雨水花園植栽篩選指標之建立與水質改善之研究

雨水花園(rain garden)為具有滯洪、淨化水質的景觀綠地形式，對都市除具維持綠化之功能，亦兼顧生態、環保效益，是低衝擊開發(low impact development, LID)的發展重點。植栽與介質是影響雨水花園淨化效益的重要元素。因此，本研究將針對臺北地區六處雨水花園(臺大桃花心木道、新生南路下窪式綠地、羅斯福路雨水花園、油杉社區LID公園、華山公園、玉東公園)進行土壤條件、植栽應用現況調查，並針對常見季節草花、可食作物做淹水試驗，以找出雨水花園滿水期適用植栽之篩選指標，最後針對常見栽培介質及耐淹植物做模擬中水氮磷淨化之探討。臺北地區雨水花園應用現況方面，以臺大桃花心木道、新生南路下窪式綠地、華山公園較符合國際上對雨水花園的定義，羅斯福路雨水花園在地上架設雨撲滿，油杉社區LID公園雨水積磚基地則將雨撲滿埋於地下，其雨水花園基地並未呈淺凹狀，而玉東公園則是窪地過深，故並不完全符合雨水花園之定義。土壤條件方面，多數調查地點的土壤硬度介於植物適合生長的範圍(10-20 mm)；土壤含水量多偏低，除新生南路下窪式綠地以及玉東公園的較低窪區，其他處於中度乾旱；土壤EC值亦均偏低、為養分缺乏的狀態；多數地區土壤pH值高，使部分植物有缺鐵黃化症狀，如杜鵑花(Rhododendron spp.)、白鶴芋(Spathiphyllum kochii Engl. et Krause)等。植物的表現可能因調查地點的光度、土壤條件等不同而有所差異，如土壤硬度過高不利巴西鳶尾(Neomarica gracilis Sprague)生長、遮陰下波士頓腎蕨(Nephrolepis exaltata (L.) Schott)生長較佳。為使雨水花園效益發揮最大，應落實淺窪地的設計理念，並且植栽選擇上除耐淹和耐旱性，尚需考量其他環境條件，才可維持良好景觀效果。針對滿水期時耐淹植栽的篩選方面，夏季草花部分以天使花(Angelonia salicariifolia Humb. & Bonpl.)表現最佳不受淹水影響，小百日草(Zinnia angustifolia Kunth)淹水期間光合作用下降，但排水後即恢復，表現次於天使花，雞冠花(Celosia argentea L.)及黃帝菊(Melampodium paludosum Kunth)則外表反應或生理反應均顯著受淹水影響且排水後無法恢復；冬季草花部分以新幾內亞鳳仙花(Impatiens hawker W. Bull)、四季秋海棠(Begonia semperflorens Link. & Otto)、一串紅(Salvia splendens Ker-Gawl.)、矮牽牛(Petunia x hybrida Hort. ex Vilm)四種參試冬季草花外表反應及生理反應皆受淹水影響，但四季秋海棠仍具有觀賞性；可食作物的部分，淹水對蕹菜(Ipomoea aquatic Forsk.)影響最小，萵苣(Lactuca sativa L.)次之，紅莧菜(Amaranthus tricolor L.)及小白菜(Brassica rapa subsp. chinensis)產量下降較多，然紅莧菜雖株高較矮但外觀品質佳。篩選的指標中以相對植生指數、葉片丙二醛濃度與植株外觀品質及可食作物的產量具有相關性，可做為日後雨水花園耐淹水植栽篩選的參考指標。中水淨化分兩個部分。介質淨化方面，黏質壤土與蛭石有最佳的總氮清除率，泥炭土和真珠石次之，發泡煉石最差；總磷清除率最佳者同總氮，發泡煉石次之，泥炭土和真珠石表現較差；另外，泥炭土+黏質壤土以1:1(體積比)可顯著降低模擬中水的總磷和總有機碳濃度。植栽淨化方面，栽種小百日草具促進總氮清除之效果；三種受試可食作物中蕹菜為總氮及總磷清除率表現最佳者，鹿角萵苣居次、皇宮菜表現最差，蕹菜亦對中水的適應力較好。所有參試植栽對於總磷的清除力皆較總氮差，可得知總氮能靠介質與植栽淨化，總磷則以介質吸附為主。介質之選用需同時考量淨化力及是否適合植栽生長，建議可針對雨水花園欲去除之污染物做不同介質的配置做進一步研究，植栽應用除景觀植物外也可選擇具淨化力的可食作物。雨水花園的植栽生長會受工法、環境氣候等影響，選擇適當植栽並搭配合適介質將使雨水花園發揮最大效益。建議未來能以雨水花園、可食地景、中水三者結合為目標作深入研究，增加雨水花園之附加價值並促進水資源永續性，期能提升雨水花園在都市中之功能與應用潛能。

關鍵字

雨水花園；中水；淹水；可食作物

並列摘要

The rain garden is a form of landscape with characteristics of rainwater retention and water pollutant removal. Apart from maintaining the greening of the city, it also takes ecological and environmental benefits into account. It is a critical part of low impact development (LID). Plants and media are important components that affect the removal efficiency of rain gardens. This study aimed to analyze the soil conditions and plant application in six rain gardens in Taipei (NTU Mahogany Road, Xinsheng South Road Bioretention Swale, Roosevelt Road Rain Garden, Taiwan Keteleeria Community LID Park, Huashan Park, and Yudong Park). In addition, responses to waterlogging of several common herbaceous flowers and vegetables were measured to find out the screening indicators for suitable plants for rain gardens in the full-water period. Last, the nitrogen and phosphorus removal efficiency of common cultivation media and waterlogging-tolerant plants to simulated reclaimed water were discussed. In terms of the application of rain gardens in Taipei, the NTU Mahogany Road, Xinsheng South Road Bioretention Swale, and Huashan Park are more in line with the international definition of rain garden. Roosevelt Road Rain Garden is equipped with rainwater tank on the ground instead of shallow depression, so do the rainwater brick-laying base of Taiwan Keteleeria Community LID Park, and there is no shallow depression in its rain garden area. As for Yudong Park, the depressed area is too deep. In respect of soil conditions, the soil hardness of most survey sites were appropriate for plant growth (10-20 mm). The soil moisture content in most survey area was low and were in moderate drought conditions, except for the Xinsheng South Road Bioretention Swale and the lower area of Yudong Park. The soil EC were quite low of most of them which indicating the lack of nutrients. The soil pH value in most survey areas was higher than the acceptable range for plant growth in (pH 4.2-6.5), especially in Huashan Park and Yudong Park. The higher soil pH value caused iron deficiency in some plants, such as azalea (Rhododendron spp.) and peace lily (Spathiphyllum kochii Engl. et Krause). Plant performance may vary depending on the luminosity and soil conditions of the survey site. For examples, harder soil is harmful to walking iris (Neomarica gracilis Sprague), boston swordfern (Nephrolepis exaltata (L.) Schott.) grows better under shading. In order to maximize the benefits of rain garden, the design of shallow depression should be implemented. Morever, in order to maintain good landscape performance, besides waterlogging-tolerance and drought-tolerance, environmental condition should be considered when selecting plants. In terms of screening for waterlogging-tolerant plants, angelonia (Angelonia salicariifolia Humb. & Bonpl) was not affected by waterlogging and showed the best performance of four summer herbaceous flowers. The photosynthesis of nArrows-leaf zinnia (Zinnia angustifolia Kunth) decreased during waterlogging, but soon recovered after back to normal watering. Celosia (Celosia argentea L.) and medallion flower (Melampodium paludosum Kunth) were significantly affected by waterlogging and could not recover, no matter in response to external reactions or physiological reactions. The the four tested winter flowers, new guinea impatiens (Impatiens hawker W. Bull), wax begonia (Begonia semperflorens Link. & Otto), salvia (Salvia splendens Ker-Gawl.), and petunia (Petunia x hybrida Hort. ex Vilm), were all affected by the waterlogging, but the wax begonia still maintained its ornamental value. As for edible crops, waterlogging had the least impact on water spinach (Ipomoea aquatic Forsk.), lettuce (Lactuca sativa L.) ranked second. The yield of ganges amaranth (Amaranthus tricolor L.) and bok-choy (Brassica rapa subsp. chinensis) decreased. However, although the height of ganges amaranth was short, it maintained good appearance quality. Among the screening indicators used for this test, relative NDVI and leaf malondialdehyde (MDA) concentration were closely related to plant quality and yield. Therefore, relative NDVI and leaf MDA concentration can be used as indicators for selection of waterlogging-tolerant plants in rain gardens. Reclaimed water purification was divided into two parts. From the media side, the clay loam and vermiculite had the best total nitrogen removal capacity, followed by peat moss and perlite, and the ceramsite was the worst. The clay loam and vermiculite also had the best phosphorus removal capacity, ceramsite was the second, peat moss and perlite performed poorly. Peat moss with clay loam by 1:1 (volume ratio) could also significantly reduce the total phosphorus and total organic carbon concentrations. From the plant side, the cultivation of narrows-leaf zinnia promoted the effect of total nitrogen removal. Among the three tested edible crops, water spinach showed the best performance of total nitrogen and total phosphorus removal, followed by lettuce, and ceylon spinach was the worst. Water spinach also had the better adaptability to simulated reclaimed water. All the tested plants had worse scavenging capacity of total phosphorus than total nitrogen. The results indicated that total nitrogen removal relied on media and plantation, as total phosphorus was mainly based on media adsorption. The choice of medium should consider both the removal capacity and suitability for plant growth. It is recommended that media configuration for different pollutants to be removed in rain garden need further study. Besides landscape plants, edible crops with pollutant removal capacity also can be apply in rain garden. The vegetation of the rain garden will be affected by construction methods, environmental conditions. Choosing appropriate vegetation and media make the rain garden play its best. It is suggested that in the future, a combination of rain garden, edible landscape, and reclaimed water can be used as an in-depth study to increase the added value of rain gardens and promote the sustainability of water resources, so as to improve the function and application potential of rain gardens in cities.