Biophysical regulation of wildfire in the mixedwood boreal forest in the context of climate change and forest management

The dominant natural disturbance in the mixedwood boreal forests of Alberta is wildfire caused by lightning. The relative importance and interactive effect of variability in biotic (vegetation) and abiotic (weather) factors in regulating wildfire is controversial, as is the extent that recent changes in climate and/or land cover have contributed to fire activity and will continue to contribute in the future.; I use empirical data to demonstrate that patterns of lightning fire initiation are controlled by variability both in weather and forest composition that exist across landscapes in the mixedwood boreal forest. Fire initiation is reduced in landscapes with increasing amounts of aspen (Populus tremuloides Michx.) but increases with area of white spruce (Picea glauca (Moench) Voss) while accounting for variation in weather and lightning conditions. These two tree species form a principal successional trajectory in the region such that the likelihood of initiation changes through space and time from this biotic, endogenous regulation.; I demonstrate that forest management influences the fire regime we currently observe and will continue to do so in the future. Initiation increases in recently harvested landscapes and decreases in recently burned landscapes; harvesting does not emulate fire from the perspective of fire itself. I use simulation experiment to show this dissimilarity could manifest as an accelerated, compound disturbance inconsistent with the natural disturbance regime.; Climate change is projected to increase fire activity in western Canada. I use fire weather data from the Canadian Regional Climate Model with parameters estimated from observed, historical data to predict the effect of changing climate on lightning fire activity and contrast these to the outcome of simulation experiments where changes in fire weather and landscape composition ( via fire, management and stand dynamics) occur simultaneously and interactively. Simulations showed forest management increases area of young, deciduous forest that reduces long-term area burned. This reduction is attractive for climate change mitigation however forest structure in harvested landscapes differs substantially from un-harvested landscapes. I examine an alternative regeneration policy to minimize these differences. Climate change increases area burned, but not enough regenerates as deciduous forest to cause long-term self-regulating feedback to fire.