|
A Technical Guide for Wine Producers
|
|
RECENT ARTICLES | WYNBOER HOME
Geology of the Breede River Valley: Worcester to Robertson
Key words: fault, geology, landscape, terroir, vineyard, soil.
Abstract
The landscapes of the Breede River Valley owe their origin to folding, followed by downward slippage of rocks to the south of the Worcester Fault. This slippage created an elongated fault valley, bordered to the north and south by the Langeberg and Riviersonderend mountains, respectively. Erosion and the meandering of the early Breede River widened the valley, which in places deposited boulders, gravel and finer materials. Most vineyards are currently established on these deposits. The undulating lowlands of the valley floor, and the slopes of the valley margin nevertheless offer a variety of soils and soil parent materials, landscape positions, slope angles and aspects for possible viticultural utilisation.
Introduction
The Breede River Valley produces a sizeable percentage of South Africa's wine output from the areas around and between the towns of Worcester and Robertson, and considerable hectarages of Sauvignon blanc and Cabernet Sauvignon have recently been planted (SAWIS, 2003). In the false colour satellite image of the Worcester-Robertson sector of the Breede River Valley presented in Figure 1, areas of intensive horticultural land utilisation are coloured in shades of red. This shows that orchards and vineyards are largely confined to areas adjacent to rivers and streams, or at the confluence of drainage systems. This distribution reflects the over-riding necessity in earlier times of planting near water. However, vineyards are now being planted in an increasingly diverse range of landscape positions, and on a wider variety of soils and soil parent materials than was formerly the case. This diversification is mainly the result of recent improvements in irrigation technology which allow wine producers freedom to select sites primarily on a basis of their viticultural potential.
Western Cape landscapes reflect the varying abilities of the underlying geological structures and materials to resist erosion. Likewise, soil properties are influenced by the chemical and physical properties of their parent materials (Wooldridge, 2003). This article outlines the geological and related features of the Breede River Valley between and around the towns of Worcester and Robertson, thereby providing a geological framework for the study and demarcation of terroirs as discussed by Carey, Archer and Saayman (2002), Saayman (2003), Turner and Creasy (2003), and Wooldridge (2003). The geology and superficial deposits of that section of the valley which lies between Worcester and Tulbagh will be discussed in a companion article.
Formation of the Breede River Valley
The Breede River Valley is an example of geological control over landscape. The Worcester Fault, which defines the course of the valley, is a set of related faults which extends eastwards from the Watervalberge west of Wolseley, through Worcester to Robertson and beyond, effectively forming a lowland link between valleys in both the north-south and the west-east limbs of the Cape Fold Belt. Vertical displacement along the fault varies from about 6 000 metres to just a few metres (Gresse and Theron, 1992). The fault line is, in consequence, marked by a discontinuous series of elongated troughs, rather than by a continuous valley. Displacement is to the south, as is the case for other, near parallel faults between the Breede River Valley and the coast. Collectively, downthrow along these faults was sufficient to displace the southern tip of the African continent below present sea level, where it now forms the Agulhas Bank.
The faulting was a response to tension in the crust associated with the rifting that culminated in the separation of South Africa from South America, and the linked Falklands Plateau, during fragmentation of the southern supercontinent of Gondwana. Since the assembly of Gondwana probably commenced between 700 and 600 million years ago (mya) (Siegfried, 1999), Gondwana, and most of the rock units described in this article were already ancient when the Breede River Valley first formed. The mountains of the Cape Fold Belt would certainly have been a prominent feature of the pre-fragmentation landscape, having been created in the Triassic by mountain building episodes between 278 and 230 mya (De Beer, 1998).
Although the rifting of Gondwana began in the late Jurassic, about 154 mya, tensional faulting probably did not peak until the early Cretaceous, between 135 and 130 mya. By 130 mya, sea floor spreading was occurring in the south-east Atlantic and Natal Valley, and a marine transgression was developing in the Algoa Basin.
Erosion of the South African continental landmass began with the initial uplift and rifting of Gondwana, and peaked after break-up. Erosion was rapid at first, facilitated by a wet climate and an extensive system of rivers which transported eroded material to the continental shelf and beyond. Since the continent stood well above sea-level, gradients were steep, at least in the coastal areas. By the end of the Cretaceous, 65 mya, the horseshoe-shaped Great Escarpment had moved 50 - 200 km inland, to within a few kilometres of its present location, and the rate at which sediment was being transported and dumped offshore had decreased considerably (Partridge, 1998).
Scarp retreat did not leave behind a flat landscape. Rather, erosion carved the geological structures and materials in ways that reflected their differing abilities to resist the erosive processes. Since sandstones resisted erosion well, the sandstone-capped mountains of the Cape Fold Belt remained as prominent features of the landscape, whereas structures composed of softer formations eroded to form valleys.
After the scarp had retreated, the landscape underwent a period of severe weathering (the African Cycle). This probably began in the late Cretaceous, around 96 mya, and continued until uplift and rejuvenation occurred in the mid Miocene (25.5 mya). The climate over this extended period, which continued to be predominantly warm and wet - though with marked shifts in sea-level and drainage base levels on land - was conducive to intense chemical weathering. This weathering, through which susceptible rocks such as granite and shale were decomposed to depths of tens of metres, is known as etching (Twidale, 1988). At the conclusion of the African cycle, the superficial materials on the African Surface were reddened and structureless (apedal), as is the case for soils of warm, high rainfall areas today. The soils were also enriched in the almost chemically inert clay mineral kaolin and, in places, armoured by silcrete or ferricrete (Partridge, 1998).
Dissection of the African Surface took place following uplift and drainage rejuvenation in the mid Miocene. Further uplift followed in the late Pliocene, about 2.5 mya. In the Western Cape these uplifts led to river incision and to dissection of the African surface, but did not form extensive new erosion surfaces (Partridge and Maud, 1987). Indeed, many of the present soil surfaces continue to lie within the zone of chemical etching of the African Cycle.
That the Western Cape continues to possess a rugged topography after scarp retreat and post-Cretaceous weathering is a reflection of the contrasts in erosion resistance which exist within and between geological structures, and between rock types. Such contrasts certainly account for the varied topography of the floor of the Breede River Valley.
The floor of the Breede River Valley is not smooth. Rather, it is characterised by ridges and undulations which reflect differential erosion of the hard and soft strata of the fold structures which cross the valley floor. Hills or ridges composed of weathered sandstone or sandy shale are therefore found in close proximity to the alluvium and gravels of past and present watercourses. These tended to meander around and sometimes through the harder structures. Some of the valley deposits, such as the Enon Conglomerates, are of considerable age. More recent alluvial deposits, which vary from boulders to fine sediments, depending on source area and degree of sorting or reworking, have also contributed to the smoothing, and to the diversity, of the soil parent materials of the Breede River Valley floor. The Worcester Fault is covered by these superficial deposits, and is nowhere visible at the surface. Neither is an identifiable fault scarp visible.
Principal rock groups of the Breede River Valley
Beds of sedimentary rock which are related, or have some other affinity, may be mapped as a single formation. Formations may, in turn, be assembled into groups. The distribution of groups in the Breede River Valley around Worcester and Robertson is shown in Figure 2.
The oldest rocks belong to the late Precambrian (older than 542 million years) Malmesbury Group. These heat- and pressure-altered and locally carbonate-rich sedimentary rocks form a narrow band to the north of the Worcester Fault (Figure 2a) along the southern footslopes of the Langeberg range. On upper slopes these clay-rich rocks often weather to form deep stony soils (Figure 3) in which the stone readily fractures along cleavage planes to form fine shards. Between Worcester and Vink River the Malmesbury metasediments are intruded by the Robertson Granite, from which potassium-rich feldspar fragments are released into the regolith (the layer of loose material which covers the bedrock). Fragments of granite also occur in association with Malmesbury rocks to the east of Worcester.
The second oldest rocks in the Breede River Valley form the heights of the Langeberg range to the north, and also of the Riviersonderend range to the south (Figure 2b). These rocks, which belong to the late Ordovician to Silurian (roughly 440 mya) Table Mountain Group, are mainly formed of erosion-resistant quartz-rich sandstones. The contact between the Table Mountain Group sediments and the underlying Malmesbury metasediments and granites is unconformable, reflecting the spread of a new sedimentary sequence (the Cape Supergroup) across an older, eroded, pre-Cape, mainly Cambrian (542-490 mya), landscape. Eventually subsidence created an elongated trough - the Cape Basin - which at times extended south from Vanrhynsdorp to the present south coastline and from west of Ceres eastward through Port Elizabeth. Sediments carried by rivers from high land to the west and north began to fill this trough in the early Ordovician, about 490 mya.
The Table Mountain Group grades upwards into the alternating sequence of mainly shale and sandstone formations which form the lower to mid Devonian Bokkeveld Group (Figure 2c). Bokkeveld sediments are of marine origin and were laid down at a time when the western and southern Cape lay beneath a sea which was at times deeper than at others. Deep, quiet conditions are usually required for deposition of the fine grained materials which, with increasing pressure, form mudstones, shales, slates and phyllites, whereas high energy environments, such as those where strong currents and wave action cause reworking of sediments and removal of fine material, are associated with sandstones. Landscapes which are underlain by tilted formations of the Bokkeveld Group tend to be subdued and to consist of an alternating sequence of hills and valleys, corresponding to outcrops of the constituent sandstone and shale bands. In arid areas, water from marine shales may be brackish.
In contrast to the darker-coloured shales of the Bokkeveld Group, the overlying upper Devonian to Carboniferous Witteberg Group (Figure 2d) is dominated by relatively mature, and therefore lighter coloured, sandstones. South-east of Worcester the Witteberg crops out in the limbs of an eroded fold.
After the Witteberg sandstones had been laid down, deposition in the Cape Trough came to a close as the axis of subsidence moved north into what is now the Karoo. The life-span of this basin, from early Ordovician (490-439 mya) to early Carboniferous (363-290 mya) was about 150 million years (Rust, 1973). In Permian to Triassic times the sediments of the Cape Basin were deformed by compression from the south and west, forming the Cape Fold Mountains.
After the Witteberg sandstones were deposited, the whole of Southern Africa was covered by the Dwyka ice sheet. In the Worcester-Robertson area glacial shales and sandstones of the late Carboniferous to early Permian (about 290 mya) Dwyka Group, and post glacial shales, siltstones and mudstones of the Permian (290-245 mya) Ecca Group, are preserved in fold cores as shown in Figure 2e.
Outcrops of the Jurassic Enon Formation (Uitenhage Group) occur at intervals along the valley floor close to the Worcester Fault (Figure 2f). These deposits, which consist of reddish conglomerate containing lenses of mudstone and sandstone, accumulated beneath the eroding scarp of the Worcester Fault about 154 mya. They were reworked by rivers and mainly accumulated in the deeper fault basins. Their distribution was probably more extensive in the past than is the case today.
Collectively, the rock groups shown in Figures 2a to 2f now outcrop as shown in Figure 2g.
Superficial deposits are unconsolidated accumulations of mineral material derived from rock by weathering and erosion. They compose the regolith. The upper metre or so of regolith is most readily accessible to vine roots (Figure 3).
The extent to which superficial deposits resemble their source rocks depends on the physical (attrition, sorting, mixing) and chemical (action of organic acids, oxidation, reduction) processes to which they are subjected during transport (if any) and before and after deposition. For the purposes of this article the superficial deposits in the Breede River Valley have been subdivided into seven broad categories as shown in Figure 4.
Of the superficial deposits, scree (talus) usually takes the form of a wedge-shaped accumulation of gravity-transported, poorly sorted rock fragments below a steep rock face from which material is intermittently shed as a result of weathering and erosion. The fragments (clasts) may vary from boulders down to granules 2 to 4 mm in diameter. Toward the base of scree slopes a matrix of gritty, coarse sand becomes increasingly prominent.
In waterways, clasts tend to become rounded. Extensive deposits of gravel to cobble sized clasts may occur in the form of outwash fans, notably in places where energetic, high-gradient mountain streams debauch onto, and spread across, the valley floor. These fan deposits may be reworked to form gravel beds or terraces. High-level gravels sometimes grade into scree.
In the Breede River Valley, alluvial deposits are generally associated with the larger tributaries and watercourses. In places, old alluvium-filled channels which were cut at times when drainage base levels were lower than at present, are buried beneath more recent deposits. Where Table Mountain sandstone predominates in the source area the alluvium consists mainly of quartzitic sand, whereas alluvial deposits derived from Malmesbury, Bokkeveld and Karoo sedimentary rocks tend to be silty. Alluvium may grade almost imperceptibly into terrace or pediment (concave erosion surfaces on flanks of hills) gravel, of which the pebbles and cobbles mostly consist of angular to rounded vein quartz and quartzite. East of Worcester the alluvium is gravelly. Where present, the thickness of boulder deposits tends to increase with height above the tributary. To the west of Worcester the remains of a gravel terrace, about 15 m above present river level, grades into the younger gravels of the Hex River fan which lies to the south and east of the town. These deposits mostly consist of well-rounded boulders and cobbles of weathering-resistant rocks such as sandstone and quartzite. Older gravels are locally cemented into a matrix that may be calcareous, iron-rich or silicaceous.
Loam and sandy loam soil parent materials probably date back to Tertiary times, and formed by weathering of fine grained shale or phyllitic bedrock, notably from the Malmesbury, Bokkeveld, Dwyka and Enon groups. Soils which developed by weathering of Dwyka sediments tend to be slightly calcareous. In contrast, most light grey to pale red sandy soils are probable of Quaternary age, and mainly consist of the weathering products of Table Mountain sandstone. They often occur in valleys and may form deep deposits between the scree slope and the alluvium of the valley floor. Deposits of building-quality sand occur along the courses of most of the rivers which drain the quartzitic mountain ranges, as between Rawsonville and Robertson. Other sand deposits have accumulated through wind action.
Soils
Soils are generated in the regolith through the process of pedogenesis, which entails interactions between mineral constituents, climate, organisms and topography over time. Lambrechts (1979) recognised ten soil associations in the Fynbos Region, using soil parent material, rockiness, soil depth and acidity as differentiating criteria. Catenas - sequences of soil forms which succeed one another in a downslope direction - may also reflect rock type.
In low rainfall areas where iron oxides tend to be retained, the schematics of Lambrechts (1979) suggest that a catena containing the soil forms (MacVicar et al.,1977): Hutton 22, Clovelly 22, Constantia 12 , La Motte 12 and Champagne 10 will develop on quartzitic fold mountain ranges. On pediments and valley floors, catenas are nevertheless likely to consist of: Clovelly 21/22, Constantia 11/12, La Motte 11/12, Fernwood 32 and Champagne 11. In contrast, Glenrosa, Swartland, Sterkspruit, Escourt and Kroonstad catenas are likely on heavy-textured or duplex soils derived from granites and shales.
Occurrence of rock groups and superficial materials in wards
The occurrences of rock groups and their major lithologies, and of the dominant superficial materials in the wards of the Worcester-Robertson section of the Breede River Valley are summarised in Table 1.
Summary
The Breede River Valley follows the line of an ancient geological fault which, in the Worcester area, cuts across the zone of intersection of the north-south and west-east limbs of the Cape Fold Belt. The valley floor has a varied topography due to differences in erosion resistance between shale and sandstone formations, and between folded structures. The undulating lowlands of the valley floor, together with the hill slopes of the valley margin offer a considerable variety of soil parent materials, landscape positions, slope angles and aspects. The viticultural potential of these sites requires critical appraisal.
Acknowledgements
The research project on which this article is based was partly funded by the Agricultural Research Council and by Winetech. This contribution is gratefully acknowledged. Portions of this work include intellectual property of the Council for Geoscience and are used herein by permission. Copyright and all rights are reserved by the said Council.
For further information contact John Wooldridge at Nietvoorbij on (021) 809-3330.
E-mail: wooldridgej@arc.agric.za.
References
CAREY, V.A., ARCHER, E. & SAAYMAN, D., 2002. Natural terroir units: what are they? How can they help the wine farmer? WineLand (February), 86-88.
DE BEER, C.H., 1998. Structure of the Cape Fold Belt in the Ceres Arc. Bulletin no. 123 of the Geological Survey of South Africa. Council for Geoscience, Pretoria, South Africa.
GRESSE, P.G., 1988. 1:250 000 geological series mapsheet 3319, Worcester. Geological Survey of South Africa, Pretoria.
GRESSE, P.G. & THERON, J.N., 1992. The geology of the Worcester area. Explanation of Sheet 3319. Department of Mineral and Energy Affairs. Geological Survey. Government Printer, Pretoria.
HARLAND, W.B., ARMSTRONG, R.L., COX, A.V., CRAIG, L.E., SMITH, A.G. & SMITH, D.G., 1990. A geological time scale 1989. Cambridge University Press, Cambridge.
LAMBRECHTS, J.J.N., 1979. Geology, geomorphology and soils. In J. Day, W.R. Siegfried, G.N. Louw & M.L. Jarman (eds). Fynbos ecology: a preliminary synthesis. South African National Science Progress Report 40, 16-26.
MACVICAR, C.N., LOXTON, R.F., LAMBRECHTS, J.J.N., LE ROUX, J., DE VILLIERS, J.M., VERSTER, E., MERRYWEATHER, F.R., VAN ROOYEN, T.H. & VON M. HARMSE, H.J., 1977. Soil Classification. A binomial system for South Africa. Department of Agricultural Technical Services, Pretoria.
PARTRIDGE, T.C., 1998. Of diamonds, dinosaurs, and diastrophism: 150 million years of landscape evolution in southern Africa. S. Afr. J. Geol. 101, 167-184.
PARTRIDGE, T.C. & MAUD, R.R., 1987. Geomorphic evolution of southern Africa since the Mesozoic. S. Afr. J. Geol. 90, 179-208.
RUST, I.C., 1973. The evolution of the paleozoic Cape Basin, southern margin of Africa. In: A.E.M. Nairn & F.G. Stehli (eds). The ocean basins and margins, vol. 1. The South Atlantic, Plenum Press, New York. pp. 247-276.
SAAYMAN, D., 2003. Practical aspects of viticultural zoning in South Africa. WineLand (April), 61-63.
SAWIS, 2003. South African Wine Industry Statistics No. 27. SAWIS, Paarl, South Africa.
SIEGFRIED, H.P., 1999. The Malmesbury batholith and its relationship to granitic plutons in the Swartland Tectonic Domain. Council for Geoscience Bulletin 125, Council for Geoscience, Pretoria, South Africa.
THERON, K., 1998. Wine of origin. Wine and Spirit Board, Stellenbosch, South Africa.
TURNER, P.T. & CREASY, G.L., 2003. Terroir. Competing definitions and applications. Wine Industry J. 18(6), 48-55.
TWIDALE, C.R., 1988. The missing link: planation surfaces and etch forms in southern Africa. In G.F. Dardis & B.P. Moon. Geomorphological studies in southern Africa. A.A. Balkema, Rotterdam, pp. 31-46.
WOOLDRIDGE, J., 2003. Geology and terroir in the Western Cape winelands. Wynland (December), 85-87.
|
|
Wynboer is incorporated in WineLand, magazine of the SA wine producers.
Visit our sister sites:
|
| UP |
COPYRIGHT (C) 2000 WineLand |