銀合歡人工林生物量估算方法之比較

The Comparison of Different Methods for Estimating the Biormass of Leucaena leucocephala Plantations

劉宣誠(Shen-Chen Liu) ； 高毓斌(Yu-Ping Kao)

《林業試驗所研究報告季刊》 2卷1期 (1987/03) Pp. 43-64

https://doi.org/10.7075/BTFRI.198703.0043

摘要

本研究係以實際生物量為比較依據，探討5種常用之生物量估算方法，應用於銀合歡人工林之可行性，期能尋求最理想之測定技術，以解析銀合歡物質生產之特性。試區設定於三民與南澳之5年生銀合歡林分，各逢機設置3處0.01公頃(10×10m)之樣區，樣區內所有立木經逐一伐倒並實測其生長及各部位重量，以求獲其實際生物量。所有樣木資料復以平均木法，分層平均木法，底面積比率法，材積比率法及廻歸估算法，分別求取各部位生物量之估值，藉以比較各方法之相對精確度。
以平均木法估算三民試區之葉部、枝部、幹部、可利用幹部及地上部等各部位生物量（乾重）時，各較實際值低估45.5%、29.3%、44.1%、73.9%及43%；對南澳試區之幹部與可利用幹部生物量則各低估23.8%及27.5%。若改以分層平均木法為估算依據，上述低估現象可獲改善，對三民試區各部位生物量之低估率僅在0.6%~1.9%之間；對南澳試區之幹部及可利用幹部生物量則各低估2.4%及2.7%。若改用底面積比率法或材積比率法，其估值與實際生物量之差額率將較分層平均木法為大。
銀合歡各部位之重量與胸徑間具存強烈之曲線關係，故簡單線型廻歸模式不宜為重量生長之解析；二次曲線式之適合指數較相對生長式為大，且其估值與實測值之差額率較相對生長式者為小。以分層平均木法所選取之樣木，經二次曲線式配置各部位重量與胸徑之廻歸關係，若據此以估算三民試區之葉部、枝部、幹部、可利用幹部及地上部生物量時，各低估7.0%、0.1%、4.2%、3.0%及3.7%；對南澳試區之幹部及利用幹部生物量則各低估0.2%及0.5%。除可利用幹部外，各部位重量與胸徑之關係式，在三民與南澳試區間均顯呈差異，未能導出其共同廻歸式；然而，幹部生物量與地上部生物量係與（林分底面積×平均樹高）呈極顯著正相關，變異係數僅各為3.2%及4.9%，藉此生物量預測式之導出，將可擴大廻歸估算法之應用價值，並供為地區性生物量調查之基礎。

並列摘要

The above-ground biomass of two 5-year-old Lcucaena leucocephala stands at San-Ming and Nan-Our was estimated by mean tree, stratified mean tree, basal area ratio, volume ratio and two regression methods. The accuracy of different methods in estimating the dry production was evaluated by comparing with the actual biomass production obtained from three study plots (l0×10m) in each site where all the trees were felled and weighed. Lowest estimates were consistently obtained with the mean tree method with the component biomass being 24-74% underestimated. The stratified mean tree method was the best method among the four non-regression methods in estimating the leucaena biomass production with the component biomass were only 0.6-2.7 underestimated. The best regression method which being based on the quadratic equations relating the component weights and DBH gave an underestimation of 0.1-7.0% for the component biomass. The quadratic regression method is recommened for estimating the above-ground biomass in leucaena plantations for its accuracy and data transferability. However, if facilities are limited, the stratified mean tree method can also provide a simple and acceptable estimation.