1
Palaeontological Heritage Impact assessment for N2 Upgrade between Grahamstown and Fish River.
Prepared for: Coastal & Environmental Services
67 African Street Grahamstown
Compiled by: Dr Robert Gess
Rob Gess Consulting, Box 40, Bathurst, 6166
Research Associate of the Albany Museum, Somerset Street, Grahamstown
September 2013
2
Contents:
page 1: Title
page 2: Contents
page 3: Background
page 3: Geology
page 9: Palaeontology page 10: Site Visit
page 23: Conclusions and Recommendations.
3
Background
Coastal and Environmental Services have been appointed to carry out a full Environmental Impact Assessment for upgrading of the N2 from north of Grahamstown to the east of the Fish River pass. Rob Gess Consulting was contracted to conduct a phase one Palaeontological Impact Assessment for this proposed development.
Geology
The area intended for development overlies strata of the upper portion of the Cape Supergroup and lowermost portion of the unconformably overlying Karoo Supergroup. In addition, portions of the Cape Supergroup rocks are capped by relict patches of Silcrete formed as a product of deep leaching during the Cretaceous.
Cape Supergroup rocks represent sediments deposited in the Agulhas Sea, which had opened to the south of the current southern African landmass, in response to early rifting between Africa and South America during the Ordivician.
The Witteberg Group is the uppermost of three subdivisions of the Cape Supergroup and was laid down during the Late Devonian and early Carboniferous.
Within the study area the Lake Mentz subgroup of the Witteberg group is exposed.
The Lake Menz Subgroup (Witteberg Group, Cape Supergroup) consists of four subunits (the Witpoort, Kweekvlei, Floriskraal and Waaipoort formations.
The Witpoort Formation, the basal member, is characterised by thick light coloured quartzites, consolidated from mature quartz sand accumulated along a sandy shoreline, under linear barrier-island conditions. These form distinctive, resilient ridges that may be traced along the full length of the Cape Fold Belt. Thin black shale layers and lenses within this formation include deposits derived from muds carried by river systems into coastal estuarine lagoons. Much of the road is routed along the top of two great arched anticlinal folds of Witpoort Formation quartzite that have proved more resistant to weathering than the overlying mudstones.
A sudden rise in relative sea level, corresponding to the Tournasion transgression at the end of the Devonian Period, terminated the sandy shoreline deposition of the Witpoort Formation, which is overlain by fine muddy sediments of the Lower Carboniferous Kweekvlei Formation. These shales generally weather to form lowered topography breaking away from the Witpoort Formation ridges.
As the basin gradually shallowed, delta front and shoreline sands of the Floriskraal Formation encroached on the underlying silts and muds of the Kweekvlei Formation.
Flouriskraal strata are in turn overlain by the Waaipoort mud and siltstone unit. This may represent a lagoonal environment behind and adjoining the prograding Floriskraal shoreline sands. It has been suggested that freshwater and lacustrine indicators, suggest that large lake- like environments may have developed on the continental margin. Localised glacially related 4
deposits including the Dirkskraal Formation, assigned to the Kommadagga Subgroup, are found overlying Lake Mentz subgroup in some parts of the Eastern Cape.
The Witteberg is terminated by an unconformity, or gap in the geological record, of possibly 30 million years, after which the massive diamictites of the Dwyka Group (Karoo Supergroup) were deposited, indicating that an ice age had passed.
This would be in keeping with palaeomagnetic interpretations which suggest that, at the time of deposition of the Witteberg Group, what is now South Africa was within the Antarctic circle.
The strata of the Karoo Supergroup were deposited within the Karoo sedimentary Basin, which resulted from shortening and thickening of the southern margin of Africa, with coeval folding and uplift of the Cape Supergroup strata along its southern margin. Lowermost Karoo strata of the Dwyka and lower Ecca Groups were affected by folding in the vicinity of the Cape Fold Belt.
The Dwyka Group (Karoo Supergroup), particularly here in the south of the basin consists almost exclusively of diamictite known as the Dwyka tillite. This is a distinctive rock type which, when freshly exposed, consists of a hard fine-grained blueish-black matrix in which abundant roughly shaped clasts are embedded. These vary greatly in both lithology and size. During the formation of the Dwyka, beginning in the late Carboniferous, southern Africa drifted over the south pole, whilst simultaneously, the world was experiencing a cold episode. Glaciers flowing into the flooded Karoo basin broke up, melted and discharged a mixture of finely ground rock flour and rough chunks of rock. These formed the matrix and clasts of the Dwyka tillite.
Early in the Permian period the ice sheets retreated and fine muds were washed into the Karoo Basin, forming the Prince Alfred and Whitehill formations of the lower Ecca Group. These interfinger, at first, with the last tillites.
Subsequent deposition of the Collingham, Ripon and Fort Brown and Waterford formations of the Ecca Group resulted from sediment carried into the Ecca Lake by rivers draining the recently upthrust Cape Mountains. These rivers formed deltas where they flowed into the Ecca Lake. Proximally the deltas tended to be sandy. Mud accumulating on the more distal front of the deltas periodically slumped and cascaded down into deep water, spreading out and depositing large layered fan shaped turbidite deposits.
As the Ecca Lake silted up a subaerial (exposed) shoreline began to develop, initially in the south east of the basin. The lake steadily shrank towards the centre of the basin, leaving behind flat silty plains across which long rivers meandered from the Cape Mountains towards the much reduced lake. Sands were deposited along the river channels whereas periodic flooding deposited muds on the broad flood plains. These in time came to form the interbedded sandstones and mudstones of the Koonap, Middleton and Balfour formations of the Adelaide Subgroup, Beaufort Group. 5
During the Cretaceous and early Tertiary Periods much of Africa was weathered down to a number of level horizons collectively known as the African Surface. The area in the vicinity of Grahamstown was reduced to a flat plain close to sea level, remnants of which are referred to as the Grahamstown Peneplane. During the Tertiary, mudstones, shales and diamictites were leached to considerable depth, transforming them into soft white kaolin clay. Silica, iron and magnesium from these rocks was carried in solution by groundwater and deposited near the ground surface due to steady evaporation of mineral rich waters. This lead to the formation of a hard mineralised capping layer, often consisting of silicified soil. Resultant silcretes are referred to as the Grahamstown Formation. Though occasional occurrences of root and stem impressions have been recorded from the Grahamstown Formation it is generally considered unfossiliferous.
With subsequent reduction of the relative sea level, deep valleys have carved back from the retreating coastline, cutting deep valleys and catchment areas into the African Surface.
6
Figure 1. Geological map of the study area based on geological survey data overlain on topography, with positions of the road route, borrow pits, quarries and info points overlain.
7
Figure 2: Stratigraphic column of the Cape Supergroup modified after Theron and Thamm (1990) following Cotter (2000). Red line indicates strata impacted by the development. 8
Figure 3: Karoo stratigraphy and biostratigraphy (after Smith et al., 2012). Red line indicates stratigraphic interval impacted by proposed development. 9
Palaeontology
The stratigraphically lowest Witteberg Group strata present belong to the Late Devonian (Famennian) Witpoort Formation ( Lake Mentz Subgroup, Witteberg Group, Cape Supergroup) This largely quatzitic unit represents mature sandy strata deposited along a linear barrier island type coast. Particularly around Grahamstown black shale lenses, interpreted as estuarine deposits preserved during brief transgressive events, have proved remarkably fossiliferous. A series of lenses at Waterloo Farm, to the south of Grahamstown, have provided southern Africa’s most important Late Devonian locality, which has yielded at least 20 taxa of fossil fish (including jawless fish (Agnatha), armoured fish (Placodermi), spiny sharks (Acanthodii), sharks (Chondrichthyes), ray finned fish (Actinopterygii) and lobe finned fishes (Sarcopterygii) including Coelacanths (Actinistia), lungfish (Dipnoi) and Osteolepiformes. Dozens of plant and algal taxa have been collected, as well as remains of giant eurypterids and other arthropods such as scorpions. These latter fossils represent the oldest known terrestrial animal remains from Gondwana. Abundant trace fossils have also been collected. The top of the Witpoort Formation coincides with the end of the Devonian and is similar in age to the end-Devonian extinction event. Witpoort Formation quartzites have yielded a range of plant stem taxa and trace fossils. Lag deposits of bone have not, as yet, been discovered, but may be expected.
The remainder of the Lake Mentz Subgroup, the Kweekvlei Formation, Floriskraal Formation and Waaipoort Formation are all Early Carboniferous in age, the 360 myo Devonian- Carboniferous boundary (and the Hangenberg Extinction event) having been passed at the top of the Witpoort Formation.
The fine muddy sediments of the Kweekvlei Formation (Lake Mentz Subgroup, Witteberg Group) are increasingly silty upwards but have, as yet, yielded only plant fragments and sparse ichnofossils. The overlying Floriskraal Formation (Lake Mentz Subgroup, Witteberg Group) is similarly characterised by sparse plant fossils and trace fossils, though acanthodian fish remains have been noted from its upper strata.
From near Port Elizabeth towards the Western Cape, thin carbon-rich lenses, as well as phosphatic and calcitic nodules within the Waaipoort Formation (Lake Mentz Subgroup, Witteberg Group) have yielded interesting, though scarce, plant fossils. More significantly, an extensive fish fauna has been described, consisting largely of primitive palaeoniscoids (ray finned fish) with, in addition, acanthodians and sharks. A eurypterid and bivalves have also been collected. Placoderm fish remains, which dominate Witpoort Formation faunal remains are entirely absent from these strata, reflecting their world wide extinction during the Hangenberg (Second Great) Extinction Event at the end of the Devonian.
An important Waaipoort Formation Locality in the Eastern Cape is on the farm Schiethoogte near the Darlington Dam (formerly Lake Mentz), north of Port Elizabeth. Here, two unique mass mortality horizons, a few centimetres apart, have been studied, containing hundreds of whole-bodied palaeoniscoid fishes of a variety of taxa, embedded in a dark silty sandstone.
Within the study area fossils are not known from the Dwyka Group (Karoo Supergroup). 10
Probably due to the lack of good outcrop in the Eastern Cape, body fossils have as yet not been found in rocks of the Whitehill Formation (lower Ecca Group, Karoo Supergroup), though invertebrate trace fossils are known from the top of the Ecca Pass. In other parts of the country the Whitehill Formation has yielded some exquisite fossils. These include Africa’s earliest known reptile, the aquatic Mesosaurus, early crustaceans, and scarce but beautifully preserved ray-finned fish.
Plant fossils are found in strata of the Ripon Formation (Ecca Group, Karoo Supergroup) and Fort Brown Formation (Ecca Group, Karoo Supergroup), for example along the cuttings of Ecca Pass, which is the type locality of the Ecca Group. These belong to the earliest appearance of the Glossopteris fauna – named after Glossopteris, an early genus of seed plant that may ultimately have included the ancestors of flowering plants. Fish trails have also been found in the vicinity of the Ecca Pass, underlying the potential that fish fossils could be recovered from these horizons. A fish fossil was reliably reported from Ecca strata near Fort Brown, but it was destroyed by construction of a roadside drainage ditch before it was collected. Actinopterygian fish have been collected from these strata in other parts of the country.
Site visit
Road route
The road route between Grahamstown and info point 1 follows the top of a resistant quartzite ridge, which is capped with silcrete closer to Grahamstown. Along this stretch there is virtually no outcrop and it does not seem likely that any disturbance of outcrop will occur.
Between info point 1 and info point 3 the route winds its way from the top of one anticlinal quartzite fold ridge to the top of a second one slightly to the north, passing through a slight decline caused by differential weathering of softer Kweekvlei Formation strata. A number of road cuttings along this route expose Witpoort Formation quartzites. These display stacked packages of fine to coarsely laminated channel and sand bar derived cross beds of mature sand. In some places shallow-water ripple surfaces were observed.
At info point 2, where the curve of the road is to be cut back in order to straighten its course, the existing roadcutting exhibits a layer of grey-weathering dark carbonaceous mudstone. This is finely layered and reminiscent of some of the fossiliferous mudstones at the productive Waterloo Farm site. Splitting of some crumbly outcrop thereof revealed the presence fossilised organic fragments, confirming the potential palaeontological importance of this layer (Figures 4 – 6).
At info point 4 a roadcutting exposes the overlying Dwyka Group diamictite consisting of a fine grained blue black matrix containing irregular angular clasts (Figure 7).
11
Figure 4: Road cutting at info point 2 with band of grey weathering carbonaceous mudstone.
Figure 5: Close up of laminated carbonaceous mudstone at info point 2.
Figure 6: Fossil fragment in carbonaceous mudstone at info point 2. 12
Figure 7: Dwyka tillite, a glacial rock consisting of a fine matrix comprised of glacial flour containing angular clasts – exposed along a cutting at info point 4.
Between info points 5 and 6 the N2, extending eastwards beyond the resistant ridges of folded Witteberg rocks, makes its way through the steep valley of the Great Fish River, which has carved its course into the less resistant strata of the lower Ecca Group.
It initially exposes sandstone dominated strata of the Ripon Formation, interbedded with fine grained dark green mudstones. As the Fort Brown Formation is encountered these mudstones become the dominant lithology, with sandstones forming much finer layers.
Particularly on fresher outcrop large scale ripple surfaces are common as well as a number of trace fossil forms, ranging from possible fish trail impressions to possible invertebrate traces.
Figure 8: Fort Brown Formation mudstones exposed along the Fish River Pass. 13
Figure 9: Inclined sheets of large scale water ripple impressions in the Fort Brown Formation.
Figure 10: Trace fossils in the Fort Brown Formation.
Figure 11: Possible fish trails in the Fort Brown Formation.
14
Alternate route
A short proposed alternative route runs along top of Fraser’s camp ridge and is underlain by extremely weathered Witteberg quartzite with patchy cover of silcrete. Here there is virtually no likelihood of palaeontological material being disturbed.
Figure 12: Boulders of silcrete disturbed along the proposed Frasers Camp alternative route.
15
Borrowpits
Three of the proposed borrow pits are located along the Ecca Pass. From north to south these are borrow pit 6, borrowpit 7 and borrow pit 5 (see Figure 1).
Borrow Pit 6 consists of an existing, partially flooded, borrow pit accessed off the main road. It is comprised of dark green mudstones of the lowermost Fort Brown Formation. These are highly weathered and crumble into elongate shards of the ‘pencil weathering’ type. Large concretions are weathered from the mudstones. As a rule the material exposed on surface is too weathered to allow proper assessment. About 500 m to the north of this borrowpit, in former decades, very fine fossilised fish trails were exposed in thin interbedded sandy layers to the left of the road. This site was subsequently destroyed during construction of a game fence.
Figure 13: View of borrow pit 6.
Figure: 14: Pencil weathering in lower Fort Brown Formation mudstones at borrow pit 6. 16
Borrow Pit 7 is situated in middle of Rippon Formation. It is also a partially flooded existing borrow pit, which contains crumbly olive green weathered mudstone with some more resistant layers. Pushed to one side was observed a Swiss Roll like structure of sandy and muddy layers created during a turbiditic subaqueous soft sediment avalanche.
Figure 15: View of borrowpit 7.
Figure 16: Swiss roll like structure of sandy and muddy layers (at borrow pit 7) formed in an underwater avalanche.
17
Borrow Pit 5 is largely cut into finely bedded strata of the Whitehill Formation, though at the back it intercepts the upper contact with the Collingham Formation mudstones. A wide range of trace fossils were noted towards the top of the Whitehill Formation. Fossil plant fragments were found to be common in the lower Collingham Formation.
Figure 17: Fine white and pink weathering strata of the Whitehill Formation in Borrow pit 5.
Figure 18: Contact between the Whitehill Formation (front) and Collingham Formation (back) in borrow pit 5. 18
The remainder of the borrow pits are within a short distance of the N2. These are discussed from west to east.
Quarry 7 is situated alongside the N2, near its highest elevation, adjacent to info point 2 (Fig.1). As noted above the roadcutting at this point is to be cut back to reduce its curvature. As this is a hilltop of relatively clean quartzite the intention is to mine it as a source of hard rock aggregate. Quartzite in the road cutting is nicely exposed and consists of fairly coarsely bedded strata. Interbedded with these is a grey muddy layer reminiscent of one of the fossiliferous strata at the Waterloo Farm site. An examination of loose material revealed the presence of small fragments of unidentified fossilised matter.
Figure 19: Coarsely bedded Witpoort Formation quartzites exposed in a roadcutting at the proposed site of Quarry 7.
Figure 20: medium grained carbon rich mudstone exposed as a layer in roadcutting adjacent to proposed quarry 7. 19
Borrow Pit 10 consists of crumbly weathered mudstones and thin sandstones of the Kweekvlei Formation. In places these are finely bedded with alternating paler mudstones and darker reddish sandstones. Some thicker sandstones up to 10cm thick occur in a band across the middle in of the outcrop.
Figure 21: Rhythmically deposited fine layers of pale mudstone and reddish sandstone ar quarry 7.
Figure 22: Coarser sandstone layers around the centre of the outcrop at quarry 7.
20
Quarry site 6 is also intended to exploit relatively clean quartzites of the Witpoort Formation. It consists of a small hill that has not previously been exploited. As a result there is no fresh outcrop that can be examined to ascertain whether or not mudstone is present. It is normal in such well weathered outcrops for only the relatively resilient clean quartzite’s to outcrop on surface.
Figure 23: Quartzitic hillslope at the proposed site of Quarry 6
Borrow Pit 18 is a proposed source of soft aggregate in a previously worked pit. This consists of leached mudstone in the form of pink and white clay mixed with lumpy, semi- consolidated silcritised soil.
Figure 24: Pink and white clay mixed with lumpy, semi-consolidated silcritised soil at borrow pit 18. 21
Borrow pit 17 is a greenfields site close to the west bank of the Fish River. It is hard to evaluate as it has no fresh outcrop on soil and the dark greenish mudstone outcropping on surface is too finely crumbly to allow for palaeontological exploration.
Figure 25: View of crumbly dark mudstones at the proposed site of borrow pit 17.
Borrow pit 15 is situated on the eastern bank of the Fish River and is comprised of grey green crumbly mudstone. Where fresh material is exposed in a small stream bed which cuts into the borrow pit area, it is found to consist of fairly well-bedded dark green shale exhibiting mats of rather poorly preserved fossil plant stems.
Figure 26: Fairly fresh dark mudstones exposed in a stream bed at the borrow pit 15 site.
22
Figure 27: Plant stems exposed as mats in stream bed at borrow pit 15 site.
Quarry 1 is an existing quarry cutting into a thick layer of extremely hard, grey massive sandstone.
Figure 28: Extremely hard sandstone layer exposed in existing excavation at Quarry 1 site.
23
Conclusions and Recommendations
1. The area of highest potential palaeontological sensitivity is the roadcutting and proposed quarry site (Quarry 7) at info point 2. Cutting back of this hill top is likely to lead to exposure and destruction of irreplaceable palaeontological heritage. It is recommended that palaeontological monitoring of this process is required. A meeting should be arranged between the engineers and the palaeontologist to establish a modus operandi. Experience suggests that initial removal of carbonaceous material to a designated on site area for regular examination by the palaeontologist may be the most feasible strategy. This has the potential to make important contributions to our understanding of the crucial Late Devonian Period.
2. Quarry 6 is also to exploit Witpoort Formation quartzites although it is currently impossible to establish whether any shale or mudstones are present due to the extremely weathered nature of the outcrop. It is recommended that a palaeontologist be required to visit this quarry at intervals during its utilisation.
3. All roadcuttings between points 1 and 3 should be visited by a palaeontologist at least once after being cut to evaluate fresh outcrop.
4. Borrow Pit 5 is intended to exploit material of the important Whitehill Formation. It is recommended that this borrowpit is visited by a palaeontologist during its exploitation to evaluate fresh outcrop.
5. Borrow pits 5,7, 15 and 17, as well as roadcuttings between info points 5 and 6 should be evaluated by a palaeontologist at the end of works, prior to any rehabilitation.
24
References
Algeo,T.J.,Scheckler, E. and Maynard, J.B. (2001). Effects of the Middle to Late Devonian Spread of Vascular Land Plants on Weathering Regimes, Marine Biotas and Global Climate.in eds Gensel, P.G. and Edwards, D. Plants Invade the Land: Evolutionary and Environmental Perspectives Columbia University Press, New York.213-236
Anderson, H.M., Hiller, N. and Gess, R.W. (1995). Archaeopteris (Progymnospermopsida) from the Devonian of southern Africa. Botanical Journal of the Linnean Society 117: 305- 320.
Anderson, M.E., Hiller, N. and Gess, R.W. (1994). The first Bothriolepis-associated Devonian fish fauna from Africa. South African Journal of Science 90: 397-403.
Anderson, M.E., Long, J.A., Gess, R.W. and Hiller, N. (1999). An unusual new fossil shark (Pisces: Chondrichthyes) from the Late Devonian of South Africa. Records of the Western Australian Museum 57: 151-156.
Cotter,E.(2000). Depositional setting and cyclic development of the lower part of the Witteberg Group (Mid- to Upper-Devonian), Cape Supergroup, western Cape, South Africa. South African Journal of Geology 103: 1-14
Council for Geosciences (Geological Survey) 1:250 000 Geological Maps: Eastern Cape: 3326 – Grahamstown
Du Toit, A.L. (1954). The Geology of South Africa. Oliver & Boyd, Edenburgh, 611p.
Evans, F.J., (1997). Palaeobiology of Early Carboniferous fishes and contemporary lacustrine biota of the Waaipoort Formation (Witteberg Group), South Africa. Masters degree thesis, University of Stellenbosch.
Gess, R.W. and Hiller, N. (1995b). Late Devonian charophytes from the Witteberg Group, South Africa. Review of Palaeobotany and Palynology 89: 417-428.
Gess, R.W. and Hiller, N. (1995a). A preliminary catalogue of fossil algal, plant, arthropod, and fish remains from a Late Devonian black shale near Grahamstown, South Africa. Annals of the Cape Provincial Museums (Natural History) 19: 225-304.
Gess, R.W. (2001). A new species of Diplacanthus from the Late Devonian (Famennian) of South Africa. Annales de Paléontologie 87: 49-60
Gess, R.W. (2003). The palaeoecology of a coastal lagoon of the Witpoort Formation (Upper Devonian, Famennian). M.Sc. thesis, University of Fort Hare.
Gess, R.W., Coates, M.I. & Rubidge, B.S. (2006). A lamprey from the Devonian period of South Africa. Nature 443: 981-984. 25
Gess, R.W. (2011). High latitude Gondwanan Famennian biodiversity patterns – evidence from the South African Witpoort formation (Cape Supergroup, Witteberg Group). PhD thesis, University of the Witwatersrand.
Hiller, N and Gess, R.W. (1996). Marine algal remains from the Upper Devonian of South Africa. Review of Palaeobotany and Palynology 91: 143-149.
Hiller, N. and Taylor, F.F. (1992). Late Devonian shore line changes: an analysis of Witteberg Group stratigraphy in the Grahamstown area. South African Journal of Geology 95: 203-212.
Long, J.A., Anderson, M.E., Gess, R.W. and Hiller, N. (1997). New placoderm fishes from the Late Devonian of South Africa, Journal of Vertebrate Palaeontology 17: 253-268.
McCarthy, T. and Rubidge, B. 2005. The story of Earth & Life; A southern African perspective on a 4.6 billion-year journey. Struik. 344p.
Theron,J.N. and Thamm, A.G. (1990). Stratigraphy and sedimentology of the Cape Supergroup in the Western Cape. Guidebook PR2, Geocongress ’90 64pp. Geological Society of South Africa.