Aeromonas Species from Nonchlorinated Distribution Systems and Their Competitive Planktonic Growth in Drinking Water

The occurrence of the bacterial genus Aeromonas in nonchlorinated drinking water in the Netherlands is regarded as an indication of elevated microbial regrowth in the distribution system. Identification of the prevalent species in 10 distribution systems by genotyping yielded seven different species, with A. rivuli, A. veronii, and A. sobria being the most dominant ones.

ABSTRACT

Aeromonas is included in the Dutch Drinking Water Decree as an indicator for elevated microbial regrowth in nonchlorinated drinking water distribution systems (DWDS). The temporal and spatial diversity of Aeromonas species in 10 DWDS and their planktonic growth characteristics with regard to different carbon sources were investigated. Genotyping of the gyrB gene of isolates showed a nonsystematic temporal and spatial variable prevalence of seven different Aeromonas species in these DWDS and no correlation with the level of assimilable organic carbon determined with Pseudomonas fluorescens strain P17 and Spirillum sp. strain NOX (AOC-P17/NOX concentration) and Aeromonas concentrations. Pure cultures of these seven species showed a high affinity for low concentrations (micrograms per liter) of individual amino acids and fatty acids, compounds associated with biomass. Growth occurred at 0.5 µg C/liter of an amino acid mixture. Growth of a mixed community of Aeromonas rivuli, Aeromonas salmonicida, Aeromonas sobria, and Aeromonas veronii in drinking water occurred in pasteurized samples; however, either no growth or decay occurred in competition with the autochthonous bacteria (nonpasteurized samples). This community also failed to grow in nonpasteurized distribution samples from a location with a clear increase in planktonic Aeromonas concentrations in the transported drinking water. For competitive planktonic growth of Aeromonas, an amino acid concentration of ≥5 µg C/liter is required. AOC-P17/NOX concentrations showed that such concentrations are not expected in Dutch drinking water. Therefore, we suspect that competitive planktonic growth is not the major cause of the observed noncompliance with the Aeromonas standard in nonchlorinated DWDS.

IMPORTANCE The occurrence of the bacterial genus Aeromonas in nonchlorinated drinking water in the Netherlands is regarded as an indication of elevated microbial regrowth in the distribution system. Identification of the prevalent species in 10 distribution systems by genotyping yielded seven different species, with A. rivuli, A. veronii, and A. sobria being the most dominant ones. Planktonic growth experiments of pure cultures confirmed former published affinity of Aeromonas for certain biomass compounds (amino and fatty acids). In competition with the autochthonous microflora, however, planktonic growth was observed only after addition of a threshold amino acid concentration of 5 µg C/liter. Based on our results and further observations, we deduced that planktonic growth of Aeromonas in the DWDS is not very likely. Benthic growth in loose deposits and planktonic release are a more plausible explanation for the observed planktonic increase of Aeromonas.

INTRODUCTION

Aeromonas spp. are ubiquitous in river and freshwater lakes (1–4) and have frequently been observed in drinking water systems (5–7). An interest in Aeromonas in nonchlorinated drinking water in the Netherlands was initiated from the 1980s, after the observation of a sudden increase of Aeromonas numbers in drinking water at the municipal Dune Waterworks of The Hague in 1984 (8). These aeromonads were observed as atypical colonies on Teepol medium for coliform detection, an observation that has been documented in literature before (9). This observation coincided with publications on the possible causal relationship between Aeromonas-associated diarrhea and Aeromonas presence in drinking water, as suggested by Australian researchers (10). Subsequently, extensive studies with phenotyping and genotyping methods demonstrated that Aeromonas isolates from fresh and drinking water environments were phenotypically and genotypically different from Aeromonas isolates from patients (11–15). In response to these studies, the Environmental Protection Agency (EPA) in the United States removed Aeromonas from the contaminant candidate list (CCL) in 2009 (16). In the Netherlands, the presence of Aeromonas in drinking water is currently not considered a health-related problem (15).

Studies on Aeromonas spp. in nonchlorinated drinking water in the Netherlands in the 1980s showed that the number of aeromonads varied between drinking water distribution systems (DWDS) (7). Numbers ranged from <1 to a few hundred CFU per 100 ml on ampicillin-dextrin agar. In addition, the Aeromonas colonies obtained were classified to the species level by biochemical characterization. Based on these biochemical characteristics, Aeromonas strains were classified as Aeromonas hydrophila, Aeromonas caviae, and Aeromonas sobria (15, 17). Nowadays, genotyping methods are included in taxonomic classification (18), and as a result, it remains unknown whether the use of genotyping methods affects the classification of Aeromonas species in DWDS in the Netherlands.

The nutritional versatility of aeromonads in drinking water has been studied before. Although bacteria require carbon (C), nitrogen (N), and phosphorus (P) as nutrients for growth, carbon is usually regarded as the growth-limiting factor for many heterotrophic microbiological processes, including growth of Aeromonas. It was shown that aeromonads have a high affinity for low concentrations (at the level of micrograms of C per liter) of certain organic carbon compounds from biomass (i.e., amino acids and long-chain fatty acids) available in the DWDS (8, 19, 20). Furthermore, aeromonads preferentially use ammonium instead of nitrate as their nitrogen source (8). These studies on the growth characteristics of pure cultures of Aeromonas in drinking water were done with Aeromonas strains that were biochemically identified as A. hydrophila (8). It was suggested that the high substrate affinity of A. hydrophila for amino acids (arginine) and long-chain fatty acids (oleate) (8) would mean that A. hydrophila competes successfully for these compounds in the DWDS. However, it remains unknown (i) whether these growth characteristics can be generalized to all Aeromonas species occurring in DWDS and (ii) whether DWDS-associated Aeromonas spp. can successfully compete for these compounds in the DWDS.

Based on the results from the previous Aeromonas research in the Netherlands and the precautionary approach for the safety of nonchlorinated drinking water supply, the authorities in the Netherlands included Aeromonas in the Dutch Drinking Water Decree as an additional indicator (beside heterotrophic plate count [HPC]) for microbial regrowth in the DWDS (21). Microbial regrowth occurs in the DWDS and is higher in nonchlorinated drinking water systems when nutrient levels in the drinking water are higher. An operational Aeromonas standard of 1,000 CFU/100 ml, obtained by growth on specific ampicillin-dextrin agar at 30°C (7), in drinking water is specified in the Dutch Drinking Water Decree. When drinking water companies do not comply with this standard, they have to minimize the growth conditions. A recent study on indicator parameters for regrowth concluded that HPCs and aeromonads are more reliable indicators for regrowth in the DWDS in the Netherlands than ATP and bacterial cell numbers (22). Another field study in the Netherlands showed that noncompliance with the Aeromonas standard in two DWDS coincided with increased HPCs (within the limits of the Drinking Water Decree), occasional coliform regrowth, and enhanced numbers of macroinvertebrates (e.g., water lice) in the DWDS (23). The results from these two studies, thus, show that Aeromonas is still useful as a regrowth indicator in nonchlorinated DWDS. However, Aeromonas is only a minor part (<0.01%) of the diverse autochthonous microflora that is present in DWDS (22, 24). This suggests that aeromonads utilize only a minor fraction of the available nutrients in drinking water for planktonic growth. Furthermore, it has been observed that Aeromonas isolates are mainly associated with sediment in the DWDS and to a lesser extent with drinking water, but not with the biofilm on the pipe wall (25, 26), demonstrating that sediment or loose deposits (consisting of small and larger [in]organic and biological suspended solids, including invertebrates) are the main niche for Aeromonas in the DWDS.

The objectives of our research were (i) to study the temporal and spatial dynamics of Aeromonas species in the DWDS of five treatment plants in the Netherlands which do not comply with the Dutch Aeromonas standard and five plants which do comply, (ii) to determine whether pure cultures of the different Aeromonas species isolated from the DWDS differ in their nutritional versatility in drinking water and how this compares to the nutritional versatility published for A. hydrophila (8), and (iii) to study the ability of these Aeromonas species to compete with the autochthonous microbial population for planktonic growth in the DWDS.

RESULTS

Identification of Aeromonas strains isolated from drinking water distribution systems.

In the summers of 2012 and 2015, the DWDS of 10 different treatment plants in the Netherlands (using surface or groundwater as source water) were sampled multiple times. The source of drinking water, assimilable organic carbon (AOC) concentration in the treated water, and compliance with the Aeromonas standard in the distribution system for each plant are given in Table 1. In total, 231 (2012) and 216 (2015) Aeromonas strains were isolated from drinking water and identified to the species level based on the gyrB gene sequence. In 2012, 12 of the 231 Aeromonas isolates could not be matched to the reference database and were omitted from further analyses (see Table S1 in the supplemental material). The other 219 isolates were classified among seven different Aeromonas species (A. hydrophila, A. punctata, A. salmonicida, A. bestiarum, A. media, A. sobria, A. veronii, and A. rivuli) (Fig. 1). A. rivuli was the most frequently isolated Aeromonas species in 2012 (74 isolates; 32% of the isolates), whereas A. sobria was found in most DWDS (8 of 9 DWDS; 16% of the isolates). In 2015, 8 of the 216 Aeromonas isolates could not be identified to the species level and were omitted from further analyses. The other 208 isolates were classified among seven Aeromonas species (A. hydrophila, A. salmonicida, A. bestiarum, A. media, A. sobria, A. rivuli, and A. veronii) (Fig. 1). A. rivuli (31%) and A. veronii (33%) were the most frequently isolated Aeromonas species in 2015, whereas A. veronii and A. sobria were found in most DWDS (4 of 5 DWDS). In addition, the Aeromonas population is variable in all DWDS, and the Aeromonas species composition within a DWDS varied between 2012 and 2015 at two of the five sampled DWDS (Table S1). The most dominant Aeromonas species were observed in drinking water produced from surface water and groundwater, indicating that the drinking water source does not impact the Aeromonas species composition. In addition, most Aeromonas species were observed in 2012 and 2015, demonstrating that year-to-year fluctuations in Aeromonas species seem to be low.

TABLE 1.

Drinking water distribution systems (DWDS) of 12 different drinking water treatment plants from which Aeromonas species were isolated

Plant	Source	Avg AOC concn (µg C/liter)		Aeromonas compliancea
		P17/NOXb	P17c
1	Surface water—reservoir	13.7	0.1	−
2	Surface water—reservoir	10.2	0.1	+
3	Surface water—reservoir	15.5	0.1	−
4	Groundwater/surface water—reservoir			−
5	Surface water—reservoir	12.6	0.1	−
6	Surface water—reservoir	16.7	1.0	−
7	Surface water—dune infiltration	3.7	0.4	+
8	Groundwater	8.4	0.7	+
9	Groundwater	9.9	0.9	−
10	Groundwater	8.2	0.3	+
11	Groundwater	4.7	0.6	−
12	Groundwater	3.8	0.4	+

^aCompliance with the Dutch Aeromonas standard (1,000 CFU/100 ml).

^bAt least 90% of this AOC was based on strain NOX, which utilizes carboxylic acids (37).

^cStrain P17 can utilize a broad range of carbon sources, including amino acids (38).

FIG 1.

Occurrence of Aeromonas species in DWDS in 2012 (open bars) and 2015 (hatched bars). The percentage of the total number of isolates represented by Aeromonas species is depicted for each species. Numbers above the bars are the numbers of DWDS in which this species was found, out of the total number of analyzed DWDS. In total, 231 (2012) and 216 (2015) Aeromonas species were isolated from the selected DWDS (Table S1).

Planktonic growth of pure cultures of Aeromonas species on low-concentration carbon sources in drinking water.

Several growth experiments with pure cultures of Aeromonas isolates from the DWDS were performed in drinking water from plant 12 (compliance with Aeromonas standard; AOC, 3.8 µg C/liter) (Table 1). Experiments were performed with and without ammonium addition and with the addition of low concentrations of several biomass compounds as a carbon source (amino acids, long-chain fatty acids, and biopolymers) to verify whether the conclusions of an earlier study (8), performed with one Aeromonas isolate, apply to only one or to all tested Aeromonas species. For these experiments, drinking water from plant 12 was selected as a blank, as no Aeromonas regrowth was observed in the respective DWDS (Table 2) and none of the seven Aeromonas species was able to grow as a pure culture in this drinking water (blank in Fig. S1). When this water was enriched with 10-µg-C/liter concentrations of oleate or arginine as a carbon/energy source and nitrate as a nitrogen source, all strains showed growth (Table 2; Fig. S1).

TABLE 2.

Growth rates and growth maxima of seven Aeromonas species^a

Carbon sourceb	Nitrogen source	A. salmonicida (M800)		A. salmonicida		A. sobria		A. bestiarum		A. hydrophila		A. rivuli		A. media		A. veronii
		V (h−1)	Nmax (CFU/ml)	V (h−1)	Nmax (CFU/ml)	V (h−1)	Nmax (CFU/ml)	V (h−1)	Nmax (CFU/ml)	V (h−1)	Nmax (CFU/ml)	V (h−1)	Nmax (CFU/ml)	V (h−1)	Nmax (CFU/ml)	V (h−1)	Nmax (CFU/ml)
Oleate	NO3	0.14	2.4 × 104	0.19	3.6 × 104	0.08	9.0 × 103	0.11	1.2 × 105	0.09	5.7 × 103	0.10	9.6 × 104	0.12	5.0 × 104	0.07	1.0 × 104
	NO3 + NH4	0.20c	1.2 × 105c	0.20	1.5 × 105	0.12	9.4 × 104	ND	ND	ND	ND	0.14	1.4 × 105	ND	ND	0.10	9.0 × 104
Arginine	NO3	0.06	3.8 × 105	0.11	1.7 × 105	0.04	4.1 × 104	0.07	3.2 × 104	0.08	2.3 × 104	0.05	2.4 × 104	0.05	2.0 × 104	0.05	1.2 × 105
	NO3 + NH4	0.06c	5.9 × 104c	0.09	1.1 × 105	0.07	7.2 × 104	ND	ND	ND	ND	0.10	9.4 × 104	ND	ND	0.07	5.9 × 104
Chitin	NO3	0.03	1.7 × 103	NG	NG	NG	NG	0.02	4.0 × 102	NG	NG	NG	NG	NG	NG	NG	NG
	NO3 + NH4	ND	ND	0.03	1.9 × 104	0.02	1.5 × 102	ND	ND	ND	ND	0.01	1.2 × 102	ND	ND	0.01	5.6 × 102
Biopolymers	NO3	NG	NG	NG	NG	NG	NG	NG	NG	NG	NG	NG	NG	NG	NG	NG	NG
	NO3 + NH4	ND	ND	0.02	4.2 × 103	0.01	8.7 × 101	ND	ND	ND	ND	NG	NG	ND	ND	0.01	2.2 × 102

^aAeromonas species were isolated from DWDS in the Netherlands. Growth experiments were performed in duplicate at 15°C. Data are averages from duplicate experiments; error bars are presented in Fig. S1. ND, not determined; NG, no growth.

^bCarbon/energy sources were used at 10 µg C/liter.

^cPreviously published results from growth experiments performed under the same conditions (8).

Growth rates and maximal growth (N_max) values differed between species. The growth rates of all seven Aeromonas species were higher with oleate than arginine as the carbon/energy source (Table 2). In addition, the growth rates of the different species with oleate ranged between 0.07 h⁻¹ and 0.19 h⁻¹, which is slightly larger than the range of growth rates with arginine as the sole carbon/energy source (0.04 to 0.11 h⁻¹). For both oleate and arginine, the highest growth rates were achieved by A. salmonicida, whereas A. veronii showed the lowest growth rates for both carbon sources. The N_max values also varied between the Aeromonas species, with a higher N_max for arginine than oleate for A. hydrophila, A. sobria, A. veronii, and A. salmonicida but a lower N_max for arginine than oleate for A. media, A. bestiarum, and A. rivuli (Table 2). In addition, the N_max for the different Aeromonas species growing with arginine varied between 2.0 × 10⁴ and 3.8 × 10⁵ CFU/ml, whereas with oleate, the N_max varied between 5.7 × 10³ and 1.2 × 10⁵ CFU/ml. The effect of ammonium as the nitrogen source was tested for A. rivuli, A. veronii, A. sobria, and A. salmonicida with oleate and arginine as carbon/energy sources. The data showed that without ammonium addition, these strains also grew on these added compounds in drinking water. However, both growth rate and N_max in the absence of added ammonium were lower for all four strains when they were grown on oleate but only for A. rivuli and A. sobria when they were grown on arginine (Table 2).

Of the seven species tested, only A. salmonicida M800 and A. bestiarum were able to grow with chitin as the sole carbon/energy source and nitrate as the sole nitrogen source, but both the growth rate (0.02 to 0.03 h⁻¹) and N_max (0.4 to 1.7 × 10³ CFU/ml) were low. None of the Aeromonas species could grow with the biopolymer mixture (amylopectin, pectin, xyloglucan, laminarin, casein, gelatin, and lectin) when only nitrate was present. When ammonium was added to the growth medium of four of these species (A. rivuli, A. veronii, A. sobria, and A. salmonicida), growth on the biopolymer mixture and chitin was observed (Fig. S1; Table 2). However, in both cases (with chitin or the biopolymer mixture as the carbon/energy source), growth rate and N_max values were considerably lower than when arginine or oleate was used as the carbon/energy source.

Aeromonas growth in produced drinking water in competition with the autochthonous microflora.

The above-described experiments were performed in pasteurized drinking water to test the capacity of pure cultures of different Aeromonas species to utilize various carbon/energy and nitrogen sources for growth. Since bacteria belonging to the genus Aeromonas are only a small fraction of the bacteria present in the drinking water system (22, 24, 27), it was hypothesized that Aeromonas cannot compete successfully with other autochthonous bacteria in drinking water. The influence of the autochthonous microflora on Aeromonas growth was investigated by adding a mixed Aeromonas inoculum (A. rivuli, A. salmonicida, A. sobria, and A. veronii; 10 to 50 CFU/100 ml) to pasteurized (autochthonous microflora eradicated) and nonpasteurized (autochthonous microflora untouched) drinking water from three drinking water treatment plants that use surface water and that were studied before (23) (Fig. 2; plants 1, 2, and 3 in Table 1). The mix of four strains was chosen for the inoculum because they were frequently observed in the DWDS and thus simulate an environmentally abundant Aeromonas population. The mixed Aeromonas inoculum showed growth in all three pasteurized drinking waters, but the growth rates differed between the different treatment plants. The level of AOC determined with Pseudomonas fluorescens strain P17 and Spirillum sp. strain NOX (AOC-P17/NOX concentration) in these three drinking waters were of the same order of magnitude and higher than for the other plants (Table 1). The calculated planktonic growth rate (V) in drinking water from plant 2 was lowest (V = 0.02 ± 0.004 h⁻¹), followed by drinking water from plant 1 (V = 0.06 ± 0.004 h⁻¹) and plant 3 (V = 0.09 ± 0.02 h⁻¹). In contrast, the mixed Aeromonas inoculum was not able to grow in nonpasteurized drinking water samples; only decay of the inoculated Aeromonas species was observed in these drinking water samples (Fig. 2; Table 1). These results demonstrate that the inoculated Aeromonas strains were able to utilize the naturally present nutrients in drinking water from these three plants but that they were unable to compete successfully for these nutrients with the autochthonous microflora present in these drinking water types.

FIG 2.

Aeromonas growth (mixture of A. rivuli, A. salmonicida, A. sobria, and A. veronii) in the drinking water distribution system of three drinking water treatment plants (plants 1, 2, and 3) with (competitive growth [CG]) and without (pure culture [PC]) the autochthonous microflora in May and July 2016. Data are averages and standard deviations (SD) (n = 2).

This finding raises the questions of whether Aeromonas growth can occur in the presence of the autochthonous microflora when carbon sources favorable for Aeromonas (e.g., amino acids and fatty acids) are present and whether there is a nutrient concentration limit below which Aeromonas growth in nonpasteurized drinking water does not occur. Therefore, Aeromonas growth was tested in nonpasteurized drinking water of plant 3, for which Aeromonas growth was observed in the DWDS (Table 1; Fig. 2). To this water, different concentrations of the amino acid and fatty acid mixtures were added. Mixtures were used instead of single compounds, as it seems unlikely that these compounds occur in nature as individual compounds. Addition of a mixture of five fatty acids at a cumulative concentration of 10 µg C/liter to the nonpasteurized drinking water of plant 3 resulted in no Aeromonas growth (Fig. 3A; Table S2), indicating that growth of Aeromonas in drinking water on fatty acids is unlikely, and no further studies were performed with fatty acids. Aeromonas growth was observed in nonpasteurized drinking water enriched with the amino acid mixture at a concentration of ≥5 µg C/liter, and at this concentration, the growth rate was 0.04 h⁻¹, with an N_max value of 16 CFU/ml (Fig. 3A; Table S2). Growth rate and N_max (Table S2; Fig. 3A) increased with the amino acid concentration, and when the amino acid mixture was at 10 µg C/liter, the growth rate and N_max were 0.12 h⁻¹ and 24 CFU/ml.

FIG 3.

(A) Growth curves of an Aeromonas mixture (A. rivuli, A. salmonicida, A. sobria, and A. veronii) with various concentrations of an amino acid mixture and 10 µg C/liter fatty acids in nonpasteurized drinking water of plant 3. Data are averages and SD (n = 2). (B) Four concentrations of the amino acid mixture were also tested in nonpasteurized drinking water of four drinking water production plants (plants 1, 3, 11, and 12). Data are averages and SD (n = 2).

The results of this first experiment in nonpasteurized drinking water were extended in drinking water samples from plant 3 and three additional plants, one also using surface water as the source (plant 1) and two plants using groundwater as the source water (plants 11 and 12). It was observed that Aeromonas growth occurred in nonpasteurized drinking water from plants 1 and 11 when a minimum cumulative amino acid concentration of 5 µg C/liter was present (Fig. 3B). Aeromonas growth in nonpasteurized drinking water from plants 3 and 12 occurred when a minimum cumulative amino acid concentration of 10 µg C/liter was present (Fig. 3B). The minimum cumulative amino acid concentration required for Aeromonas growth was not related to the AOC level of the different drinking waters (Table 1), with higher AOC levels for plant 1 (13.7 µg C/liter) and plant 3 (15.5 µg C/liter) but low levels for plant 11 (4.7 µg C/liter) and plant 12 (3.8 µg C/liter). The Aeromonas yield differed between the four drinking water types, with plant 1 having the highest N_max for Aeromonas and plant 12 having the lowest N_max (Table 3). Differences in water quality and microflora, caused by seasonal influences, may be a possible explanation for the small difference in the minimum amino acid concentration required for Aeromonas growth in the two separate experiments performed in nonpasteurized drinking water from plant 3.

TABLE 3.

Growth rates and growth maxima of a mixture of four Aeromonas species with an amino acid mixture in drinking water from 4 treatment plants in competition with the autochthonous flora^a

C concn (μg/liter)	Plant 1		Plant 3		Plant 11		Plant 12
	V (h−1)	Nmax (CFU/ml)	V (h−1)	Nmax (CFU/ml)	V (h−1)	Nmax (CFU/ml)	V (h−1)	Nmax (CFU/ml)
0.0	NG	NG	NG	NG	NG	NG	NG	NG
5.0	0.05	2.7 × 100	NG	NG	0.03	2.1 × 100	0.02	1.1 × 100
10	0.19	6.2 × 101	0.12	1.0 × 101	0.12	3.2 × 101	0.09	7.5 × 100
25	0.21	1.0 × 104	0.18	2.3 × 102	0.28	1.9 × 103	0.29	9.8 × 101

^aAeromonas species were isolated from DWDS in the Netherlands. Drinking water in plants 1 and 3 was produced from surface water; drinking water in plants 11 and 12 was produced from groundwater. Results are derived from one experiment, performed in 2020. NG, no growth.

Growth kinetics of Aeromonas species on a mixture of amino acids in drinking water.

Due to the high affinity of pure cultures of Aeromonas species from several DWDS for amino acids in drinking water without competition of the autochthonous microflora (Fig. S1) and the observation that Aeromonas is able to grow with amino acids (≥5 µg C/liter) in nonpasteurized drinking water (Fig. 3B), tests were performed to further elucidate the growth kinetics (V and N_max) of the Aeromonas mixed inoculum on the amino acid mixture without the autochthonous microflora. Therefore, a mixture of 21 individual amino acids was tested at increasing cumulative concentrations of 0.5 to 25 µg C/liter in the pasteurized drinking water of plant 12 (no-Aeromonas-growth blank) (Fig. 4; Table S2). Again, drinking water from plant 12 without carbon source addition (blank) showed no growth. Aeromonas growth was observed at the lowest cumulative amino acid concentration of 0.5 µg C/liter. Furthermore, a stepwise increase in the amino acid concentration to 25 µg C/liter led to higher N_max values and growth rates (Fig. 4A and B; Table S3). The calculated yield of 6.2 × 10⁶ CFU/µg C in the amino acid mixture of the four inoculated strains was somewhat lower than the yield of 8 × 10⁶ CFU/µg C for A. salmonicida (M800; isolate 8) obtained in a previous study (8). In nonpasteurized drinking water, supplemented with 10 µg C in the amino acid mixture/liter, the growth rate and N_max (Table 3) were clearly lower than in the pasteurized drinking water (0.29 versus 0.20 h⁻¹ and 4.9 × 10⁴ versus 5.8 × 10⁴ CFU/ml) (Table S3). Based on the yield of 6.2 × 10⁶ CFU/µg amino acid C, the percentage of utilization of amino acid mixture in the nonpasteurized tests was calculated. The percentage increased with the concentration of amino acid mixture (Table 4). At 5- and 10-µg/liter concentrations, the percentage of utilization was extremely low, <0.12% at the most. The fraction of utilization increased at 25 µg C/liter to a maximum of 7.95%.

FIG 4.

(A) Growth curves of pure cultures of four Aeromonas isolates with increasing concentrations of an amino acid mixture to a maximum of 25 µg C/liter and a mixture of fatty acids (FA; 10 µg C/liter) in pasteurized drinking water at 25°C. Data are averages and SD (n = 2). The annotation “(2)” indicates that a second separate experiment, in duplicate, was performed. (B) Growth rate (averages of two separate experiments with SD) plotted as a function of the amino acid concentration.

TABLE 4.

Percentage of amino acid carbon used by Aeromonas for growth in drinking water from plants 1, 3, 11, and 12

Amino acid C concn (µg/liter)	% amino acid carbon used by Aeromonas
	Plant 3 (1)a	Plant 3 (2)	Plant 12	Plant 1	Plant 11
5	0.00	0.00	0.00	0.01	0.01
10	0.05	0.02	0.01	0.12	0.06
25	ND	0.14	0.08	7.95	5.17

^aThe growth experiments at plant 3 were performed in separate experiments (1 and 2) in two consecutive years (2019 and 2020). ND, not determined.

A significant (P < 0.05) correlation was observed between the amino acid concentration and the growth rate (V). Based on the correlation of the substrate concentration with the growth rate (Fig. 4B), an estimated maximum growth rate of 0.4 h⁻¹, and the Monod growth kinetic equation, a substrate saturation constant (K_S) of 7 µg amino acid C/liter was calculated for the Aeromonas species in the mixed inoculum. This K_S value was of the same order of magnitude as determined earlier for reference strain A. salmonicida M800 (8). Moreover, this K_S value was of the same order of magnitude as the threshold concentration for competitive growth of Aeromonas in drinking water (Table 3).

Planktonic growth in distributed drinking water in competition with the autochthonous microflora.

In the DWDS of plants 1 and 3, the major Aeromonas increase occurred in the large transport pipes. In these pipes, an unequivocal exponential Aeromonas increase rate was observed during the year. The Aeromonas increase for an 11.6-km transport pipe from plant 3 is depicted in Fig. 5A. The increase exceeded the Dutch Aeromonas standard for drinking water of 10 CFU/ml during the residence time in this pipe. The calculated planktonic growth rate (V) for Aeromonas in this pipe section in competition with the autochthonous microflora of 0.30 h⁻¹ (Fig. 5) was of the same order of magnitude as the planktonic growth rate of pure cultures of Aeromonas on a maximum amino acid concentration of 25 µg C/liter in pasteurized drinking water (Table S3).

FIG 5.

(A) Increase in Aeromonas numbers in drinking water transported in an 11.6-km pipe in August 2016. Each sampling location (six in total) was sampled on five separate days. Data are medians, with minimum and maximum values. (B) Exponential growth rate of Aeromonas in competition with autochthonous bacteria in 28 drinking water samples from the middle and bottom of the transport pipe at 6 and 14 h of residence time in July, August, September, and October 2016.

To investigate the Aeromonas growth potential of the drinking water from this pipe in competition with the autochthonous microflora, samples from the middle or bottom of this pipe after 6 and 14 h residence time were collected in four sampling campaigns (n = 28) (Fig. 5B) and incubated for a maximum of 10 days. Results showed that in 93% of these samples, the planktonic growth rate of Aeromonas was zero or decay occurred; in two samples, low exponential growth rates of 0.04 and 0.06 h⁻¹ were observed (Fig. 5B). Thus, the unequivocal Aeromonas increase in the drinking water during the residence time in the transport pipe (Fig. 5A) was not found in the incubated drinking water samples from the same pipe section (Fig. 5B). This implies that it is highly unlikely that planktonic growth of Aeromonas in the distributed drinking water is the cause of the rapid Aeromonas increase observed in this transport pipe. Some other unknown mechanisms related to the conditions in the pipe section (i.e., flow rate, appendages, or the presence of loose deposits) are likely to be responsible for such rapid increase of aeromonads in the bulk water.

DISCUSSION

Aeromonas species identification.

Genotypic characterization of Aeromonas bacteria isolated from 10 drinking water distribution systems in the Netherlands showed that the temporal and spatial Aeromonas species composition varied. In the past, Aeromonas species in nonchlorinated drinking water in the Netherlands were identified as A. hydrophila, A. caviae, and A. sobria, using biochemical identification (15, 17). The genotype-based identification presented in this paper, however, demonstrated that the Aeromonas population was more diverse and that A. veronii and A. rivuli were the two most frequently isolated species from these nonchlorinated drinking water systems. Of the Aeromonas species that were found earlier with biochemical identification (15, 17), A. sobria was observed in 10 to 15% of the samples and A. hydrophila was isolated sporadically in our survey. Moreover, 105 isolates of four different drinking water treatment plants in the Netherlands were biochemically identified as A. hydrophila (70%) and A. caviae (30%) (15, 17), whereas the genotypic identification of these isolates resulted in A. veronii (45%), A. media (25%), A. sobria (17%), and A. salmonicida (8%). In addition, Aeromonas strain M800 was classified as A. hydrophila based on biochemical identification (8), but it belongs to A. salmonicida based on the gyrB gene sequence. A. rivuli was found only in the current study, not in earlier studies. Thus, the observed differences in the Aeromonas species community between previous studies (15, 17) and our study are caused by the discrepancy between identification methods (biochemical versus genotype identification), as has been reported by others as well (28). Our results also show that there was no relation between the Aeromonas species composition and (i) the year of the sampling campaign (2012 or 2015), (ii) the type of water source used for drinking water production (groundwater versus surface water), (iii) the AOC concentration in the drinking water, and (iv) the Aeromonas numbers in the DWDS. Thus, differences in the Aeromonas population between the DWDS do not seem to have a systematic ecological cause relevant to the drinking water situation. Further studies into Aeromonas species diversity in DWDS are therefore not expected to yield additional ecological information.

Currently, as documented in the introduction, the general view is that Aeromonas in drinking water has no health significance. This view is based on results from studies that showed that Aeromonas isolates from fresh water and drinking water were phenotypically and genotypically different from species isolated from patients (15). However, a recent study concluded that Aeromonas is highly infectious and that human susceptibility to diarrhea may be high, as in the case of undisputed enteropathogens like Salmonella and Campylobacter (29). That finding is mainly based on data from three supposed Aeromonas outbreak studies with food, which is in contrast to a published literature review where Aeromonas was not classified as a true gastrointestinal pathogen, because not a single outbreak of diarrhea could clonally be identified to be related to Aeromonas (18). Indeed, definite proof that Aeromonas was the causative agent for gastroenteritis in the three outbreak studies that were used to determine the health significant of Aeromonas in the other study (29) is missing, because these three outbreak studies have serious flaws: (i) possible fecal enteropathogenic viruses were not monitored in the food and/or stool samples, missing an important possible cause of illness (30–32); (ii) Aeromonas could not be isolated from stool samples of 80% of the patients (30); (iii) stool samples from patients were not analyzed for fecal pathogens (32); and (iv) no data are provided on analyses done on stool or food samples (31). Consequently, these supposed Aeromonas outbreak studies cannot be used to reliably determine the dose-response relationship of Aeromonas and gastroenteritis, which makes the conclusion that Aeromonas is highly infectious disputable.

In the same extended review, Aeromonas was described as a potential pathogen for other infection categories, namely, wound and soft tissue infections, blood-borne dyscrasia, and a miscellaneous “catch-all” category (18). While A. salmonicida is a well-known fish pathogen, A. hydrophila, A. caviae, and A. veronii are receiving increasing attention due to their association with the other infection categories mentioned above (18, 33). However, the epidemiological attention worldwide for Aeromonas in these diseases is low, and infections are not reported in regular health surveys of most countries, most likely because some early epidemiological studies in the United States, France, and the United Kingdom showed a very low annual incidence of Aeromonas infections (0.7 to 3.2 × 10⁻⁶) (18). In a recent Dutch study, Aeromonas spp. were also only sporadically (0.1%) isolated from wound infections in 83 hospitals (34), demonstrating that the health significance of Aeromonas in drinking water seems to be negligible.

Nutritional characteristics of the observed Aeromonas species in DWDS.

Based on the Aeromonas species we isolated from the 10 nonchlorinated DWDS, we conclude that these DWDS have a rich variety of different Aeromonas species. Due to the low nutrient levels in drinking water, growth experiments in pasteurized drinking water, in which the autochthonous bacteria have been eradicated, have been used in the past to assess the nutritional versatility of several bacterial species in drinking water, such as Klebsiella pneumoniae, Pseudomonas aeruginosa, Flavobacterium, and Aeromonas (8, 35). Our study used the same experimental method to determine the utilization of selected carbon sources for isolated Aeromonas species from the DWDS and showed that these species have nutritional preferences similar to those described in the previous study (8), although there were differences in growth rates and growth maxima for the tested substrates.

The preference for ammonium over nitrate as the nitrogen source and the utilization of amino acids and long-chain fatty acids at the level of micrograms of C per liter are common characteristics of all Aeromonas species isolated from the different DWDS. Ammonium addition in the growth experiments with oleate had more effect on the growth rate and yield of the Aeromonas spp. than ammonium addition in the growth experiments with arginine. This apparent difference is probably caused by the absence of nitrogen in oleate and the presence of nitrogen in arginine, which could be used as a nitrogen source by the different Aeromonas species during growth on arginine. Still, Aeromonas is able to grow in drinking water with low concentrations of oleate (10 µg C/liter), while no ammonium was detected. The theoretic stochiometric required ammonium concentrations in the drinking water for the usage of 10 µg C/liter arginine and oleate is approximately 1.8 µg N/liter (36), which is below the analytical detection limit of ammonium in water (0.1 mg/liter). Based on the preference of most Aeromonas species for ammonium as the nitrogen source, it might be possible that ammonium, present in drinking water at concentrations below the detection limit, were responsible for the growth response on oleate without ammonium addition. This also demonstrates that under low-nutrient conditions in DWDS, the nitrogen source is of less importance as a limiting growth factor for planktonic growth of Aeromonas than the carbon/energy sources (e.g., amino acids or fatty acids). For Aeromonas growth on high-molecular-weight compounds (e.g., biopolymers and chitin), however, ammonium appears to be essential as a nitrogen source. The relatively low growth rates and yields that Aeromonas species attain on high-molecular-weight compounds, however, indicate that competitive planktonic growth of Aeromonas on these biopolymers in drinking water is not very likely.

The growth kinetics of A. salmonicida strain M800 for growth on a mixture of 21 amino acids in drinking water has been studied before (8). Comparison of those data with the data of our study showed that when the amino acid concentration mixture was ≤5 µg C/liter, the growth rate of the Aeromonas species was similar to the growth rate for A. salmonicida strain M800 that was published before (8). At concentrations above 5 µg C/liter, the growth rate of the Aeromonas species mixture was higher than previously reported for A. salmonicida. The current study identified minor dissimilarities in nutritional characteristics (growth rates and maxima; nitrogen source and biopolymer utilization) between the tested Aeromonas species in nonchlorinated drinking water. These dissimilarities between Aeromonas species could have been the cause of the relatively high diverse and nonsystematic variation in the Aeromonas species isolated from the sampled DWDS.

Competitive planktonic growth of Aeromonas species in drinking water.

Aeromonas bacteria constitute only a very small fraction (<0.01%) of the total prokaryote community in nonchlorinated drinking water in the Netherlands (22). The low Aeromonas numbers in drinking water raise the question of whether Aeromonas can compete with the autochthonous microflora for easily biodegradable organic carbon. To assess the competitive growth of Aeromonas, growth experiments with pasteurized (autochthonous microflora eradicated) and nonpasteurized (autochthonous microflora untouched) drinking water from four different treatment plants, with or without addition of a preferred carbon source (amino acids), were done. The results demonstrated that Aeromonas was not able to compete with the autochthonous bacteria for the utilization of nutrients under the oligotrophic conditions in drinking water with AOC-P17/NOX concentrations of 3.7 to 16.7 µg C/liter. More than 90% of this AOC concentration was based on strain NOX, which utilizes carboxylic acids (37), and AOC-P17 concentrations were <1 µg C/liter (Table 1) (38).

A threshold concentration of 5 to 10 µg C/liter amino acid mixture in drinking water was required before minimal planktonic growth of Aeromonas occurred, regardless of the AOC concentration in the drinking water used in the experiment (39). The fraction of the amino acid mixture utilized by Aeromonas in competition increased with the added amino acid C concentration and only at 25 µg C/liter was the fraction >1% in two drinking waters (5.17 and 7.95%) (Table 4). Amino acid concentrations of 5 to 10 µg C/liter are not very likely to be present in nonchlorinated drinking water in the Netherlands (40), as the AOC concentration declines during distribution and is approximately 15 to 30 µg C/liter. Only a small part (<1 µg C/liter) of the total AOC concentration is expected to consist of amino acids (40), as was also demonstrated for the drinking water in the current study (Table 1), which is too low to support growth of Aeromonas in drinking water. Additional observations of a rapid exponential rate for Aeromonas in a DWDS transport pipe section in summer, but not in drinking water sampled from these transport pipes (Fig. 5A), also showed that the presence of such amino acid concentrations in drinking water during distribution is highly unlikely. Therefore, the results from our study suggest that competitive planktonic growth of Aeromonas in nonchlorinated drinking water during distribution is not a very likely cause for the increasing Aeromonas numbers in DWDS in the Netherlands. The competitive experiments were all performed under static conditions in flasks. Under the hydrodynamic conditions in DWDS, uneven distribution of nutrient concentrations may occur. Further verification of the conclusion on competitive planktonic growth is required with additional growth experiments under dynamic flow conditions. Furthermore, planktonic growth on a microscale in the loose-deposit niche of the DWDS, for instance due to cross feeding (39), is also possible and requires further attention.

The alternative mechanism for elevated Aeromonas numbers in drinking water is that growth of Aeromonas species occurs in the two other niches in DWDS, notably benthic growth in sediments (loose deposits) and/or growth in biofilms on the pipe wall, followed by transmission to the bulk water phase. The highest microbial activity in distribution networks occurs in these niches (41). It is important to note that in DWDS that distribute nonchlorinated drinking water, loose deposits were identified as the dominant niche for Aeromonas (25, 26). To identify the influence of sediment concentration and composition on the growth of Aeromonas and competition with the autochthonous microflora, an additional study is required.

Role of Aeromonas as a regrowth indicator for the DWDS.

Fecal contamination and undesired microbial growth in drinking water during distribution may result in health risks and aesthetically undesirable water quality for the consumers. Water companies in most countries control these undesirable microbial conditions by maintaining a disinfectant residual during distribution. By ensuring that a disinfectant residual, e.g., chlorine, is present in the drinking water, most pathogens in a potential fecal contamination of the drinking water can be inactivated. In addition, microbial regrowth in the DWDS can be controlled by maintaining a disinfectant residual in drinking water. However, in several European countries (e.g., Denmark, the Netherlands and parts of Germany, Belgium, Switzerland, and Austria), these issues are controlled with the production of high-quality drinking water that contains low nutrient concentrations (at the level of micrograms of carbon per liter), no disinfectant residual, and a high integrity of the distribution network. In the Netherlands, Aeromonas is, besides the HPC parameter, used as an additional legislative microbial indicator parameter for regrowth in the distribution system at a regulated standard of 1,000 CFU per 100 ml drinking water. Elucidating the cause of noncompliance with this Aeromonas standard in DWDS will enhance insights into the causes of regrowth. No association was observed between the AOC concentration of drinking water from the studied DWDS and compliance or noncompliance with the Aeromonas standard (Table 1) (8).

Regrowth in the DWDS is the result of the biological activity of (micro)organisms in water, loose deposits, and biofilm and consists of a complex food web of prokaryote and eukaryote communities (35). We have shown that Aeromonas has a preference for certain biomass components (amino acids and fatty acids). Others have observed that Aeromonas has its preferential niche in sediment (loose deposits) in the DWDS (25, 26). Consequently, Aeromonas growth can be considered an indicator of biomass accumulation in the sediment of a DWDS and thus of the microbial activity of the food web in the DWDS sediment. Biomass accumulation in the sediment originates from bacteria in the DWDS that utilize the available nutrients from the drinking water and/or pipe materials (42), which in turn are available as prey organisms for higher trophic levels.

The recycling of nutrients in such complex food webs has been defined as the microbial loop in seawater ecosystems (43). Such internal turnover of organic carbon, nitrogen, and phosphorous accumulated in the biomass has, for example, been documented for the predator-prey interaction of the ciliate Colpoda steini and Pseudomonas fluorescens in soil with variable pore sizes. Predation of the ciliate on the bacterial cells reduced P. fluorescens density but increased metabolic activity of the surviving fraction of P. fluorescens biomass due to the release of nutrients by the predator (37). For instance, proteins, amino acids, and ammonia play a major role in the carbon-nitrogen turnover (44). In the DWDS, it is expected that such food webs harbor several microorganisms that contribute to the release of internal nutrients such as bacteriophages, predatory bacteria (e.g., Bdellovibrio spp.), and predatory invertebrates. Protozoa and invertebrates (e.g., crustaceans) in the distribution system produce organic carbon, ammonium, and amino acids by sloppy feeding, excretion, and fecal pellet leaching (45). In addition, invertebrates, including Asellus aquaticus, which is visible to the naked eye, can be present in DWDS (23, 46–48) and can excrete urea and ammonia in the environment (49, 50). The biomass of these invertebrates contains large amounts of amino acids and long-chain fatty acids whose composition differs based on the composition of their diet (51). In the carbon-nitrogen turnover, proteins of consumed biomass are hydrolyzed to amino acids from which new proteins are formed. The excess amino acids are reduced and excreted as urea and ammonia and further used by nitrifying bacteria. The high abundance of ammonia-oxidizing bacteria and archaea in DWDS where nonchlorinated drinking water is distributed (52–55) demonstrates how common this nitrogen turnover process probably is in drinking water environments.

Based on these observations, it can be concluded that preferred carbon sources (long-chain fatty acids and amino acids) and ammonia-nitrogen for Aeromonas are omnipresent in these environments, including the sediment niche where Aeromonas is dominantly observed in DWDS. However, as described earlier, the threshold concentration for amino acids is most likely not reached in the drinking water (40). Furthermore, invertebrates are abundant as well in these sediments (46, 47), and it has been observed that Aeromonas can be present in the gut of chironomids and Asellidae (56–58), invertebrates that are observed in DWDS (46, 47). Currently, experiments are under way to investigate whether Aeromonas abundance in DWDS is caused by benthic growth of Aeromonas on biomass components that are available in the sediment of DWDS or through Aeromonas excretion by (certain) invertebrate species that live in the sediment fraction of DWDS. Still, the findings of our study combined with the observation that Aeromonas has its preferred niche in sediments of DWDS emphasize the significance of Aeromonas as an indicator bacterium for the potential presence of elevated biomass concentrations in sediment, but it could be less successful as a direct indicator of elevated biomass concentrations in the biofilm on the pipe wall and the loose deposits.

MATERIALS AND METHODS

Isolation and identification of Aeromonas strains.

The Aeromonas species diversity was investigated by sampling the DWDS of 10 different treatment plants in the Netherlands that distribute nonchlorinated drinking water produced from surface water or groundwater (plants 1 to 10 in Table 1). Nine of these 10 DWDS were sampled during the summer months of 2012 and five DWDS during the summer months of 2015. At four of these plants, the Aeromonas numbers were compliant with the Aeromonas standard (<1,000 CFU/100 ml), and at six of these plants, Aeromonas numbers were not compliant with the standard. During both surveys, drinking water was sampled at the kitchen tap, which was flushed until the temperature was constant for 30 s to be sure the drinking water sample contained water from the DWDS and not the premises’ plumbing system. One hundred milliliters of drinking water was filtered, and the membrane was incubated on ampicillin-dextrin agar plates for 24 h at 30°C ± 0.5°C according to NEN 6263 (59) to selectively cultivate Aeromonas. Subsequently, a maximum of three Aeromonas colonies were selected per agar plate, resulting in a total of 216 isolated Aeromonas strains in 2015 and 231 isolates in 2012. These Aeromonas isolates were then identified to the species level by amplifying the gyrB gene using the UP-1 and UP-2r primers (60). The PCR product (1,100 bp) for each Aeromonas isolate was sequenced using a previously published protocol (61). The resulting gyrB gene sequence of each Aeromonas isolate was compared to a reference database constructed in Bionumerics (Applied Maths, Sint-Martens-Latem, Belgium), which contained the gyrB gene sequences of all described Aeromonas species, to identify each isolate to the species level.

Selection of isolated Aeromonas species for growth experiments.

Seven of eight Aeromonas species identified that were isolated from the DWDS were selected for follow-up studies to determine their growth characteristics as pure cultures. The Aeromonas strains that were tested were randomly selected from DWDS with or without Aeromonas regrowth problems. One isolate each of A. media (plant 2), A. bestiarum (plant 1), A. hydrophila (plant 1), and A. salmonicida (plant 1) and two isolates each of A. veronii (plant 2 and 4), A. rivuli (plant 3 and 4), and A. sobria (plant 3 and 5) were used for growth experiments with individual (in)organic compounds. When two isolates of one Aeromonas species were used, these isolates were mixed 1:1 before inoculation in the growth experiments. To enable comparison with the results of van der Kooij and Hijnen (8), strain M800, genomically renamed A. salmonicida, was included in the study.

Growth of pure cultures on different carbon sources.

The ability of Aeromonas strains to utilize different carbon sources was determined by adding the following low-molecular weight carbon sources individually to 600 ml of nonchlorinated drinking water from plant 12 (Table 1) in duplicate: 10 µg arginine C/liter, 10 µg oleate C/liter, and two mixtures of amino acids (21 amino acids in equimolar amounts: l-alanine, l-arginine, l-asparagine, l-aspartate/l-aspartic acid, l-cysteine, l-citrulline, l-glutamate/l-glutamic acid, l-glutamine, glycine, l-histidine, l-isoleucine, l-leucine, l-lysine, l-methionine, dl-phenylalanine, l-proline, dl-serine, l-threonine, dl-tryptophan, l-tyrosine, l-valine) and of fatty acids (equimolar mixture of sodium myristate, palmitic acid, sodium stearate, sodium oleate, and arachidonic acid). The amino acid mixture was tested at final concentrations of 0.5, 1.0, 2.5, 5.0, 7.5, and 10.0 µg C/liter and the fatty acid mixture at a final concentration of 10 µg C/liter. The high-molecular-weight carbon sources were 100 µg alpha chitin C/liter or a mixture of environmental polysaccharides and proteins or biopolymers (1 µg amylopectin C/liter, 10 µg xyloglucan C/liter, 1 µg pectin C/liter, 1 µg laminarin C/liter, 1 µg casein C/liter, 1 µg gelatin C/liter, and 10 µg lectin C/liter). Three high-molecular-weight compounds (xyloglucan, chitin, and lectin) did not completely dissolve in water and were therefore tested at higher concentrations to reach a higher, but unknown, dissolved concentration.

For each condition, 600 ml drinking water was sampled in 1-liter Pyrex glass Erlenmeyer flasks (heated at 550°C for 4 h to remove all AOC), enriched with the above-mentioned carbon sources, and pasteurized at 60°C for 30 min to kill the autochthonous microflora. To sterile solutions (i.e., autoclaved for 15 min at 121°C), 1.6 mg KH₂PO₄/liter and 0.1 mg NO₃-N/liter with or without 0.1 mg NH₄-N/liter were added. All flasks were inoculated with 25 to 50 CFU of each tested Aeromonas species, confirmed by measurements directly after Aeromonas addition to the flasks, that had been pregrown in mineral medium with added glucose (0.25 mg/liter) and KNO₃ (6 mg/liter) until nutrients were depleted. Growth of each species on the different carbon sources was monitored in the Erlenmeyer flasks incubated at 15 ± 1.0°C in the dark for a maximum of 37 days. The Aeromonas colony count in these Erlenmeyer flasks was determined periodically on a nonspecific growth medium, Lab-Lemco agar (LLA 8.0), with the streak plate technique, and agar plates were incubated for 24 h at 37°C ± 0.5°C. The growth experiment was finalized when two consecutive colony counts were lower than the maximum monitored colony count in the Erlenmeyer flask.

Pure culture and competitive growth in nonchlorinated drinking water.

Aeromonas growth as pure culture (pasteurized samples) and in competition with the autochthonous microflora (no pasteurization) in nonchlorinated drinking water from two drinking water treatment plants (DWTPs) with Aeromonas regrowth during distribution (plants 1 and 3) (Table 1) and one DWTP (plant 2) with Aeromonas standard compliance was determined. For these experiments, 600 ml nonchlorinated drinking water was sampled in 1-liter AOC-free Pyrex Erlenmeyer flasks. Water was sampled in May and July 2016 at the drinking water treatment plant after the final filtration stage (biological activated carbon filters). Water was sampled in 1-liter Pyrex glass Erlenmeyer flasks (heated at 550°C as described above to remove AOC). Experiments started within 24 h of cold storage. Half of the flasks were pasteurized (30 min at 60°C), and all flasks were supplied with 1.6 mg sterile KH₂PO₄/liter and 0.4 mg NO₃ N/liter. Prior to incubation at 25°C ± 1.0°C, all flasks were inoculated with a mixture of four Aeromonas species, A. rivuli, A. salmonicida, A. sobria, and A. veronii, at 15 to 150 CFU to a low initial concentration of 1 to 10 CFU/ml. Directly after Aeromonas addition, the concentration in the water was determined. Aeromonas growth was monitored regularly during 23 days of incubation by spread plate culturing on ampicillin-dextrin agar plates and incubation for 24 h at 30°C ± 0.5°C according to NEN 6263 (59).

The growth response of Aeromonas in competition with the autochthonous microflora was also determined with the addition of a mixture of 21 amino acids as the carbon source. This mixture was similar to that used in the pure culture growth experiment described above. Amino acids were used at the same concentrations, and an extra condition of amino acids at 25 µg C/liter was included, in nonchlorinated drinking waters from plant 3 in March 2019 and again in May 2020. To examine competitive growth also in other nonchlorinated drinking waters with different chemical and microbial composition, the latter competitive growth experiment was also performed with the nonchlorinated drinking water from plants 1, 11, and 12. The competitive growth of the Aeromonas inoculum in the presence of 10 µg C/liter of the fatty acid mixture was tested just once in the nonchlorinated drinking water from plant 3 in June 2016.

Possible competitive growth of Aeromonas in drinking water from the transport pipes of the distribution network of plant 3, as a function of residence time, was investigated by determining the number of Aeromonas bacteria at residence times of 2, 6, 8, 12, 14, and 18 h. The residence time was calculated from a daily calibrated, network-specific hydraulic model and averaged over 24 h. At each sampling point, drinking water was sampled five times in August 2016. In addition, at 6 and 14 h residence time, drinking water was sampled (n = 4) from the middle and the bottom of the same transport pipes. These samples were collected by installing a stainless steel sampling pipe in the transport pipe which could be set to sample at different positions in the transport pipe. The same mixture of Aeromonas species was added to the sampled drinking water, and the competitive growth experiments were performed as described for the pure culture experiments above except for the pasteurization step, which was not executed. Growth experiments were performed in duplicate at 25°C ± 1.0°C and growth was monitored for a maximum of 10 days or less when on two consecutive measurements the colony counts were lower than the maximum monitored colony count. Growth was monitored by culturing decimal dilutions of the 0.1 ml samples on ampicillin-dextrin agar plates, which were incubated for 24 h at 30°C ± 0.5°C according to NEN 6263 (59).

Calculation of growth rate.

The data from each growth experiment were used to calculate the exponential growth rate (V; per hour). This was calculated using the following formula: V = {[logN₂ − logN₁]/[log₂(t₂ − t₁)]}, where N₁ and N₂ are the colony counts at time points t₁ and t₂.

The Monod equation was used to calculate the K_S value: V = V_max[S/(K_S + S)], where V and V_max are the specific and maximum specific growth rates of Aeromonas, S is the substrate concentration, and K_S is the value of S when V = 0.5 × V_max.

Data availability.

The gyrB sequences of one isolate for each Aeromonas species isolated were submitted to GenBank (accession numbers MW280825 to MW280832).