Articles of Interest

Results of a Branch Measurement Trial on 2 SMC Installations Stand Management Cooperative Wood Quality Measurement Cooperative

Dave Briggs, Director; Bob Gonyea, Program Operations Manager

In early 1998, the SMC Wood Quality TAC established a committee to suggest procedures for routine measurement of stem and wood quality on SMC installations. Here "stem quality" refers to bole or log characteristics such as branches, sweep, crook, etc. while "wood quality" refers to wood density, fiber length, microfibril angle, etc. This article focuses on the branching aspect of stem quality.
Anecdotal reports from the field suggest that differences in branch size are developing among treatments on SMC installations. The Wood Quality TAC felt that a procedure should be devised to measure branch characteristics and formally test for differences due to treatments. It was also felt that the procedure should become a regular part of data collection when installations are remeasured in order to gain insight on how branch characteristics change through time as stands evolve under the various regimes. This could be important if branch differences affect potential log and product quality or value sufficiently to alter conclusions one would get based solely on growth performance.

PROCEDURE

The suggested branch measurement procedure was designed to meet the following criteria

• allow data to be collected on a reasonably large sample of trees within each treatment
• be convenient to collect with minimal impact on field crew time
• be repeatable

The procedure obtains the following data on the first whorl above breast height:

• Diameter of the largest branch in the whorl.
• Number of branches in the whorl
• Number of branches in the half-internode above and below this whorl, provided that they are at least ½ the diameter of the largest whorl branch

These data will be collected from the height trees on each plot, typically at least 40 trees. When these trees are being measured for height, the person standing at the tree being measured can easily gather and record the branch data while the other person measures height.
The first whorl above breast height was chosen since it is easily reached and such branches are unlikely to have been damaged or broken by personnel accessing the bole to measure dbh. Counting internodal branches was restricted to those that are at least one-half the diameter of the largest whorl branch in order to simplify counting by excluding the many small "whiskers" and pin knots that often disappear and focus on those likely to remain and become large enough to impact products. Keeping the two counts separate will provide information on relative rates of disappearance of whorl and internodal branches as the tree ages.
It was also felt that it would be desirable to obtain more detailed branch data along the bole of a subsample of trees with the objective of establishing relationships to the bh measures. Some of this data could be gathered for a portion of the lower bole when installations are visited and could be gathered from thinnings. This would be done as time permits. In fact, some of this data already exists within branch data sets collected previously by SMC researchers.

FIELD TEST

To test the procedure and gain an understanding of its impact on field crew time, a Type I and a Type III Installation was measured during August 1998. In addition to collecting the basic data, the following data were also obtained:

• After selecting what was believed to be the largest whorl branch and measuring it's diameter, the diameters of the remaining whorl branches were also measured on half of the sample trees. This gives a sense of the distribution of branch sizes in relation to the largest and provided a means to test how often a person can select the largest whorl branch.
• Rather than take a single diameter on each branch, diameter was measured both parallel and perpendicular to the stem height axis which corresponds to knot diameter across the grain and with the grain in lumber
• An indication of whether the branch was alive or dead. In practice, measurement of height to the base of the live crown provides this information.
• For a subset of trees on each plot, the largest branch diameter and count data were obtained for a series of whorls above and below BH. This was done to gain additional insight on the effects of obtaining this detail on field crew time and to gain some preliminary insights on trends within stems receiving different treatments.

SAMPLE INSTALLATIONS

A Type III installation Located in Pierce County, Washington on Champion International land that was planted in 1986 with 2-0 stock. Measurements were obtained in 1992, 1994, 1996, and 1998. Figure 1: King Creek (Installation 910) presents 1998 height for planting densities of 100, 200, 300, 440, 680, 1210 stems per acre (plots 1-6 respectively). The 300 spacing (plot 3) has lagged behind the others. Investigations have revealed that soil differences are unlikely to be the source of this lag in height development. It has been noted that this plot falls in a shallow draw that may be a cold pocket.

A Type I installation located in Lewis County, Washington on Washington DNR land that was planted in 1980 with 3-year old stock. Measurements were obtained in 1986, 1990, and 1994. Figure 2: Longbell Road (Installation 703) present 1994 height for ISPA (~ 600 spa, Plot # 3), ISPA/2 (~300 spa, Plot # 8) a nd ISPA/4 (~150 spa, Plot # 2).

Although there are additional sample plots on both installations representing various treatments, the branch measurement trial was limited to the specific plots noted above.

RESULTS

1. Ability to select the largest whorl branch

At King Creek, 20 - 21 trees were checked on each plot to determine if the largest branch was correctly chosen. Of the 122 trees checked, the largest branch was correctly located in 111 (91%). The criterion for being "correct" was that the branch had to have the largest diameter in both measurement directions. In almost all of the 11 "misses", the branch was equal to the largest in one of the two directions; 9 of these were off by 0.1 inch; 2 were off by 0.2 inch. A similar test on the 3 plots at Longbell Road found that the largest branch was correctly located in 59 (95%) of the 62 trees checked. Of the 3 misses one was off by 0.1 inch; 2 were off by 0.2 inch. 

2. King Creek Branching

• Branch Count

Table 1 summarizes the total branch count data and Figure 3: King Creek Count Branch Frequencyillustrates the distribution for the sample trees within each planting density.Value on the x-axis is the upper limit of the class (i.e. "8" means 7 or 8 branches).

Table 1: King Creek (Inst. 910) Plots 1-6. Branch Count

Nominal Density, SPA	No. of Trees Measured	Mean Count	Standard Deviation	Minimum	Maximum
100	42	8.8	2.0	5	13
200	40	9.4	3.2	4	18
300	44	7.4	2.4	3	13
440	44	9.0	3.4	4	24
680	39	8.0	2.2	4	14
1210	41	6.9	2.0	3	11

Figure 4: King Creek Branch Counts presents the trend in mean branch count with the upper and lower lines representing one standard deviation. Although ANOVA revealed statistically significant differences (p<.001) in branch count among the planting densities, this may not be very meaningful and may be a result of the low count associated with the 300 spa plot which also exhibited poor height development (see Figure 1: King Creek Installation 910). This result is for a single installation and may not be consistent with others.

On this installation very little live crown recession has occurred so all branches are alive. It will be interesting to learn how the counts change as stands develop and crowns recede at different rates.

a) Largest Branch Diameter

Table 2 summarizes the total branch count data and Figure 5: Branch Diameter Frequencies indicates the distribution for the sample trees within each planting density. Note that the value on the x-axis is the upper limit of the branch diameter class (i.e. "1.5" means branches with a diameter greater than 1 inch up to and including 1.5 inches). The choice of half-inch branch diameter classes was based on the common use of half-inch or full-inch categories in log sorts and log grades. One can discern a trend toward larger branches as stand density decreases. For example, about 50% of the trees in the 100 and 200 spa stands have branches in the 1.5 inch class while almost none have branches in this class in the 680 and 1210 spa stands. 

Table 2: King Creek (Inst. 910) Plots 1-6. Diameter of largest branches in 1st whorl above BH

Figure 6 presents the trend in mean branch diameter with the upper and lower lines representing one standard deviation. ANOVA found a statistically significant difference (p<.001) in branch diameter. The LSD procedure indicated that there was no difference between the 100 and 200 spa densities but all others were statistically different from each other. Again, the 300 spa density continues to stand out as an anomaly.

Since very little live crown recession has occurred on this installation, it will be interesting to learn how the branch diameter changes as stands develop and crowns recede at different rates.

a) Trends along the bole

Six trees in the 100 spa and 6 trees in the 1210 spa stands had branch data collected for whorls above and below the bh whorl. These trees yielded data for 47 and 56 whorls respectively. Figure 7 presents trends in branch count and Figure 8 presents trends in branch diameter. In both figures, linear predictions are shown and help highlight some aspects of the patterns.

On the x-axis the BH whorl is labeled zero and others are numbered consecutively away from it. With this small sample, and given that there has been no crown recession, there is no trend in the branch count. This may change as the crown recedes and smaller branches self prune. Furthermore, differences may develop among the treatments related to the size the branches achieved before recession.

Figure 8 shows a trend toward smaller branch diameter at higher whorls. One would expect that higher, hence younger, branches would be smaller. One would also expect branch diameter to peak somewhere near, and most likely above, the base of the live crown and then decline through dying and dead branches that are being overgrown by stemwood.

One can observe this by noting the departures from the linear predictions. Also note that the 100 SPA has larger branches and that the trend among the whorls is steeper. This is consistent with the idea that lower branches on well-spaced trees remain vigorous and grow fast.

Note that branches taper, hence diameter at the surface of a log becomes smaller as it becomes overgrown. If there is sufficient time and growth from moment of branch death until the log is harvested and graded, this change may be important. Most models seem to predict and record the branch diameter at the stem surface at the time it dies when the crown recedes by it. This is likely to be close to the maximum diameter of the branch and use of these diameters to predict eventual log grade and product recovery may be pessimistic.

1) Longbell Road Branching

• Branch Counts

Table 3 summarizes the total branch count data. Figure 9: Longbell Road Branch Count Frequencies indicates distribution for the sample trees within each planting density.  Figure 10: Longbell Road: Branch Counts presents the trend in mean branch count with the upper and lower lines representing one standard deviation. ANOVA found no statistically significant difference (p<.001) in branch count among the planting densities. At Longbell, live crowns have already receded above measured whorl. Hence all branches are dead. Current counts and diameters are likely to be maximum values that may decline in the future through self pruning and stem wood growing over the tapering branches.

Table 3: Longbell Road (Inst. 703) Branch Count

 

Nominal Density, SPA	No. of Trees Measured	Mean Count	Standard Deviation	Minimum	Maximum
ISPA/4 (150)	44	8.6	2.6	3	13
ISPA/2 (300)	43	8.5	2.3	5	16
ISPA (600)	42	9.5	3.2	2	17

• Largest Branch Diameter

Table 4 summarizes the total branch count data and Figure 11: Longbell Road: Largest Branch Diameter Frequencies indicates the distribution for sample trees within each planting density. As in King Creek, one can discern a trend toward larger branches as stand density decreases. In fact, some ISPA/4 trees already have branches exceeding 2 inches. 

Table 4: Longbell Road (Inst. 703). Diameter of largest branches in 1st whorl above BH

Figure 12 presents the trend in mean branch diameter with the upper and lower lines representing one standard deviation. ANOVA found a statistically significant difference (p<.001) in branch diameter. The LSD procedure indicated that branch diameters of the spacings were statistically different from each other. Note that these are all dead branches. It will be interesting to learn how the branch diameter changes as through self pruning and as they become embedded within stemwood

c) Trends along the bole

Three trees in the ISPA and 3 trees in the ISPA/4 stands had branch data collected for whorls above and below the bh whorl. These trees yielded data for 26 and 29 whorls respectively. Figure 13 presents the trend in branch count and Figure 14 presents the trend in branch diameter.

Both have linear predictions included. Unlike King Creek, where all branches were alive, crown recession at Longbell was such that live branches were found only in the upper most 2-3 whorls. This may explain why there was no trend in branch count at King Creek and a trend toward more branches with whorl height at Longbell. Crown recession may have led to self pruning of some of the smallest branches in the lower whorls at Longbell.

Crown recession would also explain the trend toward increasing branch diameter with whorl height at Longbell. The base of the live crown is presently at the upper limit of whorls that were measured (whorls 5-6) at Longbell.

One would expect the largest branches in the stem to be at or somewhat above this crown base. As one drops down the stem from this point, lower branches have progressively died and the stem is already growing out over their tapered profile, hence the trend toward smaller diameters measured on the stem surface. It is interesting to note that the trend toward larger branches with whorl number is steeper in the wider spacing. If whorls could have been reached above he live crown, one would expect to find the peak branch diameter and declining diameters further up as observed at King Creek.

If one measured higher whorls one would expect to observe the trend found at King Creek.

CONCLUSIONS

This test found that the suggested branch measurement procedure can be implemented quickly and reliably. After discussion at the SMC Annual Policy Committee meeting, it has been adopted as a routine data gathering procedure each time an installation is remeasured and is incorporated into the SMC database.

Analysis of these two installations reveals some interesting trends suggesting that treatment differences are beginning to evolve, can be quantified, and hypotheses tested. However, more meaningful conclusions regarding region-wide patterns must wait until initial data is collected from more installations. When re-measurements are obtained, insight on how patterns change as the stands continue to develop should be gained.

Back to SMC Newsletter