Abstract This study was conducted to investigate the impact of different seasons (dry and rainy seasons) in 2018 on the fate, transport, and ecological risk of antibiotics drained from pig farms along the Bang Pakong River, Thailand. Twenty-one residual antibiotics were detected in pig farm effluents, drainages, and the river in both seasons. Amoxicillin, tetracycline, chlortetracycline, and tiamulin were found dominantly in the pig farm effluents, showing high detection frequencies of 83–100 % in the total number of the effluent samples. Maximum concentrations of these four dominant antibiotics in the drainage and the river were 0.12–5.20 μg L−1 in the dry season and 0.12–16.1 μg L−1 in the rainy season. Tetracycline (15.1 μg L−1) and chlortetracycline (16.1 μg L−1) were found at the highest concentrations in the rainy season. The multiple linear regression and random forest analyses between residual antibiotics in the farm effluents and the river in both seasons were performed to model dilution and removal processes of the antibiotics. The antibiotic removal processes were determined by their partition coefficients, solubility, biodegradability, and volatility. Compared to multiple linear regression, random forest models better identified the removal processes of pig farm antibiotics drained into the river. The risk quotient (RQ) of tiamulin determined based on its actual concentration in the river (0.0016 and 0.0017 in dry and rainy seasons) and its modeled concentration in river as predicted by random forest model (0.0029 and 0.001 in dry and rainy seasons) presented its potential ecological risk to fish, daphnia and green algae.

中文翻译：

---

泰国邦巴公河沿岸养猪场抗生素的命运、运输和生态风险

摘要 本研究旨在调查 2018 年不同季节（旱季和雨季）对泰国邦巴公河沿岸养猪场排放的抗生素的命运、运输和生态风险的影响。两个季节的猪场污水、排水系统和河流中均检测到 21 种抗生素残留。阿莫西林、四环素、金霉素和泰妙菌素主要存在于猪场污水中，在污水样本总数中的检测频率高达 83-100%。这四种主要抗生素在排水和河流中的最大浓度在旱季为 0.12-5.20 μg L-1，在雨季为 0.12-16.1 μg L-1。四环素（15.1 μg L-1）和金四环素（16.1 μg L-1）在雨季浓度最高。对两个季节农场污水和河流中的残留抗生素进行多元线性回归和随机森林分析，以模拟抗生素的稀释和去除过程。抗生素去除过程由它们的分配系数、溶解度、生物降解性和挥发性决定。与多元线性回归相比，随机森林模型更好地识别了排入河流的猪场抗生素的去除过程。泰妙菌素在河流中的实际浓度（旱季和雨季分别为 0.0016 和 0.0017）和随机森林模型预测的河流中模拟浓度（旱季和雨季分别为 0.0029 和 0.001）确定的风险商数（RQ）其对鱼类、水蚤和绿藻的潜在生态风险。