DhakaUniv.J.Biol.Sci.

19(2):145150,2010(July)
BACTERIALLOADINTWELVEFRESHWATERFISHESOFFOUR
FEEDINGHABITSINBANGLADESH

M.SHAMSURRAHMAN
1
,MAHMUDHASAN
1
,ZAHIDHAYATMAHMUD
ANDM.SIRAJULISLAM*

InternationalCentreforDiarrhoealDiseaseResearch,Bangladesh(ICDDR,B),
Mohakhali,Dhaka,Bangladesh

Keywords:Bacterialload,Freshwaterfishes,Feedinghabit

Abstract
Homogenized samples of skin, gill and gut of twelve freshwater fishes of
four feeding habits were used for bacteriological test. Total bacterial count
(TBC),totalcoliform(TC),faecalcoliform(FC),faecalstreptococci(FS)andtotal
Vibriolike colonies were enumerated using serial dilution and spread plate
techniques. Variation of bacterial loads among the fishes of various feeding
habits was insignificant. The TBC, TC, FC, FS and total Vibrio counts ranged
from1.720.6810
8
to7.003.3910
8
,2.491.7210
6
to6.553.0010
6
,1.58
1.2910
6
to2.761.4210
6
,4.832.0910
4
to1.190.4610
5
and2.060.67
10
3
to 3.68 2.02 10
5
, respectively among various feeding groups. However,
gillandguthadsignificantlyhigherdensityofTBC,TCandFCthanthatofthe
skin. High counts of indicator organisms indicate that market fishes of all
varietiescouldbethepotentialreservoirsintransmissionofpathogenicbacteria
responsiblefordiarrhoealdisease.

Introduction
Bacterial decomposition has been shown to be a major factor causing rapid
deteriorationoffishquality.
(1)
Shewan
(2)
suggestedthatthebacterialfloraonfishreflects
the aquatic environment which affects the quality and storage life of fishery products.
Thequalityoffishcanbeassessedbythedegreeofcontaminationwithcoliformbacteria.
Contaminationresultsmainlyfromrupturingoffishintestineduringpoorprocessingor
unhealthy washing. Several studies have suggested that intestinal microflora or
contamination of fish as a result of enteric bacteria of human or animal origin are
responsible for food spoilages.
(3)
Considerable interest has been given in studying the
occurrenceanddistributionofthevariousbacterialpollutionindicatorsinthewaterand
the associated environment. It has been suggested repeatedly that the bacterial flora of
fish might reflect the bacteriological conditions of the water andapotential indicator of
pollution.Coliforms,fecalcoliforms,andfecalstreptococciarenotnormalinhabitantsof
thegutoffishandhaveashort,unalteredsurvivalintheirintestinaltract.

*Correspondingauthor:<sislam@icddrb.org>.
1
DepartmentofFisheries,UniversityofDhaka,
Dhaka1000,Bangladesh.
146 RAHMANetal.
In Bangladesh, every day 20,000 metric tons of human excreta are deposited on
public lands and waterways and is one of the main causes of contamination of surface
water.
(4)
Duringthetimeofrainyseasonthefecalmattersofvarioussourcesarewashed
awayfromthecontaminatedlandandareultimatelycarriedintodifferentwaterbodies
(e.g. ponds, rivers, lakes etc.). Moreover, due to the poor sanitary condition of the
country most of the latrines in rural settings are directly connected to the natural water
reservoirs. Thus, virtually all the aquatic habitats of Bangladesh are heavily
contaminated with fecal coliform bacteria. Fishes which live in these polluted habitats
thus can easily intake these bacteria while feeding along with contaminated aquatic
foods.
On the other hand, fish taken from estuarine waters or rivers, ponds, lakes and
canalsmaycarrypathogenicbacteriasuchasSalmonellaspp.,Shigellaspp.,Vibriocholerae
and other water borne pathogens. Fish of good quality should have counts of total
bacterialessthan10
5
pergramandfaecalcoliforms,staphylococcishouldnotexceed10/g
and100/g,respectively.
(5)

Migrationofthebacteriafromwatertofishmaycausespoilageoffishandoutbreak
of disease. The aim of the present study is to determine the bacterial load in different
organsoffishesofdifferentfeedinghabitstounderstandthemicrobiologicalstatusofthe
aquaticenvironment.

MaterialsandMethods
Twelve fishes of four feeding habits viz., herbivores, carnivores, detrivores and
omnivores were examined each with three replicates. The fishes were identified and
groupedfollowingKumar
(6)
,Rahman
(7)
,ShafiandQuddus
(8)
,Mustafaetal.
(9)
andNargis
and Hossain.
(10)
The samples were collected from Anand Bazar, Shoarighat and
Mohakhali bazar fish markets of Dhaka city. The samples were transported to the
laboratoryusingsterilizedpolythenebagsinicebox(<5C)withintwohoursofsampling
in the morning and processed within 2 4 hr of collection following aseptic techniques.
First, the samples were washed with sterile phosphate buffer saline (PBS) to remove
sand, detritus as well as microorganisms attached to the surface of fish. After washing
the skin, gill and gut were separated aseptically. The fish parts were separately
homogenized with sterile PBS solution using homogenizer and were then used for
bacteriologicalanalysis.
SpreadplatemethodwasusedtoenumeratethetotalbacteriaasdescribedbyHasan
and Bart
(11)
using dehydrated nutrient agar (Oxoid,USA) medium. For the enumeration
oftotalandfecalcoliform,100lsamplewasseriallydilutedandspreadedonthemFC
plate. Incubation was carried out for24 hr at 37C for total coliform and at44 to 44.5C
for fecal coliform. Fecal Streptococci (FS) were enumerated on KFSA plate, which was
incubated at 37C for 48 hr. After 48 hr incubation pink colonies were selected for

BACTERIALLOADINTWELVEFRESHWATERFISHES 147
countingofFS.TotalVibriolikecolonieswascountedonTCBSplate.Coloniesofyellow
andgreencolorafter1824hrofincubationwerecountedasVibriolikeorganism.

ResultsandDiscussion
Raw fishes are highly perishable protein source that contain normal bacterial flora
fromtheirenvironmentsinadditiontothecontaminantsoccurredduringharvestingand
handling of the products. The living fishes carry populations of predominantly Gram
negativepsychrotrophicbacteriaontheirexternalskin,nearly10
2
10
3
bacteriapergram.
Coliforms could be absent or present in very low density and Salmonella, Shigella, Vibrio
andotherentericpathogensareusuallynotfoundastheseorganismsarenotthenormal
flora of the fishes or of their environment.
(5)
However, in the present study the fish
samples of different feeding habits were found highly contaminated with total aerobic
bacteria as well as total coliform, fecal coliform and fecal streptococci (Table 1). This
mightbeduetothecontaminationofsourcewaterfromwherethefisheswerecaughtor
mightbeduetosecondarycontaminationduringthetimeofhandlingaswellasstorage
offishesinicemadewithcontaminatedwater.

Table1.Bacterialdensity(cfu/g)measuredindifferentfishesoffourdifferentfeedinghabits.

Bacterialload(cfu/g)
Feeding
habit
TBC TC FC FS TotalVibriolike
coloniesonTCBS
Herbivorous 2.911.5910
8
4.523.3910
6
1.581.2910
6
5.202.1010
4
8.504.7110
3

Carnivorous 5.052.3710
8
4.681.6910
6
2.501.1910
6
8.992.9710
4
2.060.6710
3

Detrivorous 7.003.3910
8
6.553.0010
6
2.761.4210
6
1.190.4610
5
3.682.0210
5

Omnivorous 1.720.6810
8
2.491.7210
6
1.751.2410
6
4.832.0910
4
7.363.1010
3

TBC=Totalbacterialcount;TC=Totalcoliformcount;FC=Fecalcoliformcount;FS=Fecalstreptococcalcount
Means(SEM)withincolumnandrowfortotalofeachparameterofbacterialdensitywithoutlettersdenoteno
significantdifferences(ANOVA,HSD,p<0.05).
.
Therewasnoremarkablevariationoftotalbacterialcount,totalcoliformcount,fecal
coliform count, fecal streptococcus count and total Vibriolike count among the different
groupsoffeedinghabitsofthefishes.
TBCandFCcountsofthevariouskindsoffishesweremoreorlesssimilarwhilethe
TC, FS and total Vibrio counts varied considerably among the different fish samples
(Table2).TheTCcountsvariedbetween5.144.3710
5
inhilsaand1.080.4310
7
in
boal.Allthefishsamplesshowedhighcountsoffecalcoliform(FC)(rangedbetween4.16
2.2910
4
insholand5.913.9610
6
incommoncarp)andfecalstreptococci(ranged
between1.331.0810
4
inhilsaand1.971.1810
5
incommoncarp).

148 RAHMANetal.
Likewise, common carp had significantly higher density of total Vibriolike colonies
thaninsilvercarpandhilsa.ThedensitiesoftotalVibriolikecoloniesonTCBSfoundin
otherfisheswerenotdifferentfromthoseofsilvercarp,hilsaandcommoncarp.
Comparing the bacterial counts in thefish organs, gilland gut showed significantly
higherdensityofTBC,TCandFCthanthatoftheskin(Table3).However,thedensityof
FS found in the gill was the highest while the lowest density was observed on the skin.
ThedensityofFSmeasuredinthegutwasnotdifferentfromgillandgut.

Table2.Bacterialdensity(cfu/g)measuredindifferentfishessampledfromdifferentmarketsof
Dhakametropolis.

Bacterialload(cfu/g) Fish
TBC TC FC FS TotalVibriolike
coloniesonTCBS
Silvercarp 1.231.0110
8
4.342.2110
5ab
2.431.0910
5
1.770.5610
4ab
6.673.7310
1b

Catla 2.361.1410
8
2.381.0510
6ab
4.843.5510
5
1.050.5710
5a
1.190.5510
3ab

Thaipangus 5.134.6110
8
1.071.0210
7ab
4.023.8710
6
3.282.1210
4ab
2.421.3010
4ab

Shol 1.210.5810
8
2.701.2310
5ab
4.162.2910
4
6.654.3910
4ab
3.121.7310
3ab

Boal 1.070.6810
9
1.080.4310
7a
5.773.3210
6
6.593.6010
4ab
2.030.9910
3ab

Meni 3.211.3810
8
3.001.4610
6ab
1.690.7710
6
1.370.7010
5ab
1.030.4110
3ab

Mrigal 9.008.6310
8
9.286.6810
6ab
1.881.3710
6
6.423.1310
4ab
4.092.5810
3ab

Common
carp
1.120.5510
9
9.376.0810
6ab
5.913.9610
6
1.971.1810
5ab
1.100.5510
6a

Rohu 7.453.6510
7
9.984.6910
5ab
5.002.1410
5
9.497.0010
4ab
3.621.5710
3ab

Koi 2.761.9410
8
6.745.0310
6ab
4.693.6310
6
1.100.5710
5ab
1.830.8210
4ab

Tilapia 1.330.5710
8
2.271.0010
5ab
1.210.6610
5
2.141.1110
4b
3.191.7010
3ab

Hilsa 1.070.4310
8
5.144.3710
5b
4.393.9610
5
1.331.0810
4b
6.005.1310
2b

TBC=Totalbacterialcount;TC=Totalcoliformcount;FC=Fecalcoliformcount;FS=Fecalstreptococcalcount
Means(SEM)withincolumnwithdifferentlettersaresignificantlydifferent(ANOVA,HSD;p<0.05).

Table3.Bacterialdensity(cfu/g)measuredindifferentorgansoffishessampledfromdifferent
marketsofDhakametropolis.

Bacterialload(cfu/g) Organs
of
fishes TBC TC FC FS
TotalVibriolike
coloniesonTCBS
Skin 1.370.4110
7b
1.580.7010
5b
8.494.5310
4b
2.871.4010
4b
2.722.3310
3

Gill 5.912.2110
8a
8.043.0110
6a
3.831.4610
6a
1.74.4010
5a
8.937.7510
4

Gut 6.462.4510
8a
5.492.2010
6a
2.541.1710
6a
2.930.7510
4ab
1.981.3510
5

TBC=Totalbacterialcount;TC=Totalcoliformcount;FC=Fecalcoliformcount;FS=Fecalstreptococcalcount
Means(SEM)withincolumnwithdifferentlettersaresignificantlydifferent(ANOVA,HSD,p<0.05).

BACTERIALLOADINTWELVEFRESHWATERFISHES 149
Coliform, faecal coliform and faecal streptococci count of the fishes of different
feeding habitats examined in this study exceeded the acceptable limit recommended by
Food and Agricultural Organization.
(5)
This indicates that freshwater fishes collected
frompond,river,lakeandcanaletc.arepotentialsourcesofdiseasecausingorganisms.
Therefore, precautions should be taken to prevent contamination during harvesting as
well as post harvest handling of fishes. Depending on the habitat and other
environmentalfactors,awiderangeofvariationindistributionofmicroflorainfishhas
beenreported.
(2)
Thepresentstudyagreeswiththesefindings.
According to the guideline of ICMSF,
(12)
acceptable limit of total bacterial counts for
giantprawnsandwhitefishare10
6
and510
5
cfu/g,respectively.Totalcoliform,faecal
coliform,andVibriocholeraecountsare10
2
,10and0cfu/g,respectively,forboththetypes
of fish. Therefore, the bacterial loads (total bacteria, total coliform, faecal coliform and
Vibrio)foundinthisstudyfordifferentfisheswerebeyondthestandardvaluesuggested
byICMSF,whichindicatetheirunacceptabilityasfoodfrompublichealthpointofview.
Theoccurrenceoffecalcoliforminfishcouldbeareflectionofthepollutionleveloftheir
water environment in addition to their feeding habits. It has been suggested repeatedly
thatthebacterialfloraoffishmightreflectthebacteriologicalconditionsofthewaterand
thus be a potential indicator of pollution.
(3)
Gibbons
(13)
studied on fish caught from the
marineenvironmentandconcludedthatcoliformbacteriaarenotusuallyassociatedwith
thenormalintestinalfloraoffish.However,organismsofthecoliformandstreptrococcal
groups have been isolated from various species of freshwater fish caught in relatively
clean to moderately polluted waters in India
(14)
, Norway
(15)
, Canada
(16)
, United States
(17)

etc. Margolis
(16)
reported that the bacterial flora of freshwater fish depends solely upon
thefishsrecentintakeoffoodandthedegreeofcontaminationinthefoodandwater.
Theabundanceofindicatorbacteriaaswellashighbacterialloadinthefishsamples
of different feeding habits indicates a possibility of future outbreak of pathogenic
diseasesthroughingestionofcontaminatedfishincludingchancesofcrosscontamination
via kitchen utensils or by handling. The presence of high bacterial load in the fish
samples may indicate fecally contaminated water sources and also poor hygienic
practicesbyfisherman.Hence,thedatamightbeofhelpinassessinghumanhealthrisk
duetoconsumptionoffreshwaterfishesofdifferentfeedinghabits.Indeed,careshould
be taken for longterm monitoring programme for potentially pathogenic bacteria in
freshwaterfishes.

References
1. Laycock RA and LW Regier 1971. Trimethylamine producing bacteria on haddock
(Melanogrammusaeglefinus)filletsduringrefrigeratedstorage.J.FishRes.Bd.Can.28:305.
2. Shewan JM 1976. The bacteriology of fresh and spoiling fish and the biochemical changes
inducebybacterialaction.In:Proceedingoftropicalinstituteconferenceonthehandling,
processingandmarketingoftropicalfish.TropicalProductsInstitute,London,pp.5166.

150 RAHMANetal.

3. Geldreich EE and NA Clarke 1966. Bacterial pollution indicators in the intestinal tract of
freshwaterfish.Appl.Microbiol.14:429437.
4. GoB (Government of Bangldesh) 2003. Water sanitation and hygiene promotion. A national
strategyforeconomicgrowth,povertyreductionandsocialdevelopment.Presentedtothe
GED,planningcommission,5March2003,GovernmentofBangladesh.
5. FAO (Food and Agricultural Organization) 1979. Manuals of food quality control: 4.
microbiologicalanalysis.FAOFoodandNutritionpaper14/4.
6. Kumar D 1992. Fish culture in undrainable ponds A manual for extension. FAO Fisheries
technical paper No. 325. FAO (Food and Agricultural Organization of United Nations),
Rome,p.239.
7. Rahman AKM 2005. Freshwater Fishes of Bangladesh, (2
nd
ed)., Zoological Society of
Bangladesh,Dhaka,Bangladesh,28;p.394.
8. Shafi M and MMA Quddus 2001. Bangladesher matsho shampad (Fisheries of Bangladesh).
KabirPublications,Dhaka.
9. Mustafa G, ATA Ahmed and KR Islam 1980. Food, Feeding habits and fecundity of a
freshwaterperch,menifish.J.Agri.Bangladesh.5:205210.
10. NargisAandMAHossain1987.Foodandfeedinghabitofkoifish(AnabastestudineusBloch).
Bangladesh.J.Agri.12:121127.
11. HasanMandANBart2007.Improvedsurvivalofrohu,Labeorohita(HamiltonBuchanan)and
silver carp, Hypophthalmichthys molitrix (Valenciennes) fingerlings using lowdose
quinaldineandbenzocaineduringtransport.Aquacult.Res.38:5058.
12. ICMSF1982.Microorganismsinfood.Vol.2,Samplingformicrobiologicalanalysis:principles
and specific applications. International Commission on the Microbiological Specification
ofFoods(ICMSF).Univ.TorontoPress,Toronto.
13. Gibbons NE 1934. The slime and intestinal flora of some marine fishes. Contrib. Can. Biol.
Fisheries(N.S.)8:275290.
14. Venkataraman R and A Sreenivasan 1953. The bacteriology of freshwater fish. Indian J. Med.
Res.41:385392.
15. ThjqttaTandOMSomme1943.Thebacterialfloraofnormalfish.SkrifterNorskeVidenskaps
Akad.OsloMat.Naturv.4:186.
16. Margolis L 1935. The effect of fasting on the bacterial flora of the intestine of fish. J. Fisheries
Res.BoardCan.10:6263.
17. Johnson GA 1904. Isolation of Bacillus coli communis from the alimentary tract of fish and the
significancethereof.J.Infect.Diseases1:348354.

(Manuscriptreceivedon5May,2009;revisedon13January,2010)