lable at ScienceDirect Quaternary Science Reviews 306 (2023) 108030 Contents lists avai Quaternary Science Reviews journal homepage: www.elsevier .com/locate/quascirev Evidence for large land snail cooking and consumption at Border Cave c. 170e70 ka ago. Implications for the evolution of human diet and social behaviour Marine Wojcieszak a, b, *, Lucinda Backwell a, c, d, Francesco d’Errico e, f, Lyn Wadley a a Evolutionary Studies Institute (ESI), University of the Witwatersrand, Private Bag 3, WITS, 2050, Johannesburg, South Africa b Royal Institute for Cultural Heritage (RICH, KIK/IRPA), 1000, Brussels, Belgium c Grupo de Investigaci�on en Arqueología Andina (ARQAND e CONICET), Facultad de Ciencias Naturales e Instituto Miguel Lillo, Universidad Nacional de Tucum�an, Miguel Lillo 205, San Miguel de Tucum�an, Tucum�an, T4000, Argentina d Centre of Exploration for the Deep Human Journey, University of the Witwatersrand, Private Bag 3, WITS, 2050, South Africa e Universit�e de Bordeaux, CNRS, MCC, PACEA, UMR 5199, Allee Geoffroy Saint Hilaire, CS 50023, F - 33615, Pessac CEDEX, Talence, France f Centre for Early Sapiens Behaviour, Øysteinsgate 3, Postboks 7805, 5020, University of Bergen, Norway a r t i c l e i n f o Article history: Received 28 December 2021 Received in revised form 21 December 2022 Accepted 6 February 2023 Available online xxx Handling Editor: Paloma de la Pe~na Keywords: Land snail shell Infrared and Raman spectroscopy Scanning electron microscopy Experimental archaeology Broad spectrum revolution * Corresponding author. Evolutionary Studies Inst Witwatersrand, Private Bag 3, WITS, 2050, Johannesb E-mail address: marine.wojcieszak@kikirpa.be (M https://doi.org/10.1016/j.quascirev.2023.108030 0277-3791/© 2023 Elsevier Ltd. All rights reserved. a b s t r a c t Fragments of land snail (Achatinidae) shell were found at Border Cave in varying proportions in all archaeological members, with the exception of the oldest members 5 WA and 6 BS (>227,000 years ago). They were recovered in relatively high frequencies in Members 4 WA, 4 BS, 1 RGBS and 3 WA. The shell fragments present a range of colours from lustrous beige to brown and matt grey. The colour variability can occur when shell is heated. This possibility was explored here through experimental heating of giant land snail shell fragments (Achatinidae, Metachatina kraussi - brown lipped agate snail) in a muffle furnace from 200 to 550 �C for different lengths of time. Colour change, weight loss, and shattering of the heated samples were recorded. Transformation of aragonite into calcite and the occurrence of organic material was investigated by means of Infrared and Raman spectroscopy. Scanning electron microscopy was also used on selected specimens to help identify heat-induced transformation as opposed to taphonomic alteration. The identification on archaeological fragments of features produced by experi- mentally heating shells at high temperatures or for long periods has led us, after discarding alternative hypotheses, to conclude that large African land snails were systematically brought to the site by humans, roasted and consumed, starting from 170,000 years ago and, more intensively between 160,000 and 70,000 years ago. Border Cave is at present the earliest known site at which this subsistence strategy is recorded. Previous research has shown that charred whole rhizomes and fragments of edible Hypoxis angustifolia were also brought to Border Cave to be roasted and shared at the site. Thus, evidence from both the rhizomes and snails in Border Cave supports an interpretation of members of the group pro- visioning others at a home base, which gives us a glimpse into the complex social life of early Homo sapiens. © 2023 Elsevier Ltd. All rights reserved. 1. Introduction When, how, and why terrestrial molluscs became a constituent part of the diet of our ancestors are questions that are still largely unanswered and that puzzle researchers in many respects. Inver- tebrate animals represent more than 95% of earth's biodiversity itute (ESI), University of the urg, South Africa. . Wojcieszak). (Herbert and Kilburn, 2004), but they are often not studied in archaeological assemblages because they are considered marginal for the understanding of past human behaviour (Mannino, 2019). In addition, most of them are small and have little chance of surviving in the archaeological record. One of the most common remains is from soft bodied animals with mineralised exoskeletons; molluscs. Land snails may be present in archaeological sites because they occur naturally in the sediments or because they were introduced by the occupants to consume their flesh and/or to use their shells as rawmaterials or as body ornaments (Mannino, 2019). Although the mailto:marine.wojcieszak@kikirpa.be http://crossmark.crossref.org/dialog/?doi=10.1016/j.quascirev.2023.108030&domain=pdf www.sciencedirect.com/science/journal/02773791 http://www.elsevier.com/locate/quascirev https://doi.org/10.1016/j.quascirev.2023.108030 https://doi.org/10.1016/j.quascirev.2023.108030 M. Wojcieszak, L. Backwell, Francesco d’Errico et al. Quaternary Science Reviews 306 (2023) 108030 shell was often used for bead-making in African Iron Age contexts in the late Holocene, older symbolic use may have occurred too, as evidenced by an Achatina shell fragment interpreted as engraved with a criss-cross pattern, recently found at Txina-Txina, southern Mozambique, in layers dated 27.4e26.8 ka cal BP (Bicho et al., 2018). Many terrestrial molluscs bury themselves deep in the soil to avoid dehydration during dry winters (Herbert and Kilburn, 2004), so they can be found naturally in sediments, or they can be brought to a site by their numerous predators. Adverse climatic events can also cause the death of snails in their natural habitats and conse- quently their shells may be incorporated in cave and shelter de- posits through geological processes. Multiple uses of land snail shell have been investigated from an ethnoarchaeological point of view: including use as tools for agricultural, fishing, hunting or household purposes, or as decorations or ritual implements (Lubell, 2004a; Mannino, 2019). Snail shells have been discov- ered in numerous African sites, including Border Cave, the subject of this volume. At Haua Fteah Cave in Libya, North Africa, preliminary analysis suggests that small land snails were consumed occasionally from 40 to 15 ka cal BP, andmore regularly from 14 to 10 ka cal BP (Barker et al., 2012). The site of Taforalt (Grotte des Pigeons) in Morocco, Northwest Africa, records land snail consumption from 13 to 11 ka BP (Taylor et al., 2011). In Beirut in the Levant the site of Ksar ’Aqil has yielded an accumulation of land snails dated 23e22 ka BP (Mellars and Tixier, 1989). In Europe, Klissoura Cave 1 in Greece records land snails with broken lips that are interpreted as evidence of human consumption, in levels dated between 36 and 28 ka cal BP (Kuhn et al., 2010). Nearby, Franchthi Cave documents land snail consumption between 15 and 12 ka cal BP (Stiner and Munro, 2011). The site of Cova de la Barriada in Spain preserves compel- ling evidence for land snail consumption in Europe from 31.3 to 26.9 ka cal BP (Fern�andez-L�opez de Pablo et al., 2014). Here land snails were found in three different levels in association with combustion features, lithic and faunal assemblages. Charcoal analysis revealed that the snails were roasted in ambers of pine and juniper. X-Ray diffraction analysis showed that all specimens pre- sented aragonite (CaCO3) as the only mineral phase, indicating that the heating temperature was insufficient to promote the formation of calcite (Parker et al., 2010). The site of Cueva de Nerja, also in Spain, records land snails throughout the stratigraphic sequence, including Gravettian (28e19 ka) and Solutrean (22e17 ka) deposits (Aura Tortosa et al., 2012; Jord�a et al., 2011), while the Italian site of Grotta della Serratura evidences land snail consumption in Epi- gravettian layers dated 14.1e13.7 ka yrs cal BP (Martini et al., 2009). Edible land snails are often abundant in late Pleistocene and Ho- locene archaeological sites, and have been found in the Caribbean, Peru, parts of North America, East Africa, Sudan, Nigeria, the Philippines and throughout the Mediterranean region, where they become a substantial part of the human diet (Lubell, 2004b). The consumption of land snails continues today in the Mediterranean Basin in Spain, France, Italy, Portugal, Algeria, Morocco, and Tunisia. Land snails are also consumed in Nepal, Southeast Asia and Northeast India, and there is a growing demand in South America. The above review of the evidence highlights, however, what can be perceived as a conundrum in the history of our lineage's adap- tation. Terrestrial molluscs are an excellent source of nutrients, they are easy and not dangerous to collect, they can be stored for some time before being consumed, they are simple to prepare and to digest as long as one has a basic mastery of fire, certainly acquired by hominins at least 400,000 years ago. However, and contrary to what one might expect, the systematic consumption of terrestrial molluscs does not seem to date back to before 49 ka in Africa and 36 ka in Europe, and does not become an essential component of the 2 diet of most human populations before 15e10 ka. Although terrestrial molluscs, especially those of small dimension, are sub- ject in some sedimentary contexts to taphonomic processes that can damage or destroy them, it is difficult to attribute the recent exploitation of terrestrial molluscs to taphonomic causes. For this reason, the dramatic intrusion of this food source in the late Pleistocene or, depending on the region, early Holocene, has traditionally been interpreted as a consequence of an expansion of subsistence resources that would have developed at that time, probably due to demographic pressure. This “broad spectrum rev- olution” (Flannery, 1969) would have aimed at the systematic exploitation of all available resources, and in particular small prey such as small mammals, fish, reptiles and molluscs (Flannery, 1969; Stiner, 2001; Zeder, 2012). This vision has tended to attribute possible evidence of older exploitation of terrestrial molluscs to the many natural phenomena that can lead the remains of these ani- mals to be incorporated into sedimentary sequences or, at most, to sporadic consumption that does not substantially change the classical vision of the evolution of the diet of our genus. It is quite possible, however, that this mainstream narrative reflects only the best-documented and archaeologically visible part of a more complex process that has involved, since the mastery of fire, the exploitation of certain nutritious species that were only available in certain ecological settings. It is also possible that these subsistence strategies were lost in some cases and reacquired in others. Because of its particular characteristics, Achatina (giant African land snail) may have been a continuously exploited species in certain regions of Africa, andmay have been an important part of the diet of Middle Stone Age (MSA) populations long before the “broad spectrum revolution”. This point of view is supported, but not really sub- stantiated, by archaeological evidence that predates our work at Border Cave. The earliest possible habitual consumption of land snails in Africa, from Mumba-H€ole Cave in Tanzania, East Africa, from level V upper (Lubell, 2004b), dated to 49.1 ± 4.3 ka (Gliganic et al., 2012) consists of Achatina sp. shells found in association with lithic artefacts (Mehlman, 1979), though a description of the snail assemblage is yet to be published. An even older possible con- sumption of this species comes from the Kenyan site of Panga Ya Saidi (Martin�on Torres et al., 2021). These authors report large frequencies of Achatina sp. fragments, many of which bear traces of heating in the form of distinct blackening, in layers dated c. 78 ka, in which a burial of a three-year-old child was discovered. More recent evidence for possible MSA exploitation of Achatina comes from Sibudu Cave (Plug, 2004), Bushman Rock Shelter (Badenhorst and Plug, 2012) and Kuumbi Cave, Zanzibar (Faulkner et al., 2019; Shipton et al., 2016). The Achatinidae family incorporates the largest land snails and KwaZulu-Natal has a relatively rich agate snail fauna, including 15 species (Herbert and Kilburn, 2004). They are herbivores, but also consume soil, calcareous rocks, the shells of other molluscs and bone from carcasses to form their own shells made of the aragonite calcium carbonate (CaCO3) polymorph (Herbert and Kilburn, 2004). This large gastropod is still a popular food inWest African countries including Cameroon, Ghana and Nigeria (Griveaud, 2016), but not in South Africa among rural people living in Nkungwini near Border Cave (personal communication by B. Vilane to L.W. in 2016). In some parts of Africa, Achatina snails are believed to contain various curative properties, and the bluish liquid (haemolymph) obtained from the shell reportedly helps with infant development (Ugwumba et al., 2016; Munywoki, 2022). Snail meat is highly nutritious; it is rich in protein, iron, potassium, phosphorous, magnesium, selenium, calcium, copper, sodium, zinc, Omega-3 and other essential fatty acids, and various vitamins, including vitamin A, Thiamine (B1), Riboflavin (B2), Niacin (B3), vitamin B6, B12, C, D, E and K, while being low in fat (Aboua, 1990; Fagbuaro et al., 2006). M. Wojcieszak, L. Backwell, Francesco d’Errico et al. Quaternary Science Reviews 306 (2023) 108030 An adult Achatina snail may have a mass of close to a kilogram, so it provides a substantial protein source. Recent studies (Dumidae et al., 2019) suggest that humans can contract a parasitic disease caused by a worm hosted by the giant African land snail. Careful handling may prevent disease. Of course, we do not know whether the parasite is a modern phenomenon or whether humans in the past were also vulnerable to disease when handling the snails. The objective of this study is to use the results of new high- resolution excavations at Border Cave and appropriate analytical methods to test the hypothesis that Achatina snails were consumed systematically by human groups in Border Cave between about 170 ka and 70 ka ago (and occasionally thereafter) and discuss the role that this giant land snail may have played in early modern human cultural adaptations. 1.1. Background to the site Border Cave is located on a cliff, at the border between KwaZulu- Natal in South Africa and eSwatini (Fig. 1). It occurs in the Lebombo mountain range at an elevation of c. 600 m above sea level. The shelter is semi-circular in shape, approximately 50 mwide by 35 m long, and faces West. The site has been extensively excavated: first, in 1934 by Dart, but his findings were never published. In 1940 Horton extracted a vast amount of deposit, supposedly for agri- cultural purposes, which left a large pit in the middle of the site. The following year Cooke and colleagues resumed archaeological excavations and discovered an infant burial (BC3) with a Conus shell originally covered in ochre (d’Errico and Backwell, 2016), and Fig. 1. Location of Border Cave and a site plan showing the position of the various excavation excavations conducted from 2015 to 2019 along the North wall of excavation (EXC.) 3 A an according to North and East lines. Stratigraphic members and associated cultural attributions and off-white colours denote Brown Sand and White Ash members, and the sloped dividin 3 recovered additional human remains from Horton's dump, which they attributed to the Middle Stone Age (Cooke et al., 1945). Extensive excavations were conducted during the 1970s and 1980s by Beaumont and Todd and Miller (Beaumont, 1978, 1980; Beaumont et al., 1992), which led to the naming of brown sand (BS) and white ash (WA) members (Fig. 1) that make up the deposit (Butzer et al., 1978), and dating of the stratigraphic sequence (Vogel and Beaumont, 1972; Beaumont et al., 1978, 1992; Beaumont, 1980; Vogel et al., 1986; Grün and Beaumont, 2001; Grün et al., 2003; Bird et al., 2003; Millard, 2006). In recent years radiocarbon dating has been applied to archaeological finds (d’Errico et al., 2012; Villa et al., 2012; Backwell et al., 2018), and now Tribolo and colleagues (Tribolo et al., 2022) present optically stimulated luminescence ages for the sedimentary sequence and a suite of Bayesian model ages for the deposits. Bayesian model ages include all published electron spin resonance and 14C ages. Results are given at 2 sigma (95%), and the highest posterior density is given at 95%. Member 5 WA ranges between ~250 and 160 ka;Member 5 BS ranges between ~176 and 149 ka; Member 4 WA ranges between 167 and 100; Member 4 BS ranges between ~93 and 69 ka; Member 1 RGBS ranges between ~79 and 67 ka; Member 3 WA ranges between ~74 and 62 ka; Member 3 BS ranges between ~69 and 56 ka; Member 2 WA ranges between ~61 and 53 ka; Member 2 BS ranges between ~57 and 43 ka; Member 1 WA ranges between ~45 and 41 ka and Member 1 BS ranges between 43 and 22 ka. The most recent round of excavations started in 2015 under the direction of Backwell, Wadley and d’Errico (Backwell et al., 2018). Their research has confirmed that the cave records traces of s from 1934 to 2019. The orange overlay shows the position of point-plotted artefacts in d South wall of EXC. 4 A. The inset (top right) provides the square names excavated after Beaumont et al. (1992) are shown in the column on the right. The alternating tan g lines represent the dip of the deposit. M. Wojcieszak, L. Backwell, Francesco d’Errico et al. Quaternary Science Reviews 306 (2023) 108030 intermittent occupation from before 227 ka ago until 24 ka, capped by Iron Age occupation dated to about 600 BP. Their fine-scale approach identifies 10 layers where Beaumont saw one for Mem- ber 1 BS, and 15 inMember 1WA. They identify 69 layers within his single Member 2 WA and 51 in 4 WA. As opposed to the 10 mem- bers identified by Beaumont, 237 stratigraphic layers have been identified and excavated to date (Backwell et al., 2022; Stratford et al., 2022). Beaumont (1978) originally reported remarkable preservation of organic remains and the new excavations have confirmed this, with 48 layers of grass mats/bedding identified (Sievers et al., 2022). Grass bedding in Member 5 BS dated at about 170 ka, and Member 5 WA, dated to >227 ka makes it the oldest known (Wadley et al., 2020a; Esteban et al., 2023). The same is true for whole cooked starchy rhizomes, identified as Hypoxis angusti- folia, from layers dated to c. 170 ka ago (Wadley et al., 2020b). Charcoal is preserved throughout the sequence, providing a rare and comprehensive record of climate change in the region (Lennox et al., 2022). In addition to burning wood for cooking, warmth and protection against predators, the charcoal record suggests that plant species may have been selected for their latex, nutritional value, insecticidal, medicinal, and cosmetic properties (Zwane and Bamford, 2021; Lennox et al., 2022). 1.2. The Achatina record During the latest excavations, from 2015 to 2019, shell frag- ments of giant land snails (Fig. 2) were found throughout the sequence, except in the oldest members 5 WA and 6 BS. They were found in relatively high frequencies in Members 4WA, 4 BS, 1 RGBS and 3WA. Beaumont's extensive excavation of the site yielded 3529 pieces of Achatina shell, representing a minimum number of 84 individuals (Beaumont, 1978), with the highest number in Member 4 BS (n¼ 1076), as is the case with recent excavations (see Backwell et al., 2022; Fig. 7). Backwell and colleagues (Backwell et al., 2022) postulate that between about 170 ka and 70 ka ago land snails may have been a regular part of the diet of visitors to the cave, and that more recent occupants ate them only occasionally. They acknowl- edge that the land snails may have buried themselves in the cave deposit, but consider human consumption as a more likely expla- nation for their presence in the deposits because the Achatina shell assemblage is extremely large and mono-specific, the pieces are often associated with grass mats, combustion features, lithic and Fig. 2. Modern Achatinidae land snail. Such agate snails may reach a length of about 160 mm (Herbert and Kilburn, 2004). 4 faunal remains, and the shells are fragmented, with a mixture of what they consider to be burned and unburned pieces based on state of preservation and colour, though the differences may result from non-human taphonomic processes. The concentrated accu- mulation of shells in specific stratigraphic units (see Backwell et al., 2022; Fig. 7) does not conform to a random distribution of shells scattered at various depths, as would be the case if the snails had buried themselves in the deposit. The fragments vary in colour from lustrous beige to brown and matt grey. Accidental or deliberate heating of shells is said to ac- count for changes observed in the colour and chemistry of exca- vated gastropods (Lange et al., 2008; Bonizzoni et al., 2009; d'Errico et al., 2015; Milano et al., 2016, 2018). In particular, the shell of Achatina, which is essentially composed of superimposed layers of aragonite, undergoes a mineralogical phase transition into calcite, associated with morphological changes at microscopic and macroscopic scale, if the shell is submitted to substantial heating. However, change in colour and chemical composition can also be produced by non-human taphonomic processes (d’Errico et al., 2015), and this deserves further enquiry. In order to test the hy- pothesis that people at Border Cave cooked and consumed land snails, we used an experimental approach to explore colour and chemical changes to modern Metachatina kraussi shell fragments heated in a muffle furnace at various temperatures and for different lengths of time. The chemical reaction through heating was studied using vibrational spectroscopy. The macroscopic and microscopic appearance of the heated shells was documented using photo- graphs, optical and scanning electron microscopy. The results were compared to a sample of land snail shells from Border Cave in an attempt to match the features and chemical signatures resulting from heat, and to establish whether the archaeological specimens might have been similarly modified. 2. Material and methods 2.1. Experimental sample For the experimental sample, a modern brown-lipped agate snail (Achatinidae, Metachatina kraussi) shell from the Verulum region in KwaZulu-Natal was used. It was chosen because it rep- resents the same family of snails preserved in the archaeological deposits at Border Cave. The shell ranges in colour from off-white, to beige to brown (Herbert and Kilburn, 2004). It was broken into multiple pieces using a chunk of rhyolite found on the path leading to Border Cave. Pieces withmany different shapes, sizes and colours were obtained and 41 of them were selected and used to perform heating experiments. Fig. S1 shows photographs of the giant snail shell before and after crushing as well as the rhyolite used to break it. 2.2. Archaeological sample The archaeological sample studied comes from the recent ex- cavations conducted by Backwell and colleagues, and more pre- cisely from the 2017 excavation. The shell fragments are attributed to the Family Achatinidae. The attribution was made based on the relatively large size of the pieces, their curvature, and pattern made of wavy brown axial lines or stripes. Even though the sample is fragmented, we attribute the shells to Metachatina kraussi because of the pale wavy axial lines that are distinct from the bold straight ones characteristic of Achatina immaculata, the only other agate snail found in the region (Herbert and Kilburn, 2004). From the 87 ziplock bags containing 1654 shell fragments, 27 pieces showing a range of colours and state of preservationwere selected for analysis using Raman spectroscopy. Fig. 3 shows photographs of a sample of Fig. 3. Examples of Border Cave shell fragments and their stratigraphic context. a) Member 4 BS, layer Chocolate Brown Ena plan 1; b) Member 4 BS, layer Dark Brown Elmo plan 1 (Combustion Feature base) sample 3; c) Member 4 BS, layer Chocolate Brown Ena (Combustion Feature 1); d) Member 4 BS, layer Chocolate Brown Ena plan 2; e) Member 5 BS, layer Very Dark Greyish Brown Jim plan 4 (Combustion Feature) sample not selected for analysis; f) Member 4 BS, layer Dark Brown Elmo plan 1 (Combustion Feature base) sample 1; g) Member 5 BS, layer Very Dark Greyish Brown Jan sample 4; h) Member 5 BS, layer Very Dark Greyish Brown Jim sample 3 (Combustion Feature); i) Member 4 BS, layer Grass mat Ega sample 1 (lip); j) Member 5 BS, layer Very Dark Greyish Brown Jim plan 4 (Combustion Feature) sample 1 (Combustion Feature base); k) Member 4 BS, layer Dark Grey Eduardo plan 2 (Combustion Feature) sample 2; l) Member 5 BS, layer Very Dark Greyish Brown Jan plan 4 sample 1. All scale bars represent 1 cm. M. Wojcieszak, L. Backwell, Francesco d’Errico et al. Quaternary Science Reviews 306 (2023) 108030 the selected specimens fromMembers 4 BS (93.5e69.8 ka) and 5 BS (176.8e149.1 ka). Of the 27 samples selected, six specimens did not yield spectra, or the spectra recorded did not allow for carbonate polymorph discrimination using the non-invasive Raman analysis and these were subjected to invasive (but non-destructive) Infrared analysis. Table 1 lists the selected ziplock bags with the number of frag- ments, their colours, specific features (such as the presence of a lip) and the results of the analyses. A preliminary identification of heated specimens was based on the sample colour and state (cracks, if it was brittle, etc.) The full table of samples from the 2017 excavation can be found in Table S1. 2.3. Heating experiments Three experiments were performed with a muffle furnace. The temperature varied from 200 to 550 �C and the heating time from 5 min to 36 h. The first experiment was performed to determine at which temperature the mineralogical transformation from arago- nite to calcite and colour change start when heated for a short period of time (5 min) from 200 �C to 550 �C with a 50 �C incre- ment. For each temperature step three fragments of different colour were used. The fragments also exhibited different thicknesses, which can also influence the transformation. The second experi- ment explored heat exposure for an increased period of time (20 min) starting from 300 �C (since the lower temperatures showed limited change during the first experiment) to 550 �C. Two fragments per temperature step were heated to check reproduc- ibility. The last experiment focused on heat exposure time at a set temperature (350 �C) ranging from 1 h to 36 h. The rationale being that if archaeological fragments of shell were present around a fire, they could have been exposed to heat for long periods of time. The fragments were inserted when the furnace reached the desired temperature. Different shell fragments were used depending on their colour (white, beige, or brown) and some of them were cleaned with water and a soft brush to clearly see the colour change of the shell. Not all were cleaned since shell was unlikely to have been cleaned in the past. The weight loss, established at a precision 5 of 0.001 g, and the colour changewere recorded, and the specimens were photographed before and after heating. Table 2 lists the different experiments (temperature, duration, initial colour and if the sample was cleaned or not) with their respective results (colour and appearance after heating, weight loss, if the sample shattered or not and the Raman and FTIR results). 2.4. Optical microscopy Optical microscopic observations were conducted using an Olympus BX63 upright microscopewith reflected light and CellSem Dimension software. The z-stacking option of the software was used with magnification objectives of 4 � and 10 � . 2.5. Raman micro-spectroscopy Raman spectroscopy can help to determine the molecular composition of the samples. Here, all experimental samples (n ¼ 45) and archaeological specimens (n ¼ 27) were analysed on both sides, on multiple spots with a green laser covering a very localised surface area of around 1 mm. Raman spectra were recor- ded using a LabRam HR 800 spectrometer (Horiba JobineYvon), with a 514.5 nm line of a Lexel argon ion laser, equipped with an Olympus BX41 microscope attachment. A charge coupled detector cooled with liquid nitrogen collected the spectra with a spectral resolution <2 cm�1 using the 600 lines/mm grating. The spec- trometer was set to save the 80e1800 cm�1 spectral range, and the laser beam power directed at the sample was kept under 0.8 mW to avoid inducing thermal transformation. The spectra were acquired with a 100x long working distance giving a spatial resolution of around 1 mm for 8e350 s, repeated at least twice. 2.6. Infrared spectroscopy Fourier Transform Infrared spectroscopy (FTIR) was performed with a Bruker Alpha spectrometer equipped with an Attenuated Total Reflectance (ATR) module. A small piece (few millimetres wide) of the selected specimen was broken and finely ground in an Table 1 List of archaeological shell fragments excavated in 2017 and selected for the analysis. Each line refers to a ziplock bag, fromwhich one to four specimenswere selected for study. Cultural attributions according to Beaumont et al. (1992). ATR-FTIR: Attenuated Total Reflectance Fourier Transform Infrared Spectroscopy; AC: amorphous carbon; BS: brown sand; CF: combustion feature; MnOx: manganese oxide; Orga: organic matter; WA: white ash; ELSA: Early Later Stone Age. Cultural attribution Member Layer Plan Number of fragments Colour Signs of heating Note/Samples selected Raman ATR-FTIR ELSA 1 WA Grass Mat Betty 1 1 beige possibly Lip Calcite e MSA III 2 WA Very Dark Brown Dax 1 1 beige possibly e Fluorescence Bone (Apatite) MSA II 1 RGBS Very Dark Brown Eba 2 2 grey possibly Sample 1 Calcite, AC e Sample 2 Calcite, Gypsum e MSA I 4 BS Dark Grey Eduardo (CF) 2 15 beige possibly Sample 1 Aragonite, MnOx, Gypsum e Sample 2 Aragonite, AC e Sample 3 Aragonite e 4 BS Dark Grey Eduardo 2 3 1 grey, 2 beige possibly Sample 1 (beige) Aragonite e Sample 2 (grey) Calcite, AC e Sample 3 (beige) Aragonite e 4 BS Grass Mat Ega 1 2 beige no e Aragonite Aragonite, Orga, Apatite, silicates 4 BS Grass Mat Ega 1 1 beige/ brown possibly Lip Aragonite, Carotenoid Aragonite, Orga, Apatite, Silicates 4 BS Dark Brown Elmo (CF base) 1 19 10 grey, 9 beige possibly Sample 1 (grey/black) Aragonite, Calcite, MnOx, AC e Sample 2 (grey) Calcite, AC e Sample 3 (beige) Aragonite e 4 BS Chocolate Brown Ena 1 1 beige no Scratches, striations on the outer part, shiney both sides Aragonite e 4 BS Chocolate Brown Ena (CF1) 1 1 beige no Cracks visible macroscopically on the inner part Aragonite, AC e 5 BS Very Dark Greyish Brown ? 1 beige possibly e Aragonite e 5 BS Very Dark Greyish Brown Jan 1 1 grey ? e Calcite e 5 BS Very Dark Greyish Brown Jan (CF1) 3 1 beige possibly e Carbonate Calcite, Silicates (sediments) 5 BS Very Dark Greyish Brown Jan 4 2 grey possibly Sample 1 Bone e Sample 2 Calcite, AC e 5 BS Very Dark Greyish Brown Jim (CF) 3 1 beige possibly e Calcite e 5 BS Very Dark Greyish Brown Jim (CF) 4 6 2 grey, 4 beige possibly Sample 1 (beige) Carbonate Calcite, Silicates, Gypsum Sample 2 (grey) Calcite, AC e Sample 3 (grey) Fluorescence Bone (Apatite), Gypsum Sample 4 (beige) Aragonite e M. Wojcieszak, L. Backwell, Francesco d’Errico et al. Quaternary Science Reviews 306 (2023) 108030 agate mortar to obtain a homogeneous and fine powder that was pressed onto the diamond crystal. Spectra were acquired in the mid-IR range between 400 and 4000 cm�1 (Deuterated TriGlycine Sulfate detector) with 64 scans and a 4 cm�1 resolution. An at- mospheric compensationwas applied to the spectra using OPUS 7.5 software. 2.7. Scanning electron microscopy Scanning electron microscope (SEM) observations were made and images captured with a Phenom Pure generation 5 instrument from Thermo Fisher Scientific. The samples were placed on adhe- sive carbon tape. The backscattered images were acquired at a working distance of ~10 mm at 5 kV. SEM observations were per- formed on the inner and outer parts of a white experimental con- trol specimen not heated, a grey piece heated at 500 �C for 20 min, and on the inner part of three archaeological samples with different compositions. These samples are a beige sample composed of aragonite from Member 5 BS, layer Very Dark Greyish Brown Jan, not showing any signs of heating macroscopically (accession number 3038), another beige sample fromMember 5 BS (accession number 5256) composed of calcite and thus likely heated, and sample 1 (grey/black) from Member 4 BS, layer Dark Brown Elmo 6 (combustion feature base) exhibiting a mixture of calcite and aragonite. 3. Results 3.1. Experimental heating and macroscopic observations The weight loss for samples heated at the lower temperature (200e300 �C) and for a short period of time oscillated around 0.5% that likely corresponds to the loss of water. For samples heated for longer periods and at higher temperatures, the weight loss was around 1e2% and probably entailed the entire or partial loss of organic matter (Table 2). In the case of land snails, the carbonate crystals that constitute the shell are bound in an organic framework of tanned proteins that protect the shell against humic acids pre- sent in the soil (Herbert and Kilburn, 2004). Starting from 450 �C, some samples shattered and/or there was a separation of the different layers comprising the shell (Fig. S2). The effect of heat on colour was that the originally white specimens became a more snow white, while the beige and brown samples lost their colour (Table 2, Fig. 4). Starting from500 �C for at least 20min, the samples turned grey. After heating, most of the specimens lost their initial gloss and became matt. This is because the organic matter present Table 2 List of heating experiments in terms of parameters and results. ATR-FTIR: Attenuated Total Reflectance Fourier Transform Infrared Spectroscopy; AC: amorphous carbon; bh: before heating; ah: after heating; MnOx:manganese oxide; Orga: organic matter; N: no; Y: yes; ?: indeterminate compound; min: minutes; h: hours. The weight loss could not be measured for the samples which shattered (since many fragments were scattered). Temperature (�C) Heating time Sample colour bh Cleaned Sample colour & appearance ah Weight loss (%) Shattered Raman ATR-FTIR 200 5 min white x white 0.00 N Aragonite, Carotenoid Aragonite beige beige 0.13 N Aragonite Aragonite, Orga brown brown 0.12 N Aragonite, Carotenoid, ? Aragonite, Orga 250 white white 0.47 N Aragonite, Carotenoid, AC, Feldspar, Roman�echite Aragonite beige beige 0.07 N Aragonite, Carotenoid, AC, ? Aragonite, Orga brown light brown 0.27 N Aragonite, Carotenoid Aragonite, Orga 300 white white 0.50 N Aragonite, Haematite Aragonite beige beige 0.28 N Aragonite, AC Aragonite, Orga brown light brown 0.28 N Aragonite, Carotenoid, AC Aragonite, Orga 350 white white with darkening of brown areas 0.47 N Aragonite, AC Aragonite beige beige with darkening of brown areas 0.70 N Aragonite, AC Aragonite, Orga brown brown became grey & darkening of brown areas 0.46 N Aragonite Aragonite, Orga 400 white white 0.43 N Aragonite, AC Aragonite beige (a bit of brown) beige & brown became black 0.49 N Aragonite, AC, ? Aragonite, Orga brown brown became beige & is black on some areas 0.47 N Aragonite, AC, ? Aragonite, Orga 450 white white 0.51 N Aragonite, AC, ? Aragonite beige white (darkening on some areas) 0.83 N Aragonite, AC, ? Aragonite, Orga brown white 0.73 N Aragonite, AC Aragonite, Orga 500 white whiter 1.49 N Aragonite, Calcite, AC, MnOx, Anatase Calcite, Aragonite beige (a bit of brown) white (brown became black) e Y Aragonite, Calcite, AC, Haematite, ? Calcite brown white (brown became black) e Y Aragonite, Calcite, AC Calcite, Aragonite 550 white whiter e Y Calcite Calcite beige white (bit of brown left) e Y Aragonite, Calcite Calcite, Aragonite brown white/grey e Y Aragonite, Calcite, AC Calcite, Aragonite 300 20 min beige x darker outside, whiter inside 0.00 N Aragonite, AC, Maghemite Aragonite beige/brown & green spots e darker & green became brown 0.00 N Aragonite, AC, MnOx Aragonite, Orga 350 beige (a bit of brown) x whiter, brown became black 0.35 N Aragonite, Carotenoid, AC, ? Aragonite white & green spots e whiter, green became black/brown 0.53 N Aragonite, AC, ? Aragonite, Orga 400 white x whiter 0.21 N Aragonite, AC, ? Aragonite, Calcite white/brown e whiter, brown became black 0.31 N Aragonite, AC, ? Aragonite, Orga 450 beige x very white, separation of layers (at least 3 layers) 1.43 N Aragonite, MnOx Calcite beige/brown e white, brown became black, layer separation (at least 4) 1.29 N Aragonite, Calcite, AC Calcite, Aragonite 500 white x whiter, layer separation & cracking 1.90 N Aragonite, Calcite, MnOx Calcite, Aragonite beige/brown e white, slightly grey outside, layer separation, cracking 1.63 N Calcite, MnOx, Maghemite Calcite þ Aragonite 550 white x grey/white 1.22 N Calcite, MnOx Calcite, Aragonite white e grey/white e Y Calcite, AC Calcite, Aragonite 350 1 h beige/brown e whiter, brown became black 0.49 Y Aragonite, AC, MnOx, ? Aragonite, Orga 3 h beige/white e beige/whiter, outer layer detached from the inner ones 1.35 N Aragonite, AC, Haematite, Quartz, Anatase Aragonite, Orga 6 h beige/brown e whiter, brown became black 1.40 Y Aragonite, AC, Anatase, ? Aragonite, Calcite 12 h beige/brown e very white inner part, sample broke & layers bent in opposite directions 1.04 N Aragonite, Calcite, AC Aragonite, Calcite 36 h beige/brown e disappearance of many brown areas, the colour became matt, cracks, outer layer detached from the inner ones 1.19 N Calcite, AC, MnOx Aragonite, Calcite M. Wojcieszak, L. Backwell, Francesco d’Errico et al. Quaternary Science Reviews 306 (2023) 108030 in the shell structure is responsible for the glossy appearance (Nouet, 2014). Hence, the loss of gloss corresponds with the loss of organic matter, as confirmed by vibrational spectroscopy, which did not detect organic matter in matt samples. Of the 1654 archaeological fragments unearthed in 2017, 82.6% exhibit a beige colour, 15.3% are grey, 1.9% brown and only 0.2% are black (Table S1 and Fig. 3). Several samples show some black stainingwhich could be due tomanganese oxides andmany display cracks. Some of the beige specimens appear very glossy, while others are matt. The grey, brown, and black ones are always matt. 7 The loss of organic matter that makes shell appear glossy may be attributed to heat, although microbial attack of organic matter and diagenesis could also be possible causes. 3.2. Microscopic observations At a microscopic scale it is possible to observe in detail the ridges present on the in- and outside of themodern shell fragments (Fig. 5a). These correspond to the successive growth increments of the shell (Nouet, 2014). The same growth ridges are clearly visible Fig. 4. Representative colour change of the experimental Metachatina kraussi shell fragments through heating. The top left specimen was cleaned and considered as beige before heating. The bottom left specimen was not cleaned, and the adhering organic matter turned black and stained the shell pale brown. The top right specimen was considered brown before heating and lost its colour, and the bottom right specimen was white before heating and turned matt grey. M. Wojcieszak, L. Backwell, Francesco d’Errico et al. Quaternary Science Reviews 306 (2023) 108030 on the archaeological specimens (Fig. 5b) and they help to confirm our identification of the snail shell. Cracks can be observed on the modern heated specimens, which also lack gloss and have a matt surface (Fig. 5c). The same features are observed on the archaeo- logical specimens, which are similarly cracked with a matt surface (Fig. 5d). Scanning electron microscope analysis of modern heated shells reveals a plethora of desiccation cracks resulting from rapid water loss (Fig. 5e). Similar microscopic cracking of the surface was observed on archaeological specimen (Fig. 5f). SEM also revealed fusion of a shell's original microstructure after it was heated at high temperatures such as 500e700 �C (Milano et al., 2016, 2018). Other taphonomic processes caused by localised dissolution, fusion and recrystallization processes can also be observed with SEM (Toffolo, 2021) but were not observed on the Border Cave specimens. Microscopic analysis of the modern heated shells and archaeo- logical specimens from Border Cave shows that they share features resulting from exposure to heat, namely micro-cracking and a matt surface appearance. This finding, and the fact that most of the archaeological specimens derive from combustion features (see Table 1) demonstrates that these shell fragments were most likely heated. 3.3. Raman micro-spectroscopy The spectra of compounds detected on the experimental and archaeological specimens are presented in Fig. 6, and Tables 1 and 2 list the compounds obtained for each specimen analysed. Before heating, only the calcium carbonate polymorph of aragonite was detected on the fragments of the brown-lipped agate snail shell (Table 2). It is one of the polymorphs of calcium carbonate (CaCO3) with characteristic Raman peaks located around 152, 179, 189, 212, 272, 282, 702 and 1086 cm�1 (Urmos et al., 1991). On the brown parts of the shell fragments before heating, carotenoids were identified with vibrational bands located around 875, 958, 1004, 1156, 1191, 1285, 1355, 1392, 1447 and 1514 cm�1 (Maguregui et al., 2012). Spectra of amorphous carbon, quartz, haematite, maghe- mite, feldspar and manganese oxides (Mernagh, 1991; Chamritski and Burns, 2005; Prinsloo et al., 2013; Bernardini et al., 2019; Wojcieszak and Wadley, 2019) were also recorded on the experi- mental shell before the heating experiment. These minerals likely come from soil particles present on the surface of the shell frag- ments. After heating (Table 2), carotenoids were still detected on shell fragments heated for less than 20 min at 350 �C. They dis- appeared after this threshold. The transformation of aragonite into 8 calcite was detected only in fragments heated for 5min at 500 �C, in those heated for 20min above 450 �C, and those heated for 12 h at a temperature of 350 �C. Similar transition ranges in temperature and time were obtained on marine mollusc shells (Lange et al., 2008; Parker et al., 2010; d'Errico et al., 2015; Milano et al., 2016; Müller et al., 2017; Milano et al., 2018). Calcite is the most stable form of calcium carbonate at ambient conditions and can be distinguished from aragonite by the lower intensity bands in the wavenumber range below 720 cm�1. Calcite characteristic Raman bands are located around 155, 282, 713 and 1086 cm�1 (Urmos et al., 1991). The aragonite to calcite transformation through heating depends on the temperature and the type of aragonite, it can take a few days at low temperature (~100 �C) or a few minutes at higher temper- atures (>470 �C). For biogenic aragonite, the presence of intra- crystalline molecules intercalated in the atomic structure influ- ence the temperature of transition (Parker et al., 2010). For the archaeological samples (Table 1), only amorphous car- bon, gypsum, manganese oxide, aragonite and calcite were detec- ted. Amorphous carbon probably comes from the absorption of carbon from organic material surrounding the shell during the heating process (d’Errico et al., 2015) and from sediment particles. Gypsum could come from the sedimentary deposit and taphonomic processes that lead to the breakdown of organic matter, such as the introduction of fresh plant material like grass used for bedding, wood used for fuel and guano from birds and bats (Schiegl and Conard, 2006). Manganese oxides may correspond to black stains present on the surface of some fragments (Fig. 3). Dendritic struc- tures, typical of manganese oxides (Potter and Rossman, 1979), were observed on an archaeological specimen using SEM-Energy Dispersive Spectroscopy (EDS) (Fig. S3). Of the 27 archaeological specimens analysed, 12 contained aragonite. One specimen found in a layer dated to c. 60 ka still contains carotenoids. Calcite was recorded on nine specimens, often in combinationwith amorphous carbon; one was a mixture of both, and another was identified as bone. Because of high fluorescence, two samples produced spectra with no identifiable features and two others only produced spectra identifying calcium carbonate, with peaks around 1084 cm�1, but no bands distinguishing calcite from aragonite. 3.4. ATR-FTIR spectroscopy FTIR spectroscopy provides information on the molecular composition of the main compounds present in a sample; com- pounds present in low proportionwill not be detected because they Fig. 5. Microscopic images of modern (left) and archaeological (right) land snail shells. a) Growth ridges on a modern unmodified shell fragment, and b) similar growth ridges on an archaeological specimen from Member 3 BS, layer Dark Brown Elmo, plan 1 combustion feature base. c) Cracks observed on a cleaned modern shell fragment heated at 550 �C for 20 min, and d) similar irregular cracks recorded on an archaeological specimen from Member 5 BS, layer Very Dark Greyish Brown Jim, plan 3 (combustion feature). e) Uncleaned modern shell fragment heated at 500 �C for 20 min showing extreme micro-cracking of the desiccated surface, and f) archaeological specimen from Member 5 BS, layer Very Dark Greyish Brown Jim, plan 3 (combustion feature - accession number 5256) showing similar fine cracking over the entire surface of the piece. Scales in a-d ¼ 200 mm; scales in e, f ¼ 20 mm. M. Wojcieszak, L. Backwell, Francesco d’Errico et al. Quaternary Science Reviews 306 (2023) 108030 are covered by the more intense signal of the main compounds. For the experimental specimens (Fig. 7a, Table 2), only two compounds were detected on unheated samples: aragonite with characteristic bands around 700, 712, 857, 1083, 1459, 1786, 2523, 2947 and 2978 cm�1 (Tomi�c et al., 2010) and organic matter with the CH stretching vibration at 2914 cm�1. A mixture of aragonite and calcite was recorded on samples experimentally heated for 5 min at 500 �C, on samples heated for 20 min at 400 �C, and for 6 h at 350 �C. The spectra with both aragonite and calcite signals exhibit a more pronounced asymmetry of the more intense vibrational band around 1400 cm�1 in addition to a doublet at 859 and 874 cm�1, clearly indicating the presence of both compounds. The charac- teristic band of organic matter was detected in coloured (beige or brown) samples heated at the lowest temperatures and/or for the shortest times. Carotenoids could therefore be responsible for this signal. One white sample also showed evidence of organic matter, 9 but it was an uncleaned specimen, so the reading could be due to organic matter adhering to the surface of the shell. Only three specimens out of the 41 heated pieces identified only calcite, with bands around 712, 873, 1403, 1795, 2511, 2872 and 2977 cm�1 (Shillito et al., 2009; Tomi�c et al., 2010). These fragments were relatively thin. The above results suggest that the transformation of aragonite into calcite through heating depends on several param- eters, such as the dimension and shape of the specimens, and the duration and temperature of heating, as previously mentioned. It also depends on the species analysed (Yoshioka and Kitano, 1985), hence the need for species-specific reference samples. The four archaeological specimens (see above) that did not provide identifiable Raman spectra, and two additional ones, were analysed by means of ATR-FTIR (Fig. 7b). Two of them were iden- tified as bone. Two others revealed the presence of aragonite and organic matter (CH vibrations at 2916 cm�1 - data not shown) in Fig. 6. Representative Raman spectra recorded on Metachatina kraussi experimental shell fragments and Border Cave archaeological shell fragments. Tables 1 and 2 list the spectra obtained for each specimen analysed. Most of the spectra presented were recorded on heated experimental pieces, except for the spectrum of aragonite (centre), which was obtained from an unheated experimental specimen, and the spectrum of gypsum (top), which was acquired on an archaeological specimen. M. Wojcieszak, L. Backwell, Francesco d’Errico et al. Quaternary Science Reviews 306 (2023) 108030 addition to apatite and silicates, likely originating from sediment. The last two specimens contained calcite and silicates, and one of them gypsum. Fig. 7. Representative ATR-FTIR spectra recorded on a) modern Metachatina kraussi experimental shell fragments. The spectrum at the bottom corresponds to a white fragment heated at 550 �C for 5 min; the spectrum in the middle to a fragment heated at 350 �C for 6 h, and the top one to a fragment heated at 350 �C for 1 h, b) spectra obtained from Border Cave shell fragments of giant land snails. 3.5. The stratigraphic distribution of Achatinidae fragments During the excavations conducted between 2015 and 2019 snail fragments were found in all archaeological members with the exception of the deepest ones (Members 5 WA and 6 BS). The fragments are particularly abundant inMember 4 BS and, to a lesser extent, in adjacent Members 4 WA, and 1 RGBS (Fig. 8). The more recent archaeological deposits (Members 2 WA, 2 BS, 1 WA), on the other hand, show a significant reduction in number, and here most of the pieces were found on grass mats. Shell fragments from the older members are abundant in layers with combustion features. In Member 5 BS two thirds of the fragments found in 2017 came from combustion features, while a quarter of the pieces found in Mem- ber 4 BS were found in them. In this member the highest number of fragments (n¼ 252) found in one layer derive from a grassmat. This 10 supports the hypothesis that the snails were cooked in combustion features and associated with human activity but does not eliminate the possibility that the observed variations are also related to the availability of the resource in the environment or to cultural changes. Fig. 8. Bar chart showing the number of Achatina shells plotted in situ per member from 2015 to 2019. M. Wojcieszak, L. Backwell, Francesco d’Errico et al. Quaternary Science Reviews 306 (2023) 108030 4. Discussion and conclusion Our experimental results and those from previous studies indicate that shells of gastropods composed of aragonitic layers undergo a change of mineral phase when heated at high temper- ature or for a long time. The transformation of aragonite into calcite is generally associated with change in colour, the appearance of microscopic cracks, the loss of water and shell organic component, weight loss, and, at microscopic scale, the recrystallization of cal- cium carbonate from aragonite to calcite. Land snail heating for consumption does not necessarily produce a mineral phase change and the above listed features. Experiments with large edible Helix sp. indicate that roasting them on embers for 5e8minwas themost plausible technique carried out at Pupi�cina cave, Croatia. This technique leaves barely visible traces of burning (Rizner et al., 2009) and no mineral phase change when the shells were ana- lysed using XRD (Fern�andez-L�opez de Pablo et al., 2014). However, the samples analysed presented some changes in the texture of aragonite, detectable from the relative intensities of the peaks in the diffractograms. No complete or almost complete Achatina shells were recovered at Border Cave, including in the layers with few combustion fea- tures. This contradicts the possibility that live snails buried them- selves in the sediments and were accidentally heated when humans made fireplaces at the site. Shells were not brought to the site to be used as raw material, for example for the production of beads or tools. No personal ornaments and artefacts made of these shells, nor production waste have been found at Border Cave. The hypothesis that Achatinawere brought alive to the site to be cooked and consumed is consistent with the fact that fragments of this family are abundant in layers with combustion features. The pres- ence of fragments made of aragonite, a mixture of aragonite and calcite, and calcite, is also consistent with the consumption contention (although diagenesis cannot be excluded): the large size of the Achatina is such that once put to cook on embers some parts of the shell may not reach the temperature necessary for the transformation of the aragonite into calcite. In addition, as mentioned above, the Pupi�cina Cave record shows that land snails can be satisfactorily cooked at temperatures lower than those involving a mineral phase change. It is likely that, considering its larger size, to fully cook an Achatina themolluscmust be exposed to heat for longer periods of time, which leads to a transformation into calcite of only the lower part of the shell in contact with embers. 11 This would explain the mixture of burned and unburned shell fragments encountered in the same area during excavation, but add that this may be a result of diagenesis, which can be extremely localised. Combustion experiments have shown that the sediment below a fire can reach 300þ�C through vertical heat transfer (Werts and Jahren, 2007; Sievers and Wadley, 2008; Aldeias et al., 2016). This means that land snail shells in archaeological deposits, sepa- rated by thousands of years from fires made above them may be thermally altered, and exhibit the same features as intentionally roasted snail shells. We acknowledge that no distinction can be made between intentional or fortuitous heating, as was the case for the blackened Helix cincta land snail shells dating from the Meso- lithic Italian site of Edera Cave (Bonizzoni et al., 2009), and this may well have occurred with some pieces at Border Cave. However, baking of the sediment caused by overlying hearths made years later would have altered all of the fragments in a localised area, which is not the case; at Border Cave we find a mixture of trans- formed, partially transformed and unaltered shell fragments in close proximity. Heat transfer subsequent to deposition, which has probably had the effect of increasing the number of fragments made of calcite, does not explain how the snails arrived at the site in the first place, andwhy they aremore abundant in layers withmore combustion features. The most parsimonious explanation for the frequent presence of land snail shells in combustion features is that they were intentionally placed on coals, to cook them. Their occurrence in combustion features throughout the sequence, from about 170 ka ago, implies regular consumption of land snails at the site, with occupants of the shelter habitually eating them from 170 to 70 ka ago, and only occasionally thereafter, which makes Border Cave the earliest site at which the frequent consumption of land snails is recorded. This finding has implications for our view of the evolution of diet within our lineage. Our results suggest that long before the broad spectrum revolution, human populations may have consumed terrestrial molluscs and for long periods. At Border Cave, as at other Middle and Later Stone Age sites in southern and eastern Africa, it would appear that they chose a species of particular nutritional value. However, the search for and consumption of this species, with such particular characteristics, did not necessarily imply a reorganization of the subsistence strategies of the human groups that exploited it, as was certainly the casewith the advent of the broad spectrum revolution. The time dedicated to the collection of these molluscs did not impinge on that dedicated to hunting or other subsistence activities and did not require the development of new technologies. Furthermore, snails could be easily collected by all mobile members of a band, regardless of their age or gender. Future research should apply the research strategy followed at Border Cave to other African sites that have yielded Achatina frag- ments, in order to gain a clearer picture of the regions and periods in which this species was exploited. It would be particularly important to establish whether Achatina fragments appear in other archaeological records of about 170 ka ago. There is growing evi- dence for consumption of marine molluscs at coastal sites from southern Africa (Marean et al., 2007; Jerardino and Marean, 2010) and Europe (Zilh~ao et al., 2020) dated to MIS 6 and MIS 5. Evidence for the consumption of land snails in those and subsequent periods has however been scant. This introduces the possibility that different populations, living in different environments, developed distinct cultural adaptations, including or not the consumption of molluscs, with inland cultural adaptations crossing the tipping point for consumption of molluscs later than some coastal pop- ulations. These scenarios need to be tested in the future by research focusing on the shell fragments found at inland sites. This may shed light on whether changes in the quantities of snail remains may be due to differential preservation, environmental changes probably M. Wojcieszak, L. Backwell, Francesco d’Errico et al. Quaternary Science Reviews 306 (2023) 108030 reducing the availability of this resource during periods of increased aridity, or cultural change. It may be no coincidence that between about 170 and 100 ka ago, we also found many charred whole rhizomes and fragments of the edible Hypoxis angustifolia in Border Cave (Wadley et al., 2020b). The implications of this discovery reach far beyond the interesting fact that people so far back in time cooked starchy foods. The cave is perched on a cliff where Hypoxis spp. do not grow, so collectors of the edible rhizomes would have dug them elsewhere and transported them to the home base for cooking and sharing. If sharing had not been intended, the food could have been eaten in the field where it was recovered. Ethnographic records of hunter- gatherer food collection, such as those made by Marshall (1976), clearly illustrate that people eat a great many plants directly where they find them, and that hunters will also cook and eat some hunted meat in the field before carrying portions home. Thus, ev- idence from both the rhizomes and snails in Border Cave supports an interpretation of people in an economically active stage of life provisioning others (perhaps infirm bandmembers and very young children) at a home base. The evidence therefore reaches beyond subsistence strategies and gives us a glimpse into the potentially complex social life of early Homo sapiens. Author contributions M.W. and L.W. conceived the research. M.W. acquired and analysed the data and wrote the paper, with input from L.W., F.d’E. and L.B. L.W., F.d’E. and L.B. revised versions of the manuscript and made improvements to the text. Figures were produced by M.W. and L.B. All authors read and approved the final manuscript. 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. Acknowledgements This research was funded by a National Geographic Explorer grant (NGS-54810R-19), a DSI-NRF Centre of Excellence in Palae- osciences (Genus) grant (CEOOP 2020-1), and a National Research Foundation (NRF) of South Africa grant (#98824) to Lucinda Back- well. Francesco d’Errico was supported by the European Research Council (ERC) under the Herbert and Kilburn, 2004 program (QUANTA project, contract no. 951388), the Research Council of Norway through its Centres of Excellence funding scheme, SFF Centre for Early Sapiens Behaviour (SapienCE), project number 262618, the LaScArBx research programme (ANR-10-LABX-52), the Talents Programme [grant number: 191022_001], and the Grand Programme de Recherche ‘Human Past’ of the Initiative d’Excel- lence (IdEx) of theUniversity of Bordeaux. Lyn Wadley was funded by a NRF African Origins Platform grant (#98827). The Microscopy and Microanalysis Unit (University of the Witwatersrand) is acknowledged for the use of the Raman spectrometer and an op- tical microscope. We are grateful to Amafa for issuing us with the excavation permit (SAH 15/7645), and the Evolutionary Studies Institute at the University of the Witwatersrand for housing our field equipment and archaeological collections. We thank Alain Queffelec for helpful discussions during the revision of the manu- script, Bawinile Vilane, the Border Cave tour guide, and the local 12 Border Cave community for support during our field trips. We are grateful for the time taken by the anonymous reviewers to provide helpful comments and suggestions that improved the quality of this manuscript. Appendix A. Supplementary data Supplementary data related to this article can be found at https://doi.org/10.1016/j.quascirev.2023.108030. References Aboua, F., 1990. Chemical composition of Achatina fulica. Tropicultura 8 (3), 121e122. Aldeias, V., Dibble, H.L., Sandgathe, D., Goldberg, P., McPherron, S.J.P., 2016. How heat alters underlying deposits and implications for archaeological fire fea- tures: a controlled experiment. J. Archaeol. Sci. 67, 64e79. Aura Tortosa, J.E., Jord�a Pardo, J.F., P�erez Ripoll, M., Badal, E., Avezuela, B., Morales P�erez, J.V., Tiffagom, M., Wood, R., Marlasca Martín, R., 2012. El corredor costero meridional: los cazadores gravetienses de la cueva de Nerja (M�alaga, Espa~na). In: de las Heras Martín, C., Lasheras Corruchaga, J.A., Arrizabalaga Valbuena, A., de la Rasilla Vives, M. (Eds.), Pensando el Gravetiense: Nuevos Datos Para la Regi�on Cant�abrica en su Contexto Peninsular y Pirenaico. Ministerio de Edu- caci�on, Cultura y Deporte, pp. 104e113. Backwell, L.R., d'Errico, F., Banks, W.E., de la Pe~na, P., Sievers, C., Stratford, D., Lennox, S.J., Wojcieszak, M., Bordy, E.M., Bradfield, J., Wadley, L., 2018. New excavations at border cave, KwaZulu-natal, South Africa. J. Field Archaeol. 43 (6), 417e436. Backwell, L., Wadley, L., d'Errico, F., Banks, W., de la Pe~na, P., Stratford, D., Sievers, C., Laue, G., Vilane, B., Clark, J., Tribolo, C., Beaudet, A., Jashashvili, T., Carlson, K., Lennox, S., Esteban, I., Mauran, G., 2022. Border Cave: a 227,000-year-old archive from the southern African interior. Quat. Sci. Rev., 107597 Badenhorst, S., Plug, I., 2012. The faunal remains from the middle Stone age levels of bushman rock shelter in South Africa. S. Afr. Archaeol. Bull. 67, 16e31. Barker, G., Bennett, P., Farr, L., Hill, E., Hunt, C., Lucarini, G., Morales, J., Mutri, G., Prendergast, A., Pryor, A., Rabett, R., 2012. The cyrenaican prehistory project 2012: the fifth season of investigations of the Haua Fteah cave. Libyan Studies 43, 115e136. Beaumont, P.B., 1978. Border Cave. M.A. dissertation, University of Cape Town, South Africa. Beaumont, P.B., 1980. On the age of Border Cave hominids 1-5. Palaeontol. Afr. 23, 21e33. Beaumont, P.B., de Villiers, H., Vogel, J., 1978. Modern man in sub-Saharan Africa prior to 49 000 years BP: a review and evaluation with particular reference to Border Cave. S. Afr. J. Sci. 74 (11), 409e419. Beaumont, P.B., Miller, G., Vogel, J., 1992. Contemplating old clues to the impact of future greenhouse climates in South Africa. South Afr. J. Sci. 88, 490e498. Bernardini, S., Bellatreccia, F., Casanova Municchia, A., Della Ventura, G., Sodo, A., 2019. Raman spectra of natural manganese oxides. J. Raman Spectrosc. 50 (6), 873e888. Bicho, N., Cascalheira, J., Andr�e, L., Haws, J., Gomes, A., Gonçalves, C., Raja, M., Benedetti, M., 2018. Portable art and personal ornaments from txina-txina: a new later Stone age site in the limpopo river valley, southern Mozambique. Antiquity 92 (363). Bird, M.I., Fifield, L.K., Santos, G.M., Beaumont, P.B., Zhou, Y., Di Tada, M.L., Hausladen, P.A., 2003. Radiocarbon dating from 40e60 ka bp at Border Cave, South Africa. Quat. Sci. Rev. 22, 943e947. Bonizzoni, L., Bruni, S., Girod, A., Guglielmi, V., 2009. Archaeometric study of shells of helicidae from the Edera cave (northeastern Italy). Archaeometry 51 (1), 151e173. Butzer, K., Beaumont, P.B., Vogel, J.C., 1978. Lithostratigraphy of border cave, Kwa- Zulu, South Africa: a middle Stone age sequence beginning c. 195,000 B.P. J. Archaeol. Sci. 5 (4), 317e341. Chamritski, I., Burns, G., 2005. Infrared- and Raman-Active phonons of magnetite, maghemite, and hematite: a computer simulation and spectroscopic study. J. Phys. Chem. B 109 (11), 4965e4968. Cooke, H.B.S., Malan, B.D., Wells, L.H., 1945. Fossil man in the Lebombo mountains, South Africa: the ‘border cave’, ingwavuma district, zululand. Man 45 (3), 6e13. d'Errico, F., Backwell, L.R., 2016. Earliest evidence of personal ornaments associated with burial: the Conus shells from Border Cave. J. Hum. Evolution 93, 91e108. d'Errico, F., Backwell, L.R., Villa, P., Degano, I., Lucejko, J., Bamford, M., Higham, T., Colombini, M.P., Beaumont, P., 2012. Early evidence of San material culture represented by organic artifacts from Border Cave, South Africa. P. Natl. Acad. Sci. USA 109 (33), 13214e13219. d'Errico, F., Vanhaeren, M., Van Niekerk, K., Henshilwood, C.S., Erasmus, R.M., 2015. Assessing the accidental versus deliberate colour modification of shell beads: a case study on perforated Nassarius kraussianus from Blombos Cave Middle Stone Age levels. Archaeometry 57 (1), 51e76. Dumidae, A., Janthu, P., Subkrasae, C., Dekumyoy, P., Thanwisai, A., Vitta, A., 2019. Genetic characterization ofAngiostrongylus larvae and their intermediate host, Achatina fulica, in Thailand. PLoS One 14 (9), e0223257. https://doi.org/10.1016/j.quascirev.2023.108030 http://refhub.elsevier.com/S0277-3791(23)00078-1/sref1 http://refhub.elsevier.com/S0277-3791(23)00078-1/sref1 http://refhub.elsevier.com/S0277-3791(23)00078-1/sref1 http://refhub.elsevier.com/S0277-3791(23)00078-1/sref2 http://refhub.elsevier.com/S0277-3791(23)00078-1/sref2 http://refhub.elsevier.com/S0277-3791(23)00078-1/sref2 http://refhub.elsevier.com/S0277-3791(23)00078-1/sref2 http://refhub.elsevier.com/S0277-3791(23)00078-1/sref3 http://refhub.elsevier.com/S0277-3791(23)00078-1/sref3 http://refhub.elsevier.com/S0277-3791(23)00078-1/sref3 http://refhub.elsevier.com/S0277-3791(23)00078-1/sref3 http://refhub.elsevier.com/S0277-3791(23)00078-1/sref3 http://refhub.elsevier.com/S0277-3791(23)00078-1/sref3 http://refhub.elsevier.com/S0277-3791(23)00078-1/sref3 http://refhub.elsevier.com/S0277-3791(23)00078-1/sref3 http://refhub.elsevier.com/S0277-3791(23)00078-1/sref3 http://refhub.elsevier.com/S0277-3791(23)00078-1/sref3 http://refhub.elsevier.com/S0277-3791(23)00078-1/sref3 http://refhub.elsevier.com/S0277-3791(23)00078-1/sref3 http://refhub.elsevier.com/S0277-3791(23)00078-1/sref3 http://refhub.elsevier.com/S0277-3791(23)00078-1/sref3 http://refhub.elsevier.com/S0277-3791(23)00078-1/sref3 http://refhub.elsevier.com/S0277-3791(23)00078-1/sref3 http://refhub.elsevier.com/S0277-3791(23)00078-1/sref4 http://refhub.elsevier.com/S0277-3791(23)00078-1/sref4 http://refhub.elsevier.com/S0277-3791(23)00078-1/sref4 http://refhub.elsevier.com/S0277-3791(23)00078-1/sref4 http://refhub.elsevier.com/S0277-3791(23)00078-1/sref4 http://refhub.elsevier.com/S0277-3791(23)00078-1/sref4 http://refhub.elsevier.com/S0277-3791(23)00078-1/sref5 http://refhub.elsevier.com/S0277-3791(23)00078-1/sref5 http://refhub.elsevier.com/S0277-3791(23)00078-1/sref5 http://refhub.elsevier.com/S0277-3791(23)00078-1/sref5 http://refhub.elsevier.com/S0277-3791(23)00078-1/sref5 http://refhub.elsevier.com/S0277-3791(23)00078-1/sref6 http://refhub.elsevier.com/S0277-3791(23)00078-1/sref6 http://refhub.elsevier.com/S0277-3791(23)00078-1/sref6 http://refhub.elsevier.com/S0277-3791(23)00078-1/sref7 http://refhub.elsevier.com/S0277-3791(23)00078-1/sref7 http://refhub.elsevier.com/S0277-3791(23)00078-1/sref7 http://refhub.elsevier.com/S0277-3791(23)00078-1/sref7 http://refhub.elsevier.com/S0277-3791(23)00078-1/sref7 http://refhub.elsevier.com/S0277-3791(23)00078-1/sref8 http://refhub.elsevier.com/S0277-3791(23)00078-1/sref8 http://refhub.elsevier.com/S0277-3791(23)00078-1/sref9 http://refhub.elsevier.com/S0277-3791(23)00078-1/sref9 http://refhub.elsevier.com/S0277-3791(23)00078-1/sref9 http://refhub.elsevier.com/S0277-3791(23)00078-1/sref10 http://refhub.elsevier.com/S0277-3791(23)00078-1/sref10 http://refhub.elsevier.com/S0277-3791(23)00078-1/sref10 http://refhub.elsevier.com/S0277-3791(23)00078-1/sref10 http://refhub.elsevier.com/S0277-3791(23)00078-1/sref11 http://refhub.elsevier.com/S0277-3791(23)00078-1/sref11 http://refhub.elsevier.com/S0277-3791(23)00078-1/sref11 http://refhub.elsevier.com/S0277-3791(23)00078-1/sref12 http://refhub.elsevier.com/S0277-3791(23)00078-1/sref12 http://refhub.elsevier.com/S0277-3791(23)00078-1/sref12 http://refhub.elsevier.com/S0277-3791(23)00078-1/sref12 http://refhub.elsevier.com/S0277-3791(23)00078-1/sref13 http://refhub.elsevier.com/S0277-3791(23)00078-1/sref13 http://refhub.elsevier.com/S0277-3791(23)00078-1/sref13 http://refhub.elsevier.com/S0277-3791(23)00078-1/sref13 http://refhub.elsevier.com/S0277-3791(23)00078-1/sref13 http://refhub.elsevier.com/S0277-3791(23)00078-1/sref14 http://refhub.elsevier.com/S0277-3791(23)00078-1/sref14 http://refhub.elsevier.com/S0277-3791(23)00078-1/sref14 http://refhub.elsevier.com/S0277-3791(23)00078-1/sref14 http://refhub.elsevier.com/S0277-3791(23)00078-1/sref14 http://refhub.elsevier.com/S0277-3791(23)00078-1/sref15 http://refhub.elsevier.com/S0277-3791(23)00078-1/sref15 http://refhub.elsevier.com/S0277-3791(23)00078-1/sref15 http://refhub.elsevier.com/S0277-3791(23)00078-1/sref15 http://refhub.elsevier.com/S0277-3791(23)00078-1/sref16 http://refhub.elsevier.com/S0277-3791(23)00078-1/sref16 http://refhub.elsevier.com/S0277-3791(23)00078-1/sref16 http://refhub.elsevier.com/S0277-3791(23)00078-1/sref16 http://refhub.elsevier.com/S0277-3791(23)00078-1/sref17 http://refhub.elsevier.com/S0277-3791(23)00078-1/sref17 http://refhub.elsevier.com/S0277-3791(23)00078-1/sref17 http://refhub.elsevier.com/S0277-3791(23)00078-1/sref17 http://refhub.elsevier.com/S0277-3791(23)00078-1/sref18 http://refhub.elsevier.com/S0277-3791(23)00078-1/sref18 http://refhub.elsevier.com/S0277-3791(23)00078-1/sref18 http://refhub.elsevier.com/S0277-3791(23)00078-1/sref19 http://refhub.elsevier.com/S0277-3791(23)00078-1/sref19 http://refhub.elsevier.com/S0277-3791(23)00078-1/sref19 http://refhub.elsevier.com/S0277-3791(23)00078-1/sref20 http://refhub.elsevier.com/S0277-3791(23)00078-1/sref20 http://refhub.elsevier.com/S0277-3791(23)00078-1/sref20 http://refhub.elsevier.com/S0277-3791(23)00078-1/sref20 http://refhub.elsevier.com/S0277-3791(23)00078-1/sref20 http://refhub.elsevier.com/S0277-3791(23)00078-1/sref21 http://refhub.elsevier.com/S0277-3791(23)00078-1/sref21 http://refhub.elsevier.com/S0277-3791(23)00078-1/sref21 http://refhub.elsevier.com/S0277-3791(23)00078-1/sref21 http://refhub.elsevier.com/S0277-3791(23)00078-1/sref21 http://refhub.elsevier.com/S0277-3791(23)00078-1/sref22 http://refhub.elsevier.com/S0277-3791(23)00078-1/sref22 http://refhub.elsevier.com/S0277-3791(23)00078-1/sref22 M. Wojcieszak, L. Backwell, Francesco d’Errico et al. Quaternary Science Reviews 306 (2023) 108030 Esteban, I., Stratford, D., de la Pe~na, P., Backwell, L., d’Errico, F., Wadley, L., 2023. Plants, people and fire: phytolith and FTIR analyses of the post-howiesons poort occupations at border cave (KwaZulu-Natal, South Africa). Quat. Sci. 300, 107898. Fagbuaro, O., Oso, J.A., Edward, J.B., Ogunleye, R.F., 2006. Nutritional status of four species of giant land snails in Nigeria. J. Zhejiang Univ. - Sci. B. 7 (9), 686e689. Faulkner, P., Harris, M., Haji, O., Ali, A.K., Crowther, A., Shipton, C., Horton, M.C., Boivin, N.L., 2019. Long-term trends in terrestrial and marine invertebrate exploitation on the eastern African coast: insights from Kuumbi Cave, Zanzibar. J. I. Coast Archaeol. 14, 479e514. Fern�andez-L�opez de Pablo, J., Badal, E., Ferrer García, C., Martínez-Ortí, A., Sanchis Serra, A., 2014. Land snails as a diet diversification proxy during the early upper palaeolithic in Europe. PLoS One 9 (8), e104898. Flannery, K.V., 1969. Origins and ecological effects of early domestication in Iran and the Near East. In: Ucko, P.J., Dimbleby, G.W. (Eds.), The Domestication and Exploitation of Plants and Animals. Transaction Publishers, pp. 73e100. Gliganic, L.A., Jacobs, Z., Roberts, R.G., Domínguez-Rodrigo, M., Mabulla, A.Z.P., 2012. New ages for Middle and Later Stone Age deposits at Mumba rockshelter, Tanzania: optically stimulated luminescence dating of quartz and feldspar grains. J. Hum. Evol. 62 (4), 533e547. Griveaud, P., 2016. The invertebrates. In: Jolly, A., Oberl�e, P., Albignac, R. (Eds.), Key Environments: Madagascar. Elsevier. Grün, R., Beaumont, P., 2001. Border Cave revisited: a revised ESR chronology. J. Hum. Evol. 40, 467e482. Grün, R., Beaumont, P., Tobias, P.V., Eggins, S., 2003. On the age of Border Cave 5 human mandible. J. Hum. Evol. 45, 155e167. Herbert, D., Kilburn, D., 2004. Field Guide to the Land Snails and Slugs of Eastern South Africa. Natal Museum. Jerardino, A., Marean, C.W., 2010. Shellfish gathering, marine paleoecology and modern human behavior: perspectives from cave PP13B, Pinnacle Point, South Africa. J. Hum. Evol. 59 (3e4), 412e424. Jord�a, J.F., Avezuela, B., Aura, J.E., Martín-Escorza, C., 2011. The gastropod fauna of the Epipaleolithic shell midden in the Vestibulo chamber of Nerja Cave (M�alaga, southern Spain). Quat. Int. 244 (1), 27e36. Kuhn, S.L., Pigati, J., Karkanas, P., Koumouzelis, M., Kozlowski, J.K., Ntinou, N., Stiner, M.C., 2010. Radiocarbon dating results for the early upper paleolithic of Klissoura cave 1. Eurasian Prehistory 7 (2), 37e46. Lange, K., Perl�es, C., Vanhaeren, M., Reiche, I., 2008. Heat-induced modification of marine shells used as personal ornaments at the prehistoric site of Franchthi Cave, Greece: first results of a multianalytical approach. In: 9th International Conference on Nondestructive Testing of Art, Jerusalem Israel, pp. 25e30. May. art). Lennox, S., Backwell, L.R., d'Errico, F., Wadley, L., 2022. A vegetation record based on charcoal analysis from Border Cave, KwaZulu-Natal, South Africa, ~ 227 000 to ~ 44 000 years ago. Quat. Sci. Rev. (in press). Lubell, D., 2004a. Prehistoric edible land snails in the circum-Mediterranean: the archaeological evidence. In: Brugal, J.P., Desse, J. (Eds.), Petits Animaux et Soci�et�es Humaines. Du Compl�ement Alimentaire aux Ressources Utilitaires. XXIV rencon- tres internationales d'arch�eologie et d'historie d'Antibes. Editions APDCA, Antibes, pp. 77e98. Lubell, D., 2004b. Are land snail a signature for the Mesolithic-Neolithic transition. Doc. Praehist. 31, 1e24. Maguregui, M., Knuutinen, U., Trebolazabala, J., Morillas, H., Castro, K., Martinez- Arkarazo, I., Madariaga, J.M., 2012. Use of in situ and confocal Raman spec- troscopy to study the nature and distribution of carotenoids in brown patinas from a deteriorated wall painting in Marcus Lucretius House (Pompeii). Anal. Bioanal. Chem. 402 (4), 1529e1539. Mannino, M.A., 2019. Invertebrate zooarchaeology. In: Richards, M.P., Britton, K. (Eds.), Archaeological Science: an Introduction. Cambridge University Press, pp. 233e275. Marean, C.W., Bar-Matthews, M., Bernatchez, J., Fisher, E., Goldberg, P., Herries, A.I., Jacobs, Z., Jerardino, A., Karkanas, P., Minichillo, T., Nilssen, P.J., 2007. Early human use of marine resources and pigment in South Africa during the Middle Pleistocene. Nature 449 (7164), 905e908. Marshall, L., 1976. The!Kung of Nyae Nyae. Harvard University Press, Cambridge, Mass. Martini, F., Colonese, A.C., Giuseppe, Z.D., Ghinassi, M., Vetro, D.L., Ricciardi, S., 2009. Human-environment relationships during the late glacial-early Holocene transition: some examples from campania, calabria and sicily. J. Mediterr. Geog. 112, 89e94. Martin�on-Torres, M., d'Errico, F., Santos, E., �Alvaro Gallo, A., Amano, N., Archer, W., Armitage, S.J., Arsuaga, J.L., Bermúdez de Castro, J.M., Blinkhorn, J., Crowther, A., 2021. Earliest known human burial in Africa. Nature 593 (7857), 95e100. Mehlman, M.J., 1979. Mumba-H€ohle revisited: the relevance of a forgotten exca- vation to some current issues in East African prehistory. World Archaeol. 11, 80e94. Mellars, P., Tixier, J., 1989. Radiocarbon-accelerator dating of ksar 'aqil (Lebanon) and the chronology of the upper palaeolithic sequence in the Middle East. Antiquity 63, 761e768. Mernagh, T.P., 1991. Use of the laser Raman microprobe for discrimination amongst feldspar minerals. J. Raman Spectrosc. 22 (8), 453e457. Milano, S., Prendergast, A.L., Sch€one, B.R., 2016. Effects of cooking on mollusk shell structure and chemistry: implications for archeology and paleoenvironmental reconstruction. J. Archaeol. Sci.: For. Rep. 7, 14e26. Milano, S., Lindauer, S., Prendergast, A.L., Hill, E.A., Hunt, C.O., Barker, G., 13 Sch€one, B.R., 2018. Mollusk carbonate thermal behaviour and its implications in understanding prehistoric fire events in shell middens. J. Archaeol. Sci.: For. Rep. 20, 443e457. Millard, A., 2006. Bayesian analysis of ESR dates, with application to Border Cave. Quat. Geochronol. 1 (2), 159e166. Müller, P., Staudigel, P.T., Murray, S.T., Vernet, R., Barusseau, J.-P., Westphal, H., Swart, P.K., 2017. Prehistoric cooking versus accurate palaeotemperature re- cords in shell midden constituents. Sci. Rep. UK 7 (1), 1e11. Munywoki, G.N., 2022. Snail farming (Heliciculture): benefits, economic viability, challenges and prospects. Int. J. Res. Agronomy 5 (1), 9e11. Nouet, J., 2014. Caract�erisations physico-chimiques des biomin�eraux carbonat�es de Mollusques actuels et fossiles: le cas des structures entrecrois�ees. Universit�e Paris Sud - Paris XI. Ph.D. thesis. CNNT : 2014PA112066D Ctel-00984962D. Parker, J.E., Thompson, S.P., Lennie, A.R., Potter, J., Tang, C.C., 2010. A study of the aragonite-calcite transformation using Raman spectroscopy, synchrotron powder diffraction and scanning electron microscopy. CrystEngComm 12 (5), 1590e1599. Plug, I., 2004. Resource exploitation: animal use during the middle Stone age at Sibudu cave, KwaZulu-natal. South Afr. J. Sci. 100, 151e158. Potter, R.M., Rossman, G.R., 1979. Mineralogy of manganesed endrites and coatings. Am. Mineral. 64, 1219e1226. Prinsloo, L.C., Tourni�e, A., Colomban, P., Paris, C., Bassett, S.T., 2013. In search of the optimum Raman/IR signatures of potential ingredients used in San/Bushman rock art paint. J. Archaeol. Sci. 40 (7), 2981e2990. Rizner, M., Vukosavljevic, N., Miracle, P., 2009. The paleoecological and paleodietary significance of edible land snails (Helix sp.) across the Pleistocene-Holocene transition on the eastern Adriatic coast. In: McCarthan, S.B., Schulting, R., Warren, G., Woodman, P. (Eds.), Mesolithic Horizons. Papers Presented at the Seventh International Conference on the Mesolithic in Europe, Belfast 2005, II. Oxbow, Oxford, pp. 527e532. Schiegl, S., Conard, N., 2006. The Middle Stone Age sediments at Sibudu: results from FTIR spectroscopy and microscopic analyses. South. Afr. Humanit. 18 (1), 149e172. Shillito, L.-M., Almond, M.J., Nicholson, J., Pantos, M., Matthews, W., 2009. Rapid characterisation of archaeological midden components using FT-IR spectros- copy, SEM-EDX and micro-XRD. Spectrochim. Acta 73 (1), 133e139. Shipton, C., Crowther, A., Kourampas, N., Prendergast, M.E., Horton, M., Douka, K., Schwenninger, J.L., Faulkner, P., Quintana Morales, E.M., Langley, M.C., Tibesasa, R., 2016. Reinvestigation of Kuumbi cave, zanzibar, reveals later Stone age coastal habitation, early Holocene abandonment and iron age reoccupation. Azania 51 (2), 197e233. Sievers, C., Wadley, L., 2008. Going underground: experimental carbonization of fruiting structures under hearths. J. Archaeol. Sci. 35 (11), 2909e2917. Sievers, C., Backwell, L., d'Errico, F., Wadley, L., 2022. Plant bedding construction between 60,000 and 40,000 years ago at Border Cave, South Africa. Quat. Sci. Rev. 275, 107280. Stiner, M.C., 2001. Thirty years on the “broad spectrum revolution” and paleolithic demography. P. Natl. Acad. Sci. USA. 98 (13), 6993e6996. Stiner, M.C., Munro, N.D., 2011. On the evolution of paleolithic diet and landscape at Franchthi cave (peloponnese, Greece). J. Hum. Evol. 60, 618e636. Stratford, D., Clark, J., Wojcieszak, M., de la Pe~na, P., Wadley, L., d'Errico, F., Mauran, G., Sievers, C., Banks, W., Beard, T., Horn, M., Shadrach, K., Morrissey, P., Backwell, L., 2022. Geoarchaeology and zooarchaeology of Border Cave, South Africa: initial multiproxy considerations of stratigraphy and site formation processes from the Backwell et al. excavations. Quat. Sci. Rev., 107618 Taylor, V.K., Barton, R.N.E., Bell, M., Bouzougar, S., Collcutt, S., Black, S., Hogue, J.T., 2011. The Epipaleolithic (Iberomaurusian) at Grotte des Pigeons (Taforalt), Morocco: a preliminary study of the land Mollusca. Quat. Int. 244 (1), 5e14. Toffolo, M.B., 2021. The significance of aragonite in the interpretation of the microscopic archaeological record. Geoarchaeology 36 (1), 149e169. Tomi�c, Z., Makreski, P., Gaji�c, B., 2010. Identification and spectra-structure deter- mination of soil minerals: Raman study supported by IR spectroscopy and X-ray powder diffraction. J. Raman Spectrosc. 41 (5), 582e586. Tribolo, C., Mercier, N., Dumottay, C., Cantin, N., Banks, W., Stratford, D., de la Pe~na, P., Backwell, L., Wadley, L., d’Errico, F., 2022. Luminescence dating at Border Cave: attempts, questions and new results. Quat. Sci. Rev. 296, 107787. Ugwumba, C.O.A., Obiekwe, J.N., Ozor, M.U., 2016. Marketing of african giant snail (Achatina achatina) in anambra state, Nigeria. J. Dent. and Med. Sci. 15 (6), 57e66. Urmos, J., Sharma, S.K., Mackenzie, F.T., 1991. Characterization of some biogenic carbonates with Raman spectroscopy. Am. Mineral. 76 (3e4), 641e646. Villa, P., Soriano, S., Tsanova, T., Higham, T., d'Errico, F., Backwell, L.R., Lucejko, J., Beaumont, P., 2012. Border cave and the beginning of the later Stone age in South Africa. P. Natl. Acad. Sci. USA. 109 (33), 13208e13213. Vogel, J.C., Beaumont, P.B., 1972. Revised radiocarbon chronology for the Stone age in South Africa. Nature 237 (5349), 50e51. Vogel, J.C., Fuls, A., Visser, E., 1986. Pretoria radiocarbon dates III. Radiocarbon 28 (3), 1133e1172. Wadley, L., Esteban, I., de la Pe~na, P., Wojcieszak, M., Stratford, D., Lennox, S., d'Errico, F., Rosso, D.E., Orange, F., Backwell, L., 2020a. Fire and grass-bedding construction 200 thousand years ago at Border Cave, South Africa. Science 369, 863e866. Wadley, L., Backwell, L., d'Errico, F., Sievers, C., 2020b. Cooked starchy rhizomes in Africa 170 thousand years ago. Science 367 (6473), 87e91. Werts, S.P., Jahren, A.H., 2007. Estimation of temperatures beneath archaeological http://refhub.elsevier.com/S0277-3791(23)00078-1/sref23 http://refhub.elsevier.com/S0277-3791(23)00078-1/sref23 http://refhub.elsevier.com/S0277-3791(23)00078-1/sref23 http://refhub.elsevier.com/S0277-3791(23)00078-1/sref23 http://refhub.elsevier.com/S0277-3791(23)00078-1/sref23 http://refhub.elsevier.com/S0277-3791(23)00078-1/sref24 http://refhub.elsevier.com/S0277-3791(23)00078-1/sref24 http://refhub.elsevier.com/S0277-3791(23)00078-1/sref24 http://refhub.elsevier.com/S0277-3791(23)00078-1/sref25 http://refhub.elsevier.com/S0277-3791(23)00078-1/sref25 http://refhub.elsevier.com/S0277-3791(23)00078-1/sref25 http://refhub.elsevier.com/S0277-3791(23)00078-1/sref25 http://refhub.elsevier.com/S0277-3791(23)00078-1/sref25 http://refhub.elsevier.com/S0277-3791(23)00078-1/sref26 http://refhub.elsevier.com/S0277-3791(23)00078-1/sref26 http://refhub.elsevier.com/S0277-3791(23)00078-1/sref26 http://refhub.elsevier.com/S0277-3791(23)00078-1/sref26 http://refhub.elsevier.com/S0277-3791(23)00078-1/sref26 http://refhub.elsevier.com/S0277-3791(23)00078-1/sref27 http://refhub.elsevier.com/S0277-3791(23)00078-1/sref27 http://refhub.elsevier.com/S0277-3791(23)00078-1/sref27 http://refhub.elsevier.com/S0277-3791(23)00078-1/sref27 http://refhub.elsevier.com/S0277-3791(23)00078-1/sref28 http://refhub.elsevier.com/S0277-3791(23)00078-1/sref28 http://refhub.elsevier.com/S0277-3791(23)00078-1/sref28 http://refhub.elsevier.com/S0277-3791(23)00078-1/sref28 http://refhub.elsevier.com/S0277-3791(23)00078-1/sref28 http://refhub.elsevier.com/S0277-3791(23)00078-1/sref29 http://refhub.elsevier.com/S0277-3791(23)00078-1/sref29 http://refhub.elsevier.com/S0277-3791(23)00078-1/sref29 http://refhub.elsevier.com/S0277-3791(23)00078-1/sref30 http://refhub.elsevier.com/S0277-3791(23)00078-1/sref30 http://refhub.elsevier.com/S0277-3791(23)00078-1/sref30 http://refhub.elsevier.com/S0277-3791(23)00078-1/sref31 http://refhub.elsevier.com/S0277-3791(23)00078-1/sref31 http://refhub.elsevier.com/S0277-3791(23)00078-1/sref31 http://refhub.elsevier.com/S0277-3791(23)00078-1/sref32 http://refhub.elsevier.com/S0277-3791(23)00078-1/sref32 http://refhub.elsevier.com/S0277-3791(23)00078-1/sref33 http://refhub.elsevier.com/S0277-3791(23)00078-1/sref33 http://refhub.elsevier.com/S0277-3791(23)00078-1/sref33 http://refhub.elsevier.com/S0277-3791(23)00078-1/sref33 http://refhub.elsevier.com/S0277-3791(23)00078-1/sref33 http://refhub.elsevier.com/S0277-3791(23)00078-1/sref34 http://refhub.elsevier.com/S0277-3791(23)00078-1/sref34 http://refhub.elsevier.com/S0277-3791(23)00078-1/sref34 http://refhub.elsevier.com/S0277-3791(23)00078-1/sref34 http://refhub.elsevier.com/S0277-3791(23)00078-1/sref34 http://refhub.elsevier.com/S0277-3791(23)00078-1/sref34 http://refhub.elsevier.com/S0277-3791(23)00078-1/sref36 http://refhub.elsevier.com/S0277-3791(23)00078-1/sref36 http://refhub.elsevier.com/S0277-3791(23)00078-1/sref36 http://refhub.elsevier.com/S0277-3791(23)00078-1/sref36 http://refhub.elsevier.com/S0277-3791(23)00078-1/sref37 http://refhub.elsevier.com/S0277-3791(23)00078-1/sref37 http://refhub.elsevier.com/S0277-3791(23)00078-1/sref37 http://refhub.elsevier.com/S0277-3791(23)00078-1/sref37 http://refhub.elsevier.com/S0277-3791(23)00078-1/sref37 http://refhub.elsevier.com/S0277-3791(23)00078-1/sref37 http://refhub.elsevier.com/S0277-3791(23)00078-1/sref37 http://refhub.elsevier.com/S0277-3791(23)00078-1/sref38 http://refhub.elsevier.com/S0277-3791(23)00078-1/sref38 http://refhub.elsevier.com/S0277-3791(23)00078-1/sref38 http://refhub.elsevier.com/S0277-3791(23)00078-1/sref39 http://refhub.elsevier.com/S0277-3791(23)00078-1/sref39 http://refhub.elsevier.com/S0277-3791(23)00078-1/sref39 http://refhub.elsevier.com/S0277-3791(23)00078-1/sref39 http://refhub.elsevier.com/S0277-3791(23)00078-1/sref39 http://refhub.elsevier.com/S0277-3791(23)00078-1/sref39 http://refhub.elsevier.com/S0277-3791(23)00078-1/sref39 http://refhub.elsevier.com/S0277-3791(23)00078-1/sref39 http://refhub.elsevier.com/S0277-3791(23)00078-1/sref39 http://refhub.elsevier.com/S0277-3791(23)00078-1/sref39 http://refhub.elsevier.com/S0277-3791(23)00078-1/sref40 http://refhub.elsevier.com/S0277-3791(23)00078-1/sref40 http://refhub.elsevier.com/S0277-3791(23)00078-1/sref40 http://refhub.elsevier.com/S0277-3791(23)00078-1/sref41 http://refhub.elsevier.com/S0277-3791(23)00078-1/sref41 http://refhub.elsevier.com/S0277-3791(23)00078-1/sref41 http://refhub.elsevier.com/S0277-3791(23)00078-1/sref41 http://refhub.elsevier.com/S0277-3791(23)00078-1/sref41 http://refhub.elsevier.com/S0277-3791(23)00078-1/sref41 http://refhub.elsevier.com/S0277-3791(23)00078-1/sref42 http://refhub.elsevier.com/S0277-3791(23)00078-1/sref42 http://refhub.elsevier.com/S0277-3791(23)00078-1/sref42 http://refhub.elsevier.com/S0277-3791(23)00078-1/sref42 http://refhub.elsevier.com/S0277-3791(23)00078-1/sref43 http://refhub.elsevier.com/S0277-3791(23)00078-1/sref43 http://refhub.elsevier.com/S0277-3791(23)00078-1/sref43 http://refhub.elsevier.com/S0277-3791(23)00078-1/sref43 http://refhub.elsevier.com/S0277-3791(23)00078-1/sref43 http://refhub.elsevier.com/S0277-3791(23)00078-1/sref44 http://refhub.elsevier.com/S0277-3791(23)00078-1/sref44 http://refhub.elsevier.com/S0277-3791(23)00078-1/sref45 http://refhub.elsevier.com/S0277-3791(23)00078-1/sref45 http://refhub.elsevier.com/S0277-3791(23)00078-1/sref45 http://refhub.elsevier.com/S0277-3791(23)00078-1/sref45 http://refhub.elsevier.com/S0277-3791(23)00078-1/sref45 http://refhub.elsevier.com/S0277-3791(23)00078-1/sref46 http://refhub.elsevier.com/S0277-3791(23)00078-1/sref46 http://refhub.elsevier.com/S0277-3791(23)00078-1/sref46 http://refhub.elsevier.com/S0277-3791(23)00078-1/sref46 http://refhub.elsevier.com/S0277-3791(23)00078-1/sref46 http://refhub.elsevier.com/S0277-3791(23)00078-1/sref46 http://refhub.elsevier.com/S0277-3791(23)00078-1/sref47 http://refhub.elsevier.com/S0277-3791(23)00078-1/sref47 http://refhub.elsevier.com/S0277-3791(23)00078-1/sref47 http://refhub.elsevier.com/S0277-3791(23)00078-1/sref47 http://refhub.elsevier.com/S0277-3791(23)00078-1/sref47 http://refhub.elsevier.com/S0277-3791(23)00078-1/sref48 http://refhub.elsevier.com/S0277-3791(23)00078-1/sref48 http://refhub.elsevier.com/S0277-3791(23)00078-1/sref48 http://refhub.elsevier.com/S0277-3791(23)00078-1/sref48 http://refhub.elsevier.com/S0277-3791(23)00078-1/sref49 http://refhub.elsevier.com/S0277-3791(23)00078-1/sref49 http://refhub.elsevier.com/S0277-3791(23)00078-1/sref49 http://refhub.elsevier.com/S0277-3791(23)00078-1/sref50 http://refhub.elsevier.com/S0277-3791(23)00078-1/sref50 http://refhub.elsevier.com/S0277-3791(23)00078-1/sref50 http://refhub.elsevier.com/S0277-3791(23)00078-1/sref50 http://refhub.elsevier.com/S0277-3791(23)00078-1/sref50 http://refhub.elsevier.com/S0277-3791(23)00078-1/sref51 http://refhub.elsevier.com/S0277-3791(23)00078-1/sref51 http://refhub.elsevier.com/S0277-3791(23)00078-1/sref51 http://refhub.elsevier.com/S0277-3791(23)00078-1/sref51 http://refhub.elsevier.com/S0277-3791(23)00078-1/sref51 http://refhub.elsevier.com/S0277-3791(23)00078-1/sref51 http://refhub.elsevier.com/S0277-3791(23)00078-1/sref52 http://refhub.elsevier.com/S0277-3791(23)00078-1/sref52 http://refhub.elsevier.com/S0277-3791(23)00078-1/sref52 http://refhub.elsevier.com/S0277-3791(23)00078-1/sref53 http://refhub.elsevier.com/S0277-3791(23)00078-1/sref53 http://refhub.elsevier.com/S0277-3791(23)00078-1/sref53 http://refhub.elsevier.com/S0277-3791(23)00078-1/sref53 http://refhub.elsevier.com/S0277-3791(23)00078-1/sref54 http://refhub.elsevier.com/S0277-3791(23)00078-1/sref54 http://refhub.elsevier.com/S0277-3791(23)00078-1/sref54 http://refhub.elsevier.com/S0277-3791(23)00078-1/sref86 http://refhub.elsevier.com/S0277-3791(23)00078-1/sref86 http://refhub.elsevier.com/S0277-3791(23)00078-1/sref86 http://refhub.elsevier.com/S0277-3791(23)00078-1/sref86 http://refhub.elsevier.com/S0277-3791(23)00078-1/sref86 http://refhub.elsevier.com/S0277-3791(23)00078-1/sref86 http://refhub.elsevier.com/S0277-3791(23)00078-1/sref86 http://refhub.elsevier.com/S0277-3791(23)00078-1/sref86 http://refhub.elsevier.com/S0277-3791(23)00078-1/sref86 http://refhub.elsevier.com/S0277-3791(23)00078-1/sref86 http://refhub.elsevier.com/S0277-3791(23)00078-1/sref56 http://refhub.elsevier.com/S0277-3791(23)00078-1/sref56 http://refhub.elsevier.com/S0277-3791(23)00078-1/sref56 http://refhub.elsevier.com/S0277-3791(23)00078-1/sref56 http://refhub.elsevier.com/S0277-3791(23)00078-1/sref56 http://refhub.elsevier.com/S0277-3791(23)00078-1/sref57 http://refhub.elsevier.com/S0277-3791(23)00078-1/sref57 http://refhub.elsevier.com/S0277-3791(23)00078-1/sref57 http://refhub.elsevier.com/S0277-3791(23)00078-1/sref58 http://refhub.elsevier.com/S0277-3791(23)00078-1/sref58 http://refhub.elsevier.com/S0277-3791(23)00078-1/sref58 http://refhub.elsevier.com/S0277-3791(23)00078-1/sref59 http://refhub.elsevier.com/S0277-3791(23)00078-1/sref59 http://refhub.elsevier.com/S0277-3791(23)00078-1/sref59 http://refhub.elsevier.com/S0277-3791(23)00078-1/sref59 http://refhub.elsevier.com/S0277-3791(23)00078-1/sref59 http://refhub.elsevier.com/S0277-3791(23)00078-1/sref60 http://refhub.elsevier.com/S0277-3791(23)00078-1/sref60 http://refhub.elsevier.com/S0277-3791(23)00078-1/sref60 http://refhub.elsevier.com/S0277-3791(23)00078-1/sref60 http://refhub.elsevier.com/S0277-3791(23)00078-1/sref60 http://refhub.elsevier.com/S0277-3791(23)00078-1/sref60 http://refhub.elsevier.com/S0277-3791(23)00078-1/sref60 http://refhub.elsevier.com/S0277-3791(23)00078-1/sref61 http://refhub.elsevier.com/S0277-3791(23)00078-1/sref61 http://refhub.elsevier.com/S0277-3791(23)00078-1/sref61 http://refhub.elsevier.com/S0277-3791(23)00078-1/sref61 http://refhub.elsevier.com/S0277-3791(23)00078-1/sref62 http://refhub.elsevier.com/S0277-3791(23)00078-1/sref62 http://refhub.elsevier.com/S0277-3791(23)00078-1/sref62 http://refhub.elsevier.com/S0277-3791(23)00078-1/sref62 http://refhub.elsevier.com/S0277-3791(23)00078-1/sref63 http://refhub.elsevier.com/S0277-3791(23)00078-1/sref63 http://refhub.elsevier.com/S0277-3791(23)00078-1/sref63 http://refhub.elsevier.com/S0277-3791(23)00078-1/sref63 http://refhub.elsevier.com/S0277-3791(23)00078-1/sref63 http://refhub.elsevier.com/S0277-3791(23)00078-1/sref63 http://refhub.elsevier.com/S0277-3791(23)00078-1/sref64 http://refhub.elsevier.com/S0277-3791(23)00078-1/sref64 http://refhub.elsevier.com/S0277-3791(23)00078-1/sref64 http://refhub.elsevier.com/S0277-3791(23)00078-1/sref65 http://refhub.elsevier.com/S0277-3791(23)00078-1/sref65 http://refhub.elsevier.com/S0277-3791(23)00078-1/sref65 http://refhub.elsevier.com/S0277-3791(23)00078-1/sref66 http://refhub.elsevier.com/S0277-3791(23)00078-1/sref66 http://refhub.elsevier.com/S0277-3791(23)00078-1/sref66 http://refhub.elsevier.com/S0277-3791(23)00078-1/sref67 http://refhub.elsevier.com/S0277-3791(23)00078-1/sref67 http://refhub.elsevier.com/S0277-3791(23)00078-1/sref67 http://refhub.elsevier.com/S0277-3791(23)00078-1/sref68 http://refhub.elsevier.com/S0277-3791(23)00078-1/sref68 http://refhub.elsevier.com/S0277-3791(23)00078-1/sref68 http://refhub.elsevier.com/S0277-3791(23)00078-1/sref68 http://refhub.elsevier.com/S0277-3791(23)00078-1/sref68 http://refhub.elsevier.com/S0277-3791(23)00078-1/sref68 http://refhub.elsevier.com/S0277-3791(23)00078-1/sref69 http://refhub.elsevier.com/S0277-3791(23)00078-1/sref69 http://refhub.elsevier.com/S0277-3791(23)00078-1/sref69 http://refhub.elsevier.com/S0277-3791(23)00078-1/sref69 http://refhub.elsevier.com/S0277-3791(23)00078-1/sref70 http://refhub.elsevier.com/S0277-3791(23)00078-1/sref70 http://refhub.elsevier.com/S0277-3791(23)00078-1/sref70 http://refhub.elsevier.com/S0277-3791(23)00078-1/sref71 http://refhub.elsevier.com/S0277-3791(23)00078-1/sref71 http://refhub.elsevier.com/S0277-3791(23)00078-1/sref71 http://refhub.elsevier.com/S0277-3791(23)00078-1/sref71 http://refhub.elsevier.com/S0277-3791(23)00078-1/sref71 http://refhub.elsevier.com/S0277-3791(23)00078-1/sref71 http://refhub.elsevier.com/S0277-3791(23)00078-1/sref72 http://refhub.elsevier.com/S0277-3791(23)00078-1/sref72 http://refhub.elsevier.com/S0277-3791(23)00078-1/sref72 http://refhub.elsevier.com/S0277-3791(23)00078-1/sref72 http://refhub.elsevier.com/S0277-3791(23)00078-1/sref73 http://refhub.elsevier.com/S0277-3791(23)00078-1/sref73 http://refhub.elsevier.com/S0277-3791(23)00078-1/sref73 http://refhub.elsevier.com/S0277-3791(23)00078-1/sref73 http://refhub.elsevier.com/S0277-3791(23)00078-1/sref74 http://refhub.elsevier.com/S0277-3791(23)00078-1/sref74 http://refhub.elsevier.com/S0277-3791(23)00078-1/sref74 http://refhub.elsevier.com/S0277-3791(23)00078-1/sref74 http://refhub.elsevier.com/S0277-3791(23)00078-1/sref75 http://refhub.elsevier.com/S0277-3791(23)00078-1/sref75 http://refhub.elsevier.com/S0277-3791(23)00078-1/sref75 http://refhub.elsevier.com/S0277-3791(23)00078-1/sref75 http://refhub.elsevier.com/S0277-3791(23)00078-1/sref76 http://refhub.elsevier.com/S0277-3791(23)00078-1/sref76 http://refhub.elsevier.com/S0277-3791(23)00078-1/sref76 http://refhub.elsevier.com/S0277-3791(23)00078-1/sref77 http://refhub.elsevier.com/S0277-3791(23)00078-1/sref77 http://refhub.elsevier.com/S0277-3791(23)00078-1/sref77 http://refhub.elsevier.com/S0277-3791(23)00078-1/sref78 http://refhub.elsevier.com/S0277-3791(23)00078-1/sref78 http://refhub.elsevier.com/S0277-3791(23)00078-1/sref78 http://refhub.elsevier.com/S0277-3791(23)00078-1/sref78 http://refhub.elsevier.com/S0277-3791(23)00078-1/sref78 http://refhub.elsevier.com/S0277-3791(23)00078-1/sref78 http://refhub.elsevier.com/S0277-3791(23)00078-1/sref79 http://refhub.elsevier.com/S0277-3791(23)00078-1/sref79 http://refhub.elsevier.com/S0277-3791(23)00078-1/sref79 http://refhub.elsevier.com/S0277-3791(23)00078-1/sref80 M. Wojcieszak, L. Backwell, Francesco d’Errico et al. Quaternary Science Reviews 306 (2023) 108030 campfires using carbon stable isotope composition of soil organic matter. J. Archaeol. Sci. 34 (6), 850e857. Wojcieszak, M., Wadley, L., 2019. A Raman micro-spectroscopy study of 77,000 to 71,000 year old ochre processing tools from Sibudu, KwaZulu-Natal, South Af- rica. Heritage Science 7 (1), 24. Yoshioka, S., Kitano, Y., 1985. Transformation of aragonite to calcite through heating. Geochem. J. 19 (4), 245e249. Zeder, M.A., 2012. The broad spectrum revolution at 40: resource diversity, 14 intensification, and an alternative to optimal foraging explanations. J. Anthropol. Archaeol. 31 (3), 241e264. Zilh~ao, J., Angelucci, D.E., Igreja, M.A., Arnold, L.J., Badal, E., Callapez, P., Cardoso, J.L., d'Errico, F., Daura, J., Demuro, M., Deschamps, M., 2020. Last interglacial iberian neandertals as Fisher-hunter-gatherers. Science 367 (6485), eaaz7943. Zwane, B., Bamford, M., 2021. Wood charcoal from Border Cave's Member 1RGBS: evidence for the environment and plant use during MIS 5. Afr. Archaeol. Rev. 38, 657e674. http://refhub.elsevier.com/S0277-3791(23)00078-1/sref80 http://refhub.elsevier.com/S0277-3791(23)00078-1/sref80 http://refhub.elsevier.com/S0277-3791(23)00078-1/sref80 http://refhub.elsevier.com/S0277-3791(23)00078-1/sref81 http://refhub.elsevier.com/S0277-3791(23)00078-1/sref81 http://refhub.elsevier.com/S0277-3791(23)00078-1/sref81 http://refhub.elsevier.com/S0277-3791(23)00078-1/sref82 http://refhub.elsevier.com/S0277-3791(23)00078-1/sref82 http://refhub.elsevier.com/S0277-3791(23)00078-1/sref82 http://refhub.elsevier.com/S0277-3791(23)00078-1/sref83 http://refhub.elsevier.com/S0277-3791(23)00078-1/sref83 http://refhub.elsevier.com/S0277-3791(23)00078-1/sref83 http://refhub.elsevier.com/S0277-3791(23)00078-1/sref83 http://refhub.elsevier.com/S0277-3791(23)00078-1/sref84 http://refhub.elsevier.com/S0277-3791(23)00078-1/sref84 http://refhub.elsevier.com/S0277-3791(23)00078-1/sref84 http://refhub.elsevier.com/S0277-3791(23)00078-1/sref84 http://refhub.elsevier.com/S0277-3791(23)00078-1/sref85 http://refhub.elsevier.com/S0277-3791(23)00078-1/sref85 http://refhub.elsevier.com/S0277-3791(23)00078-1/sref85 http://refhub.elsevier.com/S0277-3791(23)00078-1/sref85 Evid