Modelling the sustainability and productivity of pastoral systems in the communal areas of Namaqualand

We investigated the effects of rainfall and the number of animals on changes in vegetation and on the output of milk and meat from the communal areas of Namaqualand. Previously published short- and long-term models link processes that range from the levels of tissue (in, for example, the mammary gland), to the milk yields of individual animals, to the growth and survival of their young and to long-term changes in plant species populations at the ecosystem level. These models have been used to study how different factors and management strategies affect livestock productivity and vegetation composition on a 20,000 ha rangeland in Namaqualand. First, the inter-relations between rainfall, stocking rate and productivity were studied using the short-term model. This model shows that in addition to total rainfall and stocking rate, the timing of rainfall within a year also influences doe live weight and survival to the end of the year. When the long-term model is run, using recorded rainfall, predictions of small stock numbers agree closely with livestock data recorded over the same 30-year period. One thousand replicates of 100-year runs of the long-term model were then used to study the probable impact of different upper limits to stock numbers on animal performance. Off take (sales and slaughterings) are maximal when stock numbers are limited to 2000 adults. Animal numbers increase marginally as the limit is increased above this level, but the variability between years in numbers increases. Secondly, the long-term model was used to study the long-term effects of the stocking rate strategies on rangeland condition. The model predicts that although these effects are variable, when moderately degraded range is stocked with an upper limit at the recommended level it is unable to recover to less degraded states over 100 years. Thirdly, the model was used to examine the effects of reduction in stock numbers and slaughtering of kids in a drought year on goat numbers during the subsequent 5 years. Finally, the model predicts that a 10% reduction in mean annual rainfall will lead to a 35% reduction in animal numbers over 200 years.