Assessments of the anthropogenic threats to savanna ecosystems are primarily focussed on land use change, bush encroachment, and biological invasions. There is, however, very little understanding as to the threats from atmospheric pollution. South Africa is the major emitter of CO2 on the African content while the Mpumalanga region bordering the Kruger National Park (KNP) is among the leading regions for nitrous oxide pollution in the world. It is not only increasing atmospheric pollution, but rainfall intensity is also predicted to increase for southern Africa. As savannas are nutrient limited, an increase in nitrogen deposition will have major consequences for vegetation structure and this can only be exacerbated by increased rainfall amounts. Current research suggests that these predicted increases in water and nutrients will result in increasing grass biomass and decreasing herbaceous species richness. The effects of global change drivers on savanna vegetation are also likely to propagate through to multiple trophic levels, with changes in vegetation structure cascading down to invertebrate assemblages. As invertebrates are ubiquitous, form the bulk of metazoan species diversity and biomass on earth, and play a pivotal role in many ecosystems, I discuss in the introductory chapter of this thesis why the influence of global change on these assemblages should not be ignored. In my first data chapter, Chapter 2, I examine the effect that increases in available nutrients and water may have on vegetation structure, and how this may cascade down to grasshopper and ant assemblages. I do this using a fully factorial experiment in KNP with nutrient and water additions where I assessed both herbaceous (forb and grass) and insect (ant and grasshopper) assemblages five years after resource additions began. My results show that there was a substantial increase in grass biomass while plant and insect species richness declined with water addition alone and that a combination of nutrients and water resulted in the greatest increases in grass biomass and concomitant decreases in plant and insect species richness. The effects of nutrient and water additions on the insect community assembly was primarily driven by a decrease in grasshopper species and ant abundance respectively. An analysis of ant functional traits showed that the rare ant species mediated the impact of the resource additions on the ant assemblage. Fire is inherent to savanna systems with profound effects on vegetation structure. There has, however, been relatively little research on the effects of fire on savanna invertebrate fauna. In Chapter 3 I look at the effect that fire may have on the vegetation and insect community assembly at my study site between five and eight months after the site had been burned. These results show an increase in grass biomass and decrease in plant and insect species richness with a combination of nutrients and water. My results also show that grasshopper biomass, abundance, and species richness decreased as herbaceous biomass decreased. While ant species richness decreased, ant abundance increased post-fire, primarily related to an increase in patches of bare ground. With global change, drought frequency is also expected to increase. The insect and grass assemblages, both on and off Macrotermes mounds, at two sites in the southern section of KNP had been sampled in a separate study in 2012. In Chapter 4 I describe a study where I resampled these mounds during the peak of the most severe drought in 30 years. The two sites differed in drought severity, one where the drought severity was very high and the other where severity was much lower. The objective was to determine the effects that drought may have on the grass and associated insect assemblages both on and off termite mounds. My results show that at the high severity site grass cover and biomass and grasshopper abundance decreased both on and off mounds. The overall reduction in habitat structure resulted in an increase in both ant abundance and species richness but the mound and matrix ant assemblages diverged during drought. Where the drought was less severe there was an increase in large mammal herbivores as animals moved out of the more affected areas. This increase in mammal herbivory was more evident on rather than off mounds resulting in grass biomass being lower on rather than off mounds. The cascading effect saw grasshopper abundance decrease on and increase off mounds. The mound and matrix ant assemblages did not respond to the comparatively smaller change in habitat structure. Finally, in the synthesis chapter I discuss my results in the broader context of how global change drivers such as increased nitrogen deposition may cascade down from plant to insect community assembly. At present there is very little understanding of the amounts of nitrogen being deposited in KNP or the effect that this may have. The results of my study would suggest that this increase in nitrogen deposition will have major consequences for vegetation structure and that this will cascade down to the insect assemblage. In mitigating for this, it is therefore essential that management in KNP adapt a monitoring protocol for nitrogen deposition, especially when considering that where N deposition is really high fire may not volatilise everything to allow the system to reset itself back to its original state. It is not only nitrogen deposition, but drought frequency is also likely to increase. In mitigation for this there should also be monitoring programmes to consider the effects of drought as animals may move from areas of high drought severity to areas where severity is lower. Such movement will increase grazing pressure on both low and high nutrient environments with cascading effects on vegetation structure and insect assemblages.

show more