River water samples were taken from 32 locations around the basin of Chaophraya River and its four major tributaries in Thailand to investigate resistance ratios of Escherichia coli isolates to eight antibiotic agents of amoxicillin, sulfamethoxazole/trimethoprim, tetracycline, doxytetracycline, ciprofloxacin, levofloxacin, norfloxacin and ofloxacin. Principal component analysis was performed to characterize resistance patterns of the samples. Relevancy of the obtained principal components with urban land use and fecal contamination of the river were examined. The ratio of antibiotic-resistant bacteria is likely to increase when urban land use near the sampling site exceeds a certain ratio. The resistance ratio to fluoroquinolones tends to be high in a highly populated area. Meanwhile, no significant contribution of fecal contamination was found to increase the resistance ratio. These results suggest that an antibiotic-resistance ratio is dependent on conditions of local urbanization rather than the upstream conditions, and that the major sources of antibiotic-resistant bacteria in the Chaophraya River basin are possibly point sources located in the urban area which contains a high ratio of resistant bacteria.

INTRODUCTION

Antibiotic agents are widely found in aquatic environments, from sources such as hospitals, urban sewage, livestock farming, and aquaculture (Halling-Sørensen et al. 1998; Sarmah et al. 2006). The release of antibiotics into the environment leads to the proliferation of antibiotic-resistant bacteria, owing to the selective pressure on the microorganism community (Baquero et al. 2008; Dodd 2012). In some Asian countries, for example Thailand, Cambodia and Vietnam, a wide variety of antibiotics are available over-the-counter at pharmacies without prescription; in countries such as the United States, Japan etc., their use is more tightly controlled (Udomwiboonchai et al. 2007). Animal husbandry and aquaculture in those Asian countries are also concerned as a source of antibiotic agents (Suzuki & Hoa 2012; Shimizu et al. 2013; Boonyasiri et al. 2014; Rico et al. 2014). The release of antibiotic agents and antibiotic-resistant bacteria into aquatic environments is greater in developing countries because of lower coverage of wastewater treatment systems. In the Chaophraya Delta region in Thailand, characterized by a dense canal network, rivers and canals are used for irrigation and transportation purposes, especially in rural and suburban areas (Honda et al. 2010). The presence of antibiotic-resistant bacteria in surface waters poses a health risk to the people who use water from such rivers and canals. Past studies in Southeast Asia have reported the presence of antibiotic-resistant bacteria (Kenzaka et al. 2006; Takasu et al. 2011; Boonyasiri et al. 2014), antibiotic-resistant genes (Kobayashi et al. 2007; Phuong Hoa et al. 2008; Suzuki et al. 2008; Takasu et al. 2011) and antibiotics in rivers (Le & Munekage 2004; Managaki et al. 2007; Takasu et al. 2011; Hoang et al. 2012). Most of the studies on antibiotic-resistant bacteria in the water environment in Southeast Asia targeted small, local areas, whereas the distribution of antibiotic-resistant bacteria is likely to vary depending on urbanization and potential sources of contamination in the areas surrounding the water systems. Several catchment-scale studies were reported on river basins in China and the United States, which performed characterization of river water samples by multiple antibiotic resistance and resistance gene profiles (Storteboom et al. 2010; Tao et al. 2010; Su et al. 2012). There are few studies which distinctly discussed the relationship of urban development with prevalence of antibiotic-resistant bacteria in water environment, which receives antibiotics and resistant bacteria from various sources. Our previous study reported that resistant bacteria including those with multiple-resistance prevailed in a wide area in Chaophraya River and its four major tributaries in Thailand by investigating susceptibility of almost 400 Escherichia coli isolates (Honda et al. 2011). In this study, with the aim of revealing impacts of urbanization in developing countries on prevalence of antibiotic-resistant bacteria in water environment, 32 water samples from Chaophraya River and its four major tributaries were characterized based on resistance ratios to eight antibiotic agents, their relevancy with urban land use and fecal contamination was examined.

MATERIALS AND METHODS

Sampling

River water samples taken in the previous study were used in this study (Honda et al. 2011). Samples were collected from 32 sites on the Chaophraya River and its four tributaries, the Ping, Wang, Yom and Nan rivers (Figure 1; Table 1). We selected the sampling site where the Ministry of Natural Resources and Environment, Thailand regularly monitored. Samples were collected from 17–28 January 2011 in the dry season of the target area to avoid influence of rainfall. No precipitation was observed during sampling.
Table 1

List of sampling sites and the number of isolates tested

Sampling sites
Distance to Chaophraya River mouth [km]E. coli concentration [CFU/mL]Suspended solidsn1n2
IDName[mg/L]
 Chaophraya River         
CH06 27.50 70 17 11 16 
CH08 42.72 11 13 
CH10 48.77 32 7.8 18 22 
CH12 58.77 17 7.6 17 19 
CH15 84.06 4.7 20 13 13 
CH16 96.38 2.5 14 14 13 
CH18 125.09 1.7 15 12 16 
CH19 129.70 6.1 21 14 16 
CH20 144.05 4.7 19 10 10 
10 CH21 184.97 4.3 24 
11 CH22 213.08 2.6 19 11 16 
12 CH24 229.19 5.5 20 20 23 
13 CH25 246.78 4.1 21 15 20 
14 CH26 267.39 3.9 19 12 17 
15 CH28 287.90 1.8 26 17 21 
16 CH30 332.86 5.4 61 
17 CH31 354.68 8.1 51 
  Nan River         
18 NA02 442.41 4.8 63 
19 NA07 582.73 1.8 44 10 
20 NA11 733.55 1.9 
21 NA13 965.74 1.9 10 
  Ping River         
22 PI01 385.59 1.9 58 
23 PI03 442.10 2.8 36 
24 PI04 510.90 5.2 13 
25 PI08 581.43 1.6 11 13 
26 REO1-i5 1225.56 2.0 0 
27 PI12 1246.84 42 48 12 
28 REO1-i2 1256.02 4.0 42 0 
29 PI14 1300.75 5.2 34 3 
  Wang River         
30 WA01 643.84 0.47 10 
  Yom River         
31 YO01 459.72 1.5 29 13 
32 YO04 566.76 5.4 27 11 14 
Sampling sites
Distance to Chaophraya River mouth [km]E. coli concentration [CFU/mL]Suspended solidsn1n2
IDName[mg/L]
 Chaophraya River         
CH06 27.50 70 17 11 16 
CH08 42.72 11 13 
CH10 48.77 32 7.8 18 22 
CH12 58.77 17 7.6 17 19 
CH15 84.06 4.7 20 13 13 
CH16 96.38 2.5 14 14 13 
CH18 125.09 1.7 15 12 16 
CH19 129.70 6.1 21 14 16 
CH20 144.05 4.7 19 10 10 
10 CH21 184.97 4.3 24 
11 CH22 213.08 2.6 19 11 16 
12 CH24 229.19 5.5 20 20 23 
13 CH25 246.78 4.1 21 15 20 
14 CH26 267.39 3.9 19 12 17 
15 CH28 287.90 1.8 26 17 21 
16 CH30 332.86 5.4 61 
17 CH31 354.68 8.1 51 
  Nan River         
18 NA02 442.41 4.8 63 
19 NA07 582.73 1.8 44 10 
20 NA11 733.55 1.9 
21 NA13 965.74 1.9 10 
  Ping River         
22 PI01 385.59 1.9 58 
23 PI03 442.10 2.8 36 
24 PI04 510.90 5.2 13 
25 PI08 581.43 1.6 11 13 
26 REO1-i5 1225.56 2.0 0 
27 PI12 1246.84 42 48 12 
28 REO1-i2 1256.02 4.0 42 0 
29 PI14 1300.75 5.2 34 3 
  Wang River         
30 WA01 643.84 0.47 10 
  Yom River         
31 YO01 459.72 1.5 29 13 
32 YO04 566.76 5.4 27 11 14 

CFU = colony-forming units.

n1: the number of data for susceptibility to amoxicillin, ciprofloxacin and doxytetracycline.

n2: the number of data for susceptibility to norfloxacin, levofloxacin, ofloxacin, sulfamethoxazole/trimethoprim and tetracycline.

Figure 1

Locations of sampling sites in Chaophraya River and its major tributaries. Urban land use is shown in gray. Sampling points are given as dots in white circles (⊙). Major cities are given as filled triangles (▴); A: Bangkok, B: Nonthaburi, C: Ayutthaya, D: Ang Thong, E: Sing Buri, F: Nakhon Sawan, G: Phisanulok, H: Kamphaeng Phet, I: Chiang Mai.

Figure 1

Locations of sampling sites in Chaophraya River and its major tributaries. Urban land use is shown in gray. Sampling points are given as dots in white circles (⊙). Major cities are given as filled triangles (▴); A: Bangkok, B: Nonthaburi, C: Ayutthaya, D: Ang Thong, E: Sing Buri, F: Nakhon Sawan, G: Phisanulok, H: Kamphaeng Phet, I: Chiang Mai.

Isolation of Escherichia coli strains

A total of 50 mL of a sample was diluted in series of 100 and 101 times. An amount of 1 mL of each sample was filtered using 0.45-μm cellulose-acetate membrane filters (37 mm Monitors, Advantec Toyo, Tokyo, Japan). Filters were then placed on Chromocult® Coliform Agar ES (Merck KGaADarmstadt, Germany), and incubated at 37°C for 24 h. After counting all colonies, blue-colored E. coli colonies were picked up into PERLCORE Trypto-Soy Broth (Eiken Chemicals, Tokyo, Japan), and incubated at 37°C overnight. Glycerol was added to a final concentration of 15–20% and the cultures were frozen and stored at −80°C. The number of cultures prepared from each sample varied according to the number of E. coli colonies picked up from the filter. Samples for which the number of isolates were less than three were not used in the further analysis. The procedures until colony counting were conducted within 24 hours after sampling. The agar plates after colony counting were stored in a freezer when the subsequent isolation was not immediately possible.

Antibiotic susceptibility test

Susceptibility of each E. coli isolate to eight antibiotics in four classes was tested by Kirby-Bauer disk diffusion method and used for the statistical analyses: ciprofloxacin (CIP), levofloxacin (LVX), norfloxacin (NFX), ofloxacin (OFX) in the fluoroquinolone (FQL) class, tetracycline (TC) and doxytetracycline (DTC) in the tetracycline class, amoxicillin (AMX) in the β-lactam class and sulfamethoxazole/trimethoprim (ST) in the sulfonamide class. Susceptibility to AMX, CIP and DTC was tested in this study. Susceptibility to LVX, NFX, OFX, TC and ST, which had been tested in our previous study (Honda et al. 2011), were also used for statistical analysis in this study. The isolated E. coli cultures were spread on PERLCORE Sensitivity Test Agar (Eiken Chemicals, Tokyo, Japan), and antibiotic test disks (KB Disk, Eiken Chemicals, Tokyo, Japan) were placed on the spread plates. After incubation at 37°C for 18 h, susceptibility was determined by measuring the diameter of plaques formed on the agar. When contamination was found on the test agar, the result on the contaminated agar were discarded. Therefore, the number of isolates used in calculation of the resistance ratio could be different because susceptibility to AMX, CIP and DTC were tested on the different agar plate from that of NFX, LVX, OFX, ST and TC.

Land use analysis

A ratio of urban area of 3, 5, 10-km radius around each sampling point was calculated using Geographic Information System (GIS) software (ArcGIS 10, ESRI Corp., Redlands, CA, USA). A digital vector map of land use in Thailand in 2000, using Landsat TM satellite images (Marc Souris, Institut de Recherche pour le Développement, France) was used for land use analysis.

Statistical analysis

Correlation of resistance among the tested antibiotics was evaluated using the phi (φ) coefficient, which ranges from zero to unity, corresponding to the magnitude of association between two variables. The φ coefficient was calculated based on 2 × 2 contingency tables established for each combination of two antibiotics. A 2 × 2 contingency table comprises the number of isolates which were resistant to the both antibiotic, resistant to either of the antibiotics and sensitive to the both antibiotics. Principal component analysis (PCA) was performed with IBM SPSS Statistics (Windows, Version 19.0. IBM Corp., Armonk, NY, USA) on datasets of resistance ratios to the tested antibiotics. The value of the KMO Measure of Sampling Adequacy for this set of variables is 0.756. The probability associated with the Bartlett test was <0.001.

RESULTS AND DISCUSSION

Geographical distribution of isolates resistant to each antibiotic

Antibiotic-resistant E. coli were found in relatively higher ratios in the Chaophraya River basin than in other countries (Reinthaler et al. 2003; Webster et al. 2004; Sayah et al. 2005; Laroche et al. 2009; Koczura et al. 2012). Different characteristics were found in the distribution of E. coli isolates resistant to fluoroquinolones (CIP, LVX, NFX and OFX) from those to non-fluoroquinolone antibiotic groups (AMX, ST, TC and DTC) (Figure 2). Ratios of isolates resistant to fluoroquinolones were more dependent on the locations than those of non-fluoroquinolones, because fluctuation of resistance ratios to fluoroquinolones was larger than those of non-fluoroquinolones. Relative standard deviations (RSD) of resistance ratios to fluoroquinolones CIP, LVX, NFX and OFX were 97%, 130%, 135% and 115%, respectively, while those of non-fluoroquinolones AMX, ST, TC and DTC were lower as 37%, 48%, 46% and 41%, respectively.
Figure 2

Change in resistance ratios to (a) amoxicillin (AMX), (b) sulfamethoxazole/trimethoprim (ST), (c) tetracycline (TC), (d) doxytetracycline (DTC), (e) ciprofloxacin (CIP), (f) levofloxacin (LVX), (g) norfloxacin (NFX), and (h) ofloxacin (OFX) in the Chaophraya River and its major tributaries.

Figure 2

Change in resistance ratios to (a) amoxicillin (AMX), (b) sulfamethoxazole/trimethoprim (ST), (c) tetracycline (TC), (d) doxytetracycline (DTC), (e) ciprofloxacin (CIP), (f) levofloxacin (LVX), (g) norfloxacin (NFX), and (h) ofloxacin (OFX) in the Chaophraya River and its major tributaries.

There was not a significant upstream-downstream trend found in resistance ratios to each antibiotic. Correlation coefficients between distance from river mouth and resistance ratio to each antibiotic were as low as −0.25 to −0.44. One possible reason is because the resistance ratios are affected not by upstream conditions but more local conditions around the sampling sites. Another reason is because downstream of Chaophraya River is the estuary, where river flow is affected by the tide. Compared with fluoroquinolones, resistance ratios to non-fluoroquinolones were relatively higher at most sites even upstream of the tributaries of Chaophraya River. This suggests that isolates resistant to non-fluoroquinolones were distributed widely in Chaophraya River and its tributaries.

Cross-resistance and multiple resistances

Resistance to fluoroquinolones was highly correlated (φ >0.8) (Table 2), suggesting cross-resistance within this class (Sanders 2001; Davies et al. 2003). A total of 78% of 69 isolates resistant to either of the first-generation fluoroquinolones CIP and NFX were also resistant to either or both second-generation fluoroquinolones LVX and OFX. High correlation in resistance patterns was also found among non-fluoroquinolone antibiotics: AMX/TC (φ >0.7), AMX/DTC (φ >0.7), ST/TC (φ = 0.75), and TC/DTC (φ = 0.92). Resistance to ST and TC had a high degree of correlation; 92% of ST-resistant isolates were also resistant to TC. High cross-resistance to ST and TC has also been reported by Sayah et al. (2005). A total of 96% of the 172 isolates resistant to DTC were also resistant to TC, probably because of cross-resistance to the tetracycline group. However, correlation between fluoroquinolone and non-fluoroquinolone classes was not significant (φ <0.5).

Table 2

Association between multiple resistances shown as φ coefficientsa

 β-LactamsSulfonamidesTetracyclines
Fluoroquinolones
AntibioticAMXSTTCDTCCIPLVXNFX
AMX        
ST 0.68       
TC 0.72 0.75      
DTC 0.71 0.65 0.92     
CIP 0.46 0.41 0.37 0.36    
LVX 0.34 0.37 0.35 0.29 0.83    
NFX 0.35 0.42 0.34 0.28 0.83 0.96  
OFX 0.37 0.39 0.34 0.30 0.83 0.98 0.95 
 β-LactamsSulfonamidesTetracyclines
Fluoroquinolones
AntibioticAMXSTTCDTCCIPLVXNFX
AMX        
ST 0.68       
TC 0.72 0.75      
DTC 0.71 0.65 0.92     
CIP 0.46 0.41 0.37 0.36    
LVX 0.34 0.37 0.35 0.29 0.83    
NFX 0.35 0.42 0.34 0.28 0.83 0.96  
OFX 0.37 0.39 0.34 0.30 0.83 0.98 0.95 

AMX = amoxicillin; CIP = ciprofloxacin; DTC = doxytetracycline; ST = sulfamethoxazole/trimethoprim; LVX = levofloxacin, NFX = norfloxacin; OFX = ofloxacin; TC = tetracycline.

aValues >0.7 are given in bold.

Relevancy of antibiotic-resistance patterns to urban factors

Principal component analysis was performed to characterize sampling sites by their patterns of resistance ratios to the eight antibiotics. Two major principle components corresponded to 88% of the total variance (Table 3). The first principle component (PC1) indicates an overall high resistance to all the tested antibiotics. The second principle component (PC2) indicates relatively higher resistance to fluoroquinolones than the non-fluoroquinolone class. Most sites in the Ping, Wang, Yom and Nan Rivers were plotted in the region where PC1 < 0, while the sampling sites in Chaophraya River had a wide range of PC1 (Figure 3). That suggests that resistance ratios in these tributaries were low, and that those in Chaophraya River had a large fluctuation. Meanwhile, all sampling sites except one in Chaophraya River plotted in the region where PC2 > −1.0, while most sites in the tributaries of Ping, Wang, Yom and Nan Rivers were plotted in the region where PC2 < 0. As a general trend, samples from Chaophraya River tend to have a higher resistance ratio to fluoroquinolones. There was no significant correlation was found between distance to the river mouth and each principal component (Table 4; Figure 4). Therefore, patterns of resistance seem to be affected by local features in the immediate surrounding area rather than the upstream conditions.
Table 3

Factor patterns of principle components (PC)a

AntibioticPC1PC2
AMX 0.841 −0.338 
ST 0.834 −0.308 
TC 0.733 −0.637 
DTC 0.750 −0.592 
CIP 0.749 0.382 
LVX 0.808 0.555 
NFX 0.815 0.545 
OFX 0.793 0.588 
Cumulative proportion 0.626 0.884 
AntibioticPC1PC2
AMX 0.841 −0.338 
ST 0.834 −0.308 
TC 0.733 −0.637 
DTC 0.750 −0.592 
CIP 0.749 0.382 
LVX 0.808 0.555 
NFX 0.815 0.545 
OFX 0.793 0.588 
Cumulative proportion 0.626 0.884 

AMX = amoxicillin; CIP = ciprofloxacin; DTC = doxytetracycline; ST = sulfamethoxazole/trimethoprim; LVX = levofloxacin, NFX = norfloxacin; OFX = ofloxacin; TC = tetracycline.

aMajor components making up a group are shown in bold.

Table 4

Correlation coefficients of principal components with distance from Chaophraya River mouth, E. coli concentration and urban land use. The values in parentheses represent the p-value derived from t distribution

 PC1PC2
Distance from the river mouth 0.37 (0.05) −0.14 (0.30) 
Log of E. coli concentration −0.26 (0.15) 0.12 (0.32) 
Urban land use     
 3 km-radius area −0.31 (0.09) 0.10 (0.34) 
 5 km-radius area −0.26 (0.14) 0.10 (0.34) 
 10 km-radius area −0.28 (0.12) 0.12 (0.32) 
 PC1PC2
Distance from the river mouth 0.37 (0.05) −0.14 (0.30) 
Log of E. coli concentration −0.26 (0.15) 0.12 (0.32) 
Urban land use     
 3 km-radius area −0.31 (0.09) 0.10 (0.34) 
 5 km-radius area −0.26 (0.14) 0.10 (0.34) 
 10 km-radius area −0.28 (0.12) 0.12 (0.32) 
Figure 3

Two-dimensional plot of antibiotic resistance patterns by principal component analysis.

Figure 3

Two-dimensional plot of antibiotic resistance patterns by principal component analysis.

Figure 4

Relationship between urban land use ratios in a 5-km radius area of the sampling point and resistance ratios to principal components: (a) PC1 and (b) PC2.

Figure 4

Relationship between urban land use ratios in a 5-km radius area of the sampling point and resistance ratios to principal components: (a) PC1 and (b) PC2.

A ratio of urban land use did not show linear correlation with each principal component (Table 4). However, it was also suggested that a certain extent of urbanization brings an increase in a ratio of antibiotic-resistant bacteria in river water. When a ratio of urban land use in 5-km radius area was lower than 20%, the location had a wide range of PC1 values (Figure 5(a)). This result shows that a ratio of antibiotic-resistant bacteria can occasionally become highly independent of increase in urban land use. Meanwhile, when a location has a higher ratio of urban land use than 20% in 5-km radius area, it had a higher PC1 value than −0.2 (Figure 5(a)). Therefore, it also indicates that the resistance ratio is more likely to become high at the location where a ratio of urban land use is higher than a certain level in the immediate surrounding area. This trend did not change even when urban land ratios were calculated with different radiuses of 3, 5 and 10 km. In relation to PC2 and urban land use, a sample tended to have a higher PC2 value than −0.1 when it had a higher ratio of urban land use than 70% in 5-km radius area (Figure 5(b)), All these locations were downstream of the Chaophraya River and close to the Bangkok Metropolitan. This suggests that resistance ratios to fluoroquinolones are likely to become relatively higher than those for non-fluoroquinolones especially in a highly populated area.
Figure 5

Relationship between urban land use ratios in a 5-km radius area of the sampling point and resistance ratios to principal components: (a) PC1 and (b) PC2.

Figure 5

Relationship between urban land use ratios in a 5-km radius area of the sampling point and resistance ratios to principal components: (a) PC1 and (b) PC2.

No obvious correlation was found between fecal contamination and each principal component (Table 4; Figure 6). The possible sources of fecal contamination are municipal wastewater which are untreated or treated but not disinfected, combined sewage overflow and wastewater from livestock farming and related industry. These wastewaters possibly contain antibiotic-resistant E. coli as well. However, this did not contribute to an increase in a ratio of resistant bacteria, probably because E. coli sensitive to antibiotics were also contained in the wastewater.
Figure 6

Relationship between E. coli concentrations and resistance ratios to principal components: (a) PC1 and (b) PC2.

Figure 6

Relationship between E. coli concentrations and resistance ratios to principal components: (a) PC1 and (b) PC2.

Consequently, the following relevancy was suggested between antibiotic-resistance patterns and urban factors: (i) resistant patterns seem to be affected by local features in the immediate surrounding area rather than the upstream conditions; (ii) discharge of fecal wastewater possibly increases population of antibiotic-resistant bacteria, but does not contribute to an increase of the resistance ratio; (iii) a ratio of antibiotic-resistant bacteria is likely to increase when urban land use locally exceeds a certain ratio; and (iv) a resistance ratio to fluoroquinolones tends to be high especially in a highly populated area.

Possible sources of antibiotic-resistant bacteria

Fecal wastewater is definitely the source of antibiotic-resistant bacteria. However, E. coli concentration did not relate to resistance ratio (Table 4; Figure 5). This fact implies that the major factor to determine the resistance ratio in the river water is the ratio of resistant bacteria in the discharged wastewater rather than their concentration. In this study, a high ratio of urban land use within 5-km radius area possibly had relevancy with a high resistance ratio in water body nearby. Therefore, point sources with a high resistance ratio which is located in urban area is likely to be the major sources of antibiotic-resistant bacteria in Chaophraya River basin. Municipal wastewater in Bangkok has been reported to have a higher ratio of antibiotic-resistant bacteria than other countries. Ratios of E. coli isolates resistant to fluoroquinolones, ST and TC in effluent from wastewater treatment plants in Bangkok, Thailand were reported to be more than 30% (Patchanee et al. 2011), and average resistance ratios in effluent samples in other countries have also been reported to be as high as 34% (±21%) for TC and 14% (±19%) for CIP (Reinthaler et al. 2003; Webster et al. 2004; Ferreira da Silva et al. 2007; Laroche et al. 2009; Łuczkiewicz et al. 2010; Amaya et al. 2012; Koczura et al. 2012). Hospital wastewater also possibly contributed to higher ratios of fluoroquinolone-resistant bacteria in the urban area. Clinical statistics from the National Antimicrobial Resistance Surveillance Center, Thailand (NARSCT) reported that NFX-resistance ratios in urine samples in the North region and Bangkok Metropolitan were 49% and 55%, respectively (NARSCT 2013). Because fluoroquinolones are more expensive than the non-fluoroquinolone antibiotics that were tested in this study, the use of fluoroquinolones is possibly limited to hospitals and highly populated areas where income is higher than in rural areas. Suburban and rural areas probably had fewer sources of fluoroquinolone-resistant bacteria like livestock farms, where high resistance ratios to a variety of antibiotics including fluoroquinolones have been reported (Padungtod et al. 2006). Meanwhile, sources of bacteria resistant to non-fluoroquinolones were widely present also in rural areas. Bryan et al. (2004) reported that tetracycline-resistant bacteria were isolated not only from humans and livestock but also from a variety of companion animals and wild animals.

In tropical climate regions, a large impact of domestic sewage is expected on prevalence of antibiotic-resistant bacteria in aquatic environments. In a combined sewer system, wastewater diluted by rainwater is discharged untreated when the amount of rainwater and sewage exceeds the capacity of treatment plants. In tropical climate regions, antibiotic-resistant bacteria in combined sewage overflow are often discharged into rivers and canals due to frequent heavy precipitation as showers in the rainy season. Improvement of the sewer infrastructure and systems would be beneficial in reducing sewage overflow and discharge of antibiotic-resistant bacteria into the environment. In this study, different tendency in antibiotic-resistance patterns between water receiving urban wastewater and rural drainage was suggested. For further study, comparison with rainy seasons and obtaining more information on resistance patterns of E. coli in urban wastewater would be important. Source tracking by phenotypes of E. coli (Hu et al. 2008) would also help to understand the fate of antibiotic-resistant bacteria in the target region.

CONCLUSIONS

E. coli resistant to non-fluoroquinolones (AMX, ST, TC and DTC) were distributed widely in Chaophraya River and its tributaries, while those of fluoroquinolones (CIP, LVX, NFX and OFX) were found to be more dependent on locations. A ratio of antibiotic-resistant bacteria is likely to increase when urban land use nearby exceeds a certain ratio. Especially, a resistance ratio to fluoroquinolones tends to be high in a highly populated area. Meanwhile, no significant contribution of fecal contamination was found to increase a resistance ratio. These results suggested that an antibiotic-resistance ratio is dependent on conditions of local urbanization rather than the upstream conditions, and that the major sources of antibiotic-resistant bacteria in the Chaophraya River basin is point sources located in the urban area which contains a high ratio of resistant bacteria.

ACKNOWLEDGEMENTS

This research was supported by Japan Science and Technology Agency (JST) and Japan International Cooperation Agency (JICA) through Science and Technology Partnership for Sustainable Development (SATREPS), and by Japan Society for the Promotion of Science (JSPS) through Grant-in-Aid for Scientific Research (B), (Grant Number: 24404017). We thank Regional Environment Offices, Ministry of Natural Resources and Environment, Thailand for their cooperation in sampling and experimental work.

REFERENCES

REFERENCES
Amaya
E.
Reyes
D.
Paniagua
M.
Calderón
S.
Rashid
M.-U.
Colque
P.
Kühn
I.
Möllby
R.
Weintraub
A.
Nord
C. E.
2012
Antibiotic resistance patterns of Escherichia coli isolates from different aquatic environmental sources in León, Nicaragua
.
Clin. Microbiol. Infect.
18
,
E347
E354
.
Baquero
F.
Martínez
J.-L.
Cantón
R.
2008
Antibiotics and antibiotic resistance in water environments
.
Curr. Opin. Biotechnol.
19
,
260
265
.
Boonyasiri
A.
Tangkoskul
T.
Seenama
C.
Saiyarin
J.
Tiengrim
S.
Thamlikitkul
V.
2014
Prevalence of antibiotic resistant bacteria in healthy adults, foods, food animals, and the environment in selected areas in Thailand
.
Pathog. Global Health
108
,
235
245
.
Davies
T. A.
Goldschmidt
R.
Pfleger
S.
Loeloff
M.
Bush
K.
Sahm
D. F.
Evangelista
A.
2003
Cross-resistance, relatedness and allele analysis of fluoroquinolone-resistant US clinical isolates of Streptococcus pneumoniae (1998–2000)
.
J. Antimicrob. Chemother.
52
,
168
175
.
Ferreira da Silva
M.
Vaz-Moreira
I.
Gonzalez-Pajuelo
M.
Nunes
O. C.
Manaia
C. M.
2007
Antimicrobial resistance patterns in Enterobacteriaceae isolated from an urban wastewater treatment plant
.
FEMS Microbiol. Ecol.
60
,
166
176
.
Halling-Sørensen
B.
Nielsen
S. N.
Lanzky
P. F.
Ingerslev
F.
Lützhoft
H. C. H.
Jørgensen
S. E.
1998
Occurrence, fate and effects of pharmaceutical substances in the environment—a review
.
Chemosphere
36
,
357
393
.
Honda
R.
Watanabe
T.
Masago
Y.
Chulasak
R.
Tanong
K.
Chaminda
G. G. T.
Sawaittayotin
V.
Wongsila
K.
Sienglum
C.
Sunthonwatthanaphong
V.
Poonnotok
A.
Chiemchaisri
C.
Furumai
H.
2011
Distribution of antibiotic resistant Escherichia coli in Chao Phraya River and its tributaries in Thailand
.
Environ. Eng. Res.
48
,
III_173
178
(
in Japanese
).
Łuczkiewicz
A.
Jankowska
K.
Fudala-Książek
S.
Olańczuk-Neyman
K.
2010
Antimicrobial resistance of fecal indicators in municipal wastewater treatment plant
.
Water Res.
44
,
5089
5097
.
NARSCT (National Antimicrobial Resistance Surveillance Center T)
2013
Percent Susceptibility by Specimen/Region/Year. http://narst.dmsc.moph.go.th/a8.php [accessed 13 April 2013
].
Padungtod
P.
Kaneene
J. B.
Hanson
R.
Morita
Y.
Boonmar
S.
2006
Antimicrobial resistance in Campylobacter isolated from food animals and humans in northern Thailand
.
FEMS Immunol. Med. Microbiol.
47
,
217
225
.
Patchanee
N.
Watanabe
T.
Chiemchaisri
W.
Threedeach
S.
Honda
R.
Chiemchaisri
C.
2011
Behavior of antibiotic resistant Escherichia coli in activated sludge process for municipal wastewater treatment in tropical regions
.
Proceedings of the 16th International Symposium on Health-Related Water Microbiology
,
18–23 September 2011
.
Rotorua
,
New Zealand
.
Reinthaler
F.
Posch
J.
Feierl
G.
Wust
G.
Haas
D.
Ruckenbauer
G.
Mascher
F.
Marth
E.
2003
Antibiotic resistance of E. coli in sewage and sludge
.
Water Res.
37
,
1685
1690
.
Rico
A.
Oliveira
R.
McDonough
S.
Matser
A.
Khatikarn
J.
Satapornvanit
K.
Nogueira
A. J. a.
Soares
A. M. V. M.
Domingues
I.
Van den Brink
P. J.
2014
Use, fate and ecological risks of antibiotics applied in tilapia cage farming in Thailand
.
Environ. Pollut.
191
,
8
16
.
Shimizu
A.
Takada
H.
Koike
T.
Takeshita
A.
Saha
M.
Rinawati Nakada
N.
Murata
A.
Suzuki
T.
Suzuki
S.
Chiem
N. H.
Tuyen
B. C.
Viet
P. H.
Siringan
M. A.
Kwan
C.
Zakaria
M. P.
Reungsang
A.
2013
Ubiquitous occurrence of sulfonamides in tropical Asian waters
.
Sci. Total Environ.
452–453
,
108
115
.
Udomwiboonchai
P.
Leemingsawat
C.
Visuthisakchai
S.
Thamlikitkul
V.
2007
Prevalence of antibiotic-resistant bacteria colonized in neutropenic patients at Siriraj Hospital
.
Siriraj Med. J.
59
,
344
347
.
Webster
L. F.
Thompson
B. C.
Fulton
M. H.
Chestnut
D. E.
Van Dolah
R. F.
Leight
A. K.
Scott
G. I.
2004
Identification of sources of Escherichia coli in South Carolina estuaries using antibiotic resistance analysis
.
J. Exp. Mar. Bio. Ecol.
298
,
179
195
.