Accurate estimation of mean fire interval for managing fire

Abstract

Accurate fire-history data are needed if local management of fire or costly national plans for restoring and managing fire and forest structure are to succeed. Fire-history researchers often use fire scars and the composite fire interval method to reconstruct parameters of past fire regimes, such as the population mean fire interval, but the composite method has serious limitations. We modified an alternative non-composite fire interval method, the individual-tree fire-interval method, to derive a more accurate new method, the all-tree fire-interval method. A stochastic fire-scar generating model to assess the accuracy of the new method and its predecessors was then used. Three factors (scarring ratio, population mean fire interval, and tree age) that affect accuracy were varied in the model runs. More complexity (trees with varied scarring ratio between the first scar and successive scars) also was modelled to test the robustness of the method. The all-tree fire-interval method was shown to greatly improve accuracy and provide unbiased estimates of the population mean fire interval. The method also produced encouraging results when scarring was more complex. The new all-tree fire-interval method will require further research on the rates at which trees are scarred by fire, but this would be generally beneficial to understanding fire history.