Establishing fire regime characteristics for biomes in southwestern Africa

Abstract

Fire has been occurring naturally in all ecosystems on earth before human intervention. Natural fires play a crucial role in the removal of plant litter and regeneration of certain plant species. Vegetation and wildlife mortality from fires affects species composition, habitat quality, biodiversity, seed distribution and migration patterns. In addition, human health, lives, and livelihoods can be severely threatened by fires. Studying fires requires knowledge of fire regimes, which consists of factors such as seasonality, fire intensity, fire frequency, vegetation characteristics, topography, synoptic weather patterns, climate change, soil moisture availability and land use and land cover change (LULCC). Due to the projected climatic changes and LULCC forecast for the western parts of southern Africa, it is crucial to establish the fire regime characteristics for biomes in the region. This study aimed to establish the fire regime characteristics for the biomes in southwestern Africa. Open access data were used with the exception of local weather station data obtained from the South African Weather Service (SAWS). In addition, open-source software packages for example, QGIS and selected Python libraries, were used to extract, analyse, and interpret data. Long-term datasets were extracted and analysed as well as data for five selected case studies in order to determine the relationship between synoptic weather types and fire events. In addition, the pre-fire vegetation conditions and (post-fire) burn severity and burnt areas were also examined. Fire regime characteristics found in this study related to climatic variables, and vegetation conditions prior to fire events. The remotely sensed data identified the seasonality of fires at Knysna to occur in the austral summer, at Cape Town in the austral summer and early autumn and the South-Eastern (SE) Kalahari in the austral winter and early spring. The seasonal occurrences of fires correspond with the literature on the dry seasons across the Fynbos, Savanna and Grassland Biomes. In terms of Fire Return Interval (FRI), the analysis of official records found that the two largest two fires occurred 30 years apart. While irregular FRIs were detected in the official records, the FRIs still correspond with the literature indicating a 2-10 year and 5–50-year FRI in the Fynbos biome. The majority (53%) of the largest fire events in the Western Cape occurred in the years following the El Niño phase of the El Niño Southern Oscillation (ENSO), positive phase of the Southern Annular Mode (SAM) (66%) and below normal seasonal precipitation anomalies (66%). Furthermore, the simultaneous occurrence of these teleconnections and precipitation anomalies were documented for six (40%) of the fire events, with four (27%) of the events took place after La Niña phases of ENSO, negative phases of SAM and below normal seasonal precipitation anomalies. Strong high-pressure (HP) systems were present on the day of ignition across southern Africa for all five of the case studies, and in the 2017 Knysna fire case a strong frontal low-pressure (LP) system had

also been present. In terms of fire extinction, all the fire events, with the exception of the 2021 Cape Town fire, were affected by the presence of a LP system over the region. Soil moisture volumes were mostly low prior to fire ignition, with the exception of the 2018 Knysna fire where soil moisture volumes were slightly higher than at the other sites. Vegetation was determined to be moderately healthy prior to the Knysna fires and the 2021 Cape Town fire. In contrast, vegetation was under stress prior to the 2015 Cape Town and SE Kalahari fires. The 2021 Cape Town and 2018 Knysna fire events recorded high burn severity, with the latter also classified as the largest fire in terms of burnt area. A moderate-low burn severity was documented for the 2015 Cape Town fire, while the 2017 Knysna and SE Kalahari fires were classified as low severity fires. The low severity fires consumed more hectares than the two Cape Town fires. This thesis contributes to the body of knowledge in the discipline of geography by highlighting the interconnected factors that contribute to fire events, including their occurrence over various spatial and temporal scales.