The derivation of fire hazard indices and burning prescriptions from climatic and ecological features of the fynbos biome

Doctoral Thesis


Permanent link to this Item
Journal Title
Link to Journal
Journal ISSN
Volume Title
This study deals with the development of fire behaviour prediction and fire danger rating systems, and their potential use in fynbos catchment areas. Fynbos catchments are managed for sustained water yields and for nature conservation. Wildfires have always been a problem in these areas, and early management consisted almost entirely of attempts to protect the vegetation from fire. More recently a policy of prescribed burning has been introduced. The analytical fire modelling approach adopted in the United States of America was identified as having potential for use in the management of catchment areas. This study was aimed at testing a number of hypotheses related to the adoption of this system in the management of fynbos areas. It further aimed to combine practical aspects of burning and fire control with conservation goals to provide meaningful burning prescriptions for the biome. The United States National Fire Danger Rating System (NFDRS) is based on Rothermel's analytical fire model. Vegetation characteristics are summarised in a fuel model and used in the NFDRS together with weather variables to produce fire danger indices. The physical properties of fynbos fuels were determined, and a fuel model was derived. The fuel model was tested in a series of experimental fires where it was found to predict fire behaviour parameters with acceptable accuracy. Climate data from nine weather stations were used in a preliminary study with the NFDRS to examine the variation in fire danger over the fynbos biome and to establish ranges of danger indices. The NFDRS reflected the expected seasonal trends in fire potential accurately, and indices compared well with actual fire occurrence. Recorded weather conditions during eight large wildfires were used to generate fire danger indices to determine the value of the various indices for reflecting fire danger. The energy release component was shown to be the steadiest indicator of high fire danger. The system shows good potential for use in the formal management of fynbos catchment areas. The NFDRS was then used in conjunction with the fynbos fuel model and climate data from 40 weather stations to define five distinct fire climate zones within the fynbos biome. The seasonal cycle of fire danger is pronounced in two inland zones, but mean fire danger shows little seasonal fluctuation in three coastal zones. Fires will occur in the coastal zones under occasional suitable conditions. These zones are useful fire management zones. The response of serotinous Proteaceae to season of fire was examined. Regeneration was best following late summer-early autumn fires, but some geographical variation was noticed, The results were used together with other examples from the fynbos biota to support seasonal, constraints on burning operations in conservation areas. Ecologically acceptable burning seasons coincide with periods of high fire danger, which I complicate the prescribed burning task. The conditions currently favoured by managers for burning operations were determined and used together with the seasonal constraints to define preliminary burning prescriptions. The seasonal occurrence of these prescribed conditions in the different fire climate zones was determined to examine the feasibility of burning in the ecologically acceptable season. While suitable conditions are rare 1n all seasons, it is concluded that burning could feasibly be carried out in ecologically acceptable seasons if an awareness of the conditions leading to safe fires is fostered among managers. The effects of invasion of fynbos by alien shrubs on fire hazard and potential fire behaviour were determined by deriving fuel models for invaded fynbos. Invasion increases fuel loads and changes the nature of fuel beds so that fires become difficult to control under extreme weather conditions. This provides a further argument for the eradication of such weeds from catchment areas. The results of this study can be used to refine fire management policies in fynbos mountain catchment areas. No previous work has quantified the seasonal and absolute variations in the physical fire environment of the fynbos biome or attempted to combine practical aspects of burning and fire control with conservation goals. The study also presents baseline data on fuels, fire behaviour and burning prescriptions. The results therefore represent a contribution to the understanding of fire as a physical factor in the management of catchment areas in the fynbos biome.