Models developed for the following projects could form the basis of an MSc or PhD thesis. These are multi-disciplinary projects and honours graduates in any branch of science or agriculture will be considered. Some knowledge of computer programming would be helpful.
Namaqualand livestock-ecosystem model
In collaboration with the Plant Conservation Unit at UCT we are developing a series of computer simulation models of livestock production and vegetation changes for communal and commercial rangelands in different regions, the Succulent Karoo of Namaqualand and the Savanna and Grassland biomes. You will develop a computer simulation model of diet selection and nutrient utilization by domestic livestock or wildlife for the area concerned. This is a multidisciplinary project that may involve both some experimental work in the field in addition to computer modelling.
Recent results from the rangeland modelling project have indicated the important influence of goats on the stability and sustainability of rangeland systems. Consequently rangeland models depend on accurate representation of the performance of goats and their impact on the vegetation and on the productivity of the whole system. The sub-models that simulate diet selection, nutrient utilisation, production (meat, milk, fibre), reproduction and mortality of free ranging goats need to be further developed.
Simulation models of rumen function
The end products of microbial fermentation in the rumen provide the major proportion of nutrients available to the host animal. Many microbial species are present in the rumen and most are selective fermenters. The pattern of end products produced and rate of fermentation varies with the relative proportions of species present and the food eaten. The rumen microbial population varies with the diet, amount eaten and time since a specific food was eaten. The different organisms interact with one another and with the rumen contents. Rumen models that simulate changes in the proportions of different microbial species are very complex even if only three types of organism are represented. To date no model incorporates rumen anaerobic fungi which may play important roles in the digestion of lignified rangeland plants.
Modelling spatio-temporal variability in rangelands
One of the weaknesses of our work to date has been that changes over space have largely been ignored, although a cellular automata model was developed to study the effect of distribution of water points on spatial variation in grazing pressure. There are two problems of importance where the development of suitable models could contribute tour understanding of the systems.
(1) Simulation of the development of degraded patches in semi-arid rangeland. In reality this often more complex than just the formation and spread of bare patches that has been the subject of earlier research. In the Karoo, palatable plants are also replaced by unpalatable and poisonous ones, and in grasslands and savannas bush encroachment occurs in patches that progressively increase in size.
(2) Modelling the spread and intensity of veld fires. These are influenced by density and moisture content of the standing biomass and prevailing weather especially wind.