Climate change and land management practices are causing profound changes in wildfire dynamics, particularly in coniferous forests of western North America. As a result, there is concern regarding the resilience of these forests to future fire regimes. Historical fire regimes, defined in terms of fire history studies and biophysical models, offer reference conditions against which contemporary and future fire regimes can be compared. However, spatial models of historical fire regimes lack an explicit link to resilience because they do not consider the functional traits of predominant species that determine tolerance (i.e. survival) of fire. Different tree species vary along multiple, often-correlated dimensions of fire-tolerance traits. Here we introduce a trait-based approach for spatially explicit modeling of fire regimes based on the adaptive capacity of trees to withstand fire, focusing on two correlated traits of bark thickness and tree height. We specifically ask whether biogeographic variation in these fire-tolerance traits aligns with our understanding of fire regimes based on biophysical models from the LANDFIRE program. Using information on functional traits compiled from the literature, we evaluated the fire-tolerance trait landscape across conifer forests of California and compared the mean trait values within different biophysical fire regime groups.
Results/Conclusions
Among 23 conifer species found in California, bark thickness and maximum tree height are significantly positively correlated (r=0.776, P<0.001), with thick-barked and tall species having greater fire tolerance (i.e. survival ability) than thin-barked and shorter species. The community-weighted mean trait value for both bark thickness and tree height were significantly greater in forest stands where prior biophysical modeling studies suggest frequent (<35 year) fire-return intervals (FRI’s) (mean bark thickness=1.54 cm, mean tree height = 51.4 m), compared to forests with moderate FRI’s (35-200 years, mean thickness = 1.01 cm, mean height = 31.0 m) and forests with infrequent FRI’s (>200 years, mean thickness = 0.96 cm, mean height = 31.9 m; all P<0.001). These results indicate that applying functional trait-based analyses across a landscape can yield valuable insight into the resilience of forests to changes in fire frequency and intensity. Furthermore, these continuous traits allow descriptions of fire regimes along continuous gradients of adaptive capacity rather than arbitrarily defined categories, which should greatly increase our spatial resolution for determining priority forest restoration areas and forecasting future resilience in fire-prone forests.