Abstract
Climate change impacts the boreal ecosystem through modifications of vegetation structure, composition, distribution and productivity. These changes alter the internal functioning of forests by disrupting the dynamics of natural disturbances such as fire, notably their frequency and size. Extreme weather events in recent decades have resulted in very large areas burned during some years, altering forest landscapes at northern latitudes in Canada. Large fires accounting for the majority of burned areas release massive amounts of carbon into the atmosphere and have major health consequences for people in exposed communities. They also limit the ability of Indigenous people to maintain their traditional activities by reducing ecosystem services they traditionally access. Climate projections suggest that large fires may be increasingly frequent in the coming decades, which in turn could affect forests, climate, and human societies. However, models remain uncertain and questions persist, particularly about the frequency and magnitude of extreme weather events that facilitate the onset of large forest fires.This thesis contributes to improving our understanding of the environmental drivers that have determined the dynamic of recent and past fire regimes in the boreal forest of northwestern Canada, in order to predict the interactions between climate, vegetation and future wildfires. This information will help managers and local communities anticipate future fire risk in response to climate change, in order to adapt practices and land use accordingly to limit the potentially negative effects of large forest fires. Chapter II of this thesis aimed to reconstruct the fire regime since 1965 on the territory of the Tłı̨chǫ; First Nation in the Northwest Territories (NWT), in Canada. This territory experienced one of the largest forest fires across Canada in 2014. We determine the main climatic and ecological conditions having contributed to extreme wildfire seasons (EWY) during the past few decades on the Tłı̨chǫ; territory, and we verify that the year 2014 was one of those EWY. We fixed climatic risk thresholds conducive to EWY occurrence, based on temperatures and soil organic horizon dryness indices measured during the fire season. The objective of the chapter III was to reconstruct the fire regime and vegetation dynamics on the Tłı̨chǫ; territory during the Holocene (11,700 years) and to analyze our reconstructions with regards to past climatic conditions at regional scale, in order to determine the main drivers of fire occurrence and size over time. Historical reconstructions showed that warm and dry conditions, as well as fuel availability and forest composition, were the main drivers leading to large and severe wildfires in the short and long term. In chapter IV, we simulated the risk of future EWY based on projections from different climate models coupled with two radiative forcing scenarios (RCP4.5 and RCP8.5). We used the LPJ-LMfire model to simulate the dynamics of tree biomass and burn rate over the 21st century in the NWT and the Tłı̨chǫ; territory. Model simulations showed that future burn rates will be primarily modulated by increasing temperatures and changing vegetation dynamics, particularly black spruce.