U.S. Joint Fire Science Program


Date of this Version


Document Type



Fornwalt and Kaufmann. Final report for JFSP Project Number 04-2-1-118.


US government work.


The first clear indication that unnaturally dense forest conditions existed in ponderosa pine – Douglas-fir forests of the Colorado Front Range was the Buffalo Creek Fire, a large, catastrophic wildfire that burned in 1996. Ongoing research in the Front Range indicated that the Buffalo Creek Fire likely would have burned very differently under pre-settlement forest conditions; early photographs and written descriptions, as well as fire history and stand reconstruction data, all suggested that historically these forests were characterized by a matrix of low-density forests and shrubland or grassland openings that was created and maintained by a mixed-severity fire regime. As a result of the Buffalo Creek Fire and several other large and intense wildfires in the Colorado Front Range – Bobcat Gulch and Hi Meadows in 2000, and Schoonover, Snaking, and Hayman in 2002 – managers began to plan forest restoration treatments to reduce the risk of unnaturally large, stand-replacing fires, and to return ecological sustainability to the landscape through research-guided restoration actions. Managers and others are obviously concerned about how both forest restoration treatments and intense wildfires impact the forest overstory, and these impacts are widely studied. Less studied, though, are how these disturbances impact the forest floor – in terms of both understory species composition and surface fuels that would carry a future fire. As wildfires continue to occur in the Front Range, and as forest restoration treatments continue to expand to much larger land areas, it is becoming critical for forest managers to understand how both restoration actions and wildfire influence forest understories and fuel loadings. Our objective for this project was to conduct a holistic ecosystem evaluation of the changes in forest understory communities, fuel characteristics, and potential fire behavior that result from restoration activities and wildfires. To address this objective, we established one study area within a recently treated forest, one study area within a forest recently burned by wildfire, and three study areas within untreated, unburned reference forests. All study areas were located within the Upper South Platte Watershed of the Pike National Forest, Colorado, USA, approximately 60 km southwest of Denver. Our treated study area was located in an area that was mechanically thinned in late 2002 and early 2003. The primary management objective was to reduce the risk of crown fires and to restore sustainable and ecologically appropriate overstory conditions by decreasing stand density, minimizing fuel ladders, and increasing canopy openings. Small canopy trees were sheared near ground-level using a boom-mounted hot saw on a tracked vehicle. Downed tree tops and limbs were crushed and further broken apart by driving over them with the tracked vehicle. In early 2003, where terrain and financial resources permitted, hand crews used chain saws to create additional openings and further reduce basal areas. The hand-cut trees were subsequently cut-to-length, piled, and burned. Our wildfire study area was located within the Hayman Fire, which burned 55,800 ha in 2002. Approximately half of the Hayman Fire area burned as a stand-replacing crown fire, much of it in a single day of extreme weather. The other half of the area burned as a mixed-severity fire, creating a mosaic of burn severities on the landscape. Our study area was situated in a transitional zone between these two fire behaviors, and therefore contained components of each. Three reference study areas were established near the treated and wildfire study areas to provide information about the untreated and unburned condition. The reference sites had not been disturbed by fire or other agents in the preceding five years. Each study area contained fifteen 0.1-ha plots, which were stratified by aspect (north, south, and east/west) to minimize the potential effects of topography on our findings. In each plot we measured understory species composition and cover, surface fuel loading, and overstory structure. Understory data were measured in all plots in 2004, 2005, and 2006, while surface fuels and overstory data were generally only measured in one of the three years. Prefire understory, surface fuel, and overstory data collected in 1997 were also available for the Hayman Fire plots.