Drivers of vegetation change in the eastern Karoo.

Abstract

The Nama-Karoo Biome occupies much of the western central region of South Africa and transitions into the Grassland Biome along its eastern boundary along a gradual ecotone. The area is characterised by hot summers and long, frosty winters, relatively low rainfall, peaking in mid- to late-summer, with high inter-annual variability, and botanically by a co-existence of grasses and dwarf shrubs, with grass abundance positively related to average annual rainfall that increases from west to east. Biome shifts in response to changes in rainfall pattern and grazing have been suggested but never directly examined. Major drivers of botanical composition are rainfall and grazing by livestock. Fire is rare, occurring sporadically if high rainfall allows for good grass growth. This thesis focused on understanding the influence of rainfall, grazing, low temperatures, and fire on botanical composition at Grootfontein, a site in the ecotone between the Nama-Karoo and Grassland Biomes that is home to grazing trials up to 85 years old. The following specific questions were addressed: Question 1: Over the long term, has Grootfontein shown patterns of rainfall cyclicity or experienced directional change, and how might these have influenced composition and productivity? Using data from 1888 to 2012, cyclicity in rainfall was evident for periods of approximately 20 and 60 years. Rainfall has also consistently increased since the mid-1970s, and this increase corresponds with a general pattern of increased grassiness in the eastern Karoo. Question 2: How do rainfall and grazing, alone and in interaction, influence vegetation composition in the eastern Karoo? Compositional data from the 1960s and 2010s from various treatments at two sites at Grootfontein (Camp 6 and Seligman grazing trials) show a shift from dwarf-shrub- to grass-dominated vegetation, consistent with the increased rainfall over that time. An influence of grazing, both present and historical, was evident but secondary to the effect of rainfall. In some cases, there has been a shift to grassland. Data from the Camp 6 and Seligman grazing trials from the 1940s to the 1960s further indicate a combined influence of season of grazing and of rainfall. High-intensity summer-only grazing by livestock largely extirpates grasses and allows shrubs to thrive, while summer grazing in the form of rotational grazing or continuous grazing allow for a balance of grasses and dwarf-shrubs. Severe declines in both grasses and shrubs occurred apparently in response to drought, though the exact conditions required to cause such mortality remain unclear. Plant cover data from 2008 to 2015 from the Boesmanskop grazing trial showed that consecutive years of exceptionally high rainfall increased plant cover to nearly 100%, and increased the abundance of grasses. Competitive exclusion of dwarf-shrubs by grasses was not evident. Question 3: What have been the trends in minimum temperatures, frost, and potential growth season at Grootfontein, how might these have influenced botanical composition and productivity, and is there evidence of increasing temperatures consistent with global warming? This is addressed using minimum-temperature data from 1916 to 2014. Minimum temperatures were lower than are usually reported. Variability in minimum temperatures was high, including a cooling from the 1910s to the 1950s and a warming from the 1950s to the 2010s. The length of the growing season (last frost to first frost of the subsequent season) varied considerably, and may have the potential to influence botanical composition. Question 4: What is the influence of fire in Karoo vegetation? Based on the effects of a single fire on Grootfontein, fire killed some species while most species resprouted. Grasses appeared unaffected in terms of survival, several species of dwarf shrub (notably Eriocephalus ericoides and Ruschia intricata) were killed, and will need to re-establish by seed (termed nonsprouters), while most dwarf shrub species resprouted. This resprouter/ nonsprouter dichotomy was found to be evident at a range of other fire sites in the Karoo. Heavy grazing appeared to strongly impede the recovery of burnt veld, maintaining it as a sparse grassland dominated by annual species and occasional unpalatable shrubs. Should the grassiness of the Karoo continue to increase, then fire may become more frequent thereby maintaining a grassland state. The findings allow for greater understanding of interactions among rainfall, grazing, and fire in eastern Karoo ecosystems, and these are discussed in the context of an existing state-andtransition model of eastern Karoo vegetation dynamics. The importance of long-term rainfall trends, rather than short-term variability, are highlighted. Long-term increases in rainfall will likely induce a biome shift to grassland, concomitant with a drastic reduction in dwarf-shrubs. This will likely alter both long-term carrying capacity for livestock and the type of animals that may be optimally stocked. Increased grassiness will result in the increased likelihood of fire, and if post-fire grass fuel loads remain above a critical level, a fire/grass feedback loop may be initiated whereby dwarf-shrubs are largely eliminated owing to their slow rates of growth or re-establishment. Introduction of infrequent fire will likely result in resprouter-dominated vegetation proliferating. It is demonstrated that the resilience of Karoo veld may be higher than previously thought, with severe grazing, droughts, and fire not pushing veld beyond a threshold into a state of denudation. Thus the prospects of conserving Karoo landscapes despite historical management remain high. Some key future research efforts needed to improve our understanding of Karoo ecology include the life-histories of dwarf-shrubs, the conditions of drought and herbivory under which grasses and dwarf-shrubs die, and how and when perennial dwarf-shrubs and grasses regenerate. Based on historical trends, the continued existence of long-term research trials, such as those at Grootfontein, may be under threat and should receive attention.