發展林分相對密度指數於檜木及松樹混合林

Abstract

林分密度指數可以呈現林內競爭強度，其利用決定初期栽植密度與疏伐等方式來經營林木的空間分布、控制樹冠的發展及維持森林的健康程度。最典型的林分密度指數在1933年由Reineke所建立，此指數於人工純林有很佳的研究結果，但並不適於混合林，目前可替代的林分密度指數仍被探索中。在臺灣，天然林的面積大約佔有105萬公頃，其中85 %的天然林面積屬於混合林，當前林分密度指數的相關研究多著重於單一優勢物種的天然林或是單一物種的人工林，以紅檜與柳杉的研究較多，極少研究致力於調查或發展可適用於臺灣天然混合林的林分密度指數。由於進行林分密度的管理，對於森林保護與水土保持等目標極為重要，因此，開發天然混合林的林分密度指數來制定森林經營管理方針是有必要的。本研究將臺灣第四次森林資源調查的檜木混合林及松樹混合林等資料，代入Ducey和Knapp在2010年建立的相對密度模式，為了使臺灣的資料能在此模式下有顯著的結果以及得到更完整的相對密度模式建立方針，代入的過程需要再額外加入三個步驟，包括添加1 %的隨機雜訊於所有樣木的比重值、以大於10 %的檜木或松樹胸高斷面積比例作為樣區挑選的基準值、選擇最大值、最小值的相交分位數以及各參數平均值來決定相對密度模式的參數估計。本研究不僅提供獲取某混合林林分相對密度公式的初步流程，更提供一個快速評估檜木混合林以及松樹混合林等樣區的相對密度公式，以利對此樣區進行森林經營管理之策略。

Stand Density Index (SDI) is a numerical value that captures intensity of competition within a forest stand. It is a tool for managing spatial arrangement of trees, controlling crown development and maintaining degree of forest health through decision made on initial planting density and thinning schedule. However, classic Reineke’s SDI (Reineke 1933) has been found to be unsuitable for mixed species and structurally complex forest stands. Alternative measures of SDI are being explored. Natural forests in Taiwan cover an area of approximately 1.5 million ha with 85% of them classified as mixed species forests. Current SDI research in Taiwan focuses on single-species natural and plantation forests such as Chamaecyparis formosensis and Cryptomeria japonica forests. Very few studies investigate and develop SDI for mixed species natural forests in Taiwan. Because management of these forests is crucial for conservation and protection against soil erosion, it is necessary to develop SDI for mixed species natural forests to establish guidelines for management of these forests. Therefore, based on the model by Ducey and Knapp (2010), relative SDI was developed for false cypress and pine forests using data from the 4th Taiwan National Forest Inventory. Plots with at least 10% of basal area per hectare of target species were used for model fitting. During model fitting, it was discovered that specific gravity played an important role on model convergence. Lastly, three different relative SDI models for each forest type were produced that predicted minimum, mean and maximum relative SDI for a forest stand. The major result from this study was that relative SDI could now be calculated for the mixed-species and structurally complex false cypress and pine forests that allows a manager to meet forest management strategies.