2007) 200–206www.elsevier.com/locate/aqua-online

Aquaculture 270 (

Effects of Cu2+-exchanged montmorillonite on growth performance,microbial ecology and intestinal morphology of Nile tilapia

(Oreochromis niloticus)

C.H. Hu a,⁎, Y. Xu a, M.S. Xia b, L. Xiong a, Z.R. Xu a

a College of Animal Science, Zhejiang University, The Key Laboratory of Molecular Animal Nutrition of Ministry of Education,Hangzhou, 310029, China

b Department of Earth Science, Zhejiang University, Hangzhou, 310027, China

Received 10 September 2006; received in revised form 19 January 2007; accepted 19 January 2007

Abstract

A total of 360 Nile tilapia at an average initial body weight of 3.9 g were randomly allocated to 4 treatments, each of which hadthree replicates of 30 fish/tank and used to investigate the effects of Cu2+-exchanged montmorillonite (Cu-MMT) on growthperformance, microbial ecology of the skin, gill and intestine, and intestinal morphology. The dietary treatments were: 1) basal diet,2) basal diet +1.5 g/kg MMT, 3) basal diet +30 mg/kg copper as CuSO4 (equivalent to the copper in the Cu-MMT treatmentgroup), or 4) basal diet +1.5 g/kg Cu-MMT. The results showed that supplementation with Cu-MMT significantly improvedgrowth performance as compared to control and fish fed with Cu-MMT had higher growth performance than those fed with MMTor CuSO4. Supplementation with Cu-MMT reduced (Pb0.05) the total intestinal aerobic bacterial counts and affected thecomposition of intestinal microflora with a tendency of Aeromonas, Vibrio, Pseudomonas, Flavobacterium, Acinetobacter, Al-caligence, Enterobacteriaceae decreasing as compared with control. Measurements of villus and microvillus heights at differentsites of the intestinal mucosa indicated that dietary addition of MMT or Cu-MMT improved intestinal mucosal morphology.However, the microbial ecology of the skin and gill was not affected by the addition of CuSO4, MMT or Cu-MMT.Supplementation with CuSO4 had no (PN0.05) effect on the growth performance, microbial ecology and intestinal morphology.The results showed that Cu-MMT exhibited antibacterial activity in vivo and protected intestinal mucosa from invasion ofpathogenic bacterium and toxins, resulting in a positive effect on the growth performance.© 2007 Elsevier B.V. All rights reserved.

Keywords: Cu2+-exchanged montmorillonite; Growth performance; Microbial ecology; Intestinal morphology; Nile tilapia

1. Introduction

Bacterial disease is a major problem encountered inthe rapidly developing fish farming industry. Although

⁎ Corresponding author. Feed Science Institute, Animal ScienceCollege, Zhejiang University, Hangzhou, 310029, PR China. Tel./fax:+86 571 86985607.

E-mail address: chhu@zju.edu.cn (C.H. Hu).

0044-8486/$ - see front matter © 2007 Elsevier B.V. All rights reserved.doi:10.1016/j.aquaculture.2007.01.027

vaccines are being developed and marketed, they cannotbe used alone as a universal disease control measure inaquaculture. Treatment with antibiotics continues to beimportant disease control measures in the aquacultureindustry. However, abuse or overuse of antibioticscauses various side-effects and also results in theemergence and increase of bacteria resistant to anti-biotics. Therefore, it is necessary to develop new typesof antibacterial agents that can replace antibiotics.

Table 1Ingredients and nutrient composition of diets

Ingredients % Nutrition composition %

Fish meal 10 Gross energy (kJ/g) 18.1Soybean meal 30 Crude protein 35.2Rapeseed meal 25 Crude lipid 2.5Wheat middlings 26 Fibre 6.8Corn oil 2 Ash 11.1Vitamin premix a 1Mineral premix b 5Carboxymethyl cellulose 1

a Vitamin premix (mg/ kg of diet): thiamine, 10; riboflavin, 20;pyridoxine, 10; cobalamin, 2; retinal, 4; cholecalciferol, 0.4;phylloquinone, 80; folic acid, 5; Calcium pantotheniate, 40; inositol,400; niacin, 150; tocopherol, 60; choline, 6000; ascorbic acid, 500.b Mineral premix (g/kg of diet): NaCl, 0.25; MgSO4, 3.75; KH2PO4,

8; Ca(H2PO4)2, 5; FeSO4, 0.72; (CH2CHCOO)2Ca.5H20, 0.88;ZnSO4.7H2O, 0.088; MnSO4.4H2O, 0.040; CuSO4.5H2O, 0.008;CoCl2.6H2O, 0.00025; KIO3.6H2O, 0.00075.

201C.H. Hu et al. / Aquaculture 270 (2007) 200–206

The metal-carrying inorganic antibacterial agents,being carried by non-metallic minerals, have attractedattention by durable antibacterial properties, non-resistance, and safety. Montmorillonite (MMT) is asort of aluminum silicate with 2:1 layer structure oftetrahedral and octahedral layers. Between the structuralsheets, there are exchangeable cations readily beingreplaced by other cations or compounds. Taking intoaccount that MMT has specific physical–chemicalproperties such as high surface area, strong absorptivepower, high structural stability, chemical inertia andstrong capacity to form stable suspensions, MMT maybe suitable as an antimicrobial carrier. Our recentexperiments revealed that MMT had no antibacterialactivity, while Cu2+-exchanged montmorillonite (Cu-MMT) has strong antibacterial ability on Aeromonashydrophila, Pseudomonas fluorescens, Vibrio parahae-molyticus and Escherichia coli K88 (Hu and Xia, 2005;Hu et al., 2006).

There are no data of the effects of Cu-MMTon fish invivo. In the present study, the effects of Cu-MMT ongrowth performance, microbial ecology of the skin, gilland intestine, and intestinal morphology of Nile tilapia(Oreochromis niloticus) were investigated.

2. Materials and methods

2.1. Materials

MMT is a hydrothermal product of volcano sedi-mentary rocks from the Inner Mongolia AutonomousRegion, China. Besides MMT, there were minoramounts of quartz and volcanic glass in the ore. Theraw material was dried in oven over night at 80 °C andthen milled to less than 300 mesh. The milled materialwas dispersed in water to form a 10% suspension andkept for about 10 min with stirring. Particles larger than1 μm were separated out by sedimentation while thesuspension was centrifuged to get refined MMT. Therefined MMT was dried at 80 °C followed by anothermilling to less than 300 mesh for use. The formula of thepurified MMT was [Na0.158K0.082Ca0.256Mg0.063][Mg0.376Fe

2+0.014Fe

3+0.136Al1.474][Si3.87Al0.13]O10

(OH)2·nH2O.Cu-MMTwas prepared by Cu2+-exchanged reaction.

Ten grams of the refined MMTwere mixed with 100 mlof 0.2 mol/L NaCl solution. The dispersion was kept for5 h with stirring. The Na-MMT were then separated bycentrifugation at a speed of 8000 g for about 15 min andwashed with deionized water for three times. Thewashed Na-MMT was then dispersed in 200 ml of0.05 mol/L CuSO4 solution and pH value was adjusted

to 5.0. The dispension was kept at 60 °C with stirring for6 h. After centrifugation, the sediment was washed withdeionized water for three times, dried at 80 °C overnight, then ground to a size less than 300 mesh. Thecopper concentration in Cu-MMTwas found to be 2.0%measured by atomic absorption spectroscopy.

2.2. Experimental design and fish rearing

Nile tilapia (O. niloticus) fingerlings were providedby WuYI fishery breeding field and acclimated tolaboratory conditions for 2 weeks, fed with a conven-tional diet, after which time a total of 360 Nile tilapia atan average initial body weight of 3.9 g were randomlyallocated to four dietary treatments for 56 days, each ofwhich had three replicates of 30 fish/tank. The dietarytreatments were: 1) basal diet, 2) basal diet +1.5 g/kgMMT, 3) basal diet +30 mg/kg copper as CuSO4

(equivalent to the copper in the Cu-MMT treatmentgroup), or 4) basal diet +1.5 g/kg Cu-MMT. Diets wereformulated to meet nutrient requirements suggested bythe NRC (1993) for Nile tilapia. No antibiotic wasincluded in diets (Table 1).

The experimental system was a closed recirculationsystem consisting of 12 self-cleaning aquarium (50 cmwide×100 cm long×50 cm high), sedimentation tanks, abiological filter and a UV filter to prevent cross-contamination of micro-organisms between treatments.The system was installed in an environment-controlledlaboratorymaintained at 25 °C, with a photoperiod of 12 hlight and 12 h darkness. The culture system was alsoprovided with continuous aeration through an aircompressor and heaters to keep water temperature at

Table 2Effect of Cu-MMT on growth performance of Nile tilapia1

Control CuSO4 MMT2 Cu-MMT3 SEM4

BWi5, g 4.08 3.94 3.86 3.82 0.11

BWf6, g 29.46ab 28.60b 29.65ab 34.70a 1.75

SGR7, % 3.53 b 3.54 b 3.64b 3.94a 0.05PER8 1.41b 1.46b 1.56ab 1.71a 0.08FCR9 1.38a 1.35a 1.28ab 1.20b 0.05SR10, % 94.44b 95.56ab 97.78ab 98.89a 1.12

a–bMeans within a row with different letters differ significantly(Pb0.05) when tested with Duncan's new multiple-range testfollowing analysis of variance.lData represent mean values of three replicate tanks of 30 fish each.2MMT = montmorillonite; 3Cu-MMT = Cu2+-exchanged montmorillon-ite; 4Standard error of themean; 5BWi = Initial bodyweight;

6BWf = FinalBody weight; 7SGR = Specific growth rate; 8PER = Protein efficiencyratio; 9FCR = feed conversion ratio; 10SR = percent survival.

Table 3Effect of Cu-MMTon total aerobic bacterial counts of the skin, gill andintestinel

Skin log10CFUper cm2

Gill log10CFUper gram

Intestine log10CFUper gram

Control 3.94 5.24 6.47a

CuSO4 3.98 5.17 6.24a

MMT2 3.78 4.78 5.94ab

Cu-MMT3 3.80 4.65 5.25b

SEM4 0.21 0.25 0.30a–bMeans within a line with different letters differ significantly(Pb0.05) when tested with Duncan's new multiple-range testfollowing analysis of variance.lData represent mean values of three replicate tanks of three fish perreplicate.2MMT = montmorillonite; 3Cu-MMT = Cu2+-exchanged montmoril-lonite; 4Standard error of the mean.

202 C.H. Hu et al. / Aquaculture 270 (2007) 200–206

28 °C. For water quality control, temperature anddissolved oxygen were measured daily, and weeklyanalyses were done of total ammonium, nitrite, nitrateand pH levels. The following values (±S.D.), appropriatefor tilapia cultivation, were used: temperature, 28.7±0.5 °C; dissolved oxygen, 6.3±1.5 mg/l; pH, 8.02±0.18;ammonia, 0.05±0.01 mg/l; nitrite, 0.04±0.02 mg/l;nitrate, 6.15±1.12 mg/l. The fish were fed at a rate of4% (wet weight basis) of their total biomass per day in 2portion. The daily ration was divided into two equalfeedings and fed at 9:00 and 15:00. Fish were weighedbiweekly and the daily ration was adjusted accordingly.

Growth performance indicators measured were spe-cific growth rate (SGR), feed conversion ratio (FCR),protein efficiency ratio (PER) and percent survival. Theseindicators were calculated as: SGR=100(ln(Final Bodyweight)− ln(Initial body weight)) / time (days); FCR=dryfeed intake (g) /wet weight gain (g); PER=wet weightgain (g) /protein intake (g); percent survival (%)=Finalfish number / initial fish number×100.

2.3. Isolation and identification of tilapia bacteria

At the 56th day of the feeding trial, 9 fish per treatment(3 fish per tank) were anesthetized with tricainemethanesulfonate (MS-222, Sigma, St. Louis, MO) andthen killed by cervical separation. Total counts of aerobicbacteria were determined from tilapia skin by placing a4 cm×4 cm sterile template on the fish surface (in thecentral region) and swabbing the area enclosed by thetemplate. Under aseptic conditions, samples of gill andintestinewere also taken by sterile knife. The skin, gill andintestine samples were homogenized with sterilized0.85% NaCl saline solution at room temperature. Pouredplates were prepared and that a 1.0 ml inoculum was

placed on the bottom of the Petri dish and then 20 ml ofmolten agar was poured onto and mixed with theinoculum. Total aerobic counts were determined onnutrient agar after incubation at 28 °C for 48–72 h. Thirtycolonies were randomly picked (to pick many differentphenotypes) from every sample and restreaked on nutrientagar three times to obtain pure cultures.

Isolated micro-organisms were identified by using ascheme based on Bergey's Manual (Holt et al., 1993).The following tests were conducted: Gram reaction bythe Hucker method; cell morphology by phase-contrastmicroscopy (Gerhardt, 1981); flagella type by theNishizawa and Sugawara method (Aoyama, 1976);catalase formation by dropping a 3% H2O2 solutiondirectly onto each colony; oxidase test by the Kovacmethod; and O/F-test by the Hugh and Leifson method(Harrigan, 1998).

2.4. Intestinal mucosal morphology

Light microscopy: Segments were removed from theproximal, mid and distal intestine and flushed withphysiological saline and fixed in 10% formalin. Threecross-sections for each intestinal sample (total of 9samples for each of the three intestinal segments perdietary treatment) were then prepared after staining withhematoxylin and eosin using standard paraffin embed-ding procedures. A total of 10 intact villus were selectedin triplicate for each intestinal cross-section (30 mea-surements for each sample, total of 270 measurementsfor each of the three intestinal segments per dietarytreatment). Villus height were determined using imageprocessing and analysis system (Version 1, LeicaImaging Systems Ltd, Cambridge, England).

Table 4Effect of Cu-MMT on the composition of microflora from the skin l

Aer. Pse. Fla. Mor. Aci. Ent. Mic. Oth.

Control 36.1 5.2 30.9 6.2 9.4 5.3 6.1 0.8CuSO4 31.2 4.2 30.2 5.1 12.1 6.8 8.5 1.9MMT2 32.5 3.3 30.1 6.5 10.8 6.2 9.2 1.4Cu-MMT3 33.1 3.8 30.1 6.4 9.2 4.8 10.5 2.1SEM4 2.1 0.8 1.5 0.9 1.0 0.7 1.5 0.5

Effect of Cu-MMT on the composition of microflora from the gill l

Aer. Pse. Fla. Mor. Aci. Ent. Alc Bac. Cor. Oth.

Control 11.6 11.7 34.2 11.2 9.7 4.4 6.6 5.6 3.5 1.5CuSO4 11.1 11.2 31.4 10.9 9.9 5.1 5.9 8.1 5.2 1.2MMT2 10.9 10.8 32.5 10.6 9.4 5.2 6.5 8.9 4.5 0.7Cu-MMT3 11.3 10.1 34.1 9.7 9.4 4.2 6.2 8.5 5.5 1.0SEM4 0.8 0.8 1.4 0.7 0.6 0.4 0.4 1.2 0.7 0.3

Effect of Cu-MMT on the composition of intestinal microflora l

Aer. Pse. Fla. Vib. Aci. Ent. Alc. Bac. Cor. Mic Oth.

Control 39.2a 10.4a 9.1a 5.3a 4.1a 18.6a 6.6a 1.5b 1.2c 2.5b 1.5CuSO4 37.3ab 9.9a 8.9a 4.9ab 3.9a 17.9ab 6.2a 3.4b 1.8c 3.3b 2.5MMT2 36.8ab 8.5ab 5.1b 3.8b 2.6b 15.7ab 3.6b 8.8a 7.2b 5.3a 2.6Cu-MMT3 33.5b 7.2 b 5.6b 4.1b 1.6b 15.0b 3.3b 10.4a 10.8a 5.7a 2.8SEM4 1.7 0.8 0.7 0.4 0.4 1.1 0.6 0.8 1.0 0.5 0.5a–cMeans within a line with different letters differ significantly (Pb0.05) when tested with Duncan's new multiple-range test following analysis ofvariance.lData represent mean values of three replicate tanks of three fish per replicate. The units are %. Aeromonas, Aer; Pseudomonas, Pse.; Flavobac-terium, Fla.; Acinetobacter, Aci.; Alcaligence, Alc.; Enterobacteriaceae, Ent.; Vibrio, Vib.; Bacillus, Bac.; Corynebacterium, Cor.; Moraxella,Mor.; Micrococcus, Mic.; Other, Oth.2MMT = montmorillonite; 3Cu-MMT = Cu2+-exchanged montmorillonite; 4Standard error of the mean.

203C.H. Hu et al. / Aquaculture 270 (2007) 200–206

Transmission electron microscopy (TEM): Speci-mens of the proximal, mid and distal intestine werecollected from the same sites as those used for tissuesexamined with Light microscopy. Samples were fixed in2.5% glutaraldehyde in 0.1M phosphate buffer (pH 7.0).The fixed samples were washed in the buffer, dehydratedwith graded ethanol and embedded in Spurr's resin.Three ultrathin sections for each intestinal sample (total

Table 5Effect of Cu-MMT on villus and microvillus height at different sites of the i

Site Control CuSO4

Villus height (ìm)Proximal intestine 350.29b 372.55b

Mid-intestine 637.04c 664.57bc

Distal intestine 361.37c 379.48bc

Microvillus height (ìm)Proximal intestine 7.24b 7.15b

Mid-intestine 7.89c 8.01c

Distal intestine 3.56bc 3.40c

a–cMeans within a line with different letters differ significantly (Pb0.05) whvariance.lData represent mean values of three replicate tanks of three fish per replicat2MMT = montmorillonite; 3Cu-MMT = Cu2+-exchanged montmorillonite; 4S

of 9 samples for each of the 3 intestinal segments perdietary treatment) were cut and stained with 2% uranylacetate, post-stained with 0.2% lead citrate and exam-ined in a Jeol JEM-1200 TEM at 60 kV. Mid-villusregions were identified under low magnification. Aftermid-villus regions were located, micrographs of themicrovillus border were taken at 20,000× magnification.A total of 10 well-oriented longitudinal microvilli were

ntestine of Nile tilapia1

MMT2 Cu-MMT3 SEM4

398.11ab 435.15a 17.40718.42ab 758.68a 27.02411.13ab 430.81a 16.04

8.05ab 8.53a 0.369.06ab 9.84a 0.304.28a 4.49a 0.22

en tested with Duncan's new multiple-range test following analysis of

e.tandard error of the mean.

204 C.H. Hu et al. / Aquaculture 270 (2007) 200–206

assessed for each intestinal section (30 measurements foreach sample, total of 270 measurements for each of the 3intestinal segments per dietary treatment).

2.5. Statistical analysis

Data were analyzed as a randomized complete blockusing the general linear model procedure (SAS InstituteInc., 1989., Cary, NC). A tank of fish served as theexperimental unit for all data. Differences among meanswere tested using Duncan's multiple-range tests. Effectswere considered significant at Pb0.05.

3. Results

3.1. Growth performance

The growth performance of Nile tilapia is presented inTable 2. As compared to control, supplementation withCu-MMT improved (Pb0.05) specific growth rate, pro-tein efficiency ratio, survival and decreased (Pb0.05)feed conversion ratio. While fish fed with MMT hadslightly greater growth performance than the control,the tendency was not significant. Supplementation withCuSO4 had no (PN0.05) effect on growth performance.

3.2. Microbial ecology of the skin, gill and intestine

3.2.1. Total aerobic bacterial counts of the skin, gilland intestine

The total aerobic bacterial counts of the skin, gill andintestine of Nile tilapia is presented in Table 3.Supplementation with Cu-MMT reduced (Pb0.05) thetotal intestinal aerobic bacterial counts as compared withthe control. While supplementation with MMTor CuSO4

had no influence (PN0.05) on intestinal bacterial counts.The total aerobic bacterial counts of the skin and gill wasnot affected by the addition ofCuSO4,MMTorCu-MMT.

3.2.2. The composition of microfloraThe composition of microflora of the skin, gill and

intestine is presented in Table 4. The results indicatedthat the dominant organisms were found to be Aero-monas and Flavobacterium in skin, Flavobacterium ingill, and Aeromonas and Enterobacteriaceae in intes-tine. Aeromonas constituted the largest group in bothskin and intestine, and was followed by Flavobacteriumand Enterobacteriaceae in skin and intestine, respec-tively. Meanwhile Flavobacterium constituted thelargest group in gill. The composition of microflora ofthe skin and gill was not affected by the addition ofCuSO4, MMT or Cu-MMT. Supplementation with

MMTor Cu-MMTaffected the composition of intestinalmicroflora with Aeromonas, Vibrio, Pseudomonas,Flavobacterium, Acinetobacter, Alcaligence, Entero-bacteriaceae decreasing and Bacillus, Corynebacter-ium, Micrococcus increasing as compared with control.

3.3. Intestinal mucosal morphology

Morphological measurements of the intestinal mu-cosa of Nile tilapia are presented in Table 5. Supple-mentation with Cu-MMT had higher (Pb0.05) villusand microvillus height at the proximal, mid and distalintestine mucosa as compared with the control orCuSO4. Supplementation with MMT also increased(Pb0.05) villus and microvillus height at the mid anddistal intestine mucosa as compared with the control.Supplementation with CuSO4 had no (PN0.05) effect onthe intestinal morphology as compared with control.

4. Discussion

Traditionally, MMT has been incorporated in animaldiets as a technological additive (lubricant or agglom-erant) to improve feed manufacture. However, recently arole as enhancer of the nutritive value of diets in animalshas been also proposed. Animal feed containing MMT(10–30 g/kg) has been shown to promote growth per-formance in chickens and swine (Tauqir and Nawaz,2001; Venglovsky et al., 1999). In the present study,while tilapia fed with MMT had slightly greater growthperformance than the control, the tendency was notsignificant. The reason of the lack of effects of MMTsupplementation on tilapia growth performance inthis study may be because the concentration of MMT(1.5 g/kg) was not adequate.

It is well known that MMT can adhere to bacteria.Adsorption was the main interaction between MMTand bacteria. Copper sulfate is one of the traditionalinorganic antibacterial materials with wide usage.However, Cu2+ is difficult to contact with bacteriumin water by colliding so that the usage amount must belarge for the purpose of antimicrobial. In order toincrease the antibacterial activity of Cu, we consider tocarry the Cu2+ in/on MMT with large specific surfacearea, which can adsorb bacterium rapidly after dispersedin water. Copper tended to enter interlayer position ofMMT dominantly as [Cu(AlO)n(H2O)4−n]

x+ or locate inthe ditrigonal intra-crystal hole surrounded by Si–Otetrahedron, or take a position in Al–O octahedron inMMT (Mosser et al., 1997). This made the mineral havesurplus positive charge. On the other hand, the cellularwall of the bacteria was negatively charged, so that Cu-

205C.H. Hu et al. / Aquaculture 270 (2007) 200–206

MMT particles would attract bacteria, due to theopposite static charge. The Zeta potentials of MMT,Cu-MMT, A. hydrophila, V. parahaemolyticus and P.fluorescens at pH 6.0 are −21 mV, 11.8 mV, −18 mV,−16 mV and −22 mV, respectively. The results of invitro bacterial–mineral adsorptive trial showed that thepercent of bacteria adsorbed onto MMT and Cu-MMTwas 37.4–46.5% and 94.4–99.4%, respectively. Theantimicrobial trial in vitro indicated that MMT had noantibacterial activity while Cu-MMT has strong anti-bacterial ability on the tested aquacultural pathogenicbacteria (Hu and Xia, 2005). Surplus positive charge ofCu-MMT was most probably an important factor fortheir antibacterial capability (Hu et al., 2005a; Hu andXia, 2006). In this case, the released Cu2+ would actdirectly on the bacteria adsorbed on the surface of Cu-MMT, instead of into the medium and indirectly on thebacteria. In other words, the active Cu2+ density onmineral surface was much higher than its concentrationin the solution. The high concentration of Cu2+ will actdirectly on the attracted bacteria. Summary, two possiblemechanisms for the antibacterial activity of Cu-MMTwere proposed. One involved the adsorption of thebacteria from solution and immobilization on thesurface of the Cu-MMT. Alternatively, Cu2+ coulddisassociate from the clay surface and directly exerts itsantibacterial effect on the bacteria (Hu et al., 2005a,2006).

The present study showed that supplementation withCu-MMT not only exhibited antibacterial activity invivo but also affected the composition of intestinalmicroflora with Aeromonas, Vibrio, Pseudomonas,Flavobacterium, Acinetobacter, Alcaligence, Entero-bacteriaceae decreasing while Bacillus, Corynebacter-ium, Micrococcus increasing as compared with control.Wang and Fang (1995) also reported that the intestinalbacteria counts of Bifidobacteria were significantlyincreased while those of E. coli were decreased afterdiarrhoeal children had been treated with MMT for5 days. The changes of the composition of intestinalmicroflora caused by Cu-MMT may be related to thecharacteristics of MMT. It was reported that MMTadsorbed bacteria selectively and only adhere to thepathogenic E. coli strains producing surface antigendesignated CS31A (Girardeau, 1987). Hu et al. (2006)used Caco-2 cell line to investigate the adhesion ofprobiotic or pathogenic strains to intestina epithelialcells and the effects of MMT on the bacterial adhesion.The results showed that the ability of MMT to inhibitadhesion of pathogenic strains to Caco-2 cells washigher than that of probiotic strains. The percentages ofMMT against adhesion of E. coli, Salmonella typhimur-

ium, A. hydrophila, V. parahaemolyticus to Caco-2 cellwere 54.22%, 48.41%, 60.53% and 50.64%, respec-tively, while against probiotic strains 21.49–25.64%.

The structure of the intestinal mucosa can reveal someinformation on gut health. Stressors that are present in thedigesta can lead to relatively quickly changes in theintestinal mucosa due to the close proximity of themucosal surface and the intestinal content. The addition ofMMT and Cu-MMT to the diet in the present study bothproduced a positive effect on the intestinal mucosa. Huet al. (2005b) studied the effects of Cu-MMT on A.hydrophila adhesion to epithelial cells of Nile tilapia. Theresults showed that Cu-MMT showed a strong ability toinhibition adhesion of A. hydrophila to fish skin, gill andintestinal epithelial cells and significantly reduced cellmembrane injury caused by the adhesion of A. hydro-phila. It is well known that MMT is a protector ofintestinal mucosa and can reinforce intestinal mucosalbarrier and help in the regeneration of the epithelium.MMT effectively improves gastrointestinal mucus resis-tance to various aggressions by interacting closely withthe mucous glycoproteins (Droy-Lefain et al., 1985). Theimproved intestinal mucosal morphology caused by Cu-MMT may be explained by the antibacterial activity ofCu-MMT and the MMT enhancing mucosa function.

Since copper is an essential element for all organismsincluding fish, it is often supplemented to commercialfeeds in concentrations which may reach the upper limitof toxicity. Supplementation of Cu to fish feed istherefore a balance between fulfilling the Cu require-ment and avoiding Cu toxicity. The requirement and themaximal tolerable levels of dietary copper vary widelyfor different studies on fish. Murai et al. (1981) foundreduced growth and feed conversion in channel catfish(Ictalurus punctatus) fingerlings exposed to 16 mg Cukg−1 diet for 16 weeks. In contrast, Galtin and Wilson(1986) found no growth inhibition in channel catfishfingerlings exposed to 40 mg Cu kg−1 diet for 13 weeks.For rainbow trout (Oncorhynchus mykiss), growthsuppression was not apparent at dietary Cu concentra-tions up to 350–684 mg Cu kg−1 diet (Miller et al.,1993). In Atlantic salmon (Salmo salar L.) parr, basedon the efficiency of Cu uptake from the feed, therecommended upper limit for Cu supplementation couldbe set at 35 mg Cu kg−1. However, based on the reducedgrowth it seemed possible to permit up to 500 mg Cukg−1 (Berntssen et al., 1999). Shiau and Ning (2003)found that weight gain of juvenile tilapia was highest infish given diets supplemented with 2 mg Cu per kg diet,followed by the group given 1 mg Cu per kg diet, thenthe unsupplemented control group, and was lowest inthe 20 mg Cu per kg diet group. They estimated the

206 C.H. Hu et al. / Aquaculture 270 (2007) 200–206

adequate dietary Cu concentration in growing tilapiawas about 4 mg Cu per kg diet. In the present studysupplementation with 30 mg Cu kg−1 diet had nosignificant effect on the growth performance of Niletilapia. Differences in the duration of the experimentalperiod, husbandry, and life stage of the fish make itdifficult to compare these studies. Hu et al. (2005a)investigated the Cu release from Cu-MMT in Trypticsoy broth and found the saturated Cu concentration inthe supernatant reached only 1.22–2.27% of the overallCu in Cu-MMT, suggesting the Cu nutritive value ofCu-MMT was little.

Acknowledgements

The authors gratefully acknowledge H. X. Sun, H. M.Zhang, J. M. Pan, and X. Y. Zhu for their skilled technicalassistance.

References

Aoyama, Y., 1976. (Ed.), Ikagaku Kenkyujo Gakuyukai, In:Saikingaku Jisshu Teiyo. Maruzen, Tokyo, pp. 73–158.

Berntssen, M.H.G., Lundebye, A.K., Maage, A., 1999. Effects ofelevated dietary copper concentrations on growth, feed utilisationand nutritional status of salmon (Salmo salar L.) fry. Aquaculture174, 167–181.

Droy-Lefain, M.T., Drouet, Y., Schatz, B., 1985. Sodium glycodeox-ycholate and spinability of gastrointestinal mucus: protective effectof smectite. Gastroenterology 88 (Suppl. 2), 1369.

Galtin, D.M., Wilson, R.P., 1986. Dietary copper requirements offingerling channel catfish. Aquaculture 54, 277–285.

Gerhardt, P., 1981. Manual of Methods for General Bacteriology.American Society for Bacteriology. In: Murray, R.G.E., Costilow,R.N., Nester, E.W., Wood, W.A., Krieg, N.R., Philips, G.B. (Eds.),Morphology: Light Microscopy, pp. 1–525. Washington, DC.

Girardeau, J.P., 1987. Smectite aggregation by Escherichia coli. ActaGastro-Enterologica Belgica 50, 181–192.

Harrigan, W.F., 1998. Laboratory Methods in Food Microbiology.Basic Methods: Biochemical Test for Identifying Bacterium, pp.92101–94495. San Diego, California.

Holt, J.G., Krieg, N.R., Sneath, P.H.A., Staley, J.T., Wiilliams, S.T.,1993. Bergey's Manual of Determinative Bacteriology, 9th edition.Williams & Wilkins, Baltimore.

Hu, C.H., Xia, M.S., 2005. Antibacterial effect of copper-bearingmontmorillonite on aquacultural pathogenic bacteria and discus-sion on its mechanism. Journal of the Chinese Ceramic Society 33,1376–1380 (in Chinese with English abstract).

Hu, C.H., Xia, M.S., 2006. Adsorption and antibacterial effect ofcopper-exchanged montmorillonite on Escherichia coli K88.Applied Clay Science 3, 180–184.

Hu, C.H., Xu, Z.R., Xia, M.S., 2005a. Antibacterial effect of Cu2+-exchanged montmorillonite on Aeromonas hydrophila and discus-sion on its mechanism. Veterinary Microbiology 109, 83–88.

Hu, C.H, Xia, M.S., Xiong, L., Xu, Z.R., 2005b. Effects of Cu bearingmontmorillonite on Aeromonas hydrophila adhesion to epithelialcells of Nile tilapia. Journal of Fisheries of China 29, 619–623 (inChinese with English abstract).

Hu, C.H, Xia, M.S., Xiong, L., Xu, Z.R., 2006. Effects ofmontmorillonite on bacterial adhesion to Caco-2 cells. ChineseJournal of Cell Biology 28 (1), 76–78 (in Chinese with Englishabstract).

Miller, P.A., Lanno, R.P., McMaster, M.E., Dixon, D.G., 1993.Relative contributions of dietary and waterborne copper to tissuecopper burdens and waterborne-copper tolerance in rainbowtrout (Oncorhynchus mykiss). Canadian Journal of Fisheries andAquatic Sciences 50, 1683–1689.

Mosser, C., Michot, L.J., Villierns, F., Romeo, M., 1997. Migration ofcations in Cu (II)- exchanged montmorillonite and laponite uponheating. Clays and Clay Minerals 45, 789–802.

Murai, T., Andrews, J.W., Smith Jr., R.G., 1981. Effects of dietarycopper on channel catfish. Aquaculture 22, 353–357.

National Research Council (NRC), 1993. Nutrient Requirements ofFish. National Academy Press, Washington, DC.

SAS Institute Inc., 1989. SAS/STAT User'sGuide, Version 6. SASInstitute Inc., Cary, North Carolina.

Shiau, S.Y., Ning, Y.C., 2003. Estimation of dietary copper requirementsof juvenile tilapia, Oreochromis niloticus×O. aureus. AnimalScience 77 (2), 287–292.

Tauqir, N.A., Nawaz, H., 2001. Performance and economics of broilerchicks fed on rations supplemented with different levels of sodiumbentonite. International Journal of Agriculture and Biology 3,149–150.

Venglovsky, J., Pacajova, Z., Sasakova, N., Vucemilo, M., 1999.Adsorption properties of natural zeolite and bentonite in pig slurryfrom the microbiological point of view. Veterinarni Medicina 44,339–344.

Wang, J.X., Fang, H.S., 1995. Intestinal microecological observationof smectite in the treatment of diarrhea in children. Chinese Journalof Clinical Pharmacology 11, 134–137 (in Chinese with Englishabstract).