運用乾季上游野溪之棲地環境變數評估魚類生態承載指標之研究

Abstract

本研究之目的為探討乾季時影響野溪魚類之棲地環境變數，藉以評估乾季溪流之「生態承載指標(ecological carrying capacity index, ECCI)」。為了得到台灣普遍與實際的現場理論，本研究以人為破壞較少、生態復育潛力較豐富之臺灣七條野溪為研究對象。主要研究期間於2006年1月至2008年4月間乾季時，選取214個採樣點，進行現地、採樣調查，蒐集棲地環境變數與魚類生態之數據資料，藉以建立「生態承載指標」評估公式。 現地調查結果顯示，枯水期之野溪水流量雖然較少，但大部分地點都能發現指標魚種，主要的指標魚種為：吳郭魚(或福壽魚, Nile mouthbreeder, Tilapia)、臺灣石賓(Acrossocheilus paradoxus)、短吻小鰾鮈(或短吻鐮柄魚, Microphysogobio brevirostris)、臺灣白甲魚(或臺灣鏟頜魚, Onychostoma barbatulum)、大吻鰕虎(Rhinogobius gigas)及褐吻蝦虎(Rhinogobius brunneus)等六種。依據魚類時空分布統計分析，顯示5項魚體變數(魚體最小體長、最大體長、體長差異、平均體長與魚體總數量)皆會隨著不同野溪（空間尺度）呈現顯著變異；另外，魚體最小體長、平均體長與魚體總數量等3種魚體變數也會隨著不同月分（時間尺度）而呈現顯著變異。 研究中考量10項環境變數：棲地水域深度、水寬度、水長度、酸鹼值、電導度、水溫度、水流速、累積降雨量、棲地高程（海拔）與底質特性等。由Pearson’s相關係數與三因子負荷圖顯示，水流速、累積降雨量、棲地高程（海拔）對於魚類變數有類似的影響；而水域的長度和寬度、深度(經對數轉換後)對於魚類變數亦有相似的正向影響。運用主成分分析、最大旋轉因素分析與「乘法-指數模式」之同步迴歸分析法，可由環境變數選求得「生態承載指標」之評估公式。 以往學者探討魚類生態之評估，大多要經過繁複的採樣過程。為簡化魚類生態之評估方法，本研究針對魚類之野溪棲地，以統計方式推估精簡的關係式，藉以合理、簡易的計算「生態承載指標」，了解野溪魚類的生態狀況。本研究以ECCI值判定「魚類棲地狀況」，並將魚類棲地狀況分為5級：0~2、2~4、4~6、6~8及8以上，依序表示棲地狀況為：不良(E級)、較差(D級)、普通(C級)、較佳(B級)與良好(A級)。 本研究結果可提供台灣野溪枯水期之魚類生態保育與復育計畫之參考，進而維護整個生態系平衡，才能讓生態循環綿延不絕，傳遞永續生態的理念。

The purpose of this study is to analyze the habitat environmental variables affecting dry-season creek fish, and evaluate ecological carrying capacity index (ECCI) in dry-season creeks. The research targeted on seven creeks in Taiwan which have less contamination and more ecological potential resoration. The samples were conducted in seven creeks in Taiwan. By collecting the environmental variables and the fish-distribution variables, this main research determined ECCI formula by 214 study sites during dry seasons from January 2006 to April 2008. After investigating in the field, the target fish species could be found at almost all sites in dry seasons. Six target fish species were Nile mouthbreeder (Tilapia), Acrossocheilus paradoxus, Microphysogobio brevirostris, Onychostoma barbatulum, Rhinogobius brunneus, and Rhinogobius gigas. Analyzing fish distributions by statistics, the research concluded that five fish body variables (minimum body length, maximum body length, difference length (difference between minimum body length and maximum body length), average length, and total fish number) would change notably by different creeks (spatial scale), and three fish body variables (minimum body length, average length, and total fish number) had notable differences in different dry-season months (temporal scale). This research considered 10 environmental variables: water depth of habitat, water width, water length, pH, electrical conductivity, water temperature, current water velocity, rainfall, elevation, and substrate diversity. Pearson’s correlation coefficient matrix and three-factor loading plots showed that current flow velocity, rainfall, and elevation affected on fish variables similarly, and water length (logarithmic) and water width (logarithmic) had the positive and similar influences in fish variables. Using the principal component analysis (PCA), varimax rotated factor analysis (FA), and the regression in product-exponential model (PEM), ECCI was formulated by these 10 environmental variables. In brief, ECCI was reasonable and calculated simply by the environmental variables of fish habitats. Habitat evaluation levels were classified 5 groups, namely, very poor habitat, poor habitat, fair habitat, good habitat, and then very good habitat. The results of this study can be applied extensively to evaluate the conditions of fish habitats in dry-season creeks for habitat recovery projects. This study is useful for fish conservation and restoration in dry-season creeks in Taiwan