Daily sampling of suspended sediment in the flooding Orange River was carried out at Alexander Bay from late February to the end of May 1988. Coupled with simultaneous measurements of water run off, the suspended-sediment data have been used to calculate daily variations in sediment discharge from just before the floods started, through four distinct pulses of varying magnitude and duration, namely the Bloemfontein, Fish River, Kroonstad and Upington floods. During the 3-month period, 24,3 km3 of water, 118,6% more than the mean annual runoff of 11,1 km3, was discharged, varying from 15,5 km3 in March (a record) to 5,3 km3 in April and 3,5 km3 in May. The amount of sediment, transported chiefly in suspension, was substantial but was not in proportion to the amount of water. During March 64,2 x 10(6)t was carried out to sea, the load declining sharply to 9,4 x 10(6)t in April and 7,3 x 10(6)t in May. The total sediment discharge for the three-month period amounted to 80,9 x 10(6)t, which exceeds the mean annual value of 60,4 x 10(6)t by only 40%. The texture of the mud-dominated suspended sediments was analysed by Sedigraph and clay was always dominant over silt and sand. However, the mode of the suspensate for the Bloemfontein, Kroonstad and Upington floods was medium silt and that of the Fish River flood, fine silt. Illite was the dominant clay mineral of the clay fraction but montmorillonite, derived from weathered basalts in the Drakensberg increased in abundance during the Bloemfontein and Kroonstad floods; the Fish, in contrast, is montmorillonite-poor. Near the mouth, at the height of the flood, a major channel was scoured beside the left bank, beneath the southern end of the Ernest Oppenheimer bridge (where the suspended-sediment samples were taken) and approximately 1 million tons of sediment were eroded from the right bank of the estuary. The bedload of the flood was deposited as an ephemeral flood delta up to 1,2 km offshore with an estimated mass of 3,6 million tons. The muddy suspended load was deposited on the normally sandy delta front (< 40 m) as a clearly recognizable layer of light brown 'flood mud' up to 10 cm thick. In contrast, only a veneer of 'flood mud', a few millimeters thick, was noted on the normally muddy prodelta (> 40 m). The anomalously muddy sediments (up to 97% mud) on the delta front will take several years to be redistributed by the prevailing swell. Nearly all sediment transported during the 1988 floods was derived from bank erosion and river-bed scour downstream of major dams, before the construction of which historical floods eroded poorly-managed farms, particularly in the NE Cape.
