A comprehensive study of the identity and population dynamics of filamentous bacteria in five Polish full-scale municipal wastewater treatment plants (WWTPs) with nutrients removal had been carried out for 2 years. A quantitative culture-independent, molecular method – fluorescence in situ hybridization – was applied to evaluate the structure of different filamentous bacteria populations and their temporal variations. Activated sludge was examined for the abundance of 11 groups of filamentous bacteria. On average, filaments constituted 28% of all bacteria. All samples presented a low diversity of probe-defined filamentous bacteria, usually with significant domination of Chloroflexi (with distinction to types 1851, 0803 and others) and/or Microthrix (14% and 7% of EUBmix, respectively). Haliscomenobacter hydrossis, Mycolata, Skermania piniformis and TM7 were less abundant, whereas Curvibacter, Thiothrix/021N and family Gordonia have not been detected in any of the samples. The tested WWTPs showed similarity among species found and differences in their abundance. The composition of filamentous populations was rather stable in each plant and similar to those found in other European countries. Little differences between plants were shown by multivariate analysis of variance in terms of Chloroflexi and Microthrix. No significant general correlations have been found with Pearson product-moment correlation coefficient and Spearman's rank correlation coefficient. Medium correlation strength between the presence of different filaments was recorded only for Microthrix and Skermania piniformis. Deleterious effect on settling properties of sludge (measured as sludge volume index) was found only for abundance of Microthrix; a strong linear correlation was recorded between them. However, no other correlations with wastewater and operational data were revealed.

INTRODUCTION

Activated sludge technology has been used for a hundred years; however, it still brings some serious operational problems. Main issues are bulking and foaming in wastewater treatment plants (WWTPs), which are associated with extensive proliferation of some filamentous bacteria (Jenkins et al. 2004; Nielsen et al. 2009b; Mielczarek et al. 2012). Current studies with molecular methods (e.g. fluorescence in situ hybridization (FISH)) have shown that previous identification based on morphology and chemical staining can often be unreliable (Nielsen et al. 2009b; Mielczarek et al. 2012). Also, laboratory-scale studies of the isolated microorganisms in pure cultures have to be used carefully, because data from those studies are not easy to extrapolate to full-scale conditions. Ecophysiological studies under in situ conditions on probe-defined populations (Eales et al. 2005, 2006; Kragelund et al. 2005, 2006, 2007a, b; Carr et al. 2006) allowed us to link most filamentous bacteria with substrates, electron acceptors and exoenzymes excreted by them and several other aspects.

Studies based on the FISH method revealed the most abundant and frequently occurring groups of filamentous bacteria in Italian, Danish, German and Dutch industrial and municipal WWTPs (van der Waarde et al. 2002; Levantesi et al. 2006; Kragelund et al. 2007a). Alphaproteobacteria have been found in more than half of the tested plants, along with Chloroflexi, Thiothrix sp. and Haliscomenobacter hydrossis. Other Bacteroidetes, different from H. hydrossis, were not so frequent, as well as Actinobacteria and Microthrix parvicella, which were detected only in 16% and 21% of WWTPs, respectively. Some bacteria, like Gordonia amarae, Nostocoida limicola, Leucothrix and Lepthothrix sp., occurred only in a few plants and were detected in 7%, 6%, 4% and 2% of tested samples, respectively. It seems that the composition of the population of filamentous bacteria may depend on geographic location (Chua and Le, 1994). Research by Mielczarek et al. (2012) showed that the percentage of filamentous bacteria in activated sludge was a typical feature of an object and was different among different WWTPs. However, WWTPs in Poland still lack quantitative data about filamentous bacteria. The following survey presents a comprehensive study of the identity and population dynamics of filamentous bacteria in five Polish full-scale municipal WWTPs with nutrients removal.

Aims

The aim of the study was to identify and quantify dominant filamentous bacteria in five full-scale municipal WWTPs in Poland (in the Masovian Province). Plants differed in configuration of reactors and main technological parameters of wastewater treatment process. FISH was used to investigate the population dynamics of filaments. Microbial populations were examined over a 2-year period to describe their temporal variations. All collected data were used to investigate mutual correlations among the abundance of specific filamentous bacteria and wastewater and operational parameters.

MATERIALS AND METHODS

Sampling and WWTP data

The main operational parameters of selected plants are listed in Table 1 (all data provided by the plant operators). The plants ranged in size from 18,000 to 110,000 population equivalents (PE). The fraction of industrial contribution (from food industry mainly) to the organic matter in the influent, depending on the plant, was 0–50%. All the WWTPs had biological N-removal (nitrification and denitrification) and four of them (except WWTP II) also the well-defined enhanced biological phosphorus removal (EBPR) step. In all plants but one (WWTP V) iron-based coagulants (PIX) were dosed to improve phosphorus elimination. In WWTP V polyaluminium chloride compounds (PAX) were occasionally applied, mostly during autumn and winter, to improve settling properties and to control excessive growth of filamentous bacteria. Operational problems connected with settling of activated sludge were reported in all tested plants, but foaming on the tanks was observed only in WWTP IV.

Table 1

Influent and operational parameters of the WWTPs tested in this survey

Parameter WWTP I WWTP II WWTP III WWTP IV WWTP V 
Size designed (PE) 55,400 83,000 163,500 53,040 22,500 
Size actual (PE) 73,400 99,000 110,000 76,000 18,000 
Reactor type A2O A2O AO Anaerobic tank + anoxic/aerobic oxidation ditch (ORP controller) UCT A2O 
Predenitrification Yes No No No Yes 
Presettling Yes Yes Yes Yes No 
Fermenter Yes No No No No 
Aeration Fine bubble diffusers Surface (vertical type) Surface (horizontal rotors) Fine bubble diffusers Fine bubble diffusers 
P-Precipitation Occasionally (PIX) Continuous (PIX) Occasionally (PIX) Occasionally (PIX) No 
PAX dosage No No No No occasionally (winter) 
Wastewater type (% of overall BOD5Domestic Domestic industrial (20–25%): ‘Slaughterhouse dairy’ Domestic industrial (5–10%) ‘fruit & vegetables processing, breweries, landfill leachate’ Domestic industrial (30–50%) ‘fruit & vegetables processing sugar refining’ Domestic 
SVI (mL/g) Summer 240 (184–260)   180 (168–198) 220 (185–240) 90 (80–108) 70 (66–81) 
Winter 320 (243–389)   180 (179–182) 190 (148–224) 140 (121–153) 160 (140–188) 
COD (mg/L) 1,190 (622–3,650) 940 (303–5,880) 1,530 (457–4,000) 1,440 (508–3,480) 1,420 (228–4,730) 
BOD5 (mg/L) 540 (220–1,300) 440 (178–2,770) 600 (330–1,320) 890 (360–1,900) 470 (71–920) 
N total (mg/L) 82 (37–133) 82 (41–115) 105 (10–196) 81 (36–129) 84 (18–133) 
P total (mg/L) 37 (10–124) 13 (7–32) 27 (4–69) 18 (6–64) 21 (5–63) 
BOD5/COD (mg/mg) 0.42 (0.24–0.8) 0.45 (0.31–0.60) 0.46 (0.12–0.95) 0.69 (0.36–0.96) 0.37 (0.04–0.99) 
BOD5/Ntot (mg/mg) 2.8 (1.3–6.3) 4.3 (2.4–10) 5.7 (1.6–46) 11 (2.5–66) 5.4 (0.7–20) 
BOD5/Ptot (mg/mg) 18 (5.7–36) 29 (12–63) 24 (6.8–82) 76 (8–295) 26 (4–115) 
COD/Ntot (mg/mg) 6.8 (3.1–14) 10 (5.0–21) 14 (4–115) 15 (5–107) 17 (4–81) 
COD/Ptot (mg/mg) 44 (16–121) 64 (23–158) 62 (8–375) 97 (13–555) 78 (12–328) 
pH 7.8 (7.5–8.0) 7.5 (7.1–8.2) 7.6 (7.4–8.1) 7.9 (7.0–8.6) 7.5 (7.1–7.7) 
MLSS (g/L) 3.8 (2.0–5.5) 4.8 (2.5–6.8) 5.9 (3.6–8.7) 6.2 (3.4–9.9) 4.0 (1.6–8.4) 
SRT (d) 24 (12–57) 34 (13–94) 37 (10–60) 34 (11–93) 20 (4–82) 
Sludge loading       
(gBOD5/gMLSS/d) 0.03 (0.02–0.07) 0.06 (0.03–0.13) 0.03 (0.01–0.08) 0.06 (0.02–0.16) 0.06 (0.01–0.21) 
(gCOD/gMLSS/d) 0.08 (0.04–0.20) 0.13 (0.05–0.97) 0.08 (0.02–0.22) 0.09 (0.02–0.32) 0.17 (0.01–0.72) 
Comments Severe sludge bulking poor N removal (winter) Moderate sludge bulking sludge overloadings shock loadings unstable N removal Severe sludge bulking landfill leachates and septic tanker trucks: shock loadings Minor sludge bulking severe foaming (winter) industry activity: seasonal sludge overloadings shock loadings Seasonal (winter) sludge bulking seasonal hydraulic overloading 
        
Parameter WWTP I WWTP II WWTP III WWTP IV WWTP V 
Size designed (PE) 55,400 83,000 163,500 53,040 22,500 
Size actual (PE) 73,400 99,000 110,000 76,000 18,000 
Reactor type A2O A2O AO Anaerobic tank + anoxic/aerobic oxidation ditch (ORP controller) UCT A2O 
Predenitrification Yes No No No Yes 
Presettling Yes Yes Yes Yes No 
Fermenter Yes No No No No 
Aeration Fine bubble diffusers Surface (vertical type) Surface (horizontal rotors) Fine bubble diffusers Fine bubble diffusers 
P-Precipitation Occasionally (PIX) Continuous (PIX) Occasionally (PIX) Occasionally (PIX) No 
PAX dosage No No No No occasionally (winter) 
Wastewater type (% of overall BOD5Domestic Domestic industrial (20–25%): ‘Slaughterhouse dairy’ Domestic industrial (5–10%) ‘fruit & vegetables processing, breweries, landfill leachate’ Domestic industrial (30–50%) ‘fruit & vegetables processing sugar refining’ Domestic 
SVI (mL/g) Summer 240 (184–260)   180 (168–198) 220 (185–240) 90 (80–108) 70 (66–81) 
Winter 320 (243–389)   180 (179–182) 190 (148–224) 140 (121–153) 160 (140–188) 
COD (mg/L) 1,190 (622–3,650) 940 (303–5,880) 1,530 (457–4,000) 1,440 (508–3,480) 1,420 (228–4,730) 
BOD5 (mg/L) 540 (220–1,300) 440 (178–2,770) 600 (330–1,320) 890 (360–1,900) 470 (71–920) 
N total (mg/L) 82 (37–133) 82 (41–115) 105 (10–196) 81 (36–129) 84 (18–133) 
P total (mg/L) 37 (10–124) 13 (7–32) 27 (4–69) 18 (6–64) 21 (5–63) 
BOD5/COD (mg/mg) 0.42 (0.24–0.8) 0.45 (0.31–0.60) 0.46 (0.12–0.95) 0.69 (0.36–0.96) 0.37 (0.04–0.99) 
BOD5/Ntot (mg/mg) 2.8 (1.3–6.3) 4.3 (2.4–10) 5.7 (1.6–46) 11 (2.5–66) 5.4 (0.7–20) 
BOD5/Ptot (mg/mg) 18 (5.7–36) 29 (12–63) 24 (6.8–82) 76 (8–295) 26 (4–115) 
COD/Ntot (mg/mg) 6.8 (3.1–14) 10 (5.0–21) 14 (4–115) 15 (5–107) 17 (4–81) 
COD/Ptot (mg/mg) 44 (16–121) 64 (23–158) 62 (8–375) 97 (13–555) 78 (12–328) 
pH 7.8 (7.5–8.0) 7.5 (7.1–8.2) 7.6 (7.4–8.1) 7.9 (7.0–8.6) 7.5 (7.1–7.7) 
MLSS (g/L) 3.8 (2.0–5.5) 4.8 (2.5–6.8) 5.9 (3.6–8.7) 6.2 (3.4–9.9) 4.0 (1.6–8.4) 
SRT (d) 24 (12–57) 34 (13–94) 37 (10–60) 34 (11–93) 20 (4–82) 
Sludge loading       
(gBOD5/gMLSS/d) 0.03 (0.02–0.07) 0.06 (0.03–0.13) 0.03 (0.01–0.08) 0.06 (0.02–0.16) 0.06 (0.01–0.21) 
(gCOD/gMLSS/d) 0.08 (0.04–0.20) 0.13 (0.05–0.97) 0.08 (0.02–0.22) 0.09 (0.02–0.32) 0.17 (0.01–0.72) 
Comments Severe sludge bulking poor N removal (winter) Moderate sludge bulking sludge overloadings shock loadings unstable N removal Severe sludge bulking landfill leachates and septic tanker trucks: shock loadings Minor sludge bulking severe foaming (winter) industry activity: seasonal sludge overloadings shock loadings Seasonal (winter) sludge bulking seasonal hydraulic overloading 
        

Description: BOD – biological oxygen demand; COD – chemical oxygen demand; ORP – oxidation-reduction potential; PE – population equivalent; PAX – polyaluminium chloride-based coagulants; PIX – iron based coagulants; SVI – sludge volume index; MLSS – mixed liquor suspended solids; SRT – sludge retention time; UCT – University of Cape Town; process configurations A2O – anaerobic–anoxic–aerobic, AO – anaerobic–aerobic.

Mean values and ranges (in parentheses) of influent parameters are shown.

Activated sludge samples had been collected twice a year in the period from September 2011 to March 2013 (in the beginning of March and in the end of September) from the aerobic process tank, and kept on ice until the fixation for FISH analyses (Nielsen et al. 2009b).

FISH identification

FISH analyses were performed according to Nielsen et al. (2009a). The 6-Fam labelled EUBmix oligoprobe (equimolar mixture of EUB338, EUB338II and EUB338III) was used to target the entire bacterial community. Filamentous bacteria were identified with a wide selection of 11 oligoprobes: CFXmix (equimolar concentration of GNSB-941 and CFX1223 probes, targeting phylum Chloroflexi), T0803 and CHL1851 (types 0803 and 1851, respectively, in phylum Chloroflexi), MPAmix (equimolar concentration of MPA645, MPA223 and MPA60, targeting Candidatus ‘Microthrix parvicella’ and CandidatusM. calida’), G123 T (Thiothrix eikelboomii, T. nivea, T. unzii, T. fructosivorans, T. defluvii, Eikelboom type 021N group I, II, III), Myc657 (Mycobacterium subdivision mycolata), Spin1449 (Skermania piniformis), Gor596 (family Gordonia), HHY654 (Haliscomenobacter hydrossis), Curvi997 (types 1701 and 0041/0675 of Curvibacter spp.), TM7905 (type 0041/0675 in Candidate division TM7). The specific probes were labelled with Cy3. This selection of probes was based on the recently evaluated probes for filamentous bacteria detected in activated sludge plants (Nielsen et al. 2009b; Mielczarek et al. 2012). When applicable, a hierarchical approach had been used, i.e. the more general probe was applied first and then the more specific one (e.g. Myc657 followed by Spin1449 and Gor596). Detailed information about the probes used is given in probeBase (Loy et al. 2003).

Quantification procedures were performed similarly to Mielczarek et al. (2012). Twenty separate images for each probe were captured with a Nikon Eclipse 50i microscope. ImageJ software (Collins 2007) was used to determine the biovolume of bacteria, which was relative to the pixel area of cells positive for the specific probe. The microbial abundance (expressed as % of EUBmix probe) was then quantified as a percentage of the pixel area for all bacteria positive for the EUBmix probe and was calculated as a mean of 20 separate measurements. The standard error of the mean was calculated as a standard deviation of the percentage abundance of the specific bacteria divided by a square root of 20 measurements.

Statistical measures and methods

To find the strength of relationship among quantified filamentous bacteria populations, correlation analyses (with Pearson product-moment correlation coefficient and Spearman's rank correlation coefficient) have been performed. Differentiating factors have been searched by analysis of variance (ANOVA) and multivariate analysis of variance (MANOVA) with significance level alpha 0.05. Standard statistical comparisons and graphing were performed in Microsoft Excel, correlation analyses were performed in STATISTICA™ from StatSoft®.

RESULTS

Identity and abundance of filamentous bacteria

Filamentous bacteria were abundant in all tested plants and constituted on average 28 ± 3% of all bacteria identified by EUBmix probe. The structure of filamentous bacteria community in the tested plants was similar to each other. Most often the filaments belonged to phylum Chloroflexi and Microthrix morphotype. The average abundance of phylum Chloroflexi in the tested WWTPs was 14% (expressed as a percent of EUBmix) and ranged between 3.4 and 3.5% (Figure 1). Phylum Chloroflexi was further investigated with more specific probes, T0803 and CHL1851, for abundance of types 0803 and 1851, respectively. Types 0803 and 1851 accounted for 14 and 1.4% of the whole population of Chloroflexi, respectively, whereas the rest of this phylum (about 85%) was not further identified (Figure 1).

Figure 1

The filamentous bacteria community in activated sludge of five Polish full-scale WWTPs in the period September 2011–March 2013. Abundance of individual populations was determined by quantitative fluorescence in situ hybridization with specific oligoprobes against the EUBmix probe. Bacteria belonging to Skermania piniformis, which accounted for a minor fraction, are not shown.

Figure 1

The filamentous bacteria community in activated sludge of five Polish full-scale WWTPs in the period September 2011–March 2013. Abundance of individual populations was determined by quantitative fluorescence in situ hybridization with specific oligoprobes against the EUBmix probe. Bacteria belonging to Skermania piniformis, which accounted for a minor fraction, are not shown.

In all of the tested WWTPs, bacteria classified to morphotype Microthrix were found. It was the most abundant genus among the determined filamentous bacteria and constituted on average more than 7% of the total bacterial population (varied from 1.8 to 15% among the WWTPs).

Another species with a significant contribution to overall filamentous community was H. hydrossis (average 3.6%, range 2.2–5.4%). When the number of Microthrix was dropping, the abundance of H. hydrossis increased from 0.6 to even 12%. In most sample types 0041/0675 of TM7 were also observed, with an average abundance of 0.8% (Figure 1). Mycolata and Skermania piniformis were detected only in WWTP II and III, respectively, where they constituted on average 1.6 and 1% of the bacterial community. Curvibacter (types 1701 and 0041/0675), Thiothrix/021N and family Gordonia have not been found in any of the tested samples.

Seasonal variations

Generally, the percentage of filamentous bacteria was higher after the winter season than after summer. Some small changes in the abundance of probe-defined populations in all five plants were observed during the survey (2011–2013). However, composition of the populations was rather stable in each plant despite the differences in abundance of specific bacteria (Figure 1). For instance, in WWTPs II and III the relative abundance of Chloroflexi in the filamentous bacterial population was consistently high, whereas in WWTPs I and V it was low most of the time (Figure 1). Some seasonal variation in the abundance of Microthrix was observed and confirmed by ANOVA. Microthrix was more abundant during winter (in samples collected in March). No other statistically significant seasonal pattern had been discovered for the remaining probe-defined populations.

Figure 2

Relative composition of filamentous bacterial population in the tested plants (average for 2011–2013).

Figure 2

Relative composition of filamentous bacterial population in the tested plants (average for 2011–2013).

Differences among treatment plants and intercept effect of two differentiating factors

The only statistically significant influence of WWTP on bacterial populations had been found for Chloroflexi, which was the most abundant filamentous bacteria phylum. To find out whether or not each plant had a unique filamentous population, MANOVA was carried out on all samples for all seasons. The filamentous communities were considered as a multivariate response, with treatment plant and season as explanatory variables. It proved that those two variables differentiated populations of Microthrix and Chloroflexi. The analysis showed also a tendency level for HHY654 (p = 0.093), which could be significant in further studies based on larger sample quantity.

Correlation analyses

Spearman's and Pearson's coefficients were tested to reveal mutual correlations between bacterial groups. No statistical dependence has been shown by Spearman's rank correlation coefficient. There were some weak correlations, but they were not statistically significant (Figure 3). One medium correlation (r > 0.5 with a significance level alpha = 0.011) between Microthrix and Skermania piniformis (identified by probes MPAmix and Spin1449, respectively) was obtained by the Pearson's correlation coefficient analysis. Other correlations, despite being strong, did not achieve proper significance level alpha (>0.1). Strong linear correlation has been found between sludge volume index (SVI) and probe-define population of Microthrix (r = 0.759 with a significance level alpha = 0.011). No strong relationships between the abundance of filamentous bacteria and biological oxygen demand (BOD), chemical oxygen demand (COD), total N or total P were found. Statistical analyses did not reveal meaningful correlations with other operational parameters (data not shown), similarly to the results obtained by Mielczarek et al. (2012) for Danish plants.

Figure 3

Spearman's (A) and Pearson's (B) correlation analyses showing strength of mutual relations between probe-defined groups of filaments (2011–2013).

Figure 3

Spearman's (A) and Pearson's (B) correlation analyses showing strength of mutual relations between probe-defined groups of filaments (2011–2013).

DISCUSSION

Bulking problems are reported worldwide and usually they are associated with extensive growth of filamentous microorganisms (Eikelboom 2000; van der Waarde et al. 2002; Jenkins et al. 2004; Mielczarek et al. 2012). This is the first long-term survey of the filamentous bacteria community in Polish municipal full-scale WWTPs, which was carried out with molecular identification by FISH. Recent studies have proved that the vast majority of filamentous bacteria in WWTPs could be identified by existing FISH probes (Kragelund et al. 2011; Mielczarek et al. 2012). This survey was carried out using a wide selection of 11 oligoprobes, which are widely applied for identification of those bacteria (Nielsen et al. 2009a, Kragelund et al. 2011; Mielczarek et al. 2012). In Polish WWTPs, filamentous communities constituted on average 28 ± 3% of the entire biomass as analysed by EUBmix, which is similar to the results obtained by Nielsen et al. (2010) and Mielczarek et al. (2012) for Danish plants (28% and 24%, respectively).

We showed that the structure of filamentous community in Polish WWTPs is similar to those found in other European countries, as could be anticipated due to similar climate (Wanner et al. 2009). Most filamentous bacteria detected in Polish WWTPs belonged to phyla Chloroflexi, Actinobacteria (Microthrix and Mycolata), TM7 and Bacteroidetes (H. hydrossis). However, unlike the countries mentioned in the Introduction, Microthrix parvicella was more frequent (present in all tested plants), whereas Thiothrix and filamentous Curvibacter (both of phylum Proteobacteria) were not found. The most abundant, Chloroflexi and Microthrix, on average, accounted for 14% and 7% of all bacteria, respectively. Similar results were obtained by Nielsen et al. (2010) and Mielczarek et al. (2012) for Danish municipal WWTPs with nutrients removal; Chloroflexi and Microthrix constituted in Denmark about 10% and 5%–6% of biomass targeted by EUBmix probe, respectively. Those two bacteria are the most frequently reported filaments causing bulking in activated sludge systems treating primarily municipal wastewater (Wanner et al. 2009). In previous studies, filamentous Chloroflexi was often the most abundant phylum of the filamentous organisms and constituted 1–25% (Beer et al. 2006), 8% (Kong et al. 2007) or even up to 30% of the entire bacteria population in nutrients removal plants (Morgan-Sagastume et al. 2008; Nielsen et al. 2010; Kragelund et al. 2011). Chloroflexi and Microthrix constituted the great majority (64–81%) of filamentous bacteria investigated in this survey (Figure 2). These two groups of filaments utilize different substrates – the latter takes up lipids whereas the former utilizes proteins, polysaccharides and dead cell debris. They use different substrates – there is no competition for source of energy between them, because they occupy different niches and therefore may co-dominate in sludge (Nielsen et al. 2009a).

Bacteria belonging to phylum Chloroflexi dominated in almost all of the plants investigated in this study (except WWTP I). The impact of Chloroflexi filaments on sludge floc structure is species dependent. Some bacteria from this phylum serve as backbones to which other microorganisms can adhere and form strong and dense flocs. Other morphotypes are located at the edge of the sludge flocs and protrude out into the bulk liquid, which has a negative impact on sludge settling properties (Miura et al. 2007; Kragelund et al. 2007a; Speirs et al. 2009; Nielsen et al. 2009a). The broad phylum probe (CFXmix) covers several different morphotypes, which could be investigated with more specific probes e.g. type 0803 (genus Caldilinea, Kragelund et al. (2011)) and type 1851 (close relative of Roseiflexus castenholzii, in class Chloroflexi, Beer et al. (2002)), targeted by T0803 and CHL1851, respectively. Out of these two morphotypes, the more abundant one in Polish WWTPs is type 0803, which accounted for up to 9% of all bacteria (an average abundance of 2% in the tested WWTPs). Type 0803 is common in municipal WWTPs; its straight filaments are mainly inside the flocs, but sometimes it can occur in rosettes and cause bulking (Eikelboom 2000; Jenkins et al. 2004). Type 0803 may worsen the settling properties especially during winter, when it proliferates and is responsible for creating open-floc structures (Kragelund et al. 2011). An average abundance of type 0803 in Danish full-scale municipal WWTPs is about 2.6–2.9% (Kragelund et al. 2011; Mielczarek et al. 2012), which corresponds well with the results obtained for Polish WWTPs (average 2% of EUBmix). Type 1851 is less abundant in both Polish (0–1.7%) and Danish WWTPs (average 0.5% – Kragelund et al. (2011)). Definitely, the more common population of phylum Chloroflexi is type 0092, which can reach up to 10% of all bacteria in sludge (average 3.6%) (Kragelund et al. 2011; Mielczarek et al. 2012). Interestingly, type 0092 does not respond to the EUBmix probes (Speirs et al. 2009; Nielsen et al. 2009b). Speirs et al. (2009) used probes CFX197 and CFX223, and showed that this morphotype is very common in both EBPR and non-EBPR full-scale plants in Australia. It should be stressed that for the time being about one-third of Chloroflexi cannot be identified beyond phylum level (Mielczarek et al. 2012). Unidentified population of Chloroflexi in our research was even higher (about 85%), but we did not use probes targeting type 0092, which can constitute above one-third of bacteria targeted by broad probe CFXmix (Mielczarek et al. 2012).

Microthrix can assimilate long-chain fatty acids under both aerobic and anaerobic conditions (Nielsen et al. 2009a); therefore it occurs exclusively in activated sludge plants with EBPR process (anaerobic–aerobic (AO) biomass cycling), where it is associated with both bulking and foaming incidents (Eikelboom 2000; Jenkins et al. 2004). Microthrix was the most abundant genus (above 7% of the total bacterial population) among the determined filamentous bacteria in this survey. However, in some EBPR WWTPs, which suffered from severe sludge bulking, abundance of these bacteria occasionally exceeded 20% of the entire biomass as analysed by EUBmix probe. It was correlated with SVI, which is easy to explain by the morphology and occurrence of CandidatusM. parvicella’ and C. ‘M. calida’; these characteristic coiled and twisted filaments are long, thin and usually present around flocs. Mielczarek et al. (2012) showed that Microthrix was present in all samples from 28 Danish WWTPs and its abundance varied substantially among the plants (0.3–16% of EUBmix), but it was the most abundant during winter and spring. Similar results, confirmed by ANOVA, were obtained in this study and in earlier observations (Levantesi et al. 2006; Nielsen et al. 2009a; Kragelund et al. 2011). Microthrix constituted 11–45% of all filaments in this survey, which was in some cases substantially more than in Danish WWTPs (11–22%, Mielczarek et al. 2012). The temporal changes in composition of probe-defined populations in the individual plants were minor and no statistically significant seasonal variations were noticed for other filaments, similarly to research carried out by Mielczarek et al. (2012). Another common probe-defined population in this study was H. hydrossis. These filaments accounted for 3.6% of the bacterial biovolume, which is higher than their abundance in Danish WWTPs (average 1.9%) revealed by Nielsen et al. (2010) and Mielczarek et al. (2012). We showed that when the number of Microthrix was dropping the abundance of H. hydrossis increased from 0.6% to even 12%.

Other filamentous bacteria investigated in this survey, which altogether accounted for 19–36% of filamentous community, seem to play a minor role in activated sludge. The ranges of their abundance corresponded with the respective results obtained by Nielsen et al. (2010) and Mielczarek et al. (2012), with exceptions for TM7 and Curvibacter. TM7 were less abundant (<1%) than in Danish WWTPs (5%), whereas filamentous Curvibacter, constituting on average 0.7% of the bacterial biovolume as described by Mielczarek et al. (2012), were not detected in this study.

Microbial populations were analysed by a few statistical approaches to show possible rules governing them. Mutual correlations have been searched for, but no significant general correlations have been found between the presence of different filaments. Medium correlation strength, showed by Spearman's rank correlation coefficient, was recorded only between Microthrix and Skermania piniformis. Similar analysis, carried out by Mielczarek et al. (2012), showed no strong or even medium correlation between bacteria in Danish WWTPs. In our study, we found differences in filamentous populations among treatment plants shown by ANOVA, but only for Chloroflexi. We were not able to confirm that each treatment plant had unique population composition throughout the period of FISH analyses, as was shown by Mielczarek et al. (2012). The relative abundance of filamentous species in individual plants was rather similar. Small differences between plants were noticed, mostly in the case of Chloroflexi and Microthrix (shown by MANOVA). The conclusion that filamentous community is relatively constant and similar in different plants during long-term observation contradicts the assumption of Bellucci & Curtis (2011) about chaotic microbial composition of heterotrophic bacteria despite a constant function. However, studies of Bellucci & Curtis (2011) were carried out in laboratory-scale reactors and it may be misleading to extrapolate results from laboratory-scale reactors to full-scale plants. As was shown by Muszyński et al. (2012), some of the populations identified in laboratory-scale reactors are detected mainly in those conditions and seem to play a questionable role in full-scale EBPR systems. That is why application of knowledge about the structure and functions of activated sludge community, obtained from laboratory-scale studies, can be unreliable when applied to work out the bulking control methods in full-scale WWTPs.

This is the first molecular, long-term study on the presence of filamentous bacteria in full-scale WWTPs in Poland. Prior to this study filaments were identified only to morphotypes, using conventional light microscopy. In this paper, an attempt was made to find connection between abundance of filamentous bacteria with operational data by statistical method for not only linear, but also non-linear correlations. The results of this study must be considered as reliable only on the assumption that FISH probes are sufficiently specific and target single (or closely related) species or ecotypes. Furthermore, FISH analysis does not provide enough resolution to observe fine-scale variability from plant to plant. Only methods with deep sequencing of full genomes or metagenomes can show microdiversity of bacterial groups over time or between the plants (Albertsen et al. 2011). It would also be interesting to compare the results obtained by FISH with information about microbial community structure collected with stable-isotope probing, which has been successfully used for in situ identification of denitrifying bacteria (Ginige et al. 2004), fatty acid-degrading anaerobic bacteria (Hatamoto et al. 2007) or fermenting bacteria (Nielsen et al. 2012).

CONCLUSIONS

  1. Filamentous bacteria are abundant in activated sludge of Polish full-scale WWTPs and constitute on average 28% of all bacteria. The structure of filamentous community is similar to those found in other European countries.

  2. The most abundant filamentous bacteria belong to phylum Chloroflexi and genus Microthrix (targeted by probes CFXmix and MPAmix, respectively).

  3. The relative abundance of filamentous species in individual plants is rather similar. Small differences between plants were noticed, mostly in terms of Chloroflexi and Microthrix (shown by MANOVA).

  4. No significant general correlations have been found between the presence of different filaments. Medium correlation strength was recorded only between Microthrix and Skermania piniformis (identified by probes MPAmix and Spin1449, respectively).

  5. The strong linear correlation between Microthrix and SVI confirms deleterious effect of these bacteria on settling properties of sludge. It is associated with the characteristic morphology of these twisted filaments producing a diffuse open-floc structure and protruding from flocs.

ACKNOWLEDGEMENTS

The study was financially supported by the Polish National Science Centre (grant No. N N523 736540). We would like to thank Ewa Łukomska for technical help with preparation of activated sludge samples.

REFERENCES

REFERENCES
Albertsen
M.
Hansen
L. B.
Saunders
A. M.
Nielsen
P. H.
Nielsen
K. L.
2011
A metagenome of a full-scale microbial community carrying out enhanced biological phosphorus removal
.
ISME J.
6
,
1094
1106
.
Bellucci
M.
Curtis
T. P.
2011
Research on nitrification and related processes
.
Meth. Enzymol.
496
(
B
),
269
286
.
Carr
E. L.
Eales
K. L.
Seviour
R. J.
2006
Substrate uptake by Gordonia amarae in activated sludge foams by FISH-MAR
.
Water Sci. Technol.
54
(
1
),
39
45
.
Chua
H.
Le
K. Y.
1994
A survey of filamentous foaming in activated sludge plants in Hong Kong
.
Water Sci. Technol.
30
(
11
),
251
254
.
Collins
T. J.
2007
ImageJ for microscopy
.
Biotechniques
43
,
25
30
.
Eales
K.
Nielsen
J. L.
Kragelund
C.
Seviour
R.
Nielsen
P. H.
2005
The in situ physiology of pine tree like organisms (PTLO) in activated sludge foams.
Acta Hydrochim. Et Hydrobiol.
33
(
3
),
203
209
.
Eales
K. L.
Nielsen
J. L.
Seviour
E. M.
Nielsen
P. H.
Seviour
R. J.
2006
The in situ physiology of Skermania piniformis in foams in Australian activated sludge plants
.
Environ. Microbiol.
8
(
10
),
1712
1720
.
Eikelboom
D. H.
2000
Process Control of Activated Sludge Plants by Microscopic Investigation
.
IWA Publishing
,
London
.
Jenkins
D.
Richard
M.
Daigger
G.
2004
Manual on the Causes and Control of Activated Sludge Bulking, Foaming and Other Solids Separation Problems
.
IWA Publishing, CRC Press
,
London
.
Kragelund
C.
Nielsen
J. L.
Thomsen
T. R.
Nielsen
P. H.
2005
Ecophysiology of the filamentous Alphaproteobacterium Meganema perideroedes in activated sludge
.
FEMS Microbiol. Ecol.
54
(
1
),
111
122
.
Kragelund
C.
Kong
Y.
van der Waarde
J.
Thelen
K.
Eikelboom
D.
Tandoi
V.
Thomsen
T. R.
Nielsen
P. H.
2006
Ecophysiology of different filamentous Alphaproteobacteria in industrial wastewater treatment plants
.
Microbiology
152
(
10
),
3003
3012
.
Kragelund
C.
Levantesi
C.
Borger
A.
Thelen
K.
Eikelboom
D.
Tandoi
V.
Kong
Y.
van der Waarde
J.
Krooneman
J.
Rossetti
S.
Thomsen
T. R.
Nielsen
P. H.
2007a
Identity, abundance and ecophysiology of filamentous Chloroflexi species present in activated sludge treatment plants
.
FEMS Microbiol. Ecol.
59
(
3
),
671
682
.
Kragelund
C.
Remesova
Z.
Nielsen
J. L.
Thomsen
T. R.
Eales
K.
Seviour
R.
Wanner
J.
Nielsen
P. H.
2007b
Ecophysiology of mycolic acid-containing Actinobacteria (Mycolata) in activated sludge foams
.
FEMS Microbiol. Ecol.
61
(
1
),
174
184
.
Kragelund
C.
Thomsen
T. R.
Mielczarek
A. T.
Nielsen
P. H.
2011
Eikelboom's morphotype 0803 in activated sludge belongs to the genus Caldilinea in the phylum Chloroflexi
.
FEMS Microbiol. Ecol.
76
(
3
),
451
462
.
Levantesi
C.
Rossetti
S.
Beimfohr
C.
Thelen
K.
Krooneman
J.
van der Waarde
J.
Tandoi
V.
2006
Description of filamentous bacteria present in industrial activated sludge WWTPs by conventional and molecular methods
.
Water Sci. Technol.
54
,
129
137
.
Mielczarek
A. T.
Kragelund
C.
Eriksen
P. S.
Nielsen
P. H.
2012
Population dynamics of filamentous bacteria in Danish wastewater treatment plants with nutrient removal
.
Water Res.
46
(
12
),
3781
3795
.
Morgan-Sagastume
F.
Larsen
P.
Nielsen
J. L.
Nielsen
P. H.
2008
Characterization of the loosely attached fraction of activated sludge bacteria
.
Water Res.
42
,
843
854
.
Muszyński
A.
Łebkowska
M.
Tabernacka
A.
Miłobędzka
A.
2013
From macro to lab-scale: changes in bacterial community led to deterioration of EBPR in lab reactor
.
Cent. European J. Biol.
8
(
2
),
130
142
.
Nielsen
P. H.
Kragelund
C.
Seviour
R. J.
Nielsen
J. L.
2009a
Identity and ecophysiology of filamentous bacteria in activated sludge
.
FEMS Microbiol. Rev.
33
,
969
998
.
Nielsen
J. L.
Nguyen
H.
Meyer
R. L.
Nielsen
P. H.
2012
Identification of glucose-fermenting bacteria in a full-scale enhanced biological phosphorus removal plant by stable isotope probing
.
Microbiology (Reading, England)
158
(
Pt
7), 1818–1825.
Nielsen
P. H.
Daims
H.
Lemmer
H.
(Eds.)
2009b
FISH Handbook for Biological Wastewater Treatment
.
IWA Publishing
,
London
.
Nielsen
P. H.
Mielczarek
A. T.
Kragelund
C.
Nielsen
J. L.
Saunders
A. M.
Kong
Y.
Hansen
A. A.
Vollertsen
J.
2010
A conceptual ecosystem model of microbial communities in enhanced biological phosphorus removal plants
.
Water Res.
44
(
17
),
5070
5088
.
Speirs
L.
Nittami
T.
McIlroy
S.
Schroeder
S.
Seviour
R. J.
2009
Filamentous bacterium Eikelboom type 0092 in activated sludge plants in Australia is a member of the phylum Chloroflexi
.
Appl. Environ. Microbiol.
75
,
2446
2452
.
van der Waarde
J.
Krooneman
J.
Geurkink
B.
van der Werf
A.
Eikelboom
D.
Beimfohr
C.
Snaidr
J.
Levantesi
C.
Tandoi
V.
2002
Molecular monitoring of bulking sludge in industrial wastewater treatment plants
.
Water Sci. Technol.
46
(
1–2
),
551
558
.
Wanner
J.
Kragelund
C.
Nielsen Per
H.
2009
Microbiology of bulking
. In:
Microbial Ecology of Activated Sludge
(
Seviour
R. J.
Nielsen
P. H.
eds).
IWA Publishing
,
London
, pp.
191
214
.