能高越嶺道五種不同植被類型蜘蛛多樣性之比較

Abstract

許多學者從事蜘蛛多樣性的研究，不過多數在低海拔地區進行，中、高海拔地區的研究相對較少。能高越嶺道橫跨中央山脈，海拔約1600~2900 m，全線具有多樣的植被類型。本研究目的欲了解能高越嶺道不同植被類型的蜘蛛組成與多樣性是否有差異，以及這些植被類型的蜘蛛組成如何分群。植被類型分別為松-闊葉樹混淆林、松林、鐵杉-闊葉樹混淆林、鐵杉-冷杉混淆林及草原。每種植被類型設置8個5 m × 5 m的樣區，共40個樣區。採樣使用掉落式陷阱及掃網兩種方法，在每個樣區採樣地表及灌叢活動的蜘蛛。在2010年3月至2011年3月(2010年7月及2011年2月除外)的調查共紀錄蜘蛛28科159種9155隻，其中成蛛2847隻，若蛛6308隻。掃網捕捉112種6565隻蜘蛛，掉落式陷阱捕捉65種2590隻蜘蛛。物種累積曲線顯示松-闊葉樹混淆林的採樣最接近飽和，而其他四種植被類型仍有許多物種未被採集。以ANOVA分析五種植被類型之蜘蛛群落結構結果顯示不同植被類型間的蜘蛛群落結構有顯著差異：松林、松-闊葉樹混淆林及鐵杉-闊葉樹混淆林的Shannon Index顯著高於草原及鐵杉-冷杉混淆林；鐵杉-冷杉混淆林的Simpson Index則顯著高於松-闊葉樹混淆林、松林及鐵杉-闊葉樹混淆林；鐵杉-闊葉樹混淆林的Evenness Index顯著高於鐵杉-冷杉混淆林。分層比較地表及灌叢之蜘蛛多樣性發現，灌叢之物種數與數量較地表為多，但兩者的多樣性指數沒有顯著差異。利用兩兩樣區間之Euclidean distance進行的群聚分析顯示所有樣區可區分為三群，分別是森林群、草原群與混合群。優勢種的棲地偏好分析顯示：Linyphiidae A傾向分布在鐵杉-闊葉樹混淆林及松-闊葉樹混淆林；Pardosa laura和Linyphiidae D傾向分布在草原；Linyphiidae L傾向分布在松-闊葉樹混淆林。本研究成果可作為日後環境監測的參考。Spider diversity was studied in many low altitudinal areas, whereas studies in medium or high altitudinal areas were fewer. Nengkao Cross Mountain Trail is about 1600~2900 m a.s.l., crossing through Central Mountain Range, and having divergent vegetation types. The aim of this study is to understand whether the spider composition and diversity are different among different vegetation types in Nengkao Cross Mountain Trail, and how they are clustered. The vegetation types are mixed pine-hardwood forest, pine forest, mixed hemlock-hardwood forest, mixed hemlock-fir forest, and grassland. In each vegetation type, eight 5 m × 5 m sampling plots were established. Spiders were sampled by pitfall traps and sweeping. From March 2010 to March 2011 (except July 2010 and February 2011), 9155 spiders belonging to 28 families and 159 species were recorded in Nengkao Cross Mountain Trail. Among them are 2847 adults and 6308 juveniles. Six thousand five hundred sixty five spiders belonging to 112 species were captured by sweeping, while 2590 spiders belonging to 65 species were captured by pitfall traps. Results of species accumulation curve demonstrate that the sampling was most saturated in mixed pine-hardwood forest, while still many species were not collected in the other four vegetation types. The structure of spider communities was significantly different among different vegetation types. Shannon Index of pine forest, mixed pine-hardwood forest and mixed hemlock-hardwood forest are significantly higher than those of grassland and mixed hemlock-fir forest. However, Simpson Index of mixed hemlock-fir forest is significantly higher than those of mixed pine-hardwood forest, pine forest and mixed hemlock-hardwood forest. Evenness Index of mixed hemlock-hardwood forest is significantly higher than that of mixed hemlock-fir forest. Comparing spider diversity between ground layer and bush layer, species richness and abundance at bush layer are more abundant than those at ground layer, but diversity indices are not significantly different between them. Result of cluster analysis using pair-wise Euclidean distance shows that all sampling plots can be divided into three clusters, including forest, grassland and mixed group. Habitat preference of dominant species was assessed by comparing spider abundance among different vegetation types. The results indicate that Linyphiidae A tends to distribute in mixed hemlock-hardwood forest and mixed pine-hardwood forest; Pardosa laura and Linyphiidae D tend to distribute in grassland, and Linyphiidae L tends to distribute in mixed pine-hardwood forest. Results of this study can be a reference for environmental monitoring.