The responses of grasses to fire and bush clearing in the Hluhluwe Game Reserve.
Graham, Philip Mark.
MetadataShow full item record
Contemporary and historical studies of the flora of the Hluhluwe Game Reserve (HGR), have emphasised the woody component whilst little work has been performed on the herbaceous vegetation. This is particularly true with regard to the responses of grasses to historical fire and bush clearing. This study attempted to elucidate some of these responses. Of all the variables considered in this study, woody cover, altitude, the number of fire events, geological and soil parent material are the most important affecting the abundance of grasses in this reserve. Most of these variables are not independent in their effects on grass abundance, with varying degrees of correlation between each other. Certain species appear to be restricted to particular geological substrates. Along with successional changes in the composition and cover of the woody community, due to seemingly inevitable bush encroachment, there is a parallel change in grassland communities in this reserve. In the absence of clearing, numerous fires, higher altitudes, igneous geology and soils derived from igneous parent material delays this succession, whilst sedimentary geology at lower altitudes and fire frequencies accelerates the trend to high woody cover and associated grass species. The grass communities in HGR were shown to be significantly affected by bush clearing and fire. Specifically the number of clearings and fire events, physical bush clearing during 1957 - 1963 and chemical bush clearing during 1968 - 1978. From the responses of species in relation to the various key environmental variables, viz. geological substrate, woody cover and burning and clearing, a model of species response to these variables was developed. With increased fire and bush clearing frequency, the grass communities shift from closed woodland, shade tolerant species through to more open fire climax grassland. These are also more productive communities producing palatable grasses. Validation of aspects of the model were successful - the model having a relatively high predictive capability. Further testing of the model over different substrates and under different clearing regimes is necessary. With regular fires and re-clearing in some bush cleared areas, the vegetation of this reserve should be able to be maintained as productive and diverse grasslands. In the absence of this management, the grass communities will shift towards species associated with woodlands. Bush clearing activities would appear to be most effective over sites on igneous substrate, at higher altitudes, where successional rates are slowest. This is in comparison to sites at lower altitudes over sedimentary geology.