The Pleistocene fauna of the Cape south coast revealed through ichnology at two localities

1. What they did

Along the Cape south coast of South Africa, fossil dunes hold tracks left by animals more than 100,000 years ago. Since 2007, a project has found more than 100 tracksites between Witsand and Robberg. This paper describes two large fallen blocks of rock on a remote stretch of sea cliffs east of Still Bay.

• Roberts Rock. An earlier study (Roberts and colleagues, 2008) had described elephant and “antelope” trackways on this block. The authors named it after David Roberts.
• Megafauna Rock. A block about 400 m to the west, first noticed in 2016.

Over 11 visits between 2007 and 2017, the team located the rocks with a handheld Global Positioning System (GPS) device, measured and photographed the track surfaces and the better-preserved tracks, and made 3D models of tracks on Megafauna Rock from photos (photogrammetry). They worked out which layer of the cliffs each rock came from, and so how old the tracks probably are. They then compared the tracks with those of living animals and with the extinct animals known from fossil bones.

2. What they found

Roberts Rock: a site that came and went

• In 2007 the rock looked as described in 2008. By 2009 it had split in two along a layer, exposing new surfaces with many small and medium artiodactyl tracks (two-toed hoofprints) and more elephant tracks.
• By 2016 Roberts Rock had gone into the ocean. An even larger, 12 m block had slid down next to the site. It had eight exposed surfaces, six with tracks or undertracks: elephant and artiodactyl tracks, two small three-toed bird trackways, burrow-like traces up to 10 cm across, and many smaller traces.

Megafauna Rock: four large animals

• Elephant: round or oval tracks (34 x 31 cm and 31 x 29.5 cm), the most common. The slightly oval hind foot and rounder front foot tracks leave no plausible trackmaker other than the African elephant.
• Long-horned buffalo: a large two-toed track pressed into the infill of an elephant track, wider (16 cm) than long (13.5 cm). Tracks of living Cape buffalo are longer than wide, and no living southern African hoofed animal makes tracks this shape. Animals with long, wide horns tend to make wider tracks, and the extinct long-horned buffalo had extremely long horns. Cattle farmers told the authors they can tell short-horned cattle tracks from the proportionately wider tracks of Nguni cattle.
• Rhinoceros: a single large three-toed track with rounded toes (24 cm long, 20 cm wide, 5.5 cm deep). No living southern African animal makes three-toed tracks this size except rhinoceros, although a hippopotamus’s two middle toes can sometimes merge in sand. The authors could not say whether it was a black or white rhinoceros; it is smaller than an adult white rhinoceros track.
• Giant Cape horse: a raised, symmetrical track, 27 cm long and 20 cm wide, with no split hoof and probably a “frog” at the back. That is more than double the length and width of living zebra tracks (about 10 cm long), and the only equid this large in the fossil record is the extinct giant Cape horse. The track is raised because the animal compacted the sand under its hoof, which then resisted erosion; wind can make such “pedestal tracks” on beaches today.
• Large tracks also show in cross section on the rock’s sides and as natural casts on its underside.

Firsts and context

• These are the first South African records of fossil elephant tracks (first described by Roberts and colleagues), the first probable rhinoceros track and the first giant Cape horse track of this size in southern Africa, and track evidence of the long-horned buffalo.
• By matching the layers with dated layers nearby, the rocks most likely date to Marine Isotope Stage 5e, the Last Interglacial, about 128,000 to 116,000 years ago, a warmer time with sea level 6.6 to 8 m higher than today.
• The burrows are probably too large for invertebrates and unlike root casts; the authors suggest a burrowing animal such as a golden mole, though they do not assign them with certainty.
• The team has found 14 bird tracksites on this coast, 8 of them east of Still Bay.
• All four large animals on Megafauna Rock depended on water. Tracks on many layers of one rock suggest repeated use of the area, perhaps near water such as an interdune lake like today’s Rietvlei.
• Tracksites are concentrated in zones: none were seen along stretches of more than a kilometre of seemingly suitable cliffs nearby.

3. What they concluded

• Roberts Rock and Megafauna Rock are two highlights of an area rich in trace fossils. They give a window into dune life in the Late Pleistocene and suggest an area teeming with large mammals.
• Tracks and fossil bones each have their own biases. Tracks on this coast favour large, heavy animals that left deep prints; bones come mostly from caves and rock shelters, often from hunted prey. Together they give complementary evidence of past fauna, and the tracks on Megafauna Rock independently support the bone record.
• The sites lay where the grasslands of the Palaeo-Agulhas Plain met the inland Fynbos–Strandveld–Renosterveld, so they may record animals from both habitats.
• Identifying ancient trackmakers needs knowledge of extinct species and care not to assume a trackmaker just because a track looks like that of a living species. The buffalo, rhinoceros and horse conclusions each rest on a single track.
• Using tracks to estimate population density or repeated behaviour here should be treated with caution: it would need large samples, and it is hard to tell many tracks by one animal from a few tracks by many.
• Dating Megafauna Rock directly by OSL would allow firmer conclusions about the environment at the time.
• Tracksites appear and vanish quickly as cliffs collapse; other sites have likely been lost unrecorded. Regular surveys of this coast are needed to monitor known sites and find new ones.

4. How tracking was or was not used

The whole study rests on tracks and signs identification, applied to fossil tracks. The authors identified trackmakers from track shape and size (number of toes, the gap between the two toes of hoofprints, oval versus round elephant feet, hoof “frog”, length versus width), from track measurements and photographs, and by comparing with modern tracks described in southern African tracking field guides (Walker; Liebenberg; Stuart and Stuart; Van den Heever, Mhlongo and Benadie). They also read the burrow traces and considered non-animal explanations such as root casts. No trailing was reported; the bird trackways were not measured because the rock surfaces were unstable.

5. Tracker or observer certification

Not reported. The lead author is credited with track analysis, and co-author Martin Lockley (Dinosaur Trackers Research Group, University of Colorado Denver) contributed on the ichnological content. No tracking qualification or evaluation is described. Cattle farmers’ knowledge of track width in long- and short-horned cattle came from a personal communication.

6. Other technologies, skills or innovations

• Handheld GPS to record the site locations.
• Photogrammetry: 3D models of the rhinoceros and horse tracks from 32 to 33 photos taken with a Canon PowerShot camera about 0.4 m from the surface, built in Agisoft Photoscan Professional and shown as colour height maps in CloudCompare.
• Geological fieldwork (layer thickness, colour, texture, cementation) to match each fallen block to its layer in the cliff and so to dated layers.
• Repeat visits over a decade, which revealed tracks on successive surfaces that a single visit would have missed.

7. Implications for conservation (from the paper)

• Tracksites on this coast are fragile and short-lived: Roberts Rock was lost to the sea in less than a decade, and other sites have likely been lost without being recorded.
• Repeat coastal surveys are needed to record new tracksites and follow the fate of known ones.
• Fossil tracks can add to the picture of past fauna that bones give, including extinct animals and the habitats they used, such as the grasslands of the now-submerged Palaeo-Agulhas Plain.