Development of an Ecohydraulic Model for Riparian Mangrove Expansion

大多數計畫由於忽略紅樹林生態系統與水文環境間的關係，不是以完全失敗收場，就是未能達到當初設定的目標。因此，本研究建構可以量化分析紅樹林生育地水文環境需求的「河岸紅樹林擴張之生態水利模式」，以淡水河系紅樹林水筆仔作為模式的研究案例，除驗證模式之可行性外，也根據分析成果提出紅樹林經營管理的建議。在空間分布上，模式透過鹽度、二次流強度指標（河寬/曲率半徑）與浸淹機率三個門檻，逐步篩選出紅樹林可以生長且具強勢擴張能力之生育地範圍。在時間變遷上，模式則藉由河川流態的分析成果，探究紅樹林的擴張速率是否受到過往水文事件的影響，以及區域水資源管理策略的干擾。 河系尺度下，年平均鹽度大於5 ppt（parts per thousand）的河段，水筆仔具有高度競爭優勢，多以純林形式存在；年平均鹽度介於0.1~5 ppt的河段，水筆仔僅能生長，但生長情形泰半不佳，常與其他植物混生；年平均鹽度小於0.1 ppt的河段，水筆仔在自然條件下無法生長。除此之外，當二次流強度指標（曲率半徑/河寬）小於3，淤積泥灘地得以形成，成為理想的水筆仔生育地。竹圍、關渡與社子島溼地，皆符合上述水文環境需求（年平均鹽度大於5 ppt，且二次流強度指標小於3），目前也是淡水河系水筆仔主要的分布地點。其中，挖子尾溼地係因鄰近河口，調查資料受限，分析前就已經被剔除，故不納入本論文的分析成果。灘地尺度下，挖子尾、竹圍、關渡與社子島溼地，水筆仔生長的浸淹機率上下界分別為：0.5-23.2%、4.9-28.0%、11.2-42.9% 與11.7-42.6%。根據分析成果，鹽度與浸淹機率並非彼此獨立的環境因子，鹽度乘上浸淹機率的倒數（二項式分佈的期望值）為定值，此一常數可視為植物生長最大可承載鹽分。 1978年以前，由於淡水河系開放河道採砂，關渡溼地不但變動劇烈，而且灘地高程常低於水筆仔可生長範圍（浸淹時間過長），因此水筆仔無法生長於此。停止採砂後，灘地高程則穩定地落於水筆仔可生長範圍內，加上水資源工程導致上游流量減少與小型洪水事件消失，鹽分逐漸累積於關渡溼地，成為理想的水筆仔生育地。根據流態分析的成果，發生洪峰流量小於600 cms的洪水事件，可能導致蘆葦等草本植物因洪水沖刷或長時浸淹而死亡，水筆仔得以入侵這些新生成的裸露泥灘地，加速其擴張速率。然而，若洪峰流量超過800 cms，將造成水筆仔族群的嚴重損害，隔年的水筆仔擴張速率將會趨緩。其中，1978年10月的婀拉颱風造成關渡溼地原有植物死亡，這些新生成的裸露泥灘地可能是水筆仔入侵的開端。 紅樹林的經營管理策略應根據不同的生育地屬性而擬定，包括：紅樹林生育地、潛勢紅樹林生育地與非紅樹林生育地。紅樹林於其生育地中，具有高度擴張性與排他性，不建議過度保護，應引進更為主動的管理策略。淡水河系的關渡與社子島溼地，因水筆仔向上游或河道中心的擴張潛能較高，對於臺北地區河防安全的危害程度將高於下游的挖子尾與竹圍溼地，建議持續監測，並以定期疏伐適度控制水筆仔生長的密度與範圍。另外，竹圍溼地北側與關渡溼地東側，由於立地高程已接近水筆仔的生長上界，未來可能被陸域植物取代，除持續監測植群變遷情形外，必要時需要藉由環境營造降低灘地高程，例如：引入潮溝與潮池。This study presents a new ecohydraulic model of riparian mangroves expension to quantify the relationship between mangrove habitats and its hydrological requirements. Because ecologists and engineers ignore the realities of mangrove hydrology, most mangrove often fail completely or fail to achieve the stated goals. A case study showed process model analysis and proposed a suggestion for mangrove management of the Tanshui River system. On spatial analysis, the model sieves mangrove habitats out from a tidal reach based on three indices: water salinity concentration, river secondary flow intensity-R/W (the ratio of radius of curvature to river width), and submerged probability. On temporal analysis, the model used flow regime analysis results to identify impacts of hydrolocial floods and water resource management on mangrove expansion. The results showed that the optimum conditions for K. obovata growth and dispersal exist in waters with mean annualsalinity levels that are higher than 5 ppt (parts per thousand) and the R/W lower than 3 on reach scale. K. obovata growth and dispersal is controlled by submerged probability on wetland scale: 0.5-23.2% for Watzuwei wetland, 4.9-28.0% for Zhuwei wetland, 11.2-42.9% for Guandu wetland, and 11.7-42.6% for Sherzi wetland, respectively, on wetland scale. There were two reasons why mangroves cannot invade Guandu wetland before 1978. First, instream sand mining caused instable and lower land elevation. Second, typhoon Ora on October 1978 created bare mud flats and thus K. obovata was able to grow on Guandu wetland. Discharge decreasing and small flood vanish due to past infrastructures of water resource advantaged mangrove growth and expansion. And flood (peak flow less than 600 cms) caused reed and other original vegetations died because of erision or prolonged inundation, and thus K. obovatawas able to occupy these bare mud flats. These two reasons were why K. obovata expanded so fast on Guandu wetland. However, if peak flow large than 800 cms, floods will damage mangroves and slow down mangrove expansion next few years. The model is able to distinguishmangrove areas, potential mangrove areas and non-mangrove areas. Different strategies for mangrove management shall be adopted with different categories. Mangroves are very tough species in the mangrove areas. Over-protection is not necessary. The potential mangrove areas can be rehabilitated by removing other environmental factors. Mangrove rehabilitation is not suggested within the non-mangrove areas. For mangrove management in the Tabshui River system, because mangroves on Guandu wetland and Sherzi wetland have higher expansion potential to upstream and midstream and may be invaded by terrestrial flora, they would cause more serious problem in flood-prevention.The mangrove density and area must be monitored and controlled. Plus, mangroves on northern Zhuwei wetland and eastern Guandu wetland may be replaced withterrestrial flora, becauseland elevation was close to the upper limitation for mangrove growth. The vegetation change needs to be monitored