Abstract

Actinobacteria can be one of the causes of earthy and musty odors in drinking water. In this study, the distribution and odor producing ability of actinobacteria isolated in the Han River as a source of tap water were investigated. Actinobacteria were detected in low concentration from December to February and this gradually increased in March and April. The number of actinobacteria detected was particularly high in April (63 CFU/mL), July (45 CFU/mL), and October (39 CFU/mL) due to the influence of rainfall. Actinobacteria with geosmin-producing genes were detected mainly in March and July. In contrast, actinobacteria with 2-MIB-producing genes were detected mainly in October. There was a difference in the time when actinobacteria with the geosmin and 2-MIB-producing gene were highly detected in the river. Also, the types of actinobacteria with the geosmin and 2-MIB-producing gene were different. More than 70% of the geosmin inducer gene was isolated in Streptomyces, but the 2-MIB inducer gene was detected in various genera of actinobacteria as well as Streptomyces. The detection of odorous substances in March and October when cyanobacteria were not detected, or the detected number was low, suggested that actinobacteria could be a cause of odor inducers in the Han River.

HIGHLIGHTS

  • The main cause of geosmin and 2-MIB is known to be cyanobacteria. However, odorous substances are detected even when cyanobacteria are not present. We evaluated whether actinobacteria can be a cause of odor inducers in the Han River, Korea.

  • Geosmin inducing gene was highly detected in Streptomyces during March and July. However, 2-MIB inducing gene was detected in various actinobacteria mainly during October.

  • When cyanobacteria were not detected, geosmin and 2-MIB were detected and these inducing genes were produced by actinobacteria.

  • Total colony counts of actinobacteria did not necessarily indicate the production of odor-inducing genes.

  • Our results suggest that some actinobacteria can cause odors in the Han River and the two odorous substances by actinobacteria must be managed separately. Also, it is necessary to confirm the odor inducing genes, not the number of actinobacteria, to verify the production of odorous substances by actinobacteria.

INTRODUCTION

Earthy and musty odors are one of the major reasons for distrust of tap water. Geosmin and 2-methylisoborenol (2-MIB), which are typical odorous substances, cause earthy and musty odors (Watson et al. 2008; Schrader & Summerfelt 2010). These substances are known to be produced mainly by cyanobacteria (Lin et al. 2002; Watson et al. 2008; Sun et al. 2013). Many studies have suggested the correlation between odorous substances and cyanobacteria in water (Hu et al. 2003; Sun et al. 2013). However, geosmin and 2-MIB can also be present at times when there are few or no cyanobacteria in surface water (Durrer et al. 1999; Lanciotti et al. 2003).

Actinobacteria are distributed in soil and rivers, and are bacteria that form mycelium-like fungi (Klausen et al. 2005; Zaitlin & Watson 2006). Actinobacteria have been mainly studied for their ability to decompose organic compounds and produce antibiotics (Zaitlin & Watson 2006; George et al. 2011).

There are a few studies on the distribution and production of odorous substances of actinobacteria in water (Jensen et al. 1994; Sugiura & Nakano 2000; Zuo et al. 2009,, 2010; Schrader & Summerfelt 2010; Auffret et al. 2011; Lylloff et al. 2012; Asquith et al. 2018). However, most studies have been done about Streptomyces, which are known to be the major species producing odorous substances or reported geosmin production rather than 2-MIB by actinobacteria (Sugiura & Nakano 2000; Zuo et al. 2009, 2010; Schrader & Summerfelt 2010; Auffret et al. 2011; Lylloff et al. 2012; Asquith et al. 2018).

Furthermore, several studies investigated actinobacteria numbers only, without testing their ability to produce odorous substances in water, or compared actinobacteria numbers with geosmin and 2-MIB concentrations detected in water samples, so that the production of odor-inducing substances by actinobacteria could not be confirmed (Herson et al. 1991; Lanciotti et al. 2003; Zuo et al. 2010).

Therefore, it is necessary to confirm the presence of odor causing genes, not the number of actinobacteria, and identify the kinds of actinobacteria that produce 2-MIB as well as geosmin.

The aims of this study were to investigate the seasonal characteristics of geosmin and 2-MIB producing genes present in actinobacteria, to identify various genera of actinobacteria that can produce odorous substances, and to understand the causes of odorous substances in the Han River, which is the raw material for tap water in Seoul.

MATERIALS AND METHODS

Sampling

The Han River is used as a water source in Seoul, Korea. The five water intake points are located between the Paldang dam and Jamsil weir in the Han River (Figure 1). Samples were collected monthly from January to December 2015 at the five intake points. A total of 60 samples were collected and all samples were kept at 4 °C until analysis.

Figure 1

Sampling sites (●) in the Han River, Korea.

Figure 1

Sampling sites (●) in the Han River, Korea.

Analysis of water quality parameters

Cyanobacteria, pH, chlorophyll-a, total coliforms (TC), fecal coliforms (FC), temperature, turbidity, conductivity, dissolved oxygen (DO), biochemical oxygen demand (BOD), chemical oxygen demand (COD), ammonia nitrogen (NH3-N), nitric nitrogen (NO3-N), total nitrogen (T-N), and total phosphorus (T-P) were analyzed according to the Korean standard method for the examination of water (Ministry of Environment 2012). Geosmin and 2-MIB from the water samples were extracted by headspace solid phase micro extraction (HS-SPME) using the CombiPAL SPME System (CTC Analytics, Switzerland). Gas chromatography-mass spectrometry (GC-MS; Varian 450-GC/240-MS, USA) was used for quantification of geosmin and 2-MIB according to the test method of water quality monitoring items (Ministry of Environment 2014). The discharge of Paldang dam and precipitation data were obtained from the Han River Flood Control Office (http://www.hrfco.go.kr) and the Korea Meteorological Administration (http://www.weather.go.kr), respectively.

Actinobacteria culture

The actinobacteria culture method was modified by the method described by Kuster & Williams (1964). 250 mL of water sample was concentrated by filtration through a 0.45 μm membrane filter paper, and the filter paper was suspended in 10 mL of phosphate buffer (PBS). The filter paper was treated by ultrasonic treatment for 5 min and heated at 55 °C for 5 min. 100 μL of sample were spread on Starch Casein Agar (containing 100 μg/mL cycloheximide and 10 μg/mL nalidixic acid) three times and cultured at 28 °C for more than 14 days.

PCR analysis and sequencing

DNA was extracted from actinobacteria colonies using FastDNA spin kit for soil (MP biomedicals, USA). The primers and PCR conditions are shown in Table 1. For conventional PCR analysis, primers targeting the 16S rRNA region were used, and 200 nM of each primer and 2 μL of DNA template were added to make a total of 50 μL of the PCR reaction solution (Zuo et al. 2010). PCR products were sequenced using the ABI Prism 3100 Genetic Analyzer (Applied Biosystems, USA). The 16S rRNA gene sequence of actinobacteria was registered in GenBank (http://ncbi.nlm.nih.gov) (MH745813-MH745928). Sequences obtained were aligned with the ClustalW program and the sequence alignments were used to construct phylogenic trees using the neighbor-joining method of MEGA version 6.06. The safety of the tree was assessed through bootstrap analysis (1,000 replications). PCR analysis was also performed to confirm that the separated actinobacteria possessed the gene for production of 2-MIB and geosmin. The primers used, as well as the reaction conditions, are shown in Table 2 (Giglio et al. 2008; Auffret et al. 2011).

Table 1

Conventional PCR condition

PCR typePrimerPCR condition
Conventional PCR Forward: 5′-AGAGTTTGATCCTGGCTCAG-3′
Reverse: 5′-AAGGAGGTGATCCAGCCGC-3′ 
94 °C for 10 min;
35 cycles (94 °C for 30 s; 56 °C for 30 s; 72 °C for 45 s);
72 °C for 5 min 
PCR typePrimerPCR condition
Conventional PCR Forward: 5′-AGAGTTTGATCCTGGCTCAG-3′
Reverse: 5′-AAGGAGGTGATCCAGCCGC-3′ 
94 °C for 10 min;
35 cycles (94 °C for 30 s; 56 °C for 30 s; 72 °C for 45 s);
72 °C for 5 min 
Table 2

PCR conditions for Geosmin and 2-MIB production gene identification

GenePrimerPCR condition
Geosmin Forward: 5′-TTCTTCGACGAYCACTTCC-3′
Reverse: 5′-CCCTYGTTCATGTARCGGC-3′ 
96 °C for 5 min;
30 cycles (95 °C for 30 s; 60 °C for 30 s; 70 °C for 60 s);
72 °C for 7 min 
2-MIB Forward: 5′-TGGACGACTGCTACTGCGAG-3′
Reverse: 5′-AAGGCGTGCTGTAGTTCGTTGTG-3′ 
96 °C for 5 min;
30 cycles (95 °C for 30 s; 58 °C for 30 s; 70 °C for 60 s);
72 °C for 7 min 
GenePrimerPCR condition
Geosmin Forward: 5′-TTCTTCGACGAYCACTTCC-3′
Reverse: 5′-CCCTYGTTCATGTARCGGC-3′ 
96 °C for 5 min;
30 cycles (95 °C for 30 s; 60 °C for 30 s; 70 °C for 60 s);
72 °C for 7 min 
2-MIB Forward: 5′-TGGACGACTGCTACTGCGAG-3′
Reverse: 5′-AAGGCGTGCTGTAGTTCGTTGTG-3′ 
96 °C for 5 min;
30 cycles (95 °C for 30 s; 58 °C for 30 s; 70 °C for 60 s);
72 °C for 7 min 

Data analysis

The relationships between water quality parameters and numbers of actinobacteria in water samples were analyzed by Pearson correlation analysis. Analysis of variance (ANOVA) was used to test difference in numbers of actinobacteria obtained from sampling sites. Statistical significant were considered at the 95% confidence level (Microsoft Excel 2010).

RESULTS

Presence of Actinobacteria in water samples

As a result of actinobacteria at the 5 water intake sites, the highest concentration detected was 110 CFU/mL at S2 and S4 in April and the lowest concentration detected was 2 CFU/mL at S5 in December (Figure 2). Overall, there was an average of 63 CFU/mL in April and 3 CFU/mL in December. Also, the average was 45 CFU/mL in July and 39 CFU/mL in October. In April, July, and October, there was rainfall during the sampling. The average number of actinobacteria at the 5 intake points was 19 CFU/mL (S1), 29 CFU/mL (S2), 22 CFU/mL (S3), 24 CFU/mL (S4), and 18 CFU/mL (S5). Though somewhat higher at S2 and S4, there was no significant difference by location (p > 0.05, data not shown).

Figure 2

Monthly distribution of Actinobacteria detected in 5 water intake sites.

Figure 2

Monthly distribution of Actinobacteria detected in 5 water intake sites.

Diversity of species

A total of 116 colonies was randomly selected from the water samples. Streptomyces sp. (34%) was the most common genus. Mycobacterium sp., Micromonospora sp. were also detected at 19% and 13%, respectively. In addition, Streptosporangium sp. (9%), Microbispora sp. (5%), Gordonia sp. (4%), and Sphaerisporangium sp. (3%) were also detected (Figure 3).

Figure 3

Actinobacteria genus detected in the Han River.

Figure 3

Actinobacteria genus detected in the Han River.

Streptomyces was detected continuously throughout the year, especially in summer. Mycobacterium, which was included in a broad range of actinobacteria, was also detected throughout the year and was mainly detected in spring and summer. Streptosporangium, Micromonospora, and Microbispora were detected mainly in spring. Most actinobacteria (40%) were detected in spring. In addition, actinobacteria of the most varied genus including Streptomyces, Streptosporangium, Micromonospora, Microbispora, Mycobacterium, Pseudonocardia, Sphaerisporangium, Rhodococcus, Nocardioides, and Actinomadura were detected at this time (Figure 4).

Figure 4

Seasonal distribution of Actinobacteria genus detected in the Han River.

Figure 4

Seasonal distribution of Actinobacteria genus detected in the Han River.

Detection of geosmin and 2-MIB genes

Genes producing geosmin and 2-MIB were analyzed in 116 strains. Genes producing geosmin were identified in 15 strains of Streptomyces sp. (75%), two strains of Streptosporangium sp. (10%), 2 strains of Sphaerisporangium sp. (10%), and one strain of Micromonospora sp. (5%) (Figure 3).

In phylogenetic analysis, the genus of Streptomyces were reclassified at the species level as S. lividans, S. albogriseolus, S. glauciniger, S. turgidiscabies, S. fildesensis, S. griseochromogenes, and S. achromogenes. The two strains of Streptosporangium were reclassified at the species level as S. amethystogenes and S. pseudovulgare. Sphaerisporangium sp. was reclassified at the species level as S. melleum, and Micromonospora sp. was reclassified at the species level as M. coxensis (Figure 5).

Figure 5

Phylogenetic relationships of Actinobacteria with geosmin-producing gene. Numbers of the branch nodes express bootstrap values as a percentage of 1,000 replicates.

Figure 5

Phylogenetic relationships of Actinobacteria with geosmin-producing gene. Numbers of the branch nodes express bootstrap values as a percentage of 1,000 replicates.

Genes producing 2-MIB were identified in eight strains identified as Streptomyces sp. (53.3%), two strains of Streptosporangium sp. (13.3%), one strain of Nocardioides sp., one strain of Mycobacterium sp., one strain of Micromonospora sp., one strain of Actinomadura sp., and one strain of Nonomuraea sp. In phylogenetic analysis, strains identified as Streptomyces were analyzed at the species level as S. atratus and S. lividans. The genus of Streptosporangium was analyzed at the species level as S. pseudovulgare, and the genus of Nocardioides was reclassified at the species level as N. maritimus. The genus of Mycobacterium was reclassified at the species level as M. moriokaense and Micromonospora was reclassified at the species level as M. purpureochromogenes. Actinomadura was reclassified at the species level as A. glauciflava and Nonomuraea was reclassified at the species level as N. bangladeshensis (Figure 6).

Figure 6

Phylogenetic relationships of Actinobacteria with 2-MIB-producing gene. Numbers of the branch nodes express bootstrap values as a percentage of 1,000 replicates.

Figure 6

Phylogenetic relationships of Actinobacteria with 2-MIB-producing gene. Numbers of the branch nodes express bootstrap values as a percentage of 1,000 replicates.

Geosmin-producing genes were identified as 20 isolates and 2-MIB producing genes were identified as 15 isolates. Four isolates identified as Streptomyces and Streptosporangium produced both geosmin and 2-MIB genes. The geosmin gene in actinobacteria was more abundant than the 2-MIB, but 2-MIB-producing genes were present in a variety of actinobacteria, more than geosmin (Figures 5 and 6).

Actinobacteria containing the genes for producing geosmin and 2-MIB were detected in different seasons, though genetic analysis of all actinobacteria detected in the water intakes was not conducted. Geosmin-producing genes were detected in spring and summer, mainly in March and July. In contrast, 2-MIB-producing genes in actinobacteria were detected mainly in October (Figure 7).

Figure 7

Monthly distribution of Actinobacteria with geosmin or 2-MIB production gene in the Han River.

Figure 7

Monthly distribution of Actinobacteria with geosmin or 2-MIB production gene in the Han River.

Correlation with water quality parameter

Table 3 shows the results of correlation analysis between water quality items and actinobacteria. Actinobacteria showed positive correlation with water temperature, BOD, and turbidity, negative correlation with pH, DO, conductivity, NO3-N, and T-N. Actinobacteria did not correlate with total coliforms and fecal coliforms.

Table 3

The relation coefficient between the water quality index and actinobacteria

TemperaturepHDOConductivityBODNO3-N
Actinobacteria 0.491* −0.380* −0.451* −0.493* 0.448* −0.332* 
 T-N Turbidity TC FC Discharge Precipitation 
Actinobacteria −0.314** 0.292** 0.009 0.119 0.236 0.464* 
TemperaturepHDOConductivityBODNO3-N
Actinobacteria 0.491* −0.380* −0.451* −0.493* 0.448* −0.332* 
 T-N Turbidity TC FC Discharge Precipitation 
Actinobacteria −0.314** 0.292** 0.009 0.119 0.236 0.464* 

*p < 0.01, **p < 0.05.

As a result of correlation analysis between the discharge of Paldang dam and actinobacteria, it was not statistically significant (r = 0.236, p = 0.07; Table 3). However, the discharge on the sampling date was highest in April when actinobacteria was highest, and the discharge was lowest in December when actinobacteria was lowest (Table 4, Figure 2).

Table 4

The discharge of Paldang dam and precipitation in 2015

 Month
JanFebMarAprMayJuneJulyAugSepOctNovDec
Discha. (m3/s) 127.8 128.7 128.0 179.0 127.7 79.7 81.4 81.4 80.9 83.0 85.6 53.9 
Precipi. (mm) 11.3 22.7 9.6 80.5 28.9 99.0 226 72.9 26.0 81.5 104.6 29.1 
 Month
JanFebMarAprMayJuneJulyAugSepOctNovDec
Discha. (m3/s) 127.8 128.7 128.0 179.0 127.7 79.7 81.4 81.4 80.9 83.0 85.6 53.9 
Precipi. (mm) 11.3 22.7 9.6 80.5 28.9 99.0 226 72.9 26.0 81.5 104.6 29.1 

Discha. : Discharge of Paldang dam, Precipi.: Precipitation.

The amount of precipitation in Seoul, 2015, was 792.1 mm, the lowest was 9.6 mm in March and the highest was 226 mm in July (Table 4). Unlike precipitation that was concentrated in summer (June to September) during the year, precipitation in 2015 was high in April (80.5 mm), October (81.5 mm) and November (104.6 mm) as well as in summer (Table 4). Actinobacteria concentration was also high in April (63 CFU/mL), July (46 CFU/mL); and October (39 CFU/mL); there was rainfall during the sampling (Figure 2).

Table 5 shows the correlation between the concentration of actinobacteria and precipitation on sampling day, 1 day before, and 5 days before sampling. On the day of sampling, the day before sampling, the day of and the day before sampling showed a statistically significant correlation with the actinobacteria concentration. Particularly, the precipitation on the day of sampling showed the greatest effect on the concentration of actinobacteria.

Table 5

The relation coefficient between precipitation and actinobacteria

 Precipitation
On the day of samplingThe previous day before samplingOn the day and previous dayFor 5 days before sampling
Actinobacteria 0.464* 0.272** 0.328** 0.188 
 Precipitation
On the day of samplingThe previous day before samplingOn the day and previous dayFor 5 days before sampling
Actinobacteria 0.464* 0.272** 0.328** 0.188 

*p < 0.01, **p < 0.05.

Evaluation of odorous compounds from Actinobacteria

Geosmin concentration was the highest at 46.4 ng/L in July and lowest at 1.4 ng/L in January. The concentration of 2-MIB was high in August and September, and the highest value was 37.4 ng/L in September and the lowest value was 0.2 ng/L in April (Table 6).

Table 6

Monthly distribution of actinobacteria, cyanobacteria, geosmin, and 2-MIB in the Han River

MonthActinobacteria (CFU/mL)Cyanobacteria
2-MIB (ng/L)Geosmin (ng/L)
Cell number (cells/mL)Dominant species
Jan – 4.8 1.4 
Feb 13 – 4.6 2.4 
Mar 34 – 7.4 4.4 
Apr 63 40.4 Aphanizomenon 0.2 5.8 
May 18 226 Oscillatoria 3.2 5.4 
Jun 34 18 Microcystis 2.8 10 
Jul 46 448 Anabaena 3.4 46.4 
Aug 23 6,420 Microcystis 35 23.6 
Sep 38 16,322 Microcystis 37.4 17 
Oct 39 394 Anabaena 12.2 16.4 
Nov 23 192 Phormidium 6.6 
Dec – 3.8 
MonthActinobacteria (CFU/mL)Cyanobacteria
2-MIB (ng/L)Geosmin (ng/L)
Cell number (cells/mL)Dominant species
Jan – 4.8 1.4 
Feb 13 – 4.6 2.4 
Mar 34 – 7.4 4.4 
Apr 63 40.4 Aphanizomenon 0.2 5.8 
May 18 226 Oscillatoria 3.2 5.4 
Jun 34 18 Microcystis 2.8 10 
Jul 46 448 Anabaena 3.4 46.4 
Aug 23 6,420 Microcystis 35 23.6 
Sep 38 16,322 Microcystis 37.4 17 
Oct 39 394 Anabaena 12.2 16.4 
Nov 23 192 Phormidium 6.6 
Dec – 3.8 

Cyanobacteria, which are known to be most related to the production of odor-inducing substances, were not detected in all of December to March, but were detected in high numbers in August and September, the highest numbers were 16,322 cells/mL in September, and the dominant species was Microcystis. In August and September, rainfall was low and the temperature was high, suggesting that cyanobacteria grew greatly. 2-MIB was also highly detected at this time (Table 6).

On the other hand, actinobacteria having the geosmin gene were highly detected in March and July, and actinobacteria having the 2-MIB gene were highly detected in October (Figure 7).

In March, odorous substances were detected in the Han River even though cyanobacteria were not detected at any of the sites (Figure 8, Table 6). Therefore, geosmin and 2-MIB in March could be generated by actinobacteria. In July, it was difficult to accurately estimate the effect of actinobacteria due to the detection of cyanobacteria at the 5 water intake sites. However, in October, 2-MIB was detected at the S2 site (9 ng/L), though cyanobacteria were not detected (Figure 8, Table 6). The 2-MIB inducing gene was identified in all actinobacteria randomly selected from the S2 site. We estimated that the 2-MIB was generated by actinobacteria.

Figure 8

Monthly distribution of Actinobacteria, cyanobacteria, geosmin, and 2-MIB in 5 water intake sites.

Figure 8

Monthly distribution of Actinobacteria, cyanobacteria, geosmin, and 2-MIB in 5 water intake sites.

As a result of correlation analysis between actinobacteria, cyanobacteria, geosmin and 2-MIB, cyanobacteria showed a high correlation (r = 0.676, p < 0.01) with 2-MIB, but not with geosmin (r = 0.180, p = 0.170). Actinobacteria did not show statistically significant correlation with geosmin (r = 0.218, p = 0.09) or 2-MIB (r = 0.131, p = 0.319). However, the actinobacteria and geosmin concentrations correlated (r = 0.359, p < 0.01) except for in the April data, when actinobacteria were highly detected but the detection of odor-inducing genes was low. In addition, the geosmin gene detection rate in actinobacteria and geosmin concentration in water were statistically significant (r = 0.590, p < 0.05). Therefore, simply analyzing the correlation between the number of actinobacteria and the odor-inducing substance can underestimate the production of odor-inducing substances by actinobacteria. Analysis of correlations between actinobacteria producing odor-inducing genes and odor-inducing substances could provide more accurate results.

DISCUSSION

The main cause of odors in rivers is known to be cyanobacteria (Jüttner & Watson 2007; Watson et al. 2008). However, there is a case in which odor-inducing substances are detected even when cyanobacteria is not present. Therefore, there is a problem about these causes, other than cyanobacteria. In this case, actinobacteria were considered to be one of the causes and this study was conducted. Actinobacteria were detected at high concentration in April, July, and October at 5 sampling sites of the Han River. Actinobacteria were detected at low concentration in winter (from December to February) and gradually increased during spring (March and April). In April, July, and October, it was estimated that actinobacteria in the soil flowed into the river, due to the influence of rainfall, and increased the number of actinobacteria. These results are also similar to previous studies that found Streptomyces, which accounts for most of the actinobacteria in soil, showed a high distribution rate (80.9%) in spring (April) and actinobacteria in rivers showed maximum in spring and high in rainfall (Jensen et al. 1994). As the actinobacteria showed positive correlations with water temperature, turbidity, and precipitation, it was thought that caution should be paid to the generation of odorous substances from actinobacteria during high water temperature and rainfall periods. There was no correlation between actinobacteria and total coliforms or fecal coliforms. This result indicated that it was difficult to monitor with total coliforms or fecal coliforms as an indicator bacteria of fecal contamination because actinobacteria are environmental organisms.

Actinobacteria were highest in April, but the reason the concentration of geosmin and 2-MIB was lower than that of other sampling months was estimated to be because many actinobacteria in the soil flowed into the river due to the first rainfall after the long dry season, but most of them were in the form of spores. When actinobacteria are present in the form of spores, they do not actually produce odorous substance, which is why the correlation between actinobacteria and odorous substances is not apparent (AWWA 2004). In previous studies, it was difficult to determine that actinobacteria influenced the production of geosmin and 2-MIB in water because the colony counts of actinobacteria detected in water, not the actinobacteria having the odor-inducing gene, was analyzed for the relationship with the odor-inducing substances (Lanciotti et al. 2003; Zuo et al. 2010).

In addition, many studies have been conducted mainly on Streptomyces (Zuo et al. 2009; Zuo et al. 2010; Auffret et al. 2011; Lylloff et al. 2012; Asquith et al. 2018). As suggested in our results, unlike Streptomyces, which was found to be a major actinobacteria producing geosmin, Streptomyces accounts for only 50% of the 2-MIB inducing gene in isolated actinobacteria and the 2-MIB inducing gene was generated from a variety of actinobacteria other than Streptomyces. Therefore, many papers have suggested a correlation between Streptomyces and geosmin rather than 2-MIB (Sugiura & Nakano 2000; Zuo et al. 2009; Zuo et al. 2010).

Another reason it is difficult to determine that actinobacteria influence the production of odor-inducing substances in rivers is that the odor-inducing substances produced by actinobacteria are not distributed evenly throughout the year, but have a specific seasonal distribution. Actinobacteria having the geosmin inducing gene were detected mainly in spring and summer, and actinobacteria having the 2-MIB inducing gene were detected mainly in autumn. In particular, actinobacteria with the 2-MIB-producing gene were detected more than 60% in October and the specificity of detection time was very strong.

When the cyanobacteria and actinobacteria were detected together in water, it was difficult to determine the effect of actinobacteria on the odorous substances due to the larger number of cyanobacteria, compared to that of actinobacteria. However, even in the absence of cyanobacteria, odorous substances were detected, and actinobacteria carrying genes encoding geosmin and 2-MIB were detected, suggesting that actinobacteria could be a cause of odor-inducers in the Han River. In particular, the detection of odor substances in March and October was estimated to be an effect of actinobacteria because cyanobacteria were not detected or at low levels and actinobacteria with geosmin and 2-MIB inducing genes were mainly detected in those months.

CONCLUSION

In this study, the seasonal distribution and odor producing ability of actinobacteria were investigated to evaluate whether they are one of odor inducers in the Han River, Korea. The results suggest that some actinobacteria present in the Han River can cause odors because cyanobacteria were not detected when geosmin and 2-MIB were detected and these inducer genes were produced by actinobacteria. To determine the production of geosmin and 2-MIB from actinobacteria in rivers, the seasonal characteristics of the two odorous substances must be identified separately as the main production periods of geosmin and 2-MIB from actinobacteria were different. In particular, since 2-MIB, unlike geosmin, is produced by various actinobacteria other than Streptomyces, the study of Streptomyces alone may be difficult to accurately confirm 2-MIB production by actinobacteria. Also, in order to verify the production of odor-inducing substances by actinobacteria in the aquatic environment, it is necessary to confirm the actinobacteria having odor-inducing genes because the total colony counts of actinobacteria may not necessarily indicate the production of odorous substances. Further studies are needed on how to quickly detect actinobacteria producing odorous substances in water. The application of real-time PCR can rapidly quantify actinobacteria and the genes responsible for the production of geosmin and 2-MIB because the culture method for actinobacteria takes a long time and cannot distinguish between the actively growing forms and spores.

REFERENCES

REFERENCES
American Water Works Association (AWWA)
2004
Problem Organisms in Water: Identification and Treatment
, 3rd edn.
American Water Works Association
,
Denver
,
USA
, pp.
1
5
.
Auffret
M.
,
Pilote
A.
,
Proulx
É.
,
Proulx
D.
,
Vandenberg
G.
&
Villemur
R.
2011
Establishment of a real-time PCR method for quantification of geosmin-producing Streptomyces spp. in recirculating aquaculture systems
.
Water Research
45
,
6753
6762
.
Durrer
M.
,
Zimmermann
U.
&
Juttner
F.
1999
Dissolved and particle-bound geosmin in a mesotrophic lake (Lake Zurich): spatial and seasonal distribution and the effect of grazers
.
Water Research
32
,
1673
1677
.
George
M.
,
Neethu
C.
,
Aswathi
N.
&
Hatha
A. A. M.
2011
Diversity of Bacillus and Actinomycetes in the water and sediment samples from Kumarakom region of Vembanadu lake
.
Indian Journal of Geo-Marine Sciences
40
,
430
437
.
Giglio
S.
,
Jiang
J.
,
Saint
C. P.
,
Cane
D. E.
&
Monis
P. T.
2008
Isolation and characterization of the gene associated with geosmin production in cyanobacteria
.
Environmental Science and Technology
42
,
8027
8232
.
Herson
D. S.
,
Marshall
D. R.
,
Baker
K. H.
&
Victoreen
H. T.
1991
Association of microorganisms with surfaces in distribution systems
.
Journal of the American Water Works Association
83
,
103
106
.
Hu
Q.
,
Sommerfeld
M.
,
Baker
L.
&
Westerhoff
P.
2003
Canal wall brushing – a control measure for taste and odour problems in drinking water supplies in and environment
.
Journal of Water Supply and Technology-AQUA
52
,
545
554
.
Jensen
S. E.
,
Anders
C. L.
,
Goatcher
L. J.
,
Perley
T.
,
Kenefick
S.
&
Hrudey
S. E.
1994
Actinomycetes as a factor in odour problems affecting drinking water from the North Saskatchewan River
.
Water Research
28
,
1393
1401
.
Jüttner
F.
&
Watson
S. B.
2007
Biochemical and ecological control of geosmin and 2-methylisoborneol in source waters
.
Applied Environmental Microbiology
73
,
4395
4406
.
Klausen
C.
,
Nicolaisen
M. H.
,
Strobel
B. W.
,
Warnecke
F.
,
Nielsen
J. L.
&
Jørgensen
N. O.
2005
Abundance of actinobacteria and production of geosmin and 2-methylisoborneol in Danish streams and fish ponds
.
FEMS Microbiology Ecology
52
,
265
278
.
Lanciotti
E.
,
Santini
C.
,
Lupi
E.
&
Burrini
D.
2003
Actinomycetes, cyanobacteria and algae causing tastes and odours in water of the River Arno used for the water supply of Florence
.
Journal of Water Supply and Technology-AQUA
52
,
489
500
.
Lin
T. F.
,
Wong
J. Y.
&
Kao
H. P.
2002
Correlation of musty odor and 2-MIB in two drinking water treatment plants in South Taiwan
.
Science of the Total Environment
289
,
225
235
.
Lylloff
J. E.
,
Mogensen
M. H.
,
Burford
M. A.
,
Schlüter
L.
&
Jørgensen
N. O. G.
2012
Detection of aquatic streptomycetes by quantitative PCR for production of taste-and-odour episodes in water reservoirs
.
Journal of Water Supply and Technology-AQUA
61
,
272
282
.
Ministry of Environment
2012
Korean Standard Method for the Examination of Water
.
Republic of Korea
.
Ministry of Environment
2014
Korean Test Method of Water Quality Monitoring Items
.
Seoul, Republic of Korea
.
Sun
D. L.
,
Yu
J. W.
,
An
W.
,
Yang
M.
,
Chen
C. G.
&
Zhang
S. J.
2013
Identification of causative compounds and microorganisms for musty odor occurrence in the Huangpu river, China
.
Journal of Environmental Sciences
25
,
460
465
.
Watson
S. B.
,
Ridal
J.
&
Boyer
G. L.
2008
Taste and odour and cyanobacterial toxins: impairment, prediction, and management in the Great Lakes
.
Canadian Journal of Fisheries and Aquatic Sciences
65
,
1779
1796
.
Zuo
Y.
,
Li
L.
,
Zhang
T.
,
Zheng
L.
,
Dai
G.
,
Liu
L.
&
Song
L.
2010
Contribution of Streptomyces in sediment to earthy odor in the overlying water in Xionghe Reservoir, China
.
Water Research
44
,
6085
6094
.