7 pages

Effect of initial glucose concentration and inoculation level of lactic acid bacteria in shrimp waste ensilation

of 7
All materials on our website are shared by users. If you have any questions about copyright issues, please report us to resolve them. We are always happy to assist you.
Effect of initial glucose concentration and inoculation level of lactic acid bacteria in shrimp waste ensilation
  Effect of initial glucose concentration and inoculation level of lacticacid bacteria in shrimp waste ensilation Keiko Shirai a, *, Isabel Guerrero a , Sergio Huerta a , Gerardo Saucedo a , Alberto Castillo b ,R. Obdulia Gonzalez b , George M. Hall c a  Departamento de Biotecnologı´a, Universidad Autonoma Metropolitana, D.F. Av. Michoacan y Purisima s/n Col., 09340 Vicentina, Mexico b  Departamento de Matematicas, Universidad Autonoma Metropolitana, D.F. Av. Michoacan y Purisima s/n Col., 09340 Vicentina, Mexico c  Loughborough University, Department of Chemical Engineering, UK  Received 21 June 2000; received in revised form 31 October 2000; accepted 10 November 2000 Abstract Fermentation conditions and microorganisms were determined, based on acid production, glucose concentration as carbohydrate source.Inoculation levels to obtain a stable shrimp waste silage were also determined. Shrimp waste ensilation was an efficient method of preservation, allowing the recovery of chitin and another added-value products such as pigments, proteins and enzymes. From the variouslactic acid bacteria tested,  Lactobacillus pentosus  and  Lactobacillus sp.  (B2) were the best lactic acid producers, although small quantitiesof acetic acid were detected in samples inoculated with  Lactobacillus pentosus.  Therefore B2 was chosen for the analysis of glucoseconsumption as well as for the determination of optimum inoculation levels. The best results were obtained at 10% (w/w wet basis) and 5%(v/w wet basis) respectively. Presence of starters and initial glucose concentration were critical factors in the fermentation of shrimp waste.High initial glucose and starter concentrations reduced the time and increased the amount of lactic acid produced. The fermentation patternchanged during ensilation from hetero to homofermentative. Shrimp waste ensilation prevented the growth of spoilage microorganismskeeping their microbial counts steady and pH values within the acid region. © 2001 Elsevier Science Inc. All rights reserved. Keywords:  Shrimp wastes; Response surfaces; Lactic acid fermentation; Chitin; Protein 1. Introduction Mexico is an important shrimp producer, both by catch-ing and farming. The combined production of these meth-ods generates 35,000 to 40,000 tons per year of wastes. Thismaterial is partially used as animal feed but most of it isdiscarded causing serious ecological problems. Shrimpwastes are rich in chitin, proteins, pigments and lipids,which can be recovered. However the main limitation inusing this otherwise polluting material is its high perishabil-ity [1,2].Ensilation is defined as a conservation process in whichacids added or produced inhibition pathogen growth. Acidis produced in situ during fermented silage, in which lacticacid bacteria utilize a carbohydrate source such as malt,cassava or molasses. These microorganisms can be presentin the native microflora or added as starters. Ensilation of shrimp waste allows the recovery of added-value productssuch as chitin, pigments and proteins. [3,4].Ensilation produces a change in the shrimp waste fromsemi-solid to liquid in 2 or 3 days, pH is reduced andproteases are activated. The separated liquor represents 60to 70% (wet weight) of the silage and is composed of proteins and lipids. Over 95% of the chitin present in thesubstrate remain in the sediment. The pigments are partiallysoluble in the lipids although a certain amount is attached tothe chitin fraction [5].Chitin and chitosan have numerous applications in foods,chemistry, cosmetics, paints and textiles [6,7]. The tradi- tional methods of chitin production involve the use of strongalkalis and acids, making this process ecologically aggres-sive and a source of pollution. It also promoted a certaindegree of depolymerization, reducing chitin quality [8,9].The acid/alkali process renders the protein component use-less, which otherwise can be used as fish feed. On the otherhand, with the ensilation, the recovery of chitin partially * Corresponding author. Tel.:  5-8044717/8044726; fax:  5-8044712.  E-mail address: (K. Shirai) and Microbial Technology 28 (2001) 446–4520141-0229/01/$ – see front matter © 2001 Elsevier Science Inc. All rights reserved.PII: S0141-0229(00)00338-0  purified during the fermentation besides protein for feedpurposes is feasible [10].Improvement of the lactic acid fermentation process hasbeen generally accomplished by testing one variable at atime. For example, in cassava fermentation for Gari produc-tion, a fermented food common in Africa, the addition of acontrolled starter at different proportions was studied todetermine acid production [12].Another example was lactic acid fermentation of shrimpheads in which aeration conditions were modified whilemaintaining constant sugar and inoculation levels. In thiscase the selection criterion used was the conditions at whichthe lowest pH was achieved [13]. However, this method hasthe disadvantage of requiring a large number of time-con-suming experiments and difficulty in determining interac-tions among variables. As a result mathematical modelscorrectly describing correlations between sources of varia-tion and response variables are difficult to obtain.The response surface methodology (RSM) has been de-fined as a statistical method that uses quantitative data todetermine and simultaneously solve multivariable equa-tions, graphically represented as response surfaces [14].The present study consists in describing the relationshipbetween sources of variation such as the type of inoculatedmicroorganism, glucose concentration and proportion of added inoculum and acidification of shrimp wastes sub- jected to lactic acid fermentation.In a first approach several lactic acid bacteria weretested, glucose and inoculation level was constant in orderto select the fastest acid producer bacterium. The secondpart developed a RSM describing acidification through thestudy time as a result of inoculation level and initial glucoseconcentration. The interrelationships among variables aswell as the combined effects on the response were deter-mined. Finally the fermentation pattern and microflora evo-lution were analyzed. 2. Methodology 2.1. Shrimp waste Shrimp ( Penaeus spp. ) waste consisting of heads (tho-rax) was obtained from Mexico City central seafood market.The waste was minced through a 6-mm sieve using a meatmincer (Sanitary, Chicago) and stored at  20°C until used.Our experience shows that freezing does not affect theactivity of native enzymes and any change in microbialpopulation is unimportant given the large inoculum of lacticacid bacteria added. 2.2. Microorganisms Lactic acid bacteria tested were two strains isolated fromtropical shrimp, identified as  Lactobacillus casei  strain A3and  Lactobacillus sp.  strain B2 [5]. Two commercial start-ers were also tested: Floracarn SL (  Lactobacillus pentosus and  Staphylococcus carnosus  1:2) and Lp-1 (  Lactobacillus pentosus).  Both starters were obtained from Christian Han-sen (Denmark). All starters were grown in APT broth(Difco, USA) at 30°C until a concentration of 10 8 cfu/mlwas obtained. 2.3. Starter selection The shrimp waste was mixed with 10% (w/w) wet basisglucose monohydrated and inoculated with 5% (v/w) wetbasis of the starters. The fermentation was carried out at30°C, and sampled at 0, 24 and 48 h.Possible differences among treatments were detected ap-plying an analysis of variance and Duncan’s [15] test usinga SPSS version 8.0 program (SPSS, Inc., USA 1997). Re-sponse variables were organic acid concentration (lacticacid and acetic acid), pH, total titratable acidity (%TTA)and water activity (a w. ). 2.4. Sample analysis pH of the liquor obtained during the fermentation wasmeasured in 10 ml of sample diluted 1:10 with distilledwater using a potentiometer (Conductronic pH 20, Mexico).Total titratable acidity (%TTA) determined in the dilutedsamples by titration with 0.1M NaOH to pH 8.4. Determi-nations of %TTA were done by quadruplicate, mean valueswere used for further statistical analysis since variationsamong determinations were smaller than the error term inthe analysis of variance. %TTA was expressed as a percent-age of lactic acid [16]. Glucose content was determined inthe aqueous extract as reducing sugars [17]. Cooled mirror(dewpoint) technique was used for measuring water activity(Decagon CX-1, Pullman, Washington).Concentrations of lactic and acetic acids, and glucosewere determined by HPLC. Ten grams of sample werediluted with water (1:10) and centrifuged at 10000 g for 15min. The supernatant was filtered through a 0.45   m mem-brane. One hundred fifty microliters of the filtrate wereinjected into a chromatograph (Waters 600S Massachusetts)using a column for sugars and organic acids (Sugar Shodex,4606027 SH1011, Tokyo, Japan) and eluted with 30 mM of sulphuric acid. Operation conditions were 1.0 ml/min flowrate at 24 kg/cm 2 and 50°C. The chromatograph was fittedwith a diode arrangement detector (Waters 994 Massachu-setts). Standards of sugars and organic acids were includedin the analysis (Sigma, St. Louis Missouri) [5,18] 2.5. Response surface methodology RSM was carried out on the acid production data using alinear regression approach. Three factors were defined: i)the amount of glucose added, ii) the amount of inoculationsource and iii) the fermentation time.The shrimp waste was mixed with the following glucose 447 K. Shirai et al. / Enzyme and Microbial Technology 28 (2001) 446–452  concentrations (  factor levels ): 5, 7.5, 10 and 15% (w/w wetbasis) and inoculated with  Lactobacillus spp.  (B2) at fourlevels: 0, 5, 10 and 20% (v/w wet basis). Total titratableacidity (%TTA) was measured after 0, 4, 8, 24 and 48 h of fermentation [19].Initial glucose concentration, time and inoculation levelwere the independent variables for linear regression analy-sis. The dependent variable was acid production (%TTA).The level of significance of each independent variable andsquared multiple correlation coefficient (R 2 ) were obtained.Beta coefficients were also calculated by the program [20]. 2.6. Microbiological examination The relationships between lactic acid bacteria and coli-form growth with respect to pH were studied in shrimpwaste silage added with 10% glucose and 5%  Lactobacillusspp.  cell suspension as starter. Samples were taken every24 h during 4 days, 10 g of the sample was homogenized in90 ml 0.9% saline solution (w/w) during 2 min. Decimaldilutions were prepared. The growth of microflora on fer-mented shrimp waste was determined by colony enumera-tion using de Man Rogosa Sharpe media [21] and brilliantgreen bile agar media for lactic acid bacteria and coliforms,respectively [22]. Sample pH was also determined as pre-viously described. 3. Results and discussion The preservation of fish by lactic acid fermentation de-pends upon rapid growth of competitive bacteria, acid pro-duction and inhibition of undesirable microflora by lower-ing pH values and, in some cases, other antimicrobialfactors [23].The present work studied some factors that can affect theefficiency of lactic acid fermentation on shrimp wastes as ameans of stabilization. These were fermentable carbohy-drate and growth factors availability, concentration of or-ganic acids, pH, initial lactic acid bacteria population andcompetitive microorganisms. 3.1. Screening of lactic acid bacteria Traditional fermentations depend on naturally occurringmicroorganisms in the substrate or materials involved in theprocess such as tools, equipment or human handling [24].Inoculation of suitable lactic acid bacteria ensures rapidacidification and eventual predominance of desired micro-flora able to conduct ensilation.Lactic acid bacteria were tested under fixed conditions:10% glucose (w/w) and 5% inoculum (v/w). Table 1 showspH, %TTA, a w , glucose and fermentation products obtainedin samples inoculated with the tested microorganisms ascompared to control (no inoculum). In the analysis of vari-ance procedure and Duncan’s multiple range test pH, a w ,and lactic acid were considered as response variables. ForpH a non significant model was obtained meaning that thisvariable did not show major changes due to buffering effectof proteins and minerals contained in the substrate. Thesechemicals are present in concentrations of 51 and 24% (w/wdry basis) respectively [5,11,25].Lactic acid was measured by two methods: titration andHPLC. Titration is commonly used due to its simplicity, Table 2Linear regression model for prediction of lactic acid production (percenttotal titratable acidity expressed as lactic acid) with added several initialglucose concentration (% w/w, wet basis) and inoculation levels (%v/w,wet basis) at fermentation time (hours)Model EstimatedcoefficientsSignificancelevelConstant 2.50  10  1 0.005Glucose 2 time 2.41  10  4 0.000Inoculum 2 time 1.37  10  4 0.000Inoculum 2 time 2  3.11  10  6 0.000Glucose   1.84  10  2 0.057Multiple correlation coefficient 0.906. Determination coefficient 0.821.Table 1pH, %TTA, a w , glucose and fermentation products (HPLC) determined in fermented shrimp waste at 48 hours, 30°C, with added 10% of glucose (w/wwet basis) and using several starters at 5% v/w, wet basis (10 8 ufc/ml)Microorganism pH %TTA a w  mg Glucoseg  1 fermentedshrimp wastemg Lacticg  1 fermentedshrimp wastemg Aceticg  1 fermentedshrimp waste  Lactobacillus  sp. B2 4.6 1.90 A 0.972 A,B 0.00 83  0.7 A 0.00  Lactobacillus casei  A3 5.4 1.00 B 0.969 B 12  0.2 63  0.8 B 0.00  Lactobacillus pentosus  5.5 1.16 A 0.970 A 0.00 80  0.5 A 6  0.04Floracarn SL (  Lactobacillus pentosus  and  Staphylococcus xylosus  1:2)5.8 0.88 B 0.965 D 10  0.1 68  0.09 B 4  0.03Control 5.8 0.76 B 0.978 C 1  0.03 36  0.2 C 12.3  0.3The results showed are average of 4 observations.A,B,C, Duncan Groups, which means that groups with the same letters are not significantly different.448  K. Shirai et al. / Enzyme and Microbial Technology 28 (2001) 446–452  although it is inaccurate. In order to know the concentrationof a specific acid, as well as its concentration, HPLC wasused.  Lactobacillus sp.  (B2) and  Lactobacillus pentosus, produced the largest lactic acid concentration although nosignificant difference was observed between substrates in-oculated with these bacteria. This was already observed inlactic acid determination by titration. Because  Lactobacillussp.  (B2) had a homofermentative pattern while  Lactobacil-lus pentosus  fermentation produced small amounts of aceticacid,  Lactobacillus sp.  (B2) was chosen for further work. Asthe homofermentative pattern of the  Lactobacillus sp.  strain(B2) was established, %TTA was used for further determi-nation of lactic acid produced.A decrease in a w  during fermentation was observed insamples inoculated with the studied bacteria having an ini-tial a w  0.989. This fact can be due to breakdown of largemolecules, such as proteins, to small compounds directlyaffecting water activity [5,26]. As water activity influencedmicrobial growth, the fermentation with the lowest a w  pre-sented signs of putrefaction such as off-odor and discolor-ation. This was the case of substrates fermented with  Flo-racarn SL  as well as the  control . Fig. 1. z (Total titratable acidity % w/w wet basis) is dependent on both the x (initial glucose concentration % w/w wet basis) and y (inoculation level %v/w wet basis). a) 0 h; b) 4 h; c) 8 h; d) 24 h and e) 48 h.  Experimental values,  E  Predicted values.449 K. Shirai et al. / Enzyme and Microbial Technology 28 (2001) 446–452  3.2. Glucose concentration and inoculum level Several carbohydrate sources have been studied to carryout a lactic acid ensilation of shrimp [3,11]. Glucose waschosen in this study for being a readily fermentable sugar.The use of glucose made it possible to standardize theprocess and can be a means to compare more complexcarbohydrate sources; hence the amount of sugar necessaryto ferment shrimp waste was evaluated using glucose. It hasbeen reported that the quantity of sugar influences directlythe fermentation, for instance, the minimum amount thatensures a successful fish fermentation was 5% [28,29]. The data obtained from acid production (%TTA) wereanalyzed by ANOVA procedure. Significant differenceswere observed for all factors and the majority of two-wayinteractions, only glucose and inoculum interaction was notsignificant. According to Tukey’s test, the inoculation in-creased the acid production, however with higher than 5%of inoculation levels there were no significant differences.A statistical regression model was obtained in whichfermentation time; initial glucose concentration and inocu-lation levels were included in first and second terms. Table2 shows the adjusted model better explaining the experi-mental data, all terms were significant.Initial glucose concentration, inoculum level and fermen-tation time had a significant effect on acid production.Estimated coefficients of the independent variables wereused to fit a response surface at various fermentation peri-ods. The model predicts lactic acid production during agiven time period, inoculum level and initial glucose con-centration. Data are shown in 3-dimensional graphs (Figs.1a–e). RSM gave information to determine the combinationof factors yielding the largest acid production. According tothe results inoculum level increased lactic acid productionwhen glucose concentration and time increased. However atfermentation periods of 24 and 48 h the inoculation level didnot affect acid production (Figs. 1d and 1e). All glucose andinoculation levels affected silage performance, where pHdecreased with increasing glucose concentration in the sub-strate. However when glucose concentration is higher than10% an extended lag phase is observed due to a decrease ina w  in the system (e.g. 15% a w  0.958) promoted by largeamounts of a water-binding substance, such as a sugar. Thisis not desirable due to possible putrefaction, which wasmore dramatic when the inoculation level was low, and canbe due to the initial lactic acid bacteria population, compet-ing with spoilage organisms for metabolizable sugar.Five percent inoculum and 10% initial glucose concen-tration were chosen as fixed conditions for further studiesbecause acid production was not significantly differentwhen 5 or 10% inoculum levels were applied (Figs. 1a and1e). Lower concentration of starters and carbohydratesource might reduce the cost of the process. 3.3. Fermentation pattern The fermentation pattern was followed during 48 h at theinoculum and carbohydrate concentrations described above.Figs. 2 and 3 show acid production (%TTA), pH, lactic acid(by titration) and glucose concentration (by reducing sugarmethod) with respect to time. The inoculum had an impor-tant effect on the fermentation pattern (Table 3), changingfrom heterofermentative (Control: shrimp waste, no inocu-lum) to homofermentative conversion of glucose (with in-oculum). In this case analysis was by HPLC but the results Fig. 2. Lactic acid fermentation of shrimp wastes containing 10% w/w (wetbasis) glucose, and inoculated with 5% v/w (wet basis)  Lactobacillus sp (B2): ( ■ ) pH, (  ) total titratable acidity. Control without starter: ( F ) pH,( E ) total titratable acidity.Fig. 3. Lactic acid production and glucose consumption during the fermen-tation of shrimp wastes inoculated with 10% w/w (wet basis) glucose andinoculated with 5% v/w (wet basis)  Lactobacillus sp.  (B2): ( ■ ) glucose,(  ) lactic acid. Control without starter: ( F ) glucose, ( E ) lactic acid.450  K. Shirai et al. / Enzyme and Microbial Technology 28 (2001) 446–452
Related Documents
View more...
We Need Your Support
Thank you for visiting our website and your interest in our free products and services. We are nonprofit website to share and download documents. To the running of this website, we need your help to support us.

Thanks to everyone for your continued support.

No, Thanks