探討土壤性質、鉀肥添加與淹水週期對水田土壤鉈的釋放與青鱂魚毒性影響

Tl為美國環保署優先列管的污染物之一，其生物累積性及毒性極高。近年來隨著Tl於高科技產業的廣泛應用，Tl經由人類活動進入生態環境的機會大為增加，研究發現Tl可能隨灌溉進入水田環境，並累積在水稻榖粒中，因此提高人體暴露Tl的風險，但關於Tl在水田中的移動、生物有效性及生態毒性尚未釐清。因此本研究利用青鱂魚(俗稱稻田魚，Oryzias latipes)幼魚作為模式生物，建立間歇流水式水田生物暴露系統，用以評估水田遭受鉈污染後在添加鉀肥及排水處理條件下之Tl移動性、生物有效性及毒性。此研究將一價鉈[Thallium, Tl(I)]添加至酸性(平鎮土系Pc和三坑子土系Sk)及中性 (伸港土系Su)土壤中，經三次乾溼交替後，進行120天（共三階段）高鉈土壤濃度(75-250 mg/kg)生物暴露試驗，以評估鉀(K)肥添加及排水/淹水交替後三種土壤中鉈的釋出及對稻田魚的毒性效應。第一次暴露結果顯示酸性土壤（Sk與Pc）相較於中性土壤(Su) 具較高的Tl釋放潛勢，且覆蓋水中Ｋ濃度較低，因此造成較高魚苗死亡率。加入鉀肥處理後之第二階段暴露結果，和第一階段的暴露結果相比，添加K肥對Tl誘導的生物毒性影響並不顯著。第三次暴露結果顯示，三種土壤經過排水/淹水之乾濕交替過程後，Tl的釋放量顯著降低，因此為三次生物暴露試驗中造成之毒性最低。本研究另外以Pc及Su土系進行環境相關濃度鉈污染土(2.5-15 mg/kg)之生物暴露試驗，此低濃度範圍過程皆未造成顯著魚苗死亡，在泳動行為、魚體體長體重及鉈生物累積量指標中，以魚苗生物累積性結果最為顯著，暴露於酸性土Pc之魚體累積量較高，而加鉀處理後發現兩種土暴露魚體鉈累積量皆有下降。本研究顯示土壤中酸可交換型態、覆蓋水及孔隙水(R2 > 0.8)中Tl濃度皆具有較佳預測魚體死亡率的能力。在低鉈污染土中，覆蓋水及孔隙水中(R2 > 0.81) Tl/K比例與魚體中Tl累積性有高相關性。綜上所述，本研究所建立之青鱂魚幼魚暴露系統可有效地模擬Tl於水田土壤的移動，並可用以預測受Tl污染水田生態的生物有效性與毒性。

關鍵字

鉈；水田生態暴露系統；鉀肥；排水；覆蓋水；生物毒性

並列摘要

Thallium (Tl) is a priority pollutant with high toxicity and exposure risk. With rapid development of high-tech industries, technology-critical elements such as Tl could be released into the paddy soils via discharge of manufacturing wastewater. The Tl occurrence in contaminated paddy soils was frequently detected in many studies; however, the transport, bioavailability and environmental toxicity of Tl in contaminated paddy field are less known. In this study, we aim to use medaka larvae as a model to assess the mobility, bioavailability, and toxicity of Tl in different Tl-contaminated paddy soils after treatment of K addition and drainage. Monovalent thallium was added to acidic (Pingzhen and Sankengtzu soil series) and neutral (Sangkang soil series) soils. After three time dry and wet alternations, conducting a 120-day (three stages) biological exposure test of high Tl concentration (75 - 250 mg/kg) in soils to assess the toxic effects on medaka according to soil properties, addition of potassium fertilizer, and drainage. The results of first fish exposure showed that acidic soils (Sk and Pc) have higher Tl release potential than neutral soils (Su). When the concentration of K in the overlying water was lower, it leaded to higher mortality of medaka. Compared with the first fish exposure results, the effect of adding K fertilizer on Tl-induced toxicity was not significant. The third fish exposure result showed that after short term drainage and re-irrigation, the release of Tl was significantly reduced, so it was the least toxic in the three fish exposure experiments. Otherwise, Pc and Su soils were used to conduct biological exposure experiments of environmentally relevant concentrations of Tl-contaminated soil (2.5、15mg/kg). This low Tl concentration range did not cause death of rice fish. Among the indicators of swimming behavior, body weight, body length, and thallium bioaccumulation of medaka, the bioaccumulation results of medaka were the most significant. The cumulative amount of Tl in medaka is higher when exposed to acid soil. The accumulation of Tl after K addition treatment is reduced. This study showed content of Tl in exchangeable fraction of soils, Tl concentration in overlying water and pore water (R2 > 0.8) have good ability to predict rice fish mortality. In low Tl-polluted soils, the ratio of Tl/K in overlying water and pore water were highly correlated (R2 > 0.81) with the accumulation of Tl in medaka. Based on our results, the established ecotoxicity exposure system of paddy fields with medaka fish can effectively simulate the release, bioavailability, and toxicity of Tl in different soils.