# Last Interglacial fossil elephant trackways dated by OSL/AAR in coastal aeolianites, Still Bay, South Africa

> About 140,000 years ago, elephants walked across the coastal dunes of South Africa’s southern Cape, and their footprints turned to stone. Reading these fossil trackways much as a tracker reads fresh ones, researchers found a well used game trail, a herd crossing, and two calves less than a year old walking side by side. They are the southernmost record of the African elephant so far, and hint at a wetter, more wooded coast than today.

Summary by Kersey Lawrence, Original Wisdom. Updated 9 November 2026. Page: https://www.originalwisdom.com/tracking-literature/2008-roberts-bateman-last-interglacial-fossil-elephant/

## Citation and access

- **Authors:** David L. Roberts, Mark D. Bateman, Colin V. Murray-Wallace, Andrew S. Carr and Peter J. Holmes
- **Year published:** 2008
- **Journal or source:** Palaeogeography, Palaeoclimatology, Palaeoecology
- **Full citation:** Roberts, D. L., Bateman, M. D., Murray-Wallace, C. V., Carr, A. S. and Holmes, P. J. 2008. Last Interglacial fossil elephant trackways dated by OSL/AAR in coastal aeolianites, Still Bay, South Africa. Palaeogeography, Palaeoclimatology, Palaeoecology 257(3): 261–279.
- **DOI:** 10.1016/j.palaeo.2007.08.005
- **Access status:** Publisher link only
- **Where to find it:** https://doi.org/10.1016/j.palaeo.2007.08.005

## Tags

- **Species:** Mammals > African elephant
- **Region:** Africa > South Africa
- **Tracking:** Gaits; Paths; Track measurements; Tracking; Tracks; Tracks and signs identification
- **Behaviour and ecology:** Behaviour; Movement
- **Data type:** New field data
- **Observer qualification:** Not reported
- **Discipline:** Palaeoanthropology; Taphonomy; Wildlife tracking
- **Year published:** 2008

## Detailed keywords

- **Species:** Studied: fossil footprints and trackways of the African elephant (Loxodonta africana), which the authors think very probably belong to the Cape elephant (L. africana africana). Other fossil trackways present but not studied here (to be described elsewhere): carnivore, antelope, equid, rodent and tortoise (testudinae); antelope trackways cross one elephant trackway surface. Fossils in the dune deposits: dune snails Tropidophora sp. (spelled “Trophidophora” in Table 4) and Achatina zebra, the brackish-water slug Ferissia sp., and fossil termite nests. Humans: Middle Stone Age people are known from the region (six stone artefacts in palaeosol P1, and nearby Blombos Cave), but there is no evidence they hunted elephants.
- **Location and habitat:** Sea-cliff exposures (up to 70 m high, along about 12 km of coast) of Pleistocene coastal aeolianite (wind-blown dune sand turned to rock; Waenhuiskrans Formation) east of Still Bay, on the west-southern coast of South Africa. Elephant tracks were described at five sites along about 3 km of this coast, at about 34° 20′ S. Nearby: Rietvlei, a freshwater interdune lake about 6 km east of Still Bay; Goukou River; Blombos Cave about 20 km to the west. Present-day vegetation is fynbos scrubland; the nearest forest is about 150 km east around Knysna. Ancient landscape: partly vegetated parabolic dunes with broad, gently undulating interdune areas.
- **Topics:** fossil elephant footprints and trackways (ichnology); telling hind foot (pes) from forefoot (manus) by shape; direction of travel from toe marks, an anterior bulge and push-up mounds; stride length and changes in gait; downslope sliding on dune faces; footprint walls, undertracks and overtracks; estimating shoulder height and age from hind foot length; herd and group movement; repeated use of a game trail; footprint preservation in damp dune sand (taphonomy); dating of dune deposits; ancient soils (palaeosols); ancient wind direction; Last Interglacial climate and vegetation; southernmost record of the African elephant; elephants and early modern humans.
- **Statistics or models:** No statistical tests. Optically stimulated luminescence (OSL) dating of quartz grains (single aliquot regenerative dose protocol, central age model, with an age-depth model for the cosmic dose); amino acid racemisation (AAR) dating of snail shells and whole-rock sediment (high performance liquid chromatography). Grain-size and colour measurements of the sand; a rose diagram of 83 dune foreset measurements for ancient wind direction. Shoulder height and age read off published East African graphs relating hind foot length to shoulder height and age (Western et al., 1983).
- **Data type:** New field data: footprint measurements, sediment descriptions and samples from two measured cliff sections. Dates of fieldwork are not given. The tracks were dated to Marine Isotope Stage (MIS) 5, the Last Interglacial (OSL ages of about 140 and 136 thousand years for the sand just below and above the main elephant trackway horizon). Inconsistencies in the paper: (1) the text gives the OSL age range as about 143 to 8 thousand years, but the oldest age in Table 3 is 140 ± 8 thousand years; (2) the text lists the three samples around palaeosol P1 as “Shfd05012, Shfd05011 and Shfd05012”, repeating one code (Table 3 shows Shfd05010 for the third); (3) the OSL ages around the main trackway are older than those of the palaeosol P1 that lies below it, which the authors call a minor age inversion; (4) at site 3B the elephant trackways are said to be overprinted by antelope trackways “and are therefore younger than either”, which reads the wrong way round (being overprinted would make the elephant tracks the older ones); (5) dune-building winds are called westerly throughout but “southwesterly” once in the discussion; (6) the discussion refers to Section 3.3 for the wetland evidence, which is in Section 3.4.

## Synopsis

### 1. What they did

The sea cliffs east of Still Bay, on South Africa’s southern coast, are made of old coastal dunes that have turned to rock. They hold many fossil trackways of mammals and reptiles, including the African elephant, but neither the tracks nor the rocks had been described in any detail, and their age was uncertain.

The authors:

- Described the dune rocks in two measured cliff sections: the types of layering, the ancient soils (palaeosols) and the direction of the ancient winds.

- Dated the rocks with two methods used together for the first time in South Africa: optically stimulated luminescence (OSL), which measures how long sand grains have been buried, and amino acid racemisation (AAR), which measures slow chemical change in shells and shell grains.

- Described elephant footprints at five sites along about 3 km of coast, seen in cliff faces, as casts on the underside of rock layers and as natural impressions on old ground surfaces. Where possible they measured footprint size, stride and push-up mounds, and used the length of the hind foot to estimate each elephant’s shoulder height and age from East African reference data.

- Considered how the footprints were preserved and what the elephants’ presence says about the climate, vegetation and early humans of the time.

Only the elephant tracks are described here; the other tracks are to be described elsewhere.

### 2. What they found

Age of the dunes and tracks

- The dunes built up during the Last Interglacial (Marine Isotope Stage 5), with OSL ages from about 140 to 91 thousand years. Dune building stopped at about 90 thousand years ago, and a mature soil formed over a long gap until new dunes arrived about 8 thousand years ago.

- The sand just below and above the main elephant trackway layer gave OSL ages of 140 and 136 thousand years. These are slightly older than the soil layer beneath them, a small age inversion the authors note; they read them as very early in the Last Interglacial.

- The dune sand contains shell grains reworked from older dunes, which explains why the whole-rock AAR results look older than the snail shells.

The landscape

- The dunes were large, partly vegetated parabolic dunes built mainly by westerly winter winds, as today.

- Elephant and other trackways are most common in the low-angle layered sand of broad, gently sloping areas between dunes, and rare on the steep dune faces.

- Near the base of the cliffs, iron-stained bands and a brackish-water slug point to a wetland, near where the freshwater lake Rietvlei lies today.

The elephant tracks, site by site

- Site 1: trampled layers stacked over about 4 m of cliff, showing a game trail used again and again over a long time. At the base, a 33 m stretch is so heavily trampled that prints overlap and single trackways cannot be followed, suggesting a sizeable herd crossed it. No accurate measurements were possible.

- Site 2: a fallen block with three prints as casts. Print A has the broadly oval shape of a hind foot, 38 by 27 cm and 11 cm deep, with a slight bulge at the front that shows the direction of travel. Print B is round, typical of a forefoot (33 cm across, 9 cm deep). The hind foot points to an elephant about 225 cm tall at the shoulder and about 13 to 14 years old: large, but not fully grown.

- Site 3A: two prints on a wind-rippled surface, about 75 cm apart. The round hind print (25 cm across, with a 3 cm wall around it) sits in a perfectly preserved wedge of displaced sand. The other print is blurred and stretched by a downslope slide and is probably the forefoot. The ripples formed before the elephant passed. The firm, cohesive displaced sand suggests rain had moistened the ground. This animal was about 160 cm at the shoulder and four to five years old.

- Site 3B: on the layer beneath, three elephant trackways (A, B and C) cross a rippled surface, with antelope trackways crossing them. Trackways A and B (6 and 5 prints, average stride 65 cm) run side by side and curve together, suggesting two very young elephants walking together and influencing each other. Trackway C splits off at about 30 degrees and is probably a third animal. Hind prints of 17 to 20 cm point to calves about 110 cm tall and less than a year old. Some short strides (about 25 cm) show changes in gait. Two toe marks, rarely seen in elephant prints, show the direction of travel. All three trackways show the narrow track width typical of elephants, whose legs swing beneath the body. Push-up mounds on the same side of every print show the elephants were walking and sliding down a slope.

- Site 4: prints seen only in cross-section, on steep dune faces (25 to 32 degrees), 35 m above sea level. These disturb the sand most deeply, up to 30 cm, perhaps because of uneven weight on the legs while negotiating a steep, unstable slope, or loose sand.

How the tracks were preserved

- Only dune sand moistened by rain or dew holds the shape of a footprint. The many well preserved tracks suggest the dunes were often wet.

- Westerly winds dry the windward side of a dune and blow sand onto the damp, sheltered side, burying prints there. Since dunes here were built mainly by winter westerlies, the prints were probably preserved mostly in winter.

- Deep footprints commonly left undertracks and overtracks in the layers below and above.

- Preservation depended on how damp and compact the sand was, the slope, and how soon the prints were buried. Burial deeper than about 2 m protected them from roots and burrowing animals, and rapid hardening of the shelly sand helped preserve them.

### 3. What they concluded

- These are, as far as the authors know, the only fossil footprints of the African elephant, and the only known Pleistocene elephant footprints from Africa. They are probably those of the Cape elephant, the same subspecies as the historical “Knysna elephants”.

- At about 34° 20′ S, they are the southernmost record of the African elephant so far. Sea level was lower than today, so elephants may have roamed even further south.

- Elephants seem to have crossed this dune landscape in numbers over a long period. A migration route through difficult active dunes seems unlikely; the elephants may have been drawn by water in an interdune wetland.

- Because elephants generally need plenty of browse, their presence may mean more woodland and a wetter climate than today, which fits the signs of a high water table, a wetland and frequently wetted dunes.

- Early modern humans lived in the region at the same time and almost certainly met elephants, but no elephant remains are known from Late Pleistocene cave sites along the southern coast. The authors suggest people may not yet have had the technology to hunt such formidable prey.

### 4. How tracking was or was not used

The study is a reading of fossil tracks. The authors identified elephant footprints by size and shape, told hind feet (oval) from forefeet (round), and worked out direction of travel from a front bulge, toe marks and push-up mounds. They measured prints, strides and track width, recognised changes in gait and downslope sliding, read the order of events from tracks crossing ripples and other tracks, and recognised undertracks and overtracks. Hind foot length was used to estimate shoulder height and age. They refer to the elephant footfall pattern and to tracking and field guide sources (Stuart and Stuart, 1994; Brown, 1999, The Science and Art of Tracking). No trailing in the present-day sense was involved.

### 5. Tracker or observer certification

Not reported. The footprints were interpreted by the authors, who are geologists and geographers; no tracking qualification or evaluation is described.

### 6. Other technologies, skills or innovations

- First joint use of optically stimulated luminescence (OSL) and amino acid racemisation (AAR) dating in South Africa.

- An age-depth model to correct the OSL cosmic dose for sand that built up in episodes over a long time.

- Hind foot length compared with East African data (Western et al., 1983) to estimate the shoulder height and age of each animal.

### 8. Implications for tracking (Original Wisdom's observations)

_These are Original Wisdom's observations, not the authors'._

- The fossil prints are read with the same skills trackers use today: oval hind foot versus round forefoot, direction from toe marks and the front edge, push-up mounds showing slope and pressure, stride changes showing a change in gait, and the order of tracks and ripples showing what happened first.

- Using hind foot length to estimate an elephant’s shoulder height and age turns a single print into information about the individual, here showing calves less than a year old walking side by side.

- The paper’s account of why damp sand holds detail and dry, windblown sand does not, and of undertracks below deep prints, applies equally to reading fresh tracks in sand.

## Cite this summary

Lawrence, K. 2026. Summary of: Last Interglacial fossil elephant trackways dated by OSL/AAR in coastal aeolianites, Still Bay, South Africa (Roberts and Bateman et al., 2008). Original Wisdom, Tracking Literature. https://www.originalwisdom.com/tracking-literature/2008-roberts-bateman-last-interglacial-fossil-elephant/