Middle Stone Age wood use in Rose Cottage Cave South Africa: Evidence from charcoal identifications
Introduction
The charcoal found in archaeological sites most often represents the remnants of firewood collected and burned by people. As such, species suitable for this task were selected from the environment. Charcoal fragments are thus a partial archive of past environments, and they represent past human-environment interactions. Notwithstanding these limitations, charcoal analysis is a useful tool for palaeoenvironmental reconstructions (for example, Neumann, 1999; Figueiral and Mosbrugger, 2000; Asouti and Austin, 2005; Dotte-Sarout et al., 2015; Smith et al., 2015; Picornell-Gelabert et al., 2017). Such research, together with related studies like pollen and isotopes, contribute to the debate on whether technological change in the southern African Stone Age was driven by environmental changes or not (for example, Scott and Neumann, 2018). This type of study has been conducted successfully elsewhere (Allué et al., 2012, 2017; Vidal-Mutatano et al., 2015). Several well-recognised Middle Stone Age lithic technocomplexes occur in southern Africa, for example, the Howiesons Poort, which is distinct because of the backed tools that characterise it. Archaeologists have long wondered whether the appearance and disappearance of such technology was a response to the extractive requirements of specific resources enabled by environment. An ambitious dating project (Jacobs et al., 2008) implied that this was unlikely because the Howiesons Poort technocomplex, for example, was produced within a relatively short period (as little as 5000 years according to the Jacobs et al., 2008 study) across southern Africa, where it crosscut semi-arid regions in the west, and coastal forests in the east. Rose Cottage Cave (eastern Free State, South Africa) is situated in the interior of South Africa, far from the ocean, and at relatively high altitude in a grassland environment (to be described shortly), close to mountainous Lesotho. As will shortly be explained, Rose Cottage has a long and deep sequence and it was occupied intermittently from the Middle Stone Age dated from about 96,000 years (96 ka) ago to a few hundred years ago. An extensive survey and excavation strategy in the eastern Free State did not yield evidence for any other sites with such repeated occupation over such a long period (Wadley, 1995). Indeed, the survey suggests that people did not occupy much of the eastern Free State during the Middle Stone Age. This implies that Rose Cottage occupies a special niche in the area. Its local environment in the past may have made it more similar to sites in mountainous Lesotho than to sites in the eastern Free State and this new charcoal study will make the comparison possible. Melikane, in Lesotho, has a dated charcoal record from the Middle Stone Age (Stewart et al., 2016) and is therefore an ideal site to compare with Rose Cottage Cave. Charcoal analyses of the Late Pleistocene and Holocene layers of Rose Cottage and some Lesotho sites (Wadley et al., 1992; Esterhuysen, 1996; Esterhuysen and Mitchell, 1996) imply that environmental similarities may also have existed deeper in time.
This research therefore presents a palaeo-environmental reconstruction of the landscape based on the analysis of charcoal recovered at Rose Cottage from Middle Stone Age contexts. The main aim of the paper is to study the variations in vegetation cover, through species identifications from charcoal analysis, in the warmer and cooler phases anticipated for Marine Isotope Stages (MIS) 5 to 3. MIS5b starting about 93 ka is thought to have been cool, while MIS5a starting at 84 ka ago was warm, MIS4 starting at 73 ka ago was cool and MIS3 starting at 59 ka ago was cool with warm oscillations (Shackleton et al., 2003; Svensson et al., 2008).
A brief review of charcoal analysis is covered in Methods. Limitations of the method regarding palaeoenvironmental reconstruction are, first as mentioned in the introduction, that charcoal is a byproduct of firewood collected by humans, thus is anthropogenically sourced (Shackleton and Prins, 1992; Wadley, 2001, 2013) and, secondly, charcoal is changed by taphonomic processes resulting from human activities, fragmentation during wood carbonization, compression and site conditions (Théry-Parisot et al., 2010; Vidal-Matutano et al., 2020).
It was not possible to obtain pollen data from Rose Cottage because pollen has not preserved there. Faunal remains also have not preserved in the Middle Stone Age layers, so isotope studies on tooth enamel were also not feasible. The survival of charcoal in the cave is therefore especially important since it is at present the single means of obtaining evidence about the vegetation around the site in MIS5-MIS3. Late Pleistocene and Holocene layers had better organic preservation, enabling the sampling of grazer teeth for isotopes (Smith, 1997; Smith et al., 2002). The use of isotope studies is important for reconstructing vegetation in the area because Cymbopogon-Themeda-Eragrostis C4 grassland is characteristic in the Caledon Valley area nowadays, but when temperatures dropped to about 5° cooler than today, C3-dominated Afroalpine grassland would have prevailed (Smith et al., 2002). Woody species that may have accompanied the Afroalpine grasses in colder than present periods, include Cliffortia, summer rainfall Protea species and several fynbos plants (Esterhuysen, 1996). In the Holocene, warm and dry conditions were inferred from the Olea africana firewood collected at Rose Cottage Cave and in Lesotho in the Phuthiatsana-ea-Thaba Bosiu Basin (Wadley et al., 1992; Esterhuysen, 1996; Esterhuysen and Mitchell, 1996; Wadley, 1997; Esterhuysen et al., 1999). Isotopes on sediments from Holocene layers in Rose Cottage, and from samples taken outside the cave, largely support the environmental interpretations from the charcoal (Smith, 1997; Smith et al., 2002).
A preliminary study of a small sample of charcoal from the Rose Cottage Cave, Middle Stone Age sequence yielded identifications of Leucosidea sericea, Searsia sp., Erica sp., Protea sp., Rhamnus sp. and Euclea crispa (Harper, 1994). During this early study, identifications were not attempted from the small pieces of fragmented charcoal in the deepest layer from MIS5, Lendrum (LEN, dated ~96 ka ago and, for the purposes of this charcoal study, the layer is numbered [24]), but a stream bank community was inferred from charcoal identified in ~74 ka old layer Küsel [23], and ~69 ka old layer Emby B in early MIS4, while layers Emby A to Breuil [18] were shown to have especially low species diversity, tentatively interpreted as a dry period (Harper, 1994: 41) (see Table 1 for MIS assignment, site layers, and ages). This period of low diversity coincided with MIS4 and the earliest phase of Howiesons Poort technology at Rose Cottage Cave. In contrast, the more recent MIS4, represented in layers Tobias [17] to Suzy [13] dated ~62 ka ago (Pienaar et al., 2008), were reported to have prominent Leucosidea sericea streambank communities and forest precursor taxa (Harper, 1994: 41). In western Lesotho, wetter than present conditions were interpreted from the occurrence of Leucosidea sericea charcoal (Esterhuysen and Mitchell, 1997). Since Leucosidea sericea is frost-resistant (Esterhuysen, 1996), it is not a useful temperature indicator. This type of vegetation mosaic crosscut Howiesons Poort and more recent Middle Stone Age technology and also crosscut MIS4 and MIS3, ending at about 32 ka ago (Harper, 1994).
The Rose Cottage area has not been subjected to as many environmental analyses as Lesotho. Multiproxy environmental analyses have taken place in Lesotho, where pollen, diatoms, and wetland sediments have provided palaeoenvironmental data for the late Holocene (Norström et al., 2018) and the 15 ka to 14 ka period (Fitchett et al., 2016). Lesotho's Maloti-Drakensberg Mountains may have been a refugium for human populations in dry, unstable climates because the area has better water resources than the more extreme environments farther into the interior of southern Africa (Stewart and Mitchell, 2018). At Melikane Rock Shelter, Lesotho, which is 1800 m above sea level (m a.s.l.), the earliest occupation from Layers 30 and 29 is dated ~83 to 80 ka ago in MIS5, and this is followed by a ~61 ka ago occupation in Layers 25-23 that contain a Howiesons Poort lithic assemblage (Stewart et al., 2012). Another pulse of occupation with a post-Howiesons Poort lithic assemblage is more recent than 50 ka ago in MIS3 (Stewart et al., 2012). Although Melikane may have longer occupation hiatuses than Rose Cottage, the chronology is rather similar at both high-altitude sites, making them ideal for comparison. At Melikane, both phytolith identifications and isotope analyses of sediment suggest that the site was surrounded by C3-dominated Afroalpine grassland, implying temperatures at least 5° cooler than at present (Stewart et al., 2016). At ~50 ka ago (Layers 22-16), the phytolith identifications and abundance suggest high moisture availability and more woodland than present near Melikane (Stewart and Mitchell, 2018). Stewart and Mitchell (2018) conclude that in the summer rainfall area of the interior of southern Africa, cold climatic conditions may have been accompanied by considerable moisture availability. In a later MIS3 occupation, ~43-38 ka ago, C3 grass was still prevalent, but there was an increase in C4 grass phytoliths, and drier conditions are implied by sediment isotopes (Stewart and Mitchell, 2018). Widespread aridity occurs elsewhere in southern Africa at this time, implying that the availability of water in the Melikane region may have attracted people seeking this resource.
MIS4 charcoal samples from Layer 24 (~61 ka old) at Melikane yielded Buddleja cf. salviifolia, Rhamnus sp., Leucosidea sericea and Protea sp. (Stewart et al., 2016). Since Buddleja is only common below 1600 m a.s.l. (Palmer and Pitman, 1972: 1879), people may have collected this firewood some distance from the rock shelter. In layers attributed to MIS3 occupations, between ~43 and 42 ka ago, a broader range of taxa was represented, including several evergreen ones, such as Rapanea melanophloeos, Olea europaea (Layer 13) and Heteromorpha sp. (Stewart et al., 2016). Later, at about 41 ka ago, Erica sp. was included in this group of taxa. Melikane seems to have been abandoned during final MIS5 and for most of MIS4 (Stewart and Mitchell, 2018). Rose Cottage Cave dates imply some overlap in chronology with Melikane, though the hiatuses do not seem as protracted at Rose Cottage Cave.
These high altitude sites may have been deserted during the coldest part of MIS4. It is clear that Melikane and Rose Cottage should complement each other's environmental records and, to this end, the charcoal study from Rose Cottage is really important. We suspect that long and repeated occupations at Rose Cottage may, as at Melikane, have been because the area was better watered than the interior landscape, and this is one hypothesis that we shall test.
The focus of the present analysis at Rose Cottage is thus archaeological charcoal from Middle Stone Age layers dating between ~96 and ~35 ka (Table 1). The main aim is to identify woody plant species from the charcoal fragments in the cave, thereby enabling a study of variations in the vegetation cover through time. We shall investigate the compositional differences between plant communities occurring in the predominantly warm MIS5 conditions versus the cooler ones thought to have existed in MIS3.
Section snippets
Geographical and archaeological context of Rose Cottage Cave
Rose Cottage Cave is near Ladybrand, eastern Free State, South Africa (Fig. 1), on a hill known as Platberg, at an altitude of about 1660 m a.s.l., locality 29° 13′S, 27° 28′E. A large boulder closes the 20 × 10 m cave, but allows narrow entrances on either side of the cave and a skylight near the roof.
Reference charcoal
The KwaZulu-Natal reference collection (Allott, 2005) consisted of 209 indigenous species of modern wood and charcoal, including a database of scanned electron microscope (SEM) images and anatomical features (Allott, 2005, 2006). The specimens are designated here as LFA. The collection has been extended by 98 species of indigenous trees collected from farms in Limpopo and KZN with permission from the landowners, in addition to a donation by the Larry Leach Herbarium, University of Limpopo, of
Results
Twenty-one woody taxa were identified in charcoal. The anatomical diagnostic characteristics are given in the captions of Buddleja salvifolia and Calodendrum capense (ESM Fig. A1), Celtis africana and Kiggelaria africana (ESM Fig. A2) and Maytenus albata and Protea caffra (ESM Fig. A3). In all, plants from 21 woody taxa in 17 families were amongst the 640 fragments identified. The total number of identified fragments is 90 in layers 23-24 (MIS5), 373 in layers 13-22 (MIS4) and 177 in layers
Discussion
Since only 21 woody taxa were recorded in the Middle Stone Age charcoal assemblage from Rose Cottage, we make our environmental interpretation cautiously. We are also aware that some plants crosscut vegetation types and that the relationship between vegetation communities and habitats is complex. Nonetheless, we are encouraged to think that the signal from the charcoal record is a fair reflection of available fuel woods during the periods of occupation because species diversity was also
Conclusions
Rose Cottage Cave archaeological charcoal was identified from 24 Middle Stone Age layers with ages between ~96 ka and ~35 ka and this enabled a partial reconstruction of the vegetation communities through time. Firewood burned in an archaeological site represents human selection and many of the taxa identified in the cave are well-known for their good burning properties. For this reason, we accept that taxa not favoured for fuel may be absent from the cave even though they were present in the
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.
Acknowledgements
SJL would like to express her gratitude to Professor Marion Bamford for scientific guidance over several years. The Mellon Foundation (grant to MKB), and Evolutionary Studies Institute (ESI) provided equipment and the research was made possible by funding to LW from the National Research Foundation (NRF). Opinions expressed in this paper are not necessarily those held by the NRF. Material was collected from farms with permission from the landowner for the modern charcoal reference collections
References (94)
- et al.
Changes in the vegetation and human management of forest resources in mountain ecosystems at the beginning of MIS 1 (14.7–8 ka cal BP) in Balma Guilanyà (Southeastern Pre-Pyrenees, Spain). Changement dans la végétation et gestion humaine des ressources forestières dans des écosystèmes montagnards au début du MIS 1 (14,7–8 ka cal BP) à la Balma Guilanyà (sud-est Pre-Pyrénées, Espagne)
Compes Rendus Palevol
(2012) - et al.
Multiproxy record of late Quaternary climate change and Middle Stone Age human occupation at Wonderkrater, South Africa
Quat. Sci. Rev.
(2014) Late pliocene woody vegetation of area 41, koobi fora, east turkana basin, Kenya
Rev. Palaeobot. Palynol.
(2011)Identifying the woody resources of Diepkloof Rock Shelter (South Africa) using scanning electron microscopy of the MSA wood charcoal assemblages
J. Archaeol. Sci.
(2013)- et al.
A review of charcoal analysis as a tool for assessing Quaternary and Tertiary environments: achievements and limits
Palaeogeogr. Palaeoclimatol. Palaeoecol.
(2000) - et al.
Charcoal identification in species-rich biomes: a protocol for Central Africa optimised for the Mayumbe forest
Rev. Palaeobot. Palynol.
(2012) The identification of Cryptocarya spp., Lauraceae in archaeological charcoal
South Afr. J. Bot.
(2019)- et al.
Identifying asteraceae, particularly Tarchonanthus parvicapitulatus, in archaeological charcoal from the middle stone age
Quat. Int.
(2017) - et al.
Vegetation ecology of the southern Free State: 1. Plant communities of the Zastron area
South Afr. J. Bot.
(1999) - et al.
Age dating and the orbital theory of the Ice Ages: development of a high-resolution 0 to 300 000 year chronostratigraphy
Quat. Res.
(1987)
An archaeology of fuels: social and environmental factors in behavioural strategies of multi-resource management
Quat. Int.
An investigation of climatic change in the Holocene epoch using archaeological charcoal from Swaziland, Southern Africa
J. Archaeol. Sci.
Pollen-interpreted palaeoenvironments associated with the middle and late pleistocene peopling of southern Africa
Quat. Int.
Marine Isotope Substage 5e and the Eemian Interglacial
Global Planet. Change
Charcoal analysis and the “principle of least effort” a conceptual model
J. Archaeol. Sci.
Blade technology and tool forms in the middle stone age of South Africa: the Howiesons Poort and post-howiesons Poort at Rose cottage cave
J. Archaeol. Sci.
Afromontane foragers of the late pleistocene: site formation, chronology and occupational pulsing at Melikane rockshelter, Lesotho
Quat. Int.
Late Quaternary palaeoclimates and human-environment dynamics of the Maloti-Drakensberg region, southern Africa
Quat. Sci. Rev.
Anthracology and taphonomy, from wood gathering to charcoal analysis: a review of the taphonomic processes modifying charcoal assemblages, in archaeological contexts
Palaeogeogr. Palaeoclimatol. Palaeoecol.
Neanderthal firewood management: evidence from Stratigraphic Unit IV of Abric del Pastor (Eastern Iberia)
Quat. Sci. Rev.
Disentangling human from natural factors: taphonomical value of microanatomical features on archaeological wood and charcoal assemblages
J. Archaeol. Sci.: Report
Changing environments in oxygen isotope stage 3: reconstructions using archaeological charcoal from sibudu cave
South Afr. J. Sci.
Palaeoenvironments of the Middle Stone Age at Sibudu Cave, KwaZulu-Natal, South Africa: an Analysis of Archaeological Charcoal
Archaeological charcoal as a window on palaeovegetation and wood use during the middle stone age at sibudu cave
South. Afr. Humanit.
Reconstruction of the palaeoenvironment and anthropogenic activity from the Upper Pleistocene/Holocene anthracological records of the NE Iberian Peninsula
Biological Diversity in Lesotho: a Country Study
Reconstructing woodland vegetation and its exploitation by past societies, based on the analysis and interpretation of archaeological wood charcoal macro-remains
Environ. Archaeol.
New excavations at border cave, KwaZulu-natal, South Africa
J. Field Archaeol.
Macrobotanical remains from Wonderwerk Cave (Excavation 1), oldowan to late pleistocene (2 Ma to 14 ka BP), South Africa
African Archaeol. Rev.
Charcoal from pre-holocene stratum 5, wonderwerk cave, South Africa
Palaeoecol. Afr.
A reassessment of the wooden fragment from Florisbad, South Africa
J. Archaeol. Sci.
A submerged Late Cretaceous Podocarpaceous forest, west coast, South Africa
South Afr. J. Sci.
Lesotho Highlands Water project. Consulting services for the establishment and monitoring of the instream requirements of river courses downstream of LHWP dams
Riparian and instream vegetation, Lesotho
Wood anatomy of buddlejaceae
Aliso: J. Syst. Evolutionary Botany
The wood charcoal assemblage from Elands Bay Cave southwest Cape: principles, procedures and preliminary interpretation
S. Afr. Archaeol. Bull.
The wood charcoal evidence from renewed excavations at Elands Bay Cave, South Africa
South. Afr. Humanit.
Archaeological and Palaeoecological Implications of Charcoal Assemblages Dated to the Holocene from Great Zimbabwe and its Hinterland
Not just carbon: assessment and prospects for the application of anthracology in Oceania
Archaeol. Ocean.
Statistics for Archaeologists: a Common-Sense Approach
Plant taphonomy, flora exploitation and palaeoenvironments at the Middle Stone Age site of Mwulu's Cave (Limpopo, South Africa): an archaeobotanical and mineralogical approach
Archaeol. Anthropol. Sci.
Palaeoenvironmental Reconstruction from Pleistocene to Present: an Analysis of Charcoal from the Eastern Orange Free State and Lesotho
Palaeoenvironmental and archaeological implications of charcoal assemblages from Holocene sites in western Lesotho, southern Africa
Palaeoecol. Afr.
Climatic change across the Pleistocene Holocene boundary. Results of a factor analysis of charcoal assemblages
South. Afr. Field Archaeol.
A multiproxy analysis of late quaternary palaeoenvironments, Sekhokong range, eastern Lesotho
J. Quat. Sci.
Microscopic residues as evidence of hafting on backed tools from the 60 000 to 68 000 year-old Howiesons Poort layers of Rose Cottage Cave, South Africa
South. Afr. Humanit.
The effect of carbonization on wood structure of Dalbergia violaceae, Stryphnodendron polyphyllum; Tapira guianensis, Vochysia tucanorum and Pouteria torta from the Brazilian Cerrado
IAWA J.
Cited by (3)
Human occupation of the semi-arid grasslands of South Africa during MIS 4: New archaeological and paleoecological evidence from Lovedale, Free State
2022, Quaternary Science ReviewsCitation Excerpt :In addition, the Florisian fauna and the phytolith assemblages dominated by C3 grasses below SU4 are indicative of a well-watered environment. Further to the east, at Rose Cottage Cave, charcoal assemblages from the latest MIS 5 occupation at ∼78 ka (layer Kusel, pre-HP) show the presence of deciduous woodland typical of a humid environment (Lennox and Wadley, 2022; Pienaar et al., 2008). In the Maloti-Drakensberg, the basal portion of the Melikane sequence (Layers 30–27), which includes pre-HP artifacts dated to ∼83–80 ka, yielded a phytolith assemblage dominated by C3 grass short cells, consistent with a humid and cool grassland at high altitude (Pazan et al., 2022; Stewart et al., 2012, 2016).
Past environments and human lifeways of Lesotho and the wider Maloti-Drakensberg region of Southern Africa
2022, Quaternary InternationalCave and rock shelter sediments of southern africa: A review of the chronostratigraphic and palaeoenvironmental record from marine isotope stage 6 to 1
2021, South African Journal of Geology