估計臺灣櫻花鉤吻鮭族群數量受防砂壩之影響

Estimating the Impacts of Sabo Dams on the Population of Formosan Landlocked Salmon

邱立鈞(Chiou, Li Jiun)

指導教授 ： 曾晴賢

國立清華大學/生命科學院/生物資訊與結構生物研究所/碩士(2015年)

https://doi.org/10.6843/NTHU.2015.00602

摘要

臺灣櫻花鉤吻鮭（Oncorhynchus masou formosanus）是相當珍貴的保育物種，但是在其棲息的七家灣溪流域中，卻豎立了數座防砂壩，對於環境與生態系統皆造成衝擊。本研究欲從過去20年的族群監測資料中，建立族群數量估計模型，並藉七家灣溪二號壩為案例，探討其沖毀前後的族群與環境之變動情形。此外，也希望建立族群估計模型的方法能有效率地簡化未來的調查工作。
本研究為了能針對防砂壩做單一變因的討論，先對數據做年效應的校正，以排除其他環境變因的影響。然而，在校正後的數據結果較難以達到顯著差異，這是由於許多負面的環境因子會被防砂壩放大其影響，如高水溫、颱風等，而這些影響在校正的過程中一併被剃除掉，因此校正後的顯著性與模型配適能力都較校正前來的低。
從統計檢定的結果中顯示，二號壩的沖毀對壩上游的族群有顯著的增加，原因為壩下游族群傾向上溯至水溫較低的上游，且防砂壩沖毀後的沉積物會持續影響下游河段的棲地狀況。而在三個年齡級比較上，無論是統計檢定或是迴歸分析，皆顯示小魚的數量有顯著的增加，說明了二號壩沖毀可增加成魚上溯繁殖的機會。此外，在有無防砂壩的族群數量估計模型上，亦顯示小魚在無防砂壩的時候數量有明顯增加，反而大魚與中魚甚至有些微的下降。原因為增加的幼魚造成的競爭關係並不明顯，且族群結構偏向快速成長的金字塔型結構，說明了無防砂壩的溪流對整個族群的更新能力較有幫助。
比較國外的文獻，推測二號壩沖毀後可能需要較長的恢復時間，原因在於經常性的颱風侵襲加上易崩塌的地質環境讓七家灣溪的棲地變動較大，而穩定且多樣的棲地關係著生態系統的恢復情形，因此從二號壩沖毀至今已過了十年，卻仍難斷言已得到對族群增加有顯著改善的統計結果。但即使需要較長的恢復時間，對於剩餘的防砂壩仍應予以拆除，還給臺灣櫻花鉤吻鮭原始的自然棲地。
在族群估計模型的應用研究上，發現可以透過選定的模式河段來估計全河段的族群數量。以單一河段而言，一號壩至二號壩、二號壩至三號壩、三號壩至四號壩以及五號壩至六號壩的估計能力均佳，但是對於數量變動較大的小魚族群估計誤差率較高。若選擇兩個河段做估計，以一號壩至二號壩加上二號壩至三號壩的估計能力最佳；但是考慮到調查距離，以二號壩至三號壩加上五號壩至六號壩為較有效率的模式河段。

關鍵字

臺灣櫻花鉤吻鮭 ； 七家灣溪 ； 防砂壩 ； 壩體改善 ； 族群估計

並列摘要

Formosan landlocked salmon (Oncorhynchus masou formosanus) is a valuable protected species, but its last habitat, Chichiawan Stream basin, was built up numerous sabo dams, which has impacted both environment and ecosystem. By using the population monitoring data in past 20 years, this study set up a population-estimating model, in order to compare the change of the population and environment before and after the destruction of the dam #2 in Chichiawan Stream. In addition, the model also provides an efficient way to simplify the investigating work in the future.
For the purpose of focusing on the single variable, dam, the data have to calibrate first, to eliminate the influence of other environmental impacts. However, there was less significant difference for the statistic results of the calibrated data. Since the sabo dams might amplify some negative impacts, such as high water temperature or typhoon, the calibration will reduce the significance of the results and the fitness of the models.
The results of statistic tests show that the population in the upstream increased significantly after the destruction of the dam #2, because the population in the downstream were tend to trace to the cooler upstream, and the sediments from the dam destruction will impact the physical habitats in the downstream persistently. For the comparison of three age-classes, the small fishes were increased significantly, which proved that dam #2 blocked the pathway for reproduction. In addition, the results of the population-estimating model about the dam also showed that the small fishes increased when the dam wasn’t existed. In contrast, the number of the other two age-classes decreased slightly due to the competition of the increasing small fishes. Nevertheless, the population structure became pyramid-shaped, which means that the undamming streams benefit the renewal of the population.
According to the reference, this study inferred that the destruction of the dam #2 might need more time to recover, because the recovery of ecosystem depends on stability and diversity of the habitats. Furthermore, the incessant typhoons and the easily collapse geology lead to the dramatic change of the habitats in Chichiawan Stream, so it’s difficult to find a significant difference from the statistic results, even though ten years has passed. In spite of the long recovery, we should remove the rest of the dams to return natural habitats to the salmons.
The population-estimating model can estimate the total population from the “model section”. For the single model section, dam #1 to dam #2, dam #2 to dam #3, dam #3 to dam #4 and dam #5 to dam #6 were good at estimating in all nine sections, but the error was a little bit high for small fishes. As a result, for two model sections, dam#1 to dam #2 plus dam #2 to dam #3 were the best in estimating. However, considering the investigating distance, dam #2 to dam #3 plus dam #5 to dam #6 were the most efficient model.

並列關鍵字

Oncorhynchus masou formosanus ； Chichiawan Stream ； dam ； dam removal ； population estimation

參考文獻

張石角，1989。櫻花鉤吻鮭保護區規劃。行政院農業委員會78年生態研究第010號。

曾憲郎、李政翔，2008。環境因子對櫻花鉤吻鮭族群動態的影響，農業經濟叢刊。14卷，1期，175-206。

鄧火土，1959。臺灣高地產陸封鮭魚的形態與生態。行政院農業委員會水產試驗所。

American Rivers, Friends of the Earth and Trout Unlimited, 1999. Dam Removal Success Stories: Restoring Rivers through Selective Removal of Dams that Don’t Make Sense. American Rivers, Friends of the Earth and Trout Unlimited.

賴建盛，1996。防砂壩對櫻花鉤吻鮭物理棲地影響之研究。碩士論文，國立臺灣大學地理學研究所。