Tufas indicate prolonged periods of water availability linked to human occupation in the southern Kalahari.

Detailed, well-dated palaeoclimate and archaeological records are critical for understanding the impact of environmental change on human evolution. Ga-Mohana Hill, in the southern Kalahari, South Africa, preserves a Pleistocene archaeological sequence. Relict tufas at the site are evidence of past flowing streams, waterfalls, and shallow pools. Here, we use laser ablation screening to target material suitable for uranium-thorium dating. We obtained 33 ages covering the last 110 thousand years (ka) and identify five tufa formation episodes at 114-100 ka, 73-48 ka, 44-32 ka, 15-6 ka, and ~3 ka. Three tufa episodes are coincident with the archaeological units at Ga-Mohana Hill dating to ~105 ka, ~31 ka, and ~15 ka. Based on our data and the coincidence of dated layers from other local records, we argue that in the southern Kalahari, from ~240 ka to ~71 ka wet phases and human occupation are coupled, but by ~20 ka during the Last Glacial Maximum (LGM), they are decoupled.

Introduction

A key question in human origins research is how climate change impacted early Homo sapiens population distributions across Africa. It has been hypothesized that humans did not always have the capacity to survive in arid environments [1, 2], that early human distributions were modulated by distance to [3] and availability of water [4], that people were largely restricted to wetter refugia during glacial periods [5-7], and that the occupation of arid regions was coincident with interglacial periods [8, 9]. We see this view manifested slightly differently in different regions of the African continent. For example, in eastern Africa, favourable climatic conditions and increasing water availability during MIS 5 is associated with greater mobility of Homo sapiens and their occupation of a wide variety of sites, not restricted to oases such as lake margins [7]. In North Africa, the idea of a ‘Green Sahara’ during ~130–75 ka is contested by Scerri et al. [10] who use a palaeoenvironmental model to highlight that the region was not a uniform oasis, noting that pockets of aridity persisted as barriers to mobility, even during periods of increased humidity, and that human dispersal at this time occurred along corridors connected by palaeohydrological networks. While evidence for human mobility and dispersal is generally linked to water availability [10, 11], there are some exceptions, for example, occupation at the North African site of Uan Tabu between ~60–90 ka is associated with evidence of an arid climate [12]. With recent advances in the application of dating techniques e.g. U-Th and OSL methods, to archaeological and palaeoclimate proxy deposits, improved chronological reliability and resolution allows for questions relating to human-environment interactions to be refined [13, 14].

The Kalahari Basin, in the interior of southern Africa, is a semi-arid region that has experienced significant climatic fluctuations with abundant records of both palaeoenvironment and archaeology. As such, this region provides a unique opportunity to further explore early human-environment interactions [15]. For example, in the southern Kalahari, multiple lines of evidence point to very different, much wetter periods during much of the Pleistocene [16]. Paleoenvironmental records from macrobotanical and faunal remains, pollen, and stable isotope compositions of mammalian tooth enamel, ostrich eggshells and speleothem deposits, demonstrate climatic shifts through the Pleistocene and Holocene at Wonderwerk Cave [17-24]. At Kathu Pan, several wet periods between ~160–22 ka have been identified based on sedimentary analyses [25, 26], and evidence for wetter conditions at Ga-Mohana Hill ~110–105 ka has also been reported [27]. Previous studies reveal significant complexities even at the intra-regional scale, however, due in part to the different types of proxies with variable resolutions and the variety of forcing factors at play [16]. They also reveal a complex relationship between palaeoenvironmental conditions and evidence for human occupation [15, 28]. To more fully assess the response of Homo sapiens to changes in climate and environments, more well-dated records of past environments from different proxies, preferably closely associated with archaeological records of human behaviour, are required. Here, we report one such record from the abundant carbonate deposits at Ga-Mohana Hill, identified as tufa i.e., ambient temperature, freshwater calcium carbonate precipitates, that span the last 110 thousand years.

Ga-Mohana Hill is a double-humped hillside situated on the eastern flank of the north-south trending Kuruman Hills which outcrop on the Ghaap Plateau, an elevated region in the Northern Cape province of South Africa (Fig 1). Today the area is characterised as semi-arid, with seasonal mean annual precipitation of ~300–400 mm during the austral summer months [24]. The bedrock lithology of the Ghaap Plateau, which is comprised of the Palaeoproterozoic dolomites of the Campbellrand-Malmani Subgroup [29], has undergone extensive karstification. Coupled with cross-cutting dolerite dykes, this has resulted in large groundwater compartments that are important aquifers for the region [30, 31]. Groundwater resurgence at active springs in the area, such as the Eye of Kuruman, in Kuruman (Fig 1) are a testament to this vast underground drainage network [31]. The presence and movement of these groundwaters through the dolomite host rock is a vital precursor to the formation of the tufas at Ga-Mohana Hill.

Fig 1

Map of South Africa with the location of Ga-Mohana Hill (GHN) and key palaeoenvironmental and middle stone age sites discussed in the text.

Dashed lines demarcate summer and winter rainfall zone boundaries (SRZ, WRZ), middle area experiences year-round rainfall (YRZ). Inset map shows the approximate extent of the Kalahari Basin in southern Africa and the location of the region of interest in relation to it. Figure produced in ArcGIS 10 from multiple open source datasets: Kalahari Basin extent from SASSCAL Open Access Data Centre [32]; Digital Elevation Model obtained from USGS Earth Explorer [33]; annual rainfall data from WorldClim 2 [34] and river centrelines accessed from Natural Earth vector data [35].

Recent archaeological excavations at Ga-Mohana Hill North Rockshelter have yielded a Middle Stone Age assemblage of artefacts that provides early evidence for innovation and behavioural complexity in this region at 105±3.3 ka [27, 36]. Stratified above are younger deposits dated by OSL to 31±1.8 ka and 15±0.8 ka [36]. The hillside has abundant tufa deposits which are direct evidence for the presence of water on the landscape and are amenable to radiometric dating methods, making them valuable archives of changes in environmental conditions [37-42].

In this study, we present macro- and micromorphological analyses of the Ga-Mohana tufas to assess their depositional context. Tufas are challenging materials for dating due to detrital contamination and generally low uranium concentrations [43, 44] and so samples were screened using laser ablation inductively coupled plasma mass spectrometry (LA-ICP-MS) to target optimal zones for study. This method has been used previously for dating speleothems [45], but to the best of our knowledge, this is the first tufa application. We present 33 U-Th age estimates and identify five discrete periods of tufa formation, interpreted as evidence for increased moisture and fresh water availability on the landscape during the late Pleistocene, thus providing a new record of localized climate change linked to a dated record of human occupation.

Materials and methods

Fieldwork and tufa sample collection

Ga-Mohana Hill has spiritual significance for the local communities, with visits to the shelter deliberate and rare [46]. Out of respect for this and as part of our on-going engagement with these communities, we adopted a low-impact sampling approach, with targeted samples carefully chosen after extensive survey of the 6 km area around the shelter. During this pedestrian survey, the field occurrences, positions and types of tufa were identified and mapped using a roaming Geographic Positioning System. A total of twenty-nine tufa hand samples were collected from the ~ 1 km2 Ga-Mohana hillside sampling all five tufa morphologies. Eighteen hand samples were collected using a geological hammer, mallet and chisel, marking the way-up on each sample with an arrow using permanent marker. Subsequent sampling deliberately targeted the stratigraphically older layers, closest to the host rock dolomite, to try and constrain the onset of preserved tufa formation. We used a modified Makita cordless hand drill fitted with Pomeroy Model SW-3 Miniature Water Swivel and a custom made Pomeroy 1.5” ID diamond-tipped core barrel. A total of eleven small cores were collected, 8 cm in length on average, from in-situ mound tufas, and both in-situ and ex-situ cascade tufas (S1 Table, S1–S5 Figs). The cores were set in epoxy resin and then halved lengthways with a diamond rock saw and polished. Thin sections were made from a sub-set of fourteen samples, representative of all the morphology types, for characterisation using a Zeiss AXIO polarising light microscope (S1 Table, S1–S5 Figs).

Permits for archaeological investigations at Ga-Mohana Hill were obtained from the South African Heritage Resource Agency (Permit ID 2194). The land is owned by the Baga Motlhware Traditional Council and consent was granted by them to conduct the study. All necessary permits were obtained for the described study, which complied with all relevant regulations. Additional information regarding the ethical, cultural, and scientific considerations specific to inclusivity in global research is included in the Supporting Information (S1 Checklist).

Laser ablation-inductively coupled plasma-mass spectrometer (LA-ICP-MS) pre-screening of U and Th concentrations and distributions

The aphanitic micrite layers free from detritus and inclusions, identified in thin section, were primary targets for U-Th dating. However, these layers tend to be fine, undulating and laterally variable, and so while visual evaluation of the tufas is an important first step in identifying suitable material to target for U-Th dating, it is not sufficient considering the complexity of the tufas on a microscale. We employed an additional screening step, using laser ablation inductively coupled plasma mass spectrometry (LA-ICP-MS), to measure and image the U and Th concentrations and distributions along transects within the tufa samples. Thus, layers with sufficiently high levels of 238U and low levels of 232Th, i.e. detrital thorium, can be selected, as these are the best for producing reliable age data [47].

Tufa U and Th concentrations and distributions were collected for 16 samples using laser-ablation with an Applied Spectra RESOlution SE 193nm ArFexcimer laser-ablation system coupled to an Agilent 7700x Quadrupole ICP-MS at the University of Melbourne, following the protocols outlined in Woodhead et al. [48]. High-resolution images (3200 dpi) of the samples were captured using a flat-bed scanner, used to reference the co-ordinate system of the laser cell using GeoStar software (Norris Software). Between 6 and 12 parallel lines per sample, set 62μm apart, were chosen perpendicular to the growth layers. Pre-ablation was performed twice using a 60μm spot size and stage translation speed of 150μm/s.

Trace element data for the following elements: Mg, Al, Mn, Fe, Zn, Sr, Ba, Pb, Th and U, were collected with a 60μm spot at a stage translation speed of 75μm/s, pulse rate of 10Hz, and laser fluence of ~2–3 Jcm-2. NIST SRM 612 was used for calibration, with 43Ca as an internal standard, and an estimated precision of ca <5%. NIST SRM 610 and JCp-1, a powdered coral standard, were also analysed. The raw mass spectrometry data was reduced using the Iolite software package [49, 50]. Element distribution maps for 238U and 232Th were generated in order to visualise the spatial arrangement of these trace elements through the samples [51] (S6–S8 Figs).

U-Th dating of tufa

Guided by the laser ablation results, layers with sufficiently high uranium (238U > 0.1ppm) and low thorium concentrations (232Th < 0.01ppm) were selected for U-Th analysis. A subset of 43 samples (S2 and S3 Tables, S6–S8 Figs), each with a mass of 60 ± 10 mg, were drilled from 16 tufa samples using a Dremel hand-held hobby drill and 1 mm carbide micro-drill bit. Powdered samples were dissolved in 1.5M HNO3, spiked with a mixed 236U-233U-229Th tracer equilibrated on a hotplate overnight. U and Th were separated from the calcite matrix using Eichrom TRU-spec selective ion exchange resin following established protocols [52]. The U-Th solution was dissolved in a mixture of dilute nitric and hydrofluoric acid and introduced to the Nu Instruments Plasma Multi Collector-Inductively Coupled Plasma-Mass Spectrometer via an autosampler [52, 53]. Isotope-ratio measurements for 230Th/238U and 234U/238U were calculated using an internally standardised parallel ion-counter procedure and calibrated against the secular equilibrium standard, HU-1. Reproducibility was monitored using a second in-house standard (YB-1). An a priori estimate of 1.5 ± 1.5 for the initial 230Th/232Th was applied to all the samples in order to correct for the inherent detrital component [47]. With this initial value and its uncertainty, corrected ages for all samples were calculated using Monte Carlo iterations to solve equation 1 of Hellstrom [47] and the half-life values of 234U and 230Th as reported in Cheng et al. [54]. The final age uncertainty is reported in 2σ for which the uncertainties of the measured activity ratios as well as the assumed initial 230Th/232Th are fully propagated.

Results

Tufa macro and micromorphology

The Ga-Mohana tufa system comprises five morphological components: cascades, rim pools, barrages, domes and terrace breccias (Fig 2) drawing from established classification schemes [55-58]. Tufa cascades are observed across the ~1 km2 hillside. Large (3-5m) cascades cover the tall cliffs on either side of the rock shelters. Smaller cascades bulge outward from the fronts of the dolomite steps above and below the shelters (Fig 2, S1 Fig). These cascade tufas are point-sourced, appearing to have formed from water flowing out of the dolomite bedding planes. Below the step-front cascades, sinuous tufa rims edge the flat, transverse sections of the dolomite steps. These are evidence of terraced, shallow pools, likely formed from excess water ponding below the cascades. The areas behind the rims are filled with lightly compacted sediment and debris. Curved, down-hill sloping barrage tufas, characterised by knobbly, coralloid surfaces, sit below the rim pool edges formed when water overflowed from the pools above. Meandering channels scoured in the dolomite, observed above the rock shelter, are evidence of palaeostreams and indicate periods of substantial and prolonged water flow.

Fig 2

Tufa depositional environment context and representative photographs of each of the tufa morphologies identified on the Ga-Mohana hillside.

(A) Schematic profile sketch of Ga-Mohana Hill North Rockshelter (not drawn to scale) illustrating the series of tufa deposits and the archaeological excavation. The excavation layers dated via OSL: DBSR = ~105 ka; OAS = ~31 ka and DBGS = ~15 ka; (B) cliff cascade; (C) step-front cascade; (D) sinuous rim-pool edge; (E) barrage tufa; (F) terrace breccia; (G) tufa dome.

The rockshelters and the tall cliffs adjacent to them mark a break in the hillside. At this point, cone-shaped tufa ‘noses’ jut out over the lip of the rock shelters and are spread along the overhang (Fig 2). These are interpreted as remnants of moss curtains and align with large hemispheric dome tufas below. The domes trace the dripline of the overhanging shelters and occur along the base of the cliffs adjacent to the rock shelters; they appear to be sourced from dripping and splashing waters channelled via the noses above and are composed of stacked phytoherm macro-layers reminiscent of bryophyte boundstones [55, 56]. The internal rockshelter walls are covered with clusters of small stalactitic features and calcitic crusts. Below the shelter, surface-cementations of sub-angular detrital clasts of variable sizes (0.5–20 cm) of banded ironstone and dolomite fragments occur as benches, pavements, or patches of carbonate-cemented hill-slope material on the sub-horizontal terraces between dolomite steps, similar to the surface-cemented rudites described in Pentecost and Viles [58]. In rare instances, stone artefacts are also included.

Microscale observations of the tufas reveal that, regardless of depositional setting, the tufas are composed of a few simple petrographic components: micrite, microspar, and sparite (S1–S5 Figs). Detrital clasts (quartz) and iron and manganese oxides are present to varying degrees and tend to be concentrated along thin layers in the tufas. This suggests periods of non-deposition of tufa. The variable organisation of these petrographic components within each sample results in distinct fabrics, classed as laminar, peloidal, aphanitic, and chaotic, following the scheme devised by Manzo et al. [59]. A significant biological component is evidenced by stromatolitic structures (S1 Fig), clotted micrite (S3 Fig), and primary cavities (S5 Fig).

The various tufa morphologies at Ga-Mohana Hill each represents an individual sub-environment, and together they form a continuum of linked deposits that aligns with the perched springline model of Pedley [57, 60]. This depositional environment was characterised by water emerging from bedding planes in the dolomite, flowing down the hillside via multiple divergent pathways, creating cascades on the step-fronts of the dolomite steps, generating waterfalls and moss curtains over the rock shelters, and feeding shallow pools on the flat terraces. The terrace breccia deposits hint at periods of high energy flow (e.g., flash flooding) to transport and cement substantial talus scree downslope.

U-Th chronology

The 238U concentration in the tufas is consistently low (range = 0.1 to 0.6 ppm; mean = 0.2 ppm). The 232Th concentrations were generally lower than the 238U concentrations, with most samples reflecting a wide range in 232Th concentration, between 1–100 ppb. In many instances, elevated 232Th corresponds with visually discernible detrital material (S8 Fig). Out of 43 sub-samples drilled from 18 tufa samples, we obtained 33 U-Th ages from 12 tufas (Table 1, S2 Table). Cascade, rim pool and terrace tufas exhibited high success rates; 86% of cascade samples (18 of 21), 100% of rim pools (7 of 7) and 100% of terrace breccias sampled (7 of 7) yielded resolvable U-Th age estimates, while only one of four dome samples returned a reliable age. Reliable ages tended to be unresolvable on samples with very low 230Th /232Th ratios (e.g. 230Th /232Th <7) indicating a significant detrital component (S3 Table). It was not possible to resolve reliable or precise ages for any of the barrage samples, three dome and two cascade samples, however some of the corrected ages for these samples may provide a useful upper limit age estimate, i.e. the corrected age plus the associated 2σ uncertainty (S3 Table).

Table 1

U-Th age data for tufa samples from Ga-Mohana Hill.

The samples are labelled according to the sequence they were collected in but presented in stratigraphic order. Errors on all isotope activity ratios are reported with 2σ uncertainty. All ages have been corrected to account for the effect of detrital Th assuming an estimate for initial 230Th/232Th of 1.5 ± 1.5, and calculated using the 230Th-238U decay constants of Cheng et al. [54] and equation 1 from Hellstrom [47].

Sample ID	Tufa type	238U (ng/g)	230Th/238U	2σ	234U/238U	2σ	230Th/232Th	U-Th age (ka)	2σ	% error
18–10.2	dome	174	0.088	0.001	2.443	0.006	6.5	3.0	0.9	30
GHN-2	cascade	75	0.184	0.002	2.728	0.009	36.2	7.266	0.315	4.3
GHS-5	cascade	263	0.251	0.002	1.866	0.007	4.6	10.738	4.932	45.9
17–8.1	terrace	363	0.707	0.002	1.895	0.003	88.4	48.306	0.684	1.4
17–8.2	terrace	273	0.528	0.002	1.894	0.003	238.3	34.503	0.231	0.7
17–8.3	terrace	321	0.621	0.002	1.896	0.005	3367.7	41.760	0.220	0.5
17–8.4	terrace	298	0.641	0.003	1.903	0.005	4516.0	43.230	0.270	0.6
17–8.5	terrace	307	0.618	0.003	1.895	0.005	4858.7	41.620	0.260	0.6
17–8.6	terrace	459	0.503	0.004	1.899	0.006	216.6	32.450	0.370	1.1
GHN-1	rim pool	236	0.576	0.005	1.863	0.006	716.3	39.126	0.398	1.0
GHN-1.2	rim pool	234	0.543	0.005	1.859	0.006	6818.5	36.550	0.390	1.1
GHN-1.3	rim pool	240	0.551	0.005	1.868	0.007	10846.4	37.020	0.420	1.1
GHS-6	rim pool	212	0.864	0.007	1.914	0.007	41.3	60.379	1.809	3.0
GHS-6.1	rim pool	180	0.852	0.005	1.913	0.007	50.3	59.677	1.461	2.4
GHS-6.2	rim pool	190	0.873	0.003	1.928	0.0054	251.3	61.986	0.466	0.8
GHS-6.3	rim pool	158	0.798	0.003	1.882	0.004	15.9	53.100	4.200	7.9
18–7	terrace	847	0.749	0.003	1.833	0.005	50.5	53.520	1.310	2.4
18–13.1	cascade	249	1.174	0.003	2.654	0.007	68.3	58.610	0.990	1.7
18–13.2	cascade	226	1.313	0.003	2.742	0.007	93.6	65.040	0.810	1.2
18–13.3	cascade	132	1.287	0.004	2.646	0.007	620.5	67.150	0.380	0.6
18–13.4	cascade	195	1.483	0.004	2.933	0.008	124.6	69.830	0.680	1.0
18–14.1	cascade	83	1.308	0.009	2.644	0.009	243.4	68.430	0.730	1.1
18–14.2	cascade	139	1.219	0.006	2.551	0.008	46.1	64.280	1.600	2.5
18–14.3	cascade	98	1.351	0.008	2.705	0.009	439.1	69.350	0.670	1.0
18–14.4	cascade	180	1.481	0.006	2.876	0.008	47.0	70.600	1.670	2.4
18–15.1	cascade	137	1.319	0.007	2.668	0.008	781.0	68.520	0.570	0.8
18–15.2	cascade	95	1.317	0.011	2.587	0.010	746.9	71.340	0.890	1.2
18–15.3	cascade	313	1.522	0.005	2.940	0.008	229.7	72.280	0.530	0.7
18–17.1	cascade	154	2.176	0.006	3.194	0.006	29.0	102.900	3.200	3.1
18–17.2	cascade	148	2.085	0.007	3.102	0.006	44.2	102.100	2.100	2.1
18–17.3	cascade	142	2.217	0.006	3.289	0.006	95.7	103.310	1.080	1.0
18–16.1	cascade	164	2.586	0.008	3.614	0.007	32.9	110.600	3.000	2.7
18–16.2	cascade	177	2.404	0.007	3.476	0.007	43.9	105.900	2.200	2.1

The tufa ages span the last interglacial cycle, from 110.6 ± 3.0 ka through to 3.0 ± 0.9 ka (Table 1, Fig 3). Clusters of ages, defined by distinct groups of overlapping ages and their associated 2σ uncertainties, suggests episodic growth over this time. At least five intervals of tufa formation at Ga-Mohana Hill identified at approximately 114–100 ka, 73–48 ka, 44–32 ka, 15–6 ka, and ~3 ka (Fig 3). The 2σ uncertainties associated with the ages are small; most samples are associated with errors of <3 ka (on average approximately 1 ka) except for two samples, GHS-5 and GHS-6.3, which have an uncertainty of 4.9 ka (49%) and 4.2 ka (7.9%) respectively. These larger errors are due to a high detrital thorium component (Table 1).

Fig 3

Composite plot of Ga-Mohana Hill tufa formation compared to selected global proxies over the last 120 ka.

(A) LR04 curve [61]; (B) variance of reconstructed sea surface temperatures (SST) from Indian Ocean core MD96-2048 [62]; (C) mean daily summer insolation curve for 27oS [63]; (D) OSL age data from the Ga-Mohana Hill North excavation sediments [27, 36]; (E) tufa U-Th age data with 2σ error bars presented in Table 1. The blue bars highlight clusters of overlapping tufa ages and are defined by the minimum and maximum ages in each range, calculated using the 2σ uncertainty associated with the ages. Based on the presence of the tufa deposits, these periods are inferred to represent episodes of greater water availability on the landscape.

The ages for the timing of human occupation at Ga-Mohana Hill coincides with three of the tufa forming intervals during MIS 5d, late MIS 3, and late MIS 2, indicating contemporaneous human activity and tufa formation at Ga-Mohana during those periods (Fig 3). The age certainty for the interval of tufa formation that overlaps with the MIS 2 occupation at Ga-Mohana Hill is less secure than the other intervals as it has a large error associated with it, but the human occupation falls within the 2σ uncertainty of the tufa age.

Comparison to global records

Tufa deposits occur in a variety of settings around the world, and their formation is controlled by a range of local, regional and global factors operating on different scales [38, 60, 64–66]. Here we compare the timing of tufa formation at Ga-Mohana Hill to records of global ice volume [61], austral summer insolation for 270 south [63], and changes in sea surface temperature in the southwest Indian Ocean [62] to assess the extent to which the tufa deposits reflect global-scale climate changes (Fig 3). There is no clear glacial/interglacial partitioning of tufa formation episodes, as evidenced by comparing our data to the LR04 d18O benthic stack [61] (Fig 3). This adds to growing evidence that the wet/dry, interglacial/glacial dichotomy through which much of southern African palaeoclimates has traditionally been viewed is overly simplistic [67-71]. While tufas have typically been associated with warmer and more humid interglacial climate conditions [72-75], several studies report tufa occurrences during both glacial and interglacial periods [41, 42, 66, 76, 77]. This highlights that, across the globe, regional climates respond variably to these boundary conditions, and cautions against simplistic interpretations of tufa deposits. Our record suggests tufa formation was semi-continuous across MIS 4 and MIS 3; we thus echo the conclusion of previous studies that tufa formation is not restricted to interglacial periods, nor is it a simple product of changing global climate states.

The principal conditions required for tufa formation are sufficient effective precipitation to recharge the aquifers and CaCO3 supersaturation of those waters [38, 40, 75]. Productive soil and vegetation cover is necessary to enhance the pCO2 of the percolating waters, and moderate temperatures which balance productivity, moisture and evaporation, are important secondary requirements [38, 42]. Tufa formation is thus sensitive to multiple environmental parameters, but ultimately provides direct evidence of fresh water and associated productivity on the landscape. Our record indicates that these conditions were met during five discrete time intervals (114–100 ka, 73–48 ka, 44–32 ka, 15–6 ka, and 3 ka) in the southern Kalahari over the last ~110 ka.

The limiting factor for tufa formation in semi-arid, low latitude regions is water availability [40, 78]. The moisture source for rainfall in the SRZ in South Africa originates primarily from the south western Indian Ocean [79] but the spatial and temporal variability of rainfall in this southern Kalahari region is poorly constrained. Rainfall may be modulated by summer insolation, with increased precipitation corresponding to insolation maxima [80], however, we see no simple correlation. Based on the mean summer insolation curve for 27°S [63] (Fig 3), tufa formation during the 114–100 ka and 44–32 ka intervals coincide with increasing summer insolation, while tufa formation during 73–48 ka is variable, and at a minimum during the 15–6 ka episode. The most recent tufa formation at ~3 ka does coincide with insolation maximum. Kele et al. [77] find a similarly poor correlation between the timing of tufa formation in the Kurkur-Dungal area (southern Egypt) and northern hemisphere insolation, which is thought to modulate the position of the ITCZ, i.e., high summer insolation is expected to correspond to increased precipitation. However, the southern Egypt tufa deposits do not align with high summer insolation, even when a delay in aquifer recharge is taken into account. They propose that the mechanisms driving rainfall in the region are complex and cannot be attributed to a single forcing factor, e.g. the precession-controlled motion of the ITCZ, and instead conclude that different forcing mechanisms are likely at play during interglacial vs. glacial periods. A potential explanation for the lack of correlation between our Ga-Mohana tufa record and high austral summer insolation is that direct insolation forcing has played a lesser role over the last ~50 ka due to lower amplitude changes related to declining eccentricity [81], and that after ~70 ka, high latitude changes may have had a greater influence on southern African hydroclimate [82]. This could mean that the first and last intervals of tufa formation (~114–100 ka and ~3 ka) may have been driven by low latitude mechanisms, i.e. high austral summer insolation, whereas favourable conditions for tufa formation across MIS 4 and MIS 3, which occurred across variable insolation, may have been a response to high latitude mechanisms, i.e. an increase in the global ice volume.

Following that warmer sea surface temperatures (SST) in the southwest Indian Ocean generate increased moisture and correlate to periods of greater rainfall in southeastern Africa today [79, 83, 84], one might predict that past periods of warmer SST would correspond to periods of tufa formation at Ga-Mohana. However, tufa formation occurs across a range of Indian Ocean SST [62] (Fig 3) suggesting that SST is not the driving mechanism for increased rainfall in this region. While warmer SSTs, coupled with a negative Southern Oscillation Index, is used to explain higher rainfall during the 114–100 ka interval [27], Caley et al. [62] argue that land-sea temperature gradients, rather than SSTs alone, are likely to have played an important role in modulating rainfall variability in southeastern Africa. However, with an ever-growing body of palaeoproxy data generated for sites across southern Africa, it is increasingly acknowledged that broad theoretical models evoking a primary driving mechanism do not capture the homogeneity of rainfall across the region which has experienced a spatially complex pattern of past hydroclimate variability [81, 85]. Through HadCM3 climate model simulations, Singarayer et al. [81] show that correlations between SSTs and rainfall varies spatially and temporally over southern Africa depending on the interplay of different mechanisms. This illustrates that rainfall across the region is driven by the interaction of a range of drivers that wax and wane in prominence, and suggests that the governing mechanism of rainfall variability for any particular region has likely varied over time [81].

It seems clear that tufa formation cannot be simply related to changes in global climate, as these deposits occur in a variety of settings and form in response to variable climatic conditions. However, tufas remain a valuable tool for assessing rainfall regime shifts, karst recharge processes, and changes in environmental parameters, as documented for several deposits world-wide [40, 55, 76, 78, 86–88] and may be complimentary archives to other proxies such as lake sediments and speleothem deposits [39, 60, 89].

The Ga-Mohana tufas form from emergent groundwaters. Recharge of the below surface aquifers, which feed these underground waters, is driven by either increased rainfall, less seasonal / more prolonged rainfall, or reduced effects of evaporation, or most likely a combination of all three. Little is known about the karst dynamics at this site, and so potential lags between rainfall events, aquifer recharge, residence times and tufa formation are as yet unconstrained. However, Markowska et al. [90] show that cooler temperatures and reduced rates of evaporation create favourable conditions for karst recharge in semi-arid regions. As such, tufa deposits at Ga-Mohana Hill are not simply indicators of ‘wet’ conditions but rather indicate a positive moisture balance caused by an interplay of local and regional climatic and environmental parameters. The time intervals presented here are thus interpreted as periods of prolonged water availability combined with milder temperatures and reduced rates of evaporation. These conditions were met during discrete time periods over the last ~110 ka, the timing of which arise from the interaction of multiple forcing factors.

Comparison to regional records

We compare the record of tufa formation intervals at Ga-Mohana Hill with other palaeoenvironmental records at nearby Kathu Pan and Wonderwerk Cave (Fig 4). These three sites all occur within ~60 km of each other and thus experienced comparable shifts in local hydroclimate. The tufa record at Ga-Mohana Hill indicates wet conditions during MIS 5d and MIS 4. Sediment analysis at Kathu Pan is consistent with our record; marshy conditions prevailed at Kathu Pan from ~101–80 ka, and palygorskite-coated sands indicate the presence of fluctuating water levels across five intervals between ~167–52 ka [25]. This confirms that the region was wetter during much of MIS 5 and 4. A gap in tufa formation at Ga-Mohana Hill after ~31 ka indicates less water availability during much of MIS 2. This is reflected in the development of extensive pedogenic carbonate deposits at Kathu Pan by ~23 ka, which indicate drier conditions and perhaps more seasonal rainfall compared to earlier time periods [25]. At Wonderwerk Cave, a hiatus in stalagmite growth after ~33 ka [19] is consistent with drier conditions, however, wetter conditions from ~23 to 17 ka are reflected in the pollen and stable isotope record from the same stalagmite; this evidence for wetter conditions at Wonderwerk Cave during the LGM is inconsistent with the records from Ga-Mohana and Kathu Pan. Ga-Mohana Hill documents a subsequent late glacial wet period commencing as early as ~15 ka, while at Wonderwerk Cave, slow growth of the stalagmite between ~17–13 ka signifies reduced moisture availability [19], and pedogenic carbonates at Kathu Pan, indicating dry conditions, persist at ~10 ka [25]. The most recent period of tufa formation at Ga-Mohana is at ~3 ka. Through the Holocene, fine layers of organic material alternate with calcium carbonate deposits at Kathu Pan, implying an increased amplitude of fluctuating water availability and aridity [25], and at Wonderwerk Cave, stalagmite growth resumes at ~3.5 ka, indicating wet conditions during the late Holocene [18, 19]. The earlier and latest parts of the sequences from these sites are consistent, but the record from Wonderwerk Cave indicating wetter conditions during MIS 2 is inconsistent, and some discrepancy between the records exist for the MIS 2—Holocene transition.

Fig 4

Comparison of palaeoenvironmental and archaeological records from Ga-Mohana Hill, Kathu Pan, and Wonderwerk Cave.

Archaeological occupation ages for Ga-Mohana Hill [27, 36], Kathu Pan [25, 26, 91, 92] and Wonderwerk Cave [20, 93, 94]. Pale orange bars highlight periods of occupation. Palaeoenvironment proxy data from Ga-Mohana Hill (this study), Kathu Pan [25] and Wonderwerk Cave [19, 20]. Pale blue bars highlight wet periods across the sites. Orange star marks the point at ~71 ka, before which human occupation of the region appears to have been associated with the availability of water.

The close proximity of Ga-Mohana Hill, Kathu Pan, and Wonderwerk Cave to each other makes it possible that they were utilized by the same groups of mobile hunter-gatherers, thus providing an opportunity to consider the relationships between wet periods and evidence for human occupation in this region of the southern Kalahari (Fig 4). Between ~251 and 138 ka at Wonderwerk Cave, there is evidence for both wetter conditions and human occupation [20]. Archaeological material at Kathu Pan occurs within palygorskite-coated, water-associated sediments dated to ~156 ka, ~121 ka and ~74 ka [25], with the latter being a Howiesons Poort occurrence. Also at Kathu Pan, wet, marshy conditions that are likely to have supported a significant amount of vegetation occur between ~101 and 80 ka, coupled with evidence for human occupation [25]. At Ga-Mohana Hill, tufa formation at ~114–100 ka correlates with human occupation at the site. Thus, in summary, before ~71 ka, human occupation of the region appears to have been associated with the availability of water.

After ~71 ka, the timing of human occupation and wet periods do not coincide (Fig 4). Tufas at Ga-Mohana Hill indicate that much of MIS 4 and 3 is characterised by wet conditions. The sediments at Kathu Pan continue to indicate the presence of water through much of MIS 4, although the organic-rich marsh sediments do not occur after MIS 5 [25]. However, evidence for human occupation during this time is lacking at both Kathu Pan and Ga-Mohana Hill [36]. There are Middle Stone Age deposits at Wonderwerk Cave that have not yet been securely dated that could potentially represent this period [9], but this remains unknown at this point. From ~43–32 ka, wet conditions are represented at Ga-Mohana Hill, and from ~35–33 ka at Wonderwerk Cave [19]. In contrast, there is evidence for drier conditions from ~32 ka at Kathu Pan [25]. Evidence for human occupation at Ga-Mohana Hill [36] and Kathu Pan [92] date to 31 ± 1.8 ka and 32 ± 0.78 ka, respectively. These data suggest that human occupation at these sites may overlap with a period of decreasing water availability, but the error ranges on the age estimates make it challenging to confidently assert this.

Human occupation is evident again at Ga-Mohana Hill at 14.8 ± 0.8 ka [36] and is associated with evidence for relatively wetter conditions, but at Kathu Pan, human occupation during the LGM [91, 92] is associated with evidence for relatively drier conditions. Late glacial deposits at Wonderwerk Cave indicate an association of dry conditions and human occupation [94]. Thus, MIS 2 provides very little coherence with respect to the relationship between water availability and human occupation; humans are associated with both wet and dry conditions. Through the Holocene, there is persistent evidence for human occupation despite changes in palaeoenvironmental conditions [25, 92, 94].

Discussion

In this semi-arid region with limited, seasonal rainfall and no evidence of actively precipitating tufa, the relict tufa deposits at Ga-Mohana Hill are a record of past periods of conditions favourable for tufa formation, which are primarily an indication of increased water on the landscape. We show that U-Th dating of the tufas, buoyed by the laser ablation screening method, can produce precise ages. We go on to use these ages to show that periods of tufa formation were punctuated over the last 110 ka, with five discrete time periods identified.

Our U-Th dated tufa records suggests periods of increased water availability in the southern Kalahari were not restricted to interglacials. Ga-Mohana Hill shows extensive tufa formation during much of MIS 4, a period generally assumed to be characterised by typical cold and dry glacial conditions across much of the interior of southern Africa [95]. Increased water availability during this time is supported by other palaeoenvironmental records of the Kalahari Basin, such as at Kathu Pan. At Witpan Dunes, approximately 350 km to the north west, the absence of southern Kalahari dune data during MIS 4 [96] indicates unfavourable conditions for dune accumulation, suggesting increased rainfall, decreased windiness, and a denser vegetation cover [97]. To the north, a Makgadikgadi Megalake highstand has been dated to 64.2 ± 2.0 ka suggesting there was also substantial water availability in the Middle Kalahari at that time [98]. We argue that the tufa intervals represent a southern Kalahari environment characterised by a positive hydrological balance and mild temperatures favourable for productive vegetation and soils. In accordance with other recent studies, our results challenge global generalisations of past climate change and highlight the necessity for regionally specific models [16, 21, 85].

In the southern Kalahari, early human population distributions appear to have been modulated by water availability before ~71 ka. After ~71 ka, the picture is much less clear. Despite evidence for wetter conditions, archaeological deposits dating to MIS 4 and the early part of MIS 3 have not yet been identified in the punctuated record of human occupation at Ga-Mohana Hill, nor at nearby Kathu Pan or Wonderwerk Cave. This result poses a new dilemma, in that wet conditions in this region of the southern Kalahari should have theoretically made it attractive for human occupation, but as of yet, no archaeological deposits date to this time. The time interval corresponding to MIS 2 provides little coherence with respect to the relationship between water availability and human occupation. The three records considered here do not agree on whether conditions were wetter or drier during the LGM and humans appeared to have occupied the region through both the LGM and late glacial. Others have highlighted that the palaeoenvironmental record for MIS 2 across the Kalahari Basin and surrounding regions is complex, documenting a high degree of spatial and temporal variability [16]. This lack of coherence may be in part due to the variable responses of palaeoenvironmental proxies to temperature and water availability changes, and potentially lags in responses. A shift in seasonality may also play a role, with some proxies responding to seasonality changes for precipitation, as evidenced at Kathu Pan [25], as opposed to mean annual precipitation.

We note that the absence of evidence may not be evidence of absence, and that issues with site formation, site visibility, and/or dating may instead explain why no archaeological deposits have yet been identified. The inability to fully explain this pattern is one of this study’s limitations. Further work, including the excavation and dating of new archaeological sites, will be key for further testing hypotheses that link early human population distribution patterns to water availability, potential refugia conditions, and interglacial/glacial cycling [4–6, 8, 9].

Conclusion

We have identified and described the tufa deposits at Ga-Mohana Hill, and provided a framework for reliably dating them. The observed tufa morphologies suggest a depositional environment characterised by water cascading down the dolomite steps, flowing over the rock shelters, and ponding in shallow pools. Through U-Th dating, we have produced a new, well-dated record of prolonged water availability linked to human occupation in the southern Kalahari during the Late Pleistocene. Identifying the timing and nature of human occupation in the Kalahari Desert is critical for understanding the emergence of our ability to adapt to new and extreme environments [2]. For a long time, the Kalahari Desert has been considered too arid for early human populations to persist, and evidence for occupation was assumed to represent wetter periods. Until now, a rarity of integrated palaeoenvironmental and archaeological records has largely prevented adequate testing of these assumptions. The results presented here provide evidence for prolonged periods of water availability in the southern Kalahari Basin through much of the Late Pleistocene. There is a positive association with wet conditions and human occupation before ~71 ka. However, by the LGM, water availability alone did not mediate human occupation in the southern Kalahari Desert. This may extend further back in time as the datasets for the time period between ~71 ka and the LGM become more robust. Nevertheless, this result challenges the traditional view that links wet periods to human occupation. This decoupling of human occupation and wet phases in the Late Pleistocene could reflect new social and technological adaptations that helped hunter-gatherers cope more effectively with diverse environmental conditions.

PLOS ONE inclusivity in global research questionnaire.

(PDF)

Click here for additional data file.

Photographs of cascade hand and drill core samples.

a,b) field context of sample 18–4; c) hand sample scan of 18–4 showing fine, undulating layers; d,e) field context of drill core samples 18–17 (f) and 18–16 (g); h) photograph of in-situ cascade tufa sampled with core drill; i) drill core sample 18–12; j,k) thin section photographs of sample 18–12 (in ppl) showing irregular, domal micritic laminae with lenses of microspar and micropores.

(PDF)

Click here for additional data file.

Field photographs, hand samples and thin section photomicrographs of rim pool samples.

a) Field context of sample 17–16 showing circular configuration and surface desiccation cracks; b,c) hand sample photographs showing 4cm layer of carbonate and cemented clasts on the underside; d) field context of sample GHN-1; e) hand sample scan of GHN-1; c) photograph of thin section from GHN-1 showing aphanitic fabric of biomicrite with spar-filled filamentous cavities.

(PDF)

Click here for additional data file.

Photographs of terrace breccia hand samples.

a) terrace breccia sample showing included detrital clasts and brecciated tufa clasts; b) hand sample scan of sample 18–7; c) field context and d) hand sample scan of sample 17–8 showing massive, dense micrite; e,f) thin section photographs of terrace sample 17–8 showing clotted fabric of peloidal micrite with microspar-filled void spaces.

(PDF)

Click here for additional data file.

Photographs, hand sample scans and thin section photographs of barrage tufas.

a) field photograph of sample 17–6; b) hand sample scan of sample 17–6 showing irregular, undulating and discontinuous layering; c) thin section photographs show stromatolite-type micrite crustal laminae alternating with chaotic microspar laminae with detrital and oxide inclusions (c) and discontinuous crinkly microspar laminae with overprinting of oxide precipitates (f).

(PDF)

Click here for additional data file.

Photographs of sampled domes.

a) field context of tufa dome, sampled using an angle grinder; b) dense mm-scale layers alternating with irregular, porous and friable layers; c) photomicrograph of thin section from sample in (b) showing micro laminae; d,e) dome sampled with drill-core; f) hand sample of dome core showing large cavities and porous, reticulate framework.

(PDF)

Click here for additional data file.

High resolution images of cascade (GHN2 and GHS5), terrace (17–8) and rim pool (GHN1, 18–7, GHS6) tufa samples overlain by LA-ICP-MS 238U (left) and 232Th (right) element distribution maps.

Concentrations in ppm shown in adjacent colour scales (warmer colour = higher concentration). Black circles represent approximate locations of subsamples drilled for U-Th dating prior to pre-screening, and oblong free-forms show exact locations at which subsamples were drilled for U-Th dating following LA-ICP-MS analysis. Ages associated with each subsample are given in thousands of years (ka) and are reported in Table 1.

(PDF)

Click here for additional data file.

High resolution images of cascade core samples (18–13, 18–14, 18–15, 18–16, 18–17) overlain by LA-ICP-MS 238U (left) and 232Th (right) element distribution maps.

Concentrations shown in ppm in adjacent colour scales (warmer colour = higher concentration). Black circles represent approximate locations of subsamples drilled for U-Th dating prior to pre-screening, and oblong free-forms show exact locations at which subsamples were drilled for U-Th dating following LA-ICP-MS analysis. Ages associated with each subsample are given in thousands of years (ka) and are reported in Table 1.

(PDF)

Click here for additional data file.

High resolution images of samples with unreliable and imprecise age solutions (S3 Table) with the exception of sample 18–10.2 (bottom of sample) which has an age of 3.0 ± 0.9 ka.

Black oblong outline represents material drilled for U-Th dating. Cascade samples (18–4 and 18–12), barrage samples (17–6 and 18–6) and dome core samples (18–10) overlain by LA-ICPMS 238U (left) and 232Th (right) element distribution maps. Concentrations in ppm shown in adjacent colour scales (warmer colour = higher concentration).

(PDF)

Click here for additional data file.

Tufa sample inventory. Samples labelled and listed in order of collection.

Samples with GH prefix collected in 2016, numerical prefix of other samples indicates the year they were collected (e.g. 17- = 2017).

(PDF)

Click here for additional data file.

U and Th isotope ratios measured in tufa samples with reliable and precise ages.

(PDF)

Click here for additional data file.

U and Th isotope ratios measured in tufa samples which have unreliable or imprecise age solutions.

Errors on all isotope activity ratios are reported with 2σ uncertainty. Upper limit is defined as corrected age plus 2σ uncertainty. All ages have been corrected to account for the effect of detrital Th assuming an estimate for initial 230Th/232Th of 1.5 ± 1.5, and calculated using the 230Th-238U decay constants of Cheng et al. [54] and equation 1 from Hellstrom [47].

(PDF)

Click here for additional data file.

7 Mar 2022

Submitted filename: PONE-D-22-02819_reviewer_EC.pdf

Click here for additional data file.

18 May 2022

Manuscript PONE-D-22-02819

Response to Reviewers

Dear Professor Zerboni,

Thank you for the opportunity to submit a revised draft of our manuscript Tufas indicate prolonged water availability linked to human occupation in the southern Kalahari (reference PONE-D-22-02819) for publication in the journal PLOS ONE. We appreciate the time and effort that you and the reviewers put in to providing useful feedback on our manuscript and we are grateful for the insightful comments and suggestions which definitely improve our paper.

The main issues raised by the reviewers are highlighted below, along with a summary of how we addressed them.

1) Our interpretation of the combined palaeoenvironment and archaeological records as indicating a decoupled relationship between water availability and human occupation at ~71 ka. We addressed this by clarifying and strengthening our interpretation about the relationship between water availability and human occupation. We highlight evidence that water availability and human occupation are indeed correlated until 71 ka. By ~20 ka during the LGM, humans were occupying this region of the southern Kalahari despite evidence for variable (drier) climatic conditions, as opposed to earlier times when occupation appears only during periods of increased water availability (as identified by our tufa ages). We have changed the title so that there is less emphasis on the decoupling and instead highlight the new 110,000 year long tufa record of water availability in the southern Kalahari.

2) Limitations in terms of sampling adequacy. We addressed this by making clear the limitations of this study by acknowledging the challenge to explain gaps in the archaeological record during periods when the palaeoclimate data suggests favourable conditions for human activity in the region. We point to issues with site formation, visibility and dating as potential reasons this gap persists, and highlight the need for newly excavated and well-dated sites to resolve these issues.

3) A request for an acknowledgement that the association of tufas with interglacial climate conditions is an outdated one, and a deeper discussion on the nuanced climatic conditions that tufas represent. We addressed this by including two new paragraphs to discuss the nuance around the presence of tufa deposits, and that they do not simply equate to wet conditions. Instead we argue that tufas form during both glacial and interglacial periods (citing previous studies), and in response to a complex interplay of a variety of environmental parameters, including reduced rates of evaporation and cooler temperatures.

4) The inclusion of several additional references to provide more context and to acknowledge new and relevant studies. We have addressed this by incorporating the suggested references and relevant information from the respective studies. We have thus updated our reference list.

We respond to each of the reviewers’ comments in red below. The changes are tracked within the manuscript and the line numbers referenced below refer to the revised manuscript with tracked changes.

Editor

The two reviewers greatly appreciated your work and conclusions and I also agree that your paper meets the standard of publication required by PLoS ONE, pending minor comments.

Thank you!

Among the suggestions from the two reviewers, I would remark that the tight dichotomy wet/dry (interglacial/glacial) in the formation of spring tufa (and also speleothems) in arid regions is simplistic and many studies are confirming this. In my experience in the hyperacid Sahara and arid Dhofar Mountains of Oman, I noticed that even today spring tufa form, evidently with different sedimentary patterns respect to when they formed during pluvial phase. I would appreciate (as suggested by Rev2) a deeper discussion on this topic, confirming that dealing with spring tufa requires to do beyond the equation wet->deposit, dry->no deposition.

We thank the editor for this comment, and trust that the additions made to the text (detailed in response to the comments below) sufficiently address the complexity of tufa deposits and the environmental conditions they represent. We have added several statements to the text in the “Comparison to global records” section (lines 343-438) to acknowledge that tufas form in response to a variety of climatic factors, and do not merely represent wet / dry conditions (e.g. 353-357; 430-434). In particular, we have included an additional paragraph in which we discuss the nuanced conditions required for tufa formation at Ga-Mohana Hill (lines 426-438). We have also adjusted the language we use throughout the text to move away from the simplistic implications of the term ‘wet’ when referring to the conditions that tufas represent (e.g. lines 127-129).

Reviewer #1

In overview I think this is a useful paper.

Thank you!

The substantive part of the paper is about dating tufas using sensible new methods to choose the best samples for U-Series. My suggestions are therefore relatively minor, in terms of the interpretation. This is a question of minor revisions in my opinion. The authors do a good job at showing an interesting palaeoenvironmental story in terms of tufas. But I am dubious about what this says about decoupling of humans and water. So I think this aspect of interpretation should be removed, and the paper will be stronger. The key conclusion is actually about humid conditions over a long period in Kalahari than previously known.

Thank you. We have put more emphasis on the evidence for water availability as represented by the tufas . We have clarified that we have good evidence for correlation between wet periods and human occupation before 71 ka, and a decoupling of this relationship by the LGM. There are more specific comments about this interpretation below where the Reviewer expands on their point.

Lines 40/45 – all very southern African references. Not surprising given the study area and the authors – but I think a bit of extra punch here would be given by first looking more widely at Africa before zooming in on the Kalahari. E.g. Basell (2008, QSR) on East Africa…Scerri et al. (2014, QSR) on North Africa etc. For instance, in the former, Basell argues that in East Africa humans are seemingly tethered to water sources until MIS 5, and then seemingly not. Seems relevant to me. So I suggest just an initial paragraph on such stuff.

It I also useful to think about some of complexities here…e.g. in North Africa most human occupations correlate with wet episodes…but some seemingly don’t. Uan Tabu, for instance. But this is dated by OSL, done when this was an experimental method, so maybe the dates are wrong. So as well as a bit of a wider geographical focus, a bit of intro on methods stuff would be useful.

Thank you for the reference suggestions and for raising these points. We have added a paragraph to the Introduction (lines 46-56) to provide additional context from studies in other parts of Africa. We have summarised the key conclusions presented in the suggested studies which address the relationship between water availability and Homo sapiens’ mobility in eastern and northern Africa. We have also included an introductory sentence on the methods used to investigate questions relating to human-environment dynamics (lines 56-59).

I think dating tufas is a great way to go for reconstructing humidity in the study area, just nice to situate his more widely in terms of different archives and the history of study.

HG – to think about. Sampling intensity, at one point are enough samples to reliably reconstruct human occupations and regional climate present

Thank you for this suggestion. Although we agree that it is an important consideration, restrictions relating to funding to support field work and sample analysis means that we are limited to work with what we have got so far. In terms of the number of dated tufa samples, we were constrained by the quality of material that can provide a reliable age; in total over 40 tufa samples have been collected, but only 16 of these samples met the criteria for U-Th age dating. Similarly, the number of excavated archaeological sites that date to each time period are limited. Nonetheless, we believe that these data are valuable and do shed light on patterns of the past, even if imperfectly. We have made clearer in the Discussion section what the study’s limitations are, i.e. the challenge to explain gaps in the archaeological record during periods when the palaeoclimate data suggests favourable conditions for human activity in the region (lines 558-560 and 572-575). We point to issues with site formation, visibility and dating as potential reasons this gap persists.

Using the LC-ICP-MS method seems exciting. The methods are described very clearly and all seem sensible, i.e. locating areas with less detrital contamination for dating. This sounds like a method which could be used more widely in tufa studies in the future. I am not an expert at U-Series dating so I cannot comment on specific technical aspects, but I could follow the authors’ description, it all makes sense, and seems reasonable. Likewise, the geological description of the study area is well written. I hope that a U-Series expert also reviews this paper to comment on this aspect.

Thank you.

The typical problem with studies like this is that there will be one or two age estimates for humidity in an area, and it is hard to know how representative they are. The results here show clusters of ages, suggesting wet phases ca. 114-100 ka, 73-48 ka, 44-32 ka and 15-2 ka….this mostly seems fine. However, I am not totally sure about the separation of the middle two groups. Out of interested I plotted the dates (rounded slightly, not considering uncertainty) dating to MIS 4 and 3 out of interest, and it looks like a continuous series to me, not one that is two groups. I’m dubious that this self-evidently represents two distinct humid phases….another way of looking at this…is that the ca. 48 ka age is as close to the oldest date in the 44-32 ka group as it is to the next youngest date in the 73-48 ka group….So I suggest the authors clarify their thinking a little here. To me, a simpler interpretation is that there are three clusters of ages, 114-100 ka, 73-32 ka, and 15-2 ka. I note the authors themselves suggest “our record suggests tufa formation was semi-continuous across MIS 4 and MIS 3” (line 278). It might also be worth thinking about sampling adequacy and how meaningful the inferred gaps are.

Thank you for raising this point. We define the wet episodes based on clusters of overlapping tufa ages, with the bounds defined by the minimum / maximum age calculated using the uncertainties of the ages. We have added this explanation to the manuscript (lines 319-320), and the following sentence has been added to Figure 3 caption to clarify our thinking as suggested (lines 331-334): “The blue bars highlight clusters of overlapping tufa ages and are defined by the minimum and maximum ages in each range, calculated using the 2σ uncertainty associated with the ages. Based on the presence of the tufa deposits, these periods are inferred to represent episodes of greater water availability on the landscape.” We do not infer wet periods where there is no age data, hence our statement of ‘semi-continuous’ over the MIS 4-3 period. In order to be consistent, we have adjusted the fourth episode of tufa formation to include only the two ages that overlap, and have separated the youngest age from this cluster as there is in fact no overlap. We adjusted the text to account for this fifth episode throughout.

In terms of interpretation. They describe how earlier phases of occupation correlate with wet phases, but that there is a decoupling of this after 71 ka. But I don’t really understand this, as their figure 4 suggests that no human occupations are known in the area between the end of MIS 5 and the end of MIS 3. Surely the point about humans not being tethered to water would be shown by human occupations during arid times? Their argument is that they have shown there wet condition, but no evidence for people means this de-tethering. I get what they mean, but not sure how strong an argument that is. I think a simpler interpretation is that with so few archaeological sites in the area dated, it is simply a question of sampling adequacy. If it was really so wet, why would humans not have been in the area? And they even point to possible evidence at Wonderwerk which could date to this MIS 4/3 time. It is a question of sampling adequacy. Do three sites really demonstrate the absence of people from the region? If we only had Wonderwerk and Ga-Mohana, it would appear no people were in the area between ca. 90 and 30 ka. Yet Kathu Pan shows people there at about 75 ka. Who is to say another site would not show people at 65 ka, 55 ka, etc?

Yes – this is a good point. We had included a statement about ‘absence of evidence is not necessarily evidence of absence’, and now we have made it even clearer in the discussion that this is one of the study’s limitations. There is no immediate solution to this problem – the only solution being to dig and date more sites.

That said, the three well-stratified, well-dated sites with overlapping MSA chronologies do provide a lot of value in addressing this environment-human interaction question. There are few situations like this across the African continent. For that reason, we believe these results have a lot of value. It’s true that they do pose a new dilemma – where were people during MIS 4 and the early part of MIS 3 when the southern Kalahari was wetter than it is today? We hope further investigations will help reveal the answer.

Our study has been able to demonstrate correlation between wet periods and human occupation before 71 ka, and a lack of correlation by the LGM. Humans were able to occupy diverse conditions (wet or dry) in the southern Kalahari by the LGM. Our results also suggest that this capacity may predate the LGM, pending further investigations that improve as the archaeological and palaeoenvironmental records become more robust.

It is with the end of MIS 3 and into MIS 2 that more persuasive evidence for decoupling of humans and water is evident, with humans being found in seemingly arid situations. With this evidence it could be stated that this decoupling seemed to happen around 30 ka.

Thank you. We have modified the language we use to add clarity, particularly in the Conclusion section. We now say “However, by the LGM, water availability alone did not mediate human occupation in the southern Kalahari Desert. This may extend further back in time as the datasets for the time period between ~71 ka and the LGM become more robust” (lines 608-612). We also discuss the records for human occupation and palaeoclimate at the end of MIS 3 (lines 506-512) and, although it appears that human occupation coincides with the end of a wet period and the onset of drying conditions, the errors on the available data make it difficult to confidently assert whether human occupation actually corresponds with wet or dry phases, or both.

Even if their ‘decoupling’ argument is correct, it does not really make sense. If humans were living in arid areas this would indeed say something about “new social and technological adaptations that helped hunter-gatherers cope more effectively with diverse environmental conditions” (line 426/427). But the absence of people from what are supposedly good conditions begs the question ‘why was no one there?’ rather than suggesting that some kind of adaptations are involved in people not being somewhere nice.

Point taken. Hopefully, the comments above and the edits we have made to the language resolve this concern. This statement about new adaptations applies to the record by the LGM, but we agree that between then and ~71 ka it is much less clear what was going on.

To clarify, our argument is that there is a correlation between wet periods and human occupation before 71 ka, and a lack of correlation by ~20 ka in LGM. Humans were able to occupy diverse conditions (wet or dry) in the southern Kalahari by the LGM. Our results also suggest that the capacity to occupy this region during dry periods may predate the LGM, pending further investigations that improve as the archaeological and palaeoenvironmental records become more robust.

In summary: this is a useful paper, but I suggest the authors change their interpretation somewhat.

Thanks for the helpful feedback. We have clarified our interpretation in line with the comments made here.

Reviewer #2

The study of von der Maden et al. concerns the study of fossil tufa deposits form the arid region of southern Kalahari in unraveling their connection with local archeological human origins and the impact for the palaeoenvironmental reconstructions.

I have appreciated the study, the obtained ages and the resulting data concerning implications with human settling

Many studies emphasize the close relation.

Thank you!

My only remarks are about:

Tufa morphological components (line 162): there are several morphological classifications about tufa depositional system. As done for the fabric classification (Manzo et al.), I suggest indicating the reference used for the morphological classification (definitions are available from Pentecost and Viles, 1994, Pedley 2009; Arenas-Abad et al. 2010; Jones and Renaut 2010, Capezzuoli et al., 2014),

As suggested by the reviewer, we have acknowledged the references drawn on for the tufa morphological classifications (line 221). We have also included additional statements in this section to clarify and credit sources for the terminology used to describe our observations, e.g. lines 266-267; 271; 284-285.

Comparison to global records (line 271-311): in my opinion this is paragraph is a bit ambitious, not only because the here reported bibliography is really poor (please note that there is already a review article on tufa/travertine deposition and climate…and it is not cited: Ricketts, J.W., Ma, L., Wagler, A.E. and Garcia, V.H. 2019. Global travertine deposition modulated by oscillations in climate. Journal of Quaternary Science, 34, 558–568,https://doi.org/10.1002/jqs.3144), but because the tufa deposition is really different in different setting/climatic regimes of the world. I agree with authors that the dichotomy wet/dry – interglacial/glacial view is to be considered simplistic (Line 275-276), but their consequent suggestion that tufa formation is not restricted to interglacial periods is not new!

We thank the reviewer for the reference suggestion and for raising these points. We have adjusted the wording in lines 344-348 to better capture our intention with this section, which is not to evaluate global records of tufa deposition, nor to generate a one-fits-all formula for tufa formation, but rather to assess the extent to which the timing of tufa formation at Ga-Mohana can be explained by changes in selected global climate parameters, e.g. global ice volume (using the Lisiecki-Raymo benthic stack), austral summer insolation, and sea surface temperatures in the Indian Ocean.

We have included the point that tufa formation occurs across a range of climatic settings, and that their formation is controlled by a variety of factors operating on different scales, thus not merely signalling warm and wet interglacial climate conditions (lines 344-345; 353-359; 419-423). We also adjusted the wording to acknowledge that other studies report tufa formation during both glacial and interglacial periods (lines 355; 357-359).

In addition, we have expanded our interpretations of our comparisons to the selected records of climate change by explaining what exactly these comparisons mean for tufa formation at Ga-Mohana in terms of reflecting global drivers of climate change (lines 390-398 and 406-417). We conclude with a description of the conditions the Ga-Mohana tufas represent, which goes beyond the simple deposition = wet interpretation, and instead includes an acknowledgement of the influence of other parameters (evaporation rates, temperature) involved in creating favourable conditions for tufa formation in semi-arid karst settings (lines 426-438).

For review between climate and tufa deposition, I suggest to see Andrews, 2006; Pedley, 2009, while about the former presence of active tufas as record of important rainfall regime shifts, this has been already showed in distal glacial (South Europe, Capezzuoli et al., 2010; Alexandrowicz, 2012), in semi-arid environments (Brazil, Auler & Smart, 2001; Spain, Luzon et al., 2011) and desert settings (Namibia, Viles et al., 2007; Libya, Cremaschi et al., 2010; Ethiopia, Moeyersons et al., 2006). In contrast, the presence of tufas deposits in tropical and monsoon-dominated settings testifies to an absence of destructive large wet season floods and, consequently, for reduced periods of rainfall (Carthew et al., 2003, 2006).

Thank you for this note. We are familiar with these papers and have included them as references in our new statements on lines 344-345 and 421-424, which acknowledge the multitude of studies that document the variety of settings that tufas form in, and the associated complexity of the climatic conditions they represent.

In the correct hypothesis to study the local weather conditions (rainfall and insolation) as constraining for the local tufa deposition, I suggest to also consider and compare with the Egyptian tufa described in Kele et al. 2021 (https://doi.org/10.1144/jgs2020-147) and relative discussion implementing insolation, humidity, radiometric dating and isotope (also showing that the record of climate in Egypt’s tufa is inconsistent with a simple model of palaeoclimate for this region).

Thank you for this useful paper recommendation. We have included a paragraph (lines 382-390) to compare our observation of a lack of correlation between tufa at Ga-Mohana Hill and summer insolation to the conclusion in Kele et al 2021, which is that tufa formation in southern Egypt does not correlate to insolation either, and that the mechanisms driving rainfall in the region (and subsequent tufa formation) are complex, and cannot be attributed to a single forcing factor.

Very minor trifles are indicated in the attached pdf (copied below):

Line 102: ‘pomeroy’ misspelled

Corrected.

Line 181: please check if the S1 Fig. citation is correct

We double-checked the caption and rearranged the order of the samples listed, otherwise it is correct.

Line 277: were. The association with interglaciation is now old!!

Corrected, we adjusted the wording to acknowledge that tufa forms during both glacial and interglacial periods (lines 353-354).

Line 280-281: Please pay attention to the global. Noteworthy the "simple product of changing global climates" is already claimed

We have addressed this by adjusting our wording to acknowledge that our study echoes similar conclusions to those in previous studies (lines 358-358).

Line388-389: This seems an old concept.

Agree, wording adjusted. It now reads: “In accordance with other recent studies, our results challenge global generalisations of past climate change and highlight the necessity for regionally specific models.” (lines 550-552).

Line 419: Mistake or real reference??

We made a mistake with the formatting, but it’s a real reference. It has been corrected.

Hope it helps

Enrico Capezzuoli

Journal Requirements:

When submitting your revision, we need you to address these additional requirements.

1. Please ensure that your manuscript meets PLOS ONE's style requirements, including those for file naming. The PLOS ONE style templates can be found at

https://journals.plos.org/plosone/s/file?id=wjVg/PLOSOne_formatting_sample_main_body.pdf and

https://journals.plos.org/plosone/s/file?id=ba62/PLOSOne_formatting_sample_title_authors_affiliations.pdf

2. In your Methods section, please provide additional information regarding the permits you obtained to collect samples for the present study. Please ensure you have included the full name of the authority that approved the field site access and, if no permits were required, a brief statement explaining why.

Done – a statement was added to the methods section (lines 159-162).

3. Please include a complete copy of PLOS’ questionnaire on inclusivity in global research in your revised manuscript. Our policy for research in this area aims to improve transparency in the reporting of research performed outside of researchers’ own country or community. The policy applies to researchers who have travelled to a different country to conduct research, research with Indigenous populations or their lands, and research on cultural artefacts. We note that some of the authors of your manuscript are from outside of South Africa. The questionnaire can also be requested at the journal’s discretion for any other submissions, even if these conditions are not met. Please find more information on the policy and a link to download a blank copy of the questionnaire here: https://journals.plos.org/plosone/s/best-practices-in-research-reporting. Please upload a completed version of your questionnaire as Supporting Information when you resubmit your manuscript.

Done - the Inclusivity Questionnaire is included as a Supplementary Checklist and referenced in the manuscript (lines 162-164 and 977).

4. We note that Figure 1 in your submission contain copyrighted images. 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 more information, see our copyright guidelines: http://journals.plos.org/plosone/s/licenses-and-copyright.

We require you to either (1) present written permission from the copyright holder to publish these figures specifically under the CC BY 4.0 license, or (2) remove the figures from your submission:

a) You may seek permission from the original copyright holder of Figure 1 to publish the content specifically under the CC BY 4.0 license.

We recommend that you contact the original copyright holder with the Content Permission Form (http://journals.plos.org/plosone/s/file?id=7c09/content-permission-form.pdf) and the following text:

“I request permission for the open-access journal PLOS ONE to publish XXX under the Creative Commons Attribution License (CCAL) CC BY 4.0 (http://creativecommons.org/licenses/by/4.0/). Please be aware that this license allows unrestricted use and distribution, even commercially, by third parties. Please reply and provide explicit written permission to publish XXX under a CC BY license and complete the attached form.”

Please upload the completed Content Permission Form or other proof of granted permissions as an "Other" file with your submission.

In the figure caption of the copyrighted figure, please include the following text: “Reprinted from [ref] under a CC BY license, with permission from [name of publisher], original copyright [original copyright year].”

b) If you are unable to obtain permission from the original copyright holder to publish these figures under the CC BY 4.0 license or if the copyright holder’s requirements are incompatible with the CC BY 4.0 license, please either i) remove the figure or ii) supply a replacement figure that complies with the CC BY 4.0 license. Please check copyright information on all replacement figures and update the figure caption with source information. If applicable, please specify in the figure caption text when a figure is similar but not identical to the original image and is therefore for illustrative purposes only.

We produced the figure in ArcGIS and it contains no copyrighted images. We have added text to the Figure 1 caption to acknowledge the data sources used (lines 107-110) and updated the reference list to include these.

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.

Done.

Submitted filename: Response to Reviewers.docx

Click here for additional data file.

6 Jun 2022

Tufas indicate prolonged periods of water availability linked to human occupation in the southern Kalahari

PONE-D-22-02819R1

Dear Dr. von der Meden,

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,

Andrea Zerboni, Ph.D.

Academic Editor

PLOS ONE

Additional Editor Comments (optional):

We appreciate your efforts n revising your paper and we are happy to accept for publication. One of the reviewer only suggest a very recent paper that you may consider at this stage:

Kaboth-Bahr et al (2021) on the point that climate variability does in Africa does not simplistically follow a glacial/interglacial logic:   https://www.pnas.org/doi/pdf/10.1073/pnas.2018277118

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 #1: 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 #1: Yes

**********

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

Reviewer #1: 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 #1: 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 #1: 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 #1: I think the authors have done a good job at revising their paper. I have no further comments and look forward to seeing this published.

The authors may wish to consider adding a reference to the recent paper by Kaboth-Bahr et al (2021) on the point that climate variability does in Africa does not simplistically follow a glacial/interglacial logic:

https://www.pnas.org/doi/pdf/10.1073/pnas.2018277118

**********

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 #1: No

**********

27 Jun 2022

PONE-D-22-02819R1

Tufas indicate prolonged periods of water availability linked to human occupation in the southern Kalahari

Dear Dr. von der Meden:

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

Prof. Andrea Zerboni

Academic Editor

PLOS ONE