松鼠為害對溪頭地區柳杉人工林樹高生長之影響及其空間分佈

Abstract

本研究以國立臺灣大學生物資源暨農學院實驗林管理處所轄溪頭營林區2 林班49-1 號造林地試驗區中的柳杉林分為研究材料。本研究之目的為，第一為對樣木進行樹幹解析，再以樹高估算方法估算樹高值，並以Richards 生長模式模擬樹高生長結果，建立柳杉健全木與鼠害木的樹高生長模型，比較兩者在樹高生長上之差異，第二為建立鼠害快速目測評估機制，評估鼠害之空間分佈與嚴重程度，透過林分視覺化系統呈現此研究區域中之松鼠為害情形。 研究結果顯示以三種樹高估算方法（Graves、Lenhart’s、Carmean’s）之樹高估算結果建立健全木之樹高生長模型，樹高生長率最大值在2 年內達到最高，顯示柳杉於生長初期即能達到最高生長率。鼠害木樹高生長停滯的年齡較健全木生長停滯的年齡晚，顯示鼠害造成林木樹高生長延遲。樹高生長模型參數之信賴區間在Lenhart’s 與Carmean’s 的預估結果所建立之樹高生長模型中近乎相同，顯示此兩種樹高生長模型高度重合。樹高估算方法的預估結果以Graves 方法高Lenhart’s 與Carmean’s 方法。 鼠害調查研究結果顯示，此區鼠害木百分比高達80.98% (± 7.65%)，其林分為易受害之柳杉純林，樣木遭受鼠害的比率由試驗區西南方向東北方遞增。以樹冠顏色來判斷鼠害走向，顯示鼠害發生的走向可能由試驗區西南方向東北方移動。受鼠害影響程度以樣區6 較輕微，樣區4、樣區5 較嚴重。樣區6、樣區7樣木重複遭受鼠害之比率較高。樣區4 鼠害木百分比最高，亦具有最多的鼠害分岔，顯示樣木重複受害之比率最高，且鼠害規模較大。鼠害分岔之高度有由南向北、由西向東增加之趨勢，從木材利用的角度來看，試驗區西南部的可用材少於試驗區東北部。以林分視覺化系統展示鼠害分佈之情形，可以看出鼠害分岔高度越高之樣區，其樹冠層佔樹高的百分比越低。使用林分視覺化系統對於評估林分中之鼠害情形將更有效與準確。

A Japanese cedar (Cryptomeria japonica) plantation in the 49-1 Forest Stand in the NTU Experimental Forest was studied. The purposes of this study were two-folds. Firstly, sampled trees were felled and stem analysis were carried out, tree height estimation methods were used to estimate tree height at each given age, and Richards’ growth model was used to model tree height growth for healthy trees and squirrel damaged trees, which differences in tree height growth between the two were compared. Secondly, a rapid visual evaluation mechanism for squirrel damage was established, spatial distribution and severity of squirrel damage of the forest stand were assessed, and the Stand Visualization System was used to present the conditions of squirrel damage in the research area. Results compared estimation of tree height between three height estimation methods (namely, Graves, Lenhart's, Carmean's). The estimated heights were successfully used to establish height growth models of healthy and squirrel damaged trees. The age for the fitted tree height growth curves to reach their highest growth rate was between 0 and 2 years old suggesting that Japanese cedar reached its highest growth rate in early development stage. The age at which height growth stagnate for squirrel damaged trees was later than that of healthy trees indicating that squirrel damage caused delay in tree height growth. The confidence intervals of the parameters of the fitted tree height growth models were almost the same between the two tree height growth models built from Lenhart ’s and Carmean ’s estimates, showing that the two fitted tree height growth models were highly coincident. The predictions from the tree height growth model fitted with height estimated by the Graves’ method was higher than that of Lenhart ’s and Carmean ’s methods. The results of the squirrel damage assessment showed that percentage of squirrel damage trees in this area was as high as 80.98% (± 7.65%). This forest stand is vulnerable because of its pure Japanese cedar plantation. The rate of squirrel damage increased from southwest to northeast of the forest stand. Judging the direction of squirrel damage by the color of the tree crown showed that the trend of squirrel damage might move from the southwest to the northeast of the forest stand. The impact of squirrel damage was light in sample plot 6, whereas sample plots 4 and 5 were more serious. The sample area 6 and the sample area 7 have a higher rate of repeated squirrel damage. Sample area 4 had the highest percentage of squirrel damage, and also had the largest number of branching due to squirrel damage indicating that the rate of repeated damage to sample trees was the highest and the scale of squirrel damage was relatively large. The height of the squirrel damage bifurcation tended to increase from the south to the north and from the west to the east. From the perspective of wood utilization, available materials in the southwest of the forest stand were less than those in the northeast of the forest stand. Using the Stand Visualization System to display distribution of squirrel damage, it could be seen that the sample plots with higher squirrel damage bifurcation height had lower percentage of its canopy layer in the tree height. The use of Stand Visualization Systems is more effective in assessing squirrel damage in stands.