Investigating hearth heating at Klasies River Cave during the Middle Stone Age in South Africa: insights from clumped isotopes thermometry on ostracod specimens from Klasies deposits.

Overview
This research project was developed in collaboration with the above participants. The project is designed as part of the Centre for Early Sapiens Behaviour (SapienCE) and Department of Earth Sciences at the University of Bergen (UiB) and will form part of a master’s project for Tessa Bosch. This project takes advantage of the high-quality laboratory facilities at the Department of Earth Sciences at UiB, which includes the ‘Facility for Advanced Isotopic Research and Monitoring of Weather, Climate and Biogeochmical Cycling’ (FARLAB) that hosts facilities for traditional and clumped isotope analyses, along side the close collaboration with the SapienCE project which allows the integration of data across a range of themes including palaeoclimate and archaeological records.

The Klasies River Main (KRM) site consists of a series of caves located in South Africa's Eastern Cape province. KRM has two caves (Cave 1 and Cave 2) and two associated overhangs which are known as Cave 1A and Cave 1B. Situated along the Tsitsikamma coast, the site lies between the Klasies River mouth, approximately 500 metres to the west, and Druipkelder Point, about 1 km to the east (Deacon and Geleijnse, 1988). KRM cuts into the cliffs off the Southwest Indian Ocean, forming a seaward edge of a coastal platform running along the Tsikskamma mountain range (Deacon and Geleijnse, 1988).
KRM was among the most intensively occupied sites in this Southern Cape coastal region between 120,000 and 55,000 years ago (Brenner et al., 2020). Archaeological excavations have uncovered remains showing occupation of anatomically modern humans (AMHs) and early evidence of marine and terrestrial resource utilisation (Grine et al., 2017; Rightmire and Deacon, 1991; Wurz et al., 2018). One notable example of coastal material utilisation at KRM is a 21-meter-deep shell midden within the Shell Midden One (SMONE) layer (Figure 1), which formed as a result of activities by hunter-gatherer-fishers (Wurz et al., 2018). The layers below SMONE which are called Black Occupational Series (BOS)-One. BOS-Two and BOS-Three (Figure 1) in the Sand and Shale Lower (SASL) sub-member, consist of dark moist clayey soil with abundant quartzite artefacts, fragmented charcoal, bone and some shellfish (Bentsen and Wurz, 2017). The Silty Black Soils (SBLS) sits below these layers.Finally, there is evidence of fire-related behaviour at KRM. Previous excavations at Klasies River have uncovered intact combustion features, particularly in the Witness Baulk excavation (Figure 2) within the Sand and Shell Upper (SASU) sub-member, as well as in other areas of the site (Deacon and Geleijnse, 1988; Henderson, 1992; Singer and Wymer, 1982). Unlike some other MSA sites, these combustion features are not generally bordered by hearthstones (Bentsen 2014) and no hearthstones thus far have been documented at KRM (Henderson, 1992). Analyses of faunal, botanical, micromorphological, and spatial evidence indicate that early humans at KRM consumed a diverse diet, including shellfish, plants, fish, and both marine and terrestrial mammals (Henderson, 1992; Klein, 1976; Larbey et al., 2019; Thackeray, 1988; Wurz et al., 2018). It is highly likely that some of these resources were cooked, e.g. the burnt Perna perna shells in a hearth feature indicating cooking (Henderson, 1992) as well as direct evidence pointing to the cooking of starchy tubers in hearths dating to approximately 120,000 to 65,000 years ago (Larbey et al., 2019).Within the BOS layers, there are potentially heated rock fragments displaying signs of heat exposure e.g. fracturing (Bentsen and Wurz, 2019). These potentially heat-affected quartzite from KRM look similar to colour changes from intentionally heated quartzite in fire experiments from Bentsen and Wurz (2017, 2019). These rocks would have been deliberately brought into the cave and repeatedly exposed to heat, leading to visible color changes, fractures and breaks (Bentsen and Wurz, 2019). While in the fire, these rocks would have helped retain heat, allowing the fire to burn for extended periods. This may suggest either longer occupations or a different group dynamic compared to those associated with smaller, shorter-lived fires. The previous research highlights the complexity and changeability of fire-related behaviours during the MSA.
This project contributes towards the ongoing work in understanding the use of fire and in determining potential hearth locations at Klasies River archaeological site. The samples selected for this study are drawn from both the SMONE and BOS layers. By analysing ostracods spatially distributed across layers containing artifacts that have been potentially heated and those that have not been potentially heated, it becomes possible to further the work on identifying the locations of hearths and related combustion features. This will be achieved by using clumped isotopes (see Methodology below). By examining the spatial relationships of ostracods within these layers, alongside previous analyses of heat-related features (e.g. Bentsen and Wurz, 2019), we aim to uncover insights into fire-related behaviors. This approach could also shed light on the relationship between fire technology and other MSA technologies through comparitive studies (e.g. Bentsen & Wurz, 2017; 2019).Aim/Objective
This project aims to use clumped isotopes analyses on ostracod valves from the sediment layers at KRM to help augment our understanding on the deposits at Klasies likely subject to heating by fires and their spatial extent. The working hypothesis is that the clumped isotope data will reflect a temperature signal for fire use in the site.
Clumped isotope thermometry can be used to determine whether and to what extent the ostracod carbonate material was heated above the resetting temperature of the original signal (Müller et al., 2017; Staudigel et al., 2019). Müller et al. (2017) presented heating experiments on bivalve shells and fish otoliths and demonstrated that heating treatment can be reconstructed with clumped isotope analyses. This research will apply the same approach to ostracod valves to increase our understanding of the effect of heating on their clumped isotope signature, aiding our interpretation of the archaeological ostracod valve signatures as well as the future application of this technique to archaeology. It can also provide important contextual information to other SapienCE researchers working at KRM.

First, ground truthing will need to be completed to better understand the clumped isotope signature when ostracods are heated. Therefore, we plan on using modern ostracod samples from South Africa to test the effect of heating on clumped isotope signature in ostracod samples. We plan to orchestrate heating tests on modern ostracods from South Africa (see Schmitz et al., 2024) at varying temperatures in an oven. We will then analyse these non-heated and heated modern ostracods, belonging to the same species as those found in the Klasies deposits. After the clumped isotope heating signature has been established, we will then apply this knowledge to the archaeological ostracods from KRM to determine heating extent.


These findings on heating signatures in both the modern and archaeological ostracods may provide data to other researchers working at other cave sites across the South African coastline such as Klipdrift Shelter, particularly for constraining hearth locations. Aligned with this project, clumped isotope analysis on archaeological opercula from the Gastropod species Turbo sarmaticus is being carried out by Ellie Pryor and colleagues to understand potential cooking practices at Blombos Cave site. This masters project may improve our knowledge on whether Ostracods could provide an additional heating indicator, concerning on-going work identifying potential hearth locations within both Blombos Cave and KRM sites. Methodology
Ostracods within the deposits from layers SMONE and BOS-ONE BOS-TWO and BOS-THREE (Figure 1) forming part the SASL sub-member will be used for clumped isotopes analyses. These samples have been previously examined for the presence of microfossils by Inèz Faul (Faul, 2021). The selected samples have been sieved at > 150 μm to examine for microfossils. This fraction contains a high abundance of ostracods (specifically from the species Gomphocythere obtusata) with abundance data showing the highest number of individual species within the BOS layers (Faul, 2021).
Clumped Isotopes Thermometry
This project takes advantage of the laboratory facilities at the Department of Earth Sciences at UiB, which includes the FARLAB facilities for traditional and clumped isotope analyses and the SapienCE project. Clumped isotope geochemistry is concerned with the state of ordering of rare isotopes in natural materials (Eiler, 2007). It refers to the overabundance of the 13C-18O-16O isotopologue of CO2. The surplus of this isotopologue is temperature-dependent, and it is reported with the parameter Δ47; see Eiler (2007). At equilibrium, the tendency for heavy isotopes to bond (or clump) decreases with increasing temperature. This temperature dependence allows clumped isotope compositions to serve as thermometers, providing insights into the formation temperatures of minerals and other substances. However, the formation temperature signal will be partially reset when material is heated above approximately 120 °C. This resetting behaviour is exploited in this project as it can be used as indication for strong heating, whereby experimental evidence shows the extent of resetting to increase with temperature (Müller et al, 2017, and own results). A disadvantage of clumped isotope thermometry is that each measurement is associated with large uncertainties caused by a low signal-to-noise ratio. To work around this, many replicate measurements (typically 25-30) of around 80 µg each are needed so substantial sample amounts are required. It has been verified that high enough abundances of ostracods are present in the selected material (Figure 3).
Rationale for Requested Export
This study cannot be conducted in South Africa as the analytic labs for sample preparation and analysis are located at the University of Bergen (UiB), Norway.
The permit request is for the destructive analysis and export permit.
Mode of Transport
The samples (Figure 3) will be transported via DHL courier services (arranged by Inèz Faul and Ellie Pryor) using DHL Germany.