Influence of land cover degradation on the water balance of the Northern Drakensberg high altitude mesic grasslands, South Africa.

Doctoral Degree. University of KwaZulu-Natal, Pietermaritzburg.

Mountainous regions provide vital ecosystem services, such as water provisions to low land areas. However, these regions are also considered sensitive to the effects of environmental change, due to their high levels of endemism and biodiversity. Thus, environmental change within these regions could have significant consequences beyond the extent of the region itself. An important implication of environmental change is the impact a change in land cover could have on the water balance of a catchment, especially within the headwater catchments. The Drakensberg mountains in South Africa, is such a mountain region, which is vital for its provision of ecosystem services and generation of water resources. This mountainous region is under threat from anthropogenic environmental change. The Drakensberg mountain range has been identified within South Africa as a strategic water source area (SWSA) and within the Northern-Drakensberg SWSA exists the Maloti-Drakensberg Park, which is a protected area managed by Ezemvelo KZN Wildlife and a World Heritage site. The Northern-Drakensberg mountain range natural grasslands are under threat from two forms of land cover change degradation, that of woody encroachment and following disturbance, invasion of bracken fern. Woody encroachment occurs within the natural grasslands following the removal of fire, which is of concern within the protected areas, where fire is a current management tool to maintain the natural grassland cover. Outside the protected area, disturbances due to human activities such as overgrazing and poor land management have led to a substantial level of degradation occurring, providing ideal conditions for the invasion of bracken fern. To safeguard and maintain the assurance of supply from an important SWSA of South Africa, there is the need for improved understanding on the impacts of land cover change and degradation within the Northern-Drakensberg mountain range on the water balance of the region. Thus, the overall aim of this thesis is to understand the influence of land cover change related degradation on the water balance of the Northern-Drakensberg high-altitude mesic grasslands and what this in turn means for the water supply generated from this SWSA of South Africa. To achieve the main aim, the research was scaled from a point measurement to a basin scale where management decisions are made. To improve our understanding of the impact of degradation related land cover change on the water balance an observational approach is required. Located within the Maloti-Drakensberg Park is the long-term Cathedral Peak research catchments which provided the platform for the observation component of this research. Of interest in the Cathedral Peak research catchments were three hydrologically individual catchments. Catchment III which consists of a degraded bracken fern land cover following the historical research experiments within the catchment on the impacts of commercial Pinus patula plantation on streamflow. Despite rehabilitation efforts following the removal of the plantation, the catchment transitioned to a degraded state and is currently near-completely invaded with bracken fern. Catchment VI which is under natural grassland condition, is managed with a bi-annual spring burn as proven best practice within the Drakensberg region. This catchment formed the baseline catchment for this research. Catchment IX is a woody encroached land cover following the protection of fire since the 1950s as part of research into the implications of removing fire from the natural grasslands. These three catchments provided the ideal platform from which the change in water balance components under each land cover could be monitored and investigated. The process of evapotranspiration (ET), which forms the connection between the energy and water balances, is well understood to be one of the most affected components of the water balance following a change of land cover. Therefore, ET was the focus of the initial point observation research. Due to the cost and stringency of the prominent method for ET measurement of eddy covariance (EC), an alternative more financially feasible method of surface renewal (SR) was tested over both Leucosidea sericea (woody) and Pteridium aquilinum (bracken) vegetation in comparison to EC. It was determined that the SR method, and in particular the SR dissipation theory (SRDT) method, which is independent of EC, was the best alternative and cheapest method to EC. The SR method is now used within the Cathedral Peak research catchments for long-term estimations of ET over both vegetation canopies. During this research, calibration factors (α) for the surface renewal 1 (SR1) method were determined for both the Leucosidea sericea (woody) and Pteridium aquilinum (bracken) vegetation types for winter and summer. Having confirmed the SRDT ability as an alternative to EC, the seasonal ET of Leucosidea sericea (woody) and Pteridium aquilinum (bracken) was determined. Providing first insight into the seasonality of ET over these vegetation types. Both vegetations followed a similar seasonal ET cycle to the natural grasslands, with the largest difference occurring in winter when the grasslands become dormant. It was also found that the energy balance was altered under the degraded land covers, with both forms increasing available energy and latent heat flux. Following the understanding of the seasonal change in ET through point measurements, the research focus was scaled up to the research catchment. A comparison of catchment VI and IX was conducted to identify changes in the headwater catchment water balance between the natural grassland and woody encroachment land covers. Long-term data sets of precipitation and streamflow from the Cathedral Peak research catchments, in combination with the seasonal ET observations showed that over time, as woody encroachment increased, the catchment rainfall:runoff response ratio decreased, as did streamflow under woody encroachment compared to natural grassland. Having gained an understanding of the impacts on a water balance of a headwater catchment, hydrological modelling in combination with scenario analysis was used to understand the impacts on water supply from the upper-uThukela catchment were both land cover degradation threats left unmanaged. Land cover parameters were unavailable for Leucosidea sericea (woody) and Pteridium aquilinum (bracken), and therefore were derived from understanding gained during observation. The ACRU agrohydrological model was utilised and confirmed to simulate current land cover conditions within the upper-uThukela satisfactorily at both the headwater and catchment levels. It was identified that both forms of land cover change resulted in a reduction in streamflow. This was largest for woody encroachment. The most affected flows were the low flows and winter dry period flows. Headwater catchments were also identified as the most impacted by land cover change. The key conclusions of the research were: • that the surface renewal methodology is a viable alternative for obtaining estimates of evapotranspiration over indigenous vegetation types; • that the energy balance and ET of woody vegetation in comparison to the natural grassland was significantly altered; • the importance of fire, as not only a management tool for the maintenance of the natural grasslands, but also to ensure the sustainability of the vital water resources and ensuring water security; • that there is an evident lag between the onset of degradation in the form of woody encroachment and the resultant impacts on streamflow; • the disproportionally large impact degradation related land cover change within these headwater catchments has on the downstream water balance relative to low land catchments. Following the analysis and with the understanding gained, it is recommended that the natural grasslands continue to be managed using fire, and continued protection of the natural grasslands needs to be maintained. Observation within these headwaters is key to improving the understanding of change and to drive decision making, allowing for the optimal management of the important SWSA.

Author's Publications listed on pages iv-v.