western larch and 10 percent ponderosa pine. Tree ring 

 counts indicated a stand age of about 70 to 80 years. 

 The understory was layered. The tall shrub layer con- 

 sisted of Rocky Mountain maple {Acer glabrum) and 

 some serviceberry (Amelanchier alnifolia). Below these 

 arborescent shrubs were lower shrubs including ninebark 

 (Physocarpus malvaceus) and white spirea {Spiraea 

 betulifolia), with a sparse herbaceous layer of heartleaf 

 arnica {Arnica cordifolia), pinegrass {Calamagrostis 

 rubescens), sidebells shinleaf {Pyrola secunda), western 

 rattlesnake plantain {Goodyera oblongi folia), and western 

 meadowrue {Thalictrum occidentale). There was also a 

 well developed moss layer and a large amount of litter 

 on the forest floor. 



Fire Frequency 



A fire history study for the north slope of Pattee Can- 

 yon, including the area burned in 1977, is in progress 

 (Habeck, J. R.; unpublished data). Fire scars, mostly 

 from ponderosa pine snags and stumps, provide the 

 basis for the fire chronology. Nine samples, with five to 

 nine fire scars each, have been analyzed using the 

 methods described by Arno and Sneck (1977). Between 

 approximately 1700 and 1900, the sampled area of 

 Pattee Canyon experienced fires severe enough to scar 

 the trees at an average interval of 15.8 years. Some fires 

 scarred only one of the sample trees and thus may have 

 been limited in extent or may have burned with such a 

 low severity as not to scar other trees. It appears to 

 have been more common for fires to cover large areas of 

 the north slope or even the entire north slope because 

 trees 2 or more miles apart exhibit similar fire dates 

 based on a preliminary analysis of the data. During the 

 century prior to settlement, 1780 to 1880, widespread 

 fires occurred on an average of one per decade. The fire 

 frequency in Pattee Canyon compares well with the 

 mean fire- free interval of 13 years for the period 

 1600-1910 on similar sites in the neighboring Bitterroot 

 Mountains (Arno 1980). 



Prior to settlement, fires in western Montana resulted 

 from summer hghtning storms and Indian ignitions 

 (Barrett 1981; Barrett and Arno 1982). Barrett's study 

 and Arno's earlier (1976) study suggest that most of 

 these fires were ground fires of low-to-moderate severity 

 and frequent fires precluded heavy accumulations of fuel. 



Based on fire scar analysis, tree aging, and an inter- 

 pretation of the original Federal land survey records 

 (Habeck, J. R.; unpublished data), the presettlement 

 forest in Pattee Canyon appears to have been main- 

 tained by fire as an open, nearly parkhke forest, with 

 ponderosa pine on drier ridges, western larch in moist 

 areas, and a scattering of mature Douglas-fir. Fire was 

 effectively eliminated from Pattee Canyon throughout 

 much of the 1900-77 period; selective logging removed 

 the largest trees; and the open forests changed to dense- 

 ly stocked, closed-canopied stands easily capable of sup- 

 porting high intensity crown fires. 



The Fire 



The Pattee Canyon fire started in dry grass near the 

 west end of Pattee Canyon at about 4:20 in the after- 

 noon July 16, 1977. Fire behavior was influenced by ex- 

 treme fire weather: high temperature, low relative 

 humidity, and strong, gusty winds. The Fire Danger 

 Rating was, consequently, very high. The wind-driven 

 fire spread quickly up the grassy slope and into adjacent 

 conifers, where it crowned almost immediately. The fire 

 then began its main run up Pattee Canyon toward the 

 southeast with its smoke column pushed horizontal. The 

 fire was extremely intense at this point and numerous 

 spot fires started far ahead of the front (Jerry Williams, 

 Montana State Division of Forestry, 1981; personal com- 

 munication). Self-induced winds, along with the fire's 

 direction up the hill, kept the fire away from the floor of 

 the canyon and the many homes located there. As the 

 fire burned across the north slope ravines, each ravine in 

 turn acted as a chimney, funneUng the fire up the 

 ravine. Near the upper ridge the fire met gradient winds 

 blowing the fire back upon itself. When the fire reached 

 its eastern boundary, it met the developed spot fires. 

 These spot fires then became backfires. The fire com- 

 pleted its main run by 5:15 p.m. (fig. 3). Fire suppression 

 efforts were then directed at putting out spot fires. The 

 fire was declared contained the second day (July 17) and 

 controlled the third day (July 18). 



The fire burned about 1,200 acres (486 ha) and 

 destroyed six homes (Fischer and Books 1977). It burned 

 with greatest severity across sections 4 and 3 and the 

 lower two-thirds of section 10 (fig. 3), with fire whirls in 

 the western half of sections 3 and 10. In the southern 

 (upper) portion of section 10 fire treatment was inconsis- 

 tent, with vegetation in some sectors burned severely. In 

 other areas the overstory was killed but the needles and 

 small twigs on the trees were not consumed and the duff 

 was not entirely removed from the soil. In section 11 the 

 fire consisted of numerous spot fires, which collectively 

 were much less uniform and less intense, leaving partial- 

 ly scorched crowns in some areas while confined to the 

 ground elsewhere. 



A postftre analysis of fire intensity in Pattee Canyon 

 (Rothermel 1977) was divided into two parts: probable 

 fire behavior in the grazed grass at the ignition point, 

 and probable fire behavior in surface fuels under the 

 timber (table 2). Crown fire intensity was not assessed 

 but is estimated to have been at least tenfold that of the 

 surface fire (Rothermel 1977). 



Postfire Treatments 



Immediately following the fire a team of Federal and 

 State specialists inspected the burn and prescribed 

 seeding with grass as a watershed rehabilitation 

 measure. Aerial seeding was accomplished under the 

 direction of USDA Soil Conservation Service during 

 November while the area was snow covered. One private 

 parcel on the west side of the fire was not seeded at the 

 owner's request (fig. 1). The seed mixture was varied ac- 

 cording to elevation (table 3). 



4 



