南仁山低地雨林木本植物社會之短期動態

Abstract

為瞭解南仁山低地雨林之植物組成及森林動態，本研究進行南仁山溪穀2.1ha樣區設立七年後的第一次複查，同時在鄰近地區再設置一0.64ha重複樣區，作時間軸及空間軸的比較。在重複樣區中，量測各胸高直徑?1 cm的植物個體，附以鋁牌，並標定其位置。而對2.1ha樣區複查，比對1993年之原始資料，以瞭解樣區的死亡率、新增率及直徑生長。此外，也對森林中的林隙的位置及狀態加以記錄，討論林隙在樣區中扮演的角色。 在2000年調查時，0.64ha樣區及2.1ha樣區之樹種共計有109種，分屬於43科77屬，樣區內植物底面積為44.1m2 ha-1，植株密度為3738 stems ha-1。樣區之優勢樹種為茄苳（Bischofia javanica Blume）、白榕（Ficus benjamina L.）、紅果椌木（Dysoxylum hongkongense (Tutch.) Merr.）及咬人狗（Dendrocnide meyeniana (Walp.) Chew）。樣區林隙面積為634.7 m2 ha-1，平均林隙大小為52.7±51.0 m2。TWINSPAN同時將0.64 ha及2.1 ha樣區分成四型植物社會，其中有一型和地形及海拔相關，另外三型則和不同林隙期相關。比較後顯示兩樣區極相似。 在2.1 ha樣區的森林動態方面，七年間有5種移出樣區，新加入3種，樣區現有樹種105種，分屬43科77屬。經過七年之後，雖優勢樹種名次上有少量變動，但前15名仍相同。樣區植物底面積的成長為2.63 m2 ha-1，變化量為6.75％。在密度方面則減少150 stemsha-1，變化量為-4.02％。七年間死亡率為25.04％，新增率為21.02％，死亡與新增個體主要為DBH＜4cm的植株及灌木樹種。而樣區各樹種個體數變化不顯著（p＝0.652）。樣區平均胸高直徑生長為1.88mm yr-1，七年間底面積增加量為5.8m2 ha-1，多由底面積優勢樹種所貢獻。整個樣區在1993年至2000年間樹種組成和植株數變化少，但底面積仍有成長。 比較13種樹種的生長潛力，可將之區分成：高速生長、中速生長、低速生長及其他生長四種類型，此四種類型表示其有不同的生長速度範圍，也暗示樹種的小徑級植株建立在鬱閉林下或林隙的能力。高速生長的樹種為非耐陰樹種，其更新和林隙有正相關。中速生長、低速生長及其他生長的樹種，小樹則可生長在較陰暗的林下。 比較樣區中優勢樹種（底面積前20名）、稀少種（植株數目小於10株）及常見種（植株數目大於或等於70株）時，優勢樹種以低死亡率，高新增率，或生長快速等不同的方式保持其優勢度。稀少種則在植株數上少變化。常見種相對於稀少種，有較低的死亡率和較高的新增率。因此，短期內優勢種和常見種都能繼續保持其優勢或常見的狀態。 在2.1ha樣區繪製現地林隙破空投影圖，並與1993年的林隙比較。結果顯示，林隙面積由2620m2減少為1458m2，1993年的canopy gap有69.3％樹冠高度達5m以上。根據所記錄的林隙期，將樣區的小樣方（5×5m2）加以比較。發現小樣方的死亡率及新增率會隨著林隙更新程度先增高再下降，底面積則是先減少再增加。以X2-test of goodness of fit檢驗樹種對更新生育地的選擇，其中有19種樹種更新會偏向林隙、林隙邊緣或林下，呈現具有更新生育地的分化。 綜合來看，研究樣區在時間軸及空間軸上組成變動都不大，可以以研究樣區來瞭解南仁山低地森林的組成。而在2.1ha研究樣區的森林動態中，林隙面積減少，植物底面積仍在成長，死亡率和新增率可以互補，使得植株數變動少。整個2.1ha研究樣區短期內應還會生長，但密度不會再有大變化。而林隙在樣區中為高新增和高死亡的場所，顯示其為提供樹種建立的機會地。

A permanent 2.1-ha plot, established in 1993, was recensused in 2000 to understand forest dynamics. A replicate 0.64 ha plot was also established in 2000. All trees ? 1 cm DBH were identified, measured, and mapped. In the 2.1-ha plot, all individuals have been remeasured in order to determining the rate of tree mortality, recruitment, and DBH increment. Besides, tree fall gaps were surveyed to understand gap dynamics in this forest. A total of 109 species, belonging to 77 genera and 43 families, were recorded in two plots. Dominant species were Bischofia javanica Blume, Ficus benjamina L., Dysoxylum hongkongense (Tutch.) Merr. and Dendrocnide meyeniana (Walp.) Chew. They accounted for 41.0% of the total basal area in the plots. The 2.1-ha plot was characterized by high mortality rate (25.04%, 7 yr) and high recruitment rate (21.02%, 7 yr). However, the overall change in individual numbers was small. Over time, the variation in recruitment and mortality mainly caused by individuals which DBH smaller then 4 cm and shrub species. The diameter growth of live stems and that of new recruits not only balanced the loss of basal area from dead stems, but also contributed 6.75% increment of the total basal area. For the 13 selected species, the maximum DBH increments showed four different performances, which more or less related with the ability of the tree species to establishing in the understory or gap sites. Different strategies among dominant species and common species made them keep their dominant or common status. After 7 years, the canopy gap area in study plot was decreased. There were 69.3% areas recovered from canopy gap. The mortality and recruitment rate were higher in canopy gaps than in non-gap area. By X2-test of goodness of fit, there were 19 species showing strong differentiation of regeneration habitat. The investigation conducted in 2000 indicated that the species composition of the study plots were almost the same in terms of space and time, and the total basal area of the forest will keep growing in the near future.