墾丁高位珊瑚礁森林地區不同幹擾程度棲地間底棲性蜘蛛多樣性之比較

Abstract

本研究使用50組掉落式陷阱，於墾丁高位珊瑚礁地區的五種棲地中針對底棲性蜘蛛進行採集以探討不同類型之幹擾會對其群聚結構及多樣性造成何種影響。五種棲地類型分別為原始林保留區（A）、原始林遊憩區（B）、次生林（C）、草生地遊憩區（D）及廢置之草生地遊憩區（E）。為期一年每月一次之採集共獲得2247隻成熟的個體，分屬於20科110種，其中有86種（78.2％）為新種或新紀錄種。數量較多之種類依據其季節性數量變化可區分成兩類生殖活動週期型態：乾季型（Linyphiidae B, Pardsa tschekiangensis, Phrurolithus lynx, Pardosa sp. J, Theridiidae G, Ctenus yaeyamensis, Linyphiidae L, Lycosidae E, Lycosidae L）和濕季型（Mallinella fulvipes, Mallinella shimojanai, Lycosidae B）。在棲地分析方面，Pardosa tschekiangensis和Ctenus yaeyamensis傾向分佈於草生地，Mallinella fulvipes, Mallinella shimojanai傾向分佈於原始森林，Theridiidae G傾向分佈於次生林，Ctenus yaeyamensis傾向分佈於原始森林保護區，而Lycosidae E則傾向分佈於草原。各多樣性與均勻度指數在五種棲地間並無顯著差異；各棲地之群落皆由少數優勢種及大量稀有種所組成。在110種底棲行蜘蛛中，有61種（55.5％）僅分佈於單一棲地。50組陷阱所獲之樣本經集團分析可被區分為原始森林群、次生林群、草原群與草原偏森林群。功能群分析之結果顯示，五棲地之功能群組成有極顯著差異。在原始林主要以底棲性織網群為主，在次生林中則是底棲性織網群和樹棲性織網群為主，而草原中則是主要由埋伏等待群所組成。雖然各樣區受到不同程度幹擾的影響，但是其種豐度和群聚結構均無顯著的差異。除此之外，各類型的棲地均有相似數目的獨特種，這結果顯示底棲性蜘蛛多樣性在墾丁高位珊瑚礁森林地區的異質度高，可作為經營管理的參考。

The effects various disturbances on diversity and community structures of ground spiders in Kenting National Park uplifted coral reef forest were investigated using pitfall traps. In each of the following sampling sites, ten trap stations were established evenly along the trail: primary forest (A), primary forest with tourism activities (B), secondary forest (C), grassland with tourism activities (D) and abandoned grassland (E). A total of 2237 spiders were collected, which included 20 families and 110 species. Among them 86 (78.2%) are new or new record species to Taiwan. Dominant species can be divided into two major groups according to their reproductive activity patterns generated from temporal abundance variations: dry season breeder (Linyphiidae B, Pardsa tschekiangensis, Phrurolithus lynx, Pardosa sp. J, Theridiidae G, Ctenus yaeyamensis, Linyphiidae L, Lycosidae E and Lycosidae L) and wet season breeder (Mallinella fulvipes, Mallinella shimojanai and Lycosidae B). Habitat preference of dominant species was assessed by comparing relative abundance between sampling sites. The results indicated that Pardosa tschekiangensis tends to distribute in the grassland; Mallinella fulvipes, Mallinella shimojanai and Ctenus yaeyamensis tend to distribute in the primary forests; Theridiidae G tends to distribute in secondary forest, and Lycosidae E tends to distribute in the grasslands. Shannon-Weaver function, Simpson index and evenness were not significant different between five sampling sites. Community structure of all sites was characterized by few dominant species and numerous rare species. Among 110 species obtained, 61 distributed in one sampling site only. Each sampling site exhibited between 11 to 16 unique species. Result of a cluster analysis using pair-wise Euclidean distance demonstrated that collections of 50 trap stations can be divided into four major clusters: primary forest, secondary forest, grassland I and grassland II. Foraging guild composition differed significantly among sampling sites. Primary forest site was dominated by ground level web builders, secondary forest was dominated by ground level web builders and aerial web spinners, and grassland was dominated by sit and wait ambusher. Although each sampling site receives various type of disturbance, their species richness and community structure do not differ significantly. Besides, each sampling site exhibit similar number species. These results suggest that the diversity of ground spiders in KTNP uplifted coral reef forest is quite heterogeneous, any management conduce should consider the uniqueness of each habitat type.