An analysis of selected influencing factors of wildfire regimes over the last three decades in the Boland Mountain Complex, South Africa

Abstract

This study examined the factors influencing wildfire regimes in the Boland Mountain Complex (BMC), Western Cape Province, South Africa, for over thirty years. The fynbos region, covering the BMC, is one of the world's six floral kingdoms with high levels of endemism and is both fire-adapted and fire-dependent. Inappropriate fire regimes are a considerable concern for biodiversity loss, and as such, it is crucial to understand the influencing factors to facilitate decision-making. This study aims to identify and analyse the key anthropogenic, environmental and climatic factors contributing to wildfire frequency, size, seasonality and fire return intervals over time. A mixed-methods approach was employed to evaluate various relationships by utilising a comprehensive dataset comprising fire records, including global position information, satellite imagery, land use data, and meteorological records. The results showed an increase in fire frequency and cumulative yearly hectares burnt. The cumulative hectares have a very highly significant statistical link to the number of Very Large and Large category fires. Therefore, in order to reduce fire size, the importance of the speed and scale of the suppression response must be highlighted. Increased fuel load caused by the spread of invasive alien plants (IAPs) and changes in land use as well as climatic changes mainly an increase in temperate and drought conditions facilitate more intense fire behaviour. Additionally, urban expansion have intensified the risk of fire ignitions but could also lead to an increase in reporting biasness. From a space-time composite perspective, the fire return period had an average fire return interval of 8.63 years, which is currently not conducive to maximising biodiversity. The proximity of ignitions was examined near the closest towns and transport routes, with Grabouw and track footpaths having the highest frequency. The hectares burnt for each Very Large category fire has a very highly significant statistical link to biomass and total hectares burnt in the BMC. None of the weather-related variables showed any notable links between the dependent variable of total hectares burnt per Very Large fire category. This could be due to distance from weather stations and variations in microclimates within mountainous areas. There is, however, a notable link between antecedent annual rainfall and the fire frequency and cumulative hectares burnt in the BMC for the same year. This research shows an overall trend of increased fires, highlighting key risks and specific geographical areas as hotspots. It is recommended that results from long-term monitoring, research, and input from all stakeholders must be regularly assessed and reported back into legislation, prediction, prevention, preparedness, suppression, rehabilitation recommendations, and planning. Integrated wildfire management should also be viewed holistically, considering all stakeholders' needs and capacity, coordinating all stakeholders' management efforts, and continually practising adaptive management.