History and Management of Crown‐Fire Ecosystems: a Summary and Response

Some ecosystems, such as yellow pine forests, have hada long history of frequent surface fires, but because offire suppression policy, fires have been largely excludedfrom them during the last century (Covington 2000). Un-natural fuel accumulation in these forests has increasedthe potential for large, catastrophic crown fires, and re-introduction of prescribed fire is one remedy for thiscritical fire hazard. But fire ecologists and fire managersneed to be cautious in transferring this model to all westernecosystems (Anderson et al. 1999; Gutsell et al. 2001). Al-though large, catastrophic crown fires are apparently un-natural in yellow pine forests (but cf. Shinneman & Baker1997), this is not so in other western forests and shrub-lands, and widespread prescription burning is not war-ranted everywhere.Johnson et al. (2001 [this issue]), illustrate how this yel-low-pine model has been inappropriately applied to theboreal forests of Canada, where crown fires are an inevita-ble consequence of fuel structure and burning is not age-dependent. Thus, creating a landscape-age mosaic withprescription burning will not reduce the incidence ofcrown fires in these forests (Johnson & Miyanishi 1995).The yellow-pine model has also been misapplied to thechaparral shrublands of southern California, which hasled to the erroneous conclusion that large crown fires inthese ecosystems are a modern artifact of fire-suppres-sion policy (Minnich 1983, 1989, 1995, 1998). However,large crown fires predate fire-suppression activities inCalifornia (Keeley & Fotheringham 2001 [this issue]) andother shrublands (Pyne 1991; Gill 2000). Also, it hasbeen argued that the solution to preventing large, cata-strophic fires is the use of widespread prescription burn-ing to create a landscape mosaic of different-aged patchesof vegetation (Minnich 1989, 1995, 1998). This fuel-age/mosaic model does not fit California shrublands, how-ever, because fire-suppression policy has not resulted infire exclusion, and there has been as much or more areaburned by wildfires in recent decades than before activefire suppression (Moritz 1997, 1999; Conard & Weise1998; Keeley et al. 1999). Failure to effect fire exclusionresults from the fact that fire management is challengedwith an ever-increasing rate of fire incidence which par-allels the exponential rate of human population growth(Keeley 2001) in an environment with the worst fireweather in the country (Schroeder et al. 1964). An im-portant consequence is that there has not been an unnat-ural accumulation of fuels. Recent studies show that firehazard is either independent of age (Moritz 1999) or onlyweakly dependent up to 20 years of age (Schoenberg etal. 2001); large, catastrophic fires will readily burnthrough young stands and do not require old vegetation(Fig. 1). In short, there are sufficient data to refute thecontention that chaparral “fire occurrence is constrainedin space and time by the rate of fuel accumulation and byprevious fire history” (Minnich 2001 [this issue]). Anyconstraints are weak at best.The theoretical basis behind the fuel-age/mosaicmodel lies in earlier modeling work done on Californiashrublands (Rothermel & Philpot 1973; Philpot 1974).Based on fire-spread models, it was concluded that aschaparral stands increase in age, there is a resultant in-crease in fuels, rate of fire spread, and fire size. Follow-ing suggestions by Countryman (1974), these modelswere interpreted to support a fire-management policythat relied heavily on prescription burning to producea landscape comprising a mosaic of age classes. Thethinking was that as fires burn across a landscape andencounter patches of younger age classes, they eitherdie out because of insufficient fuels or their spread