Environmental stress and landscape recovery in a semi-arid area, the Karoo, South Africa

John Boardman, Ian D L Foster, Kate M Rowntree, T M Mighall, John B Gates

Research output: Contribution to JournalArticle

Abstract

Semi-arid landscapes are vulnerable to cultivation, overgrazing and climate variability, although it is difficult to identify the relative significance of these three factors. In the South African Karoo, the ‘desertification debate’ seeks to explain a change to more shrubby vegetation in heavily grazed areas. We examine these issues in catchments where farm reservoirs provide sediment stores with 137Cs, 210Pb, geochemical and mineral magnetic signatures. Rainfall data and stocking numbers are reported and current erosion rates are estimated. Sediment accumulation in the reservoirs increases between 1935 and 1940, probably due to rises in the frequency of rainfall events of 425mm day. Significant increases in sedimentation rates (68) occur during rain-fed wheat cultivation. In an uncultivated catchment, sediment yields remain relatively high and reflect increases in erosion from hillslopes, colluvial storage, and cultivated land sources. Gully systems have acted as transport routes rather than sediment sources over the last ~70 years. Badland erosion rates average ca. 50t ha-1 yr-1. At Ganora, their development in the 1920s strongly influences peak sedimentation between 1970 and 1980. Any delay between badland initiation and increased sediment yield appears to be a function of landscape connectivity. Recovery following disturbance is occurring slowly and is likely to take ~100 years. De-stocking and better management systems are reducing erosion rates but may be offset by increases in rainfall intensity
Original languageEnglish
JournalScottish Geographical Journal
Volume126
Issue number2
DOIs
Publication statusPublished - 1 Jun 2010

Fingerprint Dive into the research topics of 'Environmental stress and landscape recovery in a semi-arid area, the Karoo, South Africa'. Together they form a unique fingerprint.

  • Cite this