The grassland dynamics of Mkambati game reserve.

A thesis submitted to the Faculty of Science, University of the Witwatersrand, Johannesburg, in fulfillment of the requirements for the degree of Master of Science.This work describes the coastal grasslands of Mkambati Game Reserve and their suitability for the present large herbivore complement. The initial description makes use of both field survey and quantitative sampling and multivariate analysis which facilitated the production of soil and vegetation maps at tre scale of 1: 10 000. Suitability of these grasslands for herbivores is assessed in terms of both quantity and quality of food available and the present herbivore impact. Local variation in soil type was considerable. Eight standard soil forms were distinguished as well as severaI others previously undescribed. Most of the soils were dystrophic although nutrient 'hot-spots' were common. This determined floristic composition of the various communities to some extent. Three grassland communities and four subcommunities were recognized which contributed 81, 5% to the total area of the reserve. The remaining area comprised forest. wetlands. exposed rock and accommodation camps. Each grassland community was associated with particular ranges of the measured environmental variables. The standing crop of the three communities was high relative to other areas of Southern Africa. Absolute amounts varied seasonally being highest in summer and lowest in winter Partitioning between the phytomass, necromass and litter components also varied seasonally. Total standing crop increased with increasing interval since the last fire. Litter breakdown rates differed between communities and species. Forage quality was poor being lowest in autumn and winter and declining with age. This was probably a result of the dystrophic soils. Crude protein and phosphorus were limiting for several months of the year. Total forage quantity was in excess of the present herbivore needs although availability declined with age and during the non-growing season. However, the poor quality reduced the amount available such that present herbivore impact was low, at all times being less than 9%.. Removal by herbivores was highest several months after a fire when absolute amounts of various nutrients were at a maxinum per unit area. Thereafter it declined to less than 1% with aging of the sward. The implications of the results for management are discussed and future research needs identified.AC 201