AN INITIAL APPROACH TO PREDICTING THE SENSITIVITY OF THE SOUTH-AFRICAN GRASSLAND BIOME TO CLIMATE CHANGE

The grassland biome in South Africa can be distinguished from neighbouring biomes on the basis of three climatic indices which are considered to be biologically meaningful. The indices are 'days of growth opportunity' based on soil water availability and calculated using a water budget model, the mean temperature on days when sufficient water is available for plant growth, and the mean temperature on days too dry for growth to occur. The difference between the growth and no-growth temperatures provides an indication of the degree of rainfall seasonality. These indices have also been used successfully to differentiate the five other biomes in South Africa (forest, fynbos, nama-karoo, savanna and succulent karoo) in a 'biome model' presented as a decision-tree. This model provides a simple method of predicting possible vegetation responses to climatic change. The possible long-term response to a scenario of climatic change, (a 2-degrees-C increase in temperature and a 15% decrease in rainfall), is illustrated as an example. Such an approach to modelling the impact of climate change has several advantages, including simplicity, appropriateness of scale and detail to the predictions of current climate models, identification of regions and processes that may require more intensive study and identification of the kind of information that will be required if scenarios of climatic change are to be translated into vegetation response. The model fails, however, to identify rates or pathways of change, or the species involved. This approach will therefore have to be complemented by more detailed dynamic models.