標題: 港灣結構物週邊底棲生物群集特性與其棲地模式之建立—以安平港為例
Cluster analysis on the benthos around breakwaters and model development of habitat evaluation procedure—An example of the An-Ping harbor
作者: 陳俊杰
Jun-Jie Chen
張憲國
Hsien-Kuo Chang
土木工程學系
關鍵字: 底棲生物;棲地模式;群集分析;港灣結構物;時空分佈;安平港;An-Pin harbor;HEP;HSI;mitigation;Similarity;Euclidean distance;hierarchical clustering;HU
公開日期: 2004
摘要: 本文研究安平港附近海域之底棲生物,以瞭解現有港灣結構物對附近海域之生態性之影響。透過連續四季的底棲生物及環境因子調查資料,來探討十八個測站的環境生態特性,及分析各測站之生物多樣性與相似度,並利用群集分析來探討底棲生物與底質的時空分佈,以瞭解底棲生物與底質環境間的關係,最後建立一套適合安平港環境之棲地評價程序(Habitat Evaluation Procedure,HEP),作為安平港生態環境之評估。 本文發現安平港附近海域所調查之底棲生物共有70種的生物,共4339個生物個體。水質均受到季節與地理因素影響。近岸底質粒徑較遠岸粒徑大。生態特性之多樣性與相似度均受到工程施工及結構物影響。由群集分析可將底質分成S1~S5五種形態,S2底質群集分佈最廣,底棲生物分成Ⅰ~Ⅳ四種形態,Ⅰ生物群集分佈最廣。最後經由HEP評估的結果與物種相關性高的環境因子為水溫、葉綠素a、中值粒徑,而各測站之HSI與最小物種數之相關係數為0.80,本文模式預測生物物種與現地調查生物物種數之迴歸相關係數為0.82。
The paper investigates the marine communities around breakwaters on benthonic habitat at An-Pin harbor using four-season field sampling of seawater and marine benthos. Species richness of benthos and their spatial similarity are evaluated by statistically analyzing 18-stations sampling data. Using cluster analysis on spatial and temporal distributions of sand sizes and benthos establish the relationship between sand sizes and benthos. Habitat Evaluation Procedure (HEP) is employed to develop a suitable ecological habitat model for An-Ping waters. This study finds 4339 individuals including 70 species at the An-Pin beach. The properties of seawater are slightly different from four seasons and all positions. Sand size at the nearshore is larger than that at offshore. Species richness of organisms and spatial similarity are significantly influenced by structure construction and existed breakwaters of two harbors. Five groups (S1~S5) are classified with the cluster analysis on sand size and of which group S2 is distributed widely. Sampling benthos are separated into four groups (Ⅰ~Ⅳ) of which the groupⅠ spreads widely. High relationship between the species of organisms and temperature, chlorophyll as well as D50 is demonstrated in the well developed HEP model. The number of species predicted with the proposed HEP model has a high correlation coefficient of 0.82 related to the observed species.
URI: http://140.113.39.130/cdrfb3/record/nctu/#GT009216537
http://hdl.handle.net/11536/72447
Appears in Collections:Thesis


Files in This Item:

  1. 653701.pdf

If it is a zip file, please download the file and unzip it, then open index.html in a browser to view the full text content.