柳杉在不同生育地及疏伐作業之材質探討

Abstract

本研究針對棲蘭山區與對高岳山區的柳杉(Cryptomeria japonica)進行非破壞（超音波和表面硬度儀）與破壞性檢測（動彈性模數和縱向壓縮強度），藉以瞭解不同疏伐處理（帶狀疏伐和下層疏伐）與同一林分中不同生長狀況（優、中、劣勢木）對機械性質、木材組織與固碳量的影響，進而建立起破壞與非破壞試驗的關係。 位於棲蘭山區不同帶狀疏伐強度的柳杉造林木試驗結果顯示：在機械性質部分，縱向超音波速度、徑向超音波速度、動彈性模數、縱向壓縮強度及木材表面刺入深度值，由分析的結果，沒有顯著性的差異存在；而在木材組織部分，年輪密度（RD）與早材密度（ED）、晚材密度（LD）、最低密度（Dmin）、最高密度（Dmax）及晚材率（LWP）之間有顯著正相關性存在。除此之外，同一地區不同條件兩樣區，在年輪密度、縱向超音波速度、木材表面刺入深度值、動彈性模數和縱向壓縮強度皆有顯著差異，同時邊材和心材在不同樣區，則有不同的結果。年輪密度與縱向壓縮強度間和動彈性模數與縱向壓縮強度間，都存在正相關，胸高直徑（DBH）與超音波波速間、木材表面刺入深度值與縱向壓縮強度間和木材表面刺入深度值與動彈性模數間，都存在負相關。 位於對高岳山區不同強度的下層疏伐對柳杉造林木的機械性質影響如下：以未處理區（不進行任何疏伐處理）的平均胸徑最小，同時其標準差又最大。伐採後原木縱向超音波波速在樹高方向與直徑方向各有四種型態的變化。此外，在對高岳山區也探討栽植密度的影響，在不同栽植密度2200株（Type A）與3000株（Type B）之間，有顯著的差異存在，且在抗壓強度與微破壞強度上，Type B皆大於Type A。成熟材與未熟材的強度不分軒輊。最後木材之抗壓強度與比重有顯著的線性正相關。在木材組織部分，栽植密度為Type B的平均管胞長度皆大於栽植密度為Type A的平均管胞長，成熟材管胞長也有此趨勢，而未成熟材管胞長則不一定；六區未成熟材管胞長與第一次疏伐後不同栽植密度下各區的管胞長度，皆沒有顯著差異，但栽植密度為Type B的管胞長度大於栽植密度為Type A的管胞長度。管胞長會由髓心向樹皮方向增加，其長度與生長年齡呈現高度線性相關，約在15-22年之間趨於緩和，之後其長只有小幅度的變動並沒有明顯的增加。未成熟材及成熟材管胞的管胞長度，兩者有顯著差異。三種定年方法算出的成熟材與未成熟材境界約在19年左右，三種方法間沒有顯著的差異。 對高岳營林區平均每公頃的材積數為354.4 m3，而此林分為36年生林分，換算為平均每年每公頃的樹幹材積固碳量約2.45 ton。而樣木整體平均製材率為39％。整體之總消耗電量而換算成釋出之碳素量為8.93 kg，淨固碳重為457.77 kg。棲蘭山的柳杉平均每年每公頃的樹幹材積固碳量約2.38 ton，而對高岳營林區和棲蘭山的柳杉，兩者每年每公頃的固碳量無差。 當含水率低於纖維飽和點時，縱向抗壓強度與含水率間呈現負線性相關。若以12％含水率為基準，含水率每增減1％時，平均縱向抗壓強度減增3.60％；若以絕乾含水率為基準時，當含水率增減1％時，平均縱向抗壓強度會減增2.52％。含水率在高於纖維飽和點時，縱向抗壓強度與含水率沒有明顯的相關性。以破壞型態則以壓碎型抗壓破壞出現的頻率為最高，當碎裂型抗壓破壞出現時，其抗壓強度較大；而剪斷型抗壓破壞出現時，其抗壓強度較小。

The study used nondestructive testing (ultrasonic instrument and Pilodyn) and destructive testing (longitudinal compressive strengths, LCS) to research the effect of different thinning treatments and grown condition (dominant, intermediate and suppressed trees) on the mechanical properties, wood anatomy and carbon storage of Japanese cedar (Cryptomeria japonica) located at Chilanshan area and Dweigaoyuei area. Additionally, the study also established the relation between nondestructive and destructive testing. The mechanical properties of Japanese cedar located at Chilanshan area with different strip thinning intensities showed that the longitudinal ultrasonic velocities, longitudinal dynamic moduli of elasticity (DMOE), radial ultrasonic velocities, radial DMOE, longitudinal compressive strengths and penetration depths subjected to stripe thinning of various intensities 16 years prior to the study were analyzed using the ANOVA method. No significant difference among the variables was detected. Positive correlations existed between the wood ring densities (RD) and earlywood densities (ED), latewood densities (LD), minmum densities (Dmin), maximum densities (Dmax), and latewood proportions (LWP), however. Samples from 2 plots at the same region but with different growth conditions tended to significantly differ in densities, longitudinal ultrasonic velocities, surface hardness penetration depths, DMOE and longitudinal compressive strengths. Also, sapwood and heartwood from the different plots exhibited different results as well. Wood density correlated positively with the DMOE and compressive strengths. Whereas, DBH of the trees showed negative correlations with the ultrasonic velocities, wood surface hardness penetration depths correlated negatively the compressive strengths of the sapwood. So was that between the wood surface hardness penetration depth and their DMOE. Japanese cedar trees from the untreated plots of the Dweigaoyuei experimental forest had the least DBHs and the greatest standard deviations. The harvested logs exhibited 4 types of ultrasonic velocity patterns each both along the steams and across the stems. Among the logs from different planting densities, there were marked difference between the Type A and Type B planting densities. All Type B logs had greater compressive strengths and micro destructive test strengths than those of Type A. There was no apparent distinction between mature wood and juvenile wood, however. The compressive strengths of the wood had a significant linear correlation with the wood densities. The influences of tending practices on the lengths of tracheids could be separated into the planting densities and thinning effects. The average tracheid lengths of planting density Type B were all greater than those of the Type A density. The tracheids of mature wood tended to maintain the trend. In juvenile wood, the pattern was indistinct. Juvenile tracheids from the 6 plots and those from the post-first thinning having different planting densities all were not significantly different. Yet, the tracheid lengths of trees from the planting density Type B were greater than those of the trees from planting density Type A. Tracheid lengths increased from pith outward to the bark, the lengths showed a highly linear correlation with the tree ages, which started to moderate between 15 to 22 years of age. Afterward, there were only minor fluctuations without notable increases. There were distinct differences between the tracheid lengths of the mature and juvenile woods. Boundary of mature and juvenile wood as determined by 3 different methods indicated that the transition occurred at ca. 19th year. There was no significant difference among the results of the methods. The timber volume stocked at the Dweigaoyuei forest averaged 354.4 m3/ha. Based on the stand age of 36 years, it translated to an annual incremental tree stem carbon sequestration rate of 2.45 ton/ha. The sampled logs had an overall lumber output of 39%, and electricity consumption equal to 8.93 kg carbon emission. There was a net carbon sequestration of 457.77 kg from the logs. On the other hand, Japanese cedar forest of Chilanshan averaged 2.38 ton/ha per year from increments of stem volumes. There were no differences between the forest carbon sequestration capacities of these 2 regional forests. When the moisture content (MC) of the logs was below the fiber saturation point, there was a negative linear correlation between the MC and the longitudinal compressive strength. If MC of 12% was taken as the basis, every percentage increase in MC caused an increase of 3.6% in the average longitudinal compressive strength. If over-dry MC was taken as the basis, then every percentage change in MC entailed a modification of average longitudinal compressive strength of 2.52%. At MC above the fiber saturation point, however, there was no apparent correlation between the variables. As for the patterns of test failures the crushing failure occurred at the highest frequency. When this occurred, a higher compressive strength was observed. Whereas shear failures often accompanied by lower compressive strengths.