地下水人工补给工程的广泛应用会对地下水系统产生复杂且长远的影响。识别地下水人工补给工程对地下水水质和微生物群落的影响，并阐明它们的相互作用机制，对指导地下水回补以及保障地下水安全都具有重要的意义。以潮白河再生水自然入渗回补区为例研究了地下水交错带中46种抗生素的分布特征及微生物群落结构与功能的变化规律。发现长期的河水入渗回补导致交错带地下水的主要离子组成向河水演化，营养盐水平与背景点相比提高了23–77%。在河水入渗回补的地下水中也检出了抗生素，但与背景点相比没有明显增加风险。吸附作用是抗生素的主要去除机制，同时低pH和高氧化还原电位有助于抗生素的去除。尽管地表补水向30 m含水层贡献了15–27%的微生物，但地下水群落组装过程仍受扩散限制（约80%）支配，而水质指标和抗生素的共同变化等仅解释了较小（约25%）的群落变异性。另外，河水入渗回补地下水中物种相互作用的强度降低，与碳、氮、锰代谢等相关的功能基因丰度也受到不同程度的影响。通过模拟实验研究了环境浓度抗生素（磺胺甲恶唑，SMX）的长期暴露（180天）对微生物群落结构和功能的影响机制。发现低浓度SMX长期暴露对微生物群落的扰动程度与暴露浓度（0.5–50 μg/L）正相关。群落中的优势菌群对SMX长期胁迫的抵抗力较强，而稀有菌群的受扰动程度随SMX暴露浓度的升高而增大。SMX长期暴露减少了群落关键物种的个数，并导致第90天时确定性过程对群落组装的贡献增加了21%。低浓度SMX长期暴露导致群落sul1和sul2抗性基因相对丰度的增加（1.2–4.3倍），而固氮基因（nifH、nifK）以及与甲苯、乙苯和二噁英降解等相关代谢通路的相对丰度减少。低浓度抗生素暴露不会影响共存有机质的代谢过程和终点，而共存有机质来源和结构的复杂性会延缓抗生素暴露带来的扰动。另外，不同丰度菌群对有机质来源和抗生素暴露复合影响的响应存在差异。通过土柱模拟实验深入探究了生物强化、碳源波动和抗生素暴露等不同进水条件下抗生素的去除效果及群落结构和功能的变化。模拟土柱对四环素类、大环内酯类和喹诺酮类等抗生素的去除率接近100%，而对磺胺类的去除率为50–70%。通过生物强化和碳源波动的方法可以提高0–10 cm深度土壤的群落多样性、降低抗生素暴露对群落的扰动并增加群落对抗生素的共代谢能力，进而强化对磺胺类抗生素的去除。此外，两种方法不会导致群落抗性基因和毒力因子的增加。

The wide use of artificial groundwater recharge has a long-term complex impact on groundwater systems. Understanding the influence of the artificial groundwater recharge practice on groundwater quality, microbial communities, and their interactions can provide insight for decision-making with regard to safe and sustainable groundwater management.Firstly, we selected a reclaimed water infiltration recharge area in the Chaobai River as a case to clarify the distribution of 46 typical antibiotics, and the variations of microbial communities in hyporheic zones. The results showed that the long-term river water infiltration recharge resulted in the evolution of the major ion composition of hyporheic groundwater towards river water. The long-term river water infiltration recharge led to the evolution of the major ion composition of groundwater towards river water, and elevated the nutrient level in groundwater by 23–77% compared with the background groundwater. Although antibiotics were also detected in hyporheic groundwater receiving reclaimed water, but the environmental risk was not further elevated. Adsorption was the main removal mechanism of antibiotics in hyporheic groundwater. Meanwhile, low pH and high redox potential contributed to the removal of antibiotics. Besides, reclaimed water contributed to 15–27% of microorganisms in 30 m groundwater, but did not alter the dominance of stochastic processes to hyporheic groundwater community assembly. Hydrochemistry, antibiotics, and other variables only explained 25% of community variation. Nevertheless, the long-term river water infiltration recharge decreased the species interaction strength, and the community functions related to carbon, nitrogen, and manganese cycles were also affected.We further clarified the chronic effect of environmentally-relevant level antibiotics (sulfamethoxazole, SMX) on microbial community structure and function through batch experiments. The results showed that the disturbance strength of long-term SMX exposure increased with increasing concentration (0.5–50 μg/L). The abundant species showed strong resistance to long-term SMX stress, while rare species were less-resistant to the increased SMX concentrations. Long-term SMX exposure reduced the number of key species and increased the deterministic community assembly process by 21%. Besides, the relative abundance of sulfonamide resistance genes (sul1, sul2) increased by a factor of 1.2–4.3, while that of nitrogen-fixing genes (nifH, nifk) and the metabolic pathways related to the toluene, ethylbenzene, and dioxin degradation decreased. In addition, the complex organic matter sources could delay the disturbance by antibiotic exposure, and the taxa with different abundances made distinct responses to organic matter sources.We set up a column system to simulate the removal of antibiotics and explore the ecological mechanisms behind it in hyporheic groundwater by different treatment pathways including antibiotic exposure, bio-augmentation, and carbon source fluctuation. In the simulated soil column system, the removal efficiencies of tetracyclines, ciprofloxacin, and macrolide antibiotics were close to 100%, while those of sulfonamides were among 50–70%. The regulation treatment of bio-augmentation and carbon source fluctuation improved the community Shannon diversity at 0–10 cm depth and reduced the disturbance of antibiotic exposure to the microbial community, which enhanced the co-metabolism of antibiotics and thereby elevated the removal efficiencies. Besides, the regulation treatment had little effect on the relative abundances of antibiotic resistance genes and virulence factors.