Shrub fuels are capable of extreme fire behaviour under conditions that are often moderate in other fuels. There is also a narrow range of conditions that determine fire success in these fuels, below which fires may ignite but hardly spread and above which they ignite and develop into fast moving and high intensity fires. This is due to the elevated dead fine fuels that dry rapidly and carry fire. Fire danger rating systems designed for forest and grassland fuels do not predict fire potential in shrub fuels very well. Fire management requires fire danger rating systems to provide accurate and timely information on fire potential for all important fuel types. Studies of fuel moisture, ignition and fire spread were carried out in the field in gorse (Ulex europaeus L.) shrub fuels to predict the moisture content of the elevated dead fuels and to define the conditions that govern fire development. The accuracy of the Fine Fuel Moisture Code (FFMC) of the Canadian Forest Fire Weather Index (FWI) System to predict moisture content of this layer was assessed. A bookkeeping method to predict moisture content was developed based on semi-physical models of equilibrium moisture content, fuel response time and the FFMC. The FFMC predicted moisture content poorly, because the FWI System is based on the litter layer of a mature conifer forest. The gorse elevated dead fuel layer is more aerated and dries faster than this conifer forest litter layer. The bookkeeping method was reliable and allowed adjustment of fuel response time based on weather conditions. Difficulties in modelling meteorological conditions under the gorse canopy limited its accuracy. Separate thresholds determined ignition and fire spread success, with both based on the elevated dead fuel moisture content. Options to improve the shrub fire danger rating system were presented based on these findings. The results are significant because they are based on data collected in the field under real conditions. Validation of these results and extension to other shrub fuels is required before the findings are used to change current models. However, the study has significantly advanced the knowledge of fire behaviour in shrub fuels and will contribute to safe and effective fire management in these fuels.

For two fire seasons in Nordegg, Alberta, a system of in-stand weather stations were arranged along a north and south aligned valley and combined with collection of destructive fine fuel moisture content data in order to quantify variations due to differences in slope and aspect. South-facing sites were found to be slightly warmer (1.5°C), less humid (5%) and received on average 20% more solar radiation than the north-facing sites during the peak burning period of the day. Based on these weather observations a difference of 1 or 2 % moisture content between north and south sites was predicted using existing theoretical relationships. A corresponding difference in observed moisture content was not identified, due to the low transmittance recorded at the in-stand sites (<10% of open solar radiation measurements), variation amongst destructive samples and logistical limits on the number of replicates collected.

For two fire seasons in Nordegg, Alberta, a system of in-stand weather stations were arranged along a north and south aligned valley and combined with collection of destructive fine fuel moisture content data in order to quantify variations due to differences in slope and aspect. South-facing sites were found to be slightly warmer (1.5°C), less humid (5%) and received on average 20% more solar radiation than the north-facing sites during the peak burning period of the day. Based on these weather observations a difference of 1 or 2 % moisture content between north and south sites was predicted using existing theoretical relationships. A corresponding difference in observed moisture content was not identified, due to the low transmittance recorded at the in-stand sites (<10% of open solar radiation measurements), variation amongst destructive samples and logistical limits on the number of replicates collected.