A. D. Clulow

Evapotranspiration from mine-affected riparian sites along the Vaal River in central South Africa

Abstract Long-term gold mining within the Witwatersrand Basin Goldfields has led to the creation of over 200 tailings storage facilities (TSFs) releasing acid mine drainage. Plumes of contaminated shallow groundwater occur downslope of these TSFs. One such area lies adjacent to the Vaal River east of the town of Orkney. Natural vegetation in the riparian soils comprises a mixture of grasslands, reeds and woodlands dominated by Vachellia karroo. The Mine Woodlands Project initiated by Witwatersrand University and AngloGold Ashanti is exploring the use of engineered woodlands for hydraulic control of plumes and immobilization of contaminants to protect the wider hydrological resource. To estimate the net increase in evapotranspiration (ET) resulting from various remedial woodlands, ET from existing vegetation types requires quantification. Growing season patterns of ET were measured and modelled in riparian grassland and two V. karroo woodland sites close to the Vaal River. Two independent estimates of annual ET from these woodland sites (674 and 787 mm) were similar to that from grassland (697 mm) and much less than the 1100–1200 mm reported from some forms of engineered evergreen woodlands established in the vicinity. We conclude that the Vachellia woodlands have low potential for hydraulic control of contaminant plumes.

The annual pattern of sap flow in two Eucalyptus species established in the vicinity of gold-mine tailings dams in central South Africa

Several hundred mine tailings dams occur in the Witwatersrand Basin Goldfields in central South Africa. Seepage of acid mine drainage (AMD) from these unlined structures is widespread, and a variety of contaminants is released into soil and groundwater. The ‘Mine Woodlands Project’ is aimed at evaluating the use of high-density tree stands over surrounding contaminant plumes to limit the spread of contaminants through hydraulic control of groundwater and enhanced uptake or immobilisation of contaminants. The annual pattern of hourly sap flow in four contiguous Eucalyptus dunnii trees (aged three years) was followed over a full year in a species trial situated near Carltonville. The annual pattern of hourly sap flow was also recorded in four contiguous sample trees (aged four years) of the clonal hybrid E. grandis × E. camaldulensis (E. G×C) at another trial near Orkney. Both species showed high sap flow rates close to reference evaporation rates in response to summer rains. Both showed greatly reduced sap flow rates during the latter half of the dry winter season. Sap flow rates only recovered after the arrival of the first spring rains. Annual sap flow (E. dunnii, 673 mm; E. G×C, 767 mm) was similar to the recorded annual rainfall at each site (E. dunnii, 629 mm; E. G×C, 795 mm), and was substantially lower than total annual reference evaporation (E. dunnii, 1 273 mm; E. G×C, 1 330 mm). We conclude that the roots of both species are not yet deep enough to access the AMD-influenced groundwater, which lies at depths of 14 and 10 m below the ground at the Carltonville and Orkney sites, respectively. Despite prolonged water deficits, both species survived well and maintained sufficient vigour to permit the quick recovery of high transpiration rates in the following summer. This resilience is essential to hasten root growth and improve the chance of contact with groundwater plumes.