The Inﬂuence of Aquaculture E ﬄ uents on the Prevalence and Biocides Resistance of Opportunistic Pseudomonas ﬂuorescens Bacteria in the Drw ˛eca River Protected under the Natura 2000 Network

: The aim of this study was to determine the e ﬀ ect of trout aquaculture e ﬄ uents on the Drw˛eca River. The count of opportunistic pathogen Pseudomonas ﬂuorescens (OPPF) in the total Pseudomonas ﬂuorescens population (TPFP) were determined by plating on King B medium and ﬂuorescence in situ hybridization. The resistance of OPPF strains to 12 antibiotics and two disinfectants was evaluated. Signiﬁcant di ﬀ erences ( p ≤ 0.05) in OPPF counts were found between seven sampling sites. OPPF counts were highest in samples collected directly downstream from three ﬁsh farms. More than 50% of these isolates demonstrated multiple-drug resistance to ampicillin, mezlocillin, cefotaxime, norﬂoxacin, tetracycline and two disinfectants (Steridial and chloramine T). Of these, 52% were resistant to high doses of cefotaxime and norﬂoxacin (MIC ≥ 256 µ g · mL − 1 ), and 65% were resistant to the maximum doses of Steridial (MIC Ste ≥ 25 mL · m − 3 ) and chloramine T (MIC Chlor ≥ 20 mg · L − 1 ). All OPPF sampled upstream from the farms were sensitive to low concentrations of CTX (cefotaxime) and NOR (norﬂoxacin) (MIC ≤ 2 µ g · mL − 1 ), Steridial (MIC Ste ≤ 5 mL · m − 3 ) and chloramine T (MIC Chlor ≤ 2.5 mg · L − 1 ). Agglomerative clustering revealed two clusters: strains from samples collected upstream and downstream from trout farms. The results indicate that aquaculture e ﬄ uents signiﬁcantly a ﬀ ect the prevalence of biocides resistant OPPF along the river continuum. 3.6% in farms Czarci sites IV Student’s t -test revealed signiﬁcant ( p = 0.000) di ﬀ erences in OPPF and TPFP counts only between water samples collected in sites I and II in the upper course of the river, and upstream from ﬁshing farm No. in Czarci Jar.


Introduction
Pseudomonas fluorescens bacteria are widespread in natural and anthropogenically altered water environments. They form the natural microbiota of ecosystems in which they play an important role in the decomposition of organic matter, circulation of elements and self-purification processes [1,2]. Some P. fluorescens strains show hemolytic activity and are classified as opportunistic P. fluorescens pathogens. These strains could be an important causative agent of epizootic diseases in fish, other wild animals, livestock and, consequently, humans [3][4][5]. Higher concentrations of potentially pathogenic strains in water bodies increases the risk of infection for aquatic organisms and contributes to their spread not only in aquatic ecosystems, but also among humans, livestock and wildlife [6,7].
The effluents evacuated from trout aquacultures into surface water bodies are potential sources of transmission of opportunistic strains that increase microbial counts in aquatic environments. Intensive trout farming increases the contamination of aquatic ecosystems that act as receptacles of effluents [8][9][10][11]. The use of high-protein fish feed to maximize weight gains in trout aquaculture promotes the growth and development of various groups, genera and species of bacteria that colonize water and fish. These microorganisms include beneficial microbiota that improve the welfare of fish, as

Study Area and Sampling Sites
The Drwęca River, a right-bank tributary of the Vistula River, is a strongly meandering lakeland river with a length of 207.2 km and an average slope of 0.41% . The river channel is highly diverse. The entire river belongs to the Drwęca River Nature Reserve, which was established to protect the aquatic environment and fish populations, in particular trout, salmon, sea trout and vimba bream habitats [24]. Drwęca is the longest ichthyological reserve in Poland with an area of 444.38 ha, and it belongs to the Natura 2000 network of protected areas in the European Union established under the Habitats Directive (92/43/EEC) [23]. The Drwęca River Valley (code PLH280001) covers an area of 2369.6 ha. The major sources of point pollution reaching the Drwęca River include household and sewage and wastewater as well as post-production water. This unique ecosystem is directly affected by surface runoffs carrying pollutants from the river's agricultural catchment area and outflows from three fish farms [14].
Samples for microbiological analyses were collected along a 15-kilometer-long segment in the upper course of the Drwęca River, which receives effluents from three moderately intensive, commercial rainbow trout farms ( Figure 1): a Broodstock Farm in Czarci Jar, and two salmonid farms in Rychnowska Wola. In the studied farms, salmonids are fed complete pelleted diets and are reared in natural and concrete ponds. The above implies that all essential nutrients are supplied with feed. Despite the fact that all facilities are properly maintained and operated, they pose the greatest risk of environmental pollution in this section of the river [25,26].
Effluents from the studied fish farms are discharged directly to the Drwęca River, which is why water samples were collected from seven sampling sites in specific locations on the river ( Figure 1): Water samples were collected in 2005-2006 at six-week intervals, at a depth of 0.3-0.5 m in the center of the river channel along the examined section of the Drwęca River. A total of 56 water samples were collected during the study.

Total Pseudomonas fluorescens Population (TPFP) in the Drwęca River
The quantitative composition of total P. fluorescens populations (TPFP) was determined by plating water samples on King B medium and incubating the samples at 30 ± 2 °C for 48-72 h according to APHA guidelines [27]. All determinations were performed in three replicates in the same sample of water. After incubation, colonies showing fluorescence under ultraviolet light were counted and passaged three times on King B medium. P. fluorescens strains were identified by fluorescence in situ hybridization (FISH). Twenty-four-hour broth cultures were fixed with neutralized formaldehyde (pH 7.4, final concentration of 4%) for 4-6 h at room temperature. The fixed cultures were gently passed through polycarbonate filters (0.2 µm pore size, 47 mm in diameter) (GTTP, Millipore) and rinsed with 10 mL of ultrapure water (MQ Millipore). The filters were dried for several minutes at room temperature. The cultures fixed on filters were subjected to the FISH procedure with the use of a Pseudo120 probe (Thermo Fisher Scientific GmbH, Ulm, Baden-Württemberg, Germany) as the specific oligonucleotide probe for the identification of P. fluorescens bacteria [28]. The subsequent steps of the FISH protocol (including hybridization conditions) and the microscopic quantification of P. fluorescens isolates were described previously

Total Pseudomonas fluorescens Population (TPFP) in the Drwęca River
The quantitative composition of total P. fluorescens populations (TPFP) was determined by plating water samples on King B medium and incubating the samples at 30 ± 2 • C for 48-72 h according to APHA guidelines [27]. All determinations were performed in three replicates in the same sample of water. After incubation, colonies showing fluorescence under ultraviolet light were counted and passaged three times on King B medium. P. fluorescens strains were identified by fluorescence in situ hybridization (FISH). Twenty-four-hour broth cultures were fixed with neutralized formaldehyde (pH 7.4, final concentration of 4%) for 4-6 h at room temperature. The fixed cultures were gently passed through polycarbonate filters (0.2 µm pore size, 47 mm in diameter) (GTTP, Millipore) and rinsed with 10 mL of ultrapure water (MQ Millipore). The filters were dried for several minutes at room temperature. The cultures fixed on filters were subjected to the FISH procedure with the use of a Pseudo120 probe (Thermo Fisher Scientific GmbH, Ulm, Baden-Württemberg, Germany) as the specific oligonucleotide probe for the identification of P. fluorescens bacteria [28]. The subsequent Water 2020, 12, 1947 5 of 15 steps of the FISH protocol (including hybridization conditions) and the microscopic quantification of P. fluorescens isolates were described previously by Gołaś et al. [29]. The fluorescence signal of the isolates hybridized with the Cy3-labeled Pseudo120 probe was viewed under a fluorescence microscope (Olympus BX61) equipped with a × 100 oil immersion lens (Olympus Poland, Warsaw, Masovia, Poland), a UV lamp (Olympus Poland, Warsaw, Masovia, Poland), DAPI and CY3 filters (Olympus Poland, Warsaw, Masovia, Poland) and a CCD camera (Olympus) (Olympus Poland, Warsaw, Masovia, Poland). The images were analyzed in the Cell Sence Dimension 2 application (Olympus).
The number of fluorescein-producing colonies that were hybridized with the Pseudo120 probe in the FISH assay [28] were taken into account in the determination of total P. fluorescens populations (TPFP) and expressed as colony-forming units (cfu) per 1 mL of the analyzed water samples. A total of 200 TPFP strains were isolated from 56 water samples and used in subsequent stages of research.

Counts of Opportunistic Pathogens P. fluorescens (OPPF) in the Drwęca River
The counts of OPPF in the 200 isolated TPFP strains were determined based on their proteolytic and hemolytic activity [30] and expressed quantitatively by the R-factor, according to the formula given by Hsu et al. [31]: R = total area clear zone (mm)/colony area (mm) The hemolytic and proteolytic activity of the examined strains was determined according to the methods proposed by Chang et al. [32] and Larsen et al. [33], respectively. The isolates exhibiting high hemolytic and proteolytic activity at R ≥ 4 were classified as OPPF. The analysis produced 63 OPPF strains, which were then tested for resistance to commonly used antibiotics and disinfectants.

Antibiotic Resistance of Opportunistic Pathogens P. fluorescens (OPPF)
The antibiotic resistance of 63 OPPF isolates was analyzed by the disc diffusion method on Mueller-Hinton medium [34]. Twelve popular antibiotics belonging to five classes were used in the analysis: penicillins, fluoroquinolones, sulfonamides, aminoglycosides and tetracyclines. The following standard therapeutic doses of the evaluated antibiotics were used: piperacillin (PRL, 100 µg), ceftazidime (CAZ, 30 µg), ampicillin (AMP, 30 µg), mezlocillin (MEZ, 75 µg), cefotaxime (CTX, 30 µg), ciprofloxacin (CIP, 5 µg), norfloxacin (NOR, 10 µg), cotrimoxazole (SXT, 25 µg), gentamicin (CN, 10 µg), tobramycin (TOB, 10 µg), tigecycline (TGC, 15 µg) and tetracycline (TE, 30µg). Twenty-four-hour cultures of bacterial strains were suspended in 0.85% saline solution to adjust suspension turbidity to 0.5 McFarland. The inocula were plated on Mueller-Hinton agar. After 30 min, discs saturated with antibiotics were applied onto the plates with a dispenser. The plates with antibiotic discs were incubated at 30 • C. The antimicrobial susceptibility of the strains was tested according to the guidelines of the Clinical and Laboratory Standards Institute [35]. The multi-drug resistance of the strains from each sampling site was assessed using the multiple antibiotic resistance (MAR) index. The MAR was calculated based on the resistance of OPPF isolates to at least two antibiotics from different groups, according to the method described by Krumperman [36]. According to Paul et al. [37], natural environments where antibiotics are present sporadically or absent are characterized by MAR ≤ 0.2, whereas environments with MAR > 0.2 are at a high risk of antibiotic exposure.

Minimum Inhibitory Concentrations (MIC) of Antibiotics
Resistance to standard doses of cefotaxime (CTX, 30 µg) and norfloxacin (NOR, 10 µg) was most commonly observed in the 63 examined OPPF isolates. Therefore, all OPPF were analyzed for minimum inhibitory concentrations of cefotaxime (MIC SXT) and norfloxacin (MIC NOR) by the gradient agar dilution method [38]. The MIC values of STX and NOR were determined at the following antibiotic concentrations: 2 µg·mL −1 , 4 µg·mL −1 , 8 µg·mL −1 , 16 µg·mL −1 , 32 µg·mL −1 , 64 µg·mL −1 , 128 µg·mL −1 , 256 µg·mL −1 and 512 µg·mL −1 . Bacterial cultures were steaked onto plates with different concentrations of CTX and NOR. The plates were incubated at 28 • C for 48-72 h, and bacterial growth was examined. The isolates that formed colonies visible to the naked eye were regarded as resistant to a given antibiotic concentration. OPPF strains that did not grow were regarded as sensitive to the tested dose of the antibiotic.

Resistance of Opportunistic Pathogens P. fluorescens (OPPF) to Disinfectants
The efficacy of the standard recommended doses of Steridial (Ste) and chloramine T (Chlor) against 63 OPPF strains was analyzed. According to Terech-Majewska et al. [21], the recommended concentrations of Steridial and chloramine T for the prevention and treatment of fish diseases range from 5 mL·m −3 to 25 mL·m −3 of water, and from 2.5 mg·L −1 to 20 mg·L −1 , respectively. The minimum inhibitory concentrations of both disinfectants (MIC Ste and MIC Chlor) were determined by the agar gradient dilution method [38]. The following concentrations were used to determine MIC Ste: 5 mL·m −3 , 10 mL·m −3 , 15 mL·m −3 , 20 mL·m −3 and 25 mL·m −3 . MIC Chlor for the studied isolates was tested at following concentrations: 2.5 mg·L −1 , 5.0 mg·L −1 , 10.0 mg·L −1 and 20.0 mg·L −1 . The resistance and sensitivity of OPPF isolates to specific concentrations of the tested biocides were determined with the use of the procedure described for MIC CTX and NOR (Section 2.2.4).

Statistical Analysis
Statistical analyses were carried out using the Statistica 13.3 software package (TIBCO Software Inc., Palo Alto, USA). One-way analysis of variance (ANOVA) was used to determine the significance of differences (p ≤ 0.05) in: (i) the abundance of TPFP and OPPF between the analyzed sampling sites, (ii) OPPF counts in isolates with specific MIC CTX, NOR, Ste and Chlor values and in river water from different sampling sites. The homogeneity of variances was determined by Levene's test. When Levene's test produced statistically significant results, the verified hypotheses were rejected. The Kruskal-Wallis test, a non-parametric version of classical one-way ANOVA, was then applied [39].
The Student's t-test was used to determine statistically significant differences (p ≤ 0.05) between the counts of total P. fluorescens populations (TPFP) and of OPPF in samples of water from different sampling sites.
The similarity of sampling sites was determined by cluster analysis with the Bray-Curtis distance measure based on the counts of bacteria with specific MIC CTX, NOR, Ste and Chlor values using PC_ORD 6.08 software. The matrix of agglomerative tree of hierarchical grouping was generated by using the elastic beta method (β = −0.250).

Counts of Opportunistic Pathogens P. fluorescens (OPPF) in the Drwęca River
The average counts of OPPF in the upper course of the Drwęca River ranged from several to several dozen cfu·mL −1 ( Table 1). The lowest counts (1 ± 1 cfu·mL −1 ) were noted in site I located at a distance of around 2 km from the river source, and the highest counts (48 ± 18 cfu·mL −1 and 56 ± 18 cfu·mL −1 ) were observed in water samples collected in sites V and VI by the effluent discharge points of fish farms No. 2 and 3 in Rychnowska Wola, respectively. Significant (p = 0.000) differences in the abundance of TPFP and OPPF between samples collected in different sites were confirmed by the Kruskal-Wallis test. Changes were also observed in the average percentage of OPPF in TPFP along the course of the river. The percentage of OPPF ranged from 11.2 ± 9.4% and 11.5 ± 7.2% in samples collected in the upper course of the river (sites I and II, respectively) to 84.8 ± 3.8%, 86.5 ± 8.0% and 89.5 ± 3.6% in samples collected directly downstream from fish farms (site III in Czarci Jar and sites IV and V in Rychnowska Wola, respectively). However, Student's t-test revealed significant (p = 0.000) differences in OPPF and TPFP counts only between water samples collected in sites I and II in the upper course of the river, and upstream from fishing farm No. 1 in Czarci Jar.

Resistance of Opportunistic Pathogens P. fluorescens (OPPF) to the Examined Antibiotics and Disinfectants
Resistance to standard doses of CTX and NOR was most frequently observed in the 63 OPPF strains isolated from the Drwęca River (Table 2). In the tested group of isolates, in 71.4% (45 isolates) were resistant to CTX and 52.4% (33 isolates) were resistant to NOR. Twenty-nine strains (46.0%) were resistant to AMC, MEZ and TE. Resistance to the remaining chemotherapeutics was observed in eight to 23 strains (12.6-36.5%). The highest number of OPPF showing resistance to most of the analyzed antibiotics were isolated from water sampled directly downstream from three fish farms (sites III, V and VI). These isolates demonstrated multi-antibiotic resistance (MAR index) and were not sensitive to at least two antibiotics belonging to different groups. The nominal values of MAR were highest in OPPF strains isolated from the samples collected in sites V and VI directly downstream from the fish farms in Rychnowska Wola (MAR 0.47 and 0.51, respectively) relative to the isolates from sites I and II in the upper course of the river (MAR 0.03 and 0.04, respectively).
An analysis of the biocidal efficacy of disinfectants revealed that 100% of the isolates (nine strains) from samples collected in the upper course of the river (sampling sites I and II) were sensitive to the minimum recommended doses of Steridial (5 mL·m −3 ) and chloramine T (2.5 mg·L −1 ) ( Table 2). Most OPPF strains isolated from water samples collected downstream from fish farms (sites III, V, VI, VII) were resistant to the recommended doses of Steridial (5-25 mL·m −3 ) and chloramine T (2.5-20 mg·L −1 ).

Minimum Inhibitory Concentrations (MIC) of Antibiotics and Disin fectants
The counts of OPPF showing resistance to the minimum inhibitory concentrations (MIC) of antibiotics (cefotaxime-CTX and norfloxacin-NOR) and disinfectants (Steridial-Ste and chloramine T-Chlor) differed between the tested bactericide and sampling sites ( Figure 2). All OPPF isolated from samples collected in sites I and II in the upper course of the Drwęca river were characterized by the lowest values of MIC CTX and MIC NOR (≤2 µg·mL −1 ). Significantly higher MIC, CTX and NOR values (from 32 to ≥ 512 µg·mL −1 ) were noted in strains from water samples collected directly downstream from the fish farms in Czarci Jar and Rychnowska Wola (sites III, V and VI) and around 6 km from the fish farms (site VII). MIC CTX and NOR value of 64 µg·mL −1 was found in 6-20% and 8-33% of these isolates, respectively. Resistance to MIC CTX and NOR values of 128-256 µg·mL −1 was observed in 25-40% and 14-20% strains, respectively. More than 52% of the 40 OPPF isolates from sites III, V and VI were characterized by MIC, NOR and CTX values of ≥256 µg·mL −1 (Figure 2A,B).
All strains from water samples collected in sites I and/or II in the upper course of the Drwęca River were sensitive to the minimum recommended doses of Steridial (5 mL·m −3 ) and chloramine T (2.5 µg·L −1 ). In contrast, the OPPF isolates from the remaining sampling sites were resistant to higher doses of both disinfectants. Out of the 40 strains isolated from water samples collected in sites III, V and VI, resistance to the maximum recommended doses of Steridial (25 mL·m −3 ) and chloramine T (20 µg·L −1 ) was observed in 72-86% and 65-92% of the isolates, respectively ( Figure 2C,D).
The Kruskal-Wallis test (N = 63, p ≤ 0.05) confirmed differences in the counts of OPPF resistant to the minimum inhibitory concentrations (MIC) of the tested antibiotics and disinfectants between sampling sites. The differences in the counts of OPPF resistant to cefotaxime (CTX), norfloxacin (NOR), Steridial (Ste) and chloramine T (Chlor) were statistically significant at p = 0.0059, p = 0.0035, p = 0.0044 and p = 0.0092, respectively.
Water 2020, 12, x FOR PEER REVIEW 9 of 15

Minimum Inhibitory Concentrations (MIC) of Antibiotics and Disin fectants
The counts of OPPF showing resistance to the minimum inhibitory concentrations (MIC) of antibiotics (cefotaxime-CTX and norfloxacin-NOR) and disinfectants (Steridial-Ste and chloramine T-Chlor) differed between the tested bactericide and sampling sites ( Figure 2). All OPPF isolated from samples collected in sites I and II in the upper course of the Drwęca river were characterized by the lowest values of MIC CTX and MIC NOR (≤2 µg·mL −1 ). Significantly higher MIC, CTX and NOR values (from 32 to ≥ 512 µg·mL −1 ) were noted in strains from water samples collected directly downstream from the fish farms in Czarci Jar and Rychnowska Wola (sites III, V and VI) and around 6 km from the fish farms (site VII). MIC CTX and NOR value of 64 µg·mL −1 was found in 6-20% and 8-33% of these isolates, respectively. Resistance to MIC CTX and NOR values of 128-256 µg·mL −1 was observed in 25-40% and 14-20% strains, respectively. More than 52% of the 40 OPPF isolates from sites III, V and VI were characterized by MIC, NOR and CTX values of ≥256 µg·mL −1 (Figure 2A,B).
All strains from water samples collected in sites I and/or II in the upper course of the Drwęca River were sensitive to the minimum recommended doses of Steridial (5 mL·m −3 ) and chloramine T (2.5 µg·L −1 ). In contrast, the OPPF isolates from the remaining sampling sites were resistant to higher doses of both disinfectants. Out of the 40 strains isolated from water samples collected in sites III, V and VI, resistance to the maximum recommended doses of Steridial (25 mL·m −3 ) and chloramine T (20 µg·L −1 ) was observed in 72-86% and 65-92% of the isolates, respectively ( Figure  2C,D).
The Kruskal-Wallis test (N = 63, p ≤ 0.05) confirmed differences in the counts of OPPF resistant to the minimum inhibitory concentrations (MIC) of the tested antibiotics and disinfectants between sampling sites. The differences in the counts of OPPF resistant to cefotaxime (CTX), norfloxacin (NOR), Steridial (Ste) and chloramine T (Chlor) were statistically significant at p = 0.0059, p = 0.0035, p = 0.0044 and p = 0.0092, respectively. The similarity of the sampling sites was evaluated by agglomerative clustering analysis with the Bray-Curtis dissimilarity based on the counts of OPPF with specific MIC CTX, NOR, Ste and Chlor values, and it produced two different clusters (Figure 3). The first cluster consisted of strains isolated from water samples collected in sites I and II in the upper course of the river. The second cluster consisted of isolates from sites III-VII which formed a network of interconnected sub-clusters. In the second group, OPPF strains isolated from the effluents discharged by fish farms in Rychnowska Wola (sites V and VI) as well as in samples collected directly downstream from fish farm in Czarci Jar and at a distance of around 6 km downstream from fish farms No. 2 and 3 (sites III and VII) were the most similar. The samples collected in site IV around 12 km from the effluent discharge points of fish farm in Czarci Jar (site III) were least similar in terms of OPPF counts with specific MIC CTX, NOR, Ste and Chlor values. The similarity of the sampling sites was evaluated by agglomerative clustering analysis with the Bray-Curtis dissimilarity based on the counts of OPPF with specific MIC CTX, NOR, Ste and Chlor values, and it produced two different clusters (Figure 3). The first cluster consisted of strains isolated from water samples collected in sites I and II in the upper course of the river. The second cluster consisted of isolates from sites III-VII which formed a network of interconnected sub-clusters. In the second group, OPPF strains isolated from the effluents discharged by fish farms in Rychnowska Wola (sites V and VI) as well as in samples collected directly downstream from fish farm in Czarci Jar and at a distance of around 6 km downstream from fish farms No. 2 and 3 (sites III and VII) were the most similar. The samples collected in site IV around 12 km from the effluent discharge points of fish farm in Czarci Jar (site III) were least similar in terms of OPPF counts with specific MIC CTX, NOR, Ste and Chlor values.

Discussion
Bacteriological contamination of rivers poses a significant problem by increasing the epidemiological risk associated with the presence of opportunistic or pathogenic bacterial strains in water. The above applies to isolates characterized by high hemolytic activity which induce cytotoxic and proinflammatory responses in epithelial intestinal cells [40] and are resistant to many commonly used antibiotics [41]. This group of bacteria also includes opportunistic pathogens P. fluorescens, which grow rapidly in aquatic environments exposed to pressure from fish aquacultures [5,22]. These observations were validated by the present study which investigated the prevalence of OPPF in the upper course of the Drwęca River. The catchment of study area of the Drwęca River

Discussion
Bacteriological contamination of rivers poses a significant problem by increasing the epidemiological risk associated with the presence of opportunistic or pathogenic bacterial strains in water. The above applies to isolates characterized by high hemolytic activity which induce cytotoxic and proinflammatory responses in epithelial intestinal cells [40] and are resistant to many commonly used antibiotics [41]. This group of bacteria also includes opportunistic pathogens P. fluorescens, which grow rapidly in aquatic environments exposed to pressure from fish aquacultures [5,22]. These observations were validated by the present study which investigated the prevalence of OPPF in the upper course of the Drwęca River. The catchment of study area of the Drwęca River was dominated by forest, agricultural areas and small villages [14]. This river ecosystem was directly affected by surface runoffs carrying pollutants from the river's agricultural catchment. Nevertheless, the results of my studies showed the spatial distribution of the studied OPPF isolates differed across sampling sites which were separated by varied distance from trout farms and effluent discharge points on the Drwęca River. These findings were confirmed by the Kruskal-Wallis test, which revealed significant differences (p = 0.0000) in bacterial counts between sampling sites. The observed transmission of studied opportunistic strains along river course suggests the major point sources for this unique ecosystem were the effluents evacuated from three trout aquacultures.
OPPF counts (0-5 cfu·mL −1 ) and their percentage in TPFP (11.2-11.5%) were lowest in the samples collected in sites I and II in the upper course of the Drwęca River. OPPF counts and their percentage in TPFP increased with distance from the river source. In water samples collected directly downstream from fish farms, OPPF counts were highest (30-76 cfu·mL −1 ) and their proportions in TPFP exceeded 80%. In total, more than 63% of the strains from all tested OPPF isolates with high hemolytic and proteolytic (R ≥ 4) activity were found in water samples from sites III, V and VI directly downstream from the effluent discharge points of the fishing farms in Czarci Jar and Rychnowska Wola. The highest OPPF counts noted in water of samples collected in V and VI downstream from the fishing farms in Rychnowska Wola corresponded with an increase of river flow rate and some physicochemical water parameters observed in study of the Drwęca River [26]. According to Teodorowicz [26] the flow water rates ranged from 0.283 to 0.393 m 3 ·s −1 at sites situated behind fish farms no. 2 and 3 in Rychnowska Wola. Equally, the highest concentrations of BOD 5 (3.2 ± 1.7 mg O 2 ·L −1 ), total phosphorus (0.20 ± 0.07 mg P·L −1 ) and ammonia nitrogen (0.18 ± 0.08 mg NH4-N·L −1 ) noted in water samples collected downstream from the fishing farms in Rychnowska Wola could have conditioned the growth the OPPF. These findings suggest that effluents from trout aquaculture directly contribute to the prevalence of OPPF in Drwęca and increase the local risk of infection for humans, fish and other living organisms.
The results of the analysis investigating the sensitivity of opportunistic pathogen P. fluorescens strains to the most commonly used drugs and disinfectants confirmed that intensive fish farming increases the epidemiological risk in the upper course of Drwęca. Most OPPF isolates were resistant to cefotaxime (71%) and norfloxacin (52%), while 46% of the isolates were resistant to ampicillin (AMC), mezlocillin (MEZ) and tetracycline (TE). Unfortunately, the information regarding the specific types and the amounts of antibiotics used in the three local fish farms over the study period of Drwęca River has not been publicly available. Nevertheless, some literature data [42,43] suggest a correlation between findings of increased bacterial resistance levels within and around inland fish farms and the antimicrobial agents used at the farms. On the other hand, Schmidt et al. [44] showed that a simple overall correlation between antibiotic usage and emergence of bacterial antibiotic resistance was not often evident, presumably because the sampling strategy was not always correlated with clinical outbreaks of disease and antibiotic therapy at the farms. According to Pirnay et al. [45], Pseudomonas species are naturally resistant to the penicillins, cephems and rifampin because they have relatively impermeable membranes, inducible efflux systems and chromosomally encoded inducible β-lactamase. However, antibiotic overuse in treatment and prevention increases drug-resistance also in P. fluorescens strains [6,7,22]. This observation is supported by the present findings and literature data, which point to different antibiotic resistance profiles of opportunistic pathogens P. fluorescens isolated from various aquatic and aquaculture environments. Research into Pseudomonas spp. isolated from 400 samples collected from 100 diseased fish demonstrated that 75% of the isolates were resistant to standard doses of cefotaxime, but were sensitive to norfloxacin, gentamicin, enrofloxacin, norfloxacin, ciprofloxacin, florfenicol, doxycycline, sulfa-trimethoprim, oxytetracycline, nalidixic acid and streptomycin [5]. Kumar and Surendran [13] analyzed the antibiotic resistance profiles of P. fluorescens isolated from brackish water and fresh water aquacultures and found that 100% of P. fluorescens isolates were resistant to oxytetracycline, nitrofurantoin, chloramphenicol and sulfamethizole. However, other studies of fish colonized by P. fluorescens isolated from fresh and brackish water demonstrated that oxytetracycline was highly effective against Pseudomon as species, whereas ciprofloxacin was less effective [46].
Among the 63 studied OPPF strains from the Drwęca River, 52% isolates were characterized by multi-drug resistance and were not sensitive to at least two antibiotics from different groups. These strains were detected in sites III, V and VI downstream from the fishing farms in Czarci Jar and Rychnowska Wola in different seasons of the year. However, these samples from different seasons did not present variety in enumeration results of multiple antibiotic resistant OPPF between water samples collected directly downstream from fish farms in Czarci Jar and Rychnowska Wola. The values of the MAR index were higher (0.27-0.51) in the strains isolated from the above sites than in the isolates sampled from sites I, II, IV and VII. The OPPF strains from sites I and II were characterized by the lowest MAR values that did not exceed 0.04, whereas the MAR values of the strains from the samples collected at a distance of around 6 km from the fish farms (sites IV and VII) were several times higher at 0.20 and 0.45, respectively. These observations indicate that most strains isolated from sites III to VII originated from high risks sources of contamination where antibiotics are often used [37]. An increase in the MAR value of strains isolated from successive sampling sites on the Drwęca River points to the spread of multi-drug resistant OPPF along the river continuum.
An analysis of the sensitivity of OPPF isolates to the minimum inhibitory concentrations (MIC) of cefotaxime and norfloxacin revealed that all strains isolated from sampling sites I and II in the upper course of the river were characterized by the lowest MIC CTX and NOR values (≥2 µg·mL −1 ). In turn, more than 50% of the isolates from water samples collected directly downstream from the fish farms (sites III, V and VI) were characterized by very high MIC CTX and NOR values (≥256 µg·mL −1 ). The MIC CTX and NOR values of the tested isolates were correlated with the MIC values of the two tested disinfectants. An analysis of the biocidal efficacy of Steridial (Ste) and chloramine T (Chlor) against OPPF revealed that all strains isolated from sites I and II were also sensitive to the lowest recommended doses of Ste (MIC ≤ 5 mL·m −3 ) and Chlor (MIC ≤ 2.5 mg·L −1 ). In turn, most OPPF isolates from water samples collected directly downstream from the fish farms (sites III, V and VI) were resistant to the maximum recommended doses of both disinfectants (MIC Ste = 25 mL·m −3 , MIC Chlor = 20 mg·L −1 ). The observed differences in the counts of OPPF isolates with specific MIC CTX, NOR, Ste and Chlor values between sampling sites were confirmed by the Kruskal-Wallis test (p ≤ 0.05) and the agglomerative clustering analysis. Two different clusters were identified based on the counts of OPPF characterized by specific values of MIC STX, NOR, Ste and Chlor. The first cluster consisted of OPPF strains isolated from water samples collected in sites I and II in the upper course of the river, and the second cluster consisted of isolates from sites located directly downstream or at a distance of 6 or 12 km from the examined fish farms.

Conclusions
The results of this study indicate that aquaculture effluents exert a significant effect on the concentrations of OPPF and their sensitivity to biocides in the upper course of the Drwęca River, which is a nature and ichthyological reserve. The effluents discharged directly from three fish farms into the analyzed fluvial ecosystem increased in the counts of the examined opportunistic pathogens P. fluorescens (OPPF) and their proportions in the total P. fluorescens population (TPFP) in water samples collected downstream from the investigated fish farms. An analysis of the resistance of OPPF isolates to 12 commonly used antibiotics and two disinfectants confirmed that aquaculture effluents increased the local risk of infection for humans, fish and other living organisms. Most OPPF strains isolated from water samples collected directly downstream from the fish farms were not only resistant to multiple antibiotics (MAR > 0.20), but also to high doses of cefotaxime and norfloxacin (MIC ≥ 256 µg·mL −1 ). Most OPPF isolates were also highly resistant to the maximum recommended doses of Steridial (MIC ≥ 25 mL·m −3 ) and chloramine T (MIC ≥ 20 mg·L −1 ), which are commonly used in the treatment and prevention of fish diseases in various aquaculture systems. The increase in the abundance of OPPF strains and their resistance to antibiotics and disinfectants (MAR, MIC) in samples of waters from successive sampling sites on the Drwęca River indicates that aquaculture effluents contribute to the spread of multi-drug and disinfectants resistant OPPF along the river continuum.