Ostrich eggshell beads from Ga-Mohana Hill North Rockshelter, southern Kalahari, and the implications for understanding social networks during Marine Isotope Stage 2.

Ostrich eggshell (OES) beads from southern African archaeological contexts shed light on past traditions of personal ornamentation, and they are also argued to provide a proxy for understanding past social networks. However, OES beads are often understudied and not reported on in detail. In particular, there has been little research on OES bead variation during Marine Isotope Stage 2 (29,000-12,000 years ago) which includes the Last Glacial Maximum when changing climatic conditions are hypothesized to have significant impact on forager social networks. Here, we present the first technological analysis of terminal Pleistocene OES beads and fragments in the Kalahari from the ~15 ka levels at Ga-Mohana Hill North Rockshelter. We contextualise these findings through comparison with coeval OES bead assemblages across southern Africa during MIS 2. Results indicate that OES beads were manufactured at Ga-Mohana Hill North during the terminal Pleistocene occupation, based on the presence of most stages of bead manufacture. The review shows that OES beads were present across southern Africa through MIS 2, suggesting that culturing of the body was an embodied and persistent practice during that time. While the importance of OES beads as decorative objects was shared by populations across southern Africa, variation in bead diameters indicate that there was stylistic variation.

Introduction

People, past and present, use jewellery, such as personal ornaments, to culture their bodies and communicate information about themselves both within and between groups. Beadwork characteristics, such as material, association, size, and location of beadworks embed messages of cultural and social significance [1,2]. The ability to create, manipulate, and communicate through symbols is argued to be critical for developing and maintaining social connections within and between groups, which confers adaptive advantages in terms of social networks and potential safety nets [3-6].

Forager groups from the Kalahari Basin have provided insight into the manufacture of ostrich eggshell (OES) beads in modern ethnographic contexts [7,8]. These studies inform a range of approaches for interpreting the manufacture of OES beads, as well as their importance within social and cultural contexts in the past [9,10]. While these analogies are essential to our understanding of OES beads and their social context, one must be wary of how far back they can be extended. The analogies are based on ethnographic data from a small number of modern forager groups, mostly restricted to the Kalahari. Modern forager groups in southern Africa have been influenced by interaction and integration in shifting socio-political landscapes [11]. While practices such as OES bead manufacture have existed for thousands of years the uses and social practices surrounding these objects will have changed.

OES beads and fragments are rarely studied in detail, despite how common they are in Holocene contexts. Jacobson was the first to note differences in the sizes of OES beads between the archaeological assemblages of forager and pastoralists groups, where foragers generally manufactured smaller beads [12]. In recent years there has been more research interest in OES beads, with further studies examining the variation in bead size during the Holocene [13] and Late Pleistocene [14] across both eastern and southern Africa. Isotopic analyses of OES beads in Lesotho and the Maloti Drakensberg, showing that OES beads were traded across large distances (>300km) as far back as 33,000 years ago (ka) [5], while stylistic comparisons suggest social networks spanning eastern and southern Africa as far back as 55,000 years ago [14].

To date, there has been little research examining the variation in OES beads made by forager groups in Marine Isotope Stage 2 (MIS 2) (29–12 ka). During MIS 2, southern Africa experienced hypervariable climatic conditions, with many intervals that were overall cooler and drier than today, and may have presented unique environmental challenges to past foragers. Due to these conditions, MIS 2 has been hypothesized to represent a period of social coalescence and connectedness, especially after 22 ka, with the spread of ‘Robberg’-designated assemblages [15].

The recovery of OES from archaeological contexts in the semi-arid Kalahari Basin is relatively rare. Previous studies of OES in the Kalahari are relatively narrow in scope and include analysis of basic characteristics such as the mean diameter of beads or the number and weight of OES fragments [16]. Here, we provide the first detailed, technological study of OES beads from a MIS 2 context in this region at the site of Ga-Mohana Hill North Rockshelter (GHN). We further contextualise this assemblage through comparisons with coeval OES bead assemblages across southern Africa. These data are complemented with the abundances of OES fragments, where available, to provide a more nuanced understanding of the use of ostrich eggs and OES bead manufacture.

Ostrich eggshell beads

OES beads first appear in the archaeological record 50–40 ka in eastern and southern Africa and shortly thereafter in China [17-21]. Beads from southern Africa tend to generally be younger than their eastern African counterparts, with the oldest OES beads from southern Africa dating to 44–41 ka at Border Cave [18]. At Spitzkloof, two bead preforms were recovered from deposits dated to > 51 ka, but as isolated finds they remain difficult to interpret [22,23]. OES bead assemblages become ubiquitous in the Holocene, with both the size of assemblages and number of sites demonstrating OES beads increasing in magnitude in comparison to the Pleistocene [13]. This increase may be linked to preservation issues, broader uptake of OES beads as decorations, or a combination of both.

Research of OES beads in archaeological contexts generally focuses on identifying variation in bead style, specifically changes or differences in OES bead technological properties, such as shape and diameter, and using these differences as proxies for inferring stylistic differences between cultural groups [12-14,18]. More recently, isotopic analyses have also been used to track the movement of OES beads across the landscape [5,24], with the results informing our understanding of the nature and extent of past exchange networks.

Jewellery comprised of OES beads are popular exchange items in! hxaro, a custom of exchange and gift-giving among Kalahari San populations, who traditionally practiced a foraging lifestyle [7,8,25]. Within the context of! hxaro, OES beads may travel hundreds of kilometres, with the gifts helping to initiate and maintain social bonds between different groups and disperse access rights to landscapes and resources [26,27]. There are acknowledged issues with direct ethnographic comparisons between recent and Palaeolithic foragers [11,28,29] however, the movement of OES beads across distances of >300 km for at least the past 30 ka, suggests that they may provide a cautious heuristic for interpreting the nature and extent of social networks in the past [5,30].

While OES beads have a long history of study, especially in southern Africa [12,31], OES fragments have not received the same level of interest. Data on OES fragments are often presented minimally, such as in summaries of the number of fragments or possibly the weight of fragments reported. While this is understandable given that there could be thousands of OES fragments at a given site, OES fragments are intrinsic to our understanding of how beads were made and traded among people through time, especially when considering where they were manufactured and the potential social and environmental constraints of their production [30]. Studies that do focus on OES fragments tend to describe those that were modified by past peoples, such as the engraved fragments from Diepkloof [32].

Marine Isotope Stage 2

Marine Isotope Stages represent a summary of changes in global temperatures and sea-levels based on deep sea sediment cores [33]. MIS with odd numbers represent warmer interglacial periods with high sea-levels, and MIS with even numbers represent cooler glacial periods with low sea-levels. Archaeologists in southern African often use these MIS as temporal frameworks for the archaeological record [15,34-36]. MIS 2 is a glacial period from 29 to 12 ka, associated with the Last Glacial Maximum (LGM, conservatively dated to ~24–18 ka), when sea levels were globally an average of 125 m lower than they are today [37,38]. As a generalization, glacial periods in Africa are characterized by drier conditions that are often thought to pose resource challenges for foragers [39,40]. However, palaeoenvironmental analyses in some regions show the opposite of that general expectation [41]. In southern Africa, palaeoenvironmental records show cool temperatures during the LGM, with warming commencing about ~17 ka, then followed by a cool period ~13 to 11 ka [42]. The effect of these temperature changes on precipitation varied across southern Africa. Based on the compilation of several proxy palaeoenvironmental archives [42], the LGM was generally drier than it is today in the east and along the south coast, but wetter in the west along the coast and into the interior. However, this general pattern is not supported at all locales [43], attesting to the importance of investigating palaeoenvironments at the local scale.

The beginning of MIS 2 correlates roughly with a shift from Middle Stone Age (MSA) to Later Stone Age (LSA) technologies; lithic assemblages characterized by points and prepared cores shift to assemblages characterized by bladelets, backed pieces, and bipolar technology. Many LSA assemblages reflect miniaturization, which is an emphasis on the production of pieces < 3–5 cm in size, potentially a response to hafting needs and/or raw material conservation [44]. However, at some sites, this shift occurred much earlier than MIS 2 [45] and the technological shift was not synchronous across southern Africa. Two technocomplexes are associated with MIS 2; the ‘Early LSA’ and the ‘Robberg’ [46]. Early LSA assemblages show a high degree of variability. Robberg-designated assemblages date to ~18–12 ka and are characterized by systematic bladelet production with rare retouched pieces, that include backed bladelets [47]. Several assemblages across southern African have been designated to the Robberg technocomplex, including Nelson Bay Cave on the south coast, Elands Bay cave on the west coast, Dikbosh 1 at the southern edge of the Kalahari, and Sehonghong in the Lesotho highlands [46].

Mackay et al. [15] hypothesise that MIS 2 was a period of social coalescence, particularly after 22 ka, with wide-spread distribution of the Robberg-designated assemblages across various environments. This is supported by the similarity in flaking systems, implement types, provisioning system, and substantial evidence for different kinds of ornamentation (ostrich eggshell beads, bone beads, and shell pendants). Using stable isotope analysis, Stewart et al. [5] show that OES beads were traded across large distances during MIS 2 into Lesotho, and outside the known range of ostriches (Struthio camelus), which indicates long distance macro-scale social networking.

Ga-Mohana Hill North Rockshelter

Ga-Mohana Hill is located in the southern Kalahari Basin, 12 km northwest of Kuruman on the eastern edge of the Kuruman Hills in South Africa (Fig 1A). The hill has a maximum elevation of 1531 masl, which is about 100-150m above the surrounding landscape. GHN, the largest shelter on the hill, is a long curved and relatively shallow shelter, facing northwest with an impressive view across the landscape, including the Kuruman River. The North of Kuruman Palaeoarchaeology Project began excavating GHN in 2016 and to date has excavated a total of 4.75 m2 of sediment across three areas of the shelter, reaching a maximum depth of 1.7 m [48]. The majority of this (4 m2) has been excavated from Area A, which reveals stratified MSA and LSA deposits (Fig 1B). The top ~10 cm of sediment is loose surface sediment rich in ash and dung, below which are three stratigraphic aggregates. From top to bottom these are Dark Brown Gravelly Silt (DBGS), Orange Ashy Silt (OAS), and Dark Brown Silt and Roofspall (DBSR). Analysis of the slope and orientation of plotted artefacts shows that the artefacts in all these stratigraphic aggregates are in near primary context [48]. Single grain OSL dating on quartz grains has shown that the stratigraphic aggregates date to 14.8±0.8 ka (DBGS), 30.9±1.8 ka (OAS), and 105.2 ±3.3 ka (DBSR) [48,49]. During excavation, all visible artefacts were piece-plotted using a total station.

Fig 1

Map showing the study area and details of Ga-Mohana Hill North Rockshelter.

A—The location of GHN relative to the boundaries of the Kalahari Desert and Basin. B–Map of GHN rockshelter with excavated areas. Red line marks profile shown in C. C -Schematic of stratigraphic boundaries, stratigraphic aggregate assignments, and optically stimulated luminescence sample locations and results. Grey-shaded areas are rock.

The focus of this paper is the OES assemblage from the DBGS deposits in Area A dated to ~15 ka. The DBGS contains numerous lithic artefacts, fragmentary faunal remains, charcoal, and rare ochre pieces. The DBGS also includes numerous OES pieces, including fragments, beads, and bead preforms. Overall, the artefact density in the DBGS is 0.42 artefacts per litre of sediment [48]. The lithic artefacts include bladelets and rare backed pieces [48] and based on this are consistent with a LSA, potentially Robberg [47], designation. The lithic artefacts are manufactured primarily from diverse locally available materials including chert and banded ironstone.

Methods

The study was limited to analysis of OES assemblage from the DBGS stratigraphic aggregate layer in order to investigate and compare MIS 2 sites with OES assemblages across southern Africa. The OES assemblage from layer DBGS at GHN was evaluated following criteria including colour, manufacturing stage, maximum diameter, maximum thickness, aperture diameter, presence of residue, fragment shape [50-54]. The maximum length, width and thickness and fragment shape were recorded for OES fragments.

Colour was attributed to each piece qualitatively, following Collins and Steele [50]. OES colour is a useful tool for understanding whether a shell might have been exposed to heat and the temperature of the heat source. The colouring of beads and fragments ranged from yellow to black (Table 1). OES becomes yellow, red, iridescent and grey when heated under oxidising conditions [32,50,55], while reducing conditions are more likely to produce blackening of the shell. Blackening of OES has not been replicated in experimental studies but is relatively common in the archaeological record [50,52,56]. All beads and fragments were examined with a hand lens (20x magnification) and pieces that showed signs of pigmentation and/or usewear were further examined using a stereo microscope at 10-100x magnification. Pigmentation was classified visually based on colour and texture. Where usewear was identified it was described in terms of location and morphology(facets or striations) following Dayet et al. [51] and Collins et al. [30]. Striations are defined as randomly oriented short marks, while facets are depressions on the surface of the bead. The shape of each fragment was recorded following Miller [10]. Striations, chips, patina and smoothing were recorded for the aperture and outer rims of beads, additionally aperture shape was recorded following Miller [10].

Table 1

Colouring of OES beads, preforms and fragments from the MIS 2 level at GHN.

Colour was assessed following protocol outlined by Collins and Steele [50]. Counts and percentages are recorded.

Colour	Beads and preforms		Fragments
	n	%	n	%
Unburned	6	21.4	7	33.3
Yellow	20	71.4	10	47.6
Red	0	0	4	19
Black	2	7.2	0	0
Iridescent	0	0	0	0
Grey white	0	0	0	0
Total	28		21

OES beads were assigned a manufacturing stage following Orton’s (2008) classification scheme, where stages II-V represent preforms (i.e., beads that have not been completed). Stage II pieces have been partially drilled, but not yet pierced through, while stage III pieces have been completely drilled. Stages IV and V are pieces that have respectively been partially trimmed and completely trimmed. From stage VI pieces are considered finished beads, where stage VI pieces are partially ground and stage VII pieces are completely ground. Finished beads are those that are ready for use as jewellery and/or decoration [30]. Beads were also assigned a manufacturing pathway following Orton [54]. The pathway was determined by presence of drilled but not ground fragments (pathway 1) or ground but undrilled fragments (pathway 2). Beads manufactured following pathway 1 are first drilled and then trimmed, whereas beads made following pathway 2 are first trimmed into round fragments and then drilled. Additionally, it was recorded whether the beads/preforms were broken or complete.

Maximum length and width were measured for all preforms and broken beads that retained less than 50% of their original circumference. Maximum diameter and aperture were measured for finished beads. Maximum thickness was measured for both preforms and beads. While preforms in stages II-V do not have apertures because the drill hole has not completely perforated the shell, the aperture of the drill hole was measured. Usewear traces were noted and identified as facets or striations [30,51].

Fragments were analysed similarly to beads, with maximum length, width and thickness recorded for each fragment. Any marking on the face of the fragments was also noted, both pigment traces and small striations. Any markings were identified as anthropogenic or taphonomic depending on their morphology. Shallow, randomly directed scratches typical of friction from fragments moving in both longitudinal and lateral directions were assigned as taphonomic markings [32]. Intentional markings, characterised as deeper marking with an U or V shaped morphology were designated as anthropogenic. Any other markings that could not confidently be ascribed as taphonomic or anthropogenic were recorded as surface modifications.

Spatial analysis was conducted to examine the patterning of the OES assemblage within the DBGS sediment both vertically and horizontally. All spatial analyses were conducted in R [57], mainly using the spatstat package [58]. The spatial patterning of all plotted OES fragments and beads within the excavation area was visually assessed by creating a relative risk surface [59], showing the probability of how likely OES artefacts are to co-occur with other artefact types.

To facilitate comparisons and an understanding of the dynamics of bead manufacture during MIS 2 in southern Africa, a regional comparative database was created. We conducted a literature review to identify well-described and chronometrically dated sites with OES assemblages within southern Africa (South Africa, Lesotho, Eswatini, Namibia, Botswana). We report the level from which the bead assemblages at each site were recovered, as well as the date, where reported, how the beads were classified (as preforms or finished beads), and metric data when available. The classification of finished beads or preforms relies on the initial authors’ identification, as much of the data was published prior to the classification schemes used in this analysis [52,54]. These published data were then compared to the GHN results. The maps to illustrate these data were made in the R statistical environment [57]. Data for the ostrich distribution were taken from the South African Bird Atlas Project 2 [60], and then interpolated to cover the entire area (southern Africa). The country border polygons were imported into R from Natural Earth, which provides free vector and raster data using the rnaturalearth package [61]. The data on OES beads and fragments from sites across southern Africa were added as a layer to each map to highlight different aspects of the data. This was done using the ggplot and scatterpie packages [62,63].

The data and code for these maps are available in the supplementary information (S2 Appendix) and in a GitHub repository (https://github.com/amyhatton/ghn_bead_paper). All necessary permits for archaeological investigations at Ga-Mohana Hill were obtained via informed written consent from the South African Heritage Resource Agency (Permit ID 2194). The land is owned by the Baga Motlhware Traditional Council and informed written consent was granted by them to conduct the study. No protected species were sampled and the study did not involve animals. All necessary permits were obtained for the described study, which complied with all relevant regulations. All specimen numbers relevant to this study are provided in S2 and S3 Tables. These specimens are currently housed in the Archaeology Department at the University of Cape Town and they will be permanently curated by the McGregor Museum, Kimberley, Northern Cape, South Africa.

Results

The OES bead assemblage from layer DBGS consists of 19 beads, 9 bead preforms, and 21 OES fragments. In total 1278 L of sediment was excavated from layer DBGS, thus the density of OES in the sediment is 0.04 pieces per litre of sediment.

Technological analysis

While the assemblage of OES beads from DBGS represents only 28 beads and preforms, almost all stages of bead manufacture are present [54] (Fig 2, S2 Table). The only stage that is absent is stage III. The presence of preforms and beads in almost all stages of manufacture indicates that OES beads were being manufactured at GHN during MIS 2. The OES beads have a mean aperture diameter of 1.4 mm and a mean external diameter of 4.4 mm (Table 2). Beads and preforms have a mean thickness of 1.6 mm. The external diameter of the OES beads from layer DGBS at GHN falls within the range noted for eastern and southern African foraging populations [13]. More than half of the finished beads (10/19, 53%) have a red residue.

Fig 2

OES beads and preforms from Layer DBGS at Ga-Mohana Hill North Rockshelter.

Find 4082 was potentially manufactured following Pathway 2. Scale divisions are 5mm.

Table 2

Technological features for the OES beads and preforms from the MIS 2 deposit at Ga-Mohana Hill North Rockshelter.

Complete beads are those identified as stage VI and VII following Orton [54]. Standard deviations are included in brackets where applicable.

Technological attributes	Finished beads	Preforms
# unbroken beads	16	1
# broken beads	3	8
total # beads	19	9
mean exterior diameter in mm (standard deviation)	4.4 (0.42)	-
exterior diameter range in mm	3.69–5.45	-
mean aperture diameter in mm (standard deviation)	1.5 (0.25)	-
aperture range in mm	1.04–2.05	-
mean thickness in mm (standard deviation)	1.6 (0.19)
thickness range in mm	1.2–1.8
# complete beads with use-wear	13	-
# of beads and preforms with residue	10
# of completed beads with both use-wear and residue	7	-

The majority of preforms have been manufactured using Pathway 1, where the bead is first perforated and then trimmed into a disk [54]. One bead may have been manufactured following Pathway 2 where an OES fragment is first trimmed and then perforated to create a bead, however Pathway 2 beads are rare in archaeological assemblages, as they are difficult to distinguish from Pathway 1 beads [54].

Twenty-one (21) OES fragments were recovered from the DBGS layer at GHN (S3 Table). The fragments have a mean length of 14.6 mm (range = 9.69 to 21.95, sd = 2.9), a mean width of 10 mm (range = 4.58 to 14.73, sd = 2.6) and a mean weight of 0.49 g (range = 0.12 to 0.84, sd = 0.2). The only two shapes represented in the fragments are polygonal and triangular. Most of the pieces were a polygon shape (n = 15), with the remaining six being a triangular shape. The OES fragments from the DBGS layer at GHN amount to 10.3 g, which represents a minimum of 1 ostrich eggshell. Commercial ostrich eggshells average at 222 g, with a range of 180–292 g [64]. Five of the 21 fragments (24%) exhibited red residue staining on their surface.

Taphonomy

Many fragments had evidence of scratches on their surface. These randomly directed shallow marks are likely of taphonomic origin from OES pressing against harder objects in the sediment, possibly through trampling [32]. Both beads and fragments have similarities in level of heat exposure for layer DBGS at GHN, with the majority in the yellow category which is indicative of OES being heated to about 200 ˚C [50]. There are however a few differences between the burning patterns of beads and preforms compared to fragments, particularly for those classified as black or red. No beads were classified as red, however 19% of the OES fragment assemblage is red, which indicates that these fragments were heated to between 300–350 ˚C. About 7% of the beads and preforms are blackened, showing that they were likely burned in an environment with limited oxygen, whereas no fragments were blackened. These differences are statistically significant, however the sample size is small (chi-squared test, χ2 = 8.59, p = 0.035).

Spatial analysis

The artefacts are uniformly spaced horizontally across excavation Area A at GHN except for the SE corner (Fig 3) where a large block of roofspall was removed during excavations. There is slight patterning in the distribution of OES; OES artefacts are most likely to occur with non-OES artefacts in the northeast corner of the excavation (Fig 3B). Vertically, artefacts are quite evenly spaced through the DBGS, but most OES artefacts occur in the upper 40 cm of the stratigraphic aggregate (Fig 3). There is one OES fragment that occurs lower than other OES artefacts, in the southeast corner. The relative risk surface shows that laterally there are two clusters where OES artefacts are likely to occur with non-OES artefacts (Fig 3A and 3C).

Fig 3

Relative risk surface for OES artefacts (fragments and beads) compared to all other plotted artefacts in DBGS layer at GHN.

Non-OES artefacts are shown as black circles while OES artefacts are shown as white squares. The darker orange indicates areas where OES artefacts and other artefacts have a higher probability of co-occurring, while lighter areas are indicative of areas where OES artefacts and other artefacts are not likely to co-occur. A- lateral view of DBGS layer looking west. B- Aerial view of DBGS layer. C- lateral view of DBGS looking north.

Regional comparison

We reviewed the literature to synthesise data on OES bead assemblages from MIS 2 (Tables 3 and S1) and situate the OES bead assemblage from GHN in regional context. OES beads and fragments were unevenly reported across sites, which makes it challenging to compare quantitatively. However, a few useful observations and comparisons can be made. The majority of assemblages are small, with fewer than 50 OES beads and preforms (Fig 6). The exceptions are Nelson Bay Cave, Boomplaas, Bushman’s Rockshelter and Dikbosch, which date to the terminal part of MIS 2 and where bead counts are in the range of 74 to 170. Interestingly, all sites within Lesotho/Maloti Drakensberg have very low counts of beads and preforms compared to sites in the rest of southern Africa (Fig 5).

Table 3

MIS 2 dated sites in southern Africa with OES assemblages.

Modern ostrich prevalence has been calculated from SABAP 2 [60].Ostrich sighting data was interpolated across southern Africa and the values included here are the mean of a 5km radius around each site location. More detailed information on the sites and OES assemblages are available in S1 Table.

Site	Age	OES assemblage description	Modern ostrich prevalence (sighting percentage)	References
Sehonghong	14.5–25 ka	9 finished beads and no preforms or fragments.	17	[65,66]
Ha Makotoko	15–29ka	12 finished beads and no preforms or fragments. Mean diameter of beads is 3.1mm.	30	[67,68]
Melkhoutboom Shelter	18.7 ka	5 finished beads and 6 preforms, there were 393 fragments of which 2 were decorated.	37	[69]
Apollo 11	13–14.5 ka	4481 fragments, but no beads or preforms.	44	[70]
Nelson Bay Cave	11.7–14.5 ka	80 finished beads and 53 preforms. 35 fragments were reported.	13	[71–73]
Bushman’s Rock Shelter	10–13 ka	83 finished beads, 101 preforms and 419 fragments. The mean bead diameter is 5.3mm.	19	[31]
Umhlatuzana	16 ka	4 finished beads, no preforms or fragments.	21	[74,75]
Rose Cottage Cave	16 ka	1 finished bead and 25 fragments.	36	[76,77]
Boomplaas	15–22 ka	63 finished beads, 13 preforms, 4025 fragments.	26	[71,78,79]
Ntloana Tsoana	14 ka	2 finished beads.	30	[68]
Spitzkloof A	23.5 ka	2 finished beads and 2 preforms. 1179 OES fragments.	34	[80]
Ga-Mohana Hill North Rockshelter	15 ka	19 finished beads, 9 preforms and 21 fragments. Mean bead diameter is 4.4mm.	48	[48]
Drotsky’s Cave (Gcwihaba Cave)	14.5 ka	2 finished beads and 197 fragments.	46	[81]
Dikbosch 1	14.5–16.5 ka	48 finished beads, 44 preforms and 12157 fragments. Mean bead diameter is 4.5mm.	69	[16,82]
Heuningneskrans	23 ka	1 finished bead.	21	[83]
Buffelskloof	27 ka	10 finished beads, 30 preforms. Mean bead diameter is 4.3mm.	25	[84]
Grassridge	11.6–13.5 ka	9 finished beads, 28 preforms and 573 fragments. Mean bead diameter is 3.5mm.	35	[30,85,86]
Txina Txina	25–29 ka	2 finished beads and 28 fragments.	17	[87]

Fig 6

Map of southern Africa showing archaeological sites dated to MIS 2 that have OES bead assemblages.

OES bead counts for sites, overlaid on a map of modern ostrich prevalence based on South African Bird Atlas Project 2 data [60]. BK–Buffelskloof, BP–Boomplaas, NBC–Nelson Bay Cave, MHB- Melkhoutboom, GRS—Grassridge Rockshelter, NT–Ntloana Tsoana, RCC–Rose Cottage Cave, HM—Ha Makotoko, SHH–Sehonghong, UMH–Umhlatuzana, BRS–Bushmans Rockshelter, HNK—Heuningneskrans, TXI—Txina Txina, GHN–Ga-Mohana Hill North, DB—Dikbosch, DC—Drotsky’s Cave (Gcwihaba Cave), APL—Apollo 11, SPZ—Spitzkloof.

Fig 5

Map of southern Africa showing archaeological sites dated to MIS 2 that have OES bead assemblages.

Pie charts show the proportion of finished beads to preforms and OES fragments at each site, overlaid on a map of modern ostrich prevalence based on South African Bird Atlas Project 2 data [60]. BK–Buffelskloof, BP–Boomplaas, NBC–Nelson Bay Cave, MHB- Melkhoutboom, GRS—Grassridge Rockshelter, NT–Ntloana Tsoana, RCC–Rose Cottage Cave, HM—Ha Makotoko, SHH–Sehonghong, UMH–Umhlatuzana, BRS–Bushmans Rockshelter, HNK—Heuningneskrans, TXI—Txina Txina, GHN–Ga-Mohana Hill North, DB—Dikbosch, DC—Drotsky’s Cave (Gcwihaba Cave), APL—Apollo 11, SPZ—Spitzkloof.

Sites in areas where modern ostrich prevalence is low have high proportions of finished beads to preforms (Fig 4). The exception is Heuningneskrans which has no preforms but is in an area with moderate ostrich prevalence. Interestingly, at BRS which is 38 km away, there is a roughly equal proportion of preforms to finished beads. The Heuningneskrans assemblage does however only consist of 1 OES bead currently.

Fig 4

Map of southern Africa showing archaeological sites dated to MIS 2 that have OES bead assemblages.

Pie charts show the proportion of preforms to finished beads at each site, overlaid on a map of modern ostrich prevalence based on South African Bird Atlas Project 2 data [60]. SPZ–Spitzkloof A, BP–Boomplaas, BK–Buffelskloof, NBC–Nelson Bay Cave, MHB–Melkhoutboon, GRS—Grassridge Rockshelter, NT–Ntloana Tsoana, RCC–Rose Cottage Cave, SHH–Sehonghong, HM–Ha Makotoko, UMH–Umhlatuzana, BRS–Bushmans Rockshelter, HNK—Heuningneskrans, TXI—Txina Txina, GHN–Ga-Mohana Hill North, DB—Dikbosch, DC—Drotsky’s Cave (Gcwihaba Cave).

Ntloana Tsoana, Ha Makotoko, Sehnghong, Umhlatuzana, and Heuningneskrans are the only sites with high proportions of finished beads compared to preforms and fragments of OES, most other sites are dominated by high proportions of preforms and fragments (Fig 5). The number of recorded OES fragments for sites across southern Africa ranges from 1 to 12157 (Fig 7), although most sites have less than 1000 fragments recorded.

Fig 7

Map of southern Africa showing archaeological sites dated to MIS 2 that have OES bead assemblages.

OES fragment counts for sites, overlaid on a map of modern ostrich prevalence based on South African Bird Atlas Project 2 data [60]. BK–Buffelskloof, BP–Boomplaas, NBC–Nelson Bay Cave, MHB- Melkhoutboom, GRS—Grassridge Rockshelter, NT–Ntloana Tsoana, RCC–Rose Cottage Cave, HM—Ha Makotoko, SHH–Sehonghong, UMH–Umhlatuzana, BRS–Bushmans Rockshelter, HNK—Heuningneskrans, TXI—Txina Txina, GHN–Ga-Mohana Hill North, DB—Dikbosch, DC—Drotsky’s Cave (Gcwihaba Cave), APL—Apollo 11, SPZ—Spitzkloof.

We compared mean bead diameter for occupations from sites that recorded this information (Fig 8) and found that the mean bead diameters are significantly different to one another (ANOVA, F = 9.935, p = 0.0124). A Tukey pairwise comparison shows that mean bead diameters differ significantly between Grassridge and Bushman’s Rock Shelter (p = 0.045), Ha-Makotoko and Bushman’s Rock Shelter (p = 0.008), and finally Ha-Makotoko and Dikbosch (p = 0.03).

Fig 8

Map of southern Africa showing archaeological sites dated to MIS 2 that have OES bead assemblages.

Mean OES bead diameters for sites, overlaid on a map of modern ostrich prevalence based on South African Bird Atlas Project 2 data [60]. BK–Buffelskloof, GHN–Ga-Mohana Hill North, DB—Dikbosch, GRS—Grassridge Rockshelter, HM-Ha Makotoko, BRS–Bushmans Rockshelter.

Discussion and conclusions

Bead manufacture

The presence of OES beads in all stages of manufacture suggests that beads were being made at GHN ~15 ka. There is however a higher proportion of beads compared to preforms, which could indicate that drilling and grinding of beads were conducted in different areas of the rockshelter. OES fragments are outnumbered by worked OES, which is interesting and may also suggest that bead manufacture took place in an as yet unexcavated part of the rockshelter, or perhaps that OES was used in a conservative manner during this period at GHN. Extending the spatial extent of this excavation will provide more clarity.

It is interesting to note that modern ostrich prevalence at GHN is quite high (48%, Table 3). Although data on changes in ostrich distribution in the late Quaternary are limited, if we assume that ostrich distribution has been relatively consistent across wet and dry phases over the last 15 ka [88], then ostrich eggs and their shells would have been regularly encountered. The relatively small number of both recovered OES beads and recovered OES fragments, however, suggest that these activities may not have been intensively practiced at GHN during this occupation and may reflect less intensive site use. Further excavation expanding the spatial extent of this occupation at GHN will further inform this hypothesis.

Of interest is the bead that was potentially manufactured using Pathway 2, as well as the black beads. Beads manufactured using Pathway 2 are rare in the archaeological record, with Orton [54] noting 5 beads conforming to Pathway 2 out of 465 from the late Holocene site KN2006/067 in Namaqualand, South Africa. One reason for this rarity, is that this manufacturing pathway is challenging to identify unless recovered in Stages II-IV. Once the hole has punctured through the OES preform it is difficult to know whether the bead was trimmed or drilled first. This bead at GHN is currently the oldest described bead that may have been manufactured following Pathway 2 in southern Africa and may suggest that diverse strategies for the manufacture of OES beads are an early characteristic of this technology. Further consideration and research are needed to better understand how differences in manufacturing strategies relate to resource availability, social context of production, and potentially inform cultural inheritance.

One bead and one preform (Stage II) recovered from DBGS were coloured black. This finding is of interest because of the absence of black colouration in the OES fragments, as well as the specific conditions noted for obtaining a black colour for OES in general [50,52,56]. The absence of OES fragments that were blackened suggests that this may reflect a deliberate behaviour for the colouration of beads, in that we would also expect to see black OES fragments if the colouration resulted from taphonomic processes. However, the data on the colouration of OES fragment assemblages is largely unreported, which limits comparisons with other sites. Further taphonomic research, especially regarding OES fragment assemblages, is required to better link bead colour to anthropogenic intent.

There is some evidence of red coloured residues located within the aperture and/or associated with wear facets of OES beads, which may indicate that the bead came into contact with an ochred surface prior to deposition. However, the presence of red coloured residues on fragments and randomly positioned on beads suggests that these residues likely result from contact with ochre within the sediment peri- or post-deposition [30].

Spatial patterning

Ethnographic data on OES bead manufacture is limited, which in turn limits our understanding of how beads were manufactured. Where OES bead manufacturing activities take place within sites is also poorly understood. It is possible that such social activities might have taken place near to hearths, however this needs to be determined by more research into spatial patterning of OES. At GHN, distinct hearth features are not identifiable in the sediments of Area A, despite evidence for burning (burned bone, charcoal). While there is some spatial patterning, in general, the distribution of OES at GHN is not strongly clustered in any one specific area.

Many of the OES artefacts are found in small clusters of 2–5 artefacts that could relate to the manufacturing process. Most of these clusters are made up of OES fragments, so it is possible that these fragments broke once they had been deposited in the sediment.

Extending the excavation area at GHN will allow for a more nuanced understanding of OES manufacture at the site and shed light on whether there was spatial partitioning of intensive OES bead manufacture in the rockshelter.

Regional context

Our literature review identified several sites that contain OES bead assemblages dating to the latter part of MIS 2 in southern Africa (Tables 3 and S1). This review identifies the presence of OES beads across much of southern Africa during this period and emphasises their importance for terminal Pleistocene foragers. OES beads reflect decorative objects and the practice of culturing the body [9], and therefore their presence suggests that this practice was embodied and persistent across the foraging groups in southern Africa during this time. This conclusion is consistent with OES bead research focusing on Holocene assemblages [13].

Moreover, this conclusion is also consistent with the argument for MIS 2 being a period of social coalescence within southern Africa [15]. Importantly, the terminal Pleistocene effectively demonstrates the first period in the southern African archaeological record where we see one type of non-lithic stylistic object occur across southern Africa. As stated above, the presence of OES beads across southern Africa indicates their cultural and social importance and further emphasises that the importance and meaning of OES bead decorative objects was transmitted and shared between populations and groups living in diverse parts of the sub-continent. In this regard, the presence of OES beads across southern Africa during the terminal Pleistocene indicates region-wide social connections, and aligns with the lithic data presented by Mackay et al. [15] for a period of social coalescence during MIS 2. The persistence, and indeed increase, in OES bead assemblages (both in terms of numbers of beads, as well as sites demonstrating beads) [10,13], speaks to the strengthening of these social connections during MIS 1.

However, there are stylistic differences in terms of bead diameter across southern Africa during the terminal Pleistocene (Fig 8). Of note are the significant differences in bead diameter between Grassridge and Bushman’s Rock Shelter, Ha-Makotoko and Bushman’s Rock Shelter, and Ha-Makotoko and Dikbosch 1. Ha-Makotoko and Bushman’s Rock Shelter have the smallest and largest bead diameters respectively, and reflect a diversity in mean bead diameter across southern Africa during late MIS 2 that ranges from 3 to more than 5 mm. These differences may reflect local stylistic variation and preferences, potentially with large-sized beads being favoured in the northeast of southern Africa, as indicated by Bushman’s Rock Shelter [31,51], intermediate-sized beads in the western part of southern Africa, as indicated by Buffelskloof, Dikbosch1, and GHN, and small-sized in the Drakensberg and sub-escarpment as indicated by Ha-Makotoko and to a lesser extent Grassridge Rockshelter (Fig 8). This bead size variation does not appear to correlate with the prevalence of ostriches (Spearman’s Rank Correlation, rho = -0.02, p-value = 1).

The absence of preforms suggests that beads were likely not manufactured at Ha-Makotoko, where bead diameters are the smallest and therefore, the majority of OES beads are argued to have been imported to the region [4,5]. The movement of OES beads over potentially hundreds of kilometres complicates our understanding of regional bead diameter diversity. Usewear and taphonomic processes may also affect bead diameter [89]. However, the nature of these processes is not yet well understood, and the OES beads from Ha-Makotoko have yet to undergo a taphonomic analysis. That being said, the OES beads from Ha-Makotoko cluster around a mean diameter of 3.1 mm, with a tight range from 3–3.2mm, and are suggestive of a preference or requirement for this smaller bead size.

These data therefore indicate potential pockets of localised stylistic variation in OES bead manufacture. This is of interest, as MIS 2 is suggested to be a period of social coalescence, in part indicated by the widespread occurrence of sites attributed to the Robberg technocomplex. However, the use of a lithic technocomplex likely masks diversity in lithic assemblage compositions during this period, and that technological variation becomes much stronger after 14 ka [15]. This pattern seems to fit with the OES bead data discussed above, in terms of increasing local influences and stylistic diversity, and suggests that while social networks were persistent during this period, there was also an increase in local stylistic innovation. The increasing diversity in material culture during this period may potentially relate to the diverse environmental responses to global cooling across southern Africa during MIS 2 [42], but more work focused on local environmental contexts at these various sites is required.

From a local perspective, the site nearest to GHN is Dikbosch 1 [16,82], located 152 km from GHN. Dikbosch 1 has four occupation layers that date to MIS 2, these are dated as contemporary and slightly earlier than GHN with a range of 14.5–16.6ka (Figs 4–8, S1 Table). OES bead sizes are similar in terms of mean diameter, which is suggestive of stylistic, and potentially social, continuity between the two sites. However, the sites differ in terms of number of beads, ratio of beads to preforms, and size of the OES fragment assemblages (Figs 4, 6 and 7; Table 3). Dikbosch 1 has a larger OES bead assemblage both in terms of number of OES beads and preforms. The ratio of beads to preforms also differs, with Dikbosch 1 demonstrating more preforms to beads, as opposed to GHN, where we see more beads than performs. Moreover, Dikbosch 1 has a much larger OES fragment assemblage. Part of this difference may relate to Dikbosch 1 being located in an area with a higher prevalence of ostriches at 69% compared to the 48% at GHN (Table 3). The scale of the difference, especially with regard to the OES fragment assemblages, also suggests that ostrich eggs and OES may have been of more importance, both as a subsistence resource, as well as a raw material, at Dikbosch 1 and that bead manufacture was likely more intensive at this site. In this respect, we suggest that OES beads were being manufactured at GHN, but that the practice was not as intensive as other sites that have been described as “bead factories” [30,54].

OES bead reporting

Also of note is the lack of standardised reporting for OES bead assemblages in the literature. This in part reflects the historical lack of attention given to OES assemblages (both beads and fragments) in archaeological research, perhaps because they were considered to be of less interpretive value than other artefacts classes. Regardless, presenting detailed descriptions of OES beads (including major attribute data), as well as OES fragments, provides important insight into past behaviours at local, regional, and sub-continental scales, as well as taphonomic information that informs site formation processes. Many sites we discuss in this review lack much of these data restricting our understanding of the role of OES and OES beads in the past. In this respect, we concur with Miller and Wang [14], Miller [10], Miller and Sawchuk [13], Collins et al. [30], and Collins [9] in arguing for greater attention to OES and OES bead assemblages, and specifically for providing crucial (and fundamental) technological data for these assemblages. We now know the practice of culturing bodies with OES beads was widespread geographically across southern and eastern Africa and China during the Late Pleistocene. Studying the manufacture and distribution of beads in detail can offer further opportunity to understand social interaction between people on the landscape during MIS 2.

Southern African MIS 2 ostrich eggshell assemblages.

Database of ostrich eggshell assemblages at archaeological sites dating to between 29 and 12ka.

(CSV)

Click here for additional data file.

Ga-Mohana Hill North Rockshelter ostrich eggshell bead data.

Technological data for each plotted find ostrich eggshell bead from the DBGS level at Ga-Mohana Hill North Rockshelter.

(CSV)

Click here for additional data file.

Ga-Mohana Hill North Rockshelter ostrich eggshell fragment data.

Technological data for each plotted find ostrich eggshell fragment from the DBGS level at Ga-Mohana Hill North Rockshelter.

(CSV)

Click here for additional data file.

Code for making the maps of southern African MIS 2 ostrich eggshell assemblages.

PDF document of RMarkdown version of the code for making the maps of southern African MIS 2 ostrich eggshell assemblages.

(PDF)

Click here for additional data file.

Code for statistical analysis of ostrich eggshell data.

PDF document of RMarkdown version of the code for statistical analysis of southern African MIS 2 ostrich eggshell assemblages and Ga-Mohana Hill North Rockshelter assemblages.

(PDF)

Click here for additional data file.

3 Jan 2022

16 Mar 2022

Response to Reviewers

Manuscript PONE-D-21-22127

Dear Dr Spinapolice,

Thank you for giving us the opportunity to submit a revised draft of the manuscript “Ostrich eggshell beads from Ga-Mohana Hill North Rockshelter, southern Kalahari, and the implications for understanding social networks during Marine Isotope Stage 2.” for publication in PLOS ONE.

We thank the reviewers for their constructive suggestions, which we addressed in this revised submission, and we believe that the manuscript is now substantially improved. Below is a detailed point-by-point response to each reviewer's comment.

Editors Comments/Journal Requirements:

1. Please ensure that your manuscript meets PLOS ONE's style requirements, including those for file naming.

Authors' response: We have made sure that the manuscript meets the requirements.

2. In your manuscript, please provide additional information regarding the specimens used in your study. Ensure that you have reported specimen numbers and complete repository information, including museum name and geographic location. If permits were required, please ensure that you have provided details for all permits that were obtained, including the full name of the issuing authority, and add the following statement: All necessary permits were obtained for the described study, which complied with all relevant regulations.' If no permits were required, please include the following statement: 'No permits were required for the described study, which complied with all relevant regulations.'

Authors' response: Supplementary Tables 2 and 3 have the specimen number, we have renamed the columns from ‘finds’ to ‘specimen number’ to indicate this. We have included more information about where the specimens are located in the methods section. And included this sentence to the methods section: “All necessary permits were obtained for the described study, which complied with all relevant regulations. Excavations at Ga-Mohana Hill were approved by the South African Heritage Resources Agency under permit ID 2194. All specimen numbers relevant to this study are provided in Supplementary Tables 2 and 3 . These specimens are currently housed in the Archaeology Department at the University of Cape Town and they will be permanently curated by the McGregor Museum, Kimberley, Northern Cape, South Africa.”

3. Please include your full ethics statement in the ‘Methods’ section of your manuscript file. In your statement, please include the full name of the IRB or ethics committee who approved or waived your study, as well as whether or not you obtained informed written or verbal consent. If consent was waived for your study, please include this information in your statement as well.

Authors' response: Ethics approval is not relevant to this study as it did not involve human subjects

4. We note that Figures 3, 4, 5, 6, and 7 in your submission contain map images which may be copyrighted. All PLOS content is published under the Creative Commons Attribution License (CC BY 4.0), which means that the manuscript, images, and Supporting Information files will be freely available online, and any third party is permitted to access, download, copy, distribute, and use these materials in any way, even commercially, with proper attribution. For these reasons, we cannot publish previously copyrighted maps or satellite images created using proprietary data, such as Google software (Google Maps, Street View, and Earth).

Authors' response: The data used to create the maps is from 2 public domain sources USGS EROS (Earth Resources Observatory and Science (EROS) Center) (public domain): http://eros.usgs.gov/# and Natural Earth (public domain): http://www.naturalearthdata.com/. The data for ostrich distribution is adapted from the South African Bird Atlas Project 2 (SABAP2) acessed at (http: //sabap2.birdmap.africa/) which is copyright under CC BY 4.0. In the text we attribute the SABAP2 data, but we have added to the figure captions to clarify this.

5. Please review your reference list to ensure that it is complete and correct. If you have cited papers that have been retracted, please include the rationale for doing so in the manuscript text, or remove these references and replace them with relevant current references. Any changes to the reference list should be mentioned in the rebuttal letter that accompanies your revised manuscript. If you need to cite a retracted article, indicate the article’s retracted status in the References list and also include a citation and full reference for the retraction notice.

Authors' response: We have added the following references to the reference list, based on citations that were added to address reviewers comments. Miller and Wang (2022); Thomas and Shaw (2002).

Reviewer 1 (anonymous)

There are very few systematic studies on ostrich eggshell beads, so these findings are a welcome addition to current research. I therefore suggest publication with minor changes. However, I recommend addressing a few issues in order to improve the paper.

Authors' response: Thank you!

1. In the supplementary tables you mention the presence of OES in the DBSR stratigraphic aggregate. Considering the age of DBSR levels, the presence of one bead with possible ochre residues, three preforms and 44 ostrich eggshell fragments is very interesting. Why were they omitted from the analysis? Did you suspect their presence results from post-depositional mixing?

Authors' response: This paper is focused only on reporting the DBGS OES. A small number of preforms and a bead are associated with the DBSR deposit, however, we need to investigate this association further. At this point we are too uncertain about this association to report such an early age for OES beads. We're worried that the small number of tiny artefacts may have fallen from the walls during excavation - further excavations will help us resolve this. The DBSR beads and OES have now been excluded from the supp files and it is made clear now that only the DBGS is being reported here. The DBAS is a strag agg in the LSA deposits in the south shelter (GHS), which is still pending publication, and should have been excluded from the supp file.

2. My second observation concerns the stages of bead manufacturing and interpretation of the material. All manufacturing stages are present but you did not mention that stages V to VII are much more frequent. This could be an interesting observation and should be taken into account in the interpretation of the material. Additionally, the fact that fragments are outnumbered by worked OES (lines 371-372) does not necessarily mean OES was used in a conservative manner. Maybe finished beads were produced at the site and then taken elsewhere.

Authors' response: We have added to the Bead Manufacture section of the discussion to further discuss the higher frequency of stages V to VII and added to the sentence about OES being used in a conservative manner.

3. Thirdly, the different types of usewear that were identified and recorded are not defined, resulting in a lack of clarity. I would therefore advise the authors to clearly define each type of usewear in the methods section.

Authors' response: We have defined the usewear types in the methods section.

4. Finally, the presence of ochre residues is mentioned throughout the text. How did you characterize these residues as ochre? You should specify in the methods section if it is a visual characterization, and which criteria were taken into account (colour, texture, etc). In order to use the term “ochre”, you should conduct chemical analyses to confirm the presence of goethite or hematite (or other iron oxides). As there are no elemental or mineralogical analyses to support this, I would advise using more general terms such as “red / yellow / coloured residues”.

Authors' response: We have added a sentence specifying that this was a visual characterisation, and changed all mentions of ochre to red/ yellow coloured residue.

OES beads section

5. Line 87. Perhaps it would be interesting not to restrict this section to beads and mention other ostrich eggshell occurrences. For instance, engraved ostrich eggshell from Diepkloof, dating back to 60 ka (Texier et al 2010) should be mentioned somewhere.

Authors' response: We have added the Texier et al. 2010 reference in this section. However we have decided not to include more information on other oes fragment occurrences as they are generally only reported if they have been modified (decorated), while the oes fragments we report on are unmodified.

Context section: Ga-Mohana Hill North Rockshelter

6. Lines 154-169. Although you cite Wilkins et al 2020, figures including the location, stratigraphy and plan of the excavated areas of the site would be welcome here.

Authors' response: We have included a figure (figure 1) to show the location and stratigraphy of the site.

7. Line 170. In the supplementary tables there are pieces from the surface and the DBSR deposits. Do you omit those pieces in your study?

Authors' response: Yes, the only pieces included in the study are from DBGS stratigraphic aggregate. This is mentioned in the last paragraph of the Ga-Mohana Hill North Rockshelter section as a lead in for the Methods section. We have also clarified this in the Methods section.

Methods

8. Line 178. In the supplementary tables you added pieces from DBSR and surface layers. Here you should clearly explain why these pieces were omitted from the analysis (see main issues above).

Authors' response: Added this sentence to clarify: “The study was limited to analysis of OES assemblage from the DBGS stratigraphic aggregate layer in order to investigate and compare MIS 2 sites with OES assemblages across southern Africa.”

9. Line 183. Which criteria did you use to identify pigmentation? See comments above concerning the identification of ochre.

Authors' response: Added this sentence: “Pigmentation was classified visually based on colour and texture.“

10. Lines 184-185. See comments on the definitions of usewear above.

Authors' response: Have added more information about the usewear classification in the methods section.

11. Lines 186-189. It would be useful to the reader if you described each manufacturing stage rather than explaining them by groups (II-V and VI-VII)

Authors' response: We have included descriptions of each manufacturing stage.

12. Lines 207-208. Data on the horizontal distribution does not appear in the supplementary tables.

Authors' response: This data is already included in figure 3.

Results

13. Lines 230 and 234, and table 1. In supplementary table 2, there is one bead and one preform from DBAS. Do you mean DBGS? If it’s a typo, you should correct the numbers in the text and table 1 (20 beads and 10 preforms from DBGS).

Authors' response: The DBAS is a strag agg in the LSA deposits in the south shelter (GHS), which is still pending publication, and should have been excluded from the supp file.

14. Line 234. “This layer”, do you mean DGBS? Again, you should clearly indicate in the methods section that you omit the pieces from DBSR and surface layers from the analysis although they appear in the supplementary tables.

Authors' response: Changed “This layer” to DBGS, and have corrected the methods section to clarify the analysis only includes DBGS

15. Lines 234-235. See comments on the stages of bead manufacturing above.

Authors' response: Have added more detail on manufacturing stages above

16. Lines 241 and 260. See comments on the characterisation of ochre residues above.

Authors' response: changed “ochreous residue” to “red coloured residue

17. Lines 241 and 260. It would be interesting to show detailed photos of these residues as they are not clearly visible in figure 1.

Authors' response: This is unfortunately not possible at the moment as the beads are stored in Cape Town and none of the authors are in South Africa at the moment.

18. Line 244-246. How many preforms show these “pathway 1” features? Maybe a figure showing a couple of detailed photos of the “pathway 1” and “pathway 2” features could be useful here.

Authors' response: Unfortunately not possible as explained above

19. Line 255. “The shape of each fragment was recorded following Miller”: this should be moved to the methods section.

Authors' response: Moved to the methods section

20. Line 259. “Each fragment was examined for residue using a hand lens”. This was already said in the methods section.

Authors' response: Removed from the methods section

21. Lines 262-265. The definition of “scratches” and criteria used to identify non-anthropogenic marks should be moved to the methods section.

Authors' response: Moved to the methods section

22. Lines 266-267. “OES colour is a useful tool for understanding whether a shell might have been exposed to heat and the temperature of the heat source […] OES becomes yellow, red, iridescent and grey when heated under oxidising conditions [48,53,62], while reducing conditions are more likely to produce blackening of the shell. Blackening of OES has not been replicated in experimental studies but is relatively common in the archaeological record [48,50,63]”: Perhaps this section should also be moved to the methods section.

Authors' response: Moved to the methods section.

23. Lines 332-333. Maybe it would be better to specify “modern” ostrich here.

Authors' response: Have added modern to the sentence to clarify.

Discussion and conclusion

24. Line 370. Again, the fact that stages V to VII are much more frequent could be an important observation.

Authors' response: Have added this sentence to highlight the higher proportion of beads to preforms. “There is however a higher proportion of beads compared to preforms, which could indicate that drilling and grinding of beads were conducted in different areas of the rockshelter.”

25. Lines 373-374. “Assuming that ostrich distribution has not changed drastically over the last 15 ka”: is there any reference that supports this?

Authors' response: Unfortunately there is not much data on past ostrich distribution, we have changed this sentence to “Although data on changes in ostrich distribution in the late Quaternary are limited, if we assume that ostrich distribution has been relatively consistent across wet and dry phases over the last 15 ka (Thomas and Shaw 2002), then ostrich eggs and their shells would have been regularly encountered.”

26. Lines 399-403. See above comments on the use of the term “ochre”.

Authors' response: Have removed ochre for description of residues on OES and replaced with red residues.

Supplementary Tables

27. You could add a column with the horizontal provenance of the pieces. Do they all come from area A?

Authors' response: The horizontal provenience is shown in figure 2 (the main square is a view from above while each of the two side panels look north and east respectively). They all come from Area A – We have added the figure (figure 1) showing the rockshelter and excavated areas which might help clear this up.

28. 4) The different types of usewear shown in the tables should be clearly defined in the methods section.

Authors' response: Added this sentence to the methods section to define use-wear “Striations are defined as randomly oriented short striations, while facets are depressions on the surface of the bead.”

29. Why is the bead ID column empty?

Authors' response: We have removed this column – not applicable for this dataset.

30. Stratigraphic aggregates column: DBAS? This stratigraphic aggregate is not mentioned in the description of the stratigraphy, or in Wilkins et al 2020: I suppose it’s a typo? You probably mean DBGS.

Authors' response: The DBAS is a strag agg in the LSA deposits in the south shelter (GHS), which is still pending publication, and should have been excluded from the supp file.

31. Width column: (mm) instead of (g)

Authors' response: Fixed.

32. “Striae”, “smoothing”, “patina”, “chip”…: all use-wear types should be defined in the methods section. Perhaps locations should also be clearly explained.

Authors' response: We have included a sentence listing all of the attributes that were recorded along with a reference to the publication where more detailed explanations can be found. Including descriptions of each of these attributes would be very wordy and many of the attribute have not been studied further in this publication.

33. Staining column: what kind of staining is it when you indicate “yes” instead of “ochre”? You should be more specific here.

Authors' response: Have updated this to be more specific, yes was referring to brown staining on the OES.

34. Staining and comments columns: ochre was not identified using chemical analysis. You should either conduct elemental and mineralogical analysis to identify these residues as ochre or simply describe them without interpreting their composition (for example: red residues).

Authors' response: We have changed all references to ochre on the beads or fragments to red residue.

35. Correct the Orton stage numbers: some of them are not in capital letters.

Authors' response: corrected.

36. Observations column: see comments on the use of the term “ochre” above.

Authors' response: Have replaced ochre with red residue throughout.

37. What do you mean by “surface modification”, as opposed to “human modification” and “non-human modifications”? This should appear in the material and methods section.

Authors' response: Surface modifications are any markings on the surface that we could not confidently ascribe as either taphonomic or anthropogenic. We have edited the column headings in the supplementary table to better describe them. As well as adding a description of surface modification to the methods section.

Reviewer 2

Thank you for allowing me to review the paper “Ostrich eggshell beads from Ga-Mohana Hill North Rockshelter, southern Kalahari, and the 2 implications for understanding social networks during Marine Isotope Stage 2.” This is a fantastic paper. The methodology is thorough, and the writing is clear. It is amazing to see the way that researchers are getting at use and manufacture of beads to look at population interactions. I really do not have any critiques. If the authors wish, they could tie the paper to the recent Nature article by Miller and Wang (2021). I recommend accepting the paper.

Authors' response: Thank you! We have added the Miller and Wang reference to the introduction and discussion.

Submitted filename: Response to reviewers.pdf

Click here for additional data file.

12 May 2022

Ostrich eggshell beads from Ga-Mohana Hill North Rockshelter, southern Kalahari, and the implications for understanding social networks during Marine Isotope Stage 2.

PONE-D-21-22127R1

Dear Dr. Hatton,

We’re pleased to inform you that your manuscript has been judged scientifically suitable for publication and will be formally accepted for publication once it meets all outstanding technical requirements.

Within one week, you’ll receive an e-mail detailing the required amendments. When these have been addressed, you’ll receive a formal acceptance letter and your manuscript will be scheduled for publication.

An invoice for payment will follow shortly after the formal acceptance. To ensure an efficient process, please log into Editorial Manager at http://www.editorialmanager.com/pone/, click the 'Update My Information' link at the top of the page, and double check that your user information is up-to-date. If you have any billing related questions, please contact our Author Billing department directly at authorbilling@plos.org.

If your institution or institutions have a press office, please notify them about your upcoming paper to help maximize its impact. If they’ll be preparing press materials, please inform our press team as soon as possible -- no later than 48 hours after receiving the formal acceptance. Your manuscript will remain under strict press embargo until 2 pm Eastern Time on the date of publication. For more information, please contact onepress@plos.org.

Kind regards,

Enza Elena Spinapolice, Ph.D

Academic Editor

PLOS ONE

Additional Editor Comments (optional):

Reviewers' comments:

Reviewer's Responses to Questions

Comments to the Author

1. If the authors have adequately addressed your comments raised in a previous round of review and you feel that this manuscript is now acceptable for publication, you may indicate that here to bypass the “Comments to the Author” section, enter your conflict of interest statement in the “Confidential to Editor” section, and submit your "Accept" recommendation.

Reviewer #2: All comments have been addressed

**********

2. Is the manuscript technically sound, and do the data support the conclusions?

The manuscript must describe a technically sound piece of scientific research with data that supports the conclusions. Experiments must have been conducted rigorously, with appropriate controls, replication, and sample sizes. The conclusions must be drawn appropriately based on the data presented.

Reviewer #2: Yes

**********

3. Has the statistical analysis been performed appropriately and rigorously?

Reviewer #2: Yes

**********

4. Have the authors made all data underlying the findings in their manuscript fully available?

The PLOS Data policy requires authors to make all data underlying the findings described in their manuscript fully available without restriction, with rare exception (please refer to the Data Availability Statement in the manuscript PDF file). The data should be provided as part of the manuscript or its supporting information, or deposited to a public repository. For example, in addition to summary statistics, the data points behind means, medians and variance measures should be available. If there are restrictions on publicly sharing data—e.g. participant privacy or use of data from a third party—those must be specified.

Reviewer #2: Yes

**********

5. Is the manuscript presented in an intelligible fashion and written in standard English?

PLOS ONE does not copyedit accepted manuscripts, so the language in submitted articles must be clear, correct, and unambiguous. Any typographical or grammatical errors should be corrected at revision, so please note any specific errors here.

Reviewer #2: Yes

**********

6. Review Comments to the Author

Please use the space provided to explain your answers to the questions above. You may also include additional comments for the author, including concerns about dual publication, research ethics, or publication ethics. (Please upload your review as an attachment if it exceeds 20,000 characters)

Reviewer #2: I reviewed this manuscript very positively the first time, and my opinion has not changed. This is a scientifically sound article, that adds insight to the field.

**********

7. PLOS authors have the option to publish the peer review history of their article (what does this mean?). If published, this will include your full peer review and any attached files.

If you choose “no”, your identity will remain anonymous but your review may still be made public.

Do you want your identity to be public for this peer review? For information about this choice, including consent withdrawal, please see our Privacy Policy.

Reviewer #2: Yes: Jamie Hodgkins

23 May 2022

PONE-D-21-22127R1

Ostrich eggshell beads from Ga-Mohana Hill North Rockshelter, southern Kalahari, and the implications for understanding social networks during Marine Isotope Stage 2

Dear Dr. Hatton:

I'm pleased to inform you that your manuscript has been deemed suitable for publication in PLOS ONE. Congratulations! Your manuscript is now with our production department.

If your institution or institutions have a press office, please let them know about your upcoming paper now to help maximize its impact. If they'll be preparing press materials, please inform our press team within the next 48 hours. Your manuscript will remain under strict press embargo until 2 pm Eastern Time on the date of publication. For more information please contact onepress@plos.org.

If we can help with anything else, please email us at plosone@plos.org.

Thank you for submitting your work to PLOS ONE and supporting open access.

Kind regards,

PLOS ONE Editorial Office Staff

on behalf of

Dr. Enza Elena Spinapolice

Academic Editor

PLOS ONE