Pharmaceuticals Load in the Svihov Water Reservoir (Czech Republic) and Impacts on Quality of Treated Drinking Water

: An important component of micropollutants are PPCPs (pharmaceuticals and personal care products). This paper contains the results of the monitoring of surface water, groundwater and wastewater in the surrounding area of the Svihov drinking water reservoir. Over the period 2017–2019, over 21,000 water samples were taken and analyzed for 112 pharmaceuticals, their metabolites, and other chemicals. The results are discussed in detail for two streams with the highest observed concentration of PPCPs (Hnevkovice, Dolni Kralovice) and two streams with the highest water inﬂow into the reservoir, representing also the highest mass ﬂow of PPCPs into the reservoir (Miletin, Kacerov). The overall analysis of the results shows that acesulfame, azithromycin, ca ﬀ eine, gabapentin, hydrochlorothiazide, ibuprofen and its metabolites, oxypurinol, paraxanthine, and saccharin (on some proﬁles up to tens of thousands ng / dm 3 ) attain the highest concentration and occur most frequently. The evaluation of raw water and treated drinking water quality showed the signiﬁcant positive e ﬀ ect of water retention in the reservoir (retention time of 413 days) and also of the treatment process, so that the treated drinking water is of high quality and contains only negligible residues of few PPCPs near the detection limit of the analytical method used.


Work Scope and Objectives
As the accuracy of analytical methods develops, it is possible to detect more and more substances belonging to the group of so-called micropollutants in natural water, occurring at very low concentration (usually in ng/dm 3 to µg/dm 3 ). A very important group of these substances are PPCPs (pharmaceuticals and personal care products)-medicines, hormones, antibiotics, cosmetics, drugs and other substances, including metabolites of the primary pollutants (Ferrer, Thurman [1], Richardson [2]). In the European Union, residues of a number of medicinal preparations have been detected in surface water and groundwater, soil, and in animal tissues at concentrations that vary depending on the medicinal preparation as well as on the nature and proximity of the source. Some painkillers, antimicrobials, antidepressants, contraceptives, and anti-parasitics are commonly found. Traces of some medical preparations may also penetrate drinking water and certain foods-European Commission [3].
The above findings led to the research on the Svihov reservoir on the Zelivka River used for drinking water supply to the population of the capital city of Prague and its surroundings, in order to observe the occurrence of PPCP substances in the largest water source in the Czech Republic. There have been long-term problems with the unsatisfactory level of municipal wastewater treatment about two thirds of the drinking water needs of the inhabitants of the capital city and its surroundings. The average long-term annual flow rate at the dam profile is 6.93 m 3 /s, and the 100year-old flow rate attains 316.0 m 3 /s (www.pvl.cz [35]).

Monitoring Methodology 2017-2018
A number of small local watercourses with a high proportion of wastewater, often poorly treated, from small municipalities around the reservoir flow into it and therefore PPCP type constituents were expected here (Morteani et al. [36], Swartz et al. [37], Kozisek et al. [38], Datel et al. [39]). The largest surface water inflows into the reservoir were identified in the project and nine main inflows with a significant proportion of wastewater from local municipalities were selected. Here, nine hydrological profiles were built ( Figure 1) characterizing the source water for the Svihov reservoir. The profiles were used for flow rate measurement and water sampling. The largest inflow into the reservoir is the Zelivka River (Miletin profile).
The flow rates were measured using a Cipoletti or Thomson spillway (sharp edge trapezoidal or triangular spillway), and the overflow height was measured using the Solinst Levelogger Edge automatic probe [40] positioned at the spillway. Control measurements (always at the time of sampling) were carried out using a measuring vessel and a stopwatch. On the streams with higher flow rates (i.e., Miletin, Kacerov), CHMI (Czech Hydrometeorological Institute) [41] data were used.
The samples were taken in one-month intervals, yet in the summer period (July-August), when we expected a decrease in flow rates and thus an increase in the concentration of the monitored substances, the sampling period was shortened to a frequency of two to four times a month. The sampling points were located below the discharge point of treated wastewater from the WWTPs (wastewater treatment plants). Simultaneously with the sampling of nine selected hydrological profiles, a sample of raw water was also taken from the reservoir at the inlet of the raw water treatment plant, and a sample of the treated drinking water entering the water supply system. Sampling was carried out in accordance with the relevant technical standards of the ISO 5667 series. A volumetric sampler on a telescopic rod was used to sample surface water, immersed in the water stream. Samples of raw and drinking water were taken from standard sampling taps in the raw water treatment plant. Within 32 sampling rounds, 21,174 water samples were taken from 20 February, 2017, to 12 November, 2018. Each sampling round was performed within one day.
At the start of the project in 2017, the laboratory offered to determine 46 different

Monitoring Methodology 2017-2018
A number of small local watercourses with a high proportion of wastewater, often poorly treated, from small municipalities around the reservoir flow into it and therefore PPCP type constituents were expected here (Morteani et al. [36], Swartz et al. [37], Kozisek et al. [38], Datel et al. [39]). The largest surface water inflows into the reservoir were identified in the project and nine main inflows with a significant proportion of wastewater from local municipalities were selected. Here, nine hydrological profiles were built ( Figure 1) characterizing the source water for the Svihov reservoir. The profiles were used for flow rate measurement and water sampling. The largest inflow into the reservoir is the Zelivka River (Miletin profile).
The flow rates were measured using a Cipoletti or Thomson spillway (sharp edge trapezoidal or triangular spillway), and the overflow height was measured using the Solinst Levelogger Edge automatic probe [40] positioned at the spillway. Control measurements (always at the time of sampling) were carried out using a measuring vessel and a stopwatch. On the streams with higher flow rates (i.e., Miletin, Kacerov), CHMI (Czech Hydrometeorological Institute) [41] data were used.
The samples were taken in one-month intervals, yet in the summer period (July-August), when we expected a decrease in flow rates and thus an increase in the concentration of the monitored substances, the sampling period was shortened to a frequency of two to four times a month. The sampling points were located below the discharge point of treated wastewater from the WWTPs (wastewater treatment plants). Simultaneously with the sampling of nine selected hydrological profiles, a sample of raw water was also taken from the reservoir at the inlet of the raw water treatment plant, and a sample of the treated drinking water entering the water supply system. Sampling was carried out in accordance with the relevant technical standards of the ISO 5667 series. A volumetric sampler on a telescopic rod Water 2020, 12, 1387 5 of 28 was used to sample surface water, immersed in the water stream. Samples of raw and drinking water were taken from standard sampling taps in the raw water treatment plant. Within 32 sampling rounds, 21,174 water samples were taken from 20 February, 2017, to 12 November, 2018. Each sampling round was performed within one day.
At the start of the project in 2017, the laboratory offered to determine 46 different pharmaceuticals; however, their number was increasing very rapidly, and by the second half of 2018, the laboratory was already able to analyze 93 different pharmaceuticals and their metabolites. In the follow-up monitoring in 2019 (see below), 20 more substances were added to 113 substances (Tables 1 and 2). The negative aspect of this development is the varying length of monitoring series duration for the individual substances.

Additional Monitoring in 2019
The monitoring results of 2017-2018 showed relatively high concentrations of PPCP substances in some local stream flow profiles. One of them was the Hnevkovice profile, which was selected for additional research. Samples from two wells in Hnevkovice were taken to compare the load in surface and groundwater, and in 2019 pre-treated wastewater from the municipal WWTP and water from the receiving stream (i.e., Hnevkovice Creek) were sampled ( Figure 2). Point S1 was chosen at the Hnevkovice Creek 20 m above the water discharge from the WWTP, and point S2 on the Hnevkovice Creek 20 m below the water discharge from the WWTP (point S2 is identical to the Hnevkovice profile from monitoring in 2017-2018). Two accessible municipal wells W1 and W2 were used for groundwater sampling. Sampling was carried out according to relevant technical standards ISO 5667 series. Wastewater is discharged from the WWTP at regular intervals of several minutes. The number and frequency of sampling was designed with respect to regular changes during the daily WWTP operation cycle. Two sampling cycles took place on 27-28 June and 21-22 October 2019. In one sampling cycle, one water sample was taken every three hours, i.e., eight samples per 24 h. Sampling of treated wastewater from the municipal wastewater treatment plant in the village of Hnevkovice was designed to determine the level and fluctuations of PPCP pollutant concentration in wastewater flowing to the Hnevkovice stream (and subsequently to the Svihov reservoir). One-off samples were also taken from the receiving stream (i.e., Hnevkovice Creek) above and below the discharge from the WWTP, in order to determine the impact of the discharged wastewater on water quality in the stream.
One-off sampling was carried out in accordance with the relevant technical standards of the ISO 5667 series. A volumetric zonal sampler on a rope was used to obtain the groundwater sample from the dug wells; water was sampled at about 0.5 m below the surface. A volumetric sampler on a telescopic rod was used to sample the wastewater and surface water and was immersed in the WWTP effluent or surface water flow.

Laboratory Work Methodology
The collected water samples were poured into glass sample bottles with ground neck/closure provided by the laboratory. The samples were stored in a refrigerated safety box according to the instruction of the laboratory and transported to an accredited laboratory, usually within a time Wastewater is discharged from the WWTP at regular intervals of several minutes. The number and frequency of sampling was designed with respect to regular changes during the daily WWTP operation cycle. Two sampling cycles took place on 27-28 June and 21-22 October 2019. In one sampling cycle, one water sample was taken every three hours, i.e., eight samples per 24 h. Sampling of treated wastewater from the municipal wastewater treatment plant in the village of Hnevkovice was designed to determine the level and fluctuations of PPCP pollutant concentration in wastewater flowing to the Hnevkovice stream (and subsequently to the Svihov reservoir). One-off samples were also taken from the receiving stream (i.e., Hnevkovice Creek) above and below the discharge from the WWTP, in order to determine the impact of the discharged wastewater on water quality in the stream.
One-off sampling was carried out in accordance with the relevant technical standards of the ISO 5667 series. A volumetric zonal sampler on a rope was used to obtain the groundwater sample from the dug wells; water was sampled at about 0.5 m below the surface. A volumetric sampler on a telescopic rod was used to sample the wastewater and surface water and was immersed in the WWTP effluent or surface water flow.

Laboratory Work Methodology
The collected water samples were poured into glass sample bottles with ground neck/closure provided by the laboratory. The samples were stored in a refrigerated safety box according to the instruction of the laboratory and transported to an accredited laboratory, usually within a time between 6-10 h, where they were refrigerated, and the bottles were stored in an inclined position. The samples were analyzed by an accredited laboratory of the state-owned enterprise Povodi Vltavy (Vltava Watershed Authority) in Plzen, according to the internal standard operating procedures (SOP O-19-A) using the LC/MS (liquid chromatography/mass spectrometry) analytical method.
One multicomponent method was used for all analytes from the group of pharmaceuticals -direct injection of the water sample after centrifugation, pH adjustment (acidification) and an addition of a mixture of isotopically labeled standards. Micropollutants were measured with Agilent Technologies 1290 Infinity II UHPLC system coupled to an Agilent 6495 Triple Quadrupole MS system using a Waters Xbridge C18 column (4.6 × 100 mm, particle size 3.5 µm) supplemented with a pre-column, with an injection volume of 50 µL (manufacturer Agilent Technologies International Pte. Ltd., Yishun Ave 7, 768923 Singapore).
Within method validation, stability tests were performed to confirm the stability of all analytes in the sample for at least three weeks when refrigerated at an inclined position. The sample had to be cooled as soon as possible after collection and then frozen.
The quality of the analytical work was assured by adding isotopically labeled standards to each sample (internal standard method). The samples were evaluated on an eight-point linear calibration curve over a concentration range of 10-2000 ng/dm 3 . In each group of samples, two blank samples and two quality control samples of the entire procedure (independent mixture of all analytes added to water) were measured and evaluated. In addition, every fifth sample was measured in duplicate (once without addition and once with the addition of an independent mixture of all analytes) and the actual yields in the given matrix were calculated from the concentration difference and included in the final concentration calculation.

Processing of Chemical Analysis Results
The PPCP analytical data processing method encounters a problem caused by the fact that a significant part of the results lies below the detection limit. One solution could be that only results above the detection limit will be presented. This would, however, result in a significant distortion of the situation and underestimation of the results obtained.
Finally, a solution on the safe side, respecting the precautionary principle, was chosen as the basic approach to the risk analysis: samples with values below the detection limit are for the purposes of display in charts and the calculation of statistical parameters used at the detection limit concentration.
Obviously, this approach has its drawbacks and presents the theoretically most unfavorable situation. However, given the importance of the research for the safety of the population, we consider this approach to be relatively the most correct one.
The evaluation of time series was based on the statistical parameter of median, and on the extreme values (MIN, MAX). The basic advantage of the median is that it is less affected by the extreme values. The time series of chemical analyses that we had display two problematic characteristics that disadvantage the use of the average: a significant proportion of samples has an unknown specific concentration value because they are below the detection limit of the analytical method, and additionally, from time to time extremely high concentrations occur, capable of greatly influencing the average value, but not the median value. The MAX values are essential for the safety of the population, as the assessment of drinking water is based on the requirement that the limits set are not exceeded.
For comparison, in the time series of concentrations in Section 3, the median, arithmetic mean, standard deviation and coefficient of variation are calculated. It can be seen that very high standard deviation values predominate, indicating a large statistical variability of the processed data sets.
The coefficient of variation (i.e., the ratio of the standard deviation to the arithmetic mean) reaches a minimum value of 33%, and in 28 out of 36 evaluated time series (sets of nine substances from four profiles) it exceeds 50%; nine sets have a coefficient of variation higher than 100%-these values make using the mean as a characteristic of the series very problematic.

Results
All results of chemical analyses are available in Table S1 in Supplementary Materials.

Monitoring 2017-2018
The monitoring results made it possible to separate the monitored streams into two basic groups: four streams with significantly higher concentrations and more frequent occurrence of PPCP substances (Hnevkovice, Kozli, Dolni Kralovice and Bernartice), and five profiles with relatively lower concentrations and less frequent occurrence of monitored substances (Radikovice, Hulice, Kacerov, Miletin and Brzotice). It has also been shown that some of the studied substances occur more frequently and in higher concentrations than others (see Discussion).
Furthermore, for the sake of clarity, we present in the charts nine PPCP substances, selected according to the highest concentration, highest frequency of occurrence in the monitoring period 2017-2018, and also according to occurrence in raw and treated drinking water. Four profiles were selected from the nine monitored profiles on local watercourses ( Figure 1): two profiles at the largest tributaries to the Svihov reservoir (Miletin and Kacerov) and two profiles with the highest concentration and the highest incidence of monitored PPCP substances (Hnevkovice and Dolni Kralovice).

Hnevkovice Profile
The sampling profile on the Hnevkovice Creek is located at the end of the village of Hnevkovice, below the discharge point from the sewage treatment plant (point S2 in Figure 2). Further, under the profile, the stream continues through a forest, and after approximately 600 m it flows into the Svihov reservoir. Raw water intake at the reservoir dam is located about 11 km from the estuary of the stream into the reservoir. The catchment area is 0.543 km 2 , rural settlement prevails (about 300 inhabitants in family houses with small farm animals), and intensively cultivated fields prevail around the village. The average rainfall is 681.8 mm, average temperature is 8.04 • C (data for the period 1961-2015), average altitude is 445.45 m above sea level, and average slope is 3.4% (TGM Water Research Institute internal database). The flow rate in the creek is strongly influenced by fluctuating effluent from the sewage water treatment plant and retention ponds, fluctuating between 0.0004-0.0056 m 3 /s (average 0.0016 m 3 /s); natural flow rate constitutes a lesser part of the total flow rate, especially under low flow conditions. The PPCP monitoring results are shown in Figure 3 and Tables 3 and 4.  according to the highest concentration, highest frequency of occurrence in the monitoring period 2017-2018, and also according to occurrence in raw and treated drinking water. Four profiles were selected from the nine monitored profiles on local watercourses ( Figure 1): two profiles at the largest tributaries to the Svihov reservoir (Miletin and Kacerov) and two profiles with the highest concentration and the highest incidence of monitored PPCP substances (Hnevkovice and Dolni Kralovice).

Hnevkovice Profile
The sampling profile on the Hnevkovice Creek is located at the end of the village of Hnevkovice, below the discharge point from the sewage treatment plant (point S2 in Figure 2). Further, under the profile, the stream continues through a forest, and after approximately 600 m it flows into the Svihov reservoir. Raw water intake at the reservoir dam is located about 11 km from the estuary of the stream into the reservoir. The catchment area is 0.543 km 2 , rural settlement prevails (about 300 inhabitants in family houses with small farm animals), and intensively cultivated fields prevail around the village. The average rainfall is 681.8 mm, average temperature is 8.04 °C (data for the period 1961-2015), average altitude is 445.45 m above sea level, and average slope is 3.4% (TGM Water Research Institute internal database). The flow rate in the creek is strongly influenced by fluctuating effluent from the sewage water treatment plant and retention ponds, fluctuating between 0.0004-0.0056 m 3 /s (average 0.0016 m 3 /s); natural flow rate constitutes a lesser part of the total flow rate, especially under low flow conditions. The PPCP monitoring results are shown in Figure 3 and Tables 3, 4.  This is a profile with the unambiguously worst water quality values, with only a minimum number of values below the detection limit. All nine studied substances have maximum concentrations in thousands of ng/dm 3 , and four substances even in tens of thousands of ng/dm 3 (acesulfame, gabapentin, oxypurinol, paraxanthine). Additionally, the median values are high, particularly with median value of 20,250 ng/dm 3 found for oxypurinol. Furthermore, other substances not evaluated in this article had a maximum concentration over 1000 ng/dm 3 (measured maximum concentration in brackets): azithromycin (1360 ng/dm 3 ), celiprolol (1690 ng/dm 3 ), clarithromycin (2600 ng/dm 3 ), furosemide (1800 ng/dm 3 ), ibuprofen-2-hydroxy (700 ng/dm 3 ), ibuprofen-carboxy (2200 ng/dm 3 ), iopromide (34,400 ng/dm 3 , but rare occurrence), caffeine (3900 ng/dm 3 ), lamotrigine (1310 ng/dm 3 ), paracetamol (1800 ng/dm 3 ), saccharin (10,000 ng/dm 3 ). Values of over 1000 ng/dm 3 were in most samples for celiprolol (7 out of 8) and in a quarter to half of the samples of ibuprofen-2-hydroxy, caffeine and saccharin. For other substances, elevated concentrations are uncommon (1-5 out of 28). It is also worth noting that for iopromide 15 samples out of 28 were below the detection limit.
When assessing the nature of the settlement and land use, this pilot area is in no way different from the others in explaining the causes of the high concentrations of PPCP. Therefore, the Hnevkovice profile was chosen for additional monitoring in 2019 (Section 3.2).

Dolni Kralovice Profile
The measured profile is located at the end of the Dolni Kralovice village under the discharge from the WWTP (Figure 1). The creek flows into the Svihov reservoir after approximately 1600 m. Raw water intake from the reservoir at the dam is located about 11 km from the estuary of the stream into the reservoir. The catchment area is 1769 km 2 , rural settlements predominate (approximately 700 inhabitants in family houses and small apartment houses, including shops, buildings with services and medical facilities). There is a large agricultural farm in the village including livestock and food industry buildings, and the surrounding land is dominated by fields with intensive crop production. The average rainfall is 670.8 mm, average temperature is 7.9 • C (data for the period 1961-2015), average altitude is 478.6 m above sea level, and the average slope is 3.4% (TGM Water Research Institute internal database). The flow in the stream is strongly influenced by fluctuating effluent from the wastewater treatment plant and retention basin fluctuating between 0.0012-0.0114 m 3 /s (average flow is 0.004 m 3 /s), natural flow constituting a minority of the total flow rate, especially under low flow conditions. The PPCP monitoring results are shown in Figure 4 and Tables 5 and 6.
Like Hnevkovice, the Dolni Kralovice profile belongs to those with the highest concentrations of PPCP substances. Oxypurinol, again, reaches the highest concentrations (maximum 20,000 ng/dm 3 , median 8205 ng/dm 3 ), whereas all other selected substances except metoprolol and tramadol have peak values over 1000 ng/dm 3 ; acesulfame, gabapentin, and hydrochlorothiazide even have median values above 1000 ng/dm 3 . Also, for some other substances (not selected for further evaluation in this article), peak concentrations of over 1000 ng/dm 3 were found (measured maximum concentration in brackets): azithromycin (1600 ng/dm 3 ), ibuprofen-2-hydroxy (12,900 ng/dm 3 ), irbesartan (1530 ng/dm 3 ), caffeine (4260 ng/dm 3 ), naproxen (1050 ng/dm 3 ), saccharin (6680 ng/dm 3 ). Values above 1000 ng/dm 3 were in the majority of samples for irbesartan (7 out of 8), but for other substances the increase in concentration is less common (1-6 out of 28).     The character of settlement and land use may explain the higher concentrations of PPCP in this catchment area. The community of Dolni Kralovice is one of the larger municipalities in the vicinity of the Svihov reservoir, and in addition the number of inhabitants is increased by employees in agricultural and food processing plants located in the catchment area. There are also service facilities, including medical facilities, where higher drug consumption is expected.

Miletin Profile
The measured profile is located on the Zelivka River, on which the Svihov reservoir is built, at the point of the Zelivka River inlet into the Svihov reservoir. It is the main inflow, providing about 70% of the total average inflow into the reservoir. The hydrological catchment belonging to the measured profile is extensive; the catchment area is 3020 km 2 . The location of raw water intake from the reservoir at the dam is located about 20 km from the estuary of the river into the reservoir. The catchment area of the upper Zelivka River has many villages and towns, with approximately 40,000 inhabitants. Land use is mixed with intensive agricultural production, commercial forests, and settlements. Larger towns include Pelhrimov (16,000 inhabitants), Pacov (5000 inhabitants) and Cervena Recice (1000 inhabitants), while other inhabitants live in several rural municipalities with less than 1000 residents. In the catchment area there are several hospitals, other medical facilities, homes for the elderly and other facilities used by services. In the catchment, the average rainfall is 671.8 mm and average temperature 8.03 • C (data for the period 1961-2015), average altitude is 433.44 m above sea level and average slope is 3.3% (TGM Water Research Institute internal database). The flow rate in the river is influenced by several reservoirs and ponds, and by the discharge of wastewater throughout the catchment. The flow rate fluctuated between 0.3-6.6 m 3 /s (long-term average flow rate of 2.8 m 3 /s according to the Czech Hydrometeorological Institute CHMI), and the share of wastewater represents a significant minority of the total flow rate (estimated at about 0.04 m 3 /s). The PPCP monitoring results are shown in Figure 5 and Tables 7 and 8.
Due to the high flow rates and large river catchment area, the water flow is diluted and delayed, resulting in lower PPCP concentration. The relatively low concentrations are offset by higher flow rates. Oxypurinol again shows the highest concentration (peak 1820 ng/dm 3 , median 483 ng/dm 3 ), and paraxanthine exhibits maximums over 1000 ng/dm 3 . When other substances are assessed, no other has a maximum concentration above 1000 ng/dm 3 .   Acesulfame  16  50  0  Diclofenac  28  20  20  Gabapentin  28  10  2  Hydrochlorothiazide  28  50  25  Ibuprofen  28  20  16  Metoprolol  28  10  17  Oxypurinol  16  50  2  Paraxanthine  26  100  1  Tramadol  28 10 5    Although the PPCP concentrations are lower on the Miletin profile than on the previous two profiles, due to the high flow rate this stream brings the largest amount of PPCP into the Svihov reservoir (Section 4). Higher quantities of transported pharmaceuticals can be logically explainedthere are many places in the large catchment area where significant mass flows of pharmaceuticals can be expected (larger cities, hospitals and other health and veterinary facilities, homes for the elderly, service facilities, etc.).

Kacerov Profile
The measured profile is located on Sedlicky Creek near the village of Kacerov at the point where the stream flows into the Svihov reservoir (edge of reservoir backwater). It is the largest tributary to the Svihov reservoir. The location of raw water intake from the reservoir at the dam is located about 6 km from the estuary of the stream into the reservoir. The monitored catchment is relatively extensive; Sedlicky Creek, with many tributaries, has a catchment area of 76 km 2 . Within the catchment area, there are several municipalities with about 2300 inhabitants. Rural settlement prevails amid the villages; there are intensively cultivated fields and commercial forests. The main municipalities are Lhota Bubenec (village with intensive livestock farming), Krivsoudov (village), Ruzkovy Lhotice (village with a small farm), Cernici (smaller agricultural farm), Mala Paseka (village), Cechtice (larger village, two large agricultural farms, two industrial areas, medical facilities and service buildings) and Chrastovice (former village, now completely abandoned). The monitored catchment area has an average rainfall of 692.1 mm, and an average temperature of 7.87 • C (data for the period 1961-2015). The average altitude is 506.34 m above sea level, and the average slope is 3.6% (TGM Water Research Institute internal database). A number of ponds and the discharge of wastewater throughout the catchment area influence the creek flow rate. The flow rate fluctuated between 0.002-0.43 m 3 /s during the measurement period, and the proportion of wastewater does not constitute a major portion of the total flow rate, except in low flow conditions. The PPCP monitoring results are shown in Figure 6 and Tables 9 and 10.   Acesulfame  15  50  0  Diclofenac  26  20  15  Gabapentin  26  10  4  Hydrochlorothiazide  26  50  11  Ibuprofen  26  20  18  Metoprolol  26  10  13  Oxypurinol  15  50  3  Paraxanthine  25  100  10  Tramadol  26 10 9    Of the selected substances, oxypurinol and hydrochlorothiazide reach the highest concentration with a maximum above 1000 ng/dm 3 , while acesulfame and paraxanthine peak just below 1000 ng/dm 3 . No other substances with concentrations above 1000 ng/dm 3 were observed on the Kacerov profile; only one azithromycin analysis (931 ng/dm 3 ) approached the 1000 ng/dm 3 limit.

Number of Samples below Detection Limit
Although the PPCP concentrations are on average even lower on the Kacerov profile than on the Miletin profile, there is also a significant PPCP mass flow into the reservoir (Section 4). The sources of PPCP in the catchment area can be well identified -seven municipalities with agricultural and industrial sites, service facilities, medical facilities, in which the number of inhabitants is slightly increased by commuting employees).

Raw and Drinking Water Monitoring
The 2017-2018 water monitoring also included quality control of raw and drinking water. Raw water is taken from the intake tower building at the Svihov reservoir dam; the treated drinking water was sampled at the point where the water leaves the treatment plant to the water supply system. Detected substances above the detection limit are summarized in Table 11 (raw water) and Table 12 (drinking water). In contrast to the previous tables and graphs, all the analyzed substances were evaluated (93 items according to Table 1) because of their importance for the overall drinking water quality. Table 11. Quality of raw water from the Svihov reservoir-all substances detected above the detection limit.

Substance
Number of Analyses Total  Only 20 substances out of 93 in raw water had concentration values above the detection limit at least once during the 2-year monitoring period, whereas 73 substances were permanently below the detection limit.

Additional Monitoring Hnevkovice
In order to investigate the local situation more precisely, additional monitoring was carried out in the community of Hnevkovice (profile with the highest detected concentration of PPCP substances in 2017-2018 monitoring), consisting of detailed sampling of wastewater at the discharge from the WWTP, stream water above and below the WWTP, and groundwater from two selected wells. Samples of water at the discharge from the WWTP were taken in three-hour intervals on 27-28 June and on 21-

Additional Monitoring Hnevkovice
In order to investigate the local situation more precisely, additional monitoring was carried out in the community of Hnevkovice (profile with the highest detected concentration of PPCP substances in 2017-2018 monitoring), consisting of detailed sampling of wastewater at the discharge from the WWTP, stream water above and below the WWTP, and groundwater from two selected wells.    The highest concentrations, observed in the spring period were found for iopromide (73,600 ng/dm 3 , spring) and oxypurinol (33,100 ng/dm 3 ). In autumn, the concentrations were mostly lower, with oxypurinol (21,500 ng/dm 3 ) and sulfamethoxazole (16,100 ng/dm 3 ) being the highest. Surprisingly, the lowest concentration of monitored substances were also observed for iopromide (72 ng/dm 3 , autumn) and ibuprofen (below the detection limit). Iopromide is the substance with the largest, but well-explained, fluctuation in concentration. Because it is used as a contrast medium, e.g., in CT scans, its extremely high and short-term concentrations are related to the rapid excretion of this substance after the performed examination. Oxypurinol is a frequently used pharmaceutical for the treatment of increased uric acid content in blood, and the associated problems with joints and kidneys. Its high concentrations indicate the prevailing elderly population in the village.
For some other substances (over 16 mentioned above in charts) maximum concentrations were also found at over ,000 ng/dm 3  The charts in Figures 10 and 11 document the change of water quality in the Hnevkovice Creek due to the discharge of treated wastewater (16 selected substances according to Figures 7-9). In the stream profile above, the WWTP (S1 profile in Figure 2) were detected only slightly above the detection limit: in June 2019 gabapentin (77.1 ng/dm 3    The highest concentrations, observed in the spring period were found for iopromide (73,600 ng/dm 3 , spring) and oxypurinol (33,100 ng/dm 3 ). In autumn, the concentrations were mostly lower, with oxypurinol (21,500 ng/dm 3 ) and sulfamethoxazole (16,100 ng/dm 3 ) being the highest. Surprisingly, the lowest concentration of monitored substances were also observed for iopromide (72 ng/dm 3 , autumn) and ibuprofen (below the detection limit). Iopromide is the substance with the largest, but well-explained, fluctuation in concentration. Because it is used as a contrast medium, e.g., in CT scans, its extremely high and short-term concentrations are related to the rapid excretion of this substance after the performed examination. Oxypurinol is a frequently used pharmaceutical for the treatment of increased uric acid content in blood, and the associated problems with joints and kidneys. Its high concentrations indicate the prevailing elderly population in the village.
For some other substances (over 16 mentioned above in charts) maximum concentrations were also found at over ,000 ng/dm 3  The charts in Figures 10 and 11 document the change of water quality in the Hnevkovice Creek due to the discharge of treated wastewater (16 selected substances according to Figures 7-9). In the stream profile above, the WWTP (S1 profile in Figure 2) were detected only slightly above the detection limit: in June 2019 gabapentin (77.1 ng/dm 3 ), in October 2019 sulfamethoxazole (10.3 The highest concentrations, observed in the spring period were found for iopromide (73,600 ng/dm 3 , spring) and oxypurinol (33,100 ng/dm 3 ). In autumn, the concentrations were mostly lower, with oxypurinol (21,500 ng/dm 3 ) and sulfamethoxazole (16,100 ng/dm 3 ) being the highest. Surprisingly, the lowest concentration of monitored substances were also observed for iopromide (72 ng/dm 3 , autumn) and ibuprofen (below the detection limit). Iopromide is the substance with the largest, but well-explained, fluctuation in concentration. Because it is used as a contrast medium, e.g., in CT scans, its extremely high and short-term concentrations are related to the rapid excretion of this substance after the performed examination. Oxypurinol is a frequently used pharmaceutical for the treatment of increased uric acid content in blood, and the associated problems with joints and kidneys. Its high concentrations indicate the prevailing elderly population in the village.
For some other substances (over 16 mentioned above in charts) maximum concentrations were also found at over 1000 ng/dm 3  The charts in Figures 10 and 11 document the change of water quality in the Hnevkovice Creek due to the discharge of treated wastewater (16 selected substances according to Figures 7-9). In the stream profile above, the WWTP (S1 profile in Figure 2) were detected only slightly above the detection limit: in June 2019 gabapentin (77.1 ng/dm 3 ), in October 2019 sulfamethoxazole (10.3 ng/dm 3 ) and telmisartan (32.2 ng/dm 3 ). Under the effluent discharge from the WWTP, concentrations of the monitored substances in the order of hundreds to thousands of ng/dm 3 were found, the highest concentration being that for oxypurinol (13,700 ng/dm 3 in spring and 14,600 ng/dm 3 in autumn).
Water 2020, 12, x 20 of 29 ng/dm 3 ) and telmisartan (32.2 ng/dm 3 ). Under the effluent discharge from the WWTP, concentrations of the monitored substances in the order of hundreds to thousands of ng/dm 3 were found, the highest concentration being that for oxypurinol (13,700 ng/dm 3 in spring and 14,600 ng/dm 3 in autumn).  High concentrations of some substances in surface and also in treated wastewater have raised the question of the extent to which these substances also occur in shallow groundwater, used by household wells to supply the population with drinking water. That groundwater is also drained into the Svihov reservoir and forms one-albeit small-water source for the reservoir (estimated inflow of 0.01-0.02 m 3 /s from the area immediately adjacent to the reservoir). Two municipal wells were selected in the center of Hnevkovice (W1, W2, Figure 2), and in October 2018 the water was analyzed for the same range of substances as surface water (93 substances, Table 1). Groundwater is significantly less contaminated, with only four substances with low concentration (Table 13). Water 2020, 12, x 20 of 29 ng/dm 3 ) and telmisartan (32.2 ng/dm 3 ). Under the effluent discharge from the WWTP, concentrations of the monitored substances in the order of hundreds to thousands of ng/dm 3 were found, the highest concentration being that for oxypurinol (13,700 ng/dm 3 in spring and 14,600 ng/dm 3 in autumn).  High concentrations of some substances in surface and also in treated wastewater have raised the question of the extent to which these substances also occur in shallow groundwater, used by household wells to supply the population with drinking water. That groundwater is also drained into the Svihov reservoir and forms one-albeit small-water source for the reservoir (estimated inflow of 0.01-0.02 m 3 /s from the area immediately adjacent to the reservoir). Two municipal wells were selected in the center of Hnevkovice (W1, W2, Figure 2), and in October 2018 the water was analyzed for the same range of substances as surface water (93 substances, Table 1). Groundwater is significantly less contaminated, with only four substances with low concentration (Table 13). High concentrations of some substances in surface and also in treated wastewater have raised the question of the extent to which these substances also occur in shallow groundwater, used by household wells to supply the population with drinking water. That groundwater is also drained into the Svihov reservoir and forms one-albeit small-water source for the reservoir (estimated inflow of 0.01-0.02 m 3 /s from the area immediately adjacent to the reservoir). Two municipal wells were selected in the center of Hnevkovice (W1, W2, Figure 2), and in October 2018 the water was analyzed for the same range of substances as surface water (93 substances, Table 1). Groundwater is significantly less contaminated, with only four substances with low concentration (Table 13). Table 13. Substances detected above the detection limit of two groundwater sources-Hnevkovice village wells (Figure 2). The results indicate that the main problem of the Svihov reservoir catchment area in terms of occurrence of PPCP substances is the municipal wastewater and the level of its treatment in local WWTPs.

Discussion of 2017-2018 Monitoring Results
The selected nine sampling profiles at the tributaries of the Svihov drinking water supply reservoir provided very different results. Regarding concentrations of the PPCP substances (93 analyzed substances in total), the Hnevkovice and Kozli profiles appear to be the worst, where the total sum of maximum concentrations of the analyzed substances is close to 200,000 ng/dm 3 . Next are the profiles Dolni Kralovice and Bernartice with a total sum of PPCP maximum concentration around 100,000 ng/dm 3 . The remaining five profiles have significantly lower concentrations of PPCP substances: Miletn, Kacerov and Radikovice with a sum of maximum concentrations around 8-10,000 ng/dm 3 , and the best quality profiles appear to be the profiles of Hulice a Brzotice with a sum of maximum concentrations of approximately 2000 ng/dm 3 .
The explanation is currently not entirely clear. The effect of dilution in higher flow rate streams (i.e., Miletin, Kacerov) is clear, but this does not apply to the cleanest profiles of Hulice and Brzotice, where there are also very low flow rates and comparable quantity of wastewater from other municipalities (Hnevkovice, Bernartice). Leaving aside the possibility of lower consumption of PPCP substances by the inhabitants of these municipalities (which does not seem very likely, as all municipalities are very similar in character), the only possible explanation may be the different efficiency of the municipal WWTPs in terms of PPCP removal.
Regarding the mass flow of PPCP substances into the Svihov reservoir, the situation is different; the monitored local streams usually have a flow rate fluctuating between 0.0002-0.005 m 3 /s (lower flow rates than long-term due to the drought since 2015). The flow rate on the Kacerov profile, however, is on the order of hundredth of m 3 /s (long-term average is 0.39 m 3 /s [35], and the average for the monitored period during the long-term drought is 0.08 m 3 /s). Additionally, the flow rate on the Miletin profile is in units of m 3 /s (long-term average 2.48 m 3 /s [35], average for the monitored period during the long-term drought is 1.300 m 3 /s). Thus, about 42% of the inflow was monitored in this project (long-term runoff average from the dam is 6.93 m 3 /s, inclusive of water-supply abstraction, [35]). Therefore, we consider the performed monitoring to be representative in terms of quality assessment of the inflowing water into the reservoir. The difference in flow rates is also reflected in the different mass flow of substances (Table 14). An accurate calculation is not possible due to large and frequent fluctuations in PPCP concentration and in WWTP effluent; therefore, the table contains only qualified parameter estimates and we are convinced that the estimates of the mass flow were accurate to the order of magnitude. In total, this represents a mass flow of approximately 180 g/day from 42% of total reservoir inflow. It can be estimated that around 400-500 g of PPCP substances flow into the Svihov reservoir per day on average (in the analyzed range, as shown in Tables 1 and 2).
Considering the frequency of occurrence of the individual PPCP substances analyzed, a group of substances that occur very often on most profiles and often at high concentration (thousands and even tens of thousands of ng/dm 3 ) can be identified: acesulfame, azithromycin, caffeine, gabapentin, hydrochlorothiazide, ibuprofen and its metabolites, oxypurinol, paraxanthine, and saccharin. In addition to these constituents, we can identify other substances also occurring in the majority of analyses, but at a lower frequency and with lower concentrations (predominantly hundreds of ng/dm 3 ): iopromide, metoprolol, paracetamol, and tramadol (Datel et al. [42,43]). Additionally, there is a group of 39 substances (out of the 112) that occur always, or usually, below the detection limit (see brown marked substances in Tables 1 and 2).
Interestingly, a good correlation between the flow rate and the concentration of PPCP substances cannot be established. This could be due to the high and fluctuating proportion of wastewater effluent from the WWTPs or the fluctuating quality of the effluent discharged (different residence time in the retention tank, different quality of sewage water entering the WWTP, probably different time of water treatment in the WWTP at peak loading and outside of it, etc.).
Other uncertainties during monitoring were the non-standard results from two sampling rounds 6-7 in 2018. For some substances, there is an inexplicable decline while for others there is an increase in the observed concentration without a clear link to the change in flow rates. If this is not an analysis error (which the laboratory has excluded), it would be possible to consider a change in the population composition in the area, e.g., due to summer holidays, which could also be reflected in the change in the way the various PPCP substances are used. It should be noted that the concentration of various pharmaceuticals measured might be related to only a few people who take these medicaments. Moreover, if these people leave temporarily and if other people taking completely different medicaments arrive for the summer holidays, this may have an impact on the changed wastewater quality. Verifying this hypothesis would, however, assume a socio-demographic study in the area.

Discussion of Addittional 2019 Wastewater Moitoring-Hnevkovice
Additional monitoring of the WWTP Hnevkovice wastewater effluent in the year 2019 confirmed the most frequently occurring substances from the previous monitoring, and in addition other substances were found in high concentration in the WWTP effluent (thousands ng/dm 3 ): celiprolol, diclofenac and diclofenac-4-hydroxy, furosemide, lamotrigine, metoprolol, sulfamethoxazole, and telmisartan. In addition, wastewater analyses have also shown frequent and high concentrations of 4-formylaminoantipyrine, benzotriazole, and sucralose, which were not analyzed during the monitoring of 2017-2018.
The results show the insufficient functionality of WWTPs in terms of removal of PPCP (even in comparison with similar neighboring municipalities with comparable conditions (Bernartice, Brzotice, Hulice). Municipalities with poor quality of treated wastewater may have WWTPs with overloaded capacity. The low efficiency of local municipal wastewater treatment can be also related to the obsolete technology at some WWTPs; small municipalities often do not have professional staff for the correct operation of WWTPs, and in some cases they have insufficient capacity, especially at peak wastewater inflows to WWTPs.
Wastewater in Hnevkovice (and other municipalities around the Svihov water reservoir) is drained separately from storm water, so that only sewage water from the households and operations flows to the municipal wastewater treatment plants. Rainwater from streets and roofs is drained by the rainwater sewer to the nearest watercourse. The quantity of wastewater is the result of household water consumption, which is currently in small municipalities around 0.07-0.08 m 3 per person per day [35]. In the Czech Republic, wastewater treatment plants are mostly two-stage mechanical-biological. The mechanical part separates larger objects and particles on the screens and separates the coarse sediment, fats and oils. Finally, the fine sediment is allowed to settle in a sedimentation tank. The biological part of the WWTP consists of activation tanks, where the wastewater is aerated, and aerobic biological treatment takes place in which bacteria and other organisms consume organic substances, nitrogen and phosphorus. Activated sludge from the bodies of microorganisms is separated in the consecutive settling tanks and the water flows for natural final treatment to the retention tank under the WWTP or directly into the stream. Most small municipal WWTPs operate in an automatic mode, but the optimum treatment efficiency depends on good maintenance and correct setup of the plant. The facility in Hnevkovice is under new construction and the community is experiencing population growth, and therefore the original capacity of the WWTP is probably also overloaded, thus reducing its efficiency [38,39].
It is not possible to directly compare the water samples at the effluent from the WWTP with the water samples in the stream below the WWTP. PPCP concentrations in wastewater are variable over time, the quantity of wastewater also changes over time, and the water flow rate in the stream is not constant, and further there is a retention tank between the WWTP and the stream. The retention tank retains water for several days for sedimentation and post-treatment processes before the water is released into the watercourse. The explanation for this phenomenon may be the considerable variation in the concentration of the analyzed substances in wastewater (depending on the intensity of use of various pharmaceuticals), including daily, weekly and seasonal cycles. The monitoring did not record these maximum concentrations (samples taken 1-2 times per month, on average). Substances can then accumulate in the retention tank and be gradually released, even in higher concentrations than represented by actual water quality coming directly from the WWTP.

Raw Water
Analysis of monitoring data from 2017-2018 shows that there is a significant difference in poorer water quality of tributaries to the reservoir and significantly better quality of water abstracted at the dam for treatment into drinking water. Therefore, there are favorable processes in the reservoir space that have a positive impact on the improvement of water quality, not only in PPCP parameters. According to the data of the water authority, the theoretical residence time of the water in the reservoir is 430 days, and the total long-term average inflow into the reservoir is 6.93 m 3 /s. The reservoir has a length of 39.1 km, and the volume of water in the reservoir is 309 million m 3 [35].
The identification of the processes in the reservoir that have an impact on PPCP concentrations was not the subject of this project. However, it may be noted that the decrease in concentrations of some of the monitored substances may certainly be related to their degradation properties, whether on the basis of biochemical or physical and chemical phenomena. The identified varying resistance to degradation is undoubtedly related to different resistance of the chemical molecules of the monitored organic substances to these processes. However, all of this is a matter of development in time, which is very rapid both in the development of knowledge about the behavior of these substances and in the development of analytical methods.
The subject of deposition of PPCP substances in sediments is also discussed in various archive sources. However, this accumulation does not seem to play a major role in the migration properties of most of the monitored substances. In addition, in this research several samples of sediments were taken for PPCP analysis in the Hnevkovice Creek estuary into the reservoir, but mostly with negative results (very small concentrations of PPCP in a solid phase).
A significant decrease in concentrations due to the retention in the reservoir occurred for the following substances: azithromycin, diclofenac, hydrochlorothiazide, ibuprofen-2-hydroxy, lamotrigine, metoprolol, telmisartan, and tramadol. On the other hand, there is a group of substances that have not displayed a significant decrease due to retention in the water reservoir: estrone, chloramphenicol, oxypurinol, progesterone, and sulfamerazine. A relatively minor effect of reservoir water retention is also shown for acesulfame, gabapentin, ibuprofen, carbamazepine, carbamazepine E, paraxanthine, ranitidine, and trimethoprim.
It can be concluded that a water reservoir with a sufficient retention time is much better for achieving a more favorable water quality, regarding the content of specific PPCP-type organic substances (and probably also pesticides, PAHs (polyaromatic hydrocarbons) and other organic degradable substances) than direct abstraction from the stream, but this influence differs for different substances.

Treated Drinking Water
For some PPCP substances, we have noticed a significant decrease during the water treatment process. The Zelivka Raw Water Treatment Plant is the largest water treatment plant for the capital city of Prague. The normal operating capacity of the treatment plant is 3 m 3 /s of drinking water, and at peak times it can produce up to 7 m 3 /s, making it one of the largest water treatment plants in Europe (www.pvl.cz [35], www.zelivska.cz [44]). The basic technology of raw water treatment is coagulation filtration with lowering of pH and dosage of aluminum sulphate. Water with precipitated impurities then undergoes one-stage filtration in open rapid sand filters (water passes through a layer of sand of the fraction 1.1-1.6 mm with a thickness of 1.6 m for about 1 h). Drinking water is finally treated with ozone and chlorine [45]. The basis of the currently developed treatment plant modernization is a new additional stage of treatment with granulated activated carbon filters to capture pesticides and other substances (e.g., medicines, hormones) in the future (Lepka [45]).
Analysis of monitoring data from 2017-2018 shows that the following substances respond well to the treatment process, mostly below the detection limit: acesulfame, hydrochlorothiazide, ibuprofen-2-hydroxy, carbamazepine, carbamazepine E, lamotrigine, oxypurinol, progesterone, telmisartan, and tramadol. It should be noted, however, that raw water already had mostly relatively low concentration (mostly in tens of ng/dm 3 ; only acesulfame and oxypurinol often had input concentration about hundreds of ng/dm 3 ), which the treatment plant handled well. On the contrary, there is a group of substances for which there was no significant decrease in water processing at the treatment plant, despite the fact that the input concentration was often very low (tens of ng/dm 3 ): estrone, chloramphenicol, ibuprofen, paraxanthine, and sulfamerazine. There was also a relatively small decrease for azithromycin, gabapentin, ranitidine, and trimethoprim. Furthermore, it is alarming that the same substances tend to pass through the water reservoir (see above) over the water retention time (estrone, gabapentin, chloramphenicol, paraxanthine, ranitidine, sulfamerazine, and trimethoprim). The effect of the treatment plant cannot be commented on with regard to substances with concentrations in raw water already below the detection limit (from those discussed in the paper, e.g., diclofenac or metoprolol).
The overall quality of the produced drinking water regarding PPCP content can be assessed as high; if we do not consider substances with exceedance only in one analysis (out of 28 analyses), which is not conclusive, it can be stated that the following constituents occur with low frequency and in low concentration in the produced drinking water: ibuprofen (3 analyses out of 28, maximum 52 ng/dm 3 , detection limit 20 ng/dm 3 ), estrone (3 analyses of 6, maximum 5 ng/dm 3 , detection limit 1 ng/dm 3 ), acesulfame (2 analyses of 28, maximum 61 ng/dm 3 , detection limit of 50 ng/dm 3 ) and azithromycin (2 analyses of 28, maximum 28 ng/dm 3 , detection limit 10 ng/dm 3 ). There are no scientific sources that would challenge the safety of drinking water with such very low, and only occasional, concentration of PPCP substances (WHO [7,8], Godoy et al. [11], Bexfiled et al. [12], Chen et al. [16], RIVM [28], Kozisek et al. [38]). The existing standard raw water treatment is therefore fully sufficient for the residual concentrations observed. However, it should be pointed out that if the substances that readily pass through the treatment process in higher concentration appear in raw water, there could be a problem with the quality of the drinking water, and the inclusion of an additional water treatment stage consisting of activated carbon filters may prove to be necessary.

Conclusions
The research has brought several valuable findings. In terms of frequency of occurrence of the individual PPCP substances analyzed, a group of substances occurring very frequently, on most profiles and often at high concentration (thousands and even tens of thousands of ng/dm 3 on some profiles) can be identified: acesulfame, azithromycin, caffeine, gabapentin, hydrochlorothiazide, ibuprofen and its metabolites, oxypurinol, paraxanthine, and saccharin. Besides these, we can name other substances also occurring in the majority of analyses, but at a lower frequency and with lower concentrations (predominantly hundreds of ng/dm 3 ): iopromide, metoprolol, paracetamol, and tramadol. Furthermore, there is a group of 39 substances (out of the 112) that are always or almost always (in 95-100% of cases) below the detection limit of the analytical method used.
According to the detected maximum concentration of PPCP substances, highly polluted local streams can be identified (Hnevkovice, Kozli, Dolni Kralovice, and Bernartice) with an average concentration of the sum of PPCP substances in thousands of ng/dm 3 (and maximums for some substances up to tens of thousands ng/dm 3 -gabapentin, oxypurinol, paraxanthine). Of these, profiles Hnevkovice and Dolni Kralovice are discussed in detail in this paper. Moderately polluted are Miletin, Kacerov and Radikovice profiles with estimated average values around 1000 ng/dm 3 . The first two of these profiles are also discussed in detail in this paper. However, these are the profiles with the highest flow rates, so even though the concentration of PPCP substances does not reach the highest values, the PPCP mass flow from these profiles into the Svihov reservoir is the highest. Hulice and Brzotice profiles have the relatively lowest concentration of PPCP substances near the limit of detection in tens or exceptionally the first hundreds of ng/dm 3 .
The additional detailed monitoring of wastewater effluent from Hnevkovice WWTP on the one hand confirmed the results of the previous monitoring and also found high concentrations of some other substances present in the stream at much lower concentration (probably due to rapid disintegration), and on the other hand, provided indication of very high and rapid fluctuations in the quality of 'treated' wastewater (e.g., iopromide 73,600 ng/dm 3 in spring and 72 only ng/dm 3 in autumn, or sulfamethoxazole 505 ng/dm 3 in spring and 16,100 ng/dm 3 in autumn).
In the future, it is strongly recommended to ensure the optimal and not overloaded operation of all local WWTPs; in some cases, it may be necessary to consider their modernization or to supplement an additional treatment stage. Further research is needed on the migration parameters and degradation processes of PPCP substances, especially those that tend to remain in the natural aquatic environment for a long time and penetrate into the raw water, and in some rare cases, into drinking water (e.g., ibuprofen, estrone, acesulfame, azithromycin).

Supplementary Materials:
The following are available online at http://www.mdpi.com/2073-4441/12/5/1387/s1, Table S1: Chemical analyses. The funders had no role in the design of the study; in the collection, analyses, or interpretation of data; in the writing of the manuscript, or in the decision to publish the results.