Development of current stand structure in dry fir-pine forests of eastern Washington

Journal of the Torrey Botanical Society, Apr-Jun 2007 by Everett, Richard, Baumgartner, David, Ohlson, Peter, Schellhaas, Richard, Harrod, Richy

Two circular plots (0.04 ha) were randomly placed within each stand to sample all live trees and small (

Larger 0.4 ha belt transects were placed within these same stands to record species, numbers, and decay class of snags, and logs greater than 22.9 cm base diameter. A total of one hundred and ninety 0.4 ha plots were established at a rate of one plot for every 8.5 ha of sampled area, as recommended by Bull et al. (1990). Snag and log data from the two plot sizes were combined on a per ha basis for each stand and summed across stands within each individual plant association group. Conversion factors for estimating dbh from the basal diameter of logs were derived by comparative sampling of these measurements on live trees of Pinus ponderosa, Larix occidentalis, and Pseudotsuga menziesii.

Increment cores, within the 0.04 ha plots, were taken at breast height (1,192 total cores) for all live trees greater than 5 cm dbh. A subsample of 1 60 trees was cored at stump height (30 cm). The early growth rate from stump to breast height (30 cm to 1 .4 m) was used to estimate time required for growth from germination (0 cm) to stump (30 cm) height (Gordon Nigh, Research Branch, B.C. Ministry of Forests, pers. comm.). The age to reach breast height was computed as the sum of the time from stump to breast height plus the estimated time from germination to stump height. This germination to breast height time was added to the age at breast height to estimate total age for live trees, snags, logs, and their establishment dates. Basal area (0.0007854*(dbh {cm})2 was computed for all living trees that exceeded breast height.

Age at mortality for snags and logs was estimated from the dbh/stump diameter-total age relationship of live trees on the same plots (n > 1000). Derived linear and/or curvilinear regressions were specific to each tree species, to each plant association group, and sometimes to specific stands, and plots as required for an acceptable r^sup 2^: (average and median r^sup 2^ = 0.76 Pseudotsuga menziesii, and r2 = 0.71 Pinus ponder osai Larix occidentalis), lack-of-fit test, and an equally distributed plot of residuals.

Time since mortality for the 3600 snags and logs on the plots was estimated from their decay state. The time required to reach each of five sequential decay states (Cline et al. 1980) was previously determined for Pseudotsuga menziesii and Pinus ponderosa from an 8 1 -year chronosequence of fires on the east slope of the Washington Cascades (Everett et al. 1999). Regression equations (r^sup 2^ = 0.77 to 0.94) used to predict years to reach the snag and log decay states were as follows: Pseudotsuga menziesii snags [years = (-2.052 2.2431*(decay state))2 and logs [years = (0.102 1 .99949*(decay state))2, Pinus ponderosa snags [years = (-0.988 1.9325*(decay state))2 and logs [years = (0.075 1.5254*(decay state))2. The plot of residuals indicated sampling errors were distributed randomly along the regression line. The median time required to achieve a specific decayed state for a snag or log was added to tree age at mortality to estimate the approximate time of tree establishment. The accuracy of this composite estimate was constrained by potential errors in estimating tree age at mortality and time to reach decay states, but our methods should have ameliorated this problem.