The large differences in densities between the inventory and the reconstruction based on GLO data cannot be reconciled by differences in diameter limits and timing of the two datasets. The reconstruction based on GLO data includes trees ⩾10 cm dbh; the BIA timber inventory includes

trees ⩾15 cm dbh. Trees 10–20 cm dbh contributed approximately 20% to total tree density across the entire study area of the reconstruction based on GLO data (Baker, 2012). In Munger, 1912 and Munger, 1917 trees 10–15 cm dbh were 17% of all trees this website ⩾10 cm dbh. Given these two data points, one can surmise that trees between 10 and 15 cm dbh constitute less than 20% of historical density. Hence, the difference of 5 cm in the diameter limit between these two studies does not account for the differences in estimated densities. Disturbances to the four township area between the time of the GLO survey and the time of the BIA inventory is also unlikely Cytoskeletal Signaling inhibitor to explain the large discrepancy between the reconstruction based on GLO data (Baker, 2012) and the BIA inventory of 1914–1922.

The original land survey of these four townships was conducted from 1866 to 1895 (blm.gov/or/landrecords/survey). The BIA inventory of this area occurred from 1915 to 1920, roughly 20–50 years after the GLO survey. A large decrease in density would not be expected unless the area was disturbed by logging, fire, or insect activity, but we found no evidence or record of such disturbances. In the late 1890s, a United States Geological Survey report recorded no logging in the four townships and classified 5% (1821 ha) of the area as “badly burned” (areas where at least 75% of the forest was burned within “white man’s occupancy of the region”) (Leiburg, 1900). Commercial logging began in this area in 1919 (NARA, 1955?) in an area inventoried in 1915. Stand-replacing fire effects (“no timber, old burn”) were noted on only five BIA timber

inventory transects (8 ha) in this area and these were in and adjacent to sites classified as dry and moist Shasta red fir (Abies magnifica) habitat types, not ponderosa pine or mixed-conifer sites. Abundant mortality C1GALT1 attributed to fire was recorded on another four BIA timber inventory transects (6.5 ha) in moist mixed-conifer. The BIA inventory record is consistent with Leiburg’s description of the area in 1890. Thus, it seems unlikely that disturbance between the time of the GLO survey and that of the timber inventory would explain the large discrepancy in reconstructed tree density based on GLO data versus recorded tree density in the timber inventory. Given the mean density of 60 ± 37 tph and the 95th percentile value of 132 tph recorded in the BIA timber inventory, we conclude that the Baker (2012) reconstruction significantly overestimates historical tree densities for this area.