柳杉栽植距離試驗之林分斷面積生長與收穫

Abstract

本研究以台大實驗林溪頭營林區173號柳杉(Cryptomeria japonica)造林地內之栽植距離試驗林分為試驗區，採用非線性混合效應模式結合2個含有3個參數的Logistic 生長模式，分析2m×2m、3m×3m、4m×4m及 5m×5m等不同栽植距離之小區胸高斷面積(BA)生長資料。第一個生長模式公式為 ，BA(t)為t時間的胸高斷面積(m2 0.1-ha-1)，asymp (m2 0.1-ha-1)為模式之漸近線，b 為形數參數，rc (year-1)為生長速率參數。第二個生長模式公式為 ，BA(t)為t時間的胸高斷面積，asymp為模式之漸近線，b1為到達反曲點的時間，b2為從反曲點到達3/4漸近線所花的時間。 分析結果為漸近線(asymp)在兩個模式下皆為混合效應。在第一個模式下，栽植距離對漸近線及形數參數b有顯著影響，但不影響速率參數rc。在第二個模式下，栽植距離對於漸近線及b1有顯著影響，但不影響b2。兩個模式所預測的漸近線結果一致。55年生時，2m栽植距離下平均胸高斷面積為最大(7.29 m2 0.1-ha-1)，5m栽植距離下平均胸高斷面積為最小(4.32 m2 0.1-ha-1)。根據第二個模式，四個栽植距離下到達反曲線的時間分別是試驗建立後的第16、19、22及24年，初始的栽植密度愈密，愈早到達生長反曲點。無論在哪一種栽植距離下，林木的內在增殖率皆為0.13，且從反曲點到達3/4漸近線的時間為8年，且各種栽植距離至50年後，胸高斷面積年生長量都接近0。 研究的結果表示最適宜的栽植密度為每公頃2500至3000株，疏伐宜於15至20年施行，輪伐期約為30年，和現今對於柳杉的撫育及經營慣例相似。

Using a nonlinear mixed-effects modeling approach, this study analyzed the basal area (BA) growth of a long-term Japanese cedar (Cryptomeria japonica, aka sugi) spacing trial in Sitou District of the Experimental Forest, National Taiwan University. The spacings analyzed were 2m×2m, 3m×3m, 4m×4m, and 5m×5m. To reveal different aspects of BA growth, two 3-parameter logistic growth models were used to analyze the data. The first model was the traditional logistic model with the form , where BA(t) is the BA (m2 per 0.1 ha) at time t (year), asymp (m2) is the asymptote of the model, b is a scale parameter, rc (year-1) is the rate parameter. The second model had the form , where BA(t) is the BA at time t, asymp is the asymptote of the model, b1 (year) is the time need to reach the inflection point, and b2 (year) is the time needed to reach approximately 3/4 of the asymptote after reaching the inflection point. The results showed that only the asymptote (asymp) in both two models should be considered as a mixed-effect. For the first model, spacing had a significant effect on both asymptote and the scale parameter b, but not on the rate parameter, rc. For the second model, spacing had a significant effect on both asymp and b1, but not on b2. The asymptote estimates from the two models were consistent. After 55 years, the narrowest spacing had the highest average BA per unit area (7.29 m2 per 0.1 ha), whereas the widest spacing had the lowest value (4.32 m2 per 0.1 ha). Based on the results from the second logistic model, the inflection point for the four spacings was reached around 16, 19, 22, and 24 years after the establishment of the spacing trial. The closer the initial spacing, the earlier the inflection point was reached. Regardless of the initial spacing, the estimated rate parameter was 0.13, and the time needed to reach 3/4 of the estimated maximum BA growth was 8 years after reaching the respective inflection points. After 50 years, the annual BA growth was close to 0, regardless of the initial spacing. This study confirmed the current silvicultural and management practices prescribed for the species, that is, the optimal planting density should be between 2500 to 3000 trees ha-1, thinning should commence between age 15 and 20, and the rotation period of the species should be around 30 years.