Simulations of historic and future vegetation and fire dynamics in sub-alpine Tasmania, Australia, using FireBGCv2: Insights from aligning simulation results with historical proxy records

Background/Question/Methods

Anthropogenic global change raises three major research questions about the influences of climate change and human land use and management on vegetation and fire regimes: (1) How have humans and climate influenced historic fire regimes? (2) How has recent climate change altered fire regimes? (3) Can knowledge of historical controls over fire regimes inform and improve fire management? Comparisons across gradients of climate and human land use are central to answering these questions. Research is now underway to do so for the western United States and Cradle Mountain National Park, Tasmania, Australia. In these locations, we employ three different types of evidence. Sedimentary pollen and charcoal records allow us to reconstruct millennial-scale fire and vegetation history at a coarse temporal grain. Tree-ring records allow us to reconstruct fire and vegetation history at a finer temporal grain. FireBGCv2 simulations then allow us to project fire and vegetation dynamics under a variety of future scenarios, while the two bodies of historical evidence allow us to independently parameterize and validate FireBGCv2. Unifying proxy records of fire activity, contemporary observations, and future projections under a wide variety of climate and land-use scenarios will further incorporate fire science into land use and management decisions.

Results/Conclusions

Recent fieldwork at Cradle Mountain National Park in Tasmania, Australia, established 22 plots to assess current vegetation and fuel loads. Data from these plots allow us to apply FireBGCv2 to a unique study site that has had a greater human influence on fire regimes than comparison sites in the western United States. Our fieldwork identified 12 different plant community types. Analysis indicates substantial disagreement between our field data and the most widely used remotely-sensed vegetation data. Parameterizing the dominant tree species – which are new to this modeling system – was a key challenge. An experimental design of five climate scenarios fully crossed with five management scenarios allows us to address our three major research questions consistently across gradients of historic climate and human land use in several FireBGCv2 study sites. For the study landscape in Tasmania, Australia, where competing hypotheses suggest different mechanisms of vegetation change, model results support an active human role in shaping vegetation across the landscape.