臺灣柳杉人工林產能管理之研究

Abstract

現今臺灣的人工林大多缺乏密度管理，其過度擁擠的林分直接或間接的造成產能停滯與木材自給率低迷的問題。本研究利用數學模式發展出一套管理柳杉人工林的工具，為幫助管理者擬定輪伐期間產能管理的策略以及評估柳杉人工林的優劣，期能幫助穩定的輸出木材。本研究首先將臺大實驗林長期試驗地與林務局永久樣區的數據集進行資料的比對與偵錯反饋，將校正後的數據分級成9個百分層級後，進一步利用Schunte生長模式擬合不同齡級各PR的「林齡級-林分胸徑」以及「林齡級-林分密度」的生長曲線組，合併曲線組為邱氏密度管理量表。同時利用有母數法與無母數法擬合柳杉人工林的胸徑大小與密度軌跡，並進行互相驗證，由此作為立木度管理圖的基礎。另一方面，考量蓄積量的計算問題，利用浮動樹高值與形數計算出不同情況的「蓄積量模擬量表」，能幫助使用者迅速在量表中，點出各種柳杉人工林林分狀態的的蓄積量。為幫助第一次疏伐規劃，合併立木度管理圖與邱氏密度管理量表衍生出「C-檢索表」。研究結果顯示利用Schunte生長模式擬合的林齡-林分胸徑曲線組與林齡-密度曲線組誤差值偏低，而且量表也能夠有效的演示文獻數據與虛擬數據的疏伐前後變化，由此可見管理者能利用邱氏密度管理量表的PR50作為生長基線評估生長優劣與擬訂疏伐計畫，也能預測各種生長性狀柳杉人工林的林分隨林齡的變化。而有母數法與無母數法都能夠有效的擬合與驗證林分發展的軌跡，為便於林業實務上的應用，建議採用有母數法。進一步應用有母數法的參數計算林分進入各階段間分界點的係數為SDI閥值，由此製作立木度管理圖，其搭配「蓄積量模擬量表」能在內業規劃輪伐期的時候快速估算疏伐前後蓄積量變化與總產量，從不同尺度檢視每階段的生長與收穫軌跡，進一步的掌控整個輪伐期的產能管理。未來期能進一步以資料庫的形式推廣，並應用在不同樹種的經營管理上。

Most of Taiwan's plantations lack of density management today, and the overcrowded stand structure either directly or indirectly leads to stagnant productivity and low timber self-sufficiency rate. This study used mathematical models to develop a set of tools for managing Japanese cedar (Cryptomeria japonica) plantations, and it aims to provide a quantified strategy for managers to evaluate Japanese cedar’s production during rotation in order to stabilize the output of. In this study, the data was compared and debugged between the long-term experimental plots of the National Taiwan University Experimental Forest and the permanent plots of the Forestry Bureau. After grading the processed data into 9 percentile ranks, the Schunte growth model was imported to fit the growth curve groups of 'age - size' and 'age - density' of each PR from different age classes, then the different curve groups were merged as the 'Chiou’s density management diagram'. Meanwhile, the size-density trajectories of Japanese cedar plantations were fitted by the parametric and nonparametric methods. The basis of 'stand density index management diagram' was the outcome of mutual authentication between those two methods. On the other hand, considering the calculation of the volume, various value of tree height and form-factor were used to draw a volume simulation diagram for different situations. This could help the user to point out volume of Japanese cedar plantations promptly. In order to assist the first thinning plan, a ' Classification Chat ' was derived by combining the Chiou’s density management diagram and stand density index management diagram. The results showed the error values of age-size curve group and age-density curve group which was fitted by Schunte growth, were relatively low. Moreover, Chiou’s density management diagram could also effectively demonstrated the changes between literature data and virtual on before and after thinning. Hence, managers could use the PR50 of Chiou’s density management diagram as the growth baseline to evaluate the growth quality, formulate thinning plans, and predict the change of Japanese cedar plantations’ various growth traits with stand age. Both parametric and nonparametric methods could effectively fit and verify the trajectories of stand development. Parametric method is suggested for facilitate the application in forestry practice. SDI threshold used the coefficients of demarcation point among each stage of the standby parametric method, and the 'stand density index management diagram' was came out from it. The combination of 'stand density index management diagram' and “volume simulation diagram” could quickly estimate volume changes and total harvest before and after thinning, view the growth and yield trajectories of each stage from different scales, and control the production capacity management of the entire rotation period. In the future, it could be promoted in a form of database and applied to the management of different tree species.