結合源頭減量及濕地治理控制農業非點源污染

Abstract

近年來，隨著農業集約化發展，提升作物產量，這也伴隨著肥料使用量的增加，進而引發了農業非點源污染的問題，農業非點源污染物主要為肥料中的氮與磷，透過降雨及灌溉所產生之逕流水及入滲水移動至地表水及地下水，造成土壤退化、優養化、水質下降等問題，然而，此類污染形成過程受地形、耕作方式、土壤質地、灌溉方式等農業條件影響，複雜的形成機制導致汙染難以管理及控制。針對農業非點源污染，常見的控制方式根據污染發生前後分為從源頭控制以及末端處理。 本研究選定小白菜作為目標作物，在源頭控制面採用不同肥料、調整施肥量以及生物炭等策略；而在末端處理方面採用人工濕地。第一階段為盆栽試驗，使用三種不同肥料 (化學肥料、緩效性肥料及有機肥料)，不同肥料劑量及有無施用生物炭進行比較，並進行暴雨模擬，探討肥料有效性以及營養鹽流失情況。第二階段是中尺度砂箱試驗，根據盆栽試驗結果篩選出表現較佳的幾種組合進行種植面積放大，進行暴雨模擬，暴雨所產生之逕流水進行人工濕地試驗，藉此探討尺度放大後的差異及濕地對於逕流水營養鹽降解效果。第三階段為現地試驗，其試驗條件與砂箱試驗相同，主要目的是探討模擬結果與實際情境之間的差異。 試驗結果顯示在盆栽試驗中，加入生物炭對三種肥料均能減少氮流失。在施肥量100%、75%以及50％情況下，化肥組降低氮流失21-31%、23-25%及11-58%，緩效肥分別為22-25%、17-21%及10-19%，有機肥分別為33-36%、23-43%及21-61%。砂箱試驗中，生物炭結合半量施肥減少的氮流失量優於全量施肥。化肥、緩效肥以及有機肥分別為11%、7%和8%。現地試驗中，半量施肥結合生物炭與傳統全量化肥相比，化肥、緩效肥以及有機肥減少11-19%、23-39%和24-40%氮流失量。濕地系統對氮具有淨化效能，氮平均去除率在25天達61.6%。然而，由於暴雨模擬產生的逕流水量較少，濕地即使有很好的氮去除效果但去除總量仍然很低，因此有效控制農業非點源污染的關鍵在於選擇緩效肥或是有機肥代替化肥、減少施肥量及結合生物炭，未來極端氣候下暴雨頻率與強度增加，導致逕流水量上升，末端處理 (例如人工濕地) 仍可以有效的淨化逕流水中的營養鹽，成為應對極端氣候的重要設施。

In recent years, intensive agricultural practices have been adopted to enhance crop yields for food production. Therefore, increase the use of fertilizer. Agricultural non-point source pollutants are mainly nitrogen and phosphorus from fertilizers, which migrate to surface water and groundwater through runoff water and infiltration water generated by rainfall and irrigation, causing problems such as soil degradation, eutrophication, and decline in water quality. The non-point source pollution control is diverse and difficult to manage because of several factors. For agricultural non-point source pollution, common control methods are divided into source control and endpoint treatment. In this study, Brassica Chinensis L.(Pak-Choi) was chosen as the model crop. The cultivation experiments were divided into three phases including pot, pilot-scale, and field experiments. In the pot experiment, biochar and three different types of fertilizers (i.e., chemical, slow-release, and organic) at different application rates were applied in this study. Simulated intense rainfalls were operated to evaluate fertilizer efficiency and nutrient loss. Based on the results of the pot experiments, several selected fertilization conditions were applied in the pilot-scale experiment for scaling up. Simulated intense rainfalls were operated and the surface runoff was introduced to the constructed wetland for further nutrient removal. Finally, the field experiment aimed to evaluate the differences between controlled simulations and actual field conditions. In recent years, intensive agricultural practices have been adopted to enhance crop yields for food production. Therefore, increase the use of fertilizer. Agricultural non-point source pollutants are mainly nitrogen and phosphorus from fertilizers, which migrate to surface water and groundwater through runoff water and infiltration water generated by rainfall and irrigation, causing problems such as soil degradation, eutrophication, and decline in water quality. The non-point source pollution control is diverse and difficult to manage because of several factors. For agricultural non-point source pollution, common control methods are divided into source control and endpoint treatment. In this study, Brassica Chinensis L.(Pak-Choi) was chosen as the model crop. For source control, strategies such as the use of different fertilizers, adjustment of fertilizer application rates, and the incorporation of biochar were adopted; for endpoint treatment, constructed wetlands were employed. The cultivation experiments were divided into three phases including pot, pilot-scale, and field experiments. In the pot experiment, biochar and three different types of fertilizers (i.e., chemical, slow-release, and organic) at different application rates were applied in this study. Simulated intense rainfalls were operated to evaluate fertilizer efficiency and nutrient loss. Based on the results of the pot experiments, several selected fertilization conditions were applied in the pilot-scale experiment for scaling up. Simulated intense rainfalls were operated and the surface runoff was introduced to the constructed wetland for further nutrient removal. Finally, the field experiment aimed to evaluate the differences between controlled simulations and actual field conditions. The experimental results indicated that adding biochar reduced nitrogen loss for all fertilizer types in pot experiments. Under 100%, 75%, and 50% fertilization levels, nitrogen loss was reduced by 21-31%, 23-25%, and 11-58% in the chemical fertilizer group; 22-25%, 17-21%, and 10-19% for slow-release fertilizers; 33-36%, 23-43%, and 21-61% for organic fertilizers, respectively. In the pilot-scale experiment, combining biochar with reduced half-rate fertilization showed better nitrogen loss reduction compared to full-rate fertilization, achieving reductions of 11%, 7%, and 8% for chemical, slow-release, and organic fertilizers, respectively. In field trials, compared to the traditional full-rate chemical fertilization method, half-rate fertilization with biochar reduced nitrogen loss by 11-19%, 23-39%, and 24-40% for chemical, slow-release, and organic fertilizers, respectively. Constructed wetland demonstrated nitrogen purification effectiveness, with an average removal rate of 61.6% achieved within 25 days. However, due to lower runoff volumes generated during simulated intense rainfalls, the total nitrogen removal was limited despite the high purification efficiency. In conclusion, effective control of agricultural non-point source pollution hinges on adopting slow-release or organic fertilizers as substitutes for chemical fertilizers, reducing fertilization rates, and applying biochar. As the frequency and intensity of extreme weather events increased, leading to higher runoff flow, endpoint treatment (i.e., constructed wetland) are able to effectively purify nutrients in runoff water, making them a critical facility for mitigating the impacts of extreme climate conditions.