Reprinte(l from JOURNAL OF FOHES'l'RY, VOL. 54, No. 9, September 1D50 PUIf.;:i;)sr:d by the FOr€;s\ Service for official use George R. Staebler Effect of Controlled Release on Growth of Individual Douglas-Fir Trees Puget Sound Research Center, Pacific Northwest Forest and Range Experiment Station, Forest Service, U. S. Department of AgricultUl'e '1'HIS paper presents preliminary results of a "spot-thinning" study in a 41-year-old, well-stocked, Site IV, Douglas-fir stand 011 the Wind River Experimental Forest in southwest Washingtoll. The ex­ periment, installed in the spring of 1952, is designed to evaluate the effect of different degrees of release on dominant, codominant, and in­ tenrediate trees. intermediate, dominant, codomi­ nant, etc. Diameters of the study trees were measured at time of treatment and again in September 1955 after 3 growing seasons. Total heights were measured in 1952, but not in 1955. More complete measure­ ments, including height, will be made after 5 years' growth record is available. Methods Results A one-chain grid was laid out in approximately 14 acres of the stand. At each grid-point, the nearest dominant, codominant, or intermediate tree was chosen that had at least three competitors. '1'0 be a competitor, a tree had to be a dominant, codominant, or inter­ mediate, and be nearer to the sub­ ject tree than D+2 feet where D equals the d.b.h. in inches of the competitor. Four treatments of three crown classes are involved: 0, 1, 2, or 3 competitors cut around 10 each elominant, codomillant, and inter­ lllc(liate trees. Thus the experiment includes 120 study trees. The num­ ber of competitors to be cut was chosen at random and trees cut were "chief" competitors, deter­ mined by observation. Orown classes of study trees were selected in rotation at consecutive grid points, i.e. dominant, codominant, In the 3 years following' treat­ ment, dominants grew more than codominants and codominants more than intermediates as mig'ht be ex­ ]1ected (Table 1). More important is the change in gTowth with treat­ ment, the data indicating a pro­ gressive increase in growth from o to 3 competitors cut. The differ­ ence is greatest between no cutting' and 1 competitor cut, and least be­ tween 2 and 3 competitors cut. ,Vithin crown classes other than dominant, growth does not follow such a neat pattern, although in every crown class the growth is least where no trees were cut. Analysis of variance shows that differences among crown classes and among' numbers of competitors cut are highly significant.l The in­ teraction between crown classes and number of" competitors cut (i.e., the tendency for release ef­ fects to change in going' from one crown class to another) was not significant.2 In a supplementary analysis, re­ gression of diameter growth over d.b.h, was calculated for each of the four treatments, pooling' the crown class data (Fig. 1). The re­ grcssion for 1 competitor cut is substantially above the control (0 competitors cut) and 2 compet­ 'One percent level test. 'Five percent level test. TABLE I.-AVERAGE 3-YEAR DIAMJilTER GROWTH BY CROWN CLASS AND NUMBE.R OF COMPETITORS CUT CompetitOl's cut Dominant Number 0 1 2 3 Avemge1 Crown class Codominant Intermediate Average' Inches .69 .43 .19 .44 .72 .29 .55 ,94 1.05 .56 .48 .33 .48 .31 .60 .62 .85 .48 .28 .55 lAverages weighted by the number of trees rcpl'esented in each crmyn anel treat­ men t clnsR. JOURNAL 568 1.6 1.4 (J) W :c u 2 1.2 I 1.0 :i a:) d Competitors Cut ,8 Z :c r- 3: 0 ,6 .4 (!) .2 .0 4 0 O.B. H. FIG. 8 12 16 PRIOR TO TREATMENT - INCHES 1.-'l'hree J'0ar diameter growth by d.b.h. fOl' each of t.he treatment.s. itors (!ut is well above 1 compet­ itor, but the 3 competitors cut re­ gression is practically identical with that for 2. When measurements were made in 1955, nine trees either had died or had suffered broken tops and were omitted from the calculation.3 These losses are almost certainly a result of thinning since they were confined to codominant and inter­ mediate trees that had been re­ leased. No further attempt at anal­ Tsis was made for this preliminary report. Growth is strongly related to d.b.h. within the crown class and 3Anal 'sis of variance was made using' (lisproportionate subclass fl'eqnencie. . The nn 11 hypothesis becomes: « Crown class and rclease do not affect the g'l'owth of t.rees tllat survive the measurement pl'I'ioil." treatment classifications. Some of the differences in growth within a crown class, apparent from Table 1, are due to larger or smaller trees in one treatment than in another. This fact is not accounted for by the analysis of variance technique. More refined analysis has not been made because of the short growth period involved and the prelimi­ nary nature of this report. Discussion This experiment is designed to use individual trees for observa­ tion, rather than some unit area as in more usual thinning trialR. It insures close control of the thin­ ning treatment applied and per­ mits evaluation of the treatment on trees occupying different posi­ tions in the stand. Consequentl T, OF FORES'rRY the results have more than usual meaning in understanding the re­ actions of trees to thinning. Con­ tinued observation should add con­ siderably to the type of informa­ tion required for the formulation of scientific marking rules. Although these preliminary re­ sults are based on only three years' growth following treatment, they already reveal interesting and sig­ nificant trends. Dominant trees, which often are thought to grow at the maximum rate possible for the site, have re­ sponded well to release in this well­ stocked, Site IV stand. In fact, Figure 1 shows that the greatest response, in terms of inches of in­ creased grovvth, occurred on the largest trees-that is, toward the dominant end of the scale. Appar­ ently the better developed crown and root system of large (domi­ nant) trees permits them to make better, or at least more immediate, use of the increased room from thinning. It must be kept in mind, how­ ever, that in dense stands such as the one in which this study is lo­ cated, dominant trees are not so well differentiated from the lower crown classes as in stands less well stocked nor, perhaps, as in stands on higher sites. That dominant trees had at least three competitors is evidence that even they were crowded. The point of diminishing returns is reached quickly as release is in­ creased. Cutting 1 competitor in­ creases growth substantially. Cut­ ting' 2 increases it still more, but not twice as much. Little added growth results from cutting 3, rather than 2, competitors. How­ ever, the effects of heavier release may be more lasting so that where the interval between thinnings is long, cutting 3 competitors may prove superior to lighter treat­ ments. About this file: This file was created by scanning the printed publication. Some mistakes introduced by scanning may remain.