An integral plume model was applied to the problems of tree death from canopy injury in dormant season hardwoods and branch embolism in Douglas fir crowns.  Our purpose was to generate testable hypotheses.  For dormant season hardwoods, effects of the plume on the crown were modeled by a branch necrosis routine involving heat transfer and thermal tolerance.  If a stem was predicted to survive stem heating from flames, aggregate branch death from plume heating determined proportional crown loss which, in turn, determined tree allocation to spring refoliation.  Hardwood mortality occurred as a function of subsequent growth.  Branch embolism in Douglas fir is hypothesized to occur in response to foliage exposures to high vapor pressure deficits within the plume occurring on a time scale too short for stomatal response. 

Results/Conclusions

For hardwoods, differences among species in bark thickness led to large differences in susceptibility to modeled stem girdling from flames.  For trees predicted to survive flame effects, branch diameter contributed significantly to differences in height of branch kill.  A parametric equation based on multiple model simulations is provided to describe branch kill height as a function of fire behavior and crown characteristics.  For Douglas fir, we speculate that branch embolism resulting from exposure to the plume may be a pervasive effect of surface fires on Douglas fir during periods when stomata are open.  Cuticular water loss was not sufficient to cause embolism.  The models were used to explore the effects of variation in fire behavior and to relate crown injury to bole injury.