See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/347464653

Earthen ridges in coastal sand dunes of the Caribbean coast, Venezuela:

Anthropogenic or natural?

Article  in  Geoarchaeology · December 2020

DOI: 10.1002/gea.21835

CITATIONS READS

0 49

10 authors, including:

Maximiliano Bezada Miguel Zavala Reyes

Universidad Pedagógica Experimental Libertador Venezuelan Institute for Scientific Research
16 PUBLICATIONS   41 CITATIONS    16 PUBLICATIONS   2 CITATIONS   

SEE PROFILE SEE PROFILE

Erika Paola Pedraza Castellanos Sergio Foghin

Simon Bolívar University Universidad Pedagógica Experimental Libertador
13 PUBLICATIONS   5 CITATIONS    4 PUBLICATIONS   2 CITATIONS   

SEE PROFILE SEE PROFILE

Some of the authors of this publication are also working on these related projects:

Estudio interdisciplinario para abordar la historia local en la Región Coriana View project

Geomorphological sedimentological and morphodynamic analysis of the tropical estuary of Osma, La Guaira, Venezuela View project

All content following this page was uploaded by González Orlando on 18 December 2020.

The user has requested enhancement of the downloaded file.

Received: 1 June 2018 | Revised: 15 October 2020 | Accepted: 28 October 2020

DOI: 10.1002/gea.21835

RESEARCH ARTICLE

Earthen ridges in coastal sand dunes of the Caribbean coast,

Venezuela: Anthropogenic or natural?

Maximiliano Bezada1,2 | Haydn Barros3 | Miguel Zavala4 | ERika Pedraza4 |

Sergio Foghin1 | Orlando González1 | Darwin Godoy1 | Eduardo Zambrano4 |
Roberto Herrera5 | Alejandra Leal4,6

1
Departamento de Ciencias de la Tierra,
Universidad Pedagógica Experimental Abstract
Libertador, Caracas, Venezuela
Earthen ridges that occur in the active Los Médanos de Coro dune fields of the Ve-
2
Minnesota Geological Survey, College of
Science and Engineering, University of
nezuelan Caribbean coast have been attributed either to anthropogenic or natural
Minnesota, Minneapolis, Minnesota, USA origins. Anthropogenic explanations have focused on their location, geometry, struc-
3
Departamento de Física, Universidad Simón ture, and compaction, in the context of archaeological sites and artifacts throughout the
Bolívar, Caracas, Venezuela
4
dunes, to suggest an origin as causeways built by the extinct Caquetío people and/or
Departamento de Estudios Ambientales,
Universidad Simón Bolívar, Caracas, their pre‐Hispanic ancestors. To assess this topic, geomorphic, physicochemical, and
Venezuela elemental analyses were carried out on the ridges and interdunal beaches and playas.
5
Dirección de Gestión Ambiental, Universidad
The configuration, elevation, and composition of the ridges, along with three bulk
Bolivariana de Venezuela, Caracas, Venezuela
6 sediment 14C dates that range 1900–3000 year BP, are here found to be most com-
Centro Universitario Regional del Este,
Universidad de la República, Rocha, Uruguay patible with a natural origin as exhumed paleochannels. Past human occupants may,
however, have used the features as causeways.
Correspondence
Alejandra Leal, Laboratorio de Palinología,
KEYWORDS
Facultad de Ciencias, Udelar, Iguá 4225,
11400 Montevideo. Caquetío, dunes, Falcón, Médanos de Coro
Email: [email protected]

Scientific editing by Nicholas Dunning

1 | INTRODUCTION peninsula (Figure 1). The dunes likely served as a camping area for the
Caquetío and their descendants, rather than a permanent settlement,
The Médanos de Coro National Park (MCNP) on the Caribbean coast due to the difficult environmental conditions of the active dunes (Oliver,
of Venezuela is located near the city of Coro in Falcón State 1989). The active dunes contain sites with a high density of pottery
(Figure 1) in an area known for stabilized and active dune fields. fragments (tiesteros), shell fragments (concheros), and lithic materials
Abundant archaeological materials associated with the dunes are (metates and manos) at the surface (Oliver, 1989; Zavala et al., 2018).
remnants of pre‐ and postcontact peoples who were bearers of Radiocarbon dates on charcoal from hearths indicate occupation in the
Dabajuroid pottery (Cruxent & Rouse, 1958; Oliver, 1989). Among MCNP in the 14th and 15th centuries A.D. Other Caquetío sites in the
them, the Caquetío people, of the Arawak linguistic branch (Arcaya, northern Falcón area date as early as AD 700–800 (Oliver, 1989).
1920), inhabited western Venezuela at the time of the Spanish Contained within interdunal playas in the MCNP active dune
conquest (Cey, 1995). These people are notable, having established field are straight, curving, to sinuous earthen ridges with rectangular
sedentary villages and complex agricultural systems along the to tabular form of varying width (Figures 2 and 3). The ridges rise
northwest coastal plain of Venezuela and in the Caribbean Antilles 0.5–0.8 m above the playa floor with typical widths of approximately
(Antczak et al., 2017; Arcaya, 1920; Strauss, 1992). 1 m expanding to over 8 m in places. Lengths of the ridges are
The active dunes examined in this study are located on the isthmus variable, with some exceeding 300 m. Notable ridges are perpendi-
that allowed human passage from the mainland to the Paraguaná cular to the active barchan dunes, although others are parallel to

Geoarchaeology. 2020;1–12. wileyonlinelibrary.com/journal/gea © 2020 Wiley Periodicals LLC | 1

2 | BEZADA ET AL.

F I G U R E 1 Location of the study site in northwestern Falcón State, Venezuela (GoogleEarthTM) [Color figure can be viewed at
wileyonlinelibrary.com]

them. In both cases, the ridges can be traced beneath the dunes et al., 2011; López‐Mazz et al., 2016). In Venezuela, the best‐
between playas (Figure 2). The abundance of archeological materials documented large earthworks of human origin are those of
along the ridges has been taken as an indication that the Caquetío or Barinas state in the western llanos. There, the initiation of
their pre‐Hispanic ancestors frequently walked along these land- earthwork construction is thought to be have been around AD
forms, thus treating them as causeways (Hernández‐Baño & Trujillo, 700–500, with the intensification of activity much later (Gassón,
1986; Morón, 2011). 1997, 2002; Redmond & Spencer, 1990; Spencer & Redmond,
It has been proposed, in particular by Hernández‐Baño (1984), 1983, 1992). Human origin of these earthworks is indicated by
Hernández‐Baño and Trujillo (1989), and Morón (2011), that these their arrangement, presence of cultural materials in their matrix,
earth ridges were intentionally constructed as causeways on the and their apparent age contemporaneous with nearby habitation
basis of the following observations: structures. Hernández‐Baño (1984), therefore, indicated that it is
quite possible that the Caquetíos also built earthworks.
1) Their location in a region that also contains well documented 2) Strategic Location: The ridges associated with archaeological
prehispanic roads: Earthworks are widespread in South America, sites are in areas subject to periodic flooding, a possible moti-
and include varied types of mounds, ridges, elevated fields, and vation for constructing causeways to facilitate passage (D′
causeways, that were built for many purposes, although the uti- Angelo‐Pietri, 1981). Today, the ridges protrude above water
lity of some of these features is still under discussion (Bracco levels of seasonal lagoons (Figure 2). This resembles the
BEZADA ET AL.
| 3

F I G U R E 2 Aerial view of the ridges in year 2015 (a) and 2011 (b) (GoogleEarthTM). View of ridge 1 in the field (c,d) [Color figure can be
viewed at wileyonlinelibrary.com]

topographic position of causeways such as those at Llanos de (Figure 2d) have been interpreted as an indication of human
Moxos in the Bolivian Amazon (Erickson, 2000, 2011). traffic (D′Angelo‐Pietri, 1981).
3) A general lack of primary sedimentary structures (D′Angelo‐
Pietri, 1981): The apparent absence of stratigraphy together with The alternative explanation is that the ridges are natural sedi-
the presence of charcoal fragments embedded in the ridge mentary features. It seems apparent that the most likely mechanism
structure have suggested to Hernández‐Baño (1984) that the for ridge formation would be fluvial processes and topographic in-
ridges were earthworks. The charcoal is seen to likely be of hu- version (Pain & Oilier, 1995). DiBiase et al. (2013) described two
man origin, as the present vegetation cover is too discontinuous inversion mechanisms—landscape inversion and deposit inversion. In
to support wildfires. the former, denudation leaves a former depression‐filling, erosion‐
4) Differential compaction along the ridges: A slight depression and resistant material standing higher than the surrounding terrain. In
greater compaction along the central axes of the ridges the latter, sediments less resistant to erosion are removed leaving
4 | BEZADA ET AL.

F I G U R E 3 Geomorphic features occurring in the active dune systems of the MCNP (a). Exhumed inverted paleochannels (b,c), and exhumed
remains of former runoff surface (d,e), along ridges 1 and 2. MNCP, Médanos de Coro National Park [Color figure can be viewed at
wileyonlinelibrary.com]

more resistant materials standing higher. An inverted channel thus (Ritter, 1967), and inverted river channels in the Sierra Nevada of
forms when the channel floor material becomes more resistant than California (Busby et al., 2016; Giegengack, 1968; Lindgren, 1911;
the surrounding terrain and becomes inverted when differential Whitney, 1880). Additional examples have been reported in Africa
erosion removes the surrounding less resistant material (Pain et al., (Butt & Bristow, 2013; Giegengack & Zaki, 2017; Giegengack,
2007). The resistance of the channel floor sediments is associated 1968; Zaki & Giegengack, 2016; Zaki et al., 2018); Australia (Kemp
with induration by cementation or compaction. & Rhodes, 2010); Europe (Cuevas et al., 2010; Friend et al., 1981);
Examples of comparable features in mountainous areas are and North America (Williams et al., 2007). Inverted channels have
stepped topography (Wahrhaftig, 1965), terrace development been important in the study of past climate, in relation to
BEZADA ET AL.
| 5

paleo‐drainage systems (Williams et al., 2007; Williams et al., the coast were carried out along parallel transects, covering the
2009; Zaki et al., 2020). entire length of the barchan dune system. A total of 100 sampling
Here we present a study of the configuration, geometry, com- points were recorded with a Garmin GPS navigator whenever a ridge
position, and geochronology of the earth ridges at MCNP and assess was found, and then data were uploaded into a Google EarthTM file
whether the features were built by the Caquetíos or are instead an producing a map of the ridge locations. The location and character-
example of inverted channels and topographic reversal. istics of archaeological material such as pottery, shells, and artifacts
were also recorded. The main ridges identified were sampled in 0.6 m
deep soil pits exposing ridge sediments and underlying deposits
2 | STUDY AREA (Figure 4). Soil pits were manually excavated and samples were
stored in aluminum boxes. Three interdunal playas were also studied
The study area is located in the southeastern part of the MCNP to compare their sediments with those of the ridges (Figure 5). Each
within the easternmost active dune field (Figure 1). The average playa was sampled by a single core obtained using a manual coring
annual precipitation in the city of Coro is less than 400 mm per year, device. Two sand samples of the surrounding barchan dunes were
and the average annual evaporation exceeds 2,300 mm per year. also collected for comparison as control samples.
With an average annual temperature of 28°C, the study area clas-
sifies as a warm desert or BWhi according to the Köppen‐Geiger
classification. Throughout 80 years of rainfall records, the annual 3.2 | Physicochemical and elemental analysis
rainfall amounts have varied between 100 and 900 mm. Rains tend
to be concentrated in October–November (Foghin, 2002; Matteucci Sediments from both ridges and playas were air‐dried and meshed
& Colma, 1986). Prevailing winds vary between ENE and ESE, with through a 2 mm sieve. Their color was characterized using a Munsell
average monthly speeds between 4.8 and 6.9 m per sec. color chart.
The current vegetation of the active dunes presents a short and Soil analyses were carried out in duplicate at each depth and site in
discontinuous cover. Where vegetation is present, the cují yaque (Pro- the Laboratory of Ecological Processes at the Universidad Simón Bolí-
sopis juliflora), and the úveda (Acacia tortuosa) dominate (Matteucci, var. Samples were dried at 105°C. Grain size analyses were completed
1987). To the south, the vegetation is more or less dense and consists by the Bouyoucos method, using sodium polyphosphate as dispersant
mainly of the cují yaque, the yabo (Cerdicium praecox), the úveda, the (Pansu & Gautheyrou, 2006). Organic carbon analyses (SOC) were done
tunas (Euphorbia grandicornis) and the cardones (Cereus sp.). by the Walkley and Black colorimetric method (Pansu & Gautheyrou,
The MCNP displays a variety of landforms (Figure 3). The 2006), and both organic matter (SOM) and inorganic carbon (SIC) were
northeastern portion of the park on the isthmus is characterized by estimated by loss‐on‐ignition (Ball, 1964), with modifications for
stabilized longitudinal and parabolic dunes (Camacho et al., 2011). carbonate‐rich soils (Wang et al., 2012). Total nitrogen (Nt) was de-
We regard these sediments as former marine beach deposits accu- termined by the Microkjeldahl method (Pansu & Gautheyrou, 2006),
mulated during the postglacial transgression, which were uplifted and available phosphorus (Pa) through an Olsen extraction. The pH was
and exposed to the wind by neotectonic processes that started ap- determined in a 1:2 soil–distilled water solution. Elemental composi-
proximately 3000 yr ago (Audemard et al., 1995; Audemard, 1996). tions were determined in the Nuclear Physics Laboratory at Simon
This indicates a late‐Holocene age for eolian deposits on the isthmus. Bolivar University by Total Reflection X‐Ray Fluorescence (TXRF), with
In the southeastern portion of the park, where the current study site added selenium as internal standard.
is located, stabilized longitudinal and parabolic dunes occur. Also present Physicochemical results are shown in Tables 1 and 2 and in
are active transverse barchan dunes with heights up to 20 m (Figure 1). Figure 4. Elemental analyses are in the supplementary material and
Eroded portions of the dunes reveal buried mineralized root casts or were included in the ordination of samples taken in the ridges and
rhizoliths (Figure 3) suggesting a history of dune activity and stabilization playas through a nonmetric dimensional scaling (NMDS). NMDS was
(Henningsen et al., 2003). Beneath the eolian deposits is an impermeable carried out in Vegan (R) after transformation and standardization of
basement of fine alluvial fan sediments that contributes to the ponding of the data. Charcoal presence/abundance analysis involved observing
water in the interdunal playas within the barchans during wet years wet samples under a light microscope, with charcoal identification
(Figure 2). based on the presence of black, opaque, angular remains (Patterson
et al., 1987), after treatment with 6% hydrogen peroxide for 96 h
(Conedera et al., 2009). At each level, charcoal was described as very
3 | METHODS abundant, abundant, scarce, or very scarce.

3.1 | Field surveys

3.3 | Geochronology
Field surveys were conducted 2014–2016 in the southeastern active
dunes of the MCNP (Figure 1). Several pedestrian surveys aligned Three bulk sediment samples for radiocarbon dating were taken in
northeast to southeast from the Coro‐Punto‐Fijo road (Figure 1) to one ridge, at depths of 0–18 cm (top), 18–40 cm (middle), and
6 | BEZADA ET AL.

F I G U R E 4 Profiles sampled in ridges 1 and 3 with Munsell colors and indications of the amount of charcoal found and the grain‐size at
different depth levels [Color figure can be viewed at wileyonlinelibrary.com]

14
40–60 cm (lower). These samples were air‐dried, roots were elimi- spectrometry for C determination at the National Accelerator
nated by flotation and sieving, 10% HCl was added until the elim- Centre (CNA, Seville, Spain). Radiocarbon dates were calibrated with
ination of carbonate, and 10% KOH was added until the elimination the IntCal13 curve (Stuiver et al., 2018) in the software Calib 7.1.,
of humic acids. The sediments were then dried and processed at the obtained from http://calib.org. Although charcoal was found in the
Gliwice Absolute Dating Methods Centre (GADAM Centre, Gliwice, profiles, the fragments were deemed to be too small and fragile to be
Poland) for graphitization and then measured by accelerator mass sampled for dating.
BEZADA ET AL.
| 7

F I G U R E 5 Distribution of ridges in the easternmost active dune field of the MCNP, with location of the three studied ridges, and the three
interdunal playas, that were studied (core 1–3) indicated [Color figure can be viewed at wileyonlinelibrary.com]

4 | RESULTS parallel to the main trade winds direction and one with a N‐S or-
ientation. The length of dune fragments in the first two systems
4.1 | Geometry of the ridges range ca. 200–600 m, but they are probably continuous and could
reach a total length of more than 1000 m. During field visits, some
Among ridges along the eastern active dune field, there are three barchans were observed to migrate to the west, corroborating ridge
main systems (Figure 5): two oriented in the WSW‐ENE direction, continuity beneath the dunes. The third system is parallel to the

TABLE 1 Physical and chemical analyses of ridge sediments, and control sample corresponding to the topsoil of adjacent active dunes
SOC SOM SIC Nt Pa
N° Relative position Depth (cm) pH % % % % mg/kg

Ridge 1 Top 0–5 8.05 (0.21) 0.57 (0.09) 3.02 (0.10) 3.45 (0.03) 0.037 (0.002) 31.70 (2.15)

Middle 5–15 8.15 (0.30) 0.40 (0.03) 2.35 (0.18) 2.00 (0.23) 0.023 (0.004) 17.32 (0.09)

Down 15–25 8.30 (0.02) 0.35 (0.01) 1.95 (0.13) 1.59 (0.01) 0.023 (0.005) 18.47 (0.18)

Bottom 25–30 7.92 (0.20) 0.29 (0.00) 0.87 (0.00) 2.03 (0.04) 0.002 (0.000) 17.34 (0.12)

Ridge 2 Top 0–5 8.17 (0.03) 0.38 (0.00) 0.98 (0.15) 1.66 (0.13) 0.007 (0.001) 17.85 (0.29)

Middle 5–15 8.43 (0.12) 0.57 (0.02) 2.48 (0.51) 2.87 (0.28) 0.027 (0.003) 20.57 (0.34)

Down 15–25 8.11 (0.07) 0.32 (0.00) 1.06 (0.20) 1.54 (0.03) 0.002 (0.003) 18.13 (0.02)

Bottom 25–30 8.17 (0.11) 0.28 (0.02) 1.27 (0.48) 2.08 (0.01) – 17.56 (0.44)

Ridge 3 Top 0–18 8.15 (0.04) 0.59 (0.01) 1.95 (0.03) 3.73 (0.00) 0.023 (0.001) 22.44 (4.08)

Middle 18–40 8.25 (0.01) 0.48 (0.03) 1.20 (0.04) 2.24 (0.04) – 18.03 (0.56)

Down 40–60 8.32 (0.01) 0.31 (0.02) 0.80 (0.08) 1.00 (0.05) – 19.95 (0.56)

Bottom 60–80 8.37 (0.01) 0.25 (0.02) 0.29 (0.03) 0.88 (0.10) – 19.50 (1.95)

Control Topsoil 0–5 8.21 (0.07) 0.32 (0.01) 0.62 (0.07) 3.04 (0.66) – 15.89 (0.03)

Abbreviations: Nt, total nitrogen, Pa, available phosphorus; SIC, soil inorganic carbon; SOC, soil organic carbon, SOM, soil organic matter.
Note: Standard deviations are shown between parentheses.
8 | BEZADA ET AL.

TABLE 2 Physicochemical analyses from the interdunal playas

SOC SOM SIC Nt Pa
N° Depth (cm) Grain‐size class pH % % % % mg/kg

Core 1 0–15 Sandy‐clay 8.68 (0.01) 0.99 (0.04) 1.37 (0.19) 6.35 (0.29) 0.064 (0.002) 28.23 (2.78)

22–35 Sandy‐clay‐loam 9.04 (0.05) 0.31 (0.03) 0.88 (0.15) 2.20 (0.03) 0.017 (0.001) 25.74 (5.14)

35–44 Sandy‐clay‐loam 9.03 (0.02) 0.29 (0.00) 0.71 (0.05) 2.17 (0.02) 0.018 (0.001) 20.28 (0.81)

94–110 Sandy‐clay‐loam 9.08 (0.11) 0.59 (0.02) 1.35 (0.18) 5.09 (0.40) 0.045 (0.001) 27.84 (3.34)

110–115 Sandy‐clay 9.04 (0.03) 0.31 (0.01) 0.90 (0.00) 7.89 (0.15) 0.038 (0.001) 16.88 (0.06)

Core 2 0–2 Clay 8.09 (0.14) 0.34 (0.00) 1.23 (0.12) 4.86 (0.37) 0.036 (0.026) 17.53 (0.56)

2–10 Clay 8.23 (0.06) 0.34 (0.01) 1.12 (0.17) 4.88 (0.19) 0.017 (0.001) 17.95 (0.31)

10–20 Clay 8.19 (0.01) 0.32 (0.01) 1.34 (0.07) 5.00 (0.18) 0.012 (0.003) 17.32 (0.07)

25–33 Clay 8.24 (0.07) 0.25 (0.01) 1.54 (0.29) 4.52 (0.50) 0.025 (0.001) 17.41 (0.78)

52–63 Clay 8.32 (0.07) 0.22 (0.05) 1.15 (0.12) 4.81 (0.05) 0.022 (0.001) 17.47 (0.14)

97–104 Clay 8.28 (0.09) 0.20 (0.01) 1.08 (0.19) 5.78 (0.47) 0.014 (0.000) 17.26 (0.25)

Core 3 0–7 Sandy‐clay‐loam 8.71 (0.21) 0.16 (0.01) 0.46 (0.07) 3.11 (0.09) – 24.05 (0.69)

7–12 Clay 8.68 (0.06) 0.25 (0.00) 1.13 (0.03) 4.72 (0.32) 0.007 (0.010) 16.81 (0.15)

12–21 Clay 8.49 (0.07) 0.25 (0.00) 1.41 (0.12) 5.00 (0.038) 0.019 (0.000) 16.38 (0.22)

21–29 Clay 8.19 (0.01) 0.34 (0.06) 1.30 (0.11) 5.21 (0.17) 0.027 (0.003) 17.87 (0.23)

78–84 Clay 8.34 (0.11) 0.21 (0.01) 0.92 (0.06) 4.26 (0.06) 0.013 (0.001) 17.22 (0.24)

150–161 Sandy‐clay‐loam 8.47 (0.05) 0.22 (0.01) 1.10 (0.08) 4.27 (0.15) – 18.43 (0.71)

Abbreviations: Nt, total nitrogen, Pa, available phosphorus; SIC, soil inorganic carbon; SOC, soil organic carbon, SOM, soil organic matter.
Note: Standard deviations are shown between parentheses.

Coro‐Punto Fijo Road and perpendicular to the trade winds direc- of the ridges and interdunal playas, although the sediments differed
tion. The southernmost fragment of the N‐S system was completely in compaction.
uncovered during the fieldwork and was found to be ca. 300 m long. Cultural remains were observed on dunes, interdunal playas, and
Several isolated mounds and small segments also were observed in ridges. The active dunes contain many sites, some apparently de-
the easternmost and southern portions of the active dune field flated, with pottery such as Dabajuroid ceramics (Cruxent & Rouse,
(Figure 5). Beyond the active dune field, ridge fragments were also 1958), shells (Donax sp., Strombus sp., Melongena sp., Valva calista,
found within the stabilized longitudinal–parabolic dune field etc.), and lithic artifacts. No artifacts were observed within ridge
(Figure 5). The ridges thus mainly occur in the depression near the deposits. Cultural remains were observed in the dunes, interdunal
Coro‐Punto Fijo Road in an area where runoff water concentrates playas, and ridges.
during the rainy season.
Ridges in these three systems, which are 0.2–0.6 m high and
0.5–1.2 m wide, range from relatively straight to somewhat curved 4.2 | Physicochemical and elemental analysis
(Figure 5). Several angular to perpendicular ridge intersections were
observed between both the WSW‐ENE and the N‐S oriented sys- Physicochemical and elemental analyses were obtained for three
tems. Earthen platforms containing subtle, undulating micro- ridges (Figure 5; Table 1). Textures ranged from sandy clay at the
topography were also recorded (Figure 3d). surface to sandy clay loam at depth. Sediment color also varied with
Field observations indicated that the ridges consisted of sand, depth (Figure 4). Clay content is higher at the surface than at depth.
silt, and clay deposits resting on sand (Figure 4). Although at first Clay content is higher at the surface than at depth. Charcoal ap-
appearing homogenous, sand percentages increase and color chan- pearing to be charred woody fragments was found at all depths
ges from darker to lighter with depth (Figure 4), and traces of mostly tending to be larger (ca. 1 mm) near the surface, to very small
planar cross‐bedding were observed (Figure 4). Within the ridges are (<0.5 mm) at depth. Charcoal was absent in the surrounding aeolian
high concentrations of small freshwater snails of genera Pomacea sp. deposit control samples. Mean values of SOC and SOM as well as
and Marisa sp. Carbonate salts were also observed in the profiles. those of total nitrogen (Nt) and available phosphorus (Pa) decreased
Grain size and color, and freshwater snails, were similar in sediments with depth in all ridges consistent with decreasing clay content
BEZADA ET AL.
| 9

(Table 1). Also, mean SOM, Nt, and Pa values in the ridges are higher and were constructed by hierarchical chiefdoms (Gassón, 2002;
than in the eolian control deposits, especially in the finer‐textured, Sánchez et al., 2017). Causeways and other pre‐Hispanic road sys-
upper parts of the ridges. tems elsewhere in northern South America have been similarly re-
Physicochemical data were also obtained on sediments from the lated to complex societies developed after the first millennium AD,
interdunal playa cores (Figure 5; Table 2). Playa sediments were and in many cases were built shortly before the Spaniards’ arrival in
more clayey than the ridges, with higher mean values of SOM, Nt, the New World (Erickson, 2011; López‐Mazz et al., 2016). Hence the
and Pa than the eolian sands, especially at shallow depths. Increases chronology established here for the ridges seems to support the
in SOM were noted in Core 1 at 94–110 cm, in Core 2 at 25–33 cm, natural origin of these structures.
and in Core 3 at 12–21 cm. Converging lines of evidence including ridge orientations aligned
NMDS ordination of ridge and core sediments (Figure 6) was with dunes and playas, location within topographic depressions, si-
based on physicochemical results (Tables 1 and 2), as well as the milar sedimentological composition of ridges and playa sediment
elemental composition obtained by TXRF (Supporting Information cores, preserved cross‐bedding, and presence of shells, suggest that
Material). In general, some elements, such as Ni, Pb, Cu, and K, were the ridges represent inverted channels. These channels likely formed
more abundant in the cores than in the ridges. Some core samples before the estimated 3 m of late Holocene uplift of the Paraguaná
also were enriched in Nt, relative to ridges. However, all samples peninsula (Audemard et al., 1995; and Audemard, 1996). In this area
from cores and ridges were ordered along the same gradient, defined of multiple generations of dunes, two ridge fragments were found
by sand and clay/silt content, SOC, Pa, and Nt content. Hence, NMDS beyond the limits of the active dunes and within a neighboring sta-
analysis failed to distinguish sediments from ridges and cores in our bilized dune field (Figure 5). The ridges are thus regarded as natural
sample. features composed of fine‐grained alluvium overlying older Eolian
sand. Although primary bedforms are preserved in places, much of
the alluvial stratigraphy has been masked by bioturbation and
4.3 | Radiocarbon dates and chronology translocation of clays and organic matter. Interdunal playas in a
former dune field likely functioned as molds that filled with sedi-
Radiocarbon dates were obtained from Ridge 3 (Table 3). The two ments where runoff and surface groundwater collected. High clay
upper strata date approximately 3000–2900 cal. yr. BP (ca. 900 BC). content in the upper part of the ridges likely reflects ponding of
The age obtained from the deepest stratum, around 1900 cal. yr BP water within the channel. This standing water would have captured
(ca. AD 100), is inconsistent with the stratigraphy and is attributed to wind‐transported sands generating a bimodal sand and clay content
translocation of older organic matter to near‐surface sediments, or like that observed today in the topsoil of interdunal playas. Upon
younger material to greater depths by eolian and/or slope processes, dune reactivation sometime after ca. 1000 BC, the barchans buried
and possibly groundwater. The δ C values from the upper strata
13
the ridge sediments, contributing to their compaction and pre-
suggest that C3 plants are the main contributors of SOM (Griffiths, servation. As the active barchans migrated, several paleochannel
1993). In contrast, the δ13C signature of the deepest stratum is segments were exhumed and adjacent, less cohesive deposits were
considerably higher, indicating a mix of organic matter coming from
C3 and C4 (CAM?) plants in the SOM. Today, the area is sparsely
vegetated, with only some annual grasses, likely C4, present on the
flanks of the active dunes, as well as in the interdunal playas.

5 | DI SCUSSION

The physicochemical characteristics and elemental composition of

the ridges in relation to the interdunal playas indicate more simila-
rities than differences (Tables 1 and 2, Figure 6), suggesting that both
sedimentary features have a similar origin. No clear trace of human
influence was seen in the analyses. For example, ridge P and N values
were relative to previously studied anthropogenic soils in the MCNP
(Zavala et al. (2018) and sediments in human‐constructed earthen
structures in the region (Gason 1997).
Sediments in Ridge 3 appear to date to ca. 1000 BC (Table 3), F I G U R E 6 Nonmetric multidimensional scaling (NMDS) of
samples taken in the ridges, cores, and control samples, with
and therefore, predate the Caquetios ethnolinguistic group in
respect to the physicochemical variables and elemental composition
northern Falcon State whose earliest settlements date to ca. AD Supplementary material: Selected elements found by TXRF in the
700–800 (Oliver, 1989) and the earliest documented causeway cores, ridges, and control samples studied (ppm). ND means under
systems in the Venezuelan llanos which date to around AD 500–700 the limit of detection. TXRF, Total reflection x‐ray fluorescence
10 | BEZADA ET AL.

TABLE 3 Radiocarbon ages, 2‐sigma calibrated dates, and d13C of bulk samples from ridge 3
14
Lab number Relative position Depth (cm) C age error d13C(°/oo) cal. yr. BP (2 s)

Gda‐4438 Top 0–18 2837 35 −32.3 2943 (2860–3058)

Gda‐4439 Middle 18–40 2869 33 −27.4 2990 (2877–3137)

Gda‐4437 Lower 40–60 1960 30 −18.6 1911 (1830–1988)

removed by deflation. The end result is an inverted topography with 2013; Mirino et al., 2018; Pain et al., 2007; Williams et al., 2007;
the paleochannels exposed in positive relief. The slight central de- Williams et al., 2009; Zaki et al., 2018).
pression along the axis of the channels that has been attributed to
human traffic (D′Angelo‐Pietri, 1981; Hernández‐Baño, 1984) could
be due to alluvial textural differences and differential compaction. 6 | CO N C L U S I O N S
With respect to climatic conditions conducive for alternating
channelized flow and dune activity, it is noted that around 3000 The composition of earth ridges at MCNP indicates they are not
years ago the region is believed to have experienced enhanced significantly different from the sediments deposited in nearby in-
precipitation variability, perhaps in response to increased El Niño terdunal playas. Moreover, the ridges are connected and form a
activity (Haugh et al., 2001). Hence, the formation of the pa- larger network, part of which underlies active barchans dunes.
14
leochannels may have been related to increased climate instability. Three C dates on ridge deposits suggest that they date ca.
The presence of charcoal within ridge deposits may be viewed by 2000–3000 years BP, a time predating Caquetíos occupation. An
some as evidence in support of human construction. However, the origin as paleochannels first buried by eolian deposits and then
charcoal pieces found in this study were considerably smaller than later exhumed by reactivation and migration of the current barchan
those found in hearths and other cultural contexts at nearby ar- system is proposed.
chaeological sites (e.g., Zavala et al., 2018). It is more likely that the The absence of artifacts within the ridges stands in stark con-
charcoal in the ridge deposits is redeposited from elsewhere. Wild- trast to unequivocal anthropogenic causeways identified elsewhere
fires along the coasts of Coro and Paraguaná tend to be limited by in Venezuela, for example, the Orinoco llanos. Charcoal embedded
low and discontinuous vegetation cover, and it is, therefore, unlikely within the ridges is attributed to natural fluvial transport possibly
that the ridge charcoal was derived from natural burning. It is also abetted by wind and ultimately derived from human burning activ-
unlikely that the charcoal originated far from the study site, since ities in the area.
macrocharcoal particles (i.e., those longer than 500 µm) are com- In summary, the size, configuration, composition, and appar-
monly locally sourced and used as indicators of in situ fire regimes ent age of the earth ridges at MCNP are here found to be
(Withlock & Larsen, 2001). Hence, we believe that the occurrence of most compatible with a natural origin as exhumed paleochannels.
charcoal embedded within the ridges is best explained by the pre- Past human occupants may, however, have used the features as
sence of human activities (burning) in the study area around causeways.
2000–3000 years ago. Thus, the charcoal was not emplaced in the
ridges by human agency but rather it was naturally transported from ACKNOWLEDGM E NTS
nearby areas of anthropogenic burning. The authors are grateful to the National Parks Institute (IN-
Our conclusion that the earth ridges at MCNP are natural does PARQUES) for a permit to study the sites, especially Lic. Gradilia
not downplay the need for their conservation. They are of potential Flores, and the Cultural Patrimony Institute (IPC). The research was
archaeological significance in that they appear to have been used as partially funded by the Council for Science and Technology of Ve-
preferred pathways by local inhabitants as indicated by the abun- nezuela (FONACIT), grant PEII N°201300280 and PE201203. We
dance of cultural remains along the ridges. Moreover, they are im- thank the International Atomic Energy Agency (IAEA) National
portant for understanding natural history and geomorphic change in Project VEN/7/005, the Gliwice Absolute Dating Methods Centre
a tectonically dynamic coastal setting. The ridges may provide clues (GADAM) in Poland, and the Centro Nacional de Aceleradores (CNA)
14
regarding local faulting and consequent changes in drainage evolu- in Spain for their support for the C dating. Prof. Nathalie Suarez
tion. Ridge deposits may hold paleoenvironmental information re- gave us advice regarding the interpretation of d13C in arid/semiarid
lated to changing precipitation regimes. More broadly, there has environments. José Oliver provided advice on an early version of this
been little reported research on exhumed paleochannels (Williams manuscript. The final version was improved by suggestions from
et al., 2007) although the identification of comparable features on Gary Huckleberry and another anonymous reviewer. Thanks to Ro-
Mars and their apparent importance for understanding the role of siris Guzman for Figures 3 and 4. The English language was improved
water in the early history of the planet has more recently en- by Prof. Max Bezada Jr. and Prof. Harvey Thorleifson of the
couraged their study on Earth (Clarke & Stocker, 2011; DiBiase et al., University of Minnesota.
BEZADA ET AL.
| 11

DATA AV AILA BILITY STATEMENT evidence for a standing body of water on the northern plains of
The data that support the findings of this study are available from Mars. Journal of Geophysical Research, 118(6), 1285–1302.
Erickson, C. (2000). An artificial landscape‐scale fishery in the Bolivian
the corresponding author upon reasonable request.
Amazon. Nature, 408, 190–193.
Erickson, C. (2011). Agency, causeways, canals, and the landscapes of
ORCID everyday life in the Bolivian Amazon. In (Eds.) Snead, J., Erickson, C.
Alejandra Leal https://orcid.org/0000-0002-3739-0757 & Darling, A., Lansdcapes of movements: trails, paths and roads in
anthropological perspective (pp. 204–231). Philadelphia: Penn
Museum Press and the University of Pennsylvania Press.
R EF E RE N C E S Foghin, S. (2002). Tiempo y clima en Venezuela: aproximación a una
Antczak, A., Urbani, B., & Antczak, M. M. (2017). Re‐thinking the geografía climática del territorio venezolano, Número 1 de
migration of Caribean‐speakers from the Middle Orinoco river to Colección Clase magistral. Universidad Pedagógica Experimental
North‐Central Venezuela (AD 800). Journal of World Prehistory, 30, Libertador, Caracas, 159 p.
131–175. Friend, P. F., Marzo, M., Nijimsn, W., & Puigdefabregas, C. (1981). Fluvial
Arcaya, P. M. (1920). Historia del Estado Falcón (1st edition.). Caracas: sedimentology in the Tertiary South Pyrenean and Ebro basins,
Tipografía Cosmos. Spain. In: Elliott, T. (Editor), Field guides to modern and ancient
Audemard, F. (1996). Late Quaternary marine deposits of the Paraguana fluvial systems in Britain and Spain: International Fluvial
peninsula, State of Falcon, northwestern Venezuela: Preliminary Conference, University of Keele. 4.1‐4.50.
geological observations and neotectonic implications. Quaternary Gassón, R. (1997). Prehispanic Intensive Agriculture, Settlement Pattern and
International, 31, 5–11. Political Economy in the Western Venezuelan Llanos. Unpublished
Audemard, F., Rodríguez, J., & Bousquet, J. (1995) Holocene tectonic doctoral dissertation, University of Pittsburgh, Pittsburgh.
uplift of La Vela anticline related to the activity of the Guadalupe Gassón, R. (2002). Orinoquia: the archaeology of the Orinoco River Basin.
thrust, northern Falcon State (Venezuela). In: Meco J, Petit‐Maire N Journal of World Prehistory, 16, 237–311.
(Ed). Climates of the Past (Pp.13‐27). Proceedings IV CLIP Meeting. Giegengack, R. (1968). Late Pleistocene History of the Nile Valley in
Servicio de Publicaciones de la Universidad de Las Palmas de Gran Egyptian Nubia. Ph.D. Thesis, Yale University.
Canaria, Giegengack, R., & Zaki, A. (2017). Inverted topographic features, now
Ball, D. (1964). Loss‐on‐ignition as an estimate of organic matter and submerged beneath the water of Lake Nasser, document a
organic carbon in non‐calcareous soils. Journal of Soil Science, 15(1), morphostratigraphic sequence of high‐amplitude late‐Pleistocene
84–92. climate oscillation in Egyptian Nubia. Journal of African Earth Science,
Bracco, R., del Puerto, L., Inda, H., Panario, D., Castiñeira, C., & García‐ 36, 176–187.
Rodríguez, F. (2011). The relationship between emergence of Griffiths, H. (1993). Carbon isotope discrimination. In (Eds.) Hall, D. O.,
mound builders in SE Uruguay and climate change inferred from Scurlock, J. M. O., Bolhar‐Nordenkampf, H. R., Leegood, R. C. &
opal phytolith records. Quaternary International, 245(1), 62–73. Long, S. P., Photosynthesis and production in a changing environment: a
Busby, C., Andrews, G., Koerner, A., Brown, S., Melosh, B., & Hagan, J. field and laboratory manual (pp. 181–192). London: Chapman & Hall.
(2016). Progressive derangement of ancient (Mesozoic) east‐west Haugh, G., Hughen, K., Sigman, D., Peterson, L., & Röhl, U. (2001).
Nevada plano paleochannels into modern (Miocene–Holocene) Southward migration of the Intertropical Convergence Zone
north‐northwest trends in the Walker Lane Belt, central Sierra through the Holocene. Science, 293, 1304–1308.
Nevada. Geosphere, 12(1), 135–175. Henningsen, P., Kandel, A., & Conard, N. (2003). The significance of
Butt, C., & Bristow, A. (2013). Relief inversion in the geomorphological calcretes and paleosols on ancient dunes of the Western Cape, South
evolution of sub‐Saharan West Africa. Geomorphology, 185, 16–26. Africa, as stratigraphic markers and paleoenvironmental indicators. In G.
Camacho, R., Salazar, S., González, L., Pacheco, H., & Suárez, C. (2011). Füleky (Ed.), Papers of the 1st International Conference on
Caracterización geomorfológica de las dunas longitudinales del Archaeology and Soils (.45‐52). BAR International S1163.
Istmo de Médanos, estado Falcón, Venezuela. Investigaciones Hernández‐Baño, A. (1984). Los Caquetíos de Falcón (1st edition.). Coro:
Geográficas, 76, 7–19. INCUDEF.
Cey, G. (1995). Viaje y descripción de las Indias (1539‐1553), Colección V Hernández‐Baño, A., & Trujillo, M. (1986). Secretos de los Médanos de Coro.
Centenario del Encuentro entre dos Mundos. Caracas: Fundación Coro: Instituto Nacional de Parques. Instituto de Cultura del Estado
Banco Venezolano de Crédito. Falcón.
Clarke, J., & Stocker, C. (2011). Concretions in exhumed and inverted Kemp, J., & Rhodes, E. (2010). Episodic fluvial activity of inland rivers in
channels near Hanksville Utah: implications for Mars. International southeastern Australia: Palaeochannel systems and terraces of the
Journal of Astrobiology, 10(3), 161–175. Lachlan River. Quaternary Science Reviews, 29(5), 732–752.
Conedera, M., Tinner, W., Neff, C., Meurer, M., Dickens, A., & Krebs, P. Lindgren, W. (1911) The Tertiary gravels of the Sierra Nevada of
(2009). Reconstructing past fire regimes: methods, applications, and California: Washington, D.C., U.S. Geological Survey, 222 p. (quoted
relevance to fire management and conservation. Quaternary Science by Busby et al 2016).
Reviews, 28, 555–576. López‐Mazz, J., Ronstain, S., & McKey, D. (2016). Cerritos, tolas, tesos,
Cruxent, J. M., & Rouse, I. (1958). ‐1959An Archaeological Chronology of camellones y otros montículos de las tierras bajas de Sudamérica.
Venezuela. Washington: Pan American Union. Revista de Arqueología, 29(1), 86–113.
Cuevas, M., Cabera, J. L., Marcuello, L., Cazo, A., Mazro, P., & Bellmunt, M. Matteucci, S. (1987). The vegetation of Falcon State. Vegetatio, 70, 67–91.
(2010). Exhumed channel sandstone networks within fluvial fan Matteucci, S., & Colma, A. (1986). Caracterización climática del Estado
deposits from the Oligo‐Miocene caspe formation, south‐east Ebro Falcón. Acta científica venezolana, 31, 63–71.
basin (north‐east Spain). Sedimentology, 57(1), 162–189. Mirino, M., Matthew, B., Fawdon, P., Davis, J., Barnes, R., Allender, E., &
D′Angelo‐Pietri, L. (1981). Características geológicas de unas estructuras en Gridrop, P. (2018). Field study of exhumed channels in Green River
forma de calzadas en los médanos de coro, Coro: Universidad Nacional and Hanksville areas, Utah, and a comparison with inverted channel
Experimental Francisco de Miranda, (UNEFM). features on Mars. 20th EGU General Assembly, EGU2018,
DiBiase, R. A., Limaye, A. B., Scheingross, J. S., Fischer, W. W., & Proceedings from the conference held 4‐13 April, 2018, Vienna,
Lamb, M. P. (2013). Deltaic deposits at Aeolis Dorsa: Sedimentary Austria, p.16047
12 | BEZADA ET AL.

Morón, C. (2011). Presencias bajo la arena: yacimientos arqueológicos en Williams, R., Chidsey, T., Jr., & Eby, D. (2007). Exhumed paleochannels in
el Parque Nacional Médanos de Coro. Boletín Antropológico, 29(81), central Utah – analogs for raised curvilinear features on Mars. In
7–26. (eds) Willis, G. C., Hylland, M. D., Clark, D. L. & Chidsey, T. C., Central
Oliver, J. (1989). The archaeological, linguistic and ethnohistorical evidence Utah‐diverse geology of a dynamic landscape (pp. 220–235). Salt Lake
for the expansion of arawakan into northwestern Venezuela and City: Utah Geological Association Publication 36.
northeastern Colombia. Ph.D. Thesis, University of Illinois. Williams, R., Irwin III, J., & Zimbelman, J. (2009). Evaluation of
Pain, C., Clarke, J., & Thomas, M. (2007). Inversion of relief on Mars. paleohydrologic models for terrestrial inverted channels: Implications
Icarus, 190, 478–491. for application to martian sinuous ridges (2009). Geomorphology, 107(3‐
Pain, C., & Oilier, C. (1995). Inversion of relief a component of landscape 4), 300–315.
evolution. Geomorphology, 12(2), 151–16. Withlock, C., & Larsen, C. (2001). 5. Charcoal as a fire proxy. In (Eds.)
Pansu, M., & Gautheyrou, J. (2006). Handbook of Soil Analysis. Mineralogical, Smol, J. P., Birks, H. J. B. & Last, W. M., Tracking Environmental
Organic and Inorganic Methods (1st edition.). Amsterdam: Springer‐Verlag. Change Using Lake Sediments: Terrestrial, Algal, and Siliceous Indicators
Patterson, W., Edwards, K., & Maguire, D. (1987). Microscopic charcoal as (3, pp. 1–23). Kluwer Academic.
a fossil indicator of fire. Quaternary Science Reviews, 6, 3–23. Zaki, A., & Giegengack, R. (2016). Inverted topography in the
Redmond, E., & Spencer, C. (1990). Investigaciones arqueológicas en el southeastern part of the Western Desert of Egypt. Journal of
piedemonte andino y los llanos altos de Barinas. Venezuela. Boletín de African Earth Sciences, 121, 56–61.
la Asociación Venezolana de Arqueología, 5, 4–24. Zaki, A., Giegengack, R., & Castelltort, S. (2020). Inverted channels in the
Ritter, D. F. (1967). Terrace development along the front of the Beartooth Eastern Sahara—distribution, formation, and interpretation to
Mountains, southern Montana. Geological Society of America Bulletin, enable reconstruction of paleodrainage networks. In (eds) Herget,
78, 467–484. J. & Fontana, A., Palaeohydrology. Geography of the Physical
Spencer, C., & Redmond, E. (1992). Prehispanic Chiefdoms of the Western Environment. Cham: Springer.
Venezuelan Llanos. World Archaeology, 24, 134–157. Zaki, A., Pain, C., Edgett, C., & Giegengack, R. (2018). Inverted stream
Spencer, C., & Redmond, E. (1983). Archaeological investigations in the channels in the Western Desert of Egypt: Synergistic remote, field
Andean Piedmont and High Llanos of Western Venezuela: A observations and laboratory analysis on Earth with applications to
preliminary report. In (Eds.) Kvietok, P. & Sandweiss, D., Recent Mars. Icarus, 309, 105–124.
Studies in Andean Prehistory and Protohistory (pp. 137–157). Ithaca: Zavala, M., Pedraza, E., Zambrano, E., Barros, H., Herrera, R., Morón, C., &
Cornell Latin American Studies Program. Leal, A. (2018). Contribución a la arqueología del Parque Nacional
Strauss, R. (1992). El tiempo prehispánico de Venezuela. Caracas: Fundación Médanos de Coro (PNMC), Venezuela: Implicaciones para los
Eugenio Mendoza. patrones de ocupación Caquetía, Interciencia 43(6), 400–408.
Stuiver, M., Reimer, P. J., & Reimer, R. W. (2018). CALIB 7.1 [WWW
program] at http://calib.org, Accessed March 27, 2018.
Sánchez, F., Fernández, J., Gassón, R., Bezada, M., & Leal, A. (2017). SU PP O RT I N G I N FO RM ATI ON
Paleoecología y ocupación humana durante el Holoceno en los Additional Supporting Information may be found online in the
Llanos del Orinoco: una revisión y nuevos datos. BioLlania, 15,
supporting information tab for this article.
297–333.
Wahrhaftig, C. (1965). Stepped topography of the southern Sierra
Nevada. Geological Society of America Bulletin, 76, 1165–1190.
Wang, X., Wang, J., & Zhand, J. (2012). Comparison of three methods for
How to cite this article: Bezada M, Barros H, Zavala M, et al.
organic and inorganic carbon in calcareous soils of northwestern
China. PLoS One, 7(8), 1–6. Earthen ridges in coastal sand dunes of the Caribbean coast,
Whitney, J. D. (1880). The auriferous gravels of the Sierra Nevada of Venezuela: anthropogenic or natural? Geoarchaeology.
California. Harvard University. Museum of Comparative Zoology 2020;1–12. https://doi.org/10.1002/gea.21835
Memoir (6), 1(quoted by Giegengack. 1968).

View publication stats