Paleoclimate reconstructions from speleothems from the De Hoop Nature Reserve, South Africa.

1. Rationale and objectives
The southern coast of South Africa is a key region to study the broader context of human
evolution with sites such as Blombos cave and Klipdrift shelter. In order to better understand the

links between climate and human evolution it is necessary to reconstruct past climate at high-
resolution near archeological sites. Speleothems (e.g. stalagmites) are formed in caves and act as

natural archives by recording past climate information such as temperature, hydroclimate and
even vegetation.As part as of the SapienCE research centre, we are analyzing speleothems collected in 2020
in Bloukrantz cave (Fig. 1; 34°27.557’S, 20°46.697’E) in the De Hoop Nature Reserve (SAHRA
permit ID 3125). These speleothem samples have proven quite unique not only by the multitude
of methods that can be applied, and hence the results that can be extracted (Maccali et al. 2023),
but also by the climate results themselves. Indeed, the first results displayed a series a short and
drier events with no associated changes in temperature. In addition, of the three thermometry
methods applied, two agree remarkably well while the third one shows a variable offset (Fig. 2).This third method, namely TEX86, rely on the biochemical fingerprint left by microbial
communities in the cave.Our objectives are twofold; our first objective is to explore West cave (34°27.049’S,
20°41.883’E) in the De Hoop Nature Reserve and subsample speleothems to potentially recover
samples to i) replicate the specific climate signal recorded in Bloukrantz’s samples and then
confirm the regionality of this signal and ii) extend the time range of our research ranging actually
from ∼ 45 to 90 ka (Fig. 3). Our second objective is to characterize the TEX86 signal specifically
for Bloukrantz cave. In order to identify the origin of the TEX86 signal in Bloukrantz cave
speleothems, both cave sediments and soil samples above the cave will be analyzed.2. Methods
Speleothems can be accurately and precisely dated by the U-Th radiometric technique that
rely on the measure of both the parent (Uranium) and daughter (Thorium) in samples. The
method is destructive as the samples need to be dissolved in acid medium before being analyzed.
U-Th dating requires radioactive material and trained staff to handle them. The samples will be
prepared in the Uranium laboratory at the University of Bergen and analyzed in the ICP laboratory
with a NuTM Plasma MC-ICP-MS.
In order to minimize damages to the speleothem samples two cores were drilled (∼2 cm wide,
∼10 cm long) at the base of two speleothems. Similarly, when coring was not possible, the base
and top of speleothems were carefully chipped off to limit damages to the samples. A total of 4
speleothems have hence been subsampled to determine the age period of their growth.
Speleothems preserve small amounts of organic matter sourced either from the overlying soil
or from the cave environment. A group of lipids derived from microorganisms, namely Glycerol
Dialkyl Glycerol Tetraethers – GDGTs, have been measured in speleothems and have shown a
strong correlation with surface air temperature. Although uncertainties on the source of the
GDGTs exists, recent studies indicate that they could be derived from microbial communities in
the cave itself (Baker et al., 2019). These microbial communities adapt to changing conditions
(e.g. changes in cave temperature) leaving distinct biochemical signatures. Samples from
Bloukrantz cave have recently been analyzed by colleagues at the Max Planck Institute for
Chemistry (MPIC) in Mainz, including the topmost sample as part of a calibration study for the
TEX86 paleothermometer in speleothems (Martinez-Garcia et al., 2022). The unusually high
amounts of GDGTs in our samples could point to the presence of a rich speleothem archaeal
community or transport of soil-derived bacterial signal into the cave, making this location distinct
from others used in the calibration. Indeed, the estimated preliminary TEX86 temperature for the
late Holocene sample is ~3.5°C lower than the other temperature methods, and the estimated
TEX86 temperatures for the older samples are on average ~7°C lower, with apparent systematic
variations over time (Fig. 3). By analyzing the cave sediment and overlying soil, we aim to further
investigate the origin of the TEX86 signal in Bloukrantz cave speleothem. The method is
destructive as it necessitates a chemical extraction of the bacteria from the soil/sediment.
Samples will be prepared and analyzed at the MPIC facility in Mainz.3. Duration and expected results
Results are expected within a year. The U-Th dating will be performed by the end of the
summer 2023. Based on the age range calculated, the speleothem could be collected for further
paleoclimate investigations in the coming years.
TEX86 analyses will be performed within a year and the results will further trace the origin of
the TEX86 signal in Bloukrantz cave speleothems and hence refine the TEX86 thermometry
method.