Bacteriophage ecology in a small community sewer system related to their indicative role in sewage pollution of drinking water.

Bacteriophage ecology in a small community sewer
system related to their indicative role in sewage
pollution of drinking water
Efrat Gino, Jeana Starosvetsky and Robert Armon*
Faculty of Civil and Environmental Engineering, Division
of Environmental, Water and Agricultural Engineering,
Technion, Haifa 32000, Israel.
Summary
In view of various studies looking for the merit of
coliphages as indicators of water pollution with
viruses originating from faecal material, a small agri-
cultural community (population of approximately
1500 inhabitants of all ages, 2–3 km from Haifa) was
selected in order to understand these bacteriophage
ecology (F-RNA and somatic coliphages) in its sewer
and oxidation pond system. Along the sewer lines,
it was possible to isolate constantly both bacter-
iophage types (F-RNA and somatic coliphages) at
102-104 plaque-forming units (pfu) ml-1. The average
numbers of somatic and F-RNA phages isolated from
oxidation pond were 103-104 pfu ml-1; however,
somatic coliphages were undetectable for several
months (April–August). Significant high correlation
(0.944 < R2 < 0.99) was found between increased
anionic detergent concentrations and F-RNA coliph-
age numbers. Infants less than 1 year old excreted
both phage types and few only F-RNA coliphages
(at high numbers > 105 pfu g-1) for up to 1 year. The
excretion of F-RNA coliphages was highly linked to
Escherichia coli F+ harborage in the intestinal track as
found in their faecal content. Finally, three bacterial
hosts E. coli F+, F– and CN13 tested for survivability in
sewage filtrate revealed that E. coli F+ had the highest
survivability under these conditions. Presence of
somatic and F male-specific phages in sewer lines of
a small community are influenced by several factors
such as: anionic detergents, nutrients, temperature,
source (mainly infants), shedding and survival capa-
bility of the host strain. Better understanding of
coliphages ecology in sewer systems can enhance
our evaluation of these proposed indicator/index
microorganisms used in tracking environmental pol-
lution of water, soil and crop contamination with
faecal material containing enteric viruses.
Introduction
Water pollution model organisms (indicators) have been
studied since the early days of the 20th century with group
variation according to the actual requirements (Bonde,
1963; Kott et al., 1973). Due to expensive and laborious
methodology involved in direct detection of enteric
viruses, the indicator systems have progressed impres-
sively (Vaughn and Metcalf, 1975; Marzouk et al., 1980;
Havelaar et al., 1993). Many viruses of human and animal
origin excreted in faeces reach the sewage system and as
a result of their known resistance to natural inactivation or
water treatment they survive better than current bacterial
indicators (Marzouk et al., 1980; Payment et al., 1985;
Rose and Gerba, 1986; Jofre et al., 1995; Bosch et al.,
1997). Since the first proposed indicator, environmental
microbiologists have not agreed on the merit of any indi-
cator microorganisms. The grounds of this disagreement
are founded on a promising onset and a disappointing
conclusion such as in the case with Pseudomonas
aeruginosa, Bifidobacterium, Aeromonas hydrophila and
Clostridium perfringens (Chang et al., 1985; Scott et al.,
2002). Briefly, some indicators are sensitive to disinfec-
tants and environmental stresses (Aeromonas, Escheri-
chia coli) while others are too sturdy (C. perfringens
spores), present at low numbers in sewage, some are
excreted by both humans and animals, some are also
pathogens (E. coli, P. aeruginosa), some are obligatory
anaerobes (Bifidobacterium), some do multiply in sewage
(most of heterotrophic bacteria). Since the late 1980s,
bacteriophages have been regarded as reliable indicators
of viral pollution of drinking water by faeces or sewage
input (Hoffmann-Berling and Mazé, 1964; Armon, 1993;
Armon and Kott, 1993; Havelaar et al., 1993). The ground
for this idea is that bacteriophages are the only group that
closely resemble human viruses therefore they are good
candidates based on morphology, genomic, presence in
human or animal faeces, highly resistant to environmental
stresses and at enough numbers to be directly
Received 12 January, 2007; accepted 31 March, 2007. *For
correspondence. E-mail cvrrobi@tx.technion.ac.il; Tel. (+972)
4 829 2377 Fax (+972) 4 829 3309. The authors would like to
dedicate this article to Mr Yosi Hermoni who passed away prema-
turely after the completion of this study.
Environmental Microbiology (2007) 9(10), 2407–2416 doi:10.1111/j.1462-2920.2007.01355.x
© 2007 The Authors
Journal compilation © 2007 Society for Applied Microbiology and Blackwell Publishing Ltd

enumerated (Armon et al., 1997; Schaper et al., 2002).
Potential new indicators from this group are somatic
coliphages (Kott et al., 1973; IAWPRC Study Group,
1991; Armon, 1992), F male-specific bacteriophages
(named also F-RNA coliphages) (Havelaar and Hoge-
boom, 1983; Durán et al., 2003) and phages infecting
Bacteroides fragilis (Jofre et al., 1995).
Among coliphages, F-RNA coliphages are as resistant
as Norwalk-like viruses (now Noroviruses group) in
extreme environments as well as to aggressive chemical
treatment such as disinfection (Keswick et al., 1985;
Durán et al., 2003; Nappier et al., 2006). Several studies
have examined the ecology of E. coli phages in sewage
and effluents; however, the results differed greatly geo-
graphically (Dhillon et al., 1976; Simkova and Cervenka,
1981; Mossel, 1982; Havelaar and Hogeboom, 1983;
Grabow et al., 1993). There are several reasons for the
above discrepancy such as: faecal content of various
populations and age that initially contribute to indicator
levels, BOD, temperature, chemical and physical treat-
ments of the sewage, animal faeces input into the
system, etc. (Lucena et al., 2004). While most of the
studies were performed on large-scale STP-s, where all
the above factors overlap, there are actually no studies
on small-scale sewage system wherein a particular
aspect can be tested and figured out. Studying small-
scale sewers with high resolution such in the present
study (manholes and lines) allowed us to look upon the
main factors impacting coliphages and their potential as
indicators of enteric viruses. The aim of the present study
was to survey a small community sewer system at
various sites to understand the ecology of coliphages
(two specific groups: E. coli somatic and F-RNA phages)
and their relationship to faecal source and the impact of
certain environmental conditions on their survival. Such a
small system allows ‘high resolution’ up to particular
manhole that collects sewage from a certain population
(i.e. infants). Additional examination of phage numbers in
faeces from infants over time and phage or host behav-
iour in sewers that contained high concentrations of nutri-
ents and detergents (i.e. kitchen sources) were also
performed. Moreover, laboratory experiments were
carried out in attempt to explain the characteristics of
these two coliphage groups under particular environ-
ments as potential indicators.
Results
Distribution of somatic and F male-specific coliphages
along sewer line manholes originating from families
habitat area (H)
Primarily somatic and F male-specific phages were moni-
tored along sewer line H at manholes during the spring
(February–March) and summer (June–August). Sewer
line H collects sewage generated by families (children and
adults) from the habitation area. Figure 1 shows the con-
centration of both phage types that were isolated in low
numbers at H1 to H2a manholes (20–> 50 pfu ml-1) with
some propensity for F male-specific phages (H2a
manhole receives sewage from a childless family). Fol-
lowing additional input, manholes H5 to H7 exhibited an
increase of one order of magnitude in both phage type
numbers. Increase in phage numbers was also observed
at each sampling site as a function of temperature
(June–August) from < 102 to < 103 pfu ml-1 along this
sewer line (data not shown).
Somatic and F male-specific phages in teenagers’
habitat with no kitchen (NK) manholes
A more distinctively characterized area sampled was the
no kitchen area inhabited by youngsters (age 13–18) with
manholes from various apartments (NK stations) (Fig. 2).
The bulk sewage from this area is mainly made of toilet
and shower sources except manholes NK1a (which col-
lects also kitchen sewage) and NK1 (which collects all NK
stations). As can be seen from Fig. 2 the somatic and F
male-specific phages numbers were very low or absent
(NK2-NK7) while NK1 and NK1a were 102–> 104 pfu ml-1.
According to these results, it was speculated that kitchen
contribution in the form of nutrients beside other factors
that will be discussed further has an impact on the two
bacteriophage type numbers and possible multiplication
in the sewer lines. Partial confirmation to these observa-
tions came from the dining room of Kibbutz Yagur that
revealed 102-104 pfu ml-1 of both phages with season
H1
1
10
100
1000
pf
u
m
l-1
Sampling Stations
Somatic coliphages
F male-specific coliphages
H2 H2a H4 H5 H6 H7
Fig. 1. Distribution of E. coli somatic and F male-specific phages in
H line sewer manholes during spring sampling tour
(February–March).
2408 E. Gino, J. Starosvetsky and R. Armon
© 2007 The Authors
Journal compilation © 2007 Society for Applied Microbiology and Blackwell Publishing Ltd, Environmental Microbiology, 9, 2407–2416

variations (data not shown). However, a more appropriate
presumption was brought up and found to be highly cor-
related to F male-specific phages namely detergent con-
centrations (mainly anionic) found in kitchen sewage.
Distribution of somatic and F male-specific coliphages
along other sewer lines manhole
Additional manholes mentioned in Fig. 10 were also
tested for the two bacteriophage groups. Samples from
elderly home (over 80 years old mainly convalescent) and
day care manholes resulted in very low numbers
< 10 pfu ml-1 of both phage types throughout the whole
sampling year. These low numbers can be attributed to
use of pampers in both cases (data not shown). Further-
more, in the communal laundry (LD1) we never isolated
phages due to outlet high water temperature (60–80°C).
The green house (GH), hall and factory contribution was
also minimal with levels of both phage types between < 1
and 102 pfu ml-1. The dairy farm (AF) contribution was
mainly on somatic phages (102-103 pfu ml-1) with very low
F male-specific phages numbers (1–10 pfu ml-1) (data not
shown).
Presence of detergents in sewage and F male-specific
phages
Previous sampling from the non-kitchen sites (NK sites)
that had shown complete absence up to very low numbers
of F male-specific phages raised the second assumption
of chemical compounds that may enhance their presence
at sites such as kitchen and residence homes. Detergents
used in domestic cleaning procedure were tested concur-
rently with F male-specific phage presence at these
various sites. Figure 3 represents the F male-specific
phage numbers at several H sampling manholes and
detergent (as methylene blue active substances, MBAS)
concentrations. Increase in detergent concentration is
well correlated with this phage type numbers. A more
accurate confirmation to this finding is shown in Fig. 4
where H points were sampled concurrently and detergent
concentration was measured in parallel to F male-specific
phage concentrations (three times per year). At all tested
manholes the correlation was highly significant revealing
the impact of detergent on F male-specific phage
numbers. This phenomenon was also looked upon
somatic phages but no correlation was observed. The
disparity among these two phage types will be discussed
later.
NK1
10
0
10
1
10
2
10
3
10
4
10
5
pf
u
m
l-
1
NK (No kitchen) stations sampling manholes
Somatic coliphages
F male-specific phages
NK1a NK2 NK3 NK4 NK5 NK6 NK7
Fig. 2. Distribution of somatic and F male-specific phages at NK
(no residential kitchen) sampling manholes receiving sewage and
grey water from 13- to 18-year-old inhabitants of this area. (NK1a
manhole is located after the communal kitchen and NK1 is the
manhole collecting sewage from the NK stations and also NK1a).
H1
100
101
102
103
104 F male-specific phages
Detergents (as MBAS)
Sampling Manholes
pf
u
m
l-1
2
4
6
C
oncentration (m
g l -1)
H2 H6 H5
Fig. 3. F male-specific phages versus detergent (as MBAS)
concentrations at several H sampling manholes.
0.0
0
500
1000
1500
2000
2500
3000
H5
R2=0.95626
H4-6
R2=0.96043
H5
R2=0.96732
H1
R2=0.98662
pf
u
m
l-1
Detergents Concentration (mg l-1)
H1
H2
H5
H4-6
0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5 5.0 5.5 6.0
Fig. 4. Correlation diagrams between F male-specific phage
numbers and anionic surfactant concentration in sewage at
different sampling sites (sewage originating from family
residences).
Bacteriophage ecology in community sewer system 2409
© 2007 The Authors
Journal compilation © 2007 Society for Applied Microbiology and Blackwell Publishing Ltd, Environmental Microbiology, 9, 2407–2416

Somatic and F male-specific coliphages in pond sewage
system
The final sewage reaches the P1 (collection pond) and is
further transferred gravitational to oxidation pond P2.
Figure 5 represents the level of somatic and F male-
specific phages enumerated throughout the year in the
oxidation pond. Somatic phages number ranged
between 2 and 3 ¥ 103 pfu ml-1 while F male-specific
phages ranged from 300 to > 104 pfu ml-1. It should be
mentioned that during the hot months (July–October)
the diurnal temperature reached 38°C increasing the
pond evaporation hence increased bacteriophages
concentration. The higher F male-specific numbers
and resistance to environmental conditions were further
substantiated by their more frequent presence in the
polishing pond (P3) (Fig. 6). This third pond used for
effluent polishing has a much lower turbidity with algal
high photosynthetic activity and therefore higher soluble
oxygen concentration. In P3 pond, somatic phages
dropped to 0–200 pfu ml-1 and F male-specific phages to
82–300 pfu ml-1.
Fig. 5. Distribution of E. coli somatic and F
male-specific phages in oxidation pond (P2)
during the whole year (from February to
December).
14
F
eb
ru
ar
y
21
F
eb
ru
ar
y
15
M
ar
ch
13
A
pr
il
22
A
pr
il
10
M
ay
15
J
un
e
28
J
un
e
12
J
ul
y
6
A
ug
us
t
17
A
ug
us
t
4
Se
pt
em
be
r
7
O
ct
ob
er
15
O
ct
ob
er
14
N
ov
em
be
r
1
D
ec
em
be
r
8
D
ec
em
be
r
100
101
102
103
104
pf
u
m
l-1
Sampling dates
Somatic coliphages
F male-specific coliphages
Fig. 6. Distribution of E. coli somatic and F
male-specific phages in polishing pond (P3)
during the whole year (from February to
December).
-- --
100
101
102
103
pf
u
m
l-1
Sampling Dates
Somatic coliphages
F male-specific coliphages
14
F
eb
ru
ar
y
21
F
eb
ru
ar
y
15
M
ar
ch
13
A
pr
il
22
A
pr
il
10
M
ay
15
J
un
e
28
J
un
e
12
J
ul
y
6
A
ug
us
t
17
A
ug
us
t
4
Se
pt
em
be
r
7
O
ct
ob
er
15
O
ct
ob
er
14
N
ov
em
be
r
1
D
ec
em
be
r
8
D
ec
em
be
r
2410 E. Gino, J. Starosvetsky and R. Armon
© 2007 The Authors
Journal compilation © 2007 Society for Applied Microbiology and Blackwell Publishing Ltd, Environmental Microbiology, 9, 2407–2416

Fate of bottle-confined somatic and F male-specific
phages in manhole
During summer tour (July–September) sewage samples
taken in the morning were left into polycarbonate bottles
in the manholes and enumerated at start and after 48 h
(Table 1). The average temperature inside the manhole
was 25 � 2°C that is bellow the required temperature for
pilli formation (> 30°C) (Novotny and Lavin, 1971) due to
intermittent sewage flow. Both phage types revealed
increased numbers following incubation that can be attrib-
uted to continuous multiplication or shedding. It can be
stated that NK2 manhole does not contain detergent,
therefore the increase in phages number can be only
attributed to multiplication or continuous excretion of
phages by the bacterial hosts.
Survival of host E. coli strains used in the present study
in Yagur sewage
The second question related to host survival under harsh
environmental conditions was experimentally performed
under laboratory conditions. Figure 7 shows that three
E. coli strains (piliated F+, plasmid cured F+ and non-
piliated CN13) exposed to and grown for 8 days in Yagur
sewage filtrate revealed different survival patterns. Start-
ing at day one and up to day three the three E. coli strains
showed similar survival. Between the third and eighth day
the fastest drop in cell numbers was observed with E. coli
CN13, followed by E. coli F– and E. coli F+. From these
results and other observations (data not shown), it seems
that E. coli F+ (piliated strain) had a better survival capa-
bility in the sewage system under tested conditions. It is
true that suspended solids larger than 0.2 mm (sand, clay
or organic aggregates) removed by filtration are known to
protect microorganisms against environmental inactiva-
tion factors (Bitton and Mitchell, 1974); nevertheless,
other soluble chemical components were not removed,
consequently the sewage chemical composition remained
unchanged.
Relationship between E. coli F+ host and F male-specific
phages in sewage and faecal material
The association between E. coli F+ host and F male-
specific phages was studied in sewage and infants’ faecal
material. The motivation for this investigation derived from
fairly low F male-specific phage numbers at all stations
while compared with final sewage collected in collection
pond (P1) and further transfer to oxidation pond (P2). It
was assumed that these phages (that below 30°C do not
multiply due to lack of pili formation) (Myhal et al., 1982)
are related somehow to bacterial host concentration.
Several H sampling stations (H2, H4, H5), kitchen
manhole (NK1a), collection (P1) and oxidation (P2) ponds
were randomly sampled for both parameters and correla-
tion analysis was applied. Figure 8 shows the correlation
analysis between E. coli F+ (piliated E. coli) and F male-
specific phages at different sampling points. Manholes
with low F male-specific phage numbers (< 100 pfu ml-1)
revealed low numbers of host E. coli F+ and those with
higher phage numbers (> 103 pfu ml-1) harboured higher
host concentrations. These results were further supported
by infants’ faecal material screening.
Finally, known from our personal observations and
other studies, F male-specific phages are not excreted in
Table 1. Fate of somatic and F male-specific phages confined in polycarbonate bottle and placed in manhole during 48 h.
Sampling date Sampling site F-RNA coliphages (pfu 100 ml-1) Somatic coliphages (pfu 100 ml-1)
10 March NK1
NK2
1700 � 45
10 � 2
480 � 23
6 � 3
12 March NK1
NK2
7000 � 134
6350 � 157
6370 � 240
150 � 18
Multiplication factor F-RNA/somatic coliphages (after 48 h)
NK1 4.1/13.2
NK2 635/25
0
100
101
102
103
104
105
106
107
108
109
1010
cf
u
10
0
m
l-1
Time (Days)
E. coli F+
E. coliF- (cured)
E. coli CN
13
1 2 3 4 5 6 7 8 9
amp
Fig. 7. Survival of three E. coli host strains in Yagur sewage filtrate
as a function of incubation time at 24 � 2°C.
Bacteriophage ecology in community sewer system 2411
© 2007 The Authors
Journal compilation © 2007 Society for Applied Microbiology and Blackwell Publishing Ltd, Environmental Microbiology, 9, 2407–2416