南仁山區低海拔亞熱帶雨林林隙更新之研究

Abstract

本研究在恆春半島的南仁山區墾丁國家公園境內低海拔亞熱帶雨林設立2.1公頃永久樣區，作自然幹擾機制、林隙特性及林隙更新的調查與分析。結果發現該樣區屬幹擾頻繁的森林，估算每年每公頃樣區將生成2.5個林隙，面積達200 m2，由此而計算得之冠層轉換率只有53年。幹擾現象推測應以颱風?主，然而由一些遺跡判斷，本研究樣區也曾經遭受人?在某程度上的幹擾。在倒樹事件中，以根拔現象和幹折現象居多，也造成較多的林隙面積。除此，發現能單獨造成林隙的死亡樹木，只有喬木種類，以楠木類、印度栲?主，死亡樹木很快便腐朽。 在植物數量及組成方面，各選定10個年齡與面積有差異的林隙及其邊緣區域，另加三個非林隙鬱閉樣區（合計23個小樣區）作調查比較。由DCA（Detrended Correspondence Analysis）作樣區分佈序列，可合理的依據林分年齡區分?年輕、中等年齡、老的林隙與它們的邊緣區域及非林隙四類型樣區，但不依林隙面積作連續性之分佈。魚眼鏡頭照像測光照量的結果說明瞭光照量隨林隙年齡的增加而減弱，以非林隙樣區的相對光照量最低。 植物調查分析結果顯示本研究樣區先驅植物種數多，分佈廣，可見生長環境適合其繁衍。此外，發現所有優勢樹種有?數甚多的小苗及小樹，更新情況良好，而林隙的出現有助於優勢喬木種類小苗及小樹的生長發育，但次喬木與灌木優勢種可自行於林下更新，林隙的更新棲位角色重要性減低。至於草本植物數量及組成在林隙及非林隙的反應，現有資料難作判斷，恐需作更多的微環境因數的調查研究。

A 2.1-ha permanent plot was established in 1992 in Nanjenshan protected lowland subtropical rain forest on Hengchun peninsula, Taiwan. My purpose in establishing this permanent plot was to study the disturbance regime, gap structure and gap-phase regeneration of this forest type. An average of 2.5 canopy gaps formed per hectare and total gap area of 200 m2 per hectare was made annually. The turnover time of the canopy layer in the forest, which was calculated from the rate of the gap formation, was about 53 years. These findings strongly suggested that this forest was exposed to disturbance in high frequency. Typhoon could be the main origin of natural disturbance in Nanjenshan area, however, some man-induced disturbances were also found, which mainly happened before the area was protected. Cases of uprooted and snapped off were more frequent among the treefall regime and about half of the total gap area was created. The more abundant fallen trees, Machilus species and Castanopsis indica, which were capable of creating canopy gaps, may decay in a short period. Sampling of 10 gaps (and 10 gap-edges with respect to each gap) different in age and size and another 3 non-gap control sites was undertaken to assess the effects of gap age and gap size on species abundance and species composition. Detrended Correspondence Analysis suggested that it was reasonable to lump sample plots into 4 groups: young gaps and gap-edges, intermediate gaps and gap-edges, old gaps and gap-edges, and non-gap control sites, but the effect of gap-size was not found. The percentage relative skylight in each sample plot estimated with black-and-white hemispherical photographs decreased significantly with gap age, the lowest percentage of skylight was found in non-gap control sites. From the analysis of species abundance and species composition, pioneers were abundant and widely distributed, mainly due to the favorable environmental conditions for their regeneration. Early recruits were abundant for most of the dominant tree species, suggesting that they were capable to sustain their dominance in this forest. The seedlings and saplings of canopy tree species showed better survival or greater recruitment in gaps but sub-canopy tree species and shrubs seemed able to regenerate in situ. The gap responses of herbaceous species showed uncertainty as to results.