Microbiological quality of fish grown in wastewater-fed
and non-wastewater-fed fishponds in Hanoi, Vietnam:
influence of hygiene practices in local retail markets
Nguyen Thi Phong Lan, Anders Dalsgaard, Phung Dac Cam and Duncan Mara
ABSTRACT
Nguyen Thi Phong Lan
Phung Dac Cam
Department of Microbiology,
National Institute of Hygiene and Epidemiology,
1 Yersin St,10,000, Hanoi,
Viet Nam
Anders Dalsgaard (corresponding author)
Department of Veterinary Pathobiology,
Royal Veterinary and Agricultural University,
Grønnega° rdsvej 15, DK-1870, Frederiksberg C,
Denmark
Tel.:+45 35 282 720
Fax: +45 35 282 757 E-mail: ad@kvl.dk
Duncan Mara
School of Civil Engineering,
University of Leeds,
Leeds LS2 9JT,
UK
Mean water quality in two wastewater-fed ponds and one non-wastewater-fed pond in Hanoi,
Vietnam was ,106 and ,104 presumptive thermotolerant coliforms (pThC) per 100 ml,
respectively. Fish (common carp, silver carp and Nile tilapia) grown in these ponds were sampled
at harvest and in local retail markets. Bacteriological examination of the fish sampled at harvest
from both types of pond showed that they were of very good quality (2 2 3 pThC g21 fresh
muscle weight), despite the skin and gut contents being very contaminated (102 2 103 pThC g21
fresh weight and 104 2 106 pThC g21 fresh weight, respectively). These results indicate that the
WHO guideline quality of #1000 faecal coliforms per 100 ml of pond water in wastewater-fed
aquaculture is quite restrictive and represents a safety factor of ,3 orders of magnitude.
However, when the fish from both types of pond were sampled at the point of retail sale, quality
deteriorated to 102 2 105 pThC g21 of chopped fresh fish (mainly flesh and skin contaminated
with gut contents); this was due to the practice of the local fishmongers in descaling and
chopping up the fish from both types of pond with the same knife and on the same chopping
block. Fishmonger education is required to improve their hygienic practices; this should be
followed by regular hygiene inspections.
Key words | coliforms, fishculture, hygiene, retail markets, Vietnam, wastewater
INTRODUCTION
Fish production in excreta-fertilized fishponds is a very
ancient practice, especially in the Far East and notably
China where the practice is believed to have been initiated
over 3,000 years ago (Zhiwen 1999). In Vietnam wastewaters
are used for aquaculture as a source of both water
and nutrients (Vo 2001). The nutrients supports the growth
of plankton and other micro-organisms which are consumed
by the fish with little additional feeding taking place.
In periurban Hanoi there are ,2,500 ha of aquaculture
ponds, over 99 percent of which are used for fish culture,
mainly carp and tilapia, with a small area (,1 percent) for
shrimp production (Mai et al. 2004). Most of the wastewater-
fed fishponds are located in Thanh Tri district in the
south of the city, where there are ,330 ha of wastewaterfed
fishponds (Vo & Edwards 2005); there is also widespread
wastewater use for rice culture which is often
alternated with fish production (Tran 2001).
In order to assure the microbiological safety of fish
raised in wastewater-fed fishponds the World Health
Organization’s guideline is that the fishpond water should
have a faecal coliform count of #1000 per 100 ml (WHO
1989); this guideline value is expected to be retained in the
new guidelines which are currently being prepared (WHO
2006). Various bodies have made recommendations for the
microbiological quality of fish rather than the fishpond
water. For example, the International Commission on
Microbiological Specifications for Foods (1986) recommended
an ‘m’ value of 11 E. coli g21 and an ‘M’ value
doi: 10.2166/wh.2007.014
209 Q IWA Publishing 2007 Journal of Water and Health | 05.2 | 2007
of 500 E. coli g21 of uncooked fresh and frozen fish flesh,
where m and M are defined as follows: if the E. coli count is
,m the quality is ‘satisfactory’; if it is .M it is ‘unsatisfactory’;
and, if no more than three out of five fish samples
have values between m and M, it is ‘acceptable’. In Vietnam
the national standards are #100 E. coli g21 of uncooked
fresh and frozen fish flesh and #3 E. coli g21 of cooked fish
flesh (Ministry of Health 1998). A comprehensive review of,
and the corresponding rationales for, microbiological
criteria for safe fish are given in Institute of Medicine (2003).
Studies on the microbiological quality of fish raised in
wastewater-fed fishponds are few with some studies
indicating that faecal bacteria may penetrate the fish flesh
when fish is grown in highly polluted water (Buras et al.
1985, 1987; Buras 1990), whereas other studies found no or
little penetration of micro-organisms in aquaculture
environments in which the fish were not stressed (Edwards
1992). Furthermore, the level of microbiological crosscontamination
and quality of wastewater-fed fish sold to
consumers at retail markets are unknown. In this paper we
report the results of an investigation into the microbiological
quality of fish from wastewater-fed and non-wastewater-
fed fishponds in Thanh Tri district of Hanoi, both at
harvest and at the point of sale in local retail markets.
METHODS
Study locations and sampling
Fishponds
The study was carried out in two wastewater-fed ponds and
one nominally non-wastewater-fed (control) pond in Yen
So commune, Thanh Tri district. The areas of the
wastewater-fed ponds were ,3 and ,15 ha and their liquid
depths were ,1.5 2 2m. Both ponds were fed with raw
wastewater directly from the Kim Nguu River through a
pumping station located in the commune; ponds also
received direct discharges of domestic wastewater from
households around the ponds. The Kim Nguu river is
essentially a wastewater canal: CEETIA (1997) found it to be
heavily polluted, with biological oxygen demand (BOD)
and chemical oxygen demand (COD) concentrations some
3 2 7 times higher than the Vietnamese permitted standard
levels (#50mg BOD l21 and #100mg COD l21 for
wastewaters discharged into water bodies used for aquaculture
and crop irrigation). Toan (2004) found thermotolerant
coliform (ThC) numbers of 3 £ 107 per 100 ml in
the inlet of a fishpond in Yen So commune fed with water
from the Kim Nguu river.
The control pond, with an area of ,14 ha and a depth
of ,1.5 2 2m, was located on the alluvial plain adjacent to
the west bank of the Red River beyond a flood-control dyke.
Red River water was used to feed the control pond as
wastewater could not be economically pumped across the
dyke. Before the control pond was selected for the study,
samples of its contents were analysed for ThC numbers
(details in Results section below).
In Yen So commune the most commonly cultured fish
are common carp (Cyprinus carpio), silver carp
(Hypophthalamichthys molitrix), and Nile tilapia (Oreochromis
niloticus). The growing season is ,10 months and
at harvest common carp weigh ,500 2 600 g, silver carp
,200 2 300 g, and tilapia ,150 2 200 g. In this study, five
individual fish of each of these three species were collected
at ,7 a.m. immediately after they had been harvested from
the wastewater-fed and non-wastewater-fed ponds. Each
fish was placed in a sterile plastic bag. At the same time the
fish samples were collected, grab samples of the fishpond
water were collected from 15 2 20 cm below the surface in
sterile 500-ml glass sampling bottles. The fish and fishpond
water samples were then protected against heat and
sunlight and transported to the laboratory within 30
minutes. Samples were kept at 4 2 58C upon arrival at the
laboratory and analyzed within six hours of collection.
Local retail fish markets
There are several retail markets within 4 km of Yen So
commune to which fish are transported in bamboo baskets
on bicycles or motorbikes early in the morning. At the
market the fish are kept alive in small aerated basins filled
with tap water (Figure 1); the same basin is used for fish
from both wastewater-fed and non-wastewater-fed ponds.
The fishmongers, who are usually women, generally sit on
small wooden chairs close to the ground. They gut and
clean the fish on small wooden chopping boards placed on
the ground (Figure 2). Normally the scales are removed and
210 Nguyen Thi Phong Lan et al. | Microbiological quality of wastewater-fed fish in Hanoi Journal of Water and Health | 05.2 | 2007
the gut removed through a cut in the side of the fish. Carp
are then chopped into pieces, placed in a polythene bag and
sold. Tilapia are de-gutted and sold as whole fish after the
scales have been removed. The same knife is used for all
stages of fish processing. The fishmongers clean the
chopping board only twice a day, generally at the end of
the morning and afternoon trading sessions.
The fish sampled at the markets were ‘tracked’ from the
fishpond at harvesting and accompanied to the market, so
that it was known which fish came from the wastewater-fed
ponds or the non-wastewater pond. At the market whole
fish were purchased and the fishmonger asked to process
each fish in the normal way (i.e., to remove the scales and
gut the fish, then chop it into pieces). Each fish processed in
this way was then placed in a sterile plastic bag and taken
immediately to the laboratory for analysis.
Microbiological examination
Fish sampled at harvest
Samples of the skin, muscle and intestinal tract of the whole
fish samples were collected separately under aseptic
conditions, as follows:
(a) skin samples were taken from a 10-cm2 (2 £ 5 cm)
central area of the fish by marking out, using a sterile
template and scalpel, the outline of the desired area and
then removing, with sterile scalpel and forceps, as thin a
layer of the skin as possible (1 2 2 mm); the skin sample
was then placed in a sterile Petri dish.
(b) flesh (muscle) samples were taken by first sterilizing the
surface with a red-hot knife blade and then removing,
with sterile scalpel and forceps, the flesh immediately
below the singed surface so that a sample could be
taken of the raw flesh below; each sample collected in
this way weighed ,5g.
(c) the whole intestinal tract of each fish was removed
aseptically with sterile scalpel and forceps.
Similar sample types (skin, flesh or intestinal tract) from
each of five fish of a single species (common carp, silver
carp or tilapia) were removed, pooled, placed in a
polyethylene bag to give a five-fish composite sample,
which was then weighed. Nine times this weight of a
solution of 0.1% peptone and 0.85% sodium chloride at pH
7.5 was added and this 1-in-10 dilution was then
homogenized in a BagMixer model VW400 stomacher
(Interscience, St Nom, France) for 30 seconds. This dilution
Figure 1 | A local retail fish market in Yen So commune.
Figure 2 | Processing of a carp at a local retail market in Yen So commune.
211 Nguyen Thi Phong Lan et al. | Microbiological quality of wastewater-fed fish in Hanoi Journal of Water and Health | 05.2 | 2007
was then used for microbiological analyses directly or
diluted further, as described below.
Fish sampled at markets
A ,10-g sample of fish flesh was taken from one of the
pieces of fish in each of five plastic bags containing the same
fish species (common carp, silver carp or tilapia). These
samples were then pooled in a polyethylene bag to give a
five-fish composite sample which was then weighed. They
were then diluted and homogenized, as described above.
Bacteriological analyses
Serial 1-in-10 dilutions to 1027 were made of each fish or
wastewater sample using the peptone-NaCl diluent. Bacteriological
analyses for presumptive ThC, enterococci and aerobic
standard plate countswere then carried outwithin 30minutes
using the procedures recommended by the Nordic Committee
on Water and Food Analysis (Danish Standards Association
1999, 2001, 2002). Spread plates of membrane lauryl sulphate
agar (MM0615 broth with 15 g L0011agar l21; Oxoid Ltd,
Basingstoke,Hampshire,England) and Slanetz&Bartley agar
(Oxoid CM0377) were used for presumptive ThC and
enterococci, respectively, with incubation at 448C for 24h
(ThC) and 48h (enterococci). Pour-plates of tryptone yeast
extract agar (Oxoid CM1012 water plate count agar) were
used for standard plate counts (SPC) following incubation at
378C for 48 h. After incubation colonies growing on the agar
plates were enumerated and the counts of cell-forming units
(CFU) per g of fish (fresh weight) and CFU per100ml of
fishpond water determined.
Statistical analyses
The student t test was used to compare the geometric mean
results from the wastewater-fed and the non- wastewaterfed
ponds, and ANOVA for those from the three fish
species. The data were analyzed in Excel 2003 (Microsoft
Corp., Seattle, WA).
RESULTS
Fishpond water
The two wastewater-fed fishponds had significantly higher
mean counts of presumptive ThC (p , 0.0001) and enterococci
(p , 0.001) than the nominally non-wastewater-fed
pond: two orders of magnitude higher for presumptive ThC
and one order of magnitude higher for enterococci; there was
no difference in the standard plate counts (Table 1). The ThC
counts inthewastewater-fed pondswere nearly three orders of
magnitude higher than the WHO (1989) guideline value of
#1000 per 100 ml, whereas those in the nominally nonwastewater-
fed pond were less than one order of magnitude
above this guideline value.
Fish sampled at harvest
There was no major significant differences (i.e., those
important froma public health perspective) in bacteriological
Table 1 | Numbers of faecal indicator bacteria and standard plate counts in wastewater-fed and non-wastewater-fed fishponds
Wastewater-fed pondsa Non-wastewater-fed pond
Bacterial group Nc Meand s n Meanc s p (t test)b
Presumptive thermotolerant coliforms 9 5.92 0.91 10 3.79 0.57 0.0001
Enterococci 7 4.41 0.68 10 3.43 0.47 0.001
Standard plate count 8 7.79 1.02 10 7.57 0.79 0.183
aThere was no significant difference in the bacterial counts in the two wastewater-fed ponds (t test: p . 0.05).
bValues in bold indicate significant differences.
cNumber of water samples analysed.
dLog geometric mean bacterial numbers per 100 ml.
212 Nguyen Thi Phong Lan et al. | Microbiological quality of wastewater-fed fish in Hanoi Journal of Water and Health | 05.2 | 2007
qualities between the skin, gut contents or flesh for the three
fish species when comparing their origin from either wastewater-
fed or non-wastewater-fed ponds (Tables 2–4). Comparison
of bacterial numbers in skin samples from the three
fish species revealed no significant differences, except in one
casewhere skin samples fromwastewater-fed silver carp had a
higher SPC than non-wastewater-fed silver carp.
Fish from both wastewater-fed and non-wastewater-fed
ponds contained similar bacterial numbers in their gut
contents: 105–106 presumptive ThC g21 and 103–105 enterococci
g21 (Table3).Amongst the fishfromthewastewater-fed
ponds common carp contained significantly higher numbers
of presumptive ThC and enterococci than silver carp and
tilapia. Common carp are primarily bottom feeders and thus
will be exposed to high bacterial numbers in pond sediment,
whereas silver carp and tilapia primarily feed in the water
column where bacterial concentrations are lower.
Fish flesh samples collected by the stringently aseptic
technique contained no or very few faecal indicator
bacteria, whereas the SPC were ,103 CFU g21 (Table 4).
No significant differences were found in bacterial
numbers between fish from the wastewater-fed and the
non-wastewater fed ponds. Thus the very limited penetration
of faecal bacteria into the fish flesh came primarily
from the fish gut.
Fish sampled at point of retail sale
The bacteriological qualities of all three fish species from
both types of fishpond were substantially worse after
handling, cleaning and purchase in the local retail markets
than that at harvest: the geometric mean presumptive ThC
and enterococci counts in the fish samples from both the
wastewater-fed and non-wastewater-fed ponds were
Table 2 | Numbers of faecal indicator bacteria and standard plate counts on the skin of fish collected immediately after harvest from wastewater-fed and non-wastewater-fed ponds
Wastewater-fed pondsa
(nb 5 20)
Non-wastewater-fed pond
(n 5 18)
Bacterial group Fish n Meand s N Mean s p (t test)c
Presumptive thermotolerant coliforms Common carp 6 2.53 0.83 6 2.46 0.90 0.44
Silver carp 6 2.30 0.43 6 2.19 1.71 0.44
Tilapia 8 2.95 1.10 6 3.27 1.18 0.69
p (ANOVA) 0.38 0.35
Enterococci Common carp 6 1.93 1.35 6 2.51 0.84 0.80
Silver carp 6 2.07 1.48 6 1.88 1.29 0.40
Tilapia 8 3.10 1.10 6 3.08 0.99 0.48
p (ANOVA) 0.20 0.18
Standard plate counts Common carp 6 5.35 0.82 6 5.01 0.44 0.20
Silver carp 6 5.47 0.69 5 4.29 1.14 0.03
Tilapia 8 5.48 1.15 4 5.50 0.61 0.51
p (ANOVA) 0.96 0.10
aThere was no significant difference in the bacterial counts on the skin of the fish harvested from the two wastewater-fed ponds (t test: p . 0.05).
bNumber of skin samples analysed.
cValues in bold indicate significant differences.
dLog geometric mean bacterial numbers g21.
213 Nguyen Thi Phong Lan et al. | Microbiological quality of wastewater-fed fish in Hanoi Journal of Water and Health | 05.2 | 2007
102–105 CFU g21 and the SPC ranged from 106–107 CFU
g21 (Table 5). Numbers of presumptive ThC and enterococci
were significant higher in silver carp than in
common carp and tilapia. In general, there was no
significant difference between the bacteriological qualities
of the fish from the wastewater-fed ponds and those from
the non-wastewater-fed ponds.
DISCUSSION
The water in the non-wastewater-fed pond receiving water
from the Red River was faecally contaminated at a level of
just under 104 presumptive ThC per 100 ml, but the quality
of the flesh of fish from this pond at harvest showed little if
any faecal contamination (maximum 2–3 presumptive ThC
g21). The flesh from fish harvested from the much more
contaminated wastewater-fed ponds (just under 106 presumptive
ThC per 100 ml) contained similar levels of
presumptive ThC and was thus of an equally satisfactory
microbial quality. Thus very few faecal indicator bacteria
penetrated into the fish flesh even in the highly faecal
polluted wastewater-fed fish pond. However the SPC of the
fish flesh was 102–104 CFU g21, indicating that bacterial
penetration did occur, but at similar levels in the wastewater-
fed and non-wastewater-fed ponds.
A limited number of other studies have investigated the
association between microbiological qualities of the fishpond
water and the fish in both laboratory environments
and functioning waste-fed aquaculture ponds. A few studies,
mainly conducted in Israel, have suggested a threshold
bacterial concentration in the fishpond water above which
Table 3 | Numbers of faecal indicator bacteria and standard plate count in the gut of fish collected immediately after harvest from wastewater-fed and non-wastewater-fed ponds
Wastewater-fed pondsa
(nb 5 20)
Non-wastewater-fed pond
(n 5 18)
Bacterial group Fish N Meand s n Mean s p (t-test)c
Presumptive thermotolerant coliforms Common carp 6 5.33 1.12 6 6.19 0.85 0.91
Silver carp 6 4.67 0.91 6 4.65 0.95 0.48
Tilapia 8 5.17 1.07 6 4.62 1.47 0.21
p (ANOVA)c 0.53 0.04
Enterococci Common carp 6 3.75 1.01 6 5.11 1.35 0.96
Silver carp 6 3.36 0.64 6 3.25 0.62 0.39
Tilapia 8 3.42 0.60 6 2.57 0.90 0.02
p (ANOVA) 0.63 0.001
Standard plate counts Common carp 6 7.97 0.73 6 8.32 0.21 0.85
Silver carp 6 7.42 0.48 6 7.85 0.75 0.86
Tilapia 8 7.58 0.63 6 7.60 0.85 0.52
p (ANOVA) 0.30 0.20
aThere was no significant difference in the bacterial counts in the gut of the fish harvested from the two wastewater-fed ponds (t test: p . 0.05).
bNumbers of individual fish analysed.
cValues in bold indicate significant differences.
dLog geometric mean bacterial numbers g21.
214 Nguyen Thi Phong Lan et al. | Microbiological quality of wastewater-fed fish in Hanoi Journal of Water and Health | 05.2 | 2007
bacteria enter the edible muscle tissues of fish and thus
increase the risk of exposure for consumers of the fish.
Buras and co-workers (Buras et al. 1985, 1987; Buras 1990)
reported such a threshold concentration of total culturable
bacteria in fishpond water of 1 2 5 £ 106 per 100 ml, but
this seems to have been due to a major malfunction in the
wastewater treatment plant which introduced such high
organic loadings into the receiving fishpond that the fish
were extremely stressed and only just able to survive (P.
Edwards, personal communication, 2005). A study in
Thailand reported an SPC range in septage-fed fishponds
of 1.8 £ 105–2.0 £ 106 per 100 ml of pond water which
produced fish with minimal bacterial penetration into their
flesh (Edwards et al. 1984). Exposure of 132 healthy tilapia
to fishpond E. coli concentrations of up to 106 cfu per
100 ml from wastewater sources led to little or no detectable
bacterial or bacteriophage penetration into their flesh
(Fattal et al. 1988, 1993). In the United States Hejkal et al.
(1983) found a maximum of 25 faecal streptococci per 100 g
of fish muscle even though the gut contained .105 per
100 g. The fish in the Buras studies were grown under
conditions of high stress which is atypical of normal
aquaculture ponds; thus the penetration of micro-organisms
into the fish flesh in this study may have been an
exceptional case. The results of the current study, together
with other studies on well-managed ‘normal’ wastewaterfed
fishponds (reviewed by Edwards 1992), suggest that the
maximum permissible number of faecal indicator bacteria in
wastewater-fed fishponds should be less than that which
would lead to significant contamination of the fish flesh.
However, further research is needed to assess how many
orders of magnitude are needed to provide a realistic (i.e.,
Table 4 | Numbers of faecal indicator bacteria and standard plate count in the flesh of fish collected immediately after harvest from wastewater-fed and non-wastewater-fed ponds
Wastewater-fed pondsa,b
(nc 5 20)
Non-wastewater-fed pond
(n 5 18)
Bacterial group Fish n Meane s n Meanc s p (t-test)d
Presumptive thermotolerant coliforms Common carp 6 0.41 0.28 6 0.30 0 0.17
Silver carp 6 0.30 0 6 0.30 0 2
Tilapia 8 0.30 0 6 0.41 0.28 0.86
p (ANOVA)d 0.32 0.39
Enterococci Common carp 6 0.41 0.28 6 0.30 0 0.17
Silver carp 6 0.51 0.53 6 0.30 0 0.17
Tilapia 8 0.30 0 6 0.41 0.28 0.86
p (ANOVA) 0.48 0.39
Standard plate count Common carp 6 3.03 0.94 6 3.13 0.97 0.57
Silver carp 6 3.40 1.20 6 2.65 0.55 0.09
Tilapia 8 2.72 0.46 6 3.88 0.78 0.99
p (ANOVA) 0.38 0.03
aThere was no significant difference in the bacterial counts in the flesh of the fish harvested from the two wastewater-fed ponds (t test: p . 0.05).
bOnly three of the 20 fish examined had measurable numbers of presumptive ThC and enterococci per g of flesh (zero colony formation was recorded as ,2g21).
cNumbers of individual fish analysed.
dValues in bold indicate significant differences.
eLog geometric mean bacterial numbers g21.
215 Nguyen Thi Phong Lan et al. | Microbiological quality of wastewater-fed fish in Hanoi Journal of Water and Health | 05.2 | 2007
not over-restrictive) safety factor for faecal bacterial
indicator numbers in wastewater-fed fishponds.
Our results indicate that these fish flesh qualities were
satisfactory in terms of their faecal bacterial indicator
counts and complied with the recommendations of the
International Commission on Microbiological Specifications
for Foods (1986) and the Vietnamese Ministry of
Health (1998). They are in partial agreement with the fish
quality classification scheme developed by Buras et al.
(1987); this proposed “that in the case of fish grown in
wastewater, the quality of the fish should be determined by
the presence of any bacteria in the muscles” and “that the
indicators should be bacteria that grow on nutrient and
mFC agar, and the bacteriological quality should be
expressed as: 0 2 10 bacteria ml21, very good; 10 2 30
bacteria ml21, medium quality; more than 50 bacteria ml21,
not acceptable” [sic; it is not clear what quality was to be
assigned for 31 2 49 bacteria ml21]. Thus, based on this
scheme, the fish flesh qualities at harvest were ‘very good’
on the basis of their E. coli counts but ‘not acceptable’ on
the basis of their SPC (Table 4). It is difficult to imagine a
wastewater-fed (or river-water-fed) aquaculture situation in
which the “nutrient and mFC agar” counts are the same as
this would imply that all (or essentially all) the bacteria
present were faecal coliforms/E. coli. We thus accept the
classification of Buras et al. (1987) but only in terms of
the E. coli counts g21 of fish flesh and not in terms of the
SPC g21.
Table 5 | Numbers of faecal indicator bacteria and standard plate counts in fish samples (flesh, skin, bone) from wastewater-fed and non-wastewater-fed ponds purchased at local
retail markets
Wastewater-fed pondsa
(nb 5 52)
Non-wastewater-fed pond
(n 5 64)
Bacterial group Fish n Meand s n Mean s p (t test)c
Presumptive thermotolerant coliforms Common carp 10 2.89 0.69 20 3.45 1.80 0.82
Silver carp 20 4.23 1.35 20 4.28 1.28 0.55
Tilapia 22 3.49 0.78 24 3.73 0.95 0.81
p (ANOVA)c 0.004 0.15
Enterococci Common carp 10 2.68 0.92 20 3.23 1.29 0.87
Silver carp 20 4.33 1.26 20 3.70 0.69 0.02
Tilapia 22 3.68 0.63 24 3.54 0.57 0.22
p (ANOVA) 0.0003 0.24
Standard plate count Common carp 10 6.49 0.54 20 6.65 1.31 0.64
Silver carp 19 6.93 0.52 19 6.91 0.93 0.46
Tilapia 22 6.93 0.65 24 6.99 0.95 0.58
p (ANOVA) 0.11 0.57
aThere was no significant difference in the bacterial counts in the fish harvested from the two wastewater-fed ponds (t test: p . 0.05).
bNumbers of individual fish analysed.
cValues in bold indicate significant differences.
dLog geometric mean bacterial numbers g21.
216 Nguyen Thi Phong Lan et al. | Microbiological quality of wastewater-fed fish in Hanoi Journal of Water and Health | 05.2 | 2007
Surprisingly, the fish from both the wastewater-fed and
the non-wastewater fed ponds contained similar bacterial
concentrations in their gut contents (,105–106 presumptive
ThC g21 and ,103–105 enterococci g21). This may be
partly explained by the relatively high numbers of presumptive
ThC) in the non-wastewater pond (,104 per
100 ml). Our findings of high numbers of faecal bacteria in
the gut content are in agreement with several studies
reviewed by Edwards (1992). Fish samples (muscle, skin,
bone) purchased at retail markets contained from ,1,000
to ,20,000 presumptive ThC g21, irrespective of whether
the fish originated from the wastewater-fed or the nonwastewater
fed ponds. This indicates that significant faecal
cross-contamination occurred at the markets during handling
and processing of the fish for human consumption. As
noted by Buras et al. (1987), “exposure to pathogens can
occur when fish are handled and cleaned. During the
digestive tract removal, the content is usually spilled and
contaminates the intestinal cavity of the fish and the hands
of the handler. Casual rinsing does not prevent contamination”.
Clearly, local environmental health officers/assistants
need to educate local fishmongers so that (a) they are
aware of the risks for faecal contamination of the fish
products and possible occupational health risks of their
unhygienic practices and (b) they are then able to
implement and sustain improved hygiene practices; they
also need to be regularly inspected by the local environmental
health officers/assistants to ensure that their fishhandling
and cleaning practices are always hygienic.
Finally, it should be noted that only the bacteriological
quality of fish from wastewater-fed and non-wastewater-fed
fishponds was investigated by the use of bacterial indicators.
Thus, the possible occurrence and food-safety aspects of
fishborne zoonotic parasites, in particular trematode parasites,
bacterial and viral pathogens, and any bio-accumulation
of toxic chemicals in the wastewater, were not
assessed.
CONCLUSIONS
† Fish grown in both wastewater-fed and nominally nonwastewater-
fed fishponds with presumptive ThC counts
of ,106 and ,104 per 100 ml, respectively, were of very
good quality at harvest (2 2 3 presumptive ThC g21 of
flesh). This indicates that the current WHO guideline
value for wastewater-fed aquaculture (#1000 E. coli per
100 ml of fishpond water) is quite restrictive as it
represents a high factor of safety of three orders of
magnitude.
† Grossly unhygienic fish handling and cleaning practices
at the local retail markets caused significant recontamination
(102 2 105 presumptive ThC g21) of the fish
grown in both wastewater-fed and nominally nonwastewater-
fed ponds.
† The fishmongers in the local retail markets should be
informed about their unhygienic fish handling and
cleaning practices, and how these can be improved to
reduce the faecal cross-contamination of the fish they
sell.
ACKNOWLEDGEMENTS
We are grateful to the Danish International Development
Agency (Danida) which financially supported this research
through its Enhancement of Research Capacity (ENRECA)
programme through the project “Sanitary Aspects of
Drinking Water and Wastewater Reuse in Vietnam” (grant
no. 104.Dan.8.L). The study also received financial support
from the EU INCO-DEV project “Production in Aquatic
Peri-urban Systems in Southeast Asia” (PAPUSSA) (project
no:ICA4-2001-10072).
REFERENCES
Buras, N. 1990 Bacteriological guidelines for sewage-fed fish
culture. In Wastewater-fed Aquaculture. Proceedings of the
International Seminar on Wastewater Reclamation and Reuse
for Aquaculture, Calcutta, India, 6 2 9 December 1988 (ed.
P. Edwards & R. S. V. Pullin), pp. 223–226. Environmental
Sanitation Information Center, Asian Institute of Technology,
Bangkok, Thailand.
Buras, N., Duek, L. & Niv, S. 1985 Reactions of fish to microorganisms
in wastewater. Applied and Environmental
Microbiology 50, 989 2 995.
Buras, N., Duek, L., Niv, S., Hepher, B. & Sandbank, E. 1987
Microbiological aspects of fish grown in treated wastewater.
Water Research 21, 1–10.
217 Nguyen Thi Phong Lan et al. | Microbiological quality of wastewater-fed fish in Hanoi Journal of Water and Health | 05.2 | 2007
CEETIA 1997 Investigations into Environmental Pollution Caused
by Industrial Activities in Hanoi. (Report for the Hanoi
People’s Committee). Center for Environmental Engineering
and Technology in Urban and Industrial Areas, Hanoi
University of Civil Engineering, Hanoi, Vietnam.
Danish Standards Association 1999 Environmental Quality –
Enumeration of Enterococci – Colony Count on Solid
Medium. (Danish Standard No. DS 2401:1999). Danish
Standards Association, Charlottenlund, Denmark.
Danish Standards Association 2001 Water Quality – Detection and
Enumeration of Escherichia coli and Coliform bacteria – Part
1: Membrane Filtration Method. (Danish Standard No. DS/EN
ISO 9308-1:2001). Danish Standards Association,
Charlottenlund, Denmark.
Danish Standards Association 2002 Water Quality – Enumeration
of Culturable Micro-organisms – Colony Count by Inoculation
in a Nutrient Agar Culture Medium. (Danish Standard No.
DS/EN ISO 6222/Till.1:2002). Danish Standards Association,
Charlottenlund, Denmark.
Edwards, P. 1992 Reuse of Human Wastes in Aquaculture: A
Technical Review. Water and Sanitation Program, The World
Bank, Washington, DC.
Edwards, P., Pacharaprakiti, C., Kaewpaitoon, K., Rajput, V. S.,
Ruamthaveesub, P., Suthirawut, S., Yomjinda, M. & Chao,
C. H. 1984 Re-use of Cesspool Slurry and Cellulose
Agricultural Residues for Fish Culture. (AIT Research Report
No. 166). Asian Institute of Technology, Bangkok, Thailand.
Fattal, B., Dotan, A., Tchorsh, Y., Parpari, L. & Shuval, H. I. 1988
Penetration of E. coli and F2 bacteriophage into fish tissues.
Schriftenreihe des Vereins fu¨ r Wasser-, Boden- und
Lufthygiene 78, 27–38.
Fattal, B., Dotan, A., Parpari, L., Tchorsh, Y. & Cabelli, V. J. 1993
Microbiological purification of fish grown in fecally
contaminated commercial fish pond. Water, Science and
Technology 27(7 2 8), 303–311.
Hejkal, T. W., Gerba, C. P., Henderson, S. & Freeze, M. 1983
Bacteriological, virological and chemical evaluation of a
wastewater 2 aquaculture system. Water Research 17,
1749–1755.
Institute of Medicine 2003 Scientific Criteria to Ensure Safe Food.
National Academies Press, Washington, DC.
International Commission on Microbiological Specifications for
Foods 1986 Micro-organisms in Foods 2 –Sampling for
Microbiological Analysis: Principles and Specific
Applications, 2nd ed. Blackwell Scientific Publications,
Oxford, England.
Mai, T. P. A., Ali, M., Hoang, L. A. & To, T. T. H. 2004 Urban and
Peri-urban Agriculture in Hanoi: Opportunities and
Constraints for Safe and Sustainable Food Production.
(Technical Bulletin No. 32). AVRDC – The World Vegetable
Center, Tainan, Taiwan.
Ministry of Health 1998 Recommended Microbiological Limits for
Food. (Circular No. 91/429). Ministry of Health, Hanoi,
Vietnam.
Toan, T. Q. 2004 A Study of the Efficiency of Wastewater-fed
Fishponds as a Low-cost Wastewater Treatment System.
(internal report). National Institute of Occupational and
Environmental Health, Hanoi, Vietnam.
Tran, V. V. 2001 Sewage water aquaculture in Hanoi: current status
and further development. In Wastewater Reuse in Agriculture
in Vietnam: Water Management, Environment and Human
Health Aspects. IWMI Working Paper No. 30 (ed.
L. Raschid-Sally, W. van der Hoe & M. Ranawaka), pp.
24–25. International Water Management Institute, Colombo,
Sri Lanka.
Vo, Q. H. 2001 Wastewater reuse through aquaculture in Hanoi:
status and prospects. In Wastewater Reuse in Agriculture in
Vietnam: Water Management, Environment and Human
Health Aspects. IWMI Working Paper No. 30 (ed.
L. Raschid-Sally, W. van der Hoe & M. Ranawaka), pp.
20–23. International Water Management Institute, Colombo,
Sri Lanka.
Vo, Q. H. & Edwards, P. 2005 Wastewater reuse through urban
aquaculture in Hanoi: status and prospects. In Urban
Aquaculture (ed. B. Costa-Pierce, A. Desbonnet, P. Edwards &
D. Baker), pp. 103–117. CABI Publishing, Wallingford,
England.
WHO 1989 Health Guidelines for the Use of Wastewater in
Agriculture and Aquaculture. (Technical Report Series No.
778). World Health Organization, Geneva, Switzerland.
WHO 2005 Guidelines for the Safe Use of Wastewater, Excreta and
Greywater. Volume 3: wastewater and excreta use in
aquaculture. World Health Organization, Geneva,
Switzerland.
Zhiwen, S. 1999 Rural Aquaculture in China. (RAP Publication No.
1999/22). Regional Office for Asia and the Pacific, Food and
Agriculture Organization, Bangkok, Thailand.
Available online January 2007
218 Nguyen Thi Phong Lan et al. | Microbiological quality of wastewater-fed fish in Hanoi Journal of Water and Health | 05.2 | 2007
and non-wastewater-fed fishponds in Hanoi, Vietnam:
influence of hygiene practices in local retail markets
Nguyen Thi Phong Lan, Anders Dalsgaard, Phung Dac Cam and Duncan Mara
ABSTRACT
Nguyen Thi Phong Lan
Phung Dac Cam
Department of Microbiology,
National Institute of Hygiene and Epidemiology,
1 Yersin St,10,000, Hanoi,
Viet Nam
Anders Dalsgaard (corresponding author)
Department of Veterinary Pathobiology,
Royal Veterinary and Agricultural University,
Grønnega° rdsvej 15, DK-1870, Frederiksberg C,
Denmark
Tel.:+45 35 282 720
Fax: +45 35 282 757 E-mail: ad@kvl.dk
Duncan Mara
School of Civil Engineering,
University of Leeds,
Leeds LS2 9JT,
UK
Mean water quality in two wastewater-fed ponds and one non-wastewater-fed pond in Hanoi,
Vietnam was ,106 and ,104 presumptive thermotolerant coliforms (pThC) per 100 ml,
respectively. Fish (common carp, silver carp and Nile tilapia) grown in these ponds were sampled
at harvest and in local retail markets. Bacteriological examination of the fish sampled at harvest
from both types of pond showed that they were of very good quality (2 2 3 pThC g21 fresh
muscle weight), despite the skin and gut contents being very contaminated (102 2 103 pThC g21
fresh weight and 104 2 106 pThC g21 fresh weight, respectively). These results indicate that the
WHO guideline quality of #1000 faecal coliforms per 100 ml of pond water in wastewater-fed
aquaculture is quite restrictive and represents a safety factor of ,3 orders of magnitude.
However, when the fish from both types of pond were sampled at the point of retail sale, quality
deteriorated to 102 2 105 pThC g21 of chopped fresh fish (mainly flesh and skin contaminated
with gut contents); this was due to the practice of the local fishmongers in descaling and
chopping up the fish from both types of pond with the same knife and on the same chopping
block. Fishmonger education is required to improve their hygienic practices; this should be
followed by regular hygiene inspections.
Key words | coliforms, fishculture, hygiene, retail markets, Vietnam, wastewater
INTRODUCTION
Fish production in excreta-fertilized fishponds is a very
ancient practice, especially in the Far East and notably
China where the practice is believed to have been initiated
over 3,000 years ago (Zhiwen 1999). In Vietnam wastewaters
are used for aquaculture as a source of both water
and nutrients (Vo 2001). The nutrients supports the growth
of plankton and other micro-organisms which are consumed
by the fish with little additional feeding taking place.
In periurban Hanoi there are ,2,500 ha of aquaculture
ponds, over 99 percent of which are used for fish culture,
mainly carp and tilapia, with a small area (,1 percent) for
shrimp production (Mai et al. 2004). Most of the wastewater-
fed fishponds are located in Thanh Tri district in the
south of the city, where there are ,330 ha of wastewaterfed
fishponds (Vo & Edwards 2005); there is also widespread
wastewater use for rice culture which is often
alternated with fish production (Tran 2001).
In order to assure the microbiological safety of fish
raised in wastewater-fed fishponds the World Health
Organization’s guideline is that the fishpond water should
have a faecal coliform count of #1000 per 100 ml (WHO
1989); this guideline value is expected to be retained in the
new guidelines which are currently being prepared (WHO
2006). Various bodies have made recommendations for the
microbiological quality of fish rather than the fishpond
water. For example, the International Commission on
Microbiological Specifications for Foods (1986) recommended
an ‘m’ value of 11 E. coli g21 and an ‘M’ value
doi: 10.2166/wh.2007.014
209 Q IWA Publishing 2007 Journal of Water and Health | 05.2 | 2007
of 500 E. coli g21 of uncooked fresh and frozen fish flesh,
where m and M are defined as follows: if the E. coli count is
,m the quality is ‘satisfactory’; if it is .M it is ‘unsatisfactory’;
and, if no more than three out of five fish samples
have values between m and M, it is ‘acceptable’. In Vietnam
the national standards are #100 E. coli g21 of uncooked
fresh and frozen fish flesh and #3 E. coli g21 of cooked fish
flesh (Ministry of Health 1998). A comprehensive review of,
and the corresponding rationales for, microbiological
criteria for safe fish are given in Institute of Medicine (2003).
Studies on the microbiological quality of fish raised in
wastewater-fed fishponds are few with some studies
indicating that faecal bacteria may penetrate the fish flesh
when fish is grown in highly polluted water (Buras et al.
1985, 1987; Buras 1990), whereas other studies found no or
little penetration of micro-organisms in aquaculture
environments in which the fish were not stressed (Edwards
1992). Furthermore, the level of microbiological crosscontamination
and quality of wastewater-fed fish sold to
consumers at retail markets are unknown. In this paper we
report the results of an investigation into the microbiological
quality of fish from wastewater-fed and non-wastewater-
fed fishponds in Thanh Tri district of Hanoi, both at
harvest and at the point of sale in local retail markets.
METHODS
Study locations and sampling
Fishponds
The study was carried out in two wastewater-fed ponds and
one nominally non-wastewater-fed (control) pond in Yen
So commune, Thanh Tri district. The areas of the
wastewater-fed ponds were ,3 and ,15 ha and their liquid
depths were ,1.5 2 2m. Both ponds were fed with raw
wastewater directly from the Kim Nguu River through a
pumping station located in the commune; ponds also
received direct discharges of domestic wastewater from
households around the ponds. The Kim Nguu river is
essentially a wastewater canal: CEETIA (1997) found it to be
heavily polluted, with biological oxygen demand (BOD)
and chemical oxygen demand (COD) concentrations some
3 2 7 times higher than the Vietnamese permitted standard
levels (#50mg BOD l21 and #100mg COD l21 for
wastewaters discharged into water bodies used for aquaculture
and crop irrigation). Toan (2004) found thermotolerant
coliform (ThC) numbers of 3 £ 107 per 100 ml in
the inlet of a fishpond in Yen So commune fed with water
from the Kim Nguu river.
The control pond, with an area of ,14 ha and a depth
of ,1.5 2 2m, was located on the alluvial plain adjacent to
the west bank of the Red River beyond a flood-control dyke.
Red River water was used to feed the control pond as
wastewater could not be economically pumped across the
dyke. Before the control pond was selected for the study,
samples of its contents were analysed for ThC numbers
(details in Results section below).
In Yen So commune the most commonly cultured fish
are common carp (Cyprinus carpio), silver carp
(Hypophthalamichthys molitrix), and Nile tilapia (Oreochromis
niloticus). The growing season is ,10 months and
at harvest common carp weigh ,500 2 600 g, silver carp
,200 2 300 g, and tilapia ,150 2 200 g. In this study, five
individual fish of each of these three species were collected
at ,7 a.m. immediately after they had been harvested from
the wastewater-fed and non-wastewater-fed ponds. Each
fish was placed in a sterile plastic bag. At the same time the
fish samples were collected, grab samples of the fishpond
water were collected from 15 2 20 cm below the surface in
sterile 500-ml glass sampling bottles. The fish and fishpond
water samples were then protected against heat and
sunlight and transported to the laboratory within 30
minutes. Samples were kept at 4 2 58C upon arrival at the
laboratory and analyzed within six hours of collection.
Local retail fish markets
There are several retail markets within 4 km of Yen So
commune to which fish are transported in bamboo baskets
on bicycles or motorbikes early in the morning. At the
market the fish are kept alive in small aerated basins filled
with tap water (Figure 1); the same basin is used for fish
from both wastewater-fed and non-wastewater-fed ponds.
The fishmongers, who are usually women, generally sit on
small wooden chairs close to the ground. They gut and
clean the fish on small wooden chopping boards placed on
the ground (Figure 2). Normally the scales are removed and
210 Nguyen Thi Phong Lan et al. | Microbiological quality of wastewater-fed fish in Hanoi Journal of Water and Health | 05.2 | 2007
the gut removed through a cut in the side of the fish. Carp
are then chopped into pieces, placed in a polythene bag and
sold. Tilapia are de-gutted and sold as whole fish after the
scales have been removed. The same knife is used for all
stages of fish processing. The fishmongers clean the
chopping board only twice a day, generally at the end of
the morning and afternoon trading sessions.
The fish sampled at the markets were ‘tracked’ from the
fishpond at harvesting and accompanied to the market, so
that it was known which fish came from the wastewater-fed
ponds or the non-wastewater pond. At the market whole
fish were purchased and the fishmonger asked to process
each fish in the normal way (i.e., to remove the scales and
gut the fish, then chop it into pieces). Each fish processed in
this way was then placed in a sterile plastic bag and taken
immediately to the laboratory for analysis.
Microbiological examination
Fish sampled at harvest
Samples of the skin, muscle and intestinal tract of the whole
fish samples were collected separately under aseptic
conditions, as follows:
(a) skin samples were taken from a 10-cm2 (2 £ 5 cm)
central area of the fish by marking out, using a sterile
template and scalpel, the outline of the desired area and
then removing, with sterile scalpel and forceps, as thin a
layer of the skin as possible (1 2 2 mm); the skin sample
was then placed in a sterile Petri dish.
(b) flesh (muscle) samples were taken by first sterilizing the
surface with a red-hot knife blade and then removing,
with sterile scalpel and forceps, the flesh immediately
below the singed surface so that a sample could be
taken of the raw flesh below; each sample collected in
this way weighed ,5g.
(c) the whole intestinal tract of each fish was removed
aseptically with sterile scalpel and forceps.
Similar sample types (skin, flesh or intestinal tract) from
each of five fish of a single species (common carp, silver
carp or tilapia) were removed, pooled, placed in a
polyethylene bag to give a five-fish composite sample,
which was then weighed. Nine times this weight of a
solution of 0.1% peptone and 0.85% sodium chloride at pH
7.5 was added and this 1-in-10 dilution was then
homogenized in a BagMixer model VW400 stomacher
(Interscience, St Nom, France) for 30 seconds. This dilution
Figure 1 | A local retail fish market in Yen So commune.
Figure 2 | Processing of a carp at a local retail market in Yen So commune.
211 Nguyen Thi Phong Lan et al. | Microbiological quality of wastewater-fed fish in Hanoi Journal of Water and Health | 05.2 | 2007
was then used for microbiological analyses directly or
diluted further, as described below.
Fish sampled at markets
A ,10-g sample of fish flesh was taken from one of the
pieces of fish in each of five plastic bags containing the same
fish species (common carp, silver carp or tilapia). These
samples were then pooled in a polyethylene bag to give a
five-fish composite sample which was then weighed. They
were then diluted and homogenized, as described above.
Bacteriological analyses
Serial 1-in-10 dilutions to 1027 were made of each fish or
wastewater sample using the peptone-NaCl diluent. Bacteriological
analyses for presumptive ThC, enterococci and aerobic
standard plate countswere then carried outwithin 30minutes
using the procedures recommended by the Nordic Committee
on Water and Food Analysis (Danish Standards Association
1999, 2001, 2002). Spread plates of membrane lauryl sulphate
agar (MM0615 broth with 15 g L0011agar l21; Oxoid Ltd,
Basingstoke,Hampshire,England) and Slanetz&Bartley agar
(Oxoid CM0377) were used for presumptive ThC and
enterococci, respectively, with incubation at 448C for 24h
(ThC) and 48h (enterococci). Pour-plates of tryptone yeast
extract agar (Oxoid CM1012 water plate count agar) were
used for standard plate counts (SPC) following incubation at
378C for 48 h. After incubation colonies growing on the agar
plates were enumerated and the counts of cell-forming units
(CFU) per g of fish (fresh weight) and CFU per100ml of
fishpond water determined.
Statistical analyses
The student t test was used to compare the geometric mean
results from the wastewater-fed and the non- wastewaterfed
ponds, and ANOVA for those from the three fish
species. The data were analyzed in Excel 2003 (Microsoft
Corp., Seattle, WA).
RESULTS
Fishpond water
The two wastewater-fed fishponds had significantly higher
mean counts of presumptive ThC (p , 0.0001) and enterococci
(p , 0.001) than the nominally non-wastewater-fed
pond: two orders of magnitude higher for presumptive ThC
and one order of magnitude higher for enterococci; there was
no difference in the standard plate counts (Table 1). The ThC
counts inthewastewater-fed pondswere nearly three orders of
magnitude higher than the WHO (1989) guideline value of
#1000 per 100 ml, whereas those in the nominally nonwastewater-
fed pond were less than one order of magnitude
above this guideline value.
Fish sampled at harvest
There was no major significant differences (i.e., those
important froma public health perspective) in bacteriological
Table 1 | Numbers of faecal indicator bacteria and standard plate counts in wastewater-fed and non-wastewater-fed fishponds
Wastewater-fed pondsa Non-wastewater-fed pond
Bacterial group Nc Meand s n Meanc s p (t test)b
Presumptive thermotolerant coliforms 9 5.92 0.91 10 3.79 0.57 0.0001
Enterococci 7 4.41 0.68 10 3.43 0.47 0.001
Standard plate count 8 7.79 1.02 10 7.57 0.79 0.183
aThere was no significant difference in the bacterial counts in the two wastewater-fed ponds (t test: p . 0.05).
bValues in bold indicate significant differences.
cNumber of water samples analysed.
dLog geometric mean bacterial numbers per 100 ml.
212 Nguyen Thi Phong Lan et al. | Microbiological quality of wastewater-fed fish in Hanoi Journal of Water and Health | 05.2 | 2007
qualities between the skin, gut contents or flesh for the three
fish species when comparing their origin from either wastewater-
fed or non-wastewater-fed ponds (Tables 2–4). Comparison
of bacterial numbers in skin samples from the three
fish species revealed no significant differences, except in one
casewhere skin samples fromwastewater-fed silver carp had a
higher SPC than non-wastewater-fed silver carp.
Fish from both wastewater-fed and non-wastewater-fed
ponds contained similar bacterial numbers in their gut
contents: 105–106 presumptive ThC g21 and 103–105 enterococci
g21 (Table3).Amongst the fishfromthewastewater-fed
ponds common carp contained significantly higher numbers
of presumptive ThC and enterococci than silver carp and
tilapia. Common carp are primarily bottom feeders and thus
will be exposed to high bacterial numbers in pond sediment,
whereas silver carp and tilapia primarily feed in the water
column where bacterial concentrations are lower.
Fish flesh samples collected by the stringently aseptic
technique contained no or very few faecal indicator
bacteria, whereas the SPC were ,103 CFU g21 (Table 4).
No significant differences were found in bacterial
numbers between fish from the wastewater-fed and the
non-wastewater fed ponds. Thus the very limited penetration
of faecal bacteria into the fish flesh came primarily
from the fish gut.
Fish sampled at point of retail sale
The bacteriological qualities of all three fish species from
both types of fishpond were substantially worse after
handling, cleaning and purchase in the local retail markets
than that at harvest: the geometric mean presumptive ThC
and enterococci counts in the fish samples from both the
wastewater-fed and non-wastewater-fed ponds were
Table 2 | Numbers of faecal indicator bacteria and standard plate counts on the skin of fish collected immediately after harvest from wastewater-fed and non-wastewater-fed ponds
Wastewater-fed pondsa
(nb 5 20)
Non-wastewater-fed pond
(n 5 18)
Bacterial group Fish n Meand s N Mean s p (t test)c
Presumptive thermotolerant coliforms Common carp 6 2.53 0.83 6 2.46 0.90 0.44
Silver carp 6 2.30 0.43 6 2.19 1.71 0.44
Tilapia 8 2.95 1.10 6 3.27 1.18 0.69
p (ANOVA) 0.38 0.35
Enterococci Common carp 6 1.93 1.35 6 2.51 0.84 0.80
Silver carp 6 2.07 1.48 6 1.88 1.29 0.40
Tilapia 8 3.10 1.10 6 3.08 0.99 0.48
p (ANOVA) 0.20 0.18
Standard plate counts Common carp 6 5.35 0.82 6 5.01 0.44 0.20
Silver carp 6 5.47 0.69 5 4.29 1.14 0.03
Tilapia 8 5.48 1.15 4 5.50 0.61 0.51
p (ANOVA) 0.96 0.10
aThere was no significant difference in the bacterial counts on the skin of the fish harvested from the two wastewater-fed ponds (t test: p . 0.05).
bNumber of skin samples analysed.
cValues in bold indicate significant differences.
dLog geometric mean bacterial numbers g21.
213 Nguyen Thi Phong Lan et al. | Microbiological quality of wastewater-fed fish in Hanoi Journal of Water and Health | 05.2 | 2007
102–105 CFU g21 and the SPC ranged from 106–107 CFU
g21 (Table 5). Numbers of presumptive ThC and enterococci
were significant higher in silver carp than in
common carp and tilapia. In general, there was no
significant difference between the bacteriological qualities
of the fish from the wastewater-fed ponds and those from
the non-wastewater-fed ponds.
DISCUSSION
The water in the non-wastewater-fed pond receiving water
from the Red River was faecally contaminated at a level of
just under 104 presumptive ThC per 100 ml, but the quality
of the flesh of fish from this pond at harvest showed little if
any faecal contamination (maximum 2–3 presumptive ThC
g21). The flesh from fish harvested from the much more
contaminated wastewater-fed ponds (just under 106 presumptive
ThC per 100 ml) contained similar levels of
presumptive ThC and was thus of an equally satisfactory
microbial quality. Thus very few faecal indicator bacteria
penetrated into the fish flesh even in the highly faecal
polluted wastewater-fed fish pond. However the SPC of the
fish flesh was 102–104 CFU g21, indicating that bacterial
penetration did occur, but at similar levels in the wastewater-
fed and non-wastewater-fed ponds.
A limited number of other studies have investigated the
association between microbiological qualities of the fishpond
water and the fish in both laboratory environments
and functioning waste-fed aquaculture ponds. A few studies,
mainly conducted in Israel, have suggested a threshold
bacterial concentration in the fishpond water above which
Table 3 | Numbers of faecal indicator bacteria and standard plate count in the gut of fish collected immediately after harvest from wastewater-fed and non-wastewater-fed ponds
Wastewater-fed pondsa
(nb 5 20)
Non-wastewater-fed pond
(n 5 18)
Bacterial group Fish N Meand s n Mean s p (t-test)c
Presumptive thermotolerant coliforms Common carp 6 5.33 1.12 6 6.19 0.85 0.91
Silver carp 6 4.67 0.91 6 4.65 0.95 0.48
Tilapia 8 5.17 1.07 6 4.62 1.47 0.21
p (ANOVA)c 0.53 0.04
Enterococci Common carp 6 3.75 1.01 6 5.11 1.35 0.96
Silver carp 6 3.36 0.64 6 3.25 0.62 0.39
Tilapia 8 3.42 0.60 6 2.57 0.90 0.02
p (ANOVA) 0.63 0.001
Standard plate counts Common carp 6 7.97 0.73 6 8.32 0.21 0.85
Silver carp 6 7.42 0.48 6 7.85 0.75 0.86
Tilapia 8 7.58 0.63 6 7.60 0.85 0.52
p (ANOVA) 0.30 0.20
aThere was no significant difference in the bacterial counts in the gut of the fish harvested from the two wastewater-fed ponds (t test: p . 0.05).
bNumbers of individual fish analysed.
cValues in bold indicate significant differences.
dLog geometric mean bacterial numbers g21.
214 Nguyen Thi Phong Lan et al. | Microbiological quality of wastewater-fed fish in Hanoi Journal of Water and Health | 05.2 | 2007
bacteria enter the edible muscle tissues of fish and thus
increase the risk of exposure for consumers of the fish.
Buras and co-workers (Buras et al. 1985, 1987; Buras 1990)
reported such a threshold concentration of total culturable
bacteria in fishpond water of 1 2 5 £ 106 per 100 ml, but
this seems to have been due to a major malfunction in the
wastewater treatment plant which introduced such high
organic loadings into the receiving fishpond that the fish
were extremely stressed and only just able to survive (P.
Edwards, personal communication, 2005). A study in
Thailand reported an SPC range in septage-fed fishponds
of 1.8 £ 105–2.0 £ 106 per 100 ml of pond water which
produced fish with minimal bacterial penetration into their
flesh (Edwards et al. 1984). Exposure of 132 healthy tilapia
to fishpond E. coli concentrations of up to 106 cfu per
100 ml from wastewater sources led to little or no detectable
bacterial or bacteriophage penetration into their flesh
(Fattal et al. 1988, 1993). In the United States Hejkal et al.
(1983) found a maximum of 25 faecal streptococci per 100 g
of fish muscle even though the gut contained .105 per
100 g. The fish in the Buras studies were grown under
conditions of high stress which is atypical of normal
aquaculture ponds; thus the penetration of micro-organisms
into the fish flesh in this study may have been an
exceptional case. The results of the current study, together
with other studies on well-managed ‘normal’ wastewaterfed
fishponds (reviewed by Edwards 1992), suggest that the
maximum permissible number of faecal indicator bacteria in
wastewater-fed fishponds should be less than that which
would lead to significant contamination of the fish flesh.
However, further research is needed to assess how many
orders of magnitude are needed to provide a realistic (i.e.,
Table 4 | Numbers of faecal indicator bacteria and standard plate count in the flesh of fish collected immediately after harvest from wastewater-fed and non-wastewater-fed ponds
Wastewater-fed pondsa,b
(nc 5 20)
Non-wastewater-fed pond
(n 5 18)
Bacterial group Fish n Meane s n Meanc s p (t-test)d
Presumptive thermotolerant coliforms Common carp 6 0.41 0.28 6 0.30 0 0.17
Silver carp 6 0.30 0 6 0.30 0 2
Tilapia 8 0.30 0 6 0.41 0.28 0.86
p (ANOVA)d 0.32 0.39
Enterococci Common carp 6 0.41 0.28 6 0.30 0 0.17
Silver carp 6 0.51 0.53 6 0.30 0 0.17
Tilapia 8 0.30 0 6 0.41 0.28 0.86
p (ANOVA) 0.48 0.39
Standard plate count Common carp 6 3.03 0.94 6 3.13 0.97 0.57
Silver carp 6 3.40 1.20 6 2.65 0.55 0.09
Tilapia 8 2.72 0.46 6 3.88 0.78 0.99
p (ANOVA) 0.38 0.03
aThere was no significant difference in the bacterial counts in the flesh of the fish harvested from the two wastewater-fed ponds (t test: p . 0.05).
bOnly three of the 20 fish examined had measurable numbers of presumptive ThC and enterococci per g of flesh (zero colony formation was recorded as ,2g21).
cNumbers of individual fish analysed.
dValues in bold indicate significant differences.
eLog geometric mean bacterial numbers g21.
215 Nguyen Thi Phong Lan et al. | Microbiological quality of wastewater-fed fish in Hanoi Journal of Water and Health | 05.2 | 2007
not over-restrictive) safety factor for faecal bacterial
indicator numbers in wastewater-fed fishponds.
Our results indicate that these fish flesh qualities were
satisfactory in terms of their faecal bacterial indicator
counts and complied with the recommendations of the
International Commission on Microbiological Specifications
for Foods (1986) and the Vietnamese Ministry of
Health (1998). They are in partial agreement with the fish
quality classification scheme developed by Buras et al.
(1987); this proposed “that in the case of fish grown in
wastewater, the quality of the fish should be determined by
the presence of any bacteria in the muscles” and “that the
indicators should be bacteria that grow on nutrient and
mFC agar, and the bacteriological quality should be
expressed as: 0 2 10 bacteria ml21, very good; 10 2 30
bacteria ml21, medium quality; more than 50 bacteria ml21,
not acceptable” [sic; it is not clear what quality was to be
assigned for 31 2 49 bacteria ml21]. Thus, based on this
scheme, the fish flesh qualities at harvest were ‘very good’
on the basis of their E. coli counts but ‘not acceptable’ on
the basis of their SPC (Table 4). It is difficult to imagine a
wastewater-fed (or river-water-fed) aquaculture situation in
which the “nutrient and mFC agar” counts are the same as
this would imply that all (or essentially all) the bacteria
present were faecal coliforms/E. coli. We thus accept the
classification of Buras et al. (1987) but only in terms of
the E. coli counts g21 of fish flesh and not in terms of the
SPC g21.
Table 5 | Numbers of faecal indicator bacteria and standard plate counts in fish samples (flesh, skin, bone) from wastewater-fed and non-wastewater-fed ponds purchased at local
retail markets
Wastewater-fed pondsa
(nb 5 52)
Non-wastewater-fed pond
(n 5 64)
Bacterial group Fish n Meand s n Mean s p (t test)c
Presumptive thermotolerant coliforms Common carp 10 2.89 0.69 20 3.45 1.80 0.82
Silver carp 20 4.23 1.35 20 4.28 1.28 0.55
Tilapia 22 3.49 0.78 24 3.73 0.95 0.81
p (ANOVA)c 0.004 0.15
Enterococci Common carp 10 2.68 0.92 20 3.23 1.29 0.87
Silver carp 20 4.33 1.26 20 3.70 0.69 0.02
Tilapia 22 3.68 0.63 24 3.54 0.57 0.22
p (ANOVA) 0.0003 0.24
Standard plate count Common carp 10 6.49 0.54 20 6.65 1.31 0.64
Silver carp 19 6.93 0.52 19 6.91 0.93 0.46
Tilapia 22 6.93 0.65 24 6.99 0.95 0.58
p (ANOVA) 0.11 0.57
aThere was no significant difference in the bacterial counts in the fish harvested from the two wastewater-fed ponds (t test: p . 0.05).
bNumbers of individual fish analysed.
cValues in bold indicate significant differences.
dLog geometric mean bacterial numbers g21.
216 Nguyen Thi Phong Lan et al. | Microbiological quality of wastewater-fed fish in Hanoi Journal of Water and Health | 05.2 | 2007
Surprisingly, the fish from both the wastewater-fed and
the non-wastewater fed ponds contained similar bacterial
concentrations in their gut contents (,105–106 presumptive
ThC g21 and ,103–105 enterococci g21). This may be
partly explained by the relatively high numbers of presumptive
ThC) in the non-wastewater pond (,104 per
100 ml). Our findings of high numbers of faecal bacteria in
the gut content are in agreement with several studies
reviewed by Edwards (1992). Fish samples (muscle, skin,
bone) purchased at retail markets contained from ,1,000
to ,20,000 presumptive ThC g21, irrespective of whether
the fish originated from the wastewater-fed or the nonwastewater
fed ponds. This indicates that significant faecal
cross-contamination occurred at the markets during handling
and processing of the fish for human consumption. As
noted by Buras et al. (1987), “exposure to pathogens can
occur when fish are handled and cleaned. During the
digestive tract removal, the content is usually spilled and
contaminates the intestinal cavity of the fish and the hands
of the handler. Casual rinsing does not prevent contamination”.
Clearly, local environmental health officers/assistants
need to educate local fishmongers so that (a) they are
aware of the risks for faecal contamination of the fish
products and possible occupational health risks of their
unhygienic practices and (b) they are then able to
implement and sustain improved hygiene practices; they
also need to be regularly inspected by the local environmental
health officers/assistants to ensure that their fishhandling
and cleaning practices are always hygienic.
Finally, it should be noted that only the bacteriological
quality of fish from wastewater-fed and non-wastewater-fed
fishponds was investigated by the use of bacterial indicators.
Thus, the possible occurrence and food-safety aspects of
fishborne zoonotic parasites, in particular trematode parasites,
bacterial and viral pathogens, and any bio-accumulation
of toxic chemicals in the wastewater, were not
assessed.
CONCLUSIONS
† Fish grown in both wastewater-fed and nominally nonwastewater-
fed fishponds with presumptive ThC counts
of ,106 and ,104 per 100 ml, respectively, were of very
good quality at harvest (2 2 3 presumptive ThC g21 of
flesh). This indicates that the current WHO guideline
value for wastewater-fed aquaculture (#1000 E. coli per
100 ml of fishpond water) is quite restrictive as it
represents a high factor of safety of three orders of
magnitude.
† Grossly unhygienic fish handling and cleaning practices
at the local retail markets caused significant recontamination
(102 2 105 presumptive ThC g21) of the fish
grown in both wastewater-fed and nominally nonwastewater-
fed ponds.
† The fishmongers in the local retail markets should be
informed about their unhygienic fish handling and
cleaning practices, and how these can be improved to
reduce the faecal cross-contamination of the fish they
sell.
ACKNOWLEDGEMENTS
We are grateful to the Danish International Development
Agency (Danida) which financially supported this research
through its Enhancement of Research Capacity (ENRECA)
programme through the project “Sanitary Aspects of
Drinking Water and Wastewater Reuse in Vietnam” (grant
no. 104.Dan.8.L). The study also received financial support
from the EU INCO-DEV project “Production in Aquatic
Peri-urban Systems in Southeast Asia” (PAPUSSA) (project
no:ICA4-2001-10072).
REFERENCES
Buras, N. 1990 Bacteriological guidelines for sewage-fed fish
culture. In Wastewater-fed Aquaculture. Proceedings of the
International Seminar on Wastewater Reclamation and Reuse
for Aquaculture, Calcutta, India, 6 2 9 December 1988 (ed.
P. Edwards & R. S. V. Pullin), pp. 223–226. Environmental
Sanitation Information Center, Asian Institute of Technology,
Bangkok, Thailand.
Buras, N., Duek, L. & Niv, S. 1985 Reactions of fish to microorganisms
in wastewater. Applied and Environmental
Microbiology 50, 989 2 995.
Buras, N., Duek, L., Niv, S., Hepher, B. & Sandbank, E. 1987
Microbiological aspects of fish grown in treated wastewater.
Water Research 21, 1–10.
217 Nguyen Thi Phong Lan et al. | Microbiological quality of wastewater-fed fish in Hanoi Journal of Water and Health | 05.2 | 2007
CEETIA 1997 Investigations into Environmental Pollution Caused
by Industrial Activities in Hanoi. (Report for the Hanoi
People’s Committee). Center for Environmental Engineering
and Technology in Urban and Industrial Areas, Hanoi
University of Civil Engineering, Hanoi, Vietnam.
Danish Standards Association 1999 Environmental Quality –
Enumeration of Enterococci – Colony Count on Solid
Medium. (Danish Standard No. DS 2401:1999). Danish
Standards Association, Charlottenlund, Denmark.
Danish Standards Association 2001 Water Quality – Detection and
Enumeration of Escherichia coli and Coliform bacteria – Part
1: Membrane Filtration Method. (Danish Standard No. DS/EN
ISO 9308-1:2001). Danish Standards Association,
Charlottenlund, Denmark.
Danish Standards Association 2002 Water Quality – Enumeration
of Culturable Micro-organisms – Colony Count by Inoculation
in a Nutrient Agar Culture Medium. (Danish Standard No.
DS/EN ISO 6222/Till.1:2002). Danish Standards Association,
Charlottenlund, Denmark.
Edwards, P. 1992 Reuse of Human Wastes in Aquaculture: A
Technical Review. Water and Sanitation Program, The World
Bank, Washington, DC.
Edwards, P., Pacharaprakiti, C., Kaewpaitoon, K., Rajput, V. S.,
Ruamthaveesub, P., Suthirawut, S., Yomjinda, M. & Chao,
C. H. 1984 Re-use of Cesspool Slurry and Cellulose
Agricultural Residues for Fish Culture. (AIT Research Report
No. 166). Asian Institute of Technology, Bangkok, Thailand.
Fattal, B., Dotan, A., Tchorsh, Y., Parpari, L. & Shuval, H. I. 1988
Penetration of E. coli and F2 bacteriophage into fish tissues.
Schriftenreihe des Vereins fu¨ r Wasser-, Boden- und
Lufthygiene 78, 27–38.
Fattal, B., Dotan, A., Parpari, L., Tchorsh, Y. & Cabelli, V. J. 1993
Microbiological purification of fish grown in fecally
contaminated commercial fish pond. Water, Science and
Technology 27(7 2 8), 303–311.
Hejkal, T. W., Gerba, C. P., Henderson, S. & Freeze, M. 1983
Bacteriological, virological and chemical evaluation of a
wastewater 2 aquaculture system. Water Research 17,
1749–1755.
Institute of Medicine 2003 Scientific Criteria to Ensure Safe Food.
National Academies Press, Washington, DC.
International Commission on Microbiological Specifications for
Foods 1986 Micro-organisms in Foods 2 –Sampling for
Microbiological Analysis: Principles and Specific
Applications, 2nd ed. Blackwell Scientific Publications,
Oxford, England.
Mai, T. P. A., Ali, M., Hoang, L. A. & To, T. T. H. 2004 Urban and
Peri-urban Agriculture in Hanoi: Opportunities and
Constraints for Safe and Sustainable Food Production.
(Technical Bulletin No. 32). AVRDC – The World Vegetable
Center, Tainan, Taiwan.
Ministry of Health 1998 Recommended Microbiological Limits for
Food. (Circular No. 91/429). Ministry of Health, Hanoi,
Vietnam.
Toan, T. Q. 2004 A Study of the Efficiency of Wastewater-fed
Fishponds as a Low-cost Wastewater Treatment System.
(internal report). National Institute of Occupational and
Environmental Health, Hanoi, Vietnam.
Tran, V. V. 2001 Sewage water aquaculture in Hanoi: current status
and further development. In Wastewater Reuse in Agriculture
in Vietnam: Water Management, Environment and Human
Health Aspects. IWMI Working Paper No. 30 (ed.
L. Raschid-Sally, W. van der Hoe & M. Ranawaka), pp.
24–25. International Water Management Institute, Colombo,
Sri Lanka.
Vo, Q. H. 2001 Wastewater reuse through aquaculture in Hanoi:
status and prospects. In Wastewater Reuse in Agriculture in
Vietnam: Water Management, Environment and Human
Health Aspects. IWMI Working Paper No. 30 (ed.
L. Raschid-Sally, W. van der Hoe & M. Ranawaka), pp.
20–23. International Water Management Institute, Colombo,
Sri Lanka.
Vo, Q. H. & Edwards, P. 2005 Wastewater reuse through urban
aquaculture in Hanoi: status and prospects. In Urban
Aquaculture (ed. B. Costa-Pierce, A. Desbonnet, P. Edwards &
D. Baker), pp. 103–117. CABI Publishing, Wallingford,
England.
WHO 1989 Health Guidelines for the Use of Wastewater in
Agriculture and Aquaculture. (Technical Report Series No.
778). World Health Organization, Geneva, Switzerland.
WHO 2005 Guidelines for the Safe Use of Wastewater, Excreta and
Greywater. Volume 3: wastewater and excreta use in
aquaculture. World Health Organization, Geneva,
Switzerland.
Zhiwen, S. 1999 Rural Aquaculture in China. (RAP Publication No.
1999/22). Regional Office for Asia and the Pacific, Food and
Agriculture Organization, Bangkok, Thailand.
Available online January 2007
218 Nguyen Thi Phong Lan et al. | Microbiological quality of wastewater-fed fish in Hanoi Journal of Water and Health | 05.2 | 2007
No comments:
Post a Comment