Informations et ressources scientifiques
sur le développement des zones arides et semi-arides

Accueil du site → Master → Afrique du Sud → 1998 → Effect of stocking rate and rainfall on rangeland dynamics and cattle performance in a semi-arid savanna, KwaZulu-Natal

University of KwaZulu-Natal (1998)

Effect of stocking rate and rainfall on rangeland dynamics and cattle performance in a semi-arid savanna, KwaZulu-Natal

Fynn, Richard Warwick Sinclair

Titre : Effect of stocking rate and rainfall on rangeland dynamics and cattle performance in a semi-arid savanna, KwaZulu-Natal

Auteur : Fynn, Richard Warwick Sinclair

Université de soutenance : University of KwaZulu-Natal

Grade : MS, Grassland science, (1998)

Considerable understanding of the functioning of semi-arid systems is still needed to enable range managers to formulate management policies, with a degree of confidence. Long term data sets that encompass a wide range of interactions between the various major components of a semi-arid system (vegetation, herbivory, animal performance, landscape and rainfall), are unfortunately rare but essential to provide sufficient depth of data to adequately test various hypotheses about rangeland dynamics. This study comprises an analysis of a ten year data set derived from two cattle stocking rate trials in the semi-arid savanna of KwaZulu-Natal. Statistical analysis revealed that the most pronounced and rapid compositional change was due to rainfall, but that stocking rates between 0.156 and 0.313 AU ha-1 had an important effect as well. Sites on steeper slopes with heavy stocking rates, exhibited the greatest amount of compositional change between 1986 and 1996 (40 Euclidean points in heavy stocking rate treatments on slopes vs. 21-24 Euclidean points in heavy stocking rate treatments on flatter land, or 11-24 Euclidean points in low stocking rate treatments). Heavy stocking rates in conjunction with low rainfall tended to cause decreases in densely tufted perennial grasses and increases in annuals and weakly tufted perennials. Multiple regression analysis revealed that seasonal peak grass production (measured as disc height) declined between 1986 and 1996 only at those sites on steeper slopes with heavy stocking rates. The camps that declined in productivity also underwent the greatest degree of compositional change. The decline in grass productivity in certain high stocking rate camps did not translate into a decline in cattle performance. Depending on rainfall, cattle gained on a seasonal basis between 112 and 241 kg at low stocking rates, 82 and 225 kg at medium stocking rates and 84 and 217 kg at high stocking rates. Rainfall, compared with stocking rate, accounted for the greatest amount of variance in seasonal peak grass production and cattle performance. Cattle performance had a strong curvilinear response to rainfall, which also proved to be a better predictor of cattle performance than grass biomass. There were no clear trends in soil physical and chemical characteristics between low and high stocking rates that could provide convincing evidence that loss of soil nutrients was an important mechanism of range degradation. The total standing crop of plant nitrogen but not of phosphorus tended to decline at high stocking rates. Plant nutrient and van Soest analyses suggested that forage quality was higher at heavy stocking rates. The results of this study generally supported traditional concepts of rangeland dynamics with regard to rainfall and grazing effects on compositional change and seasonal grass production. The results were important in being able to show quantitatively that heavy stocking rates result in a decline in grass production and that this effect is dependent on an interaction between stocking rate and landscape position or slope, and that there is a link between a decline in seasonal grass production and compositional change. The results also highlighted areas for future research that would be useful for furthering our understanding of various aspects of rangeland dynamics and mechanisms of degradation


Version intégrale (22,16 Mb)

Page publiée le 9 juillet 2014, mise à jour le 1er janvier 2021