Tundra fires have important ecological impacts through interactions with vegetation, wildlife, permafrost, and carbon cycling. Despite this importance, little is known about natural variability in tundra burning and links to climate and vegetation change. We used observational and paleoecological data from tundra ecosystems in Alaska to study how fire regimes varied in relation to climate and vegetation across a range of time scales. Specifically, we used boosted regression tree modeling to quantify the link between annual area burned in Alaskan tundra and summer climate over the past 60 years. Pollen and high-resolution charcoal records from lake sediments were used to reconstruct fire history over the past 2000-6000 years in two regions of northern Alaska and examine variability in past fire regimes in relation to climate and vegetation.
Results/Conclusions
Over the past 60 years growing season temperature and precipitation explained over 90% of the variability in annual area burned in Alaskan tundra. Nonlinear relationships between climate and area burned imply that tundra systems could experience dramatic increases in area burned as temperatures rise. Paleoecological data suggest that such change may be underway, as charcoal records reveal that a record-setting fire in 2007 on Alaska’s North Slope was the first such event for at least 5000 years. In other regions of Alaska climate and vegetation are more conducive to fire, and tundra has burned as frequently as modern boreal forests. In north-western Alaska, fire return intervals inferred from four high-resolution sediment charcoal records averaged between ~150 and 250 years since 6000 years BP. Contrasting temporal patterns of burning among sites suggest that local controls of fire moderated or acted independently of regional climate drivers. Pollen records from these same sites suggest that species composition influenced fire frequency over millennial time scales. Alder was consistently negatively correlated with fire frequencies, whereas spruce, grasses, sedges and birch were positively correlated with fire frequencies. Our results place modern tundra fires in the context of natural variability, provide critical fire-history information for fire managers, and suggest that vegetation change will likely interact with climate to influence the probably of burning in future tundra ecosystems.