Floodplain Formation and Environmental Change in the Cape Region, South Africa

During the last decades significant rainfall events resulted in widespread and serious flooding of large areas of the Little Karoo, Cape Province (Fig. 1). Consequences of the heavy rainfalls were flooding and dam failure in several river basins and the onshore of the southwest Cape region. In the course of the so-called "Laingsburg flood" on January 1981, large parts of the city of Laingsburg in the Great Karoo had been destroyed and more than 100 people died. The flood had been a direct consequence of long-lasting and wide-ranging rainfall in the Karoo with a 48-hour total of more than 200 mm. However, severe floods were also be triggered by less intensive rainfall, if they followed a period of wet conditions.

The Little Karoo is characterised as a dry, semi-arid region. The area is drained by the Gourits River and its main tributaries Groot and Olifants River. The valleys and their floodplains are up to 200-300 m wide and the floodplain deposits are several meters thick sediment layers above the valley floor (Fig. 2). They are clearly distinguishable from Pleistocene gravel terraces, which are also morphologically differentiable due to altitude differences of partly over 10 m.

High sediment loads during past flood events caused the accumulation of floodplains up to several meters. The floodplain deposits of the Gourits River and its tributaries are generally 5 – 8 m thick sediment layers above the channel. They are mainly composed of sands and silts that are partly laminated and are infrequently interbedded with gravel layers or lenses. Layers of loamy to silty sands, silts, and loams appear dark, indicating relatively higher soil moisture and/or contents of organic matter (Fig. 3). Dependent upon the characteristics and thickness, the percentage of organic carbon, the position within the stratigraphic sequence, and, with respect to radiocarbon ages, the organic matter-rich layers generally are indicators of in-situ soil formation on older land surfaces or eroded and reworked soils. Radiocarbon dating of fossil wood found in several profiles predominantly result in modern ages. Samples that not have modern ages are dated between 210±50 years BP and 510±50 years BP. No fossil wood with an age older than 510±50 years BP was found in the sampled floodplain profiles.

In general, the floodplain sediments show a well developed vertical structure. The upper sediment section, 2.5-3.3 m thick, is composed of solid, densely bedded grey-yellow fine sands with hardly visible horizontal bedding. In general, a horizontal differentiation is macromorphologically not possible. These sediments are mainly of modern age, partly deposits of centennial flood events and show sedimentation rates of 2-3 m in ca. 50 years. Situated below, fluvial sediments of interbedded gravels, clays and silty sands follow down to the gravel base. In part, they are stratified by organic horizons and inclusions and show sedimentation rates of ca. 0.3 m in 100 years. The radiocarbon dating of numerous organic horizons as well as fossil wood show that the sedimentation during the older phase occurred between 1215 and 875 years BP at the base, and 670 and 15 years BP at the top of this sequence. During the deposition period of the lower floodplain section, over 600 to 800 years, no flood sediments were accumulated that correspond to the sediment of the upper section in grain structure and thickness.

Essentially, the erosion and denudation processes in the southern Cape region are controlled through fluvial dynamics of the large valleys. Among other reasons, this is due to the fact that different planation surfaces are each aligned to the higher river terraces as well as to the floodplain deposits, respectively to the present valley floor. Thus, the material that structures the river terraces and floodplain deposits must be, at least partly, considered to be sediments correlative to the denudation processes affecting the adjacent land surfaces.

However, the question arises how the deposition of the lower sequence of alluvial sediments is to be explained, which has been stratified by organic inclusions, and from which conclusions can be drawn with respect to their formation as well as to the factors controlling erosion and sedimentation. A possible model of formation of the floodplain deposits might be as follows: The aggradation of the lower (gravel) terrace was followed by incision resulting in the present-day valley floor, primarily through flood events and simultaneously, strong lateral erosion. At least locally, the bedrock was affected by fluvial incision. At low-level and mean discharge, the alluvium drift was temporary deposited on the valley floor until the next flood event. Not until the younger Holocene, according to the radiocarbon dating since 1215 to 875 years BP did the deposition of fine alluvial sediments start, which may have occurred in the context of moderate flood events. In part, sediment was also eroded from the hill slopes and deposited as hill-wash. As a result of these sedimentation processes, the formation of floodplains was initiated and the deposits grew in thickness to 2-3 m. As the floodplains grew, they could only be inundated at increasingly higher water levels, and were covered by, to some extent, thick sediments. Sedimentation like this, due to extremely high flood events, cannot be ascertained for the time before the 20th century, at the earliest.

According to the present state of knowledge, there is little evidence that the change in sedimentation is the result of climate variations. However, anthropogenic impact upon the fluvial morphodynamics in the southern Cape region is indicated by the settlement history (Fig. 4). In the pre-colonial period, the land use was that of the hunters and gatherers of the Khoi San, and partly intensive herding by the Khoi Khoi. Archaeological investigations suggest initial sedentary herding in the study area around AD 396 - 476, i.e. circa 400 years before the deposition of floodplain sediments started. For this pre-colonial period, degradation of varying, and in places possibly high intensity of the natural vegetation cover up to the start of soil erosion is assumed. The degradation of the vegetation cover due to pasture farming increased soil erosion and thus the influx of fine material into the fluvial systems, which could afterwards be deposited as alluvial floodplain sediment. A significant increase in pasturing and thus of soil erosion in the study area was initiated through the immigration of white settlers, especially connected with the introduction and intensification of sheep and goat herding. Furthermore, a strong population growth in the Cape region between 1750 and 1815 went along with an increase of livestock. At first, between 1750 and 1850, the Europeans lived as nomadic herders in the arid areas. However, after 1850 they became more sedentary. After capacity limits of the pastures had been reached by around 1865, ring fences and rotating pastures became compulsory at the beginning of the 20th century. Subsequently, by 1981 the livestock in the Karoo decreased by up to 50 % in response to the increasing dryness and thus lower productivity of the pasture land.

Deacon, H.J., 1995. Two Late Pleistocene-Holocene archaeological depositories from the southern Cape, South Africa. South African Archaeological Bulletin 50, 121-131.

Holmgren, K., Lee-Thorp, J.A., Cooper, G.J., Lundblad, K., Partridge, T.C., Scott, L., Sithaldeen, R., Talma, A.S., Tyson, P.D., 2003. Persistent millennial-scale climatic variability over the past 25,000 years in southern Africa. Quaternary Science Reviews 22, 2311-2326.

Meadows, M.E., Sugden, J.M., 1993. Late Quaternary environmental changes in the Karoo, South Africa. In Dardis, G.F. and Moon, B.P. eds., Geomorphological studies in Southern Africa. Proceedings of the Symposium on the Geomorphology of Southern Africa/Transkei, 337-353.

Scott, L., Bousman, C.B., Nyakale, M., 2005. Holocene pollen from swamp, cave and hyrax dung deposits at Blydefontein (Kikvorsberge), Karoo, South Africa. Quaternary International 129, 49-59.

Tyson, P.D., Lindesay, J.A., 1992. The climate of the last 2,000 years in southern Africa. The Holocene 2, 271-278.