臺灣對於木質藤本結構與分布的資訊較為缺乏,多為森林植群生態研究。本研究地區位於臺東縣達仁林場,屬於東南氣候區的低海拔次生林。本研究設置60個400 m2的方形樣區,以及60個100 m2的小區,量測16項環境因子以及覆蓋度與胸徑等植物介量,並使用典型對應分析探討環境因子與木質藤本分布的相關性。藉由變異數分析測試(1)不同徑級木質藤本所攀附的樹木胸徑、不同徑級樹木所承載的木質藤本覆蓋度是否有顯著差異;(2)不同地形、攀爬機制之間的木質藤本胸徑、支持木胸徑是否有顯著差異;(3)樹冠層與灌木層樹木之間的木質藤本承載率是否有顯著差異。並以卡方檢定測試物種間的相關性,探討木質藤本與支持木之間的相關程度。調查結果共記錄 40 科 64 屬 86種木質藤本。以溪谷的植物多樣性最高,木質藤本株數最多,樹木直徑最大;坡地的木質藤本胸徑最大;稜線的植物多樣性最低,木質藤本和樹木胸徑最小。典型對應分析的結果顯示樹木與草本覆蓋度等生物因子,對於木質藤本分布的解釋能力最大。變異數分析的結果顯示,不同徑級木質藤本攀附的支持木質胸徑有顯著差異,支持木的胸徑越大,木質藤本的胸徑也越大。不同攀爬機制的木質藤本能夠攀附的支持木胸徑也有所差異,吸附性木質藤本能攀附的支持木胸徑顯著高於依靠性與纏繞性木質藤本。樹木與木質藤本的種間相關測試結果顯示,木質藤本與樹木多為正相關,香楠是與最多木質藤本有關的樹木;南五味子是與最多樹木相關的木質藤本。樹冠層樹木的木質藤本承載率顯著低於灌木層樹木,平均承載率最高的樹木是倒卵葉楠、最低的是杏葉石櫟與臺灣光蠟樹。
There was a paucity of relative knowledge about liana structure and distribution in Taiwan, most studies focused on forest vegetation ecology. Study area were located at Tajen experimental forest station belongs to lowland secondary forest of southern east climate. Vegetation was sampled in relev#westeur042#s of standard size of 400 m2, a total of 60 relev#westeur042#s and 100 m2 subplots in each relev#westeur042#. In each relev#westeur042#, we measure 16 environmental factors and plant parameter such as breast height diameter (DBH) and coverage. Using canonical correspondence analysis (CCA), the relationship between environmental factors and liana distribution could be figured out. Analysis of variance (ANOVA) was used to test significant difference among groups (1) the host tree DBH of different diameter level lianas and the liana loads of different diameter level trees; (2) the liana DBH and host tree DBH within different topography; (3) the liana loads of tree layer and shrub layer. Correlation of liana and host tree was investigated through Pearson's chi-squared test. We recorded 86 liana species that belonged to 64 genera and 40 families. Among the different topographies studied, valley had the highest plant diversity and liana individuals; slope had the largest DBH liana, the plant diversity and plant DBH of ridge was the lowest. Results of CCA indicated the coverage of trees and herbs were the most explainable environmental factors of liana distribution. Results of ANOVA showed there was the significant differences among different diameter level lianas, the bigger host tree DBH, the bigger liana DBH. Lianas of different climbing mechanisms also had significant differences in the host tree DBH; the host tree DBH of adhesive lianas was significantly higher than twining and leaning lianas. Species correlation tests showed tree-liana correlations most of the time. Machilus zuihoensis had relation to much more liana species, indicating it might be an ideal host tree for lianas; Kadsura japonica had relation to much more tree species. The liana loadings of tree layer trees were significantly lower than shrub layer trees. The highest average liana loadings tree was Machilus obovatifolia. The lowest average liana loadings trees were Lithocarpus amygdalifolius and Fraxinus griffithii.