Quaternary International xxx (xxxx) xxx–xxx

Contents lists available at ScienceDirect

Quaternary International
journal homepage: www.elsevier.com/locate/quaint

Environmental reconstruction spanning the transition from hunter/

gatherers to early farmers in Colombia: paleopedological and archaeological
indicators from the pre-ceramic sites Tequendama and Aguazuque
A.V. Triana-Vegaa,∗, S. Sedovb, J. Salinas-Aceroc, D. Carvajal-Contrerasd, C. Moreanoe,
M. Tovar–Reyesf, E. Solleiro–Rebolledob, J. Díaz-Ortegab
a
Posgrado en Antropología Universidad de los Andes, Bogota, Colombia
b
Instituto de Geología, UNAM, Ciudad Universitaria, 04510, Mexico, D. F., Mexico
c
Postgraduate Program in Anthropology, University of Berkeley, California, USA
d
Facultad Estudios de Patrimonio Cultural, Programa de Arqueología, Universidad Externado de Colombia, Bogotá, Colombia
e
Posgrado en Arqueología, Universidad Nacional del Centro de la Provincia de Buenos Aires, Buenos Aires, Argentina
f
Universidad Nacional de Colombia, Bogotá, Colombia

ARTICLE INFO ABSTRACT

Keywords: Complex geoarchaeological research of the archaeological sites with continuous long-term occupation could
Holocene provide insight into environmental change and human-landscape interaction during the transition from hunter-
Paleoenvironment gatherer to horticulture economy in the north-west of South America. We performed a paleopedological study in
Paleosol the new sections at Tequendama (rockshelter site) and Aguazuque (open-air site) two key pre-ceramic sites of the
Pre-ceramic site
Savannah of Bogotá region of Colombia, combined with traditional archaeological and paleontological in-
Micromorphology
Pedogenic carbonates
vestigations. The chronological scales of the sections based on a large number of 14C dates show that their
occupation periods are complementary and together cover the period from the terminal Pleistocene to late
Holocene. The material of the archaeological layers has a colluvial origin with large tephra input. Presence of
volcanic components influence pedogenesis supporting development of granular structure, dark humus accu-
mulation, and Andic properties that could not be used for paleoenvironmental reconstruction.
Micromorphological observations however show abundant evidence of pedogenic carbonate accumulation that
we interpret as evidence of gradual aridization during the middle Holocene. This conclusion is supported by the
paleontological data, an upward increase of the frequency of Cavia sp. remains, and agrees with the regional
palynological proxies. Aridization was the background environmental trend during the transition period from a
hunter-gatherer economy to an early horticulture and plant and animal domestication.

1. Introduction Ecuador) Center. In consequence, it was influenced by the flow of

Colombia occupies the north-western corner of the South American
continent. This territory plays an important role in the cultural history
of the Western Hemisphere. At the time of early human dispersal during
the terminal Pleistocene, Colombia formed a key part of the migration
pathways of the first hunter-gatherers into South America (Aceituno
et al., 2013)In the earlier period of the transition to subsistence system,
based on agriculture and sedentary population pattern, this territory
underwent a complex socioeconomic evolution (Aceituno and Loaiza,
2018). It is located between two major centers of the origin of culti-
vated plants defined by Vavilov in 1926 (Vavilov, 1992): The Central
American (Mexico/Guatemala) Center and the South American (Perú/
exogenous domesticated plants and technologies. On the other hand
within this territory local processes of domestication and horticultural
development could take place (Piperno, 2011).
These socioeconomic processes occurred under changing climatic
and environmental conditions since Pleistocene-Holocene transition
through early and middle Holocene. These changes varied within a
mosaic of ecosystems, produced by complex geological and geomor-
phological structure of the territory. The general outline of the early
cultural history (Aceituno et al., 2013; Aceituno and Loaiza, 2018) as
well as the environmental evolution (Marchant et al., 2002)are traced
on the regional scale. However the local scenarios of the human-land-
scape interactions are still few. Complex geoarchaeological and soil-

∗
Corresponding author.
E-mail address: angelicatriana177@gmail.com (A.V. Triana-Vega).

https://doi.org/10.1016/j.quaint.2018.09.048
Received 2 March 2018; Received in revised form 31 August 2018; Accepted 30 September 2018
1040-6182/ © 2018 Elsevier Ltd and INQUA. All rights reserved.

Please cite this article as: Triana, A., Quaternary International, https://doi.org/10.1016/j.quaint.2018.09.048
A.V. Triana-Vega et al.
Fig. 1. Map of the study region, the stars indicate the studied sites.
archaeological research could provide a set of indicators for such re-
constructions. The sections of sites with long-term semi-continuous
occupation are most valuable for the proposed approach.
The sites of Tequendama and Aguazuque were hotspots for the
study of pre-ceramic societies in the Savannah of Bogotá region. Large-
scale excavations in the rock shelters of Tequendama were carried out
by (Correal and Van der Hammen, 1977)The researchers reported
evidence of human occupation since the terminal Pleistocene until the
late Holocene, including lithics, human remains, paleofauna materials,
as well as plant remains dated back to more than 10 BP (Correal and
Van der Hammen, 1977). During late 1980's Correal excavated the site
of Aguazuque where he documented ancient occupation between 8 and
2.5 BP. This site yielded a great number of lithic artefacts, faunal ma-
terial, several human burials, and several charred plant remains. At this
stage of the research, the conclusion was made that Tequendama was
occupied by migrating hunter-gatherer communities, characterised by
high animal protein consumption, whereas the population of Agua-
zuque was more sedentary and depended heavily upon plant resources.
In general, it is supposed that the transition to agriculture and seden-
tary occupation was accompanied by development of more complex
social organization. However, recently the complexity of hunter and
gatherer societies has been grossly re-evaluated. These groups could
maintain quite sophisticated social system and behaviour, develop or-
ganised sustainable economy based on division of labour and modify
regional ecosystems (Aceituno and Loaiza, 2014; Smith, 2001; Solich
2
Quaternary International xxx (xxxx) xxx–xxx
and Bradtmöller, 2017)The re-opening of excavation at both sites in
2014 provided the opportunity to sample for complex geoarchaeolo-
gical investigations. In particular, a paleopedological study of the in-
dicators of ancient soil development in the sequence of archaeological
sediments exposed in the excavation sections. Paleosols, soils formed
under environmental factors of the past different from the present day
ones, are encountered on the surface (relict soils) or overlain by
younger sediments (buried paleosols or fossil soils) in various terrestrial
geosystems (Catt, 1990)They contain valuable paleoecological proxies
recorded in a set of persistent pedogenetic properties and features that
are integrated within the concept of Soil Memory (Targulian and
Goryachkin, 2004)The resulting paleopedological records provide re-
latively high spatial resolution that is quite important for regional to
local paleoecological reconstructions. In particular paleosols linked to
the archaeological sites and materials could be useful for deciphering
environmental settings of ancient cultures as well as various aspects of
human-landscape interactions (Holliday, 2004)Microscopic indicators
provided by micromorphological studies are of special importance for
soil-archaeological research (Macphail and Goldberg, 2017) allowing
researchers to identify different stages of soil development and dis-
criminate between natural and human induced processes even when
pedogenetic features are incipient and/or affected by posterior diag-
enesis.
In this paper, we present the first results of the paleopedological
study of the archaeological layers accompanied by new archaeological
A.V. Triana-Vega et al.
and paleontological data. The obtained multiproxy record was inter-
preted to trace the local environmental evolution and the history of
human-landscape interactions in the period of transition from hunter/
gatherer to agricultural subsistence systems in the region.
2. Study region
The Savannah of Bogotá is an extensive elevated plain located in the
south of the Cundiboyacense plateau, in the geographic center of
Colombia. Its average altitude is about 2600 m.a.s.l. This region consists
of two physiographic zones: the high plateau and the surrounding
mountainous zone; climate is characterised by bimodal precipitation
and an altitudinal differentiation of temperatures. The latter controls
the distribution of the vegetation types that are subdivided into two
major altitudinal levels: Andean and high Andean (Van der Hammen,
1992, 1957). In particular, the area of the archaeological sites of Te-
quendama and Aguazuque has a mean annual temperature between 12°
and 18 °C and an annual precipitation varying from 500 to 1000 mm.
Tequendama archaeological site (4 °31′59.881″N and 74 °16′30.895″
W; 2574 m a.s.l.) is located in the estate of the same name and the site
of Aguazuque (4° 36′ 32,200 “N and 74° 16′50,375″ W; 2577 m a.s.l.) in
the Hacienda Fute. Although, both sites are quite close one to another
and at similar elevations, their vegetation cover differs: Tequendama is
characterised by a dry montane forest while humid mountain forest
vegetation predominates in Aguazuque (Correal, 1990; Correal and Van
der Hammen, 1977) (Figs. 1 and 2).
The sites are located at the periphery of the Basin of Bogotá which
has been studied from geological and geomorphological standpoint for
more than a century (Acosta and Ulloa, 2002, 2001; Hettner, 1892;
Hubach, 1931; Montoya, 2005; Pérez and Salazar, 1978) (Fig. 3). The
basament of the study area is formed by the Upper Cretaceous Gua-
dalupe Formation that consists predominantly of sandstones inter-
bedded with siltstones and clays. Above these rocks, a thick sequence of
Pliocene-Quaternary lacustrine deposits developed in the central part of
the Basin, the upper part is defined as Sabana Formation. The overlying
Late Quaternary Chia formation includes already large proportion of
alluvial materials (Helmens and Van der Hammen, 1994; Torres et al.,
2005). The lacustrine sequence incorporates multiple layers of pyr-
oclastic materials from Andean volcanoes (Riezebos, 1978).
Fig. 2. View of Tequendama archaeological site.
3
Quaternary International xxx (xxxx) xxx–xxx
In the Basins periphery, in which the study sites sit, the lacustrine
sequence is substituted by fluvial and colluvial deposits. They were first
described by (Julivert, 1961) as Quaternary “red and brown silts (limos
rojos y marrones)” capped by the “Black soil complex (complejo de suelos
negros)“. Later, a part of these footslope deposits were defined as the Rio
Siecha Formation (Helmens and Van der Hammen (1994).
Geomorphic position of the Tequendama site, that is a rock shelter
associated with a niche in the base of a sandstone escarpment (Fig. 2)
implies a major contribution of colluviation to sediment accumulation.
The open-air site of Aguazuque is located on a levelled terrace-like
surface more distant from the footslope (Fig. 3). More complex de-
positional environment is supposed here involving colluvial as well as
fluvial processes.
3. Materials and methods
3.1. Fieldwork
The sites of Tequendama and Aguazuque were excavated between
the 1970 and 1980 by Correal and Van der Hammen (1977) and Correal
(1990). From these excavations, human burials, bone remains of fauna,
lithic material, and ceramics were obtained. From these materials the
major contribution about the hunter-gatherers and early sedentary
groups in the Savannah of Bogota has been documented. The archae-
ological excavations continued in 2014 when five sections in Te-
quendama site and four in Aguazuque site were opened. The excavation
method was conducted using stratigraphic levels. All artefacts and
ecofacts were collected on the field by hand and soil samples were
collected to recover cultural materials using fine meshes. Each material
was identified with an individual code and separated by type material.
This phase of research focused on the geoarchaeological investigations,
including physical, chemical, micromorphological, and archae-
obotanical analyses of the soil and sediment layers.
Studied profiles appeared poorly differentiated in the field, thus the
identification of soil diagnostic horizons and soil type according to in-
ternational classifications is difficult. In most cases we could not detect
also clear boundaries between different sedimentary strata. That is why
the sampling scheme was based on the archaeological stratigraphy. In
this way, samples were collected from each archaeological stratigraphic
A.V. Triana-Vega et al.
Fig. 3. View of Aguazuque archaeological site.
level in both sites: small undisturbed soil blocks to prepare thin sec-
tions, and bulk samples for physical-chemical analysis and phytolith
extraction. In the field, a hydrochloric acid test was done to evaluate
the presence of carbonates (Reeuwijk, 2002), and the phenolftaleine
solution after NaF treatment to detect allophane using the methodology
of the FAO (FAO, 2009).
3.2. Laboratory analyses
The chronological scales for both studied sections integrated a
number of radiocarbon dates accumulated over decades of research at
the sites. For Tequendama we used some of the conventional radio-
carbon dates of the charcoal samples collected in 1977 from each ar-
chaeological layer and processed in the Radiocarbon laboratory in
Groningen (Correal and Van der Hammen, 1977). At Aguazuque we
included the C14 ages of human bones obtained by Correal (1990).
Additionally, the dating of human burial found in Aguazuque (strati-
graphic level 7A) in 2014 was carried out in the laboratory Interna-
tional Chemical Analysis Inc. (ICA) to be presented in this paper. All
dates are given in calibrated 14C years BP unless noted. The calibration
was done using Calpal 1.5. Table 1 integrates all utilized dates the ages
with the index GrN are from earlier publications whereas those labelled
Col AAA were obtained from the materials of 2014 excavations.
For micromorphological observations, sediment blocks with un-
disturbed structure were air dried and later impregnated with Crystal
resin. After solidification, they were cut with a diamond disk, mounted
on the glass slides, polished and covered by cover glass for microscopic
observations. These sections were investigated under an Olympus BX51
petrographic microscope using the Image - pro plus software and de-
scribed following the terminology of (Stoops, 2003). Some aspects of
identification and interpretation of micromorphological features in the
archaeological contexts are quite specific as compared to natural soils.
We utilized methodological approaches to micromorphological re-
search in the archaeological sites described by (Macphail and Goldberg,
2017) whereas some specific microscopic components were identified
using recent synthesis by (Nicosia and Stoops, 2017).
The bulk samples were dried at a constant temperature of 105 °C for
preparation for physical and chemical analyses.
For grain size distribution, the cementing agents were eliminated;
4
Quaternary International xxx (xxxx) xxx–xxx
organic matter with 30% H2O2 and carbonates with 0.5N HCl. The sand
fraction (2–0.02 mm) was separated by sieving, while silt (0.02–0.002)
and clay (< 0.002 mm) fractions by gravity sedimentation, following
the guidelines of USDA (2004).
Magnetic susceptibility (χ) was determined at low (LF) and high
(HG) frequencies: 0.47 and 4.7 kHz using a MS2B Bartington magnet-
ometer. Samples were previously dried at 60 °C and crushed in an agate
mortar to homogenize the material, and then placed in 8 cm3 acrylic
boxes. The frequency dependence of magnetic susceptibility was ob-
tained by: χdf% = [(χLF-χHF)/χLF]*100). The evaluation of the mag-
netic susceptibility (χ) has proven to be useful to differentiate soil
horizons from sediments and thus relate them to soil forming factors,
while the χdf% is used as an indicator of the ultrafine magnetic particles
formed by pedogenesis (Dearing et al., 1997).
pH and electric conductivity were measured in water with the soil-
to-solution ratio 1:2.5 after agitation for 3 h in a Thermo Scientific
Orion Versastar Pro. The measurement process was repeated twice in
each sample and was calibrated with the standards of pH 4.7 and 10
respectively.
The contents of organic matter and the inorganic carbon (carbo-
nates) were estimated by loss of ignition (LOI) at 450 °C and 800 °C,
respectively. This test was carried out in porcelain crucibles first heated
to 100 °C for 24 hours to obtain the initial weight of the dry sample
without humidity. Then the weight is obtained after heating at a tem-
perature of 450 °C for five hours (when decomposition of organic
matter is accomplished), and after heating for two hours at 800 °C
(when breakdown of carbonates takes place). The calculation of carbon
was made based on the differences in weight after heating at 450 °C and
800 °C, based on methodology of (Jackson, 1958).
Moreano, and Carvajal completed faunal identifications at
Universidad de los Andes Laboratory, Instituto de Ciencias Naturales
laboratory and Universidad Externado de Colombia's Archaeological
Laboratory. These institutions housed comparative collections that in-
clude more than 91 families and 40 genera of vertebrate and in-
vertebrate fauna. Relying on anatomical landmarks, all remains were
identified at the lowest possible taxonomic level. Specialized literature
was used to avoid confusions on the nomenclature (Chaix and Méniel,
2005; France, 2008; Olsen, 1982; Von Den Driesch, 1976). The samples
came from our 2014 excavations from four columns in Aguazuque and
A.V. Triana-Vega et al. Quaternary International xxx (xxxx) xxx–xxx

Table 1
Radiocarbon dates of the archaeological sites.
Arqueological site Stratigraphics level 14Ca conventional age 14Cb calibrated age (2 sigma) Dated Laboratory code References
material

Tequendama 9 Grainy Brown 2225 ± 35 B.P. 2182 BC: 2308 BP charcoal Col 159 GrN-6536 Correal and Van der
Below datum (50 cm) Hammen, 1977
Tequendama 8A Gray something 6990 ± 110 B.P. 7719–7926 B:P 7844–7979 BP charcoal Col 163 GrN-6728 Col Correal and Van der
brownish 7090 ± 75 B.P 164 GrN-6729 Hammen, 1977
Below datum (1.70 cm)
Tequendama 5B Light Brown 10130 ± 150 B.P. 11456 B:C 12080 BP 12327 B:C charcoal Col 176 GrN-6732 Col Correal and Van der
Below datum (3.40 cm) 10590 ± 90 B.P 12680 BP 12580 B:C 13111 BP 167 GrN-6505 Col 170 Hammen, 1977
10920 ± 260 B.P GrN-6539
Tequendama 3 Dark Brown 22250 ± 470 B.P. 26013 B:C 27543 BP charcoal Col 178 GrN-6579 Correal and Van der
Below datum (3.80 cm) Hammen, 1977
Tequendama 7A Brownish gray to 6080 ± 40 BPc 6897 B:C 7001 BP Human Col-AAA
brownish gray Bone
Below datum
(118–145 cm)
Aguazuque 52 Dark gray franco 2725 ± 35 B.P 2789 B:C 2856 BP Human GrN 14.479. Col. 594 Correal, 1990
Below datum (25 cm) Bone
Aguazuque 42 Franco carmelite 3850 ± 35 B.P 4200 B:C 4365 BP Human GrN 14.478. Col. 593 Correal, 1990
Below datum (75 cm) Bone
Aguazuque 41 Franco Sandy 4030 ± 35 B.P 4453 B:C 4550 BP Human GrN 12.930. Col. 477 Correal, 1990
Below datum (1.20 cm) Bone
Aguazuque 3 Franco reddish Sandy 5025 ± 40 B.P 5713 B:C 5866 BP Human GrN 14.477. Col. 1592 Correal, 1990
Below datum (135 cm) Bone
Aguazuque 52 Dark gray franco 3600 ± 40 BP c
3865 B:C 3965 BP Human Col-AAA
Below datum (84 cm) tooth

a
All dates are given in uncalibrated 14C years BP.
b
Calibration was done using Calpal 1.5.
c
The dates with code col-AAA are the dates obtained in the year 2017 for this research.

five columns in Tequendama. All excavated columns were field-
screened using 1 mm sieves. Each sample was processed and collected
all zooarchaeological remains, the remains of animals recovered from
fine sieves were not yet identified. In this study, the term “identified”
means that animal remains could be attributed to the lowest taxonomic
level possible below Superclass, i.e. family, order, genus or species
given the limitations of the available reference collections and ob-
servable morphological variation. Some badly fragmented bones could
only be classified as “unidentified”, which means that such remains
were only recognized as Mammalia).
The measures of relative abundance calculated included the number
of identified specimens (NISP), minimum number of individuals (MNI).
MNI was calculated using conservative criteria, such as symmetry and
fusion. Taphonomic attributes such as thermal alteration and staining
were registered.
4. Results
4.1. Morphological and archaeological description
In the Tequendama profile, six stratigraphic levels were dis-
criminated relying on the morphological characteristics of the sedi-
ments and those of the corresponding archaeological materials. The
profile stratification was quite complex because differences among
layers are sometimes subtle (Van der Hammen y Correal, 1977). The
basal layer (layer 4) was distinguished by its light yellowish colour and
silty-loamy texture; no archaeological materials were encountered
within this level. Above it, the layer 5B was found, which had a darker,
brown colour; its texture was loamy with an admixture of fine sand; few
lithic tools and bones were extracted. The overlying layer 7A had quite
different morphology: the colour was dark gray-brown, defined by the
humus pigment, the texture was sandy-silty loam, coarser than the
underlying strata. This layer produced the most abundant and variable
archaeological findings including one human feminine burial. Layer 8A
had similar properties however the number of artefacts decreased in
this layer; the upper limit of this layer was marked by a dark stratum,
probably an incipient humus topsoil horizon. The uppermost layer 9
and the recent surface topsoil horizon had a gray colour and a well
developed granular structure. The amount of the archaeological mate-
rial further decreased, however, different artefacts were found only in
this horizon: ceramic fragments and the remains of a stone floor
(Figs. 4b and 5 a,b and c).
The stratigraphic sequence of the Aguazuque site was also char-
acterised by rather weak morphological differences among the layers.
The basal layer 2 having light-brown colour and sandy-loamy texture
had not produced archaeological findings. The overlying loamy red-
dish-gray layer 3 already contained numerous bones but fewer lithics.
The frequency of lithic artefacts as well as the faunal remains increased
in the brown-coloured layers 4.1 and especially in the layer 4.2.
Additionally on this layer 4.2, Correal described a carbonized seed of
Cucurbita pepo and carbonized fragmented Ipomea batata (Correal,
1990. figure 69.3–69.4: 248). The dark reddish-gray layer 5.1 contained
the highest number of bones, including unarticulated human remains,
at the same time lithics were less abundant. Nevertheless, in the over-
lying layer 5.2 a well preserved secondary burial was excavated, which
provided a 14C date cal 3865–3965 B.P. The uppermost layer 6 con-
tained the remains of a fireplace, however, the overall amount of the
archaeological material was small; this layer contains recent artefacts
including glass fragments (Figs. 6b and 7 a,b and c).
The typology of the lithic artefacts in the Tequendama and
Aguazuque was rather uniform. Lithic debitage flakes were the most
abundant at both sites followed by cores, scrapers, hammerstone, and
knifes. However the frequencies of different types varied within the
profiles. In Tequendama the debitage fragments and scrapers were the
most common in the layer 9 whereas hammerstone and knives in-
creased slightly in the layers 7A and 8A. The site of Aguazuque was
characterised by a greater frequency of knives, especially in the layers
5.2 and 4.1. Also in all layers 3–4 scrapers were quite common.
The paleontological materials which form a large part of findings
will be characterised in a special section below.

5
A.V. Triana-Vega et al.
Fig. 4. Distribution of archaeological and paleontological materials in Tequendama site (a) Stratigraphic scheme of the profile; (b) Archaeological material; (c)
Faunal distribution.
4.2. Radiocarbon dating
During the large scale excavations carried out in years 1970–1980
by Correal and Van der Hammen (1977), and Correal (1990) a large set
of conventional radiocarbon datings was obtained from all stratigraphic
levels of both profiles. Recently two additional dates were obtained
from two human burials encountered during the excavations of 2014.
All the available datings are integrated in Table 1 showing that the ages
within the same layer were quite similar, whereas along the profiles
they produced a normal sequence without inversions. This set of
radiocarbon datings provided the basis for the chronological scale of
the studied sites. The archaeological materials in Tequendama corre-
sponded to the period from 10 to 7 BP and then after a long hiatus
continued around 2 BP. In Aguazuque several archaeological layers
were formed during the interval 6–2.5 BP. Thus the two studied sites
complement each other to produce a continuous archaeological record
Fig. 5. Section morphology and selected archaeological findings at the site of Tequendama. (a) view of the profile; (b) Stone floor associated with level 9; (c) female
individual found at level 7A. (d) Deer's pelvis (Odocoileus virginianus).
6
Quaternary International xxx (xxxx) xxx–xxx
covering the major part of the Holocene.
4.3. Micromorpholgical observations
In the Tequendama section we observed well developed granular
microstructure combined with small subangular blocks in the majority
of horizons. Granular aggregates were partly coalescent generating
microareas with spongy fabric. The layer 7A demonstrated a different
structural pattern: large very compact angular blocks were dominant,
separated by fractures (Fig. 8e). Fine clayey material of the basal layer 4
had light yellow-brown colour (Fig. 8c), whereas the overlying horizons
had a dark gray-brown colour due to abundant organic pigment. In the
coarse (sand and silt) fraction, quartz was the most common mineral,
however in all horizons we observed volcanogenic components: vol-
canic glass (Fig. 8d), plagioclases, and pyroxenes. Phytoliths were fre-
quent among silt particles. Very few primary carbonates were present
A.V. Triana-Vega et al. Quaternary International xxx (xxxx) xxx–xxx

Fig. 6. Distribution of archaeological and paleontological materials in the site of Aguazuque (a) Stratigraphic scheme of the profile; (b) Archaeological material; (c)
Faunal distribution. (d) Guinea pig's mandible (Cavia sp).

within the coarse material throughout the profile; some of the carbo-
nate particles were fragments of terrestrial mollusc shells.
Neoformed pedogenic carbonates in the layers 8A, 5B y 4 were re-
presented by the micritic and microsparitic nodules, hypocoatings and
local impregnations of the groundmass, as well as few incrustations of
the plant tissues (Fig. 8f). Only in the lower layers 5B and 4 we ob-
served very few small rounded ferruginous nodules (Fig. 8g).
In all horizons (including the basal layer, which produced no ar-
chaeological materials) we detected large charcoal particles with well-
preserved cellular structure. Bone fragments, some of them with strong
brown pigmentation and fracturing due to burning, were found in all
layers. Both charcoal and bone had the largest size and the highest
abundance in the 7A horizon (Fig. 8ab).
Additional thin section of a rock fragment from the roof of the rock
shelter had shown a typical arkose sandstone composition. It consisted
predominantly of quartz with admixture of feldspars and was quite well
sorted; its cement was clayey. No carbonate minerals were found within
this rock (Fig. 8h).
The micromorphological characteristics of Aguazuque profile were
quite similar to that of Tequendama concerning granular structure
(Fig. 9g), dark humus pigmentation of the fine material, presence of
volcanic components and primary carbonates (Fig. 9f) in the coarse
fraction, as well as inclusions of charcoal and bones (Fig. 9b); only here
however, we encountered thin fish bones (Fig. 9c). Secondary carbo-
nates were also present throughout the profile including incrustations
of plant tissue fragments, observed in the layer 51 (Fig. 9h). Phytoliths

Fig. 7. Section morphology and selected archaeological findings at the site of Aguazuque. (a) view of the profile; (b) Secondary burial of a female individual
associated with level 5 (2); (c) Knife with bilateral wear associated with level 4 (1).

7
A.V. Triana-Vega et al. Quaternary International xxx (xxxx) xxx–xxx

Fig. 8. Micromorphological features of the Tequendama archaeological site (a) Fragment of bone, level 7A; (b) Fragment of charcoal, level 7A; (c) Granular structure
with charcoal and phytoliths, level 4; (d) Volcanic glass, 5B; (e) Burned and compacted soil, level 7A; (f) Calcified plant tissue 5B; (g) Iron oxide nodule level 5B; (h)
Sandstone structure, roof of the rock shelter. a,b,c,d,e,g – plain polarized light; f,h – crossed polarizers.

are common in all horizons, sometimes aligned according their original
position in the plant tissue (Fig. 9e). Specific microscopic features in
Aguazuque were found: concentrations of sorted sand particles without
fine material in the layers 52, 42, and 3 (Fig. 9d) as well as fragment of
surface crust with vesicular voids in the horizons 41 and 2 (Fig. 9a).
Distribution of the micromorphological characteristics along the
studied sections is summarized in the Table 2
4.4. Physical and chemical characteristics of study profiles
The particle size distribution within Tequendama profile (Fig. 10a)
was characterised by a predominance of the silt fraction in all horizons
(Fig. 10b). A considerable increase of sand was observed in the layer
7A, where also a clear maxima of the χ and the χdf% were detected
(Fig. 10c). The pH values varied between 6.8 and 8.4 pointing to
neutral to moderately alkaline soil environment showing a modest
downward increase (Fig. 11b). The electric conductivity general ten-
dency showed higher values in the upper part of the section (Fig. 11b).
The LOI yielded quite low values. At 450° (indicative of organic matter)
varied between 1 and 2%, with slight increase upwards. LOI at 800°
(used to estimate carbonate content) was less than 1%, with irregular
fluctuations (Fig. 11c).
In the Aguazuque section (Fig. 12a), the particle size composition
showed high proportion of sand in the inferior and middle layers
(Fig. 12b), whereas the silt content increased in the upper strata. The χ
values are quite uniform along the profile with only minor variations,
however, χdf% had a clear maximum in the near surface, layers 5.2 and
6 (Fig. 12b). The pH values are higher than in Tequendama – between 7
in the upper part and 10 in the lower part (Fig. 13a), the latter pointed
to strongly alkaline environment. Simultaneously electric conductivity
increased significantly downwards (Fig. 13b). LOI at 450° and 800°
(Fig. 13c) produced similar values to Tequendama profile, both with
the observed tendency to increase at the top of the section.
4.5. Paleontological materials
Of the 1157 bones recovered from 1 mm sieve, 374 were identified
to at least family level. As observed in figure X and figure X, the relative

Fig. 9. Micromorphological features of the Aguazuque archaeological site: (a) Surface crust fragment, level 4 (1); (b) Fragment of burned bone, level 4 (1); (c) fish
bone fragment, level 4 (1); (d) sand cluster, level 2; (e) oriented phytoliths, level 2; (f) Shell fragment, level 2; (g) Granular soil structure, level 4 (1); (h) Calcitic
nodule, level 5 (1). a,b,c,e,g – plain polarized light, d,f,h = crossed polarizers.

8
A.V. Triana-Vega et al. Quaternary International xxx (xxxx) xxx–xxx

frequencies (counts NISP) suggests the zooarchaeological assemblage

on both sites are dominated by deer (Odocoileus virginianus) and guinea

carbonates
Secondary pig (Cavia sp) which are present in all layers. In Tequendama site, deer
remains are more prevalent on layers 7A and 5B, whereas guinea pig is

xxx
xxx
xxx
common on layers 7A, 8A and 9 (Fig. 4c). Zooarchaeological data from

xx

xx
x

x
x
Aguazuque suggest that deer bones are more common on the oldest
layers 3, 41 and 42, and their numbers decrease in the most recent
superficial
Feature pedogenetic

layers (Fig. X). The guinea pig's bones show high numbers in the layers
crusts

42 and 51. Our preliminary observations, without having allometric

data, suggest that there is an increase in the size of the animals. Guinea

x
pig's bones from Aguazuque are more common in Aguazuque than in
oxides

Tequendama.
Iron

x
x

x
Other animals represented on the sites are armadillo (Dasypodidae)
found on Tequendama's layer 7A and opossum (Didelphidae) on
Aguazuque's layer 5.1. Finally, the data also shows high counts of
carbonates
Primary

unidentified specimens. This is the result of several factors such as the

conservation and fragmentation, the presence of immature specimens,
xxx

xxx
xxx
xxx
xx
xx

xx

xx
x

x
x
x
x

heat treatment and the limitations of reference collections. The taxa

represented from 2014 excavations are similar to those reported by
Sandstone

Correal (1977, 1990). More detailed analyzes will allow more accurate
fragment

access to species and their proportions.

x

x
x

5. Discussion
Sedimentary material

cluster
Sand

5.1. Chronostratigraphy of Aguazuque and Tequendama in the context of

xx

xx

xx

the regional quaternary geology

Vocanic

Although the Savannah of Bogota is a paleolake basin, Aguazuque

Glass

xxx

xx

xx

xx

and Tequendama are located at the basin periphery in an elevated

x

x
x
x
x
x
x
x
x

position within a high terrace and thus do not correspond to the la-
Diatom

custrine paleoenvironment (Julivert, 1961). Due to the vicinity of the

x

mountains, the sites are affected by colluvial processes (with con-

siderable volcanic input) that cause continuous soil aggradation and
Calcified plant

development of paleosol-sedimentary sequences which host archae-

fragments

ological materials. The dates of the basal layer of Tequendama profile

correspond to the terminal Pleistocene, the overlying strata are formed
xx
x

in the early-middle Holocene, however the sequence presents a major

Mollusk

hiatus between 7000 and 3000 BP. It should be noted that the layer 7A,
immediately below this hiatus, is the richest in microartefacts and
x

structural changes related to human impact. It seems that during and

Spores
Distribution of micromorphological features in Tequendama and Aguazuque section.

immediately after its developmentcolluvial sedimentation slowed down

xx

xx
x

limiting new sediment supply and causing artefact concentration. This

Charcoal

brake of sedimentation can be linked to the Middle Holocene aridiza-

tion (discussed below). This gap is partly covered by a complementary
xxx

xxx
xxx
xxx

xxx
xx

xx

Aguazuque section that presents a set of archaeological layers 3–5.1

x

x
x
x
x

which give radiocarbon ages between 5000 and 3000 BP. These ages fit
ttissues

into the “lost” interval at Tequendama.

Plan

xxx
xxx
xxx
xxx
xx

Within the traditional regional scheme of the Quaternary deposits

x

by Julivert (1961), the soil-sedimentary sequences of both sites corre-

Phytoliths

spond to the uppermost “Black Soil Complex (complejo de suelos ne-

gros)“; only the basal stratum of Tequendana is related to the “Red and
xxx
xxx
xxx
xxx
xxx
xxx
xx
xx

xx
xx
xx
xx
xx
Biological materials

Brown Silts (limos rojos y marrones)” The correlation with the more
FishBone

developed Pliocene-Quaternary stratigraphy by Helmens and Van der

Hammen (1994) shows the link between the studied sequences with the
xxx
x

uppermost Chia Formation of the basin bottom and the uppermost part
Bone

of the Rio Siecha Formation, at the peripheral footslope area. In the

xxx
xx
xx
xx
xx
xx

xx

xx
xx
xx

xx
xx
x

detailed stratigraphic study of the central part of the Basin of Bogotá by

Torres et al. (2005), the Chia Formation is characterised as Late Pleis-
Stratigraphic

tocene/Holocene unit formed by a “mix of clay, sand and gravel of

fluvial origin” overlying lacustrine deposits of the Sabana Formation
Level

with a clear inconformity – that points to the recent degradation of the

CV

CV
8A
7A
5B

52
51
42
41
9

3
2

lake and possible the partial loss of the upper lacustrine layers, due to
Tequendama
Tequendama
Tequendama
Tequendama
Tequendama
Tequendama

erosion. It is interesting that these authors report “interbedded paleo-

Aguazuque
Aguazuque
Aguazuque
Aguazuque
Aguazuque
Aguazuque
Aguazuque
Aguazuque

soil” incorporated within the sediments of Chia formation (Torres et al.,

Table 2

2005).
Site

As we see, various studies of the Quaternary deposits of the Basin of

9
A.V. Triana-Vega et al. Quaternary International xxx (xxxx) xxx–xxx

Fig. 10. Physical characteristics of the Tequendama profile. (a) Stratigraphic scheme of the profile; (b) Texture (% of sand, silt and clay fractions); (c) Magnetic
Susceptibility: low frequency (SI) and frequency dependence (%).

Bogota mention paleosols as a prominent element of the upper Late
Quaternary sedimentary unit. A comparison of the paleopedological
phenomena at archaeological sites with paleosol levels in the natural
sedimentary sequences has relevance for developing the regional ped-
ostratigraphic scheme. Pleistocene paleosols in the region have been
first studied by Jungerius (1976) who related their development to the
glacial-interglacial cycles. Special investigation (including radiocarbon
dating) of Late Quaternary paleosols of the Eastern Cordillera of Co-
lombia, including the Bogotá, was carried out by (Fölster et al., 1977;
Fölster and Hetsch, 2017). In several exposures they observe sequences
of paleosols which show contrasting morphological differences: the
lower ones, corresponding to the Middle Pleniglacial, are pale

Fig. 11. Chemical characteristics of the Tequendama profile. (a) Stratigraphic scheme of the profile; (b)Conductivity μ s; (c) pH; (d) Loss of ignition.

10
A.V. Triana-Vega et al. Quaternary International xxx (xxxx) xxx–xxx

Fig. 12. Physical characteristics of the Aguazuque profile. (a) Stratigraphic scheme of the profile; (b) Texture; (c) Susceptibility.

hydromorphic soils (type a) whereas the upper paleosols formed during
the Upper Pleniglacial (type b) and the Holocene (type c) have dark
“humiferous” Ah horizons. We correlate the dark Holocene paleosols of
Tequendama and Aguazuque profiles with the type c paleosols of
Fölster et al. (1977).
Some of the paleosols reported by (Jungerius, 1976)and by Fölster
et al. (1977) are developed on pyroclastic sediments. In general
multiple volcanic ash layers are known to be a prominent element of
the Quaternary sedimentary sequences of the Eastern Cordillera region
(Riezebos, 1978)the key strata are provided with reliable fusion-track
dates of volcanic zircons (Andriessen et al., 1993). In both studied
sections tephrous materials (in particular porous volcanic glass) are
detected in all layers though they do not form any individual stratum
being mixed by colluvial processes with the components derived from

Fig. 13. Chemical characteristics of the Aguazuque profile. (a) Stratigraphic scheme of the profile; (b) Conductivity μ s; (c) pH; (d) Loss of ignition.

11
A.V. Triana-Vega et al.
local non-volcanic rocks.
Finally, it is important to compare the studied profiles with other
synchronous archaeological sites known in the Savannah of Bogotá. In
this concern the rock shelters of El Abra corridor present the closest
analogue especially with Tequendama site regarding both type and time
of occupation (Correal et al., 1966; Hurt et al., 1972; Van der Hammen,
1978) The sections in El Abra (especially El Abra III and IV) cover
longer time span. Their lower strata produce a set of radiocarbon dates
older than 20000 BP. C3 stratigraphic unit dated back to 12,400 ± 16
BP was considered as the oldest layer containing artefacts, however
(Muttillo et al., 2017)state that the lithics recovered from this unit, in El
Abra II are unworked and they originate from natural breakdown of the
rock shelter wall. These authors argue that true worked tools appear in
the upper younger part of unit C dated to 10000 BP and even there they
could be redeposited. The Tequendama profile corresponds to the units
C, D and E of El Abra sequence.
5.2. Pedogenetic and paleoecological interpretation of the soil-sedimentary
sequences
The sediments were continuously reworked by pedogenetic pro-
cesses that generated a rich “soil memory” record of the environmental
evolution (Targulian and Goryachkin, 2004). Extraction of this record
however meets with a major obstacle: the necessity to discriminate
between the properties produced by natural soil forming processes and
human-induced features that overprint the natural pedogenesis. Mi-
cromorphology permits us to identify a set of features that are clearly
related to the natural pedogenetic processes even when they are com-
bined with the ancient antropic effects. Micromorphological scanning
of the pedosedimentary sequences at the archaeological sites, with
long-term (millennial) occupation history, provides the “backbone” of
the paleopedological record of environmental change then enriched
with the quantitative laboratory soil characteristics. This approach was
successfully applied at the world-famous Kostenki European Palaeo-
lithic open-air sites (Sedov et al., 2010; Velichko et al., 2009) and at
Krems-Wachtberg (Terhorst et al., 2014) A well developed granular
structure with high inter-aggregate porosity giving rise to spongy fabric
together with the uniform colouring with dark organic pigment, points
to high biological activity and humus accumulation. These morpholo-
gical features are typical for the mollic horizons of Chernozems and
Phaeozems of grassland ecosystems (Pawluk, 1985)Nevertheless, the
estimated humus content by LOI (450°) is not high enough to meet the
requirements for a mollic diagnostic horizon.
However in the case study, pedogenetic effects of the tephrous
components, especially volcanic glass (encountered in all the horizons
of both profiles), should be taken into account. Quick alterations of
these components produce allophanes that further give rise to accu-
mulation of stable organo-mineral complexes and aggregation of the
soil material (Shoji et al., 1993)The resulting soils known as Andosols
are also characterised by strong dark humus pigmentation and stable
granular structure, similar to that of the Chernozems (Sedov et al.,
2010). However these properties of Andosols can form under humid
forest ecosystems and are not indicative of the grasslands. Strong po-
sitive reaction to allophane test, observed in most horizons, confirm the
andic properties of the studied paleosols.
Diversity of pedogenetic passways to the Mollic horizon develop-
ment renders uncertain and ambiguous paleoecological interpretation
of dark-coloured paleosols in the regions affected by volcanic sedi-
mentation (among them Savannah of Bogotá). consider late Pleistocene
dark buried Ah horizons in Colombian Cordillera to be an evidence of
aridization and grassland expansion during the last glaciation and do
not consider the possibility of Andosol development. Sedov et al. (2001)
while studying tephra-paleosol sequences of Central Mexico casts doubt
on the interpretation of Fölster et al. (1977), pointing to the possibility
that buried dark Ah horizons formed on tephra can belong to an An-
dosol formed in a humid forested environment. This ambiguity
12
Quaternary International xxx (xxxx) xxx–xxx
motivates us to look for additional paleopedological evidences for a
correct interpretation of the mollic-type features of the studied paleo-
sols.
Neoformed calcite is another important indicator of paleopedo-
genesis observed in the studied profiles in all horizons that helps to
refine paleoenvironmental interpretation. This component is re-
sponsible for a strong reaction with HCl demonstrated by most horizons
in the course of field description as well as for the high pH values. They
are presented in most cases by the micritic nodules, hypocoatings, or
local impregnation of groundmass and sometimes by calcified plant
tissues. All these neoformations are typical for pedogenic carbonates
precipitation from the migrating soil solute (Durand et al., 2010). The
combination of observed properties and features: Mollic horizon with
andic properties and neoformed carbonates allows us to classify the
dark paleosols of Tequendama and Aguazuque as Calcic Chernozems
(Andic) (WRB, 2014).
It should be stressed that high accumulation of the neoformed mi-
gratory carbonates are typical for the soils with water deficit formed
under dry climatic conditions. They are not expected within the recent
soil development under present day semi-humid climate. We interpret
these prominent pedofeatures as reliable indicators of the a past drier
environment. Besides carbonates, the Aguazuque section may contain
some more soluble salts in the lower horizons – their presence is evi-
denced by the increase of electric conductivity. Very high pH values
suggest that these salts are soluble carbonates, these components are
also typical for arid environments. As far as they are associated with the
archaeological levels dated back to 9000–3000 BP we conclude that
drier environment persisted over large part of the Early-Middle
Holocene and hosted the early sedentary population of the region.
Low weathering status reflected in relatively low clay content and
presence of pale minerals in the sand fraction (including relatively fresh
volcanic glass) agrees with the conclusion of the drier conditions during
the formation of most part of the sequence. Only in the basal layer of
Tequendama profile we observe higher accumulation of the yellow-
brown clayey fine material that points to stronger accumulation of
weathering products. Together with the ferruginous nodules indicative
of redoximorphic processes these features point to relatively moist soil
environment.
Comparison of this interpretation with the existing palynological
proxies looks ambiguous. The late Glacial (Guantiva interstadial) is
known to be the period of humidization and expansion of forest after
cold and dry Last Glacial Maximum (Helmens and Van der Hammen,
1994), this agrees with the characteristics of the basal layer of Te-
quendama section that indicates moister conditions. The early and
middle Holocene is mostly characterised as the period of maximal de-
velopment of the forest vegetation (dominated by Quercus) under warm
and mild climate (Helmens and Van der Hammen, 1994; Torres et al.,
2005; Van Geel and Van der Hammen, 1973). However, in more de-
tailed scheme of the Holocene climate change by Van der Hammen
(1992) the Middle Holocene (7500–5500 BP) is characterised as the
period of drier climate and decreasing lake level in the Savannah of
Bogotá. Further aridization took place in the subsequent interval 5500
to 3000 BP and only afterwards in the Late Holocene the tendency is
inverted towards moister conditions. A similar regional trend towards
mid-Holocene aridity is reported by (Marchant et al., 2002)An in-
dependent proxy – stable carbon isotope ratio of soil organic matter in
Savannah of Bogotá - also evidenced the spread of open vegetation with
large proportion of C4 grasses due to drying throughout the Holocene
that starts as early as 9500 B.P. (Guillet et al., 1988). We assume that
the accumulation of neoformmed carbonates corresponses to a dry
period during the middle Holocene (perhaps started in the early Ho-
locene). Later the climate changed to more humid conditions but this
humidization was not long and intensive enough to leach the carbo-
nates. Paleontological results provide an additional evidence for the
proposed scenario. The remains of deer (Odocoileus virginianus), the
animal typical for the gallery forests, combined with grasslands are
A.V. Triana-Vega et al.
abundant in the lower strata and quickly decrease upwards in the stu-
died profiles from its maximum in the 7A layer of Tequendama. At the
same time, the relative abundance of guinea pig (Cavia sp.), an animal
that prefers open landscapes, increases in the upper strata reaching
maximum in 5.1 of Aguazuque (Alberico et al., 2000)This trend could
be linked to the aridization occurring throughout the Middle Holocene.
However (Martínez-Polanco, 2016, 2011),mentions that both
guinea pig (Cavia sp) and deer (Odocoileus virginianus) are very adap-
table animals, even in their diet, which allows them to live in different
types of habitats. The change in proportions may be related to changes
in the social organization of human groups.
5.3. Archaeological and micromorphological records of occupation in the
context of history of human-landscape interaction
The sites of Tequendama and Aguazuque complement each other to
produce a continuous record of the Holocene human occupation until
the beginning of ceramic period in the Savannah of Bogotá. The con-
tinuous occupation of the sites for several thousand years produces not
only rich archaeological material but also a severe impact on the soils
developed on the occupation land surfaces. The quantity and variety of
the archaeological findings together with the abundance of the an-
thropogenic materials observed in the soil thin sections provide in-
formation about the intensity of the human occupation.
Micromorphological observations provide identification of anthropic
components and features and semi-quantitative estimation of their
abundances. Micromorphology is also used for evaluation of human
impact and its interaction with natural pedogenetic and sedimentary
processes in various archaeological contexts, e.g. paleolithic cultural
layers (Schilt, 2017).
In Tequendama the earliest artefacts are encountered in the layer
5B, whereas the basal layer 4 dated back to 12 BP has not yielded any
findings. Nevertheless, micromorphological observations have shown
frequent bone microfragments (some of them burnt) and charcoal in
this layer, these are interpreted as the evidence of human activity
(Courty et al., 1989)We conclude that humans could already inhabit
this rock shelter in the terminal Pleistocene. If this is true, this occu-
pation corresponds to the earliest known phase of early human dis-
persal in the northwest of South America (Aceituno et al., 2013)This
phase, represented by very few sites, is characterised by the occurrence
of archaeological materials together with the remains of Pleistocene
megafauna (Correal, 1993, 1981; Correal et al., 2005) The layer 4 of
Tequendama until now have not received adequate attention of ar-
chaeologists being considered as sterile; taking into account our mi-
cromorphological observations more extensive excavations into this
layer are recommended.
The maximal abundance of human-induced components and prop-
erties was observed in the layer 7a dated back to 7500–9500 BP. It is
marked not only by the numerous bone (including burnt) and charcoal
fragments some of them having very large, but also by complete
transformation of structure. There are few very few Biogenic granular
microareas, large compact blocks of fragmented shape which in-
corporate plant debris and small charcoal are dominant, pore space is
reduced: all these features point to anthropogenic compaction due to
trampling (Nicosia and Stoops, 2017) Structural changes caused by
trampling in the living floors corresponding to the occupation surfaces
are quite persistent and are clearly identified even in the sediment se-
quences of palaeolithic sites (Vallverdú-Poch and Courty, 2012)We also
interpret the maximum of magnetic susceptibility and frequency de-
pendence as human-induced feature related to burning. The same layer
produced the widest variety of the lithic tools, animal bones and a re-
markable human burial. This intense occupation corresponds to the
period of climate warming in the Early Holocene. The number of
findings decrease in the overlying layer 8A and afterwards a clear
hiatus is observed both in pedosedimentary and archaeological records
between the layers 8A (near 7 BP) and 9 (about 2 BP), the latter already
13
Quaternary International xxx (xxxx) xxx–xxx
containing ceramic material. As mentioned above this gap caused by
the sharp decrease of the colluvial sediment supply coincides with the
inferred drier phase in the middle Holocene.
The gap observed in Tequendama is filled with the records obtained
in the open-air site of Aguazuque: the 14C dates of the layers 3, 4.2, 4.1
and 5.1 fall within the interval 5000–3000 BP. Especially the layer 5.1
demonstrated maximum of macro- and microscopic human-induced
materials, which demonstrate certain diversification of the subsistence
practices. The specific lithic assemblage: high frequency of anvil stone,
cores, and cobble may indicate more intensive processing of vegetal
materials (Correal, 1990). Besides bone and charcoal fragments we
found other components of possibly anthropic origin. Oriented phyto-
lith concentrations can originate from plant materials brought by hu-
mans (e.g. in form of mats), it can also be evidence of weathered faecal
remains (Nicosia and Stoops, 2017). Fragments of surface crust are
likely to be a result of the structure degradation on the bare anthro-
pogenically disturbed soil surface. We suppose that the clusters of pure
sorted sand grains without admixtures of fine components and pig-
menting materials, surrounded by heterogeneous dark-coloured
groundmass observed in stratigraphic levels 52, 42 and 3 of the Agua-
zuque profile may also be human-induced. These clusters can hardly be
interpreted as sedimentary structures as far as they do not form mi-
crolayers or seams typical for sedimentation processes. Also it should be
noted that mineralogical composition of these concentrations is com-
pletely different from that of sand fraction in the groundmass. The latter
has high proportion of volcanic minerals: volcanic glass, plagioclases
and pyroxenes whereas the sand clusters are dominated by quartz.
Another possible explanation is that they are weathered and disin-
tegrated clasts of Cretaceous sandstone common in the region. However
there are no sandstone outcrops nearby Aguazuque. Sand was widely
used by ancient humans in various constructions, industrial, and do-
mestic activities and frequently forms microscopic concentrations in
archaeological pedosediments, e.g. (Sedov et al., 2017) described sand
clusters in the thin sections from the fill of epipaleolithic tell in Turkey.
As mentioned above, the Aguazuque site is characterised also by the
increase of Cavia sp. bones from lower to upper stratigraphic levels,
reaching maximum in the 5.1.layer. The abundance and also the larger
size of the bones suppose domestication of this animal between 5 and 3
BP. Earlier excavations yielded the charred macroremains of Cucurbita
pepo and Oxalis tuberosa. These observations agree with the idea that
the Middle Holocene was a period of the transition from hunter-gath-
erer practices to agricultural economy with an important component of
horticulture based both on the local and exogenous plants (Aceituno
and Loaiza, 2018)Archaeobotanical investigations in the region pro-
duced pollen, starch grains and the macroremain record of the use of
domestic and introduced cultivated plants in the middle Holocene.
Archaeological evidence from other regions in Colombia have
shown the manipulation of plants by early populations during the
middle and late Holocene. For example, by 6.5–5 BP phytolith and
starch grains extracted from ceramics and lithics have marked the use
of Palms, Poaceae, Cyperaceae, Annonaceae, Cucurbitaceae, Zea mays,
Manihot spp., Amaranthus Sp, and Impomea Sp (Aceituno, 2001; Castillo
and Aceituno, 2006)This data gives us an understanding of the types of
plants that may have been manipulated and domesticated during this
time period and in this region (Aceituno, 2001b; Castillo and Aceituno,
2006).
In the Middle Cauca Valley, on the western side of the Central
Cordillera do Colombia with humid premontane vegetation, there are
various early sites between 10120 ± 70 and 4180 ± 70 BP that have
been excavated (Aceituno and Loaiza, 2007)Archaeobotanical evi-
dence, including phytoliths, starch grains, and pollen obtained from
lithic artefacts, suggest that the introduction and the domestication of
plants like Xanthasoma spp., Manihot spp., Discorea spp., Phaseolus spp.,
and Zea mays may have occurred during the Middle Holocene with a
radiocarbon date of approximately 5 (Aceituno and Lalinde, 2011;
Aceituno and Loaiza, 2008, 2014). Zea mays pollen has been found in
A.V. Triana-Vega et al.
deposits ranging between 7 and 5 BP and the microbotanical remains
extracted from artefacts from the same sites also contain evidence of the
use of palms (Aceituno and Loaiza, 2008). Archaeological studies con-
ducted in the Central Cordillera at the site of Palestina (8-6 BP) report
the presence of macrobotanical remains of Arecaceae, Persea americana,
Fabaceae, Zea mays, and Cyperaceae (Morcote, Beltrán and Peña, 2010;
Júyar, 2014). Startch grains extracted form grinding stones of the same
site provided evidence for the use of Zea mays, Discorea spp., Calathea
Sp, and Poaceae plant (Herrera et al., 2016).
Isotopic research conducted on individuals from the sites of
Tequendama and Aguazuque have evidenced a change in diet during
the occupation of these sites from hunting and gathering strategies
towards more intensive plant collection, to finally the presence of cul-
tivated plants in the Late Holocene (Van der Hammen et al., 1990;
Cárdenas, 2002; Delgado, 2018).
The uppermost layers of both sites contain abundant ceramic frag-
ments and correspond to the period of well-developed agriculture,
which is also reflected in the palinological records (Helmens and Van
der Hammen, 1994).
6. Conclusions
All presented archaeological and paleoecological evidence leads us
to the conclusion that the transition from hunter/gatherer subsistence
system to the sedentary occupation based on agricultural practices oc-
curred gradually under the environmental conditions transforming to-
wards aridization. Climate and environmental change from the
Terminal Pleistocene to early and middle Holocene could have a sti-
mulating effect on the search for novel approaches to the resource
management. Micromorphological observations provide a set of well-
defined indicators of natural pedogenetic processes as well as ancient
human impact in the sequence of archaeological layers of the study
sites. They could serve as an instrument of archaeological prospection.
In particular, they validate the archaeological perspectives in the basal
layer of Tequendama profile earlier considered to be sterile.
Acknowledgement
We are grateful to the Stahl Grant (UC Berkeley), the Instituto
Colombiano de Antropología e Historia, the Fundación Nacional de
Investigaciones arqueológicas and Universidad de los Andes for pro-
viding funds and support for field and laboratory research. We are in-
debted to Universidad Nacional Autónoma de México and especially to
the Institute of Geology where the soil analysis of this research were
carried out. Finally, we thank the owners of Hacienda Tequendama and
Hacienda Fute for their support on the fieldwork.
References
Aceituno, F., 2001. Ocupaciones tempranas del bosque tropical subandino en la cordillera
centro Occidental de Colombia. Facultad de Geografía e Historia, Universidad
Complutense de Madrid, Madrid Tesis doctoral inédita.
Aceituno, F., Lalinde, V., 2011. Residuos de almidones y el uso de plantas durante el
holoceno medio en el cauca medio (Colombia). Caldasia 33.
Aceituno, F., Loaiza, N., 2008. Rastreando los orígenes de la agricultura en la vertiente
oriental del Cauca medio. In: Ecología Histórica: Interacciones Sociedad-ambiente a
Distintas Escalas Socio-temporales. Universidad Tecnológica de Pereira Universidad
del Cauca Sociedad Colombiana de Arqueología., Pereira, pp. 68–73.
Aceituno, F.J., Loaiza, N., 2018. The origins and early development of plant food pro-
duction and farming in Colombian tropical forests. J. Anthropol. Archaeol. 49,
161–172. https://doi.org/10.1016/j.jaa.2017.12.007.
Aceituno, F.J., Loaiza, N., 2014. Early and middle Holocene evidence for plant use and
cultivation in the middle cauca river basin, cordillera central (Colombia). Quat. Sci.
Rev. 86, 49–62. https://doi.org/10.1016/j.quascirev.2013.12.013.
Aceituno, F.J., Loaiza, N., 2007. Domesticación del bosque en el Cauca medio colombiano
entre el Pleistoceno final y el Holoceno medio. Archaeopress, Oxford.
Aceituno, F.J., Loaiza, N., Delgado-Burbano, M.E., Barrientos, G., 2013. The initial human
settlement of Northwest South America during the Pleistocene/Holocene transition:
synthesis and perspectives. Quat. Int. 301, 23–33. https://doi.org/10.1016/j.quaint.
2012.05.017.
Acosta, J., Ulloa, C., 2002. Mapa Geológico del departamento de Cundinamarca. Instituto
14
Quaternary International xxx (xxxx) xxx–xxx
de investigación e información geocientífica, minero-ambiental y nuclear.
Ingeominas, Bogota.
Acosta, J., Ulloa, C., 2001. Geología de la plancha 227 La Mesa. Instituto de investigación
e información geocientífica, minero-ambiental y nuclear. Ingeominas, Bogota.
Alberico, M., Cadena, A., Hernández-Camacho, J., 2000. Mamíferos (Synapsida: theria)
de Colombia. Biota Colomb. 34.
Andriessen, P.A.M., Helmens, K.F., Hooghiemstra, H., Riezebos, P.A., Van der Hammen,
T., 1993. Absolute chronology of the Pliocene-Quaternary sediment sequence of the
Bogota area, Colombia. Quat. Sci. Rev. 12, 483–501. https://doi.org/10.1016/0277-
3791(93)90066-U.
Cárdenas, F., 2002. Data on Prehispanic Feeding in the Sabana de Bogotá. Colombian
Institute of Anthropology and History, Colombia. Bogotá.
Castillo, N., Aceituno, F., 2006. El Bosque Domesticado, el Bosque Cultivado: un Proceso
Milenario en el Valle Medio del Río Porce en el Noroccidente Colombiano. Lat. Am.
Antiq. 17, 561–578. https://doi.org/10.2307/25063072.
Catt, J.A., 1990. Paleopedology manual Quaternary International. 6, 1–95.
Chaix, L., Méniel, P., 2005. Manual de Arqueozoología. Ariel prehistoria.
Correal, G., 1993. Nuevas evidencias culturales pleistocénicas y megafauna en Colombia.
vol. 8 Fundación de Investigaciones Arqueológicas Nacionales Banco de la Republica,
Bogotá. Año número 1.
Correal, G., 1981. Evidencias culturales de la fauna Pleistocénica en Colombia. Fundación
de Investigaciones Arqueológicas Nacionales Banco de la República, Bogotá.
Correal, G., 1990. Aguazuque: evidencias de cazadores, recolectores y plantadores en la
altiplanicie de la Cordillera Oriental. Bogotá, Colombia. Fundación de
Investigaciones Arqueológicas Nacionales Banco de la República, Bogotá.
Correal, G., Guitérrez, J., Calderón, K., Villada, D., 2005. Evidencias arqueológicas y
megafauna extinta en un salado del Tardiglacial Superior. Boletín de Arqueología 20
Boletín de Arqueología 20.
Correal, G., Van der Hammen, T., 1977. Investigaciones arqueológicas en los abrigos
rocosos del Tequendama: 12.000 Años de historia del hombre y su medio ambiente
en la altiplanicie de Bogotá. Fondo de promocion de la cultura del banco Popular,
Bogota.
Correal, G., Van der Hammen, T., Lerman, E., 1966. Investigaciones arqueologicas en los
abrigos rocosos de El Abra. Imprenta Nacional XIV, Colombia.
Courty, M.A.G., Goldberg, P., Macphail, R., 1989. Soils and Micromorphology in
Archaeology, vol. 77. Univ. Press Camb., pp. 39–59.
Delgado, M., 2018. Stable isotope evidence for dietary and cultural change over the
Holocene at the Savannah of Bogotá region, Northern South America. Archaeol.
Anthropol. Sci. 10, 817–832. https://doi.org/10.1007/s12520-016-0403-3.
Dearing, J.A., Bird, P.M., Dann, R.J.L., Benjamin, S.F., 1997. Secondary ferrimagnetic
minerals in Welsh soils: a comparison of mineral magnetic detection methods and
implications for mineral formation. Geophys. J. Int. 130, 727–736. https://doi.org/
10.1111/j.1365-246X.1997.tb01867.x.
Durand, N., Monger, H.C., Canti, M.G., 2010. Calcium carbonate features. In:
Interpretation of Micromorphological Features of Soils and Regoliths. Elsevier, pp.
149–194. https://doi.org/10.1016/B978-0-444-53156-8.00009-X.
FAO, 2009. Guía para la descripción de suelos. Organizacion de las Naciones Unidas para
la aricultura y la alimentacion, Roma.
Fölster, H., Hetsch, W., 2017. Paleosol sequences in the eastern cordillera of Colombia.
Quat. Res. 9, 238–248. https://doi.org/10.1016/0033-5894(78)90071-6.
Fölster, H., Hetsch, W., Schrimpff, E., 1977. Late Quaternary paleosols in the western and
central Cordillera of Colombia. Palaeogeogr. Palaeoclimatol. Palaeoecol. 21,
245–264. https://doi.org/10.1016/0031-0182(77)90037-2.
France, D.L., 2008. Human and Nonhuman Bone Identification: a Color Atlas, New. CRC
PR INC, Boca Raton.
Guillet, B., Faivre, P., Mariotti, A., Khobzi, J., 1988. The 14C dates and 13C/12C ratios of
soil organic matter as a means of studying the past vegetation in intertropical regions:
examples from Colombia (South America). Palaeogeogr. Palaeoclimatol. Palaeoecol.
65, 51–58. https://doi.org/10.1016/0031-0182(88)90111-3.
Helmens, K.F., Van der Hammen, T., 1994. The Pliocene and Quaternary of the high plain
of Bogotá (Colombia): a history of tectonic uplift, basin development and climatic
change. Quat. Int. 21, 41–61. https://doi.org/10.1016/1040-6182(94)90020-5.
Herrera, L., Moreno, C., Peña, O., 2016. Datos de un estudio sobre la ocupación humana
en la cordillera Central de Colombia: el Proyecto Arqueológico Aerocafé (Palestina,
Caldas). Bol. Mus. Oro 0, 103–173.
Hettner, A., 1892. Die Kordillere von Bogotá. (Mit Karten und Profilen). Ergänzungsheft
104 22, 131.
Holliday, V.T., 2004. Soils in Archaeological Research. Oxford University Press, New
York.
Hubach, E., 1931. Exploración de la región de Apulo - san Antonio - viotá. Bol. Minas Pet.
4 (25–27), 41–60.
Hurt, W.R., Hammen, T. v. d., Urrego, G.C., 1972. Preceramic sequences in the el Abra
rock-shelters, Colombia. Science 175, 1106–1108. https://doi.org/10.1126/science.
175.4026.1106.
Jackson, M.L., 1958. Soil Chemical Analysis. Verlag: Prentice Hall, Inc., Englewood
Cliffs, NJ.
Julivert, M., 1961. Observaciones del cuaternario sobre la Sabana de Bogota. Bol. Geol.
7, 35.
Jungerius, P.D., 1976. Quaternary landscape development of the Rio magdalena basin
between neiva and Bogotá (Colombia). Palaeogeogr. Palaeoclimatol. Palaeoecol. 19,
89–137. https://doi.org/10.1016/0031-0182(76)90044-4.
Júyar, X., 2014. Use of Plants in Forage Groups of Palestina Caldas (unpublished un-
dergraduate thesis). University of Caldas, Manizales, Colombia.
Macphail, R.I., Goldberg, P., 2017. Applied Soils and Micromorphology in Archaeology,
first ed. Cambridge University Press. https://doi.org/10.1017/9780511895562.
Marchant, R., Behling, H., Berrío, J.,C., Cleef, A., Duivenvoorden, J., Hooghiemstra, H.,
A.V. Triana-Vega et al.
Kuhry, P., Melief, R., Schreve-Brinkman, E., Van Geel, B., Van der Hammen, T., Van
Reenen, G., Wille, M., 2002. Pollen-based biome reconstructions for Colombia at
3000, 6000, 12000, 15000 and 18000 14C yr ago: late Quaternary tropical vegetation
dynamics. Late Quat. Trop. Veg. Dyn. J. Quat. Sci. 17 113e129.
Martínez-Polanco, M.F., 2016. El Cuy (Cavia Sp.), Un Recurso Alimenticio Clave en
Aguazuque, Un Sitio Arqueológico de la Sabana de Bogotá, Colombia. Lat. Am. Antiq.
27, 512–526. https://doi.org/10.7183/1045-6635.27.4.512.
Martínez-Polanco, M.F., 2011. La biología de la conservación aplicada a la
zooarqueología. La sostenibilidad de la cacería del venado cola blanca, Odocoileus
virginianus (Artiodactyla, Cervidae) en Aguazuque. Antipodas 99–118. https://doi.
org/10.7440/antipoda13.2011.06.
Montoya, D., 2005. Ministerio de minas y energía instituto colombiano de geología y
minería ingeominas. Geología de la Sabana de Bogotá. Ingeominas, Bogotá.
Morcote-Ríos, G., Beltrán, D., Peña, P., 2010. Determination of Seeds and Archaeological
Plant Remains of the Aerocafé Archaeological Project. Report presented to the
Aerocafé Archaeological Project (Unpublished). Aerocafé Archaeological Project,
Bogotá.
Muttillo, B., Lembo, G., Rufo, E., Peretto, C., Lleras Pérez, R., 2017. Revisiting the oldest
known lithic assemblages of Colombia: a review of data from el Abra and tibitó
(Cundiboyacense plateau, eastern cordillera, Colombia). J. Archaeol. Sci. Rep. 13,
455–465. https://doi.org/10.1016/j.jasrep.2017.04.018.
Nicosia, C., Stoops, G. (Eds.), 2017. Archaeological Soil and Sediment Micromorphology.
Wiley, Hoboken, NJ.
Olsen, S., 1982. An Osteology of Some Maya Mammals. Peabody Museum of Archaeology
&, Cambridge, Mass.
Pawluk, S.B.,L., 1985. Micromorphology of selected mollic epipedons SSSA special pub-
lication. Soil Micromorphol. Soil Classif. 15, 63–83.
Pérez, G., Salazar, A., 1978. Estratigrafía y facies del Grupo Guadalupe, vol. 10. pp. 6–85.
Piperno, D.R., 2011. The origins of plant cultivation and domestication in the new world
tropics patterns, process, and new developments. Curr. Anthropol. 52, S453–S470.
https://doi.org/10.1086/659998.
Reeuwijk, L.P., 2002. Procedures for Soil Analysis. International Soil Reference and
Information Center, Wageningen.
Riezebos, P.A., 1978. Petrographic aspects of a sequence of quaternary volcanic ashes
from the laguna de Fuquene area, Colombia, and their stratigraphic significance.
Quat. Res. 10, 401–424. https://doi.org/10.1016/0033-5894(78)90029-7.
Schilt, F., 2017. Micromorphology of an upper paleolithic cultural layer at grub-
Kranawetberg, Austria. J. Archaeol. Sci. Rep. 14. https://doi.org/10.1016/j.jasrep.
2017.05.041.
Sedov, S., Stoops, G., Shoba, S., 2010. Regoliths and soils on volcanic ash. In:
Interpretation of Micromorphological Features of Soils and Regoliths. Elsevier, pp.
275–303. https://doi.org/10.1016/B978-0-444-53156-8.00013-1.
Sedov, S.N., Aleksandrovskii, A.L., Benz, M., Balabina, V.I., Mishina, T.N., Shishkov, V.A.,
Şahin, F., Özkaya, V., 2017. Anthropogenic sediments and soils of tells of the Balkans
and Anatolia: composition, genesis, and relationships with the history of landscape
and human occupation. Eurasian Soil Sci. 50, 373–386. https://doi.org/10.1134/
S1064229317040093.
Sedov, S., Solleiro, E., Gama, J., Vallejo, E., González, A., 2001. Buried palaeosols of the
15
Quaternary International xxx (xxxx) xxx–xxx
Nevado de Toluca: an alternative record of Late Quaternary environmental change in
central Mexico. J. Quat. Sci. 16, 375–389. https://doi.org/10.1002/jqs.615.
Shoji, S., Nanzyo, M., Dahlgren, R., 1993. first ed. Volcanic Ash Soils, vol. 21.
Smith, B.D., 2001. Low-level food production. J. Archaeol. Res. 9, 1–43.
Solich, M., Bradtmöller, M., 2017. Socioeconomic complexity and the resilience of hunter-
gatherer societies. Quat. Int. 446, 109–127. https://doi.org/10.1016/j.quaint.2017.
06.064.
Stoops, G., 2003. Guidelines for Analysis and Description of Soil and Regolith Thin
Sections. Soil. Sci. Soc. Am. Inc Madison Wis, USA, pp. 184.
Targulian, V.O., Goryachkin, S.V., 2004. Soil memory: types of record, carriers, hierarchy
and diversity. Rev. Mex. Ciencias Geol. 21, 1–8.
Terhorst, B., Kühn, P., Damm, B., Hambach, U., Meyer-Heintze, S., Sedov, S., 2014.
Paleoenvironmental fluctuations as recorded in the loess-paleosol sequence of the
Upper Paleolithic site Krems-Wachtberg. Quat. Int. 351, 67–82. https://doi.org/10.
1016/j.quaint.2013.03.045.
Torres, V., Vandenberghe, J., Hooghiemstra, H., 2005. An environmental reconstruction
of the sediment infill of the Bogotá basin (Colombia) during the last 3 million years
from abiotic and biotic proxies. Palaeogeogr. Palaeoclimatol. Palaeoecol. 226,
127–148. https://doi.org/10.1016/j.palaeo.2005.05.005.
Vallverdú-Poch, J., Courty, M.-A., 2012. Microstratigraphic analysis of level J deposits: a
dual paleoenvironmental-paleoethnographic contribution to paleolithic archeology
at the abric Romaní. In: Carbonell i Roura, E. (Ed.), High Resolution Archaeology and
Neanderthal Behavior. Springer Netherlands, Dordrecht, pp. 77–133. https://doi.
org/10.1007/978-94-007-3922-2_4.
Van der Hammen, T., Urrego, G., Klinken, G., 1990. Stable isotopes and diet of prehistoric
man in the savanna of Bogotá (an initial study). Archeol. Bull. FIAN 5 (2), 3–10.
Van der Hammen, T., 1992. Historia, Ecología Y Vegetación. Corporación Colombiana
para la Amazonia, Araracuara.
Van der Hammen, T., 1978. Stratigraphy and environments of the upper quaternary of the
el Abra corridor and rock shelters (Colombia). Palaeogeogr. Palaeoclimatol.
Palaeoecol. 25, 111–162. https://doi.org/10.1016/0031-0182(78)90075-5.
Van der Hammen, T., 1957. Estratigrafía palinológica de la Sabana de Bogotá (Cordillera
Oriental de Colombia). BoI. Geol. (Bogotá) 5 (2), 187–203.
Van Geel, B., Van der Hammen, T., 1973. Upper quaternary vegetational and climatic
sequence of the fuquene area (Eastern Cordillera, Colombia). Palaeogeogr.
Palaeoclimatol. Palaeoecol. 14, 9–92. https://doi.org/10.1016/0031-0182(73)
90064-3.
Vavilov, N.I., 1992. Origin and Geography of Cultivated Plants. Cambridge University
Press, Cambridge.
Velichko, A.A., Pisareva, V.V., Sedov, S.N., Sinitsyn, A.A., Timireva, S.N., 2009.
Paleogeography of kostenki-14 (markina gora). Archaeol. Ethnol. Anthropol. Eurasia
37, 35–50. https://doi.org/10.1016/j.aeae.2010.02.002.
Von Den Driesch, A., 1976. A Guide to the Measurement of Animal Bones from
Archaeological Sites. Institut für Palaeoanatomie, Domestikationsforschung und
Geschichte der Tiermedizin. Institute Universität München.
WRB, 2014. World Reference Base for Soil Resources 2014: International Soil
Classification System for Naming Soils and Creating Legends for Soil Maps. FAO,
Rome.