Document [original]

Climate variability in a 3.8 Ma old sedimentary record from the hyperarid

Atacama Desert

Felix L. Arens

a,*

, Jenny Feige

a,b

, Alessandro Airo

a,b

, Christof Sager

a,b

, Lutz Hecht

Lucas Horstmann

c,d

, Felix E.D. Kaufmann

, Johannes Lachner

e,f

, Thomas Neumann

Norbert Nowaczyk

, Ferry Schiperski

, Peter Steier

, Alexandra Stoll

h,i

, Ulrich Struck

Bernardita Valenzuela

, Friedhelm von Blanckenburg

k,l

, Hella Wittmann

, Lukas Wacker

Dirk Wagner

c,n

, Pedro Zamorano

, Dirk Schulze-Makuch

a,c,p

Technische Universit¨

at Berlin, Zentrum für Astronomie und Astrophysik, 10623 Berlin, Germany

Museum für Naturkunde, Leibniz-Institut für Evolutions- und Biodiversit¨

atsforschung, 10115 Berlin, Germany

GFZ German Research Centre for Geosciences, Section Geomicrobiology, 14473 Potsdam, Germany

Department Experimental Phycology and Culture Collection of Algae (EPSAG), Albrecht-von-Haller-Institute for Plant Sciences, Georg August University G¨

ottingen,

37073 G¨

ottingen, Germany

University of Vienna, Faculty of Physics, Isotope Physics, 1090 Vienna, Austria

Helmholtz-Zentrum Dresden-Rossendorf (HZDR), Institute of Ion Beam Physics and Materials Research, 01328 Dresden, Germany

Technische Universit¨

at Berlin, Institut für Angewandte Geowissenschaften, Fachgebiet Angewandte Geochemie, 10623 Berlin, Germany

Laboratorio de Microbiología Aplicada, Centro de Estudios Avanzados en Zonas ´

Aridas, La Serena 1720256, Chile,

Instituto de Investigaci´

on Multidisciplinario en Ciencia y Tecnología, Universidad de La Serena, La Serena 1720256, Chile

Laboratorio de Microorganismos Extrem´

ofilos, Instituto Antofagasta, Universidad de Antofagasta, Antofagasta 1240000, Chile,

GFZ German Research Centre for Geosciences, Section Earth Surface Geochemistry, Telegrafenberg, 14473 Potsdam, Germany

Fachbereich Geowissenschaften, Freie Universit¨

at Berlin, 12249 Berlin, Germany

Ion Beam Physics, ETH Zürich, 8093 Zürich, Switzerland

University of Potsdam, Institute of Geosciences, 14476 Potsdam, Germany

Departamento Biom´

edico, Facultad de Ciencias de la Salud, Universidad de Antofagasta; Antofagasta 1240000, Chile

Department of Plankton and Microbial Ecology, Leibniz Institute of Freshwater Ecology and Inland Fisheries, 16775 Stechlin, Germany

GFZ German Research Centre for Geosciences, Section Paleoclimate and Landscape Evolution, 14473 Potsdam, Germany

ARTICLE INFO

Editor: Prof. Liviu Matenco

Keywords:

Hyperaridity

Sedimentology

Pedogenic processes

Paleoclimate

Atacama Desert

ABSTRACT

The hyperarid Atacama Desert is one of the driest and oldest deserts on Earth, rendering it a valuable climate

archive. However, unraveling its past climate is particularly challenging and the few studied paleoclimate re-

cords of the region reveal strong temporal and spatial variabilities. To enhance our understanding of these dy-

namics we investigated a sedimentary record in the Yungay valley located in the southern hyperarid Atacama

Desert. We employed paleomagnetic and radiocarbon dating, and for the first time for Atacama Desert sediments,

a meteoric

Be/

Be based method for determining the depositional age. The respective 4.20 m deep profile

comprises a lower alluvial fan deposit with a maximum age of 3.8 ±0.8 Ma, and an upper 1.84 m thick clay pan

deposit that has accumulated over the last 19 ka. Different proxies including grain size, salt concentration, and

elemental composition indicate an aridity increase around 2.3 Ma ago and repeated dry and wet phases during

the late Pleistocene and the Holocene. The latter climatic shifts can be assigned to variabilities of the South

American Summer Monsoon and El Ni˜

no Southern Oscillation with moisture sources from the Atlantic and the

Pacific Ocean, respectively. This study provides deeper insights into the heterogeneous climate of the hyperarid

Atacama Desert and underlines the importance of interdisciplinary investigations to decipher climate systems

and their effect on potential habitable regions in such an extreme environment.

* Corresponding author.

E-mail address: [email protected] (F.L. Arens).

Contents lists available at ScienceDirect

Global and Planetary Change

journal homepage: www.elsevier.com/locate/gloplacha

https://doi.org/10.1016/j.gloplacha.2024.104576

Received 9 November 2023; Received in revised form 13 August 2024; Accepted 5 September 2024

Global and Planetary Change 242 (2024) 104576

Available online 7 September 2024

1. Introduction

The temperate Atacama Desert, Chile, is one of the most arid deserts

on Earth and has persisted since the late Triassic (Clarke, 2006). The

onset of hyperaridity occurred between the late Oligocene and early

Miocene (e.g., Dunai et al., 2005;Evenstar et al., 2017). Furthermore,

evidence has been found that since then multiple periods of increased

precipitation interrupted the hyperaridity (e.g., Jordan et al., 2014;

Ritter et al., 2018a). However, these punctuations occurred rather

locally (Ritter et al., 2018b). Despite the overall hyperarid conditions,

very rare but heavy rain events can cause flash floods and result in the

temporary flooding of closed drainage systems, creating clay pans (a.k.a.

playas) (Amundson et al., 2012;Pfeiffer et al., 2018;Ritter et al., 2018a,

2019;Diederich et al., 2020;Wennrich et al., 2024) (Fig. 1). There are

numerous basins located in the Coastal Cordillera that incorporate clay

pans (10

–10

m in diameter), with deposition records spanning

10s–100s ka and high sedimentation rates of 10s of m Ma

−1

, compared

to the millions of years old surrounding alluvial fans with much slower

sedimentation rates of ~1 m Ma

−1

(Jungers et al., 2013;Ritter et al.,

2019;Diederich et al., 2020;Wennrich et al., 2024). Moreover, their

Fig. 1. Overview maps of the study area. A) South America with arid and hyperarid climate zones (Houston and Hartley, 2003) and the main moisture sources (blue

arrows) relevant for the Atacama Desert: tropical continental (TrC), temperate continental (TeC), tropical maritime (TrM), temperate maritime (TeM) (Houston and

Latorre, 2022). Black rectangle shows zoom-in of B. B) The topographic map of the Atacama Desert with the mean annual precipitation (mm a

−1

) is shown as isohyets

(white lines) after Ritter et al. (2019) and the border of the summer- and winter-precipitation dominated areas (white dashed line) after Houston (2006b). Location of

recent Huidobria fruticosa shrubs (green rectangle) and location of paleo records discussed in this study (dots) (1. Diederich et al., 2020; 2. Ritter et al., 2018b; 3.

Ritter et al., 2019; 4. Wang et al., 2015; 5. Gonz´

alez-Pinilla et al., 2021; 6. Vargas et al., 2006; 7. Diaz et al., 2012; 8. Placzek et al., 2014; 9. Amundson et al., 2012;

10. Pfeiffer et al., 2018; 11. Quade et al., 2008; 12. S´

aez et al., 2016; 13. Wennrich et al., 2024; 14. Maldonado et al., 2005) Black rectangle shows zoom-in of C. C)

Landsat-8 satellite image with 200 m contour lines (white lines) of the Yungay valley and the investigated clay pan, here referred to as the ‘Herradura’clay pan (black

rectangle, showing zoom-in of D) with its 188 km

large catchment area (red outline), as well as the adjacent Aguas Blancas salar in the east (white line pattern). The

black dashed line indicates a fault going through the valley (Domagala et al., 2016). D) Drone orthophoto of the 0.1 km

large Herradura clay pan with the main pit

(black arrow) and the four adjacent pits (gray arrows). From the south, recent (2017) run-off channels (blue arrows) have introduced new sediment into the clay pan,

also flooding the road B-55. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)

F.L. Arens et al. Global and Planetary Change 242 (2024) 104576

catchment areas are hydrologically disconnected from the Andes and

therefore hold records of these local environments within the hyperarid

core of the Atacama Desert (Diederich et al., 2020). Hence, they provide

an opportunity to gain a more comprehensive picture of the climatic and

ecological evolution within the Atacama Desert and to understand the

regional variabilities (e.g., Amundson et al., 2012;S´

aez et al., 2016;

Ritter et al., 2019;Diederich et al., 2020;Wennrich et al., 2024)

(Fig. 1B). These paleo environmental studies are also important to gain

insights into the preservation of biological matter and the former

habitability under hyperarid conditions, which can help to understand

the environmental conditions and the habitability of Mars (Davila and

Schulze-Makuch, 2016;Azua-Bustos et al., 2023).

The younger clay pan deposits can be dated by multiple geochro-

nological methods including magnetostratigraphy, radiocarbon, and

optically stimulated luminescence (OSL) dating (Ritter et al., 2019;

Diederich et al., 2020). Dating the millions of years old alluvial fan

deposits, however, is more challenging, since radiocarbon and OSL are

limited to maximal ages of 60 and 350 ka, respectively (e.g., Murray and

Olley, 2002;Bartz et al., 2020). Biostratigraphy is also not applicable in

the hyperarid core of the Atacama Desert due to the lack of microfossils

while magnetostratigraphy is a challenge due to the coarse grain size

and the typically discontinuous sedimentation of alluvial deposits.

Hence, the depositional ages of these Ma-old sediments are in large parts

only constrained by individual marker horizons, such as volcanic ashes

or ignimbrites that can be dated with Ar

–

Ar or U

–

Pb methods (e.g.,

Ewing et al., 2006;Rech et al., 2006).

More compelling are dating methods with cosmogenic radionuclides

such as

Be that is produced from secondary cosmic rays reacting with

O and N (Lal and Peters, 1967). In-situ produced

Be forms in the

minerals themself and yields exposure ages, that e.g., have been used to

constrain the cessation of alluvial activity correlating with the latest

onset of hyperaridity ~2 Ma ago (Amundson et al., 2012;Jungers et al.,

2013). Depositional ages, however, can be directly determined through

meteoric

Be, which is produced in the Earth’s atmosphere (Raisbeck

and Yiou, 1984). Its half-life of 1.387 Ma (Chmeleff et al., 2010;Kor-

schinek et al., 2010) allows for depositional age dating within the range

of 0.2–14 Ma (Bourles et al., 1989). Thus far, only a single study in the

Atacama Desert has used

Be soil concentrations for incremental time-

resolved stratigraphy of alluvial deposits (Wang et al., 2015). However,

the depositional process and sediment type, such as changing sedi-

mentation rates, varying grain sizes and compositions, affect the final

Be concentrations and therefore reduce the accuracy of the deposi-

tional ages (Bourles et al., 1989;Wittmann et al., 2012). This inaccuracy

can be compensated through the measurement of the stable isotope

deriving from bedrock weathering and using

Be/

Be ratios for depo-

sitional dating (e.g., Bourles et al., 1989). This approach relies on the

assumption that

Be and

Be have similar depositional mechanisms

(Wittmann et al., 2012).

In this study, we deployed a unique set of methods across multiple

disciplines, including a novel dating technique of meteoric

Be/

dating for millions of years old soils based on alluvial fans formed under

hyperarid conditions, and

C dating coupled with magnetostratigraphy

for the clay pan deposits. Combined with geochemical, sedimentological

and phytologic analysis on sediment deposits in the Yungay valley

(24◦S, 71◦W; 1010 m above sea level) a more comprehensive paleo re-

cord on the regional climate variability in the Atacama Desert is pro-

vided over a million-years range.

2. Regional setting

Numerous factors collectively contribute to the extreme aridity of the

Atacama Desert (Houston and Hartley, 2003;Houston, 2006b;Garreaud

et al., 2010). Besides its position in the subtropical high-pressure belt,

the distance to the Atlantic Ocean as a main moisture source and the rain

shadow effect that developed with the orogenesis of the Andean

Cordillera increases the aridity (Houston and Hartley, 2003). The cold

Humboldt Current causes further aridification by cooling the above-

lying air and thereby reducing its moisture capacity (Garreaud et al.,

2010). This process causes a temperature inversion and leads to the

formation of persistent stratus cloud cover, which can reach elevations

of up to 1200 m above sea level (a.s.l.). In the coastal region, these

clouds create fog, providing an important water source for the local

vegetation (Houston, 2006b;Cereceda et al., 2008;Latorre et al., 2011).

The Coastal Cordillera (<2000 m a.s.l.) is effectively blocking the fog

from penetrating further inland, except for a few topographical corridors

where it can migrate several kilometers into the desert. The Andean

Cordillera, bordering the Atacama Desert in the east, receives mean

annual precipitation of 300 mm a

−1

at 5000 m a.s.l. which decreases

with lower elevation to <20 mm a

−1

at 2300 m a.s.l. In the hyperarid

core of the Atacama Desert located between 19◦S and 25◦S mean annual

precipitation can be <1 mm a

−1

(Houston and Hartley, 2003). Lastly, the

high evaporation rates amplify the prevailing dry climate conditions

(Houston, 2006a). The main sources of precipitation in the Atacama

Desert are from the Atlantic Ocean and the Pacific Ocean, which are

influenced by the El Ni˜

no Southern Oscillation (ENSO) (Houston,

2006b). The Atlantic moisture occurs as South American Summer

Monsoon from the north-northeast (tropical continental) and from the

southeast (temperate continental) during the summer and especially

during La Ni˜

na events (Fig. 1A). The Pacific moisture occurs as extra-

tropical cyclones from the southwest (temperate maritime) and tropical

precipitation anomalies from west and northwest (tropical maritime)

during the winter and especially during El Ni˜

no (Houston, 2006b;de

Porras et al., 2017). The different proximity of these sources leads to a

divide into two regions of summer-dominated precipitation and winter-

dominated precipitation, where precipitation is minimal at their

boundary (Houston, 2006b;Quade et al., 2008) (Fig. 1B). Records in the

Atacama Desert and adjacent Andean Cordillera have shown that these

precipitation sources have changed within orbital (i.e., Milankovitch

cycles) and millennial timescales throughout the late Pleistocene and

Holocene (e.g., Placzek et al., 2009;S´

aez et al., 2016;Ritter et al., 2019).

This dynamic also influenced the position of the boundary (Houston and

Latorre, 2022).

The dominating geomorphological features of the Atacama Desert, i.

e., alluvial fans and stream incisions, are mostly formed during more

moist, but still arid periods millions of years ago (Amundson et al., 2012;

Jordan et al., 2014;Ritter et al., 2018b). In contrast, during the recent

stage of hyperaridity depositional processes are reduced around two

orders of magnitude with overall sedimentation rates of 0.5–5 m Ma

−1

(Davis et al., 2014;Sanchez et al., 2021). Moreover, seismic activity

became relevant for the net sediment transport on sloped surfaces (Sager

et al., 2020), and atmospheric dust and salt accumulation dominated,

leading to a soil volume expansion typical for the Atacama Desert

(Ericksen, 1981;Michalski et al., 2004;Ewing et al., 2006).

The saline fraction of atmospheric input with a current depositional

rate of ~4 g m

−2

−1

is largely not washed away by flash floods after

minor rain events but rather infiltrates differentially into the soil (Ewing

et al., 2006;Owen et al., 2013). The most soluble components (e.g.,

chlorides or nitrates) can migrate downward up to a few meters,

whereas the shallow subsurface is dominated by sulfates (Ewing et al.,

2006;Ewing et al., 2008;Arens et al., 2021). These soils show a typical

stratification beginning with the surficial ~30 cm thick gypsic horizon

(locally termed chusca) composed of surface clasts lying on loose sand-

rich sediment with embedded losas (local name for palm-sized, sphe-

roidal aggregates) underlain by the vesicular layer (porous and fragile

horizon), both highly concentrated in sulfates. Below lies a multiple

meters thick gravelly petrogypsic horizon, firmly cemented by sulfates

(costra), followed by a petrosalic horizon (caliche) recognizable by its

high chloride and nitrate content (Ericksen, 1981;Finstad et al., 2014).

2.1. Study area

The investigated clay pan (24.095◦S, 70.028◦W) is here referred to as

F.L. Arens et al. Global and Planetary Change 242 (2024) 104576

Herradura clay pan named after the adjacent mountain in the north

(Fig. 1, S1). The ~0.1 km

large clay pan and its current catchment area

of 188 km

is located in the eastern Coastal Cordillera, 60 km southeast

of Antofagasta (Fig. 1B). The surrounding mountains are built up by

igneous bedrock, reaching over 2300 m a.s.l. and their slopes are

covered by alluvial fans which have accumulated atmospheric dust and

salts (Amundson et al., 2012;Domagala et al., 2016). The Herradura

clay pan is situated at an elevation of 1010 m a.s.l. and is part of in an

endorheic basin with its spill point leading into the Aguas Blancas salar

(960 m a.s.l.) located ~8 km further east (Fig. 1C). The sedimentary

profile was sampled down to 4.20 m in a preexisting pit located in the

center of the clay pan. This pit is surrounded by four ~1 m deep pits at

50 m distance, which display the same stratigraphy supporting that the

investigated profile represents the overall buildup of the clay pan

(Fig. 1D).

3. Methods

3.1. Sampling

The profile samples were taken after laterally excavating the pit by

more than 50 cm for collecting pristine samples. In total three different

sample sets were taken in two field campaigns in 2018 and 2019: Sample

set A was taken in 10 cm increments and a vertical thickness of roughly

5 cm from the surface down to bottom of the pit. Sample set B was taken

in a similar manner but regarding the identified sedimentary horizons

within the upper 185 cm. Both sets were stored in polyethylene sample

bags. Sample set C was collected with cubic plastic boxes (internal size

2.0 ×2.0 ×1.6 cm) at a standard distance (center to center) of 2.5–3.0

cm for the upper 187 cm. For each undisturbed sample cube of set C

azimuth, inclination, and rotation were determined with a geological

compass. This sampling technique was not applicable for sediments

below 187 cm being coarse-grained and salt cemented. All sets were

stored at room temperature.

3.2. pH and electrical conductivity (EC)

The pH and EC measurements were conducted in the Astrobiology

laboratory of the Technische Universit¨

at Berlin. Subsamples (5 g) from

sample set A were dried at 60 ◦C and sieved dry to remove the >2 mm

grain size fraction before further analysis. Ultra-pure water was added in

1:2.5 and 1:5 (w/w) ratio to the processed subsamples, respectively, and

shaken for 60 min before measuring the pH-value (pH meter 691,

Metrohm) and EC (GMH 3400, Greisinger).

3.3. Grain size distribution

The grain size analysis was conducted at the Department of Applied

Geochemistry of the Technische Universit¨

at Berlin. The non-soluble

grain fraction of subsamples from sample set A and B (100 g) was wet

sieved with distilled water employing 5 mesh sizes: <0.063 mm (clay

and silt); 0.063–0.25 mm (fine sand); 0.25–0.63 mm (medium sand);

0.63–2 mm (coarse sand); >2 mm (pebbles and cobbles). Subsequently,

the <0.063 mm fractions were further suspended in water and separated

into clay and silt fraction by centrifugation (Heraeus Megafuge 1.0) with

settling distance of 10 cm and relative centrifugal force of 210 xgfor 3

min. Mean grain size, standard deviation, skewness, and kurtosis were

calculated with GRADISTAT (Blott and Pye, 2001).

3.4. Mineralogy

For the bulk mineralogy, subsamples from sample set A and B were

crushed, homogenized and further ground to powder. X-ray diffraction

(XRD) analysis was performed by using a D2 Phaser XRD device (Bruker)

at the Department of Applied Geochemistry of the Technische Uni-

versit¨

at Berlin. The X-ray source is a CuK

anode operating at 30 kV and

10 mA. A step interval of 0.1◦(2Θ) with a step-counting time of 1 s was

used in a scanning range from 3◦to 80◦(2Θ). For clay mineral identi-

fication smear slides of the clay fractions were measured in air-dried,

ethylene glycol saturated, and 550 ◦C preheated configuration. A

semi-quantitative evaluation of the results was conducted with the

software package “Diffrac.eva”(Bruker) and the “Powder Diffraction

File Minerals 2019”(File and others, 2000).

3.5. Ion chromatography (IC)

Cation and anion analysis was performed on a SYKAM Compact IC

system (Sykam Chromatographie, Fürstenfeldbruck, Germany) at the

GFZ German Research Centre for Geosciences, Potsdam. Sample prep-

aration was done according to Genderjahn et al. (2018) with subsamples

of set A dried at 50 ◦C for 12 h and sieved to <2 mm grain size. The

leaching was performed in a 1:5 ratio of 5 g soil to 25 mL Milli-Q water

and diluted according to their electric conductivity prior to ion chro-

matographic measurement. Standards for the detection of cations (Roth,

Multi-Element IC Standard Solution, covering lithium, sodium, ammo-

nium, potassium, magnesium, and calcium) and anions (Sykam,

covering fluoride, chloride, nitrite, bromide, nitrate, phosphate, and

sulfate) were used. The samples were run in triplicates to estimate the

error based on the standard deviation.

3.6. Micro X-ray fluorescence (

-XRF)

A vertical profile consisting of single point

-XRF analyzes every ca.

3 mm was obtained on sample set C with a Bruker Tornado Plus spec-

trometer at the Museum für Naturkunde Berlin. Single points were

measured with an Rh X-ray tube set at 50 kV and 600

A with an X-ray

spot size of 160

m and a measuring time of 60 s. 10 measurements per

sample were averaged to obtain a ~1 cm resolution.

3.7. Concentration and stable isotope analysis of carbon and nitrogen

Subsamples from set A were dried at 60 ◦C and ground to powder

before applying the isotope analysis.

3.7.1. TC, TN, δ

, and δ

Stable isotope analysis and concentration measurements of total ni-

trogen (TN) and total carbon (TC) was performed simultaneously with a

THERMO/Finnigan MAT V isotope ratio mass spectrometer, coupled to a

THERMO Flash EA 1112 elemental analyzer via a THERMO/Finnigan

Conflo IV- interface in the stable isotope laboratory of the Museum für

Naturkunde Berlin. Stable isotope ratios are expressed in the conven-

tional delta notation (δ

C, δ

N) relative to atmospheric nitrogen

(Mariotti, 1983) and Vienna PeeDee Belemnite standard (VPDB). The

standard deviation for repeated measurements of lab standard material

(peptone) is generally better than 0.15 ‰for δ

C and δ

N. Standard

deviations of concentration measurements of replicates of our lab

standard are <3 % of the concentration analyzed.

3.7.2. δ

carbonate

and δ

carbonate

For oxygen and carbon isotope measurements of carbonates

approximately 100–400

g of sample material was put into a clean 10

mL exetainer. After sealing the exetainer with a septum cap (caps and

septa for LABCO exetainer 438b) the remaining air was removed by

flushing the exetainer with He for 6 min at a flow of 100 mL per minute.

After flushing, approximately 30

L of anhydrous phosphoric acid was

injected through the septum into the sealed exetainer by using a

disposable syringe. Approximately after 1.5 h of reaction time at 50 ◦C

the sample was ready for isotope measurement. The oxygen and carbon

isotopic composition in the CO

in the headspace was measured using a

Thermo Finnigan GASBENCH II coupled online with a Thermo Finnigan

Delta V isotope ratio mass spectrometer. Reference gas was pure CO

(4.5) from a cylinder calibrated against the VPDB standard by using

F.L. Arens et al. Global and Planetary Change 242 (2024) 104576

IAEA reference materials (NBS 18, NBS 19). Isotope values are shown in

the conventional delta-notation (δ

O, δ

C) in ‰(VPDB). Reproduc-

ibility of replicate measurements of lab standards (limestone) is gener-

ally better than 0.10 ‰(one standard deviation (1 SD)).

3.8.

Be/

Be ratios

Subsamples (0.5 g) from the sample set A were dried at 105 ◦C and

leached for 6 h using 0.04 M NH

OH⋅HCl in 25 % (v/v) acetic acid

(Bourles et al., 1989) in the Helmholtz Laboratory for the Geochemistry

of Earth Surface (HELGES) at GFZ, Potsdam. This procedure dissolves

meteoric

Be adsorbed onto any particulate matter (i.e., clay particles)

without dissolving the (coarse grained) silicate fraction containing in-

situ produced

Be. Samples and processing blanks were prepared

further according to Wittmann et al. (2015). An aliquot was taken for the

determination of stable natural

Be as well as other cations (Al, Ca, Fe, K,

Mg, Mn, Na) using ICP-OES. The samples were spiked with 150

g of

carrier to determine (

Be/

Be)

carrier

ratios using accelerator mass

spectrometry (AMS) measurements, after anion and cation exchange

chromatography, subsequent alkaline precipitation, oxidation, and

mixing with BeO:Nb in a 1:6 weight ratio.

The (

Be/

Be)

carrier

AMS ratios were determined at the Vienna

Environmental Research Accelerator (VERA) facility of the University of

Vienna, Austria, with a method described in Steier et al. (2019). The

ratios were measured relative to the Dresden-AMS (DREAMS) standard

material SMD-Be-12, with a

Be/

Be ratio of (1.704 ±0.030) ×10

−12

(Akhmadaliev et al., 2013). After blank ratio (1 ×10

−15

to 4 ×10

−14

)

subtraction, (

Be/

Be)

carrier

were converted into natural

Be/

Be ra-

tios. Fluxes Φof

Be within each sediment layer with thickness Δd and

corresponding age interval Δt were calculated with Φ(

)=C

(

)×

×Δd/Δt, where C(

) refers to the

Be concentration

corrected for radioactive decay, and

is the estimated layer density of

1.5 g cm

−3

3.9. Seed identification and semi-quantification

The plant seeds were extracted from a subsample of sample set A and

B by using their buoyancy in water. After stirring the water sediment

mixture thoroughly, the buoyant seeds were picked from the water

surface and dried at 60 ◦C. For comparison seeds from living shrubs were

collected close to the catchment area. The recovered seeds were inves-

tigated under light microscope and using a JEOL-6610LV scanning

electron microscope at the Museum für Naturkunde Berlin for identifi-

cation. Additionally, seeds from sample set A were counted for semi-

quantification.

3.10.

C dating

Radiocarbon (

C) of the seeds from sample set A and B, the only

source of sufficient organic matter in the profile, was measured at the

AMS laboratory of the ETH Zürich. The small samples containing only

between 10 and 100

g of carbon were measured on a MICADAS type

system equipped with a gas ion source for the direct

C determination

after sample combustion (Ruff et al., 2010;Wacker et al., 2013),

avoiding any potential sample contamination due to the laborious step

of sample graphitization. The associated calibrated ages were obtained

using the software Calib Rev. 8.1.0 (Stuiver and Reimer, 1993) in

conjunction with the calibration curve SHCal20 (Hogg et al., 2020).

3.11. Magnetostratigraphy

The paleo and rock magnetic analyses were applied to the samples

from set C at the GFZ, Potsdam. The natural remanent magnetization

(NRM) was measured and demagnetized with a 2G Enterprises 755 SRM

long-core magnetometer with an in-line tri-axial alternating field

demagnetizer. An anhysteretic remanent magnetization (ARM) was

imparted to discrete samples with a separate 2G Enterprises 600 single-

axis alternating field demagnetizer, with an alternating field amplitude

of 100 mT and a superimposed static field of 0.05 mT. Measurements

and stepwise demagnetization were performed with the long-core

magnetometer after applying field amplitudes up to 65 mT in order to

determine the median destructive field (MDF), and to provide data for

estimation of the relative paleo-intensity (rPI) variations using ARM

normalization. The latter was calculated using the NRM/ARM intensity

ratio after demagnetization at 30mT each. Low-field magnetic volume

susceptibility (κ

) measurements were performed with an AGICO

Kappabridge MFK1-A. The paleomagnetic methods and results are

described in more detail in the supplements as well as their results (Fig.

S2, S3). The calibrated

C age ranges were used to calibrate the mag-

netostratigraphic data. Finally, the magnetic field intensity, declination,

and inclination were used for the age tuning by comparing available

magnetostratigraphic records and models (Mix et al., 2003;Pav´

on-

Carrasco et al., 2014;Liu et al., 2020).

4. Results and interpretations

The stratigraphy of the studied profile is divided into two intervals:

The upper fine-grained deposits extend from 0 to 184 cm depth, fol-

lowed by coarse-grained deposits down to the base of the profile at a

depth of 420 cm (Table S1, S2; Figs. 2, 3, S4).

4.1. Clay Pan (CP)

4.1.1. CP geochemistry and sedimentology

The Herradura clay pan experienced recent flooding in June 2017

(Fig. S5). Subsequent evaporation has resulted in the formation of mud

chips and desiccation polygons (Fig. S6). Eight months after the flooding

event, exposed curled-up mud chips have been intensely eroded at their

edges (Fig. S1E). A distinct base of the recent depositional layer is not

visible, as the grain size distribution of the surface layer is analogous to

the stratigraphy of the upper 184 cm. This interval is characterized by

laminated layers dominated by clayey silt and interbedded sand, and

each layer exhibits unimodally grain size distributed (Table S1; Fig. 3B,

S4). The similarity between the upper interval stratigraphy and the

recent deposition suggests that the upper interval also consists of clay

pan deposits (CP), where episodic overland flow deposits are associated

with rain events of varying intensity. The CP stratigraphy can be divided

into a lower CP interval (104–184 cm depth) with few sand-rich between

thick silt-rich layers and an upper CP interval containing desiccation

cracks at the surface, sand-filled streamlets at ~45 cm depth, and overall

more abundant sand-rich layers. The siliciclastic mineralogy of the CP

deposits determined through semi-quantitative XRD is dominated by

quartz, feldspars, and phyllosilicates including smectite, illite, chlorite,

and sepiolite; showing no systematic change in content with depth

(Fig. 2C; S7).

The

-XRF element abundances and ratios are here used as envi-

ronmental proxies (e.g., Diederich et al., 2020;Liu et al., 2023). The Zr/

Rb ratio and the magnetic susceptibility (κ

) correlate with the grain

size derived from sieving (Fig. 3). The Rb/Sr ratio is used in lake sedi-

ments as a proxy for weathering within the catchment area, since Rb is

less readily released from bedrock by chemical weathering than Sr (Liu

et al., 2023). Generally, the Rb/Sr in the CP interval follows the grain

size proxies (κ

and Zr/Rb) but between 110 and 140 cm depth it shows

significantly lower values (Fig. 3C, D). The XRD results reveal that

gypsum (CaSO

⋅2H

O) is the dominant mineral of the water-soluble

fraction, referred to as solubles ranging between 8 and 18 wt%. Ca

and S abundances from

-XRF show a moderate positive correlation of

=0.66, indicating that Ca is not only present in gypsum but might be

present in other minerals such as plagioclase, phyllosilicates, and car-

bonates (Fig. 2C, 3E).

Salinity by EC measurements together with Na

, Cl

−

, and NO

−

concentration by IC analysis are elevated between 40 and 150 cm depth,

F.L. Arens et al. Global and Planetary Change 242 (2024) 104576

reaching a maximum at ~100 cm, where halite (NaCl) was detected by

XRD. The distribution of the nitrate, presumably nitratine (NaNO

), is

mirrored by the total nitrogen content (TN). The δ

values, being

close to the atmospheric value of 0 ‰, similar to most of the nitrate

deposits within the soils of the Atacama Desert, originate dominantly

from photochemical reactions in the atmosphere (Fig. S8) (B¨

ohlke et al.,

1997;Michalski et al., 2004). In the CP interval, the salinity peak around

100 cm depth is presumably not caused by rainwater infiltration down

to this depth as it is interpreted for typical soils based on Ma-old alluvial

deposits in the Atacama Desert, as the clay pan sediment are less

permeable and the clay pan setting is dominated by rapid evaporation

(Ewing et al., 2006, 2008;Finstad et al., 2016). Rather, it could indicate

a major shift in environmental conditions, which is also seen in the

change in sedimentology.

The TC is dominated by carbonates (calcite) detected with XRD

(Fig. 2). Due to the lack of bedrock- or plant-derived soil carbonate in the

catchment area (Amundson et al., 2012;Domagala et al., 2016), we

interpret the elevated carbonate content in the CP deposits to result from

atmospheric CO

uptake during flooding events followed by in-situ

carbonate precipitation during evaporation (Quade et al., 2007). The

bulk

value is a mix of carbonate and organic carbon and reflects the

overall carbonate dominance as the values are on average positive (1.15

‰) and close to the δ

carbonate

values (Fig. S8). The mean δ

carbonate

value of 3.5 ‰reflects the formation of carbonate at isotopic equilib-

rium with atmospheric CO

(Quade et al., 2007), suggesting that the

carbonate is primarily of abiotic origin. The mean δ

carbonate

value of

−1.5 ‰is close to the values of rainwater (−1.2 to −4.4 ‰) near 1000

m a.s.l. or fog (−1.9 ‰) (Aravena et al., 1989;Jordan et al., 2019) and

typical for evaporitic environments (Quade et al., 2007).

4.1.2. CP chronology

No microfossils (e.g., diatoms) or phytoliths or anything else suitable

for biostratigraphy was found in the profile. The only macroscopic

biological remains found are plant seeds within the upper 104 cm depth,

occurring at higher concentrations in the most fine-grained horizons

with the highest abundance at 3.5 cm and 40 cm depth (Fig. 3F). All

seeds are from the C3-plant Huidobria fruticosa, based on the comparison

with modern seeds found at the western slopes of Cerro de las Tetas, 15

km southwest of the Herradura clay pan, but none have been reported in

its current catchment area (Fig. 1, S9) (Placzek et al., 2014). Within the

fine sand fraction we have also found glass shards, indicative for crypto

tephra, along the entire CP interval, however, these were not suitable for

further age dating.

The depositional ages of the upper CP deposits (<104 cm depth)

were determined through radiocarbon age dating of H. fruticosa seeds

(Table 1; Fig. S9), which are maximum ages, as the transport duration is

Fig. 2. Sedimentological and geochemical results covering the full profile. A) Stratigraphic column, including sedimentological features based on field observations.

The lithic clasts density in the AF interval indicates the observed degree of clast support i.e., soluble content. B) Grain size distribution of the lithic fraction and

soluble content of the total sample. Uni- and bimodal grain size distribution is shown in black and gray columns. C) Semi-quantitative mineral content. D) Na

, Cl

−

ion concentration and electrical conductivity (EC). E) Total carbon (TC) and total nitrogen (TN) concentration.

F.L. Arens et al. Global and Planetary Change 242 (2024) 104576

unknown.

These ages were used to calibrate the magnetostratigraphic chro-

nology. The results of paleomagnetic analyses show that the samples are

very suitable for recording the paleomagnetic signal (Fig. S2, S3). The

NRM intensity, which is fairly high, ranges from ~70 to 700 mAm

−1

(Fig. 4A). Obtained relative paleo intensity (rPI) estimates, slope (NRM/

ARM) and NRM(30 mT)/κ

show a fairly good match (Fig. 4B). The

directions, inclinations and declinations of the characteristic remanent

magnetization (ChRM), together with the corresponding maximum

angular deviation (MAD) values are shown in Fig. 4C-E. Samples shown

in fig. S2 are marked by red diamonds in the MAD record (Fig. 4E).

Relative grain size changes are quite moderate (Fig. 4F). Coarser (finer)

particles are linked to higher (lower) ratio of the ARM against the

saturated remanent magnetization (SIRM), the latter represented by the

low field magnetic susceptibility (κ

) (Fig. 4F, H). The content of he-

matite is fairly constant, represented by S-ratios around 0.92 (Fig. 4G),

which is also indicated by the saturated and hard remanent magneti-

zation (Fig. S3). No reversed inclinations are observed nor geomagnetic

events like the Mono Lake or the Laschamp could be identified. The age

model was tuned by comparing available magnetostratigraphic records

and models yielding the depositional ages of the entire CP deposits with

its onset 19 ka ago (Mix et al., 2003;Pav´

on-Carrasco et al., 2014;Liu

et al., 2020) (Figs. 5, 6). The resulting sedimentation rate from 90 to 184

cm depth is ~10 cm ka

−1

, followed by a rapid decrease to ~2 cm ka

−1

within 85–90 cm depth, which was either caused by reduced sedimen-

tation or enhanced erosion (hiatus). At 0–85 cm depth a higher sedi-

mentation rate of ~20 cm ka

−1

was calculated.

The

Be/

Be ratios show a low variation over the CP interval

(Table 2;Fig. 7, S10A). Radioactive decay remains insignificant due to

the young age of the CP compared to the

Be half-life. We derive an

average

Be flux of Φ(

)=1.2 ×10

atoms cm

−2

−1

and a cal-

cium sulfate flux of Φ(CaSO

)=30 g m

−2

−1

(Table 3).

Fig. 3. Additional results covering the CP interval. A) Stratigraphy, for legend see Fig. 2. B) Grain size distribution and soluble content for the CP interval, for legend

see Fig. 2. C) Zr/Rb ratio (red line) and magnetic susceptibility (κ

) given in SI units (10

−6

) (black line) (see Fig. 4). D) Rb/Sr ratio. E) Ca and S concentrations. F)

Huidobria fruticosa seed abundance (purple dots), purple circles indicate that no seeds were found. (For interpretation of the references to colour in this figure

legend, the reader is referred to the web version of this article.)

Table 1

Radiocarbon dating results. Bold values were used for calibration of the magnetostratigraphic data.

Sample label Depth [cm] ETH lab code C [

g] F

C age (BP) [a] Age [cal BP] (68 %) [a] Prob. [%]

AC18-PP-2-5 3.5 111,657 40 0.896 ±0.009 880 ±90

670–800 88

870–900 12

AC19-PP-25 25 108,627 21 0.724 ±0.010 2600 ±110

2470–2480 3

2490–2760 97

AC19-PP-35 35 108,628 29 0.701 ±0.008 2850 ±90

2790–2820 13

2850–3010 87

AC18-PP-40 40 111,658 100 0.707 ±0.005 2790 ±60

2770–2880 87

2900–2920 13

AC-18-PP-80 80 111,659 20 0.582 ±0.014 4340 ±190

4580–4600 2

4620–5060 70

5110–5130 4

5180–5280 15

AC18-PP-90 90 111,660 12 0.304 ±0.042 9560 ±1100

9550–12,190 95

12,220–12,460 5

AC19-PP-95 95 108,629 16 0.351 ±0.012 8410 ±270 8990–9670 100

F.L. Arens et al. Global and Planetary Change 242 (2024) 104576

4.2. Alluvial Fan (AF)

4.2.1. AF geochemistry and sedimentology

The lower interval (184–420 cm) is composed of poorly sorted clay-

to gravel-sized siliciclastic sediments (Fig. 2B, C, Table S1, S2) with

clasts up to 5 cm in diameter. Thus, we interpret the parent material of

the lower interval as a distal alluvial deposit. The siliciclastic clasts are

heavily cemented by fairly soluble salts, mainly gypsum and minor

anhydrite (Fig. 2C). The soluble content decreases with depth, being

between 184 and 215 cm depth ~ 50 wt% where gypsum and anhydrite

Fig. 4. Down-core plots of various paleo- and mineral magnetic parameters from Herradura sediments: A) NRM intensity (black) and ARM intensity (brown), the

latter as a measure of the concentration of fine-grained magnetite, B) estimates of relative paleo intensity by the slope (NRM/ARM) of common alternating field

demagnetization steps (black) and NRM after 30 mT normalized by low-field magnetic susceptibility (κ

) (blue), C) ChRM inclination, D) ChRM declination, E)

precision of ChRM determination measured by the MAD, F) estimation of magnetic grain size (relative) using ARM over SIRM, G) S-ratio as a measure of relative

hematite contribution, and H) overall concentration of magnetic minerals represented by low field magnetic susceptibility (κ

). NRM/ARM/ChRM: natural/

anhysteretic/characteristic remanent magnetization, MAD: maximum angular deviation, SIRM: saturated isothermal remanent magnetization. Red diamonds in E)

mark samples with their detailed demagnetization results shown in fig. S2. Dashed vertical blue lines in C) and D) indicate inclination and declination, respectively,

of a pure geocentric axial dipole, calculated for the study site. (For interpretation of the references to colour in this figure legend, the reader is referred to the web

version of this article.)

F.L. Arens et al. Global and Planetary Change 242 (2024) 104576

have formed a vesicular texture which is only visible fragmentarily (Fig.

S1F). In 215–325 cm depth the soluble content averages 42 wt% (range

6–68 wt%) and below it decreases to 16 wt% in average (range 4–29 wt

%) (Table S1; Fig. 2B). These features resemble the locally common

inactive alluvial fan deposits (AF), accumulating dust and cemented by

salts during prolonged hyperaridity (e.g., Ericksen, 1981;Ewing et al.,

2006). The fragmented vesicular texture, being presumably part of the

near surface chusca soil layer, could indicate that the former surface of

the soil has been partially degraded during the initial clay pan flooding

events and subsequent compaction. Below the deposits resembles a

costra soil layer, which can be divided into an upper costra (215–325 cm)

layer being more intense cemented by gypsum as the lower costra soil

layer (325–420 cm). Its transition is also displayed in the grain size

distribution shifts from a unimodal lower costra to bimodal in the upper

costra and chusca soil layer, with an additional local maximum in the silt

fraction, and a higher standard deviation of the grain size distribution

indicating a poorly sorting (Table S2; Fig. S6). Due to these features, we

here define the chusca and upper costra soil layer as the upper AF interval

and the lower costra soil layer as lower AF interval.

No significant amounts of highly soluble salts (e.g., chlorides or ni-

trates) were detected that would indicate a caliche horizon, typical for

these deposits (Ericksen, 1981), and has been found in the Yungay valley

in depth of around 1–2 m, which would be an equivalent depth in our

profile of 3–4 m depth (Ewing et al., 2006;Arens et al., 2021). Either the

caliche horizon has initially formed below 420 cm depth, or the highly

soluble salts have been leached during the initial floodings of the basin.

Total carbon (TC) and total nitrogen (TN) content decrease in the AF

interval to values near or even below the detection limit (0.0124 wt%)

Fig. 5. Magnetostratigraphic results (light yellow background) including relative paleo intensity (slope (NRM/ARM)), declination (ChRM), and inclination (ChRM),

along with the comparison with marine records (Mix et al., 2003;Liu et al., 2020) and geomagnetic field models (Pav´

on-Carrasco et al., 2014). VADM =Virtual axial

dipole moment. Tie points are indicated by vertical dashed lines. (For interpretation of the references to colour in this figure legend, the reader is referred to the web

version of this article.)

F.L. Arens et al. Global and Planetary Change 242 (2024) 104576

and thus the isotope analysis reaches its limits, however a shift in

Carb

can be observed from the upper to the lower AF interval (325

cm depth).

4.2.2. AF chronology

While the

Be remains roughly constant in the CP deposits, it

sharply decreases at the transition to the AF deposits (Fig. 7). We

attribute this decrease to the exponential decay of

Be with depth. To

correct for the grain size dependence of

Be, the

Be concentration is

considered which correlates with the mean grain size of the fractions

used for the Be extraction (R

=0.61) (Fig. S10D-F). Consequently, the

Be/

Be ratio is applied for depositional age dating. The law of radio-

active decay, N(t) =N

t=19ka

−λd/S

, is fitted to the data, where λis the

decay rate of

Be, N(t) is the measured

Be/

Be ratio within each soil

layer and d the depth of each horizon (Fig. 8A). We assume the former

AF surface at 184 cm depth to correspond with the beginning of the CP

deposition 19 ka ago. Thus, the former AF surface

Be/

Be ratio N

t=19 ka

derived from the best fit yields (0.78 ±0.16) ×10

−8

. The corresponding

average sedimentation rate S =63 ±13 cm Ma

−1

is used for deriving the

depositional ages t =d/S for each layer. This results in the lowest AF

horizon covered in the profile at 420 cm depth having an age of 3.8 ±

0.8 Ma (Fig. 8B). Since no ash layer nor glass shards were found in the AF

interval for additional Ar/Ar or K/Ar dating, we consider dating using

the

Be/

Be ratio to be the most robust approach compared to e.g., only

using the

Be concentrations. However, for comparison, we also present

alternative dating approaches and their results in the supplemental

materials. Two alternative age models are illustrated for comparison,

using 1) the

Be concentrations alone, and 2) analogously to Wang et al.

(2015), the

Be concentration corrected for the water-soluble fraction

(Fig. S11), both yielding significantly higher ages than the

Be/

approach.

5. Discussion

The Atacama Desert provides an extraordinary opportunity to study

the dynamics of extreme arid landscapes persisting over geological

Fig. 6. CP depositional dating results. A) Depositional ages from radiocarbon

dating of H. fruticose seeds (yellow dots with gray uncertainty bars) and

magnetostratigraphy (gray line) with tie points (black dots and dashed vertical

line) (Fig. 5). (For interpretation of the references to colour in this figure

legend, the reader is referred to the web version of this article.)

Table 2

Analytical results of meteoric

Be data of the entire profile.

Sample Depth

[cm]

Age

[ka](CP)

[Ma]

(AF)

Be/

[10

−8

atoms

−1

]

[10

atoms

−1

]

[10

atoms g

−1

]

CP interval

AC18-

PP-0-2 10.17 ±

0.2 2.39 ±0.09 5.64 ±

0.20

2.236 ±

0.098

AC18-

PP-2-5 3.5 0.61 ±

0.2 2.47 ±0.06 5.81 ±

0.14

2.359 ±

0.013

AC18-

PP-10 10 1.35 ±

0.2 2.74 ±0.08 5.48 ±

0.12

1.998 ±

0.032

AC18-

PP-20 20 2.03 ±

0.2 2.62 ±0.06 4.96 ±

0.11

1.895 ±

0.006

AC18-

PP-30 30 2.58 ±

0.2 2.60 ±0.06 3.75 ±

0.09

1.442 ±

0.01

AC18-

PP-40 40 2.81 ±

0.2 2.75 ±0.08 6.64 ±

0.18

2.417 ±

0.012

AC18-

PP-50 50 3.58 ±

0.2 2.92 ±0.07 5.64 ±

0.14

1.935 ±

0.007

AC18-

PP-60 60 3.79 ±

0.2 3.04 ±0.11 5.87 ±

0.13

1.933 ±

0.055

AC18-

PP-70 70 4.18 ±

0.2 2.94 ±0.08 6.53 ±

0.17

2.222 ±

0.025

AC18-

PP-80 80 4.73 ±

0.2 3.06 ±0.08 6.71 ±

0.15

2.196 ±

0.026

AC18-

PP-90 90 8.35 ±

0.2 2.81 ±0.07 6.14 ±

0.14

2.19 ±

0.011

AC18-

PP-100 100 10.08 ±

0.2 2.57 ±0.07 5.90 ±

0.15

2.292 ±

0.008

AC18-

PP-110 110 11.39 ±

0.5 2.98 ±0.07 6.16 ±

0.13

2.066 ±

0.013

AC18-

PP-120 120 11.84 ±

0.5 3.03 ±0.07 5.22 ±

0.11

1.721 ±

0.01

AC18-

PP-130 130 12.96 ±

0.5 2.86 ±0.06 4.66 ±

0.10

1.63 ±

0.006

AC18-

PP-140 140 14.46 ±

0.5 2.80 ±0.04 6.13 ±

0.08

2.187 ±

0.013

AC18-

PP-150 150 15.23 ±

0.5 2.89 ±0.07 6.26 ±

0.16

2.168 ±

0.002

AC18-

PP-160 160 15.84 ±

0.5 2.76 ±0.04 5.41 ±

0.07

1.961 ±

0.006

AC18-

PP-170 170 17.12 ±

0.5 2.72 ±0.04 5.50 ±

0.08

2.019 ±

0.015

AC18-

PP-180 180 18.33 ±

0.5 2.73 ±0.04 5.06 ±

0.07

1.853 ±

0.007

AF interval

AC18-

PP-190 190

0.11 ±

0.02 1.835 ±0.041

2.697 ±

0.05

1.469 ±

0.018

AC18-

PP-200 200

0.27 ±

0.05 0.92 ±0.035

1.064 ±

0.04

1.156 ±

0.006

AC18-

PP-210 210

0.43 ±

0.09 0.524 ±0.015

0.552 ±

0.013

1.053 ±

0.017

AC18-

PP-220 220

0.59 ±

0.12 0.314 ±0.009

0.414 ±

0.012

1.319 ±

0.001

AC18-

PP-230 230

0.75 ±

0.15 0.438 ±0.012

0.585 ±

0.015

1.337 ±

0.009

AC18-

PP-240 240

0.9 ±

0.18 0.995 ±0.022

0.971 ±

0.021

0.976 ±

0.007

AC18-

PP-250 250

1.06 ±

0.22 0.415 ±0.015

0.343 ±

0.012

0.828 ±

0.004

AC18-

PP-260 260

1.22 ±

0.25 0.283 ±0.009

0.371 ±

0.012

1.313 ±

0.009

AC18-

PP-270 270

1.38 ±

0.28 0.244 ±0.008

0.562 ±

0.018

2.301 ±

0.018

AC18-

PP-280 280

1.54 ±

0.31 0.348 ±0.008

0.836 ±

0.018

2.401 ±

0.018

AC18-

PP-290 290

1.7 ±

0.35 0.736 ±0.016

0.881 ±

0.019

1.197 ±

0.006

AC18-

PP-300 300

1.85 ±

0.38 0.484 ±0.013

0.604 ±

0.016

1.248 ±

0.008

AC18-

PP-310 310

2.01 ±

0.41 0.271 ±0.01

0.259 ±

0.01

0.957 ±

0.004

AC18-

PP-320 320

2.17 ±

0.45 0.147 ±0.008

0.106 ±

0.006

0.72 ±

0.002

(continued on next page)

F.L. Arens et al. Global and Planetary Change 242 (2024) 104576

timescales. Therefore, chronology is crucial. The AF interval of the

Herradura record allows to reach back into the late Pliocene, while the

CP interval covers the late Pleistocene and Holocene including the

Marine Isotope Stage 1 (MIS 1, 0–14 ka) and part of the MIS 2 (14–19 ka)

(Fig. 9). Due to the sedimentation rates of the AF interval (~1 m Ma

−1

)

versus the CP interval (~100 m Ma

−1

) and their overall age difference,

different dating methods for each are required to disentangle their

respective climate histories. Furthermore, the deposition processes of

both intervals are fundamentally different. The CP is dominated by a

single process of episodic basin floodings during rare rain events. In

contrast, the AF has formed by discontinuous alluvial deposition and by

continuous accumulation of atmospheric dust and salts over millions of

years, resulting in a complex soil structure (Ewing et al., 2006;Arens

et al., 2021) (Fig. 9). Within the entire Atacama Desert only three clay

pans have been dated so far, only going back 10s and 100s ka (Ritter

et al., 2019;Diederich et al., 2020;Wennrich et al., 2024). In contrast to

the scattered clay pans, alluvial fans are almost ubiquitous in the Ata-

cama Desert and reach ages of over several million years. However,

dating alluvial deposits remains challenging and could strongly benefit

from new approaches like the here applied meteoric

Be/

Be method,

which is discussed in the following.

5.1. Alluvial fan record

For the millions-year-old AF interval common dating methods e.g.,

ash deposits, biostratigraphy, or magnetostratigraphy are not applicable

due to the absence of ash (including crypto tephra), fossils, or fine-

grained fluviatile deposits (e.g., Ewing et al., 2006;Rech et al., 2006;

Ritter et al., 2019). In-situ cosmogenic nuclide dating methods have

been widely used to date surface ages and burial ages in the Atacama

Desert (e.g., Ewing et al., 2006;Amundson et al., 2012;Jungers et al.,

2013). Surface disturbance, erosion, and complex exposure history can

significantly influence the dating outcome of depositional ages which

warrants caution and requires good knowledge of the depositional

environment (Jungers et al., 2013). As an alternative approach to dating

these deposits, we here applied meteoric

Be/

Be ratio-based method

providing a continuous record of depositional ages. However, when

Be/

Be ratio-based dating is applied to Atacama Desert sediments

several aspects have to be considered.

First, it is essential to note that

Be is released from rocks through

weathering, while

Be originates from the atmosphere (von Blancken-

burg et al., 2012). Unlike in the CP interval, the decay-corrected

and

Be concentration of the AF deposits do not correlate well (R

0.20, Fig. S10B) and it cannot be excluded that this is due to variations in

Be input in this environmental setting. Thus, the continuous sedimen-

tation in this environmental context is only a first approximation of the

AF chronostratigraphy. In fact, the alluvial fan deposition occurs

episodically with hiatus in between, whereas the atmospheric input of

salt and dust including

Be is added more continuously to the soil.

Especially the former surface of the alluvial fan deposit where the

stratigraphy transitions into the clay pan deposits, could signal such a

hiatus, as cosmogenic nuclide dating has yielded surface sediments of

alluvial fans in the Yungay valley with exposure ages of 0.3–1.1 Ma

(Jungers et al., 2013;Placzek et al., 2014). The high

in the upper

samples of the AF interval can also be indicative of a mixture with the

clay pan deposits, being fine grained and concentrated with

Be during

the initial clay pan flooding (Fig. 9, Table S4).

A size-depended particle migration after deposition due to rain

events is possible, for which the

Be related to nm-sized atmospheric

dust is more mobile than the

Be mainly related to

m-sized silicates.

Table 2 (continued)

Sample Depth

[cm]

Age

[ka](CP)

[Ma]

(AF)

Be/

[10

−8

atoms

−1

]

[10

atoms

−1

]

[10

atoms g

−1

]

AC18-

PP-330 330

2.33 ±

0.48 0.154 ±0.006

0.217 ±

0.008

1.415 ±

0.004

AC18-

PP-340 340

2.49 ±

0.51 0.151 ±0.057

0.153 ±

0.058

1.012 ±

0.004

AC18-

PP-350 350

2.65 ±

0.54 0.112 ±0.005

0.129 ±

0.006

1.156 ±

0.008

AC18-

PP-360 360

2.8 ±

0.58 0.138 ±0.004

0.126 ±

0.003

0.914 ±

0.005

AC18-

PP-370 370

2.96 ±

0.61 0.38 ±0.013

0.259 ±

0.009

0.68 ±

0.003

AC18-

PP-380 380

3.12 ±

0.64 0.073 ±0.012

0.029 ±

0.005

0.393 ±

0.001

AC18-

PP-390 390

3.28 ±

0.67 0.147 ±0.014

0.056 ±

0.005

0.379 ±

0.002

AC18-

PP-400 400

3.44 ±

0.71 0.156 ±0.013

0.149 ±

0.013

0.957 ±

0.002

AC18-

PP-410 410

3.6 ±

0.74 0.37 ±0.018

0.313 ±

0.015

0.845 ±

0.001

AC18-

PP-420 420

3.75 ±

0.77 0.174 ±0.005

0.104 ±

0.003

0.594 ±

0.006

Fig. 7. Beryllium isotope concentrations and ratios (not corrected for radio-

active decay) along the profile depth.

Average

Be fluxes are 3 orders of magnitudes lower than in the CP interval,

with Φ(

)=9.3 ×10

atoms cm

−2

−1

, while the average calcium sulfate

flux is one order of magnitude lower, with Φ(CaSO

)=0.31 g m

−2

−1

(Table 3).

Table 3

Sedimentation rates, CaSO

- and

Be-fluxes for the different intervals and

reference data. All fluxes including

Be and calcium sulfate fluxes calculated

analogous to the

Be fluxes are given in supplement tables S3 and S4.

Sedimentation-

rate

[m Ma

−1

]

Φ(CaSO

)

[g m

−2

−1

]

Φ(

)

[atoms cm

−2

−1

]

Atacama Desert

atmospheric

deposition

0.5–1.5

(Ewing et al.,

2006,

Owen et al.,

2013)

(3.7 ±0.6) ×10

(Wang et al.,

2015)

CP (0–184 cm) 100

(20−200)

(range 2–105)

1.2 ×10

(range 2.5 ×

–3.2 ×10

)

AF (184–420 cm) 0.63

0.31

(range

0.04–0.65)

9.3 ×10

(range 1.3 ×

–2.7 ×10

)

F.L. Arens et al. Global and Planetary Change 242 (2024) 104576

Furthermore, it is conceivable that seismic shaking results in an analo-

gous downward migration (Sager et al., 2020). This differential migra-

tion could explain the observed large variability in the

Be/

Be ratio

and the resulting dating. Hence, the chronostratigraphic interpretation

of these hyperarid settings remains challenging due to potential post-

depositional migration of soil constituents.

Thus far, only a single study, 150 km further north of the Herradura

clay pan, determined a chronostratigraphy using single meteoric

concentration. This hyperarid location, analogous to our AF interval,

yielded a maximum age of 6.6 ±0.4 Ma at 225 cm depth (Wang et al.,

2015). Applying the Wang et al. (2015) dating method, which is not

corrected for

Be concentration, to our data yields a maximum age of 6.8

±1.1 Ma at an analogous depth compared to 3.8 ±0.8 Ma with

Be/

Be (Fig. S11). We attribute this large difference in maximum age

to grain size variations affecting the single

Be concentrations, but not

the

Be and

Be concentrations (Wittmann et al., 2012). Given the

correlation between

Be and

Be in the CP interval and the correlation

between

Be and the grain size, we recommend to use

Be/

Be dating

for the following paleoenvironmental reconstruction of the AF interval.

Based on the

Be/

Be dating the average accumulation rates of

CaSO

and

Be in the AF interval (Φ(CaSO

)=0.31 g m

−2

−1

;

Φ(

)=9.3 ×10

atom cm

−2

−1

) are in the lower end of the at-

mospheric deposition rates calculated for the Atacama Desert, which, for

Be, are already one magnitude lower than the global average (5.7 ×

atom cm

−2

−1

) (Ewing et al., 2006;Field et al., 2006;Wang et al.,

2015) (Table 2). This can either point to particularly arid conditions in

the Yungay valley or to partially removal by erosion.

The transition of lower to upper AF deposits between 320 and 330

cm depth, marked by an increase in soluble content, occurred 2.3 ±0.5

Ma ago. This could be a result of a reduced salt leaching or an increasing

salt input in the upper AF interval. A higher input could be caused by a

more frequent fog occurrence, being a major salt source in the Atacama,

introducing salts from the ocean into the desert (Voigt et al., 2020). The

current situation with an upper fog boundary in this region ~1200 m a.s.

l. allows fog to penetrate the Yungay valley, however, the stable isotope

signature of the sulfates within the adjacent soils indicate only a minor

influence of the fog as a salt supplier (Arens et al., 2021). Synchronously

a drainage division occurred in the Yungay valley during the late Plio-

cene which was caused by local alluvial activity, that could have

influenced the salt input into the Yungay valley (Amundson et al., 2012).

The AF interval is likely part of these local alluvial infills and could be

affected by this division, becoming hydrologically disconnected from

the Andes. The higher CaSO

content in the upper interval is presumably

related to a lower moisture availability during its formation i.e., a cli-

matic change towards more arid conditions. This has been shown for

soils along an aridity gradient within the Atacama Desert, where sulfate

content correlates with aridity (Ewing et al., 2006). The shift from a

clast- to a matrix-supported texture (Fig. 9), as well as the grain size

distribution shifting from unimodal to bimodal could point also to an

overall more arid condition (Fig. S4): The unimodal distribution in the

lower AF interval points to a single transport mechanism dominated by

alluvial transport, while the bimodal grain size distribution in the upper

AF interval could point to a secondary input, most likely atmospheric

dust (Vandenberghe, 2013). We assume that initially the alluvial ma-

terial was clast-supported and underwent volumetric expansion through

gypsum and minor lithic dust uptake. In addition, the upper AF interval

contains remnants of a chusca horizon, which is a pedogenic feature

typically for the hyperarid Atacama Desert and a good proxy for

persistent and prolonged hyperarid conditions (Pfeiffer et al., 2021).

This soil structure also indicates a very slow, or non-deposition of al-

luvial material (lithic clasts), and in this sense presents a hiatus. This

potential transition to a drier climate at 2.3 ±0.5 Ma is consistent with

other paleo record data available for the Atacama Desert based on

radiometric dating (including cosmogenic nuclides) situating this

climate shift between 2 and 3 Ma ago (Hartley and Chong, 2002;Ewing

et al., 2006;Amundson et al., 2012;Jordan et al., 2014;Wang et al.,

2015;Ritter et al., 2018b). The climatic change is linked to an interplay

between global cooling during the Pliocene resulting in a northern

hemisphere glaciation ~2.75 Ma ago and long-term variability in the El

Ni˜

no Southern Oscillation (ENSO), which stabilized ~2 Ma ago to pre-

sent day condition (Ravelo et al., 2004;Amundson et al., 2012). For later

wet phases which have been identified in paleo records further north we

find no indication (Jordan et al., 2014;Wang et al., 2015), highlighting

the complexity of the climate within the Atacama Desert. Another

reason could be the poorly constrained chronology in the AF interval

and its broad resolution.

Future investigation should apply additional dating methods like

Al/

Be exposure dating and

Be/

Ne burial dating to achieve a

Fig. 8. AF depositional dating results. A)

Be/

Be ratio (black dots) and their regression line (black line) as a function of depth for the AF interval. The band of

uncertainty (gray) is calculated with a confidence interval of one standard deviation (1 SD). B) Depositional age as function of depth for the AF interval. The

correlation including the 1 SD band of uncertainty is derived from the sedimentation rate.

F.L. Arens et al. Global and Planetary Change 242 (2024) 104576

more precisely constrained chronology of alluvial fan deposits. A com-

bined dating approach could also give information about potential hi-

atus within the deposition history and post depositional soil processes.

The exposure dating should be able to identify whether there is a sig-

nificant hiatus and its duration at the border of the AF and CP intervals.

Further efforts should be made to find adjacent ash layers which can be

linked to the stratigraphy in the field or crypto tephra within the stra-

tigraphy itself, to apply K/Ar or Ar/Ar dating for tuning the cosmogenic

nuclide dating methods.

5.2. Clay pan record

The onset of the clay pan is related to the formation of an endorheic

basin, preventing further drainage into the lower-lying Aguas Blancas

salar. We assume that the basin was subsequently flooded. However, the

initial CP deposits recorded in our profile do not necessarily coincide

with the base of the entire clay pan. Hence, the clay pan together with

the basin could have formed earlier. Reasons for the formation could be

outflow blocking through tectonic events, as known for other clay pans

in the hyperarid Atacama Desert (Ritter et al., 2019;Diederich et al.,

2020) and with a potential fault lying only ~1 km east of the clay pan

(Domagala et al., 2016) (Fig. 1C). Alternatively, also alluvial activity can

cause drainage division, which has been described for the Yungay valley

before (Amundson et al., 2012).

For the CP deposits, which are too young for the meteoric

Be dating

method, we observed a good correlation between the

Be and

concentration (R

=0.78) yielding a relatively constant ratio

throughout its deposition (Fig. S10A). This has also been shown for

Fig. 9. A) Sketch of the Herradura stratigraphic buildup. 3.8–2.0: Ma arid conditions, active alluvial fan; 2.0–0.02 Ma: hyperarid conditions, reduced alluvial ac-

tivity, accumulation of atmospheric dust, Atacama typical soil structure (caliche); 19 ka–present: basin formation and clay pan deposition by episodically flooding

after rare rain events. Legend as in Fig. 2. B) Chronological columns with the Herradura record next to relevant paleo records including identified dry and wet phases

like the Late Holocene (L. H.) wet phase and Central Andean Pluvial Events (CAPE I and CAPE II), Marine Isotope Stages (MIS), and the corresponding geological time

episodes (ages in ka, black ticks in the CP interval represent 1 ka time interval, red ticks in the AF interval represents 1000 ka time interval). (For interpretation of the

references to colour in this figure legend, the reader is referred to the web version of this article.)

F.L. Arens et al. Global and Planetary Change 242 (2024) 104576

various lacustrine and marine deposits over the past 8 Ma (Lebatard

et al., 2010), and 12 Ma, respectively (Willenbring and von Blancken-

burg, 2010). The fluxes for calcium sulfate are a hundred times higher

(Φ(CaSO

)=30 g m

−2

−1

) and for

Be a thousand times higher

(Φ(

)=1.2 ×10

atom cm

−2

−1

) in the CP interval compared to

the AF interval, which can be explained by the different depositional

environments, depositing more or less efficient fine-grained material

which is associated to the Beryllium and the calcium sulfate. In a clay

pan environment the dominating input of fine-grained sediment are

flash floods or overland flows from the entire catchment area. In

contrast, the alluvial fan is a less efficient depositional environment for

fine-grained material, which is dominantly transported further down-

stream. Remarkably, the CP depositional

Be flux also exceeds the

fluxes derived from other materials collected around the globe, such as

soil profiles, riverine solids, and marine records, by an order of magni-

tude (e.g., Wang et al., 1996;Deng et al., 2021). Therefore, a detailed

reconstruction of the paleoclimate aids in a better understanding of this

unique environment.

The Yungay valley has undergone multiple climatic changes within

the last 19 ka that can be deciphered from the CP record. We base our

paleo reconstruction of the CP interval on different proxies including the

grain size distribution, which is being used as a proxy for precipitation

within the catchment area of a clay pan (e.g. Ritter et al., 2019;Die-

derich et al., 2020;Wennrich et al., 2024). Silt-dominated fine-grained

layers represent phases with minor flooding events and sand-dominated

coarse-grained layers represent phases with moderate floodings caused

by heavy rain events developing higher transport energies. These shifts

in moisture availability are supported by the following additional

proxies: 1) The shift in sedimentation rates in the Holocene, dropping

after the Pleistocene down to ~2 cm ka

−1

and increasing over the last 5

ka supports the dry phase in the early Holocene and more intense and/or

more frequent flash floods during the late Holocene (Fig. 6A). 2) The

elemental ratio Rb/Sr (Fig. 3C), indicative for weathering (e.g., Liu

et al., 2023), reflect the pattern of the grain size distribution. Moreover,

the Rb/Sr ratio reveals a sustained period of even lower Rb/Sr ratio

14.5–12.4 ka ago, supporting increased water availability during this

period. 3) The increase of seed abundance during the late Holocene

(Fig. 3E) coincides with other records in the coastal area, like the

presence of rodent middens and alluvial fan activity (Vargas et al., 2006;

Diaz et al., 2012). 4) The increase of NO

−

and Cl

−

begins with the

transition from the Pleistocene to the Holocene, indicating a input from

incised salt-containing alluvial fans by more abundant or more intense

flash floods, capable of leaching these salts. Eroded alluvial fans are

especially abundant in the southeastern catchment area (Fig. S5C).

These channel incisions have been observed as a response in the study

area after recent extraordinary rain events (Pfeiffer et al., 2021).

Together, three wet phases can be identified in the CP interval (Fig. 9A,

B). The two earlier wet phases (14.5–12.5 ka and 11.0–9.5 ka) and a ~ 5

ka long third phase during the late Holocene, while within the last

centuries the aridity tends to increase again to present day conditions.

The variety of available paleo records around the hyperarid core of

the Atacama Desert, including rodent middens, pollen, groundwater

discharge deposits, and paleo shoreline reconstruction reveal a hetero-

geneous precipitation pattern throughout the Atacama Desert during the

late Pleistocene and Holocene (e.g., Grosjean et al., 2001;S´

aez et al.,

2016;de Porras et al., 2017;Pfeiffer et al., 2018;Gonz´

alez-Pinilla et al.,

2021). These records show that the intensity and frequency of wet

conditions were primarily influenced by the proximity to the moisture

sources. In the northern latitudes (16–20◦S) in areas fed by water from

the eastern mountains, the records suggest that the N-NE South Amer-

ican Summer Monsoon significantly influences the regional rainfall

patterns, with four wet phases: the Tauca phase (18–14 ka), the Coipasa

phase (13–10 ka), early and late Holocene phases. Conversely, in the

eastern mountains of the southern latitudes (23–28◦S), the data in-

dicates a SE summer monsoon, coupled with SW Pacific extratropical

moisture influences brought by cutoff lows and moisture conveyor belts,

with only three wet phases: the Central Andean Pluvial Events (CAPE I:

15.9–13.8 ka; CAPE II: 12.7–9.7 ka) and a late Holocene wet phase. The

late Pleistocene wet phases are attributed to a complex interaction be-

tween North Atlantic sea surface temperature, anomalies in the Pacific

sea surface temperature leading to La Ni˜

na-like states, and increased

local summer insolation, all influencing the South American summer

monsoon (Gonz´

alez-Pinilla et al., 2021). For the Holocene, in the eastern

Atacama Desert both in the north (Gonz´

alez-Pinilla et al., 2021) and in

the south (S´

aez et al., 2016), the records imply that the orbital forcing

became more dominant, influencing the ENSO, resulting in Eastern

Pacific sea surface temperature fluctuation (S´

aez et al., 2016;Gonz´

alez-

Pinilla et al., 2021).

In the western sectors of the hyperarid core of the Atacama Desert

which do not receive runoff from the Andes mountains, the influence of

the major moisture sources becomes less clear, as records are sparse.

Here, clay pans have been proven suitable as valuable records of pre-

cipitation proxies including sediment grain size, element ratios, and

biomarkers including diatoms, phytoliths, and lipids (Ritter et al., 2019;

Diederich et al., 2020;Wennrich et al., 2024). In the northern part, at

20◦S, the Huara clay pan is influenced by the N-NE South American

Summer Monsoon (tropical continental) and the SW Pacific extra-

tropical (temperate maritime) moisture sources (Diederich et al., 2020).

In the central Atacama Desert, at 21.5◦S, the PAG clay pan reveals a

precipitation pattern asynchronous to the South American Summer

Monsoon but can be linked with the sea surface temperature offshore

North Chile and Peru, connected to the ENSO (tropical maritime) (Ritter

et al., 2019). Further south of the Yungay valley, in the Paranal clay pan

(24.5◦S, 2231 m a.s.l.), precipitation occurrence and intensity based on

clay pan flooding analyzed via satellite imaging over the past 30 years

(Wennrich et al., 2024). These flooding events can be correlates with

recent ENSO variabilities indicating it as the main driver for heavy rain

events in the catchment area. During the late Pleistocene and Holocene

the chronostratigraphy of this clay pan deposit roughly links coarse-

grained sediments, elevated diatom, phytolith, and lipid biomarker

concentration within the deposits to enhanced moisture availability

which could be related with the CAPEs (Wennrich et al., 2024).

The Herradura record lies in a key position to provide a link between

southern and northern clay pan records helping to refine the spatial

influences and temporal variability of the different weather systems

affecting the Atacama Desert, such as the South American summer

monsoon and the ENSO (S´

aez et al., 2016;Houston and Latorre, 2022).

This is particularly evident in the well constrained chronology of the

upper 104 cm CP interval, combining radiocarbon dating with magne-

tostratigraphy. Below, we have to solely rely on the magneto-

stratigraphy, resulting in a reduced precision of the chronology.

However, the commonality of the CP record with the adjacent records

suggests a climatological connection between the Yungay valley, the

southern Pre-Cordillera, and the southern Coastal Cordillera (e.g.,

Grosjean et al., 2001;Maldonado et al., 2005;S´

aez et al., 2016;Wenn-

rich et al., 2024). The two earlier wet phases identified in the CP record

roughly coincide with the CAPEs (Quade et al., 2008), indicating a

mutual moisture source during this period that affected the Yungay

Valley. The temporal offset could be due to the distance from the

moisture source of the CAPEs, which had only limited influence on this

area. However, the less constrained chronology in the lower CP interval

warrants further investigation to verify this. During the Holocene, the

sedimentation rate increased from ~2 to 20 cm ka

−1

at around 5 ka

which aligns with records around the southern Atacama Desert and

western slopes of the Andes which can be linked to ENSO regime shifts

(S´

aez et al., 2016;Gonz´

alez-Pinilla et al., 2021).

The comparison of the CP record with other available clay pan re-

cords within the Atacama Desert, the records further north (PAG ~280

km north, Huara ~445 km north) show significant differences in their

precipitation history (Ritter et al., 2019;Diederich et al., 2020). Espe-

cially within the late Holocene, the lack of sedimentation or a deflation

resulting in a hiatus for the last 6 to 8 ka in both records indicating no or

F.L. Arens et al. Global and Planetary Change 242 (2024) 104576

very little flash floods, synchronous to the highest sedimentation rate

within the CP underlines the heterogeneity (Ritter et al., 2019;Diederich

et al., 2020). The most adjacent Paranal clay pan, ~50 km south of the

Yungay valley shows similarities with the CP record despite of its poorly

constrained chronology. Both records share the same sequence of wet

and dry phases in the late Pleistocene and in the Holocene. Especially the

diatom and phytolith record of the Paranal clay pan shows significant

similarities with the seed record, indicating an abundance increase in

the late Holocene and a drop towards the present day (Wennrich et al.,

2024). However, the intensity of the rain events in the late Pleistocene

wet phase seems to differ significantly, where the Herradura clay pan is

dominated by sand/silt versus the Paranal deposit encompasses gravel,

indicating alluvial debris flow into the clay pan during this period

(Wennrich et al., 2024). This points to less extreme rain events affecting

the Herradura clay pan compared to the Paranal clay pan during this

period. Reasons for the different precipitation intensities could be the

distance to the southern moisture sources (temperate maritime, i.e., cut-

off lows and temperate continental, i.e., SE South American summer

monsoon). In general, also the elevation difference of over 1000 m be-

tween both clay pans presumably plays an important role, as precipi-

tation increases with increasing elevation exceeding 20 mm a

−1

at 2300

m a.s.l. (Houston and Hartley, 2003). The comparison of clay pan re-

cords within the hyperarid core of the Atacama Desert supports the

heterogeneous precipitation pattern in the region (Ritter et al., 2019;

Diederich et al., 2020;Wennrich et al., 2024). Furthermore, the higher

sedimentation rates of the southern clay pans point to the overall higher

precipitation within their catchments, indicating a northward increase

of aridity for the last 19 ka. Unfortunately, the northern and southern

records do not overlap for long timespans, which warrants further in-

vestigations on clay pan records in the south, reaching further back in

time to better comprehend the climate controls in the Atacama Desert

within its past to allow predictions for future climate change.

6. Conclusion

The challenge of reconstructing the climatic history of the Atacama

Desert over millions of years was addressed by a meteoric

Be/

dating method, novel for soils, which formed under hyperaridity and

capable of correcting for grain-size effects on the determined deposi-

tional ages. In contrast, the overlying CP deposits are dominated by

relatively rapid deposition by flash floods, being too young for the

Be/

Be dating method. However, these fine-grained sediments are

suitable for magnetostratigraphy and with the occurrence of the

H. fruticosa seeds it was possible to conduct radiocarbon dating. Caution

is advised as the chronology is limited, as the lower CP interval relies

solely on the magnetostratigraphy and lacks independent tie points. This

is also true for the AF interval, where we only have the

Be dating.

Nonetheless, both record types contribute to unraveling the evolution of

the depositional environment. The alluvial fan deposits, in particular,

integrate long-term climate signals spanning millions of years, while the

clay pan deposits offer insights into episodic events on local and regional

scales over millennial and orbital timescales. Together with different

paleo environmental proxies, the combined record provides new dating

tools and perspectives on the climate in the hydrologically discrete

Coastal Cordillera responding to the temporally and spatially variable

climate system of the Atacama Desert over the course of 3.8 Ma. Hence,

our study is of great importance for studies investigating habitability

under extreme aridity and biomarker preservation (e.g. Schulze-Makuch

et al., 2018;Azua-Bustos et al., 2023), which require a profound un-

derstanding of soil processes and past environmental conditions as well

as their regional variability in the Atacama Desert.

CRediT authorship contribution statement

Felix L. Arens: Writing –review &editing, Writing –original draft,

Visualization, Validation, Resources, Project administration,

Methodology, Investigation, Formal analysis, Data curation, Conceptu-

alization. Jenny Feige: Writing –review &editing, Writing –original

draft, Methodology, Investigation, Funding acquisition, Formal analysis,

Data curation. Alessandro Airo: Writing –review &editing, Writing –

original draft, Supervision, Project administration, Investigation,

Conceptualization. Christof Sager: Writing –review &editing, Meth-

odology, Investigation, Formal analysis, Data curation. Lutz Hecht:

Methodology, Data curation. Lucas Horstmann: Writing –review &

editing, Methodology, Formal analysis. Felix E.D. Kaufmann: Meth-

odology, Formal analysis. Johannes Lachner: Methodology, Formal

analysis, Data curation. Thomas Neumann: Writing –review &editing,

Validation, Resources. Norbert Nowaczyk: Writing –review &editing,

Resources, Methodology, Formal analysis, Data curation. Ferry Schi-

perski: Writing –review &editing, Methodology, Data curation. Peter

Steier: Methodology, Formal analysis, Data curation. Alexandra Stoll:

Writing –review &editing, Methodology, Investigation. Ulrich Struck:

Methodology, Formal analysis. Bernardita Valenzuela: Writing –re-

view &editing, Resources. Friedhelm von Blanckenburg: Writing –

review &editing, Data curation. Hella Wittmann: Writing –review &

editing, Methodology, Data curation. Lukas Wacker: Writing –review

&editing, Methodology, Formal analysis, Data curation. Dirk Wagner:

Writing –review &editing, Resources. Pedro Zamorano: Writing –

review &editing, Resources. Dirk Schulze-Makuch: Writing –review &

editing, Supervision, Resources, Funding acquisition.

Declaration of competing interest

The authors declare that they have no known competing financial

interests or personal relationships that could have appeared to influence

the work reported in this paper.

Data availability

Data will be made available on request.

Acknowledgement

J.F., A.A. and C.S. acknowledge partial funding by the European

Union (ERC, NoSHADE, 101077668). DSM acknowledges support by the

European Research Council Advanced Grant Habitability of Martian

Environments (#339231). Views and opinions expressed are however

those of the author(s) only and do not necessarily reflect those of the

European Union or the European Research Council Executive Agency.

Neither the European Union nor the granting authority can be held

responsible for them.

J.F. and C.S. thank the BMBF (Bundesministerium für Bildung und

Forschung, project 05K16KTB) for partial funding. The authors

acknowledge the support by the RADIATE project funded by the Euro-

pean Union’s Horizon 2020 research and innovation program under

grant agreement No. 824096 (Radiate Proposal 19001796-ST).

PZ and BV thank for support by MINEDUC-UA Project code

ANT1856.

We thank Laura Jentzsch, Matthias Müller, Jacob Heinz, Friedrich

Trepte, Edit Airo, Paloma Morales and M´

aximo Gonzalez for their help

during the fieldwork.

Appendix A. Supplementary data

Supplementary data to this article can be found online at https://doi.

org/10.1016/j.gloplacha.2024.104576.

References

Akhmadaliev, S., Heller, R., Hanf, D., Rugel, G., Merchel, S., 2013. The new 6 MV AMS-

facility DREAMS at Dresden. Nucl. Instrum. Methods Phys. Res., Sect. B 294, 5–10.

Amundson, R., Dietrich, W., Bellugi, D., Ewing, S., Nishiizumi, K., Chong, G., Owen, J.,

Finkel, R., Heimsath, A., Stewart, B., Caffee, M., 2012. Geomorphologic evidence for

F.L. Arens et al. Global and Planetary Change 242 (2024) 104576

the late Pliocene onset of hyperaridity in the Atacama Desert. Geol. Soc. Am. Bull.

124, 1048–1070.

Aravena, R., Suzuki, O., Pollastri, A., 1989. Coastal fog and its relation to groundwater in

the IV region of northern Chile. Chemical Geology: Isotope Geoscience section 79,

83–91.

Arens, F.L., Airo, A., Feige, J., Sager, C., Wiechert, U., Schulze-Makuch, D., 2021.

Geochemical proxies for water-soil interactions in the hyperarid Atacama Desert,

Chile. CATENA 206, 105531.

Azua-Bustos, A., Fair´

en, A.G., Gonz´

alez-Silva, C., Prieto-Ballesteros, O., Carrizo, D.,

S´

anchez-Garcínez, M.´

A., Escudero, C., Mu˜

noz-Iglesias, V., others, 2023. Dark

microbiome and extremely low organics in Atacama fossil delta unveil Mars life

detection limits. Nat. Commun. 14, 808.

Bartz, M., Walk, J., Binnie, S.A., Brill, D., Stauch, G., Lehmkuhl, F., Hoffmeister, D.,

Brückner, H., 2020. Late Pleistocene alluvial fan evolution along the coastal Atacama

Desert (N Chile). Glob. Planet. Chang. 190, 103091.

Blott, S.J., Pye, K., 2001. GRADISTAT: a grain size distribution and statistics package for

the analysis of unconsolidated sediments. Earth Surf. Process. Landf. 26, 1237–1248.

B¨

ohlke, J.K., Ericksen, G.E., Revesz, K., 1997. Stable isotope evidence for an atmospheric

origin of desert nitrate deposits in northern Chile and southern California, USA.

Chem. Geol. 136, 135–152.

Bourles, D., Raisbeck, G.M., Yiou, F., 1989. 10Be and 9Be in marine sediments and their

potential for dating. Geochim. Cosmochim. Acta 53, 443–452.

Cereceda, P., Larrain, H., Osses, P., Farías, M., Ega˜

na, I., 2008. The spatial and temporal

variability of fog and its relation to fog oases in the Atacama Desert, Chile. Atmos.

Res. 87, 312–323.

Chmeleff, J., von Blanckenburg, F., Kossert, K., Jakob, D., 2010. Determination of the

10Be half-life by multicollector ICP-MS and liquid scintillation counting. Nucl.

Instrum. Methods Phys. Res., Sect. B 268, 192–199.

Clarke, J.D.A., 2006. Antiquity of aridity in the Chilean Atacama Desert. Geomorphology

73, 101–114.

Davila, A.F., Schulze-Makuch, D., 2016. The last possible Outposts for Life on Mars.

Astrobiology 16, 159–168.

Davis, M., Matmon, A., Placzek, C.J., McIntosh, W., Rood, D.H., Quade, J., 2014.

Cosmogenic nuclides in buried sediments from the hyperarid Atacama Desert, Chile.

Quat. Geochronol. 19, 117–126.

de Porras, M.E., Maldonado, A., Pol-Holz, R. de, Latorre, C., Betancourt, J.L., 2017. Late

Quaternary environmental dynamics in the Atacama Desert reconstructed from

rodent midden pollen records. J. Quat. Sci. 32, 665–684.

Deng, K., Wittmann, H., Hsieh, M.-L., Yang, S., von Blanckenburg, F., 2021. Deposition

and retention of meteoric 10Be in Holocene Taiwan river terraces. Quat. Sci. Rev.

265, 107048.

Diaz, F.P., Latorre, C., Maldonado, A., Quade, J., Betancourt, J.L., 2012. Rodent middens

reveal episodic, long-distance plant colonizations across the hyperarid Atacama

Desert over the last 34,000 years. J. Biogeogr. 39, 510–525.

Diederich, J.L., Wennrich, V., Bao, R., Büttner, C., Bolten, A., Brill, D., Buske, S.,

Campos, E., Fern´

andez-Galego, E., G¨

odickmeier, P., Ninnemann, L., Reyers, M.,

Ritter, B., Ritterbach, L., Rolf, C., Scheidt, S., Dunai, T.J., Melles, M., 2020. A 68 ka

precipitation record from the hyperarid core of the Atacama Desert in northern

Chile. Glob. Planet. Chang. 184, 103054.

Domagala, J.P., Escribano, J., La Cruz, R. de, Saldias, J., Joquera, R., 2016. Cartas Blanco

Encalada y Pampa Remiendos, Region de Antofagasta. Servicio Nacional de Geología

B´

asica, 187–188.

Dunai, T.J., Gonz´

alez, L., G., Juez-Larr´

e, J, 2005. Oligocene–Miocene age of aridity in the

Atacama Desert revealed by exposure dating of erosion-sensitive landforms. Geology

33, 321–324.

Ericksen, G.E., 1981. Geology and Origin of nitrate deposition in Atacama Desert. U.S.

Geological Survey Professional Paper 37:1188.

Evenstar, L.A., Mather, A.E., Hartley, A.J., Stuart, F.M., Sparks, R.S.J., Cooper, F.J.,

2017. Geomorphology on geologic timescales: Evolution of the late Cenozoic Pacific

paleosurface in Northern Chile and Southern Peru. Earth Sci. Rev. 171, 1–27.

Ewing, S.A., Sutter, B., Owen, J., Nishiizumi, K., Sharp, W., Cliff, S.S., Perry, K.,

Dietrich, W., McKay, C.P., Amundson, R., 2006. A threshold in soil formation at

Earth’s arid–hyperarid transition. Geochim. Cosmochim. Acta 70, 5293–5322.

Ewing, S.A., Yang, W., DePaolo, D.J., Michalski, G., Kendall, C., Stewart, B.W.,

Thiemens, M., Amundson, R., 2008. Non-biological fractionation of stable ca

isotopes in soils of the Atacama Desert, Chile. Geochim. Cosmochim. Acta 72,

1096–1110.

Field, C.V., Schmidt, G.A., Koch, D., Salyk, C., 2006. Modeling production and climate-

related impacts on 10Be concentration in ice cores. J. Geophys. Res. Atmos. 111.

File, P.D., others, 2000. International Centre for Diffraction Data. Swarthmore, Pa,

p. 19081.

Finstad, K., Pfeiffer, M., Amundson, R., 2014. Hyperarid Soils and the Soil Taxonomy.

Soil Sci. Soc. Am. J. 78, 1845–1851.

Finstad, K., Pfeiffer, M., McNicol, G., Barnes, J., Demergasso, C., Chong, G.,

Amundson, R., 2016. Rates and geochemical processes of soil and salt crust

formation in Salars of the Atacama Desert, Chile. Geoderma 284, 57–72.

Garreaud, R.D., Molina, A., Farias, M., 2010. Andean uplift, ocean cooling and Atacama

hyperaridity: a climate modeling perspective. Earth Planet. Sci. Lett. 292, 39–50.

Gonz´

alez-Pinilla, F.J., Latorre, C., Rojas, M., Houston, J., Rocuant, M.I., Maldonado, A.,

Santoro, C.M., Quade, J., Betancourt, J.L., 2021. High-and low-latitude forcings

drive Atacama Desert rainfall variations over the past 16,000 years. Sci. Adv. 7,

eabg1333.

Grosjean, M., van Leeuwen, J.F.N., van der Knaap, Willem, Oscar, Geyh, M.A., Ammann,

B., Tanner, W., Messerli, B., La Nú˜

nez, Valero-Garc´

es, B.L., Veit, H, 2001. A 22,000

14C year BP sediment and pollen record of climate change from Laguna Miscanti (23

S), northern Chile. Glob. Planet. Chang. 28, 35–51.

Hartley, A.J., Chong, G., 2002. Late Pliocene age for the Atacama Desert: Implications for

the desertification of western South America. Geol 30, 43.

Hogg, A.G., Heaton, T.J., Hua, Q., Palmer, J.G., Turney, C.S.M., Southon, J., Bayliss, A.,

Blackwell, P.G., Boswijk, G., Bronk Ramsey, C., et al., 2020. SHCal20 Southern

Hemisphere Calibration 0–55 000 years cal BP. Radiocarbon 62, 759–778.

Houston, J., 2006a. Evaporation in the Atacama Desert: an empirical study of spatio-

temporal variations and their causes. J. Hydrol. 330, 402–412.

Houston, J., 2006b. Variability of precipitation in the Atacama Desert: its causes and

hydrological impact. Int. J. Climatol. 26, 2181–2198.

Houston, J., Hartley, A.J., 2003. The central Andean west-slope rainshadow and its

potential contribution to the origin of hyper-aridity in the Atacama Desert. Int. J.

Climatol. 23, 1453–1464.

Houston, J., Latorre, C., 2022. The role of the non-stationary Andean Dry Diagonal in

paleoclimate reconstructions. Hydrol. Process. 36.

Jordan, T.E., Kirk-Lawlor, N.E., Blanco, N.P., Rech, J.A., Cosentino, N.J., 2014.

Landscape modification in response to repeated onset of hyperarid paleoclimate

states since 14 Ma, Atacama Desert, Chile. Bulletin 126, 1016–1046.

Jordan, T.E., Herrera, L., C., Godfrey, L.V., Colucci, S.J., Gamboa P., C., Urrutia M., J.,

Gonz´

alez L., G., Paul, J.F, 2019. Isotopic characteristics and paleoclimate

implications of the extreme precipitation event of March 2015 in northern Chile.

andgeo 46, 1.

Jungers, M.C., Heimsath, A.M., Amundson, R., Balco, G., Shuster, D., Chong, G., 2013.

Active erosion–deposition cycles in the hyperarid Atacama Desert of Northern Chile.

Earth Planet. Sci. Lett. 371-372, 125–133.

Korschinek, G., Bergmaier, A., Faestermann, T., Gerstmann, U.C., Knie, K., Rugel, G.,

Wallner, A., Dillmann, I., Dollinger, G., von Gostomski, C.L., others, 2010. A new

value for the half-life of 10Be by heavy-ion elastic recoil detection and liquid

scintillation counting. Nucl. Instrum. Methods Phys. Res., Sect. B 268, 187–191.

Lal, D., Peters, B., 1967. Cosmic ray produced radioactivity on the Earth. Kosmische

Strahlung II/Cosmic Rays II, 551–612.

Latorre, C., Gonz´

alez, A.L., Quade, J., Fari˜

na, J.M., Pinto, R., Marquet, P.A., 2011.

Establishment and formation of fog-dependent Tillandsia landbeckii dunes in the

Atacama Desert: evidence from radiocarbon and stable isotopes. J. Geophys. Res.

116.

Lebatard, A.-E., Bourl`

es, D.L., Braucher, R., Arnold, M., Duringer, P., Jolivet, M.,

Moussa, A., Deschamps, P., Roquin, C., Carcaillet, J., others, 2010. Application of the

authigenic 10Be/9Be dating method to continental sediments: reconstruction of the

Mio-Pleistocene sedimentary sequence in the early hominid fossiliferous areas of the

northern Chad Basin. Earth Planet. Sci. Lett. 297, 57–70.

Liu, J., Nowaczyk, N.R., Panovska, S., Korte, M., Arz, H.W., 2020. The Norwegian-

Greenland Sea, the Laschamps, and the Mono Lake Excursions Recorded in a Black

Sea Sedimentary Sequence Spanning from 68.9 to 14.5 ka. Journal of Geophysical

Research: Solid. Earth 125 e2019JB019225.

Liu, L., Yu, K., Li, A., Zhang, C., Wang, Le, Liu, X., Lan, J., 2023. Weathering Intensity

Response to Climate Change on Decadal Scales: A Record of Rb/Sr Ratios from

Chaonaqiu Lake Sediments. Western Chinese Loess Plateau, Water, p. 15.

Maldonado, A., Betancourt, J.L., Latorre, C., Villagran, C., 2005. Pollen analyses from a

50 000-yr rodent midden series in the southern Atacama Desert (25◦30

′

S). J. Quat.

Sci. 20, 493–507.

Mariotti, A., 1983. Atmospheric nitrogen is a reliable standard for natural 15N

abundance measurements. Nature 303, 685–687.

Michalski, G., B¨

ohlke, J.K., Thiemens, M., 2004. Long term atmospheric deposition as the

source of nitrate and other salts in the Atacama Desert, Chile: New evidence from

mass-independent oxygen isotopic compositions. Geochim. Cosmochim. Acta 68,

4023–4038.

Mix, A.C., Tiedemann, R., Blum, P., 2003. Proceedings of the Ocean Drilling Program

Initial Reports Leg 202.

Murray, A.S., Olley, J.M., 2002. Precision and accuracy in the optically stimulated

luminescence dating of sedimentary quartz: a status review. Geochronometria:

Journal on Methods &Applications of Absolute Chronology 21.

Owen, J.J., Dietrich, W.E., Nishiizumi, K., Chong, G., Amundson, R., 2013. Zebra stripes

in the Atacama Desert: Fossil evidence of overland flow. Geomorphology 182,

157–172.

Pav´

on-Carrasco, F.J., Osete, M.L., Torta, J.M., Santis, A.P.D., 2014. A geomagnetic field

model for the Holocene based on archaeomagnetic and lava flow data. Earth Planet.

Sci. Lett. 388, 98–109.

Pfeiffer, M., Latorre, C., Santoro, C.M., Gayo, E.M., Rojas, R., Carrevedo, M.L.,

McRostie, V.B., Finstad, K.M., Heimsath, A., Jungers, M.C., de Pol-Holz, R.,

Amundson, R., 2018. Chronology, stratigraphy and hydrological modelling of

extensive wetlands and paleolakes in the hyperarid core of the Atacama Desert

during the late quaternary. Quat. Sci. Rev. 197, 224–245.

Pfeiffer, M., Morgan, A., Heimsath, A., Jordan, T., Howard, A., Amundson, R., 2021.

Century scale rainfall in the absolute Atacama Desert: Landscape response and

implications for past and future rainfall. Quat. Sci. Rev. 254, 106797.

Placzek, C., Quade, J., Betancourt, J.L., Patchett, P.J., Rech, J.A., Latorre, C.,

Matmon, A., Holmgren, C., English, N.B., 2009. Climate in the dry Central Andes

OVER geologic, millennial, and interannual timescales 1. Ann. Mo. Bot. Gard. 96,

386–397.

Placzek, C., Granger, D.E., Matmon, A., Quade, J., Ryb, U., 2014. Geomorphic process

rates in the Central Atacama Desert, Chile: Insights from cosmogenic nuclides and

implications for the onset of hyperaridity. Am. J. Sci. 314, 1462–1512.

Quade, J., Rech, J.A., Latorre, C., Betancourt, J.L., Gleeson, E., Kalin, M.T.K., 2007. Soils

at the hyperarid margin: the isotopic composition of soil carbonate from the

Atacama Desert, Northern Chile. Geochim. Cosmochim. Acta 71, 3772–3795.

F.L. Arens et al. Global and Planetary Change 242 (2024) 104576

Quade, J., Rech, J.A., Betancourt, J.L., Latorre, C., Quade, B., Rylander, K.A., Fisher, T.,

2008. Paleowetlands and regional climate change in the Central Atacama Desert,

northern Chile. Quat. Res. 69, 343–360.

Raisbeck, G.M., Yiou, F., 1984. Production of long-lived cosmogenic nuclei and their

applications. Nucl. Inst. Methods Phys. Res. B 5, 91–99.

Ravelo, A.C., Andreasen, D.H., Lyle, M., Olivarez Lyle, A., Wara, M.W., 2004. Regional

climate shifts caused by gradual global cooling in the Pliocene epoch. Nature 429,

263–267.

Rech, J.A., Currie, B.S., Michalski, G., Cowan, A.M., 2006. Neogene climate change and

uplift in the Atacama Desert, Chile. Geol 34, 761–764.

Ritter, B., Binnie, S.A., Stuart, F.M., Wennrich, V., Dunai, T.J., 2018a. Evidence for

multiple Plio-Pleistocene lake episodes in the hyperarid Atacama Desert. Quat.

Geochronol. 44, 1–12.

Ritter, B., Stuart, F.M., Binnie, S.A., Gerdes, A., Wennrich, V., Dunai, T.J., 2018b.

Neogene fluvial landscape evolution in the hyperarid core of the Atacama Desert.

Sci. Rep. 8, 13952.

Ritter, B., Wennrich, V., Medialdea, A., Brill, D., King, G., Schneiderwind, S.,

Niemann, K., Fern´

andez-Galego, E., Diederich, J., Rolf, C., Bao, R., Melles, M.,

Dunai, T.J., 2019. Climatic fluctuations in the hyperarid core of the Atacama Desert

during the past 215 ka. Sci. Rep. 9, 5270.

Ruff, M., Fahrni, S., G¨

aggeler, H.W., Hajdas, I., Suter, M., Synal, H.-A., Szidat, S.,

Wacker, L., 2010. On-line Radiocarbon Measurements of Small Samples using

Elemental Analyzer and MICADAS Gas Ion Source. Radiocarbon 52, 1645–1656.

S´

aez, A., Godfrey, L.V., Herrera, C., Chong, G., Pueyo, J.J., 2016. Timing of wet episodes

in Atacama Desert over the last 15 ka. The Groundwater Discharge Deposits (GWD)

from Domeyko Range at 25◦S. Quat. Sci. Rev. 145, 82–93.

Sager, C., Airo, A., Arens, F.L., Rabethge, C., Schulze-Makuch, D., 2020. New types of

boulder accumulations in the hyper-arid Atacama Desert. Geomorphology 350,

106897.

Sanchez, C., Regard, V., Carretier, S., Riquelme, R., Blard, P.-H., Campos, E., Brichau, S.,

Lupker, M., H´

erail, G., 2021. Neogene basin infilling from cosmogenic nuclides (10

be and 21 Ne) in Atacama, Chile: Implications for palaeoclimate and supergene

copper mineralization. Basin Res. 33, 2549–2571.

Schulze-Makuch, D., Wagner, D., Kounaves, S.P., Mangelsdorf, K., Devine, K.G., de

Vera, J.-P., Schmitt-Kopplin, P., Grossart, H.-P., Parro, V., Kaupenjohann, M.,

Galy, A., Schneider, B., Airo, A., Fr¨

osler, J., Davila, A.F., Arens, F.L., C´

aceres, L.,

Cornejo, F.S., Carrizo, D., Dartnell, L., DiRuggiero, J., Flury, M., Ganzert, L.,

Gessner, M.O., Grathwohl, P., Guan, L., Heinz, J., Hess, M., Keppler, F., Maus, D.,

McKay, C.P., Meckenstock, R.U., Montgomery, W., Oberlin, E.A., Probst, A.J.,

S´

aenz, J.S., Sattler, T., Schirmack, J., Sephton, M.A., Schloter, M., Uhl, J.,

Valenzuela, B., Vestergaard, G., W¨

ormer, L., Zamorano, P., 2018. Transitory

microbial habitat in the hyperarid Atacama Desert. Proc. Natl. Acad. Sci. USA 115,

2670–2675.

Steier, P., Martschini, M., Buchriegler, J., Feige, J., Lachner, J., Merchel, S.,

Michlmayr, L., Priller, A., Rugel, G., Schmidt, E., others, 2019. Comparison of

methods for the detection of 10Be with AMS and a new approach based on a silicon

nitride foil stack. Int. J. Mass Spectrom. 444, 116175.

Stuiver, M., Reimer, P.J., 1993. Extended 14C data base and revised CALIB 3.0 14C age

calibration program. Radiocarbon 35, 215–230.

Vandenberghe, J., 2013. Grain size of fine-grained windblown sediment: a powerful

proxy for process identification. Earth Sci. Rev. 121, 18–30.

Vargas, G., Rutllant, J., Ortlieb, L., 2006. ENSO tropical–extratropical climate

teleconnections and mechanisms for Holocene debris flows along the hyperarid coast

of western South America (17◦–24◦S). Earth Planet. Sci. Lett. 249, 467–483.

Voigt, C., Klipsch, S., Herwartz, D., Chong, G., Staubwasser, M., 2020. The spatial

distribution of soluble salts in the surface soil of the Atacama Desert and their

relationship to hyperaridity. Glob. Planet. Chang. 184, 103077.

von Blanckenburg, F., Bouchez, J., Wittmann, H., 2012. Earth surface erosion and

weathering from the 10Be (meteoric)/9Be ratio. Earth Planet. Sci. Lett. 351,

295–305.

Wacker, L., Fahrni, S.M., Hajdas, I., Moln´

ar, M., Synal, H.-A., Szidat, S., Zhang, Y., 2013.

A versatile gas interface for routine radiocarbon analysis with a gas ion source.

Nuclear Instruments &Methods in Physics Research Section B-beam Interactions

With Materials and Atoms 294, 315–319.

Wang, L., Ku, T.L., Luo, S., Southon, J.R., Kusakabe, M., 1996. 26Al-10Be systematics in

deep-sea sediments. Geochim. Cosmochim. Acta 60, 109–119.

Wang, F., Michalski, G., Seo, J.-H., Granger, D.E., Lifton, N., Caffee, M., 2015. Beryllium-

10 concentrations in the hyper-arid soils in the Atacama Desert, Chile: Implications

for arid soil formation rates and El Ni˜

no driven changes in Pliocene precipitation.

Geochim. Cosmochim. Acta 160, 227–242.

Wennrich, V., B¨

ohm, C., Brill, D., Carballeira, R., Hoffmeister, D., Jaeschke, A.,

Kerber, F., Maldonado, A., May, S.M., Olivares, Lester, Opitz, S., Rethemeyer, J.,

Reyers, Mark, Ritter, B., Schween, J.H., Sevinç, F., Steiner, J., Walber-Hellmann, K.,

Melles, M., 2024. Late Pleistocene to modern precipitation changes at the Paranal

clay pan, Central Atacama Desert. Glob. Planet. Chang. 233, 104349.

Willenbring, J.K., von Blanckenburg, F., 2010. Meteoric cosmogenic Beryllium-10

adsorbed to river sediment and soil: applications for Earth-surface dynamics. Earth

Sci. Rev. 98, 105–122.

Wittmann, H., von Blanckenburg, F., Bouchez, J., Dannhaus, N., Naumann, R.,

Christl, M., Gaillardet, J., 2012. The dependence of meteoric 10Be concentrations on

particle size in Amazon River bed sediment and the extraction of reactive 10Be/9Be

ratios. Chem. Geol. 318, 126–138.

Wittmann, H., von Blanckenburg, F., Dannhaus, N., Bouchez, J., Gaillardet, J., Guyot, J.-

L., Maurice, L., Roig, H., Filizola, N., Christl, M., 2015. A test of the cosmogenic 10Be

(meteoric)/9Be proxy for simultaneously determining basin-wide erosion rates,

denudation rates, and the degree of weathering in the Amazon basin. J. Geophys.

Res. Earth 120, 2498–2528.

F.L. Arens et al. Global and Planetary Change 242 (2024) 104576