U.S. Joint Fire Science Program


Date of this Version


Document Type



Final Report: JFSP Project Number: 06-3-1-17


US government work.


Knowledge of past fire regimes associated with mountain big sagebrush-dominated landscapes is inadequate for scientifically-based land management that requires assessment of departures from historic conditions. Widely utilized estimates of fire frequency for sagebrush ecosystems are largely based upon few studies using fire-scarred proxy trees positioned at the forest/shrubland ecotone. These studies, all conducted in the northern half of the species distribution, generally fail to adequately address questions of fire behavior across the fuels threshold at the forest/woodland-shrubland ecotone. Alternatively, post-fire rates of succession have been used to suggest fire frequencies compatible with big sagebrush recovery. Minimum and maximum fire-free intervals can be inferred based upon the time required for big sagebrush population recovery and succession to tree dominance, respectively. Published studies of mountain big sagebrush post-fire recovery are also limited primarily to higher latitudes and seldom consider the long timeframes required for trees to establish in burned areas on semi-arid landscapes. We had three objectives to address these deficiencies: (1) We developed estimates of historical fire frequency for mountain big sagebrush communities at 10 sites in the eastern Great Basin, upper Colorado Plateau and intervening mountains and highlands (southern half of the species distribution) using fire chronologies from proximal fire-scarred trees. Proxy-derived estimates were evaluated based upon proximity of fire-scarred trees to mountain sagebrush communities. We assessed the likelihood that these estimates accurately reflect historic fire frequency for mountain sagebrush communities across the study region. (2) We developed field-based estimates of recovery rate for mountain big sagebrush using a chronosequence approach (n = 27 burn sites of known fire year distributed throughout western and central Utah). Recovery time extrapolated from regression models averaged 37 years. The precipitation regime immediately following the fire year proved the single most important predictor of recovery. Recovery rate was positively associated with precipitation in the summer before fire and also with course textured soil. (3) We developed a grid-based model of fire and sagebrush recovery via seed dispersal, establishment probability, and time to reproductive maturity to simulate the predicted long-term response of mountain big sagebrush to alternative fire regime scenarios. Model scenarios included a range of fire frequencies and fire extent. The response variable of interest was the percent of the virtual landscape dominated by mountain big sagebrush. Model outputs were highly sensitive to mountain big sagebrush life history trait parameters. Dispersal distance and time to reproductive maturity parameters were evaluated for their influence on model output. Slow rates of recovery were associated with the shortest dispersal distance (10 m) and the longest time to reproductive maturity (4 yrs). Fire regimes characterized by small fire sizes (e.g. 2 ha) are analogous to fire regimes with larger fires that are patchy and that leave unburned islands every 2 ha with a viable seed source. In this regard it was redundant to model establishment from seedbank that survived the fire as this would effectively reduce the influence of fire size on persistence of sagebrush—our main response of interest. Under fire rotations commonly cited as typical of mountain big sagebrush (i.e. 30-80 years) (Baker 2006, 2011), simulations revealed that long-term convergence on 20% composition or greater is only compatible with small fire sizes, smaller than 5 ha for 30 yr rotations and smaller than 30 ha for 80-yr fire rotations.