Surface waters are becoming increasingly contaminated by pharmaceutically active compounds (PhACs), which is a potential risk factor for drinking water quality owing to incomplete riverbank filtration. This study examined the efficiency of riverbank filtration with regard to 111 PhACs in a highly urbanized section of the river Danube. One hundred seven samples from the Danube were compared to 90 water samples from relevant drinking water abstraction wells (DWAW) during five sampling periods. The presence of 52 PhACs was detected in the Danube, the quantification of 19 agents in this section of the river was without any precedent, and 10 PhACs were present in >80% of the samples. The most frequent PhACs showed higher concentrations in winter than in summer. In the DWAWs, 32 PhACs were quantified. For the majority of PhACs, the bank filtration efficiency was >95%, and not influenced by concentrations measured in the river. For carbamazepine lidocaine, tramadol, and lamotrigine, low (<50%) filtration efficiency was observed; however, no correlations were observed between the concentrations detected in the Danube and in the wells. These frequently occurring PhACs in surface waters have a relatively even distribution, and their sporadic appearance in wells is a function of both space and time, which may be caused by the constantly changing environment and micro-biological parameters, the dynamic operating schedule of abstraction wells, and the resulting sudden changes in flow rates. Due to the changes in the efficiency of riverbank filtration in space and time, predicting the occurrence and concentrations of these four PhACs poses a further challenge to ensuring a safe drinking water supply.

Small streams are crucial but vulnerable elements of ecological networks. To better understand the occurrence of pharmaceutically active compounds (PhACs) in streams, this study focused on the occurrence, distribution, and environmental risk of 111 PhACs and 7 trace elements based on a total of 141 water and sediment samples from small streams located in the urbanizing region of Budapest, Hungary. Eighty-one PhACs were detected in the aqueous phase, whereas sixty-two compounds were detected in the sediment. Carbamazepine (CBZ) was the most frequently identified PhAC in water, and was found in 91.5% of all samples. However, the highest concentrations were measured for lamotrigine (344.8 μg·L^-1) and caffeine (221.4 μg·L^-1). Lidocaine was the most frequently occurring PhAC in sediment (73.8%), but the maximum concentrations were detected for CBZ (395.9 ng·g^-1) and tiapride (187.7 ng·g^-1). In both water and sediment, more PhACs were found downstream of the wastewater treatment plants (WWTPs) than in the samples not affected by treated wastewater, even though no relationship was observed between the total amount of treated wastewater and the number of detected PhACs. The PhAC concentrations were also independent of the distance from the WWTP effluents. PhAC-polluted samples were detected upstream of the WWTPs, thereby suggesting the relevance of diffuse emissions in addition to WWTP outlets. The most frequently detected PhACs in the sediment were usually also present in the water samples collected at the same place and time. The varying concentrations of PhACs and the fluctuating water-sediment properties resulted in a lack of correlation between the general chemical properties and the concentrations of PhACs, which makes it difficult to predict PhAC contamination and risks in urbanized small streams. The environmental risk assessment indicated that diclofenac had the highest risk in the sampling area.

The role of the chemical properties of Pharmaceutically Active Compounds (PhACs) in their sorption behaviour and consequently in their fate and mobility is of major environmental interest, but a comprehensive evaluation is still lacking. The sorption of nine PhAC molecules with distinct physico-chemical properties on soils and goethite was described using linear, Freundlich and Langmuir models and the relationship between the chemical structures of the compounds and the parameters of the adsorption was evaluated using redundancy analysis (RDA). The latter showed that the sorption of the pharmaceuticals was determined by the intrinsic chemical characteristics of the molecules, as shown by the 35% value of constrained variability. For the hydrophobic estrogens, E1, E2 and EE2, the logD value and the number of hydrogen bond sites were found to be the main controlling factors for adsorption, indicating that hydrophobic interaction and hydrogen bonding are the dominant sorption mechanisms. The π energy of the molecules also proved a very important parameter, governing the retention of PhACs in soils, especially in the case of carbamazepine, oxazepam and lamotrigine. The main controlling factor for ionic compounds, such as diclofenac sodium, tramadol or lidocaine, is the fraction of PhACs present as charged species, revealing the importance of Coulomb forces. The results of this study will allow semi-quantitative predictions to be made on how the molecular structure governs the sorption of PhACs and which sorption mechanism could be involved.

Decreased water retention and increased runoff and soil loss are of special importance concerning soil degradation of hilly crop fields. In this study, plots under ploughing (conventional) tillage (PT) and conservation tillage (CT; 15 years) were compared. Rainfall simulation on 6 m² plots was applied to determine infiltration and soil loss during the growing season. Results were compared with those measured from 1200 m² plots exposed to natural rainfalls in 2016. Infiltration was always higher under CT than PT, whereas the highest infiltration was measured under the cover crop condition. Infiltration under seedbed and stubble resulted in uncertainties, which suggests that natural pore formation can be more effective at improving soil drainage potential than can temporary improvements created by soil tillage operations. Soil erodibility was higher under PT for each soil status; however, the seedbed condition triggered the highest values. For CT, soil loss volume was only a function of runoff volume at both scales. Contrarily, on PT plots, some extreme precipitation events triggered extremely high soil loss owing to linear erosion, which meant no direct connection existed between the scales. Improved soil conditions due to conservation practice are more important for decreasing soil loss than the better surface conditions.