile
organic acid phases. Organic samples were combusted under
vacuum to produce CO2. All 14C dates were calibrated with the
Intcal04 atmospheric curve (Reimer et al., 2004) using OxCal 3.10
(Bronk Ramsey, 1995, 2001, 2005). Dates for each core were
modeled as a sequence of non-overlapping events, which slightly
modified each probability distribution. The weighted average of
each decadal probability distribution was used as the central-point
estimate to develop the ageedepth model (Telford et al., 2004).
The age models for cores SOC05-2 and SOC05-3 were developed
using four AMS 14C age determinations each and a date of
55 cal BP for the top sample in each core. Given the small error
associated with the AMS dates, the resulting age models were fit
with a constrained cubic smoothing spline, suggesting a basal date
of 6600 cal BP for core SOC05-2 and 4750 cal BP for core SOC05-3.
The ageedepth curve for SOC05-2 indicates that sedimentation rate
was fairly constant throughout most of the core, except after
600 cal BP when the rate increased rapidly and remained high until
present. The ageedepth curve for core SOC05-3 indicates that the
sedimentation rate was also constant throughout most of the core,
except for a period of more rapid deposition between ca. 4100 and
3800 cal BP. Overall these paleoecological sequences cover the last
6600 cal BP and provide a context for evaluating evidence from
early archaeological sites.
Table 1
Radiocarbon ages and calibrated dates for Soconusco cores.
Lab #
Provenance
OS-53396
OS-53397
AA63353
AA63354
AA63355
OS-53403
OS-53398
AA63359
Core
Core
Core
Core
Core
Core
Core
Core
SOC05-2,
SOC05-2,
SOC05-2,
SOC05-2,
SOC05-3,
SOC05-3,
SOC05-3,
SOC05-3,
220e222 cmbs
308e310 cmbs
407e408 cmbs
467e468 cmbs
308 cmbs
404e405 cmbs
505e507 cmbs
565e566 cmbs
Type
14
Error
Black Peat
Black Peat
Wood
Black Peat
Black Peat
Black Peat
Black Peat
Black Peat
590
3710
5279
5789
2481
3520
3680
4176
30
25
44
44
39
25
35
34
C Age
Cal BP (2 Sigma)
Range
Weighted average
660e530
4150e3970
6190e5930
6720e6470
2730e2360
3870e3700
4150e3900
4840e4580
600
4050
6070
6590
2560
3790
4020
4720
Author's personal copy
3404
D.J. Kennett et al. / Journal of Archaeological Science 37 (2010) 3401e3411
3.2. Pollen analysis
3.5. Phytoliths
Sediment samples of 1e5 ml were processed following standard
techniques (Erdtman, 1960; Faegri et al., 1989). European Lycopodium spp. spores were added as an exotic tracer. Pollen samples
were mounted using glycerine and identifications were made using
a Nikon compound stereomicroscope at 400 magnification.
Identifications were confirmed by using published keys and the
extensive pollen reference collection at Washington State University. A minimum 200-grains was counted per sample. The overall
goal of the pollen study was to determine broad vegetational
formations and changes in them through time caused by human
modification of vegetation and/or other significant perturbations.
Therefore, we employed 200-grain counts, which are considered to
be reflective of these kinds of changes.
Sediment samples from SOC05-3 were processed to extract
phytoliths using standard techniques (Piperno, 2006b). Incremental samples from the earliest deposits in SOC05-2 were also
analyzed. The samples displayed variable phytolith content ranging
from very high to very poor. Many samples that demonstrated
negligible to no phytolith content had large amounts of biogenic
silica from diatoms, sponges, and other aquatic organisms. This
indicates that phytolith recovery procedures worked properly and
that sediment deposition occurred in a non-terrestrial environment
in which there was no mechanism such as inflowing streams to
bring in phytoliths from vegetation in the surrounding area.
Where phytolith content was good assemblages usually contained a diversity of arboreal and herbaceous plants characteristic
of mature and successional tropical forest growth. Phytoliths were
identified through comparison to a modern reference collection of
more than 2000 species of Neotropical plants, including domesticated species such as maize and squashes and their closest wild
relatives that is housed in Piperno’s laboratory at the Smithsonian
Institution. Percentages of phytoliths in each sample are based on
counts of at least 100 phytoliths. Frequencies of phytoliths showing
evidence of charring, which serve as an index of plant burning,
were made independently of the phytolith count.
3.3. Charcoal analysis
Samples of 1 cm3 were taken every 5 cm for macroscopic
charcoal analysis as a proxy for fire activity. Charcoal samples were
soaked in sodium hexametaphosphate for >24 h, washed through
nested sieves of 250 and 125 mm mesh size, and the residue was
transferred into gridded petri dishes and counted under a stereomicroscope. Only charcoal particles >125 mm in minimum diameter were tallied because previous studies indicate that large
particles are not transported far from the source and thus are an
indicator of local fire activity (Whitlock and Millspaugh, 1996;
Clark et al., 1998). Charcoal particles were identified and tallied
as either woody or herbaceous (e.g., grass) based on the presence
of stomatal openings and comparison to burned reference material
(Fig. 2).
3.4. Weight-loss on ignition
Weight-loss on ignition (LOI) was used to determine changes in
the amount of organic material in the sediment (Dean, 1974).
Samples analyzed for organics and carbonates were taken
approximately every 5 cm. Approximately 1 g of sample was dried,
then fired first to 550 C for one hour, then weighed, then fired to
1000 C for another hour. The first firing removes all organic
carbon, so the percentage weight loss indicates the percentage of
organic carbon in the sample. Weight loss between the first and
second firing results from CO2 driven off from carbonate at
temperatures above 850 C. CO2 percentages are converted to
calcite (CaCO3) percentages.
4. Results
SOC05-2 provides the longest continuous paleoecological
sequence for the region, spanning the middle and late Holocene
(w6600e200 cal BP). Two major stratigraphic units are visible in
the 4.7 m core, a coarse sandy horizon (0e1.8 m) underlain by an
organic-rich clay/peat deposit extending to the base of the
sequence (1.8e4.7 m). Our ageedepth model suggests that the
upper stratum was deposited rapidly, perhaps a product of recent
landscape modification coupled with a series of catastrophic floods
that have impacted the region in recent decades. The deepest
deposits date to 6600 cal BP and the uppermost portion of the
organic-rich stratum (w1.8 m) dates to 500 cal BP. Accumulation
rates were highest (w0.615 mm/yr) in the lower part of the core
and slow after 4000 cal BP (w0.349 mm/yr; excluding coarse sandy
horizon).
We first present the pollen data because they provide the
general paleoecological framework for interpreting the charcoal
and phytolith records. Pollen, phytoliths and charcoal are carried
and deposited into the sedimentary record in different ways (wind,
Fig. 2. Photographs of modern maize leaf charcoal (A) and an example of the herbaceous charcoal identified throughout the SOC05-2 and SOC05-3 cores (B).
Author's personal copy
D.J. Kennett et al. / Journal of Archaeological Science 37 (2010) 3401e3411
water, etc.). Each record is sensitive to a somewhat different set of
environmental parameters and the variability in these records is
explored in the discussion section within the context of the broader
paleoecological and archaeological data for this region.
SOC05-2 is the longer of the two records and well-preserved
fossil pollen was found from 180 to 470 cmbs (6600e500 cal BP).
Concentration values were generally high throughout the core, and
pollen was usually well preserved. Two distinct pollen zones are
apparent in this core, based largely on the abundance of red
mangrove (Rhizophora) pollen in the basal section and disturbance
vegetation (e.g., Asteraceae, Cheno-Ams) in the upper zone.
Zone 1 occurs from 470 to 310 cmbs (6600e4100 cal BP), and is
marked by high percentages of red mangrove pollen (Fig. 3A). The
abundance of salt-tolerant red mangrove pollen indicates that the
area was a brackish mangrove swamp. Lesser amounts of sedge
3405
(Cyperaceae) and cattail (Typha) at intervals hints at the presence of
less marine or brackish environments, but these may have been
somewhat spatially removed from the coring locality. Both nonmangrove arboreal elements and herbs are significantly reduced in
this zone. The few arboreal types encountered in this basal section
(pine [Pinus], oak [Quercus] and Moraceae [probably breadnut]),
likely represent types transmitted over significant distances, and
they are thought to be non-local in origin.
Zone 2 occurs from 310 to 180 cmbs (4100e400 cal BP), and
shows a dramatically different environment. Here we see a significant drop in red mangrove pollen, indicating that the environment
had become much less brackish or marine-influenced. At this time,
there is a corresponding increase in sedge, cattail, and water lily
pollen (Nymphea), as well as a spike in Salvinia spores (a floating
freshwater aquatic weed), confirming that freshwater now
Fig. 3. (A) Pollen diagram showing percentages through time of major taxa identified in SOC05-2. Dots indicate presence of taxa. (B) Pollen diagram showing percentages through
time of major taxa identified in SOC05-3. Dots indicate presence of taxa.
Author's personal copy
3406
D.J. Kennett et al. / Journal of Archaeological Science 37 (2010) 3401e3411
dominated the region where the core was collected. Forests in the
area change from a mangrove forest to arboreal associations characterized by palms, Moraceae, Sapindaceae and other genera.
Disturbance vegetation is most noticeable in Zone 2, including
pollen from weeds (Asteraceae and Cheno-Ams) and grasses
(Poaceae). These types proliferate in disturbed environments
including human settlements and agricultural fields. Slight
increases in oak and especially pine pollen may also reflect a more
open environment, where long-distance pollen can more easily
blow into the sediments of the area. Cultigen pollen is poorly represented in this core, and only occurs after 1000 years ago. These
include maize (Zea mays) and possibly manioc (Croton/Manioc),
although positive identification of the latter is difficult and this
grain may simply represent a wild member of the family. The
presence of cultigen pollen indicates that farming was taking place
in the area. Zea pollen is usually present in low numbers even if
fields are close because they are so heavy and poorly dispersed, and
other cultivars are poor pollen producers. One would expect
percentages of maize and other cultigens to increase with closer
proximity to ancient fields. Pollen evidence therefore points to
forest clearing consistent with agriculture during the late Archaic
(4100e3800 cal BP).
Additional evidence for forest disturbance during the late
Archaic Period comes from SOC05-3 (Fig. 3B). This 5.7 m sediment
core was drilled 35 km southwest of SOC05-2 in the seasonally
inundated wetland surrounding the Acapetahua Estuary. The
accumulation of sediment at this location was rapid (w1.608 mm/
yr) and provides a detailed paleoecological record from 4700 to
2300 cal BP that complements the longer sequence from SOC05-2.
Well-preserved fossil pollen was present in most of the SOC05-3
sequence, although sediments above 280 cm (2300 cal BP) were
oxidized. In addition, sediments from 328 to 380 cm
(2800e3600 cal BP) and from 400 to 416 cm (w3800 cal BP) were
largely devoid of fossil grains. Past drying episodes during or
shortly after the deposition of sediments from these zones are
likely to be responsible for the oxidation of fossil pollen at these
times. Concentration values of fossil pollen grains from other
portions of the core, however, were generally high, attesting to the
excellent state of preservation of these grains. Four distinctive
zones are apparent in this pollen record, reflecting past conditions
in the coring area.
Zone 1 (564e525 cmbs [4700e4200 cal BP]) is represented by
high percentages of sedge (Cyperaceae) and cattail (Typha) pollen,
indicating a freshwater or slightly brackish depositional environment. Red mangrove (Rhizophora) pollen is low through this zone,
suggesting these plants were present in the area, but somewhat
removed from the location where the core was collected. Disturbance indicators through this zone are high, as indicated by ChenoAms and Asteraceae grains, consistent with human disturbance in
the local area, although cultigen pollen is lacking in the samples.
Zone 2 (525e475 cmbs [4200e3900 cal BP]) is marked by
a dramatic, but temporary increase in red mangrove pollen. The
environment continues to be dominated by sedge and cattail
pollen, suggesting that any changes were largely temporary.
Cheno-Am and to a lesser extent Asteraceae pollen are noticeably
reduced, and grass (Poaceae) pollen increased in this zone. The
initial appearance of Zea mays pollen along with increases in grass
pollen occurs at the end of this interval at 3900 cal BP along with
a single cotton (Gossypium) pollen grain. The morphological characteristics of this pollen type do not tell us if it was a domesticated
or a wild form.
Charcoal concentrations in both SOC05-2 and SOC05-3
complement the pollen record and provide additional information
about burning and forest clearing. The charcoal concentrations (#
particles/cm3) and the ratio of herbaceous/total charcoal (%) are
shown in Fig. 4. Fire activity occurs throughout the SOC05-2
Fig. 4. Charcoal, LOI, and carbonate data for cores SOC05-2 and SOC05-3.
Author's personal copy
D.J. Kennett et al. / Journal of Archaeological Science 37 (2010) 3401e3411
sequence, but is highest from 1200 to 500 cal BP. The charcoal
assemblage through this interval contains high percentages of
herbaceous charcoal. Herbaceous charcoal is similar in character to
modern maize burned for comparative purposes, but other tropical
grasses have similar structures and therefore may have been
burned as well. Other periods of very high fire activity occur
between w4300 and 4200 cal BP, 5200 cal BP, and between 5800
and 5600 cal BP. These periods have relatively high herbaceous
charcoal values, consistent with the burning of agricultural fields
(Walsh et al., 2010). Fire activity in SOC05-3 occurs throughout the
sequence, but was highest for a brief period at w2600 cal BP, and
then consistently high between w4750 and 3800 cal BP, with
relatively high percentages of herbaceous charcoal consistent with
the clearing of agricultural plots.
The earliest evidence for forest disturbance and maize farming
comes from the phytolith record near the base of SOC05-2 and
dates to 6500 cal BP (Fig. 5A). Three samples analyzed from the
bottom of the core all had well-preserved phytoliths. Numerous
diatoms and sponge spicules were also present. In the lowermost
sample at 467e469 cmbs there were high concentrations of grasses.
Many are from bamboos (e.g., Chusquea, tall saddles), and only
a few are from weedy grass varieties (e.g., bilobates representing
the Panicoideae and Chloridoideae). No burned or unburned maize
phytoliths were observed.
Between 464 and 460 cmbs (6500 cal BP) significant changes
take place in the phytolith record. Weedy grasses (e.g., Chloridoideae and Panicoideae) along with Heliconia, an archetypal plant
of early successional forests, are dominant in the record, and these
phytoliths are commonly burned. Maize leaf and cob phytoliths,
including burned examples, were present in the samples, further
suggesting deliberate land clearance associated with slash and burn
systems. High percentages of Podostemaceae phytoliths, a plant
that grows only on rocks in freshwater streams, indicate that the
phytolith assemblage originated from interior dryland agricultural
fields and arrived in the mangrove swamp via water transport.
Phytoliths diagnostic of the ears of teosinte are absent in these and
all other sediments examined, further reinforcing the identification
of maize phytoliths and also indicating that Zea pollen in these
sequences is probably from maize introduced into the area rather
than from wild stands of teosinte that may have once existed in the
region and are no longer present. The records indicate that by
w6500 cal BP, significant landscape modification by maize farmers
involving forest clearing and burning occurred in the area.
The phytolith content of SOC05-3 was highly variable (Fig. 5B).
Some samples had no phytoliths while others possessed them in
high number. Many samples with few phytoliths contained a mass
of biogenic silica from diatoms, sponges, and other aquatic organisms. Apparently in these contexts there was no mechanism such as
inflowing water to transport terrestrial phytoliths to the coring
locale. The samples with a considerable phytolith content show
clear evidence for landscape modification for farming, with low
frequencies of arboreal phytoliths and high frequencies of weedy
grasses and other disturbance taxa (e.g., Heliconia). The phytoliths
commonly are burned and maize phytoliths occur in the late
Archaic Period deposits by 4100 cal BP.
5. Discussion
Evidence for forest disturbance, burning, and the cultivation of
maize extends back to 6500 cal BP (Fig. 6). Weedy grasses, evidence
for burning (charcoal and burned phytoliths) and the presence of
maize leaf and cob phytoliths in these deposits are all indicative of
purposeful land clearing associated with a slash and burn system of
cultivation. Teosinte does not grow naturally in this area today and
the absence of ear phytoliths of this wild grass throughout the 6600
3407
year sequence indicates that the maize present in the record is
a cultigen introduced from outside the area. The dominance of red
mangrove in the pollen record in the early deposits coupled with
only weak evidence for forest disturbance suggests that these
agricultural fields were located some indeterminate distance away
from the coring location. The abundance of Podostemaceae phytoliths in this early assemblage suggests that the burned maize
phytoliths were carried by water to the coring location from agricultural fields that are now buried somewhere upstream on the
coastal plain. Maize pollen was not identified in these earliest levels
even with extended counting of multiple pollen slides, a result that
conforms with the phytolith data because pollen will not survive as
well as phytoliths once eroded from land surfaces and deposited
into other sedimentary contexts. We will design future work to
target sediments of this age and earlier that are closer to potential
agricultural fields.
The high fire activity evident in the charcoal record between
5800 and 5600 cal BP and again at 5200 cal BP is likely of anthropogenic origin and is consistent with field burning associated with
slash and burn agriculture. Fires are triggered by other mechanisms
(e.g., lightning) in the tropics (Cardoso et al., 2008; Goldammer,
1990), but the charcoal during these two periods is predominantly herbaceous, rather than arboreal, and its structure is
remarkably similar to modern burned maize leaves and not similar
to many other known grasses. However, we cannot rule out the
possibility that other lowland Neotropical grasses were the source
of the charcoal without a more comprehensive comparative
collection. Cultigens were not identified during the fire episodes in
the pollen record and the phytolith record does not span this
interval, but similar patterns of burning associated with cultigens
were found to occur to the south along the Pacific coastal plain of
Guatemala by 5500 cal BP (Neff et al., 2006). The confluence of data
suggests that these fires are related to slash and burn cultivation.
Evidence for maize, forest disturbance, and burning at
6500 cal BP along the Pacific Coast of southern Mexico is consistent
with evidence from Central and South America for early dispersal of
this domesticate from the Balsas region through the lowland
Neotropics (Pearsall et al., 2004; Dickau et al., 2007; Zarillo et al.,
2008). Genetic and biogeographic data indicate that maize was
domesticated from an annual species of teosinte [Zea mays subsp.
parviglumis Iltis and Doebley] presently native to the Central Balsas
River valley of Mexico (Doebley, 2004; Wang et al., 2005). Microsatellite genetic work suggests that this occurred at w9100 cal BP
(Matsuoka et al., 2002), and recent archaeobotanical and paleoecological research in that region indicates that its cultivation had
begun by 8700 cal BP and that slash and burn cultivation systems
incorporating maize were present by 7200 cal BP (Piperno et al.,
2007, 2009). Early dispersal through the seasonally dry lowland
Neotropics south of Mexico is indicated by its appearance in
Panama and northern portions of South America by w7800 and
6000 cal BP, respectively (Piperno and Pearsall, 1998; Piperno et al.,
2000; Dickau et al., 2007; Zarillo et al., 2008).
Although maize is the only cultigen evident in our Soconusco
records, it is likely that a mixture of cropsdseed, root, and
treedwas being grown in the region by 6500 cal BP. Large labor
investments are required to clear forests and burn them and it is
unlikely that people would invest this time and effort for a single
cultigen. It is also clear that other indigenous Mexican cultigens,
such as squashes, were domesticated by 8700 cal BP (Piperno et al.,
2009; Smith, 1997a). Many crop plants are difficult to detect in
paleoecological records because of poor production and/or taxonomic specificity of their pollen and phytoliths, leading to their
under-representation in these kinds of records.
The period from 4700 to 3800 cal BP is when five of the six
known shell mounds in the Soconusco region accumulated and this
Author's personal copy
3408
D.J. Kennett et al. / Journal of Archaeological Science 37 (2010) 3401e3411
Fig. 5. (A) Phytolith diagram showing percentages through time of major taxa identified in SOC05-2. Dots indicate presence of taxa. Percentages of burned phytoliths were
calculated independently from other phytoliths in the sum. (B) Phytolith diagram showing percentages through time of major taxa identified in SOC05-3. Dots indicate presence of
taxa. Percentages of burned phytoliths were calculated independently from other phytoliths in the sum. Phytolith preservation was generally poor in this core.
Author's personal copy
D.J. Kennett et al. / Journal of Archaeological Science 37 (2010) 3401e3411
3409
Fig. 6. Pollen, phytolith, and macroscopic charcoal during the middle and late Holocene. Dots or squares indicate taxa is present. Inset shows selection of phytoliths from SOC05-02
(464e465 cmbs) indicating disturbance and slash and burn farming at from 6500 cal BP A) Podostemaceae; B) Maize cob phytolith; C) Heliconia (burned); D) Heliconia (unburned).
corresponds in age to the high sustained fire activity evident in our
cores. The pollen record indicates forest disturbance and maize
phytoliths by w4100 cal BP. This is also the same interval that maize
phytoliths are first found in the Archaic Period shell mounds (Jones
and Voorhies, 2004). Therefore, data from various sources all point
to the presence of slash and burn farmers living in close proximity
to this seasonally inundated estuarine zone. Clearing forests for
cultivating plants implies that crops supplied a significant number
of calories to diets and that people had intensified their farming
systems beyond simpler forms of “dooryard” horticulture practiced
next to residences.
It thus appears that high sedimentation rates on this flat coastal
plain have obscured early settlements of people who were practicing slash and burn cultivation. This evidence coincides with
intensive exploitation of resources from the coastal estuarine zone.
Large shell mounds started accumulating between 7500 and
5500 cal BP along the interior edges of the coastal brackish-water
lagoons. Small fish vertebrae recovered during excavations of these
shell mounds suggest the use of nets and other sophisticated
fishing technology (e.g., weirs; Voorhies, 2004) by these early
farmers. Future work will be necessary to determine if the accumulation of these shell mounds at 7500 cal yrs BP directly relates to
the adoption of slash and burn agriculture in this area.
Complementary use of slash and burn cultivation and intensive
shellfish
harvesting
persisted
in
this
region
until
w3800e3600 cal BP when shell mound accumulation ceased.
Stable oxygen isotope shellfish harvesting profiles suggest relaxed
predation on shellfish populations after w4100 cal BP with the first
appearance of maize phytoliths in these shell mounds (Jones and
Voorhies, 2004). The cessation of intensive mollusk harvesting
after w3800 cal BP occurs with widespread evidence for increases
in forest disturbance, burning and a greater reliance upon maize
and likely other crops (Neff et al., 2006). Human settlements
expanded rapidly along the coast after this time and evidence for
the emergence of institutionalized social inequality appears in the
archaeological record (Clark and Cheetham, 2002). The abandonment of intensive shellfish harvesting strategies may be more
related to these larger societal changes than the adoption of maize
at the end of the Archaic as once was hypothesized (Kennett and
Voorhies, 1996).
Additional paleoecological records are needed to extend
geographically the evidence presented here and determine the
overall extent of slash and burn farming systems on this coastal
plain during the earlier stages of the Archaic Period. When we
started this project it was unclear how deep these peri-coastal
sedimentary sequences would extend below the modern ground
surface. It turns out that this is highly variable between coring
locations and in some instances they extended deeper than the
viable range of our coring technology. Additional work with more
sophisticated coring technology may extend these records a little
earlier than 6600 cal BP, but sedimentary sequences in coastal
settings are not likely to extend back beyond 7700 cal BP when sealevel started to stabilize.
6. Conclusions
The paleoecological evidence presented in this paper supports
the view that people living along the Pacific coast of southern
Mexico employed systems of slash and burn cultivation during the
Archaic Period. Episodic burning and forest disturbance together
with the remains of maize are all evident in these records between
6500 and 3800 cal BP (and not surprisingly later into the Formative,
Author's personal copy
3410
D.J. Kennett et al. / Journal of Archaeological Science 37 (2010) 3401e3411
Classic and Post-Classic Periods). Future work targeting sediments
dating to before 6700 cal BP will be required to determine when
slash and burn practices were first established in this region, but
the available data conform with evidence from Panama and South
America for maize dispersal from the Balsas region of Mexico
through the lowland Neotropics during the early stages of the
middle Holocene.
These new data alter our view of Archaic Period adaptive
strategies along the Pacific coast of Southern Mexico. We now
hypothesize that settlements of farmers (i.e., people who were
deriving a significant portion of their dietary calories from crop
plants) are buried beneath the alluvium on the coastal plain. These
settlements were probably more seasonally stable than we once
thought, given the labor investments associated with growing and
storing maize. These observations will improve our predictive
settlement models as the hunt for Archaic Period archaeological
sites continues on the coastal plain. The most fertile lands on the
coastal plain are not in the peri-coastal wetlands where soils are
saline and seasonally inundated. Well-drained alluvial soils are
most productive closer to the Sierra Madre where rainfall is more
seasonally predictable. The strength and character of the paleoecological signatures reported here suggest that settlements were
located some distance away from the wetlands earlier in the
Archaic. By the Late Archaic Period the evidence for burning, forest
disturbance and maize cultivation suggests that agricultural
communities had expanded across the coastal plain and were
positioned closer to the wetlands. This is consistent with intensified
exploitation of shellfish during the Late Archaic indicated by the
increased number of shellfish processing localities. Overall, the
combination of slash and burn farming and the exploitation of
estuarine resources was a powerful combination that persisted in
this area for much of the Archaic Period.
Acknowledgments
This research was funded by the National Science Foundation
(BCS-0211215) and the Smithsonian National Museum of Natural
History. We thank the two anonymous reviewers for their useful
comments. A special thanks goes to John Clark and the New World
Archaeological Foundation for their logistical support during our
field campaign. The paper was written with fellowship support
from the AHRC Centre for the Evolution of Cultural Diversity,
Institute of Archaeology, University College London (DJK).
References
Blake, M., Chisholm, B.S., Clark, J.E., Mudar, K., 1992a. Non-agricultural staples and
agricultural supplements: early formative subsistence in the Soconusco,
Mexico. In: Price, T.D., Gebauer (Eds.), Transitions to Agriculture. Prehistory
Press, Madison, WI, pp. 133e151.
Blake, M., Chisholm, B.S., Clark, J.E., Voorhies, B., Love, M.W., 1992b. Prehistoric
subsistence in the Soconusco region. Curr. Anthropol. 33, 83e94.
Blake, M., 2006. Dating the initial spread of Zea mays. In: Staller, J., Tykot, R., Benz, B.
(Eds.), Histories of Maize: Multidisciplinary Approaches to Prehistory, Linguistics, Biogeography, Domestication, and Evolution of Maize. Academic Press,
New York, pp. 55e68.
Bronk Ramsey, C., 1995. Radiocarbon calibration and analysis of stratigraphy: the
OxCal program. Radiocarbon 37, 461e474.
Bronk Ramsey, C., 2001. Development of the radiocarbon program OxCal. Radiocarbon 43, 355e363.
Bronk Ramsey, C., 2005. OxCal v. 3. http://c14.arch.ox.ac.uk/embed.php?File¼oxcal.
html 10.
Cardoso, M.F., Nobre, C.A., Lapola, D.M., Oyama, M.D., Sampaio, G., 2008. Long-term
potential for fires in estimates of the occurrence of savannas in the tropics.
Global Ecol. Biogeogr. 17, 222e235.
Chisholm, B., Blake, M., 2006. Diet in the prehistoric Soconusco. In: Staller, J.,
Tykot, R., Benz, B. (Eds.), Histories of Maize: Multidisciplinary Approaches to
Prehistory, Linguistics, Biogeography, Domestication, and Evolution of Maize.
Academic Press, New York, pp. 55e68.
Clark, J.E., 1986. Excavaciónes en el Cerro de las Conchas, Municipio de Huixtla,
México. Informe Preliminar Entregado al Consejo de Arqueologia. Instituto
Nacional de Antropología e Historia, Mexico, D.F.
Clark, J.E., Blake, M., 1994. The power of prestige: competitive generosity and the
emergence of rank societies in lowland Mesoamerica. In: Brumfiel, E.M.,
Fox, J.W. (Eds.), Factional Competition and Political Development in the New
World. Cambridge University Press, Cambridge, pp. 17e30.
Clark, J.E., Cheetham, D., 2002. Mesoamerica’s tribal foundation. In: Parkinson, W.A.
(Ed.), The Archaeology of Tribal Societies. International Monographs in
Prehistory, Ann Arbor, MI, pp. 278e339.
Clark, J.S., Lynch, J., Stocks, B.J., Goldammer, J.G., 1998. Relationships between
charcoal particles in air and sediments in west-central Siberia. The Holocene 8,
19e29.
Dean Jr., W.E., 1974. Determination of carbonate and organic matter in calcareous
sediments by loss on ignition comparison with other methods. J. Sed. Res. 44,
242e248.
Dickau, R., Ranere, A.J., Cooke, R.G., 2007. Starch grain evidence for the preceramic
dispersals of maize and root crops in tropical dry and humid forests of Panama.
Proc. Natl. Acad. Sci. USA 104, 3651e3656.
Doebley, J.F., 2004. The genetics of maize evolution. Ann. Rev. Gen. 38, 37e59.
Drucker, P., 1948. Preliminary notes on an archaeological survey of the Chiapas
coast. Middle Am. Res. Records 1, 151e169.
Erdtman, G., 1960. The acetolysis method: a revised description. Sven. Botan. Tid.
54, 561e564.
Faegri, K., Kaland, P.E., Krzywinski, K., 1989. Textbook of Pollen Analysis. John Wiley
and Sons, New York.
Flannery, K.V., 1982. The Early Mesoamerican Village. Academic Press, New York.
Gasco, J., Voorhies, B., 1989. The ultimate tribute: the role of the Soconusco as an
Aztec tributary. In: Voorhies, B. (Ed.), Ancient Trade and Tribute: Economies of
the Soconosco Region of Mesoamerica. University of Utah Press, Salt Lake City,
pp. 48e94.
Goldammer, J.G. (Ed.), 1990. Fire in the Tropical Biota: Ecosystem Processes and
Global Challenges. Springer-Verlag, BerlineHeidelbergeNew York.
Jones, J.G., Voorhies, B., 2004. Humaneplant interactions. In: Coastal Collectors in
the Holocene: The Chantuto People of Southwest Mexico. University Press of
Florida, Gainesville, FL.
Kennett, D.J., Voorhies, B., 1996. Oxygen isotopic analysis of archaeological shells to
detect seasonal use of wetlands on the southern Pacific coast of Mexico. J. Arch.
Sci. 23, 689e704.
Kennett, D.J., Voorhies, B., Martorana, D., 2006. An ecological model for the origins
of maize-based food production on the Pacific Coast of southern Mexico. In:
Kennett, D.J., Winterhalder, B. (Eds.), Behavioral Ecology and the Transition to
Agriculture. University of California Press, Berkeley, pp. 103e136.
Kennett, D.J., Voorhies, B., Wake, T., Martinez, N., 2008. Long-term effects of human
predation on marine ecosystems in Guerrero, Mexico. In: Rick, T.,
Erlandson, J.M. (Eds.), Human Impacts on Marine Environments. University of
California Press, Berkeley, pp. 103e124.
Love, M.W., 1999. Ideology, material culture, and daily in preclassic Mesoamerica:
a Pacific Coast perspective. In: Grove, David C., Joyce, Rosemary A. (Eds.), Social
Patterns in Pre-Classic Mesoamerica. Dumbarton Oaks, Washington D.C., pp.
127e153.
Love, M.W., 2002. Early complex society in Pacific Guatemala: settlements and
chronology of the Rio Naranjo, Guatemala. Papers of the New World Archaeological Foundation, No. 66. Provo.
Matsuoka, Y., Vigouroux, Y., Goodman, M.M., Sanchez, G.J., Buckler, E., Doebley, J.,
2002. A single domestication for maize shown by multilocus microsatellite
genotyping. Proc. Natl. Acad. Sci. USA 99, 6080e6084.
Neff, H., Pearsall, D.M., Jones, J.G., Arroyo, B., Collins, S.K., Friedel, D.E., 2006. Early
Maya adaptive patterns: mid-late Holocene paleoenvironmental evidence from
Pacific Guatemala. Lat. Am. Antiq. 17, 287e315.
Pearsall, D.M., Chandler-Ezell, K., Zeidler, J.A., 2004. Maize in ancient Ecuador:
results of residue analysis of stone tools from the Real Alto site. J. Arch. Sci. 31,
423e442.
Piperno, D.R., 2006a. Latin American biogeography: causes and effects. In:
Proceedings of the 51st Annual Systematics Symposium. Missouri Botanical
Garden Press, St. Louis, pp. 274e293.
Piperno, D.R., 2006b. Phytoliths: A Comprehensive Guide for Archaeologists and
Paleoecologists. AltaMira Press, New York.
Piperno, D.R., Flannery, K.V., 2001. The earliest archaeological maize (Zea mays L.)
from Highland Mexico: new accelerator mass spectrometry dates and their
implications. Proc. Natl. Acad. Sci. USA 98, 2101e2103.
Piperno, D.R., Pearsall, D.M., 1998. The Origins of Agriculture in the Lowland
Neotropics. Academic Press, San Diego.
Piperno, D.R., Ranere, A.J., Holst, I., Hansell, P., 2000. Starch grains reveal early root
crop horticulture in the Panamanian tropical forest. Nature 407, 894e897.
Piperno, D.R., Moreno, J.E., Iriarte, J., Holst, I., Lachniet, M., Jones, J.G., Ranere, A.J.,
Castanzo, R., 2007. Late Pleistocene and Holocene environmental history of the
Iquala Valley, Central Balsas watershed of Mexico. Proc. Natl. Acad. Sci. USA 104,
11874e11881.
Piperno, D.R., Ranere, A.J., Holst, I., Iriarte, J., Dichau, D., 2009. Starch grain and
phytolith evidence for early ninth millennium B.P. maize from the Central
Balsas River Valley, Mexico. Proc. Natl. Acad. Sci. USA 106, 5019e5024.
Pohl, M.E.D., Piperno, D.R., Pope, K.O., Jones, J.G., 2007. Microfossil evidence for preColumbian maize dispersals in the neotropics from San Andrés. Proc. Natl. Acad.
Sci. USA 104, 6870e6875.
Author's personal copy
D.J. Kennett et al. / Journal of Archaeological Science 37 (2010) 3401e3411
Pohl, M.D., Pope, K.O., Jones, J.G., Jacob, J.S., Piperno, D.R., deFrance, S.D., Lentz, D.L.,
Gifford, J.A., Danforth, M.E., Josserand, J.K., 1996. Early agriculture in the Maya
lowlands. Lat. Am. Antiq. 7, 355e372.
Pope, K.O., Pohl, M.E.D., Jones, J.G., Lentz, D.L., von Nagy, C., Vega, F.J., Quitmyer, I.R.,
2001. Origin and environmental setting of ancient agriculture in the lowlands of
Mesoamerica. Science 292, 1370e1373.
Ranere, A.J., Piperno, D.R., Holst, I., Dickau, R., Iriarter, J., 2009. The cultural and
chronological context of early Holocene maize and squash domestication in the
Central Balsas River Valley, Mexico. Proc. Natl. Acad. Sci. USA 106, 5014e5018.
Reimer, P.J., Baillie, M.G.L., Bard, E., Bayliss, A., Beck, J.W., Bertrand, C.J.H.,
Blackwell, P.J., Buck, C.E., Burr, G.S., Cutler, K.B., Damon, P.E., Edwards, R.L.,
Fairbanks, R.G., Friedrich, M., Guilderson, T.P., Hogg, A.G., Hughen, K.A.,
Kromer, B., McCormac, G., Manning, S., Bronk Ramsey, C., Reimer, R.W.,
Remmele, S., Southon, J.R., Stuiver, M., Talamo, S., Taylor, F.S., van der Plicht, J.,
Weyhenmeyer, C.E., 2004. IntCal04 terrestrial radiocarbon age calibration,
0e26 cal kyr BP. Radiocarbon 46, 1029e1059.
Smalley, J., Blake, M., 2003. Sweet beginnings: stalk sugar and the domestication of
maize. Curr. Anthropol. 44, 675e703.
Smith, B.D., 1997a. The initial domestication of Cucurbita pepo in the Americas
10,000 years ago. Science 276, 932e934.
Smith, B.D., 1997b. Reconsidering the Ocampo caves and the era of incipient
cultivation in Mesoamerica. Lat. Am. Antiq. 8, 342e383.
Smith, B.D., 2001. Low level food production. J. Archaeol. Res. 9, 1e43.
Telford, R.J., Heegaard, E., Birks, H.J.B., 2004. The intercept is a poor estimate of
calibrated radiocarbon age. The Holocene 14, 296e298.
3411
Voorhies, B., 1976. The Chantuto people: an Archaic period society of the Chiapas
Littoral, Mexico. Papers of the New World Archaeological Foundation, No. 4l.
Provo, UT: Brigham Young University.
Voorhies, B., 1989. Settlement patterns in the western Soconusco: methods of site
recovery and dating results. In: Bové, F.J., Heller, L. (Eds.), New Frontiers in the
Archaeology of the Pacific Coast of Mesoamerica. Arizona Research Papers 39,
pp. 329e369.
Voorhies, B., 2004. Coastal Collectors in the Holocene: The Chantuto People of
Southwest Mexico. University of Florida, Gainsville.
Voorhies, B., Kennett, D.J., 1995. Buried sites on the Soconusco, Chiapas, Mexico. J.
Field Archaeol. 22, 65e79.
Voorhies, B., Kennett, D.J., Jones, J.G., Wake, T.A., 2002. A Middle Archaic archaeological site on the west coast of Mexico. Lat. Am. Antiq. 13, 179e200.
Walsh, M., Pearl, C.A., Whitlock, C., Bartlein, P.J., Worona, M., 2010. An 11,000-yearlong fire and vegetation history from Beaver Lake, central Willamette Valley,
Oregon. Quat. Sci. Rev. 29, 1093e1106.
Wang, H., Nussbaum-Wagler, T., Li, B., Zhao, Q., Vigouroux, Y., Faller, M.,
Bomblies, K., Lukens, L., Doebley, J.F., 2005. The origin of the naked grains of
maize. Nature 436, 714e719.
Whitlock, C., Millspaugh, S.H., 1996. Testing the assumptions of fire-history studies:
an examination of modern charcoal accumulation in Yellowstone National Park,
USA. The Holocene 6, 7e15.
Zarillo, S., Pearsall, D.M., Raymond, J.S., Tisdale, M.A., Quon, D.J., 2008. Directly
dated starch residues document early formative maize (Zea mays L.) in tropical
Ecuador. Proc. Natl. Acad. Sci. USA 105, 5005e5111.