Donaldson, Jason Ernest2019-05-202019-05-202018https://hdl.handle.net/10539/27065A thesis submitted to the Faculty of Science, University of the Witwatersrand, Johannesburg in fulfillment of the requirements for the degree of Doctor of Philosophy Johannesburg 2018Traditionally, landscape structure and variability are thought to be largely determined by climatic and edaphic conditions (‘green world’ view). However, it is now accepted that herbivore densities and fire regimes play an important role in shaping vegetation structure due to their role as consumers. In savannas the roles of fire and herbivory have been recognised for a long time but their interactions and feedbacks to vegetation are still being elucidated. Savanna trees in the same environment show evidence for specialisation to either frequently burnt or heavily grazed communities suggesting that alternative system states, representing either a ‘brown’ or ‘black’ world (heavily grazed vs. frequently burnt), can occur in savanna landscapes. When it comes to the grass layer we know that grazers can create and maintain patches of high-quality grazing habitat and exclude fires locally by removing grass biomass and limiting fuel for fires (brown world). Comparatively, large fires disperse grazing animals over broad areas of the landscape as they respond to high quality graze available on burn scars in what has been termed pyric herbivory. The “thinning” of large mammal herds over wide ranges after a fire decreases concentrated grazing and allows grasses to accumulate biomass and ultimately encourages repeat burning (black world). Thus, both consumers have self-enhancing feedback loops and we would expect grass communities to respond to these: different species with divergent functional traits, either associated with frequent fire or heavy grazing, should be associated with each consumer state. Abiotic conditions, soils and rainfall, influence the strength of these feedbacks and alter the dominance of fire vs. herbivory across environmental gradients. This implies that rainfall variability, and the occurrence of drought, should influence these feedbacks at a particular location. How drought affects the competition between these consumers by shifting the conditions governing feedbacks is not well understood, but is becoming increasingly important in southern Africa where drought events are expected to become more frequent and severe with climate change. The main aim of this study was to investigate the processes involved in the establishment and break-down of grass-consumer feedbacks and understand how these dictate the balance of fire and grazer driven grass communities in an African savanna. Specifically, I used a landscape scale experiment setup in the Kruger National Park to test how easily a system switch from fire- to grazer-dominated can be initiated in a landscape, and how the functional traits of the grasses drive these switches. I then tracked the impacts of a severe two season drought on the established consumer driven feedbacks within the study system and analysed the knock-on implications of these changes in grass community structure and function at broader scales. The experimental results indicate that altering fire management practices and burning repeated small (<25 ha) fires concentrated grazers. After three seasons (2013-2015) this resulted in a structural shift in grass communities that maintained local grass height in a short-cropped state and excluded fire. Once the dominant consumer pressure shifted there were notable changes in the local grass species composition and novel grass communities had functional traits that favoured repeat grazing (low C : N ratios and high leaf moisture content). Thus, feedbacks between grazers and grass communities established within a relatively short period. This work adds to a growing number of studies that highlight the importance of feedbacks in the establishment of dominant consumer processes within savannas. Drought altered the forage use patterns of grazers as biomass on grazing-lawns became limited and animals were forced to feed in low-quality tall-grass areas or migrate away from the study site completely. This more intensive grazing in the broader landscape where fire is usually the dominant consumer lead to a convergence in the grass community trait space between lawns and tall-grass areas. This suggests that drought has a homogenising effect on grass community trait space across fire-driven and grazer-driven grass community patches, at least in the short-term. I found that grass-grazer feedbacks did remain at broad scales with lawns re-establishing after the drought, but that these areas were severely restricted in the study system (<4% cover) and only occurred where strong drivers of grazer feedbacks (water availability) repeatedly concentrated grazing. Moreover, drought reduced the basal area of tussock grass, and created large areas of seemingly bare ground in the grazed patches. But grass communities recovered quickly – and within 3 months after the drought broke productivity was at its maximum. My results highlight the importance of grass-consumer feedbacks in dictating the structure and function of African savannas. Clearly, fire and herbivory need to be considered in conjunction and understanding the state of a given grass community can only be done by assessing the governing grass-consumer relationship. Drought represents a good example of this; changes in the short-term abiotic conditions within a savanna appear to have little impact on savanna function and structure unless it drives changes in the positive feedback loops that govern the distribution of grazer-driven and fire-driven grass communities within the landscape. Fires remain one of the only major management tools in large protected savannas and my thesis underlines the need to manage them by considering how fire influences grass-grazer dynamics and not simply use fire as a ubiquitous method for improving graze quality in the short-termOnline resource (166 leaves)enThesis