Printer Friendly
The Free Library
23,403,340 articles and books


INCIDENCE OF DEFECTS BY TREE CHARACTERISTICS IN RADIATA PINE RANDOM-WIDTH BOARDS.

RADO GAZO [+]

ROBERT BEAUREGARD [+]

ABSTRACT

This paper links radiata pine (Pinus radiata D. Don.) characteristics that can be observed by tree growers to the quality of random-width boards for appearance markets. The objective was to analyze the potential of observed standing-free characteristics as predictors of board quality parameters. A database of 259 random-width radiata pine boards (10.45 [m.sup.3] or 4,429 board feet) was used for the analysis. These boards came from 10 different clones of New Zealand radiata pine. Thirty-one percent of the boards came from pruned logs while the remainder came from unpruned logs. Eight defect categories were identified. Although this visual-estimation approach yields only limited results, several board quality parameters, such as trends in frequency and average area of defects can be predicted. For example, the smaller the tree diameter at breast height, the higher the frequency of intergrown knots in boards from both pruned and unpruned logs of that tree; a large tree diameter at breast height signals large a reas of clearwood in sideboards from pruned logs, etc.

The overall objective of this project was to optimize remanufacturing technologies to improve recovery from radiata pine (Pinus radiata D. Don) logs and enhance product value. Several types of questions are commonly asked by people involved in breeding, growing, and processing radiata pine in New Zealand when discussing improvement of recovery.

Geneticists ask which tree clones and what clone characteristics are most suitable or desired for specific products or markets. Tree growers, on the other hand, ask which tree characteristics and what parts of the tree (log height class) are most desirable for specific products (lumber, veneer, paper, etc.). Sawmillers want to know which sawing strategy is optimal in terms of recovery and product value, and remanufacturers want to know what recovery yields they can expect, which processing method is most suitable for a specific order, or what distribution of part sizes can be obtained from radiata pine boards of given characteristics.

Quality assessment of standing trees could help growers to merchandise trees selectively for a given market and product. Given the broad scope of this project, the publication of results has been broken down into several papers, each dealing with a specific area. This paper has the objective of establishing the relationships between standing-tree parameters and the quality of random-width boards. Standing-tree parameters were expressed by diameter at breast height and visual assessment of internode length and branch size. Board quality was expressed by frequency of defects, average defect area for each defect type, and percentage of clear surface area. A previous paper [1] analyzed the random-width board quality as it relates to various clone characteristics while upcoming papers will present relationships with log characteristics.

METHODOLOGY

SAMPLE MATERIAL

To address some of the questions being asked, a database was constructed of random-width radiata pine lumber. Gazo et al. [2] describes reasons for and detailed methodology of developing this database.

The trees selected for these studies were required to show a range of qualities typical of the crop being harvested now and in the near future. A unique early test in Compartment 1350 of Kaingaroa Forest, planted in 1968, had the capability of providing trees of desired quality. The selected trees were believed to be representative of the main body of the population as well as of more extreme values of tree characteristics present in current plantations. In November 1995, at age 27 years, 20 trees were harvested.

Four logs were cut from each tree; these included the pruned butt log, the second log, one intermediate log, and the top log. The butt logs were on average about 4 m long and the rest of the logs were about 4.9 m long. A top log from one tree was not suitable for sawmill processing and one extra intermediate log was taken from two trees. This resulted in 20 butt logs, 20 second logs, 22 intermediate logs, and 19 top logs: a total of 81 logs.

Processing of these logs in Vanner Sawmill at Reporoa resulted in 392 random-width boards. The first 40 logs (first replicate set of 10 clones) were live sawn into boards 40 mm thick (Fig. 1). This method targets the U.S. 5/4-inch random-width shop boards. The second 41 logs (second replicate set of 10 clones) were cant sawn by flat sawing 40-mm boards leaving the central cant. Logs with a small-end diameter between 200 to 300 mm were sawn with a 100-mm cant and larger logs with a 200-mm cant. The cants were reduced to 100- by 40-mm Australian structural stock before drying. One butt log was of inferior quality and didn't yield any random-width boards. This log was excluded from further analysis. Boards were dried in a commercial dry kiln on a high-temperature schedule and surfaced on both faces.

All boards were then digitized. Digitizing consisted of recording the board dimensions (width and length), defect positions, and defect types in an orthogonal coordinate system for each face of each board. Intergrown knots, partially intergrown knots, loose knots/holes, spike knots, bark pockets/blemishes, resin pockets, needle flecks, and wane were classified as separate defect categories.

Cant-sawn logs did not yield any random-width boards from the "cant" zone of the log; the cants were immediately processed into structural lumber. Therefore, the random-width boards that came from the same "cant" zone of live-sawn logs were excluded (Fig. 1) from the statistical analysis in order to provide a sample of two replicates (trees) per clone with comparable wood characteristics. The exclusion of the center zone from the analyses in this paper was also desirable since in radiata pine sawing, it is common practice that only the sideboards are aimed at the production of random-width boards. Boards from the center cants are most commonly used as dimension lumber for framing purposes. Hence, from the original 392 boards, 133 boards were excluded. This in turn decreased the variation between the replicates and increased the likelihood of establishing differences between trees of different characteristics. Furthermore, separate analyses were performed on groups of pruned and unpruned logs. Thus 259 boards were included in the analysis. The total volume of boards in the database is 17.10 [m.sup.3] (7,245 board ft.). The volume of boards included in this analysis is 10.45 [m.sup.3] (4,429 board ft.). The distribution of the trees, logs and boards by various tree characteristics is shown in Table 1.

INCIDENCE OF DEFECTS

In order to analyze the occurrence of defects in random-width radiata pine lumber, two variables were considered: 1) frequency of defects; and 2) average area of the defects. The defect frequency was defined as the number of occurrences of a defect per [m.sup.2] of board surface area computed on a per-tree basis. The number of each type of defects in each board was recorded and divided by the board surface area of each tree. The board surface area was calculated as length of the board times the largest width of the board. Defect frequency was calculated for all defect types except wane since wane can be found on every random-width board. The defect area was defined as the area of a defect type, in [cm.sup.2] per [m.sup.2] of board surface area, computed on a per-tree basis. The average defect area was calculated for all defect types, including wane. The total defect area per defect type was calculated for each tree. This defect area was then divided by the total board surface area of the tree in question. Whe n calculating frequency and area of the defects, only defects on the worse face were considered.

CLEAR SURFACE AREA

Percentage of clear surface area was calculated as the ratio of board clear area (board surface area minus total defect area) to board surface area. In addition, for boards from the pruned butt logs, two clearwood-related variables were introduced. The first was the yield of clear boards defined as the ratio of the surface area of the boards with no defects other than wane and the surface area of all the boards. The second was the yield of boards with no knots defined as the ratio of the surface area of the boards with no knots to the surface area of all the boards. These two variables allowed the analysis of the production of clearwood after pruning.

ESTIMATING STANDING-TREE CHARACTERISTICS

Diameter at breast height (DBH) in millimeters was the only measured parameter. Trees with DBH less than 400 mm were classified as small; those with DBH between 400 and 600 mm were medium; and those with DBH greater than 600 mm were large.

Branch size is a visual estimate of the average diameter of all branches at the stem of a given tree. This estimate is made while standing several meters away from a tree. The small branch size value was assigned if the perceived average branch size was less than 30 mm, medium if the perceived average branch size was between 30 and 60 mm, and large if the perceived average branch size was greater than 60 mm.

Internode length is a visual estimate of the average distance between the whorls on the stem of a given tree. This estimate is also made while standing several meters away from a tree. The short internode length was assigned if the perceived average internode length was shorter than 600 mm, medium if the perceived average internode length was between 600 mm and 1000 mm, and large if the perceived average internode length was longer than 1000 mm. A crew of experienced observers measured and estimated these parameters.

RESULTS

DIAMETER AT BREAST HEIGHT

Frequency of defects. -- In boards from pruned logs, the frequency of intergrown knots in trees with small DBH was higher (p = 0.06) than in trees with medium and large DBH (Table 2). The tree DBH could be used to predict the trend in frequency of intergrown knots in boards from pruned logs. The smaller the DBH, the higher the frequency of intergrown knots/[m.sup.2] of board surface area.

In boards from unpruned logs, the frequency of intergrown knots in trees with small DBH (15.051 defects/[m.sup.2]) was significantly higher (p = 0.0001) than in trees with medium (6.542 defects/[m.sup.2]) and large (4.344 defects/[m.sup.2]) DBH (Table 2). The tree DBH could be used to predict the trend in frequency of intergrown knots in boards from unpruned logs. The smaller the DBH, the higher the frequency of intergrown knots/[m.sup.2] of board surface area.

In boards from unpruned logs, the frequency of needle flecks in trees with large DBH was marginally higher (p = 0.06) than in trees with small and medium DBH (Table 2). The tree DBH could be used to predict the trend in frequency of needle flecks in boards from unpruned logs. The lowest frequency of needle flecks/[m.sup.2] of board surface area was from trees with medium DBH, followed by small DBH. DBH did not have a considerble effect on the frequency of the other defects described.

Clear surface area. -- In boards from pruned logs, the percentage of clear area in trees with small DBH (65.62%) was significantly lower (p = 0.0008) compared to trees with medium (78.99%) and large (79.37%) DBH (Table 3). The tree DBH could be used to predict the trend in the percentage of clear area in boards from pruned logs. The smaller the DBH, the lower the percentage of clear surface area.

This trend could be explained by the geometry of logs, where the small DBH interacts with taper to give a much higher proportion of wane. The average pruned logs from trees with a small DBH tapered more (8.9 mm/m) than the pruned logs from trees with medium (6.6 mm/m) and large (8.0 mm/m) DBH. The difference in proportion of wane in pruned logs between trees with medium and large DBH is much smaller because in this case the bigger taper counteracts the size effect. The percentage of clear area in boards from unpruned logs was not affected by tree DBH.

Average defect area. -- In boards from pruned logs, the average area of wane in trees with a small DBH (3346 [cm.sup.2]/[m.sup.2]) was significantly higher (p = 0.005) than in trees with medium (2057 [cm.sup.2]/[cm.sup.2]) and large (1938 [cm.sup.2]/[cm.sup.2]) DBH (Table 3). The tree DBH could be used to predict the trend in average area of wane in boards from pruned logs. The smaller the DBH, the higher the average area of wane/[cm.sup.2] of board surface area. This defect type was the cause of the trend observed previously in clearwood area.

In boards from unpinned logs, the average size of needle fleck area in trees with large DBH (308.74 [cm.sup.2]/[cm.sup.2]) was significantly larger (p = 0.01) than in trees with medium (20.41 [cm.sup.2]/[cm.sup.2]) and small (42.20 [cm.sup.2]/[cm.sup.2]) DBH (Table 3). The magnitude of difference in size of a needle fleck area can be explained by the relationships between the tree geometry and saw patterns used. According to Kininmonth [3], needle fleck tends to be absent outside the inner two to four growth rings. In the case of larger trees, being of the same age as smaller trees, these four inner rings will occupy a wider zone inside the logs and are likely to show outside the center zone that was excluded in this analysis.

In boards from unpinned logs, the average area of intergrown knots in trees with small DBH (166.63 [cm.sup.2]/[cm.sup.2]) was only marginally larger (p = 0.06) than in trees with medium (104.77 [cm.sup.2]/[cm.sup.2]) and large (110.68 [cm.sup.2]/[cm.sup.2]) DBH (Table 3). The tree DBH could be used to predict the trend in average area of intergrown knots in boards from unpinned logs. The smaller the DBH, the higher the average area of intergrown knots/[cm.sup.2] of board surface area. Knots were more frequent in trees with small DBH and they were smaller in size. This made the area only marginally different. DBH did not have a significant effect on the average area of other defects.

BRANCH SIZE

Frequency of defects. -- In boards from unpinned logs, the frequency of intergrown knots in trees with small branches (11.114 defects/[cm.sup.2]) was significantly higher (p = 0.005) than in trees with medium (6.113 defects/[cm.sup.2]) and large (4.051 defects/[m.sup.2]) branches (Table 4). Branch size could be used to predict the trend in frequency of knots in boards from unpruned logs. The smaller the branches, the higher the frequency of intergrown knots/[m.sup.2] of board surface area.

In boards from unpruned logs, the frequency of loose knots in trees with small branches (2.991 defects/[m.sup.2]) was significantly higher (p = 0.04) than in trees with medium (2.382 defects/[m.sup.2]) and large (0.518 defects/[m.sup.2]) branches (Table 4). Branch size could be used to predict the trend in frequency of loose knots in boards from unpruned logs. The smaller the branches, the higher the frequency of loose knots/[m.sup.2] of board surface area. Branch size did not have a significant effect on the frequency of other defects.

Clear surface area. -- The percentage of clear area in boards from both pruned and unpruned logs was not affected by tree branch size. The branch size was not a suitable predictor of the percentage of clear board surface area (Table 5).

Average defect area. -- The average defect area in boards from both pruned and unpruned logs was not affected by tree branch size. The branch size was not a suitable predictor of the average defect area (Table 5).

INTERNODE LENGTH

Frequency of defects. -- In boards from unpruned logs, the frequency of partially intergrown knots in trees with short internodes (2.743 defects/[m.sup.2]) is higher (p = 0.04) than in trees with medium (1.902 defects/[m.sup.2]) and long (1.345 defects/[m.sup.2]) internodes (Table 6). The tree internode length can be used to predict the trend in frequency of partially intergrown knots in boards from unpruned logs. The shorter the internodes, the higher the frequency of partially intergrown knots per [m.sup.2] of board surface area.

In boards from unpruned logs, the frequency of loose knots in trees with short internodes (3.076 defects/[m.sup.2]) was significantly higher (p = 0.005) than in trees with medium (1.963 defects/[m.sup.2]) and long (0.53 8 defects/[m.sup.2]) internodes (Table 6). The tree internode length could be used to predict the trend in frequency of loose knots in boards from unpruned logs. The shorter the internodes, the higher the frequency of loose knots/[m.sup.2] of board surface area.

In boards from unpruned logs, the frequency of bark pockets and blemishes was marginally different (p = 0.08) across trees with various internode lengths (Table 6). The highest frequency of bark pockets and blemishes was in boards from unpruned logs from trees with medium internodes. The tree internode length did not have a significant effect on the frequency of other defects.

Clear surface area. -- The percentage of clear area in boards from both pruned and unpruned logs was not affected by tree internode length. The tree internode length was not a suitable predictor of the percentage of clear board surface area (Table 7).

Average defect area. -- In boards from unpruned logs, the average area of loose knots in trees with short (20.38[cm.sup.2]/[m.sup.2]) and medium (20.62 [cm.sup.2]/[m.sup.2]) internodes was significantly larger (p = 0.04) than in trees with long (4.43[cm.sup.2]/[m.sup.2]) internodes (Table 7). The tree internode length could be used to predict the trend in average area of loose knots in boards from unpruned logs. The smallest average area of loose knots/[m.sup.2] of board surface area is in boards from unpruned logs from trees with long internodes.

In boards from unpruned logs, the average area of bark pockets and blemishes in trees with short (8.52 [cm.sup.2]/[m.sup.2]) and long (8.52 [cm.sup.2]/[m.sup.2]) internodes was significantly smaller (p = 0.03) than in trees with medium (19.05 [cm.sup.2]/[m.sup.2]) internodes (Table 7). The tree internode length could be used to predict the trend in average area of bark pockets and blemishes in boards from unpruned logs. The highest frequency of bark pockets and blemishes was in boards from unpruned logs from trees with medium internodes. The tree internode length did not have a significant effect on the area of other defects.

The database was also analyzed with explanatory variables such as the size of the defect core in pruned logs. Size of the defect core was found not to be significantly related to defect frequencies or areas. Multiple factor Analyses of Variance (ANOVAs) were performed, but no satisfactory model could be derived to predict the behavior of the radiata pine resource in terms of defect frequencies and areas on random-width boards using every available tree characteristics. Finally, analyses were carried out in which all knot types were grouped in one single category and resin pockets, bark pockets and blemishes were grouped into another. However, no other significant results than what has already been presented could be achieved.

Overall, the tree DBH, visually assessed branch size, and internode length were not very strong predictors of frequency, average area of defects, and the percentage of clear surface area. DBH and internode length seemed to be more suitable than branch size. It was easier to predict the quality of the boards from unpruned logs than from pruned logs. It was also easier to predict the frequency and average area of defects and more difficult to predict the percentage of clear area. The incidence of knots, especially intergrown, partially intergrown, and loose knots was easier to predict than the incidence of bark pockets, blemishes, and needle flecks. The incidence of spike knots and resin pockets was impossible to predict using tree characteristics included in the analysis. The frequency of knots in boards from unpruned logs of small branch trees was higher than in boards from other logs. In pruned logs, half of the boards without any knots were not clear boards. Bark pockets and blemishes were by far the most frequent defects in those boards.

CONCLUSIONS

The objective was to analyze the potential of standing-tree characteristics that could be observed by tree growers as predictors of board quality parameters. These parameters included DBH, branch size, and internode length. DBH was measured and branch size and internode indices were visually estimated on standing trees according to current grower practice. This visual assessment approach produces only marginal results; however, several board quality parameters, such as trends in frequency and average area of knots could be predicted, mostly for boards from unpruned logs. Smaller trees proved to have smaller, more numerous branches. Hence, knot frequencies were significantly higher for smaller trees, but the effect on knot area was only marginal.

In our analyses, several tests came out non-significant with p-values between 0.06 and 0.09. It is still unclear if these tests failed to show differences because of a small sample size or because of the sampling strategy with insufficient range in the spread of tree characteristics, particularly branch size. The study has confirmed the expected relatively high correlation between tree size and branch size.

The study leaves many questions unanswered. Future reports will explore the data to determine the grade distribution, defect frequencies, and areas within trees. The results will be analyzed by log height class, board position within the log, small- and large-end log diameter, taper, sweep, log branch and internode indices (measured). The difference between the two sawing strategies will also be assessed. The database of digitized random-width boards will also be used to analyze the component recovery potential of this resource.

The authors are, respectively, Assistant Professor, Purdue Univ., 1200 Forest Prod. Building, West Lafayette, IN 47907-1200; Manager, Forintek Canada Corp., 319 Franquet, Sainte-Foy, QC, Canada, G1P 4R4; and Research Scientists, Forest Res. Inst., Private Bag 3020, Rotorua, NZ. This paper was received for publication in March 1999.

(+.) Forest Products Society Member.

(1.) Beauregard, R., R. Gazo, M.O. Kimberley, J. Turner, S. Mitchell, and A. Shelbourne. 1999. Clonal variation in the quality of radiata pine random-width boards. Wood and Fiber Sci. 31(4): 222-234.

(2.) Gazo, R., S. Mitchell, and R. Beauregard. 1998. Development of a database and its use to quantify incidence of defects in random-width Pinus radiata boards. New Zealand J. of Forest Sci. 28(2): 254-269.

(3.) Kininmonib, J.A. and L.J. Whitehouse. 1991. Properties and Uses of New Zealand Ridiata Pine. Vol. 1-Wood Properties. Kininmonth and Whitehouse, eds. New Zealand Forest Res. Inst., Rotorua, New Zealand. p.2.2.
 Random-width boards
 (side only) by tree
 characteristics.
 No. of No.of No.of
 trees logs boards Volume Volume
Pruned logs ([m.sup.3]) (BF)
 Diameter at breast height
 Small 4 4 8 0.226 95.6
 Medium 12 12 45 1.621 687.0
 Large 4 4 31 1.345 570.1
 Branch size
 Small 8 8 21 0.628 266.2
 Medium 9 9 45 1.748 741.0
 Large 3 3 18 0.815 345.5
 Internode length
 Short 12 12 48 1.804 764.3
 Medium 4 4 14 0.424 179.5
 Long 4 4 22 0.965 408.9
 Total 20 20 84 3.192 1,352.7
Unpruned logs
 Diameter at breast height
 Small 4 12 25 0.834 353.3
 Medium 12 36 92 3.566 1,511.1
 Large 4 13 58 2.860 1,211.9
 Branch size
 Small 8 24 51 1.784 756.2
 Medium 9 27 88 3.717 1,575.1
 Large 3 10 36 1.758 745.1
Internode length
 Small 12 36 99 3.992 1,691.8
 Medium 4 12 31 1.136 481.5
 Large 4 13 45 2.131 903.1
 Total 20 61 175 7.259 3,076.3
 Frequency of defects by
 diameter at breast height.
 Defect frequency
Diameter at Intergrown Partially Loose Spike
breast height knot intergrown knot knot/hole knot
Pruned logs (no. of defect/
 [m.sup.2] of boards
 Small 2.618 0.757 0.345 --
 Medium 0.508 0.243 0.189 --
 Large 0.542 0.310 0.118 --
 ANOVA [a] p = 0.06 p = 0.23 p = 0.60 --
 Average 0.849 0.338 0.198 --
Unpruned logs
 Small 15.051 2.821 2.873 0.115
 Medium 6.542 2.344 2.532 0.029
 Large 4.344 1.624 1.260 0.011
 ANOVA [a] p = 0.0001 [**] p = 0.29 p = 0.27 p = 0.16
 Average 7.804 2.295 2.346 0.043
Diameter at Bark Resin Needle
breast height pocket/blemish pocket fleck
Pruned logs
 Small 1.231 0.149 0.146
 Medium 1.314 0.149 0.070
 Large 1.231 0.328 0.064
 ANOVA [a] p = 0.81 p = 0.70 p = 0.77
 Average 1.369 0.186 0.080
Unpruned logs
 Small 1.719 0.058 0.063
 Medium 2.012 0.009 0.041
 Large 2.112 0.034 0.230
 ANOVA [a] p = 0.90 p = 0.32 p = 0.06
 Average 1.973 0.024 0.083
(a.)(**.)=highly significant difference (.01 level).
 Clearwood and defect area by
 diameter at breast height.
 Defect area
Diameter at Clearwood Intergrown Partially Loose
breast height area knot intergrown knot knot/hole
Pruned logs (%) ([cm.sup.2]/[m.sup.2]
 Small 65.62 12.68 6.19 0.48
 Medium 78.99 4.99 2.37 0.95
 Large 79.37 2.40 4.26 4.04
 ANOVA [a] p = 0.0008 [**] p = 0.37 p = 0.57 p = 0.32
 Average 76.96 5.66 3.37 4.52
Unpruned logs
 Small 74.34 166.63 27.32 15.39
 Medium 75.79 104.77 44.63 19.17
 Large 77.00 110.68 38.24 13.31
 ANOVA [a] p = 0.70 p = 0.06 p = 0.39 p = 0.66
 Average 75.74 118.32 39.89 17.24
Diameter at Spike Bark Resin Needle
breast height knot pocket/blemish pocket fleck Wane
Pruned logs
 Small -- 5.44 0.11 67.16 3346
 Medium -- 11.72 1.38 22.47 2057
 Large -- 10.20 4.06 99.54 1938
 ANOVA [a] -- p = 0.82 p = 0.46 p = 0.42 p = 0005 [**]
 Average -- 10.41 1.74 45.75 2235
Unpruned logs
 Small 1.92 6.16 0.04 42.20 2306
 Medium 1.09 11.53 0.06 20.41 2219
 Large 0.17 8.71 0.11 308.74 1820
 ANOVA [a] p = 0.67 p = 0.54 p = 0.84 p = 0.001 [**] p = 0.19
 Average 1.07 9.89 0.07 82.43 2157
(a.)(**.)= highly significant difference (.01 level).
 Frequency of defects by branch size.
 Defect frequency
Branch Intergrown Partially Loose Spike
size knot intergrown knot knot/hole knot
Pruned logs (no. of defects/
 [m.sup.2] of boards)
 Small 1.333 0.601 0.259 --
 Medium 0.436 0.206 0.179 --
 Large 0.955 0.123 0.116 --
 ANOVA [a] p = 0.48 p = 0.15 p = 0.76 --
 Average 0.849 0.338 0.198 --
Unpruned logs
 Small 11.114 2.639 2.991 0.070
 Medium 6.113 2.393 2.382 0.032
 Large 4.051 1.083 0.518 0.000
 ANOVA [a] p = 0.005 [*] p = 0.08 p = 0.04 [*] p = 0.44
 Average 7.804 2.295 2.346 0.043
Branch Bark Resin Needle
size pocket/blemish pocket fleck
Pruned logs
 Small 1.468 0.107 0.182
 Medium 1.438 0.309 0.017
 Large 0.934 0.000 0.032
 ANOVA [a] p = 0.79 p = 0.36 p = 0.11
 Average 1.369 0.186 0.080
Unpruned logs
 Small 2.215 0.029 0.031
 Medium 2.139 0.018 0.118
 Large 0.831 0.028 0.119
 ANOVA [a] p = 0.24 p = 0.93 p = 0.44
 Average 1.973 0.024 0.083
(a.)(*.)=significant difference (.05 level);
(**.)=highly significant difference (.01 level).
 Clearwoad and defect area by branch size.
 Defect area
Branch Clearwood Intergrown Partially Loose
size area knot intergrown knot knot/hole
Pruned logs (%) ([cm.sup.2]/[m.sup.2])
 Small 72.82 6.29 6.04 0.80
 Medium 79.33 5.17 2.00 2.57
 Large 79.49 5.63 1.24 0.08
 ANOVA p = 0.11 p = 0.98 p = 0.30 p = 0.51
 Average 76.96 5.66 3.37 1.52
Unpruned logs
 Small 76.15 127.20 32.30 20.02
 Medium 74.42 109.73 51.25 18.99
 Large 78.59 120.42 26.05 4.57
 ANOVA p = 0.34 p = 0.76 p = 0.09 p = 0.11
 Average 75.74 118.32 39.89 17.24
Branch Spike Bark Resin Needle
size knot pocket/blemish pocket fleck Wane
Pruned logs
 Small -- 6.04 1.56 67.31 2629
 Medium -- 15.28 2.47 43.07 1996
 Large -- 5.99 0.00 03.52 2034
 ANOVA -- p = 0.41 p = 0.71 p = 0.68 p = 0.11
 Average -- 10.41 1.74 45.75 2235
Unpruned logs
 Small 1.19 8.35 0.02 21.10 2175
 Medium 1.32 13.18 0.10 135.77 2226.94
 Large 0.00 4.15 0.10 85.98 1899
 ANOVA p = 0.23 p = 0.23 p = 0.34 p = 0.34 p = 0.52
 Average 1.07 9.89 0.07 82.43 2157
 Frequency of defects by internode length.
 Defect frequency
Internode Intergrown Partially Loose Spike
length length intergrown knot knot/hole knot
Pruned logs (no. of defects/
 [m.sup.2] of boards)
 Short 1.333 0.601 0.259 --
 Medium 0.436 0.206 0.179 --
 Long 0.955 0.123 0.116 --
 ANOVA [a] p = 0.44 p = 0.18 p = 0.12 --
 Average 0.849 0.338 0.198 --
Unpruned logs
 Short 9.086 2.743 3.076 0.057
 Medium 7.404 1.902 1.963 0.043
 Long 4.358 1.345 0.538 0.000
 ANOVA [a] p = 0.15 p = 0.04 [*] p = 0.005 [**] p = 0.54
 Average 7.804 2.295 2.346 0.043
Internode Bark Resin Needle
length pocket/blemish pocket fleck
Pruned logs
 Short 1.468 0.107 0.182
 Medium 1.438 0.309 0.017
 Long 0.934 0.000 0.032
 ANOVA [a] p = 0.33 p = 0.16 p = 0.20
 Average 1.369 0.186 0.080
Unpruned logs
 Short 2.113 0.033 0.068
 Medium 2.712 0.000 0.124
 Long 0.815 0.021 0.089
 ANOVA [a] p = 0.08 p = 0.62 p = 0.81
 Average 1.973 0.024 0.083
(a.)(*.)=significant differnce (.05 level);
(**.)=highly significant difference (.01 level).
 Clearwood and defect area by internode length
 Defect area
Intermode Clearwood Intergrown Partially Loose
length area knot intergrown knot knot/hole
 (%) [cm.sup.2]/[m.sup.2]
Pruned logs
 Short 74.10 8.12 5.40 2.61
 Medium 81.76 0.32 0.05 0.00
 Long 80.00 4.23 1.11 0.06
 ANOVA [a] p = 0.07 p = 0.38 p = 0.18 p = 0.35
 Average 76.96 5.66 3.37 1.52
Unpruned logs
 Short 76.02 123.34 41.73 20.38
 Medium 72.99 108.36 49.47 20.62
 Long 77.63 113.23 24.79f 4.43
 ANOVA [a] p = 0.30 p = 0.85 p = 0.25 p = 0.04 [*]
 Average 75.74 118.32 39.89 17.24
Intermode Spike Bark Resin Needle
length knot pocket/blemish pocket fleck Wane
Pruned logs
 Short -- 15.11 3.01 53.55 2501
 Medium -- 3.42 0.00 67.42 1752
 Long -- 4.49 0.00 2.64 1987
 ANOVA [a] -- p = 0.27 p = 0.34 p = 0.65 p = 0.08
 Average -- 10.41 1.74 45.75 2235
Unpruned logs
 Short 0.89 8.52 0.09 95.48 2107
 Medium 2.69 19.05 0.00 61.23 2439
 Long 0.00 8.52 0.08 64.49 2025
 ANOVA [a] p = 0.35 p = 0.03 [*] p = 0.69 p = 0.9l p = 0.32
 Average 1.07 9.89 0.07 82.43 2157
(a.)(*.)=significant difference (.05 level).
COPYRIGHT 2000 Forest Products Society
No portion of this article can be reproduced without the express written permission from the copyright holder.
Copyright 2000 Gale, Cengage Learning. All rights reserved.

 Reader Opinion

Title:

Comment:



 

Article Details
Printer friendly Cite/link Email Feedback
Author:GAZO, RADO; BEAUREGARD, ROBERT; KIMBERLEY, MARK; MCCONCHIE, DON
Publication:Forest Products Journal
Date:Jun 1, 2000
Words:5539
Previous Article:A FIELD TRIAL OF PIGMENT-EMULSIFIED CREOSOTE: 11 YEAR DATA.
Next Article:Great Lakes Section meeting focused on powder coating.

Terms of use | Copyright © 2014 Farlex, Inc. | Feedback | For webmasters