畜牧廢水灌溉對土壤與葉菜中小分子有機污染物分布之影響

Abstract

畜牧廢水中富含植物生長所需之巨量營養素，可部分取代化學合成肥料再利用於農業灌溉，同時減少畜牧廢水排放造成自然水體污染。然而畜牧廢水中含有的多重小分子有機污染物，可能經由灌溉途徑進入農田土壤，甚至污染農作物。為了全面性地揭示農業環境的污染概況，本研究利用高解析質譜法對土壤及青江菜進行廣泛篩查，探討畜牧廢水再利用對於農業土壤與葉菜中小分子有機污染物分布的影響。 本研究於真實農田栽種青江菜，將農田分為3組，分別灌溉地下水（控制組）、適量畜牧廢水（? 400公噸/公頃/年）、過量畜牧廢水（適量的3倍）。田間試驗總共進行三期，於每期青江菜採收時，同時採集表層土壤（每組n = 6）。均質的土壤與青江菜樣本（5 g）經過溶劑萃取與管柱淨化，以資料獨立擷取模式收集樣本中小分子（m/z 70－1,100）的高解析質譜資訊，並與化學資料庫中超過3,000種化合物（如：藥物、個人保健品、動物用藥品等）的圖譜比對。針對檢出率高（任一組 ? 3次）的污染物，利用同位素標記化學品標準化波峰面積後，以偏最小平方判別法建立污染物整體分布模型，並以交叉驗證與置換試驗結果評估模型的預測能力及可信度。另外利用Kruskal Wallis檢定與Dunn事後檢定觀察個別污染物相對濃度變化。最後針對土壤中因施灌畜牧廢水而改變相對濃度（p < 0.05）的指標污染物進行化學鑑定，與已知物比對標準為相對質量誤差 ? 5 ppm及至少有2個碎片離子與已知化合物圖譜相符。 我們在三期土壤（n = 54）與青江菜（n = 53）中分別檢出94與90種污染物。土壤中污染物整體分布隨畜牧廢水施灌量與栽種採收期次而改變（交叉驗證Q2 ? 0.5且Q2/R2 ? 0.5，置換試驗p < 0.05），有17種污染物（包括：抗生素、防腐劑、除草劑等）的相對濃度隨畜牧廢水施灌量不同具顯著差異（p < 0.05），其中11種污染物於灌溉畜牧廢水之土壤中濃度較高（1.52－7.31倍，p < 0.05），事後檢定結果顯示有8種污染物（如：防腐劑）僅在過量灌溉時濃度顯著增加。有4種抗生素（lincomycin、tiamulin、tilmicosin、oxytetracycline）在控制組中均未檢出，證明畜牧廢水為這些抗生素的唯一污染源。反之，有6種污染物於控制組濃度較高（1.70－4.52倍，p < 0.05），包含農業廣泛使用之除草劑（atrazine及simazine），可能為畜牧廢水改變土壤中微生物相，加速部分污染物降解所致。此外經過三期的栽種試驗，本研究未發現土壤中污染物含量因畜牧廢水重複施灌而增加，豬隻飼育隨季節變化可能是造成各期之間濃度差異的主因。另外在本研究的試驗條件下，亦未發現青江菜中任何污染物的含量，隨著施灌畜牧廢水而增加，可見小分子有機污染物雖然經由灌溉進入土壤，但可能較不易被青江菜吸收。 本研究成功應用高解析質譜法在土壤及青江菜中篩查出多類小分子有機污染物，並證實畜牧廢水灌溉可能導致小分子有機污染物進入土壤，且具有劑量－效應關係。由於多數指標污染物的含量僅在過量灌溉時明顯增加，本研究建議合理化施灌畜牧廢水，並持續監測抗生素與防腐劑等指標污染物於畜牧廢水與農業環境中的變化趨勢，以同時保護環境水體與促進農業永續發展。

Livestock wastewater is rich in macronutrients that plants require and can therefore irrigate farmland. Such reuse may reduce the use of synthetic fertilizers and the wastewater discharge into natural water. However, multiple small-molecule contaminants (micropollutants) in livestock wastewater may enter the soil and contaminate crops through irrigation. Aiming to reveal the contamination profile in the agricultural environment, this study employed high-resolution mass spectrometry (HRMS) to perform wide-scope screening of micropollutants in soil and pak choi and investigate the effects of livestock wastewater irrigation on the distribution of micropollutants in agricultural soil and a leafy vegetable. Pak choi was planted on real farmland. The farmland was divided into three groups, irrigated with groundwater only (control group), a rational amount (? 400 tons/hectare/year), and an excessive amount (three times the rational) of livestock wastewater, respectively. Three trials were conducted. Pak choi and topsoil samples were collected at each harvest (n = 6 in each group). A homogenized soil or pak choi sample (5 g) was processed through solvent extraction and cartridge cleanup. The HRMS data (m/z 70－1,100) were acquired in data-independent acquisition mode. The molecular features were matched to over 3,000 compounds included in chemical databases, such as pharmaceuticals, personal care products, and veterinary drugs. For micropollutants with high detection rates (? 3 times in any group), their peak areas were normalized to those of isotope-labeled chemicals. The overall distribution trend of micropollutants was illustrated using partial least squares-discriminant analysis (PLS-DA). The model prediction ability and reliability were evaluated with cross-validation and permutation test, respectively. The difference in the relative concentration of each micropollutant was compared using the Kruskal Wallis test and Dunn post hoc test. The micropollutants that the relative concentrations differed (p < 0.05) by the amount of livestock wastewater irrigated were further identified by matching with known compounds according to the criteria: (1) relative mass error ? 5 ppm, and (2) number of identical fragments ? 2. Ninety-four and 90 micropollutants were found in soil (n = 54) and pak choi (n = 53) samples of three harvests, respectively. The distribution of micropollutants in the soil changed with the amount of livestock wastewater and harvests (cross-validation Q2 ? 0.5 and Q2/R2 ? 0.5, permutation test p < 0.05). The relative concentrations of 17 micropollutants, such as antibiotics, preservatives, and herbicides, significantly changed with the wastewater amount (p < 0.05). Among the 17 micropollutants, 11 exhibited higher relative concentrations in the soil irrigated with livestock wastewater (1.52－7.31 times, p < 0.05). The post hoc test results further demonstrated that for eight of the 11 micropollutants, such as preservatives, a significant increase (p < 0.05) was only found after excessive irrigation. Moreover, four antibiotics (lincomycin, tiamulin, tilmicosin, and oxytetracycline) were absent in the control group, indicating that livestock wastewater was the only contamination source. On the contrary, six micropollutants were more abundant in the control group (1.70－4.52 times, p < 0.05) than in the others, including herbicides (atrazine and simazine) commonly used in agriculture. Livestock wastewater may change the microbiota in soil, promoting the degradation of some micropollutants. Nevertheless, after the three trials, none of the micropollutants increased with the repeat application of livestock wastewater. The differences in micropollutant concentrations among harvests could be primarily attributed to the seasonal changes in swine breeding. Under the experimental conditions of this study, there was no observable change in any micropollutant in pak choi after irrigating livestock wastewater. The result may imply that pak choi would not tend to absorb the micropollutants in the soil translocated from livestock wastewater. This study successfully applied HRMS to screening multiclass micropollutants in soil and pak choi. The screening results and the dose-response relationship proved that livestock wastewater irrigation could result in micropollutant contamination in soil. Given that most marker micropollutants only significantly increased after excessive irrigation, we recommend rational fertilization using livestock wastewater. Also, monitoring marker micropollutants, such as antibiotics and preservatives, in livestock wastewater and the agricultural environment is suggested to protect the natural water and promote sustainable agriculture development.