2. The crown canopy was closed or nearly closed at the time of sampling. Stumps were absent; and mortality, if evident, was due to suppression and was less than 5 percent of the plot volume. Volume growth patterns for the plot determined from stem analysis of at least 12 sample trees across the range of diameter classes indicated that the highest periodic annual volume increment had occurred within the last 5 to 15 years. These factors indicate that there was no severe competition between trees in the past.

3. The dominant trees on the plot did not contain a group of narrow annual rings, which would indicate stress in the past.

4. Trees were not infected with disease, and no visible signs of insect defoliation were present.

5. Dominant trees did not exhibit crook in the bole, and internodal lengths did not indicate past top damage. Some plots were rejected after sampling because abrupt breaks in the height growth curves suggested substantial top damage 30 or more years earlier, even though this damage was not apparent at the time of selection.

6. Clumps of trees were not sampled. The plots were in homogeneous stands, and each plot had a buffer strip equivalent in width to tree height.

Diameter at breast height (d.b.h.) was measured for each tree on each plot, and 12 to 15 trees of each species on each plot were felled to determine past periodic annual increments for the plot. Included in this group were the three largest diameter trees. If taller trees existed on the plot, they were felled and their stem analysis data used

for construction of the site index and height growth curves but not in the volume determinations. Past volume and basal area growth for nonsample trees on the plot were predicted from their basal areas. Deliberately using the tallest tree in forming the predictive equations would have produced a biased overestimate of volume growth.

The three to five tallest Douglas-fir trees at the time of sampling were sectioned at a 1-foot stump, at 4.5 feet (bh), 10 feet, and then at 10-foot intervals up the stem after total height was measured. Sections at ground line were also taken from at least two of the largest diameter trees, two of the smallest diameter trees, and one tree with close to the mean diameter. Rings were counted for all sections and recorded for the appropriate height.

Curve Construction

An age of 50 years at breast height (4.5 feet) was chosen as the index age. For each plot, heights of the three to five tallest trees were plotted as a function of bh age4 for each tree on a single sheet of graph paper. The bh age for each tree was used as the independent variable in the initial plotting rather than average bh age because height growth from 0 to 4.5 feet seems to be greatly influenced by competition immediately adjacent to a seedling and perhaps by early animal damage. Thus, it is possible for dominant trees to

be the same age at ground line but to differ as much as 10 years at 4.5 feet. Use of an average bh age for plotting heights results in an underestimate of the height growth potential of the site under management. Shifts in the tree of maximum height for its bh age among the sample trees occurred on 50 percent of the plots. Freehand curves were drawn for each tree and the highest points at each decadal age interval were used in subsequent construction of curves. Site index for each plot was defined as the tallest height at bh age 50. This procedure resembles that of Dahms (1963), except that Dahms used an average age for the plot as the independent variable in plotting the course of height growth for the tallest trees.

line of the methods of curve construction. Some understanding of curve construction leads to an appreciation of how they should be used so the appendix is recommended reading even for the occasional user.

Estimating Site Index

Site index curves are used to indicate the potential productivity of land. The curves here give the estimated height of the tallest tree when the breast high age of that tree is 50 years. Many of the same plot qualifications used in this study are applicable in selecting plots for measuring site index. The following steps and procedures are recommended for estimating the site index of a managed stand.

1. Select suitable plots with the following characteristics:

(a) Even-aged at the ground line (practically, there are no older remanents from earlier stands and the present stand is one storied).

(b) No visible signs of disease or insect attack to reduce height growth.

(c) No narrow ring groups to indicate suppression.

(d) Consistent internodal lengths on taller trees.

(e) No remnant understory vegetation or suppression mortality to indicate competition early in the life of a stand.

2. Establish boundaries of a 1/5-acre plot with a prespecified shape.

3. Measure the height of the three to five tallest trees on the plot.