The Oldowan horizon in Wonderwerk Cave (South Africa): Archaeological,

geological, paleontological and paleoclimatic evidence

Michael Chazan, D. Margaret Avery, Marion K. Bamford, Francesco Berna, James Brink,

Yolanda Fernandez-Jalvo, Paul Goldberg, Sharon Holt, Ari Matmon, Naomi Porat, Hagai

Ron, Lloyd Rossouw, Louis Scott, Liora Kolska Horwitz

Supplemental Material

1. Micromorphology, FTIR, and Microspectroscopy (µFTIR) (Figure S1-S10)

2. Grass Phytoliths (Figure S11-S13)

3. Stratum 12 Fauna (Table S1-S2, Figure S14-S15)

1. Micromorphology, FTIR, and Microspectroscopy (µFTIR)

A. Micromorphology

Methods

Samples were taken as intact blocks, wrapped with tissue paper, and bound tightly with plastic

wrapping tape (Goldberg and Macphail 2006). After collection, the samples were initially treated

at the Boston University MicroStratigraphy Laboratory. There, they were partially unwrapped

and oven dried for several days at 60°C and then impregnated with unpromoted polyester resin,

diluted with styrene at a ratio of 7:3 and catalyzed with methyl-ethyl-ketone peroxide (MEKP).

After the resin had gelled to a firm consistency, the samples were returned to the oven overnight

at 60°C. They were then cut to size (50 × 75 mm by 10 mm thick) using a rock saw; in instances

where the sampled blocks were relatively long (e.g., sample 22), two thin sections (A, upper, and

B, lower) were made from each one. The cut blocks were then sent to Spectrum Petrographics

(Vancouver, WA, USA) where they were processed into petrographic thin sections 30 μm thick.

The thin sections were examined with binocular and petrographic microscopes in plane-polarized

(PPL) and cross-polarized (XPL) light at magnifications ranging from 20× to 200×. Descriptive

nomenclature follows that of Stoops (2003) and Courty et al. (1989). Samples come from the

base of the profile shown in Fig. 2 (in text). A close-up of the samples and thin section macro

scans are shown in Fig. S1.

References

Goldberg, P., Macphail, R., 2006. Practical and Theoretical Geoarchaeology. Blackwell

Publishing, Oxford.

Courty, M.A., Goldberg, P., Macphail, R. I., 1989. Soils and Micromorphology in Archaeology.

Cambridge Manuals in Archaeology. Cambridge University Press, Cambridge.

Stoops, G., 2003. Guidelines for Analysis and Description of Soil and Regolith Thin Sections.

Soil Science Society of America, Madison, WI.

Figure S1. Shown here are locations of 2005 samples 9, 10, and 11 within Archaeological

Stratum 12 in the lower part of the profile (see Fig. 2 of main text, scale 10 cm.). The

stratigraphic units are shown at the right-hand side of the profile. Note the following in the thin

section scans, which measure 50x75 mm: (a) finer sandy laminations within sample 11 and the

darker bed rich in ironstone at the base (arrow); (b) bioturbation which disrupts the bedding in

the lower part of sample 11, and which also forms darker brown, cm-thick bands across thin

sections 10B and 10C (arrow); (c) a cm-thick white band at the base of thin section WW-05-9

(arrow) which is primarily of the phosphate mineral dahllite (carbonate hydroxylapatite) but also

contains localized radiating crystals of montgomeryite; (d) pale yellow spots in the very upper

part of sample WW-05-9 (arrow), which are composed of dahllite as determined by FTIR (see

below). Details of some of these features are shown in the photomicrographs below (Figs. S2-

S6).

Figure S2. Photomicrograph of slide WW-05-11B showing fine laminated sand at the bottom

with some graded bedding, overlain by rounded mm-size grains of ironstone. Plane-polarized

light (PPL).

Figure S3. Fragment of a disrupted and rotated piece of a slaking crust from sample WW-05-

10B (PPL).

Figure S4. Circular, radiating crystals of montgomeryite with the white band at the base of

sample WW-05-9. This white matrix is dahllite and represents diagenetic alteration of an

originally calcareous flowstone (PPL).

Figure S5. Rounded pale brown aggregates composed of silty clay with quartz inclusions from

the upper part of sample WW-05-9, in lithostratigraphic Unit 8 (PPL).

Figure S6. Elongated, almost vertical piece of bone in the center surrounded by quartz sand

from sample WW-05-10C; a piece of ironstone can be seen left of the bone fragment (PPL).

Figure S7. Micromorphology sample WW-09-11 from unit underlying lithostratigraphic Unit 9.

This sample consists of finely laminated very fine silts and clay that are highly contorted and

deformed. They are locally stained by iron and perforated by mm-sized burrows [‘passage

features’ (Stoops, 2003)] that contain fine quartz sand and aggregates of the finely laminated

silt/clay. Some of the burrows locally exhibit areas with fine sparry calcite.

Figure S8. Detailed photomicrograph of upper part of sample WW-09-11 showing contorted

finely laminated silt and clay.

Figure S9. Contorted and finely laminated silts and clays with ~cm-wide passage feature

running diagonally across the photograph from sample WW-09-11.

B. Fourier Transform Infrared Spectroscopy (FTIR) and Microspectroscopy (μFTIR)

Sediment samples were taken from Squares Q32 and R32 during excavation by our team in 2007

and 2009. Sediments in Q32 were homogenous red sands with burrow features consistent with

Sedimentary Unit 9. Excavation was by 5cm spits and samples were taken throughout the

sequence. In Square R32, Beaumont (pers. comm.) suspected the presence of a hearth based on

the occurrence of mottled black and white sediments in the top 3cm overlying red sands of

lithostratigraphic Unit 9. Field observations by Goldberg cast doubt on the identification of

these sediments as a hearth feature, which appear more consistent with a surface that has

undergone complex diagenesis. This surface was also observed during excavation by the late

Prof. Hillary Deacon. Excavation was limited to the top five centimeters of north half of Square

R32. The contact with the underlying sands was found to be sharp. This lens is tentatively

correlated with the white sediments found at the interface between lithostratigraphic Units 8 and

9 in the main section.

Methods

Powdered sediment samples were analyzed by FTIR spectroscopy using a Thermo-Nicolet

Nexus 470 IR spectrometer. Representative FTIR spectra were obtained by grinding a few tens

of micrograms of sample with an agate mortar and pestle. About 0.1 mg or less of the sample

was mixed with about 80 mg of KBr (IR-grade). A 7 mm pellet was made using a hand press

(Qwik Handi-Press, Spectra-Tech Industries Corporation) without evacuation. The spectra were

collected between 4000 and 400 cm-1 at 4 cm-1 resolution. Samples processed in thin section

were analyzed by FTIR microspectroscopy using a Thermo-Spectra-tech Continuum IR

microscope attached to the Thermo-Nicolet Nexus 470 IR spectrometer. Spectra of particles with

diameter of about 100 µm were collected in total reflectance or transmission mode with a

Reflectocromat 15x objective between 4000 and 450 cm-1 at 8 cm-1 resolution.

Results

FTIR investigations provided information about the composition of the clay minerals and the

authigenic phosphates contained in the sediments of lithostratigraphic Units 8 and 9. All

sediments show diffuse presence of kaolinite and smectite. Amongst the authigenic minerals

carbonate, hydroxyl apatite and montgomeryite have been locally detected. FTIR analysis of

several bone fragments (Fig. S10) shows that a significant portion of these bone are likely to

have been heated at a temperature of, or above, 500° C and below 800° C as suggested by Berna

(2010).

Reference

Berna, F. 2010. In: Scientific Methods and Cultural Heritage. An Introduction to the Application

of Materials Science to Archaeometry and Conservation Science, G. Artioli, Ed. Oxford

University Press, Oxford, pp. 364-367.

Figure S10. At top: Photographs of bone fragments found in Stratum 12 from Excavation 1 at

Wonderwerk Cave. At bottom: Representative Infrared spectrum showing IR absorptions at 630,

1090, and 3570 cm-1 (circled in red) characteristic of bone carbonate hydroxyl apatite burnt at or

above 500 °C. The mineral transformation resulting in the sharpening of the phosphate anti-

symmetric stretch band at 1020-1100 cm-1 with splitting in two distinct peaks at ca. 1040 and

1090 cm-1 and the appearance of 630 cm-1 band are characteristic of high temperature processes

and are not known to form as a consequence of low temperature geochemical processes (Berna

2010).

2. Grass Phytoliths

Eight soil samples were extracted from archaeological Strata 12-11. Approximately 50g of each

sample was used for each phytolith extraction. Essential steps included deflocculation, removal

of clays by means of sedimentation and the elimination of carbonates using HCl in low

concentration (10%). Phytolith extraction involved mineral separation with a heavy liquid

solution of sodium polytungstate (S.D. = 2.3). Fractions were mounted on microscope slides in

glycerin jelly and scanned under a Nikon 50i polarizing microscope at x500 magnification.

Phytolith counts were based on systematic scanning following standardized transects for each

slide. Counts were normalized as percentages of the short-cell sum for each sample, based on the

total count of all the morphotypes per slide. Preliminary results indicated that except for one

sample, grass phytoliths were generally well-represented throughout the sequence (Fig. S11).

Interpretation of the results was based on previous phytolith studies that have emphasized the

importance of studying modern phytolith assemblage variability in order to interpret fossil

assemblages on a regional scale, through ‘vegetative’ rather than ‘floristic’ reconstructions of

grass communities (Fredlund and Tieszen 1994), i.e. to assign ecological meaning to short cell

phytoliths by comparing a range of ecological preferences in modern grasses with the phytoliths

that they produce, rather than just using phytoliths to discriminate between grass subfamilies,

tribes or genera. This was accomplished by constructing a model of proportional representation,

based on the relative occurrence and rate of production of eleven grass silica short-cell phytolith

morphotypes produced in three hundred and nine modern grass species (National Museum

Bloemfontein Phytolith Database). One hundred short-cell phytoliths were counted per

specimen, each specimen representing one species. Phytolith morphotype proportions were

tabulated according to their distribution and the ecological adaptation of the grasses that produce

them. From the data, phytolith profiles were created for seasonal rainfall variability and habitat

preference (biomes), showing meaningful correlation between C4-grass phytolith abundance and

<500 mm. annual summer rainfall conditions (r2 = 0.6427, r = 0.8017, p = 0.0010). Saddle-

shaped and Variant 1 bilobate (Variant 1 bilobates have a central portion or neck equal or greater

than one third of total length of body) morphotypes associate with warm, locally mesic to dry

and arid summer rainfall growing conditions related to Nama-Karoo or Savanna grass

communities, while trapeziform-shaped morphotypes (including trapezoid, rondel, oblong- and

reniform-shaped morphotypes) affiliate with cooler, winter rainfall growing conditions related to

Succulent Karoo and Fynbos grass communities (Fig. S12). Ordination through correspondence

analysis (CA) was performed to aid interpretations. Based on comparison with modern

morphotype ecology, the phytolith composition of lithostratigraphic Unit 9 correlates with a

warm Savanna to Nama-Karoo type grassland, while the top of lithostratigraphic Unit 8 shows a

slight shift that becomes pronounced in overlying Stratum 11, towards a more arid environment

with cooler growing conditions analogous to Succulent Karoo conditions (Figs. S13-13; Fig. 5 in

text).

Reference

Fredlund, G.G., Tieszen, L.T. 1994. Modern phytolith assemblages from the North American

Great Plains. Journal of Biogeography 21, 321 – 335.

Figure S11. Total number of grass short-cell phytoliths counted per sample.

Figure S12. Ternary graph showing the relationship between modern grass short-cell

morphotype production and three different seasonal rainfall parameters prevalent in the southern

African region.

Figure S13. Percentage diagram of grass silica short-cell phytolith retrieved from Strata 11 and

12. Counts were normalized as percentages of the short-cell sum for each sample, based on the

total count of all the morphotypes per slide.

3. Stratum 12 Fauna

SPECIES NISP

REPTILIA

Chelonia Tortoise indet. 4

Ophidia Snake indet. 2

Reptilia indet. 9

AVES

Struthio eggshell 8

Indet. (small) 12

MAMMALIA

Rodentia Hystrix sp. 7

Pedetes sp. 4

Lagomorpha Lepus capensis 1

Pronolagus sp./ Bunolagus sp. 3

Lagomorpha indet. 4

Primates Cercopithecidae indet. 1

Carnivora

Canidae Indet. (large-medium) 1

Indet. (small-medium) 3

Indet. (small) 1

Felidae Indet. (small-medium) 1

Hyracoidea Procavia transvaalensis 9

Procavia antiqua 7

Procavia sp. 26

Ungulata Indet. (large-medium) 11

Perissodactyla Equidae indet. 22

Hipparionini indet. 1

Artiodactyla

Bovidae Caprinae indet. (large.) 1

Alcelaphini indet. 9

Antilopini indet. 1

Pelea capreolus 1

Indet. (large) 1

Indet. (large-medium) 45

Indet. (small-medium) 57

Indet. (small) 10

TOTAL 262

Table S1. Large vertebrate remains recovered from Wonderwerk Cave, Stratum 12, according to

number of identified specimens (NISP).

Order: Family (no. genera) Common Name Stratum 12

Afrosoricida: Chrysochloridae (2) golden moles 0.15

Macroscelidea: Macroscelididae (2) elephant shrews 7.90

Soricomorpha: Soricidae (3) shrews 14.77

Chiroptera: various (6) bats 1.44

Rodentia: Gliridae (1) dormouse 0.06

Rodentia: Nesomyidae (6) various mice 25.47

Rodentia: Muridae: Gerbillinae (3) gerbils 28.39

Rodentia: Muridae: Murinae (8) various mice 15.82

Rodentia: Muridae: Otomyinae (2–3) vlei & bush karoo rats 4.92

Rodentia: Bathyergidae (2) molerats 1.07

MNI 14226

Table S2. Proportional representation of micromammal Orders and Families in samples from

Wonderwerk Cave Excavation 1, Stratum 12. Taxonomy after Wilson & Reeder (2005). MNI =

Minimum Number of Individuals. This was calculated based on sided mandibles and maxillae.

Reference

Wilson, D.E. and Reeder, D.M. (eds.) 2005. Mammal Species of the World. Baltimore: Johns

Hopkins University Press.

Figure S14. Scanning electron microscopy (SEM) image of a partially blackened rodent mandible

from Wonderwerk Stratum 12 (see Fig. S15). a. The mandible, when observed under the SEM

(backscattered electron mode, BSE-SEM) shows whiter patches (more metallic nature, higher

emission) than the rest of the bone (more organic, more greyish). b. The square in the general view

of the mandible that was chemically analyzed using energy dispersive spectrometry (EDS). c. The

chemical analysis of the area of the purple box in b (WDK-04) shows a high manganese peak. This

analysis shows that the composition of the white patches is manganese. d. Chemical analysis of the

point WDK-05 marked by a purple cross in b located beyond the white patch. This analysis gives a

calcium phosphate spectrum of the bioapatite composition of the bone and lacks the manganese

peak found in the analysis of WDK-04.

Figure S15. Comparison between the scanning electron microphotograph (a) and the optical light

photograph (b) of the detailed area marked by a box in the microphotograph. When the mandible is

observed under the SEM, using backscattered electron mode, the black spots visible in the optical

light photo disappear and do not show higher emission (whiter color) as would be expected if the

black spots had metallic (manganese) composition. On the contrary, the spots have identical

organic composition (no metallic, grayish colour under the BSE-SEM) on both the bone and the

dentine of the tooth. This can happen only due to another taphonomic agent- burning- that changes

the bone to black but does not alter the bioapatite composition.