to the present (latest fires in 1889, 1919, and 1953). 

 Plot L-2 had experienced the same fire history through 

 1919. In contrast, the Whitehorse site (L-3) had its 

 last fire in 1897. Average pre-1900 fire intervals were 

 26 years at Whitehorse and 31 years at Sawmill (fig. 2). 



Abies grandis/Linnaea borealis is the common moist 

 habitat type that supports old growth ponderosa pine 

 stands on the Bitterroot and Lolo National Forests. 

 However, most stands are unsuitable for age-class 

 reconstruction because of past logging, heavy insect 

 and disease mortality, small size, or irregular shape. 

 After considerable searching we located and sampled 

 one stand (B-4) at the 5,000-ft level on an east-facing 

 slope near Canyon Creek. This site lies immediately 

 west of Hamilton (elevation 3,570 ft) in the Bitterroot 

 Valley. Fire scar analysis indicated that pre-1900 

 fire intervals were relatively short, averaging only 

 13 years (fig. 2); this is probably due in part to a long 

 history of Indian burning in the Bitterroot Valley 

 (Barrett and Amo 1982). 



Our other moist-site plots were located in the Swan 

 Valley of northwestern Montana, on the Flathead 

 National Forest. They are small remnants of a for- 

 merly extensive old growth ponderosa pine forest situ- 

 ated in a broad glaciated valley on gently undulating 

 topography (<8 percent slopes), about 3,600 ft eleva- 

 tion. These stands occur in the Pseudotsuga menziesiil 

 Vaccinium caespitosum h.t. This and the Abies grandis/ 

 Linnaea borealis h.t. represent the most productive 

 sites that were dominated historically by serai com- 

 munities of ponderosa pine (Pfister and others 1977). 

 Western larch (Larix occidentalis) and lodgepole pine 

 (Pinus contorta var. latifolia) were often serai associ- 

 ates of ponderosa pine. We placed detailed plots in 

 the two largest intact stands and "supplementary 

 plots" with less-intensive age £md fire history data in 

 two smaller stands. Each plot was 1 to 3 miles from 

 its nearest neighbor. Fire scar analysis revealed av- 

 erage pre-1900 fire intervals of 25 to 31 years (fig. 2; 

 Freedman and Habeck 1985). Plot F-2 had an 18-year 

 average for its very brief period of record, 1817-1889. 



Methods of Stand Reconstruction 



We placed a 100-m (328-ft) square plot within each 

 small sample stand in a location where topography 

 was relatively uniform. This plot size seemed ade- 

 quate to encompass age-group patterns found in other 

 studies of stand structure (Bonnicksen and Stone 

 1981; Cooper 1960; Stephenson and others 1991; 

 West 1969; White 1985). On each plot we recorded 

 species and diameter at breast height (d.b.h.) for all 

 trees that originated before 1900, referred to here as 

 "overstory" trees. Relatively large understory trees 

 that might possibly have originated before 1900 were 

 also increment bored to ensure that all living pre-1900 

 trees were identified and sampled. The position of 



each overstory tree was recorded in Cartesiein coordi- 

 nates using 100-m (328-ft) tapes placed along the con- 

 tour (X-axis) and down the fall line of the slope (Y-axis). 

 On the relatively flat Flathead National Forest plots, 

 the X- and Y-cixes were oriented along compass beeir- 

 ings. Additional tapes were placed at 25-m intervals 

 to divide the plot into 16 "cells." Tree positions were 

 determined by sighting at right angles fi"om the near- 

 est tapes (representing the X and Y axes) to each tree 

 using a double right-angle prism or a mirror-sighting 

 hand compass. 



The uneven microtopography in plot B-4 discouraged 

 tree mapping with tapes, as described earher. Instead, 

 we mapped trees in B-4 with a survey laser instrument 

 (Laser Technology, Inc. Criterion 400). Beginning at 

 the lower left comer of the plot, we made a closed-loop 

 traverse through eight sighting points within the plot. 

 The distance and azimuth to each tree were recorded 

 from one of these eight points, allowing computation 

 of the Cartesian position of the tree within the plot. 



We chose 1900 as a basis for characterizing histori- 

 cal stand conditions because studies in ponderosa pine/ 

 fir forests generally indicate a disruption in the his- 

 torical pattern of frequent fires shortly after that time 

 (Amo 1988). Results of such comparative analyses 

 can be influenced by the year chosen to represent his- 

 toric or presettlement conditions. The bias associ- 

 ated with using a single year to typify historic condi- 

 tions is offset somewhat by sampling several stands 

 with different disturbance histories, as we did. Also, 

 records of fire history in ponderosa pine forests of 

 western Montana show a consistent pattem of frequent 

 fire from 1900 back to about 1500; beyond that thresh- 

 old few hving trees are available to sample (Amo 1976; 

 Barrett and Amo 1982). Stand structure comparisons 

 might be made for several different dates prior to 1900, 

 but the quahty of projected stand information declines 

 for earlier dates because of decay and effects of past 

 fires in consuming dead trees. 



To recreate an approximation of the circa-1900 

 stand, diameters of living overstory trees were re- 

 duced by their post-1900 radial growth (x 2 = diam- 

 eter), measured on increment borings. Also, over- 

 story trees that apparently had died since 1900 were 

 recorded by species, diameter, and position (X and Y 

 coordinates). Standing dead trees were assumed to 

 have died after 1900. Fallen trees whose boles were 

 largely disintegrated were assumed to have died be- 

 fore 1900. We verified the approximate accuracy of 

 some of these judgments by examining increment 

 cores showing growth release dates on trees immedi- 

 ately adjacent to the dead trees. We also inspected 

 the condition of trees in a similar stand that had 

 been felled and left on the ground in the 1880's. 



All "understory" trees (less than 90 years of age and 

 greater than 4.5 ft tall) in each cell were inventoried 



4 



