Bioresource Technology 97 (2006) 2414–2420

Acid-hydrolysis of fish wastes for lactic acid fermentation

Min-Tian Gao *, Makoto Hirata, Eiichi Toorisaka, Tadashi Hano

Department of Applied Chemistry, Oita University, Oita 870-1192, Japan

Received 6 August 2005; received in revised form 27 September 2005; accepted 1 October 2005Available online 15 November 2005

Abstract

In this study, two acid-hydrolysis processes, process A and process B, were proposed to produce low-cost nutrients for the productionof lactic acid. Process A was a direct way to hydrolyze protein with diluted acid while process B was process A plus fish wastes pretreat-ment (an extraction by water). The two methods could both treat fish wastes to be suitable nutrient sources for promoting lactic acidproduction. As the pretreatment indicated some favorable effect on fish waste hydrolyzate (FWH), process B increased lactic acid pro-ductivity by 22%. Compared with 20 g/L yeast extract (YE), 6.8% FWH hydrolyzed by process B had more efficiency in lactic acid pro-duction, indicating that process B was suitable to produce high performance nutrients for lactic acid production and FWH hydrolyzed byprocess B would be an substitute for YE.� 2005 Elsevier Ltd. All rights reserved.

Keywords: Lactic acid; Fermentation; Nutrient sources; Acid-hydrolysis; Fish waste hydrolyzate (FWH)

1. Introduction

In recent years, much attention has been paid to the greatpotential of L-lactic acid in the manufacture of biodegrad-able plastics and approximately 90% of L-lactic acid is pro-duced by lactic acid bacteria fermentation every year(Hofvendahl and Hagerdal, 2000; Wang et al., 2002). Lacticacid bacteria exist generally in nutritious habitats so thatthey have developed a typical metabolism that is devoidof most biosynthetic activities (Hugenholtz and Kleereb-ezem, 1999). Therefore, they need specific minerals, vita-mins, specific peptides and some unknown nutrients toensure their optimum growth. Commonly, yeast extract(YE) is used in laboratory scale fermentations as a nutrientsource. As YE is not economically attractive, it is desired tofind some new nutrients suitable for an industrial processand to replace YE. Generally, the proteins in nutrients arehydrolyzed into peptides and amino acids before used forlactic acid production. Enzymatic hydrolysis is an alterna-tive approach for protein recovery and finds wide use in

0960-8524/$ - see front matter � 2005 Elsevier Ltd. All rights reserved.

doi:10.1016/j.biortech.2005.10.002

* Corresponding author. Tel./fax: +81 97 554 7901.E-mail address: gmttiger@cc.oita-u.ac.jp (M.-T. Gao).

food industry because it brings about products of high func-tionality and nutritive value. It has, however, the disadvan-tages in slow reaction rate and high cost due to therequirements of feedstock pretreatment, enzyme produc-tion and enzyme recovery. From an economic viewpoint,the protein hydrolyzates produced by enzymatic hydrolysisare not applicable to the nutrient sources in the productionof low-priced lactic acid. On the contrary, acid-hydrolysishas the advantages of low-cost, short hydrolysis time andsimple operation so that it is applicable to industrial pro-cesses. Some nutrients, such as casamino acids (Yoo et al.,1997), soybean hydrolyzate (Hsieh et al., 1999; Kwonet al., 2000) and ram horn protein hydrolyzate (Kurbanogluand Kurbanoglu, 2003), have been used for lactic acid pro-duction after hydrolyzed with acids. The utilization of thosenutrients could reduce nutrient cost to some extent. Theywere, however, in low performance in lactic acid productionrelative to the nutrients hydrolyzed with enzymes. It couldbe considered to be the destruction of nutrients in severehydrolysis conditions since conventional acid-hydrolysisfor protein recovery was performed generally with concen-trated acids. Moreover, concentrated acids contributed tohigh concentration salts after neutralization, inhibiting

M.-T. Gao et al. / Bioresource Technology 97 (2006) 2414–2420 2415

fermentation. The acid-hydrolyzed nutrients had, therefore,to be used in combination with YE or peptone to increasethe efficiency of fermentation. To produce hydrolyzates ofhigh performance from nutritious wastes, two acid-hydroly-sis processes were proposed in this study. Fish wastes werechosen as a nutrient source for lactic acid production due tothe background that fishing industry creates large amountof wastes every year in Japan and elsewhere so there isincreasing demand for effective and ecological techniquesto treat the wastes. Some researches have been conductedin the production of fish protein hydrolysate (Shahidiet al., 1995; Guerard et al., 2002; Martone et al., 2005),the production of single cell protein from shrimp-shellwastes (Ferrer et al., 1996), the lactic acid fermentationfor ensilation of shrimp wastes (Evers and Carroll, 1998;Shirai et al., 2001), the composting of fishery processingwastes with peat (Martin, 1999) and the production of pep-tones from autohydrolyzed fish viscera (Vazquez et al.,2004). However, most of them were treated by enzymatichydrolysis and there were few studies on using fish wastesas a nutrient source for lactic acid production. Thus, fishwastes were hydrolyzed with the proposed acid-hydrolysisprocesses in this study and the processes were examinedthrough evaluating the performance of fish waste hydroly-zate (FWH) in lactic acid production.

2. Methods

2.1. Microorganism

Lactobacillus rhamnosus (NBRC 3863), a kind of L-lacticacid producer, was used in this study because it provideshigh productivity and yield to lactic acid production(Gao et al., 2004, 2005; Hirata et al., 2005). The strainwas stored at �80 �C. Before fermentation, the culturefor inoculation was carried out for 24 h in an incubatorat 37 �C under anaerobic condition. The medium (NissuiPharmaceutical Co., Ltd.) for inoculum preparation con-tained the following compositions (per liter). YE: 5.5 g;peptone: 12.5 g; glucose: 11.0 g; KH2PO4: 0.25 g;K2HPO4: 0.25 g; CH3COONa: 10.0 g; MgSO4: 0.1 g;MnSO4: 0.05 g; FeSO4: 0.05 g.

2.2. Fermentation medium

Two fermentation media were used in this study. Thecontrol medium contained, per liter, 20 g YE (Difco),100 g glucose, 0.1 g NaCl, 0.5 g K2HPO4 and 2.0 g MgSO4

and the experimental media contained, per liter, 100 g glu-cose, 0.1 g NaCl, 0.5 g K2HPO4, 2.0 g MgSO4 and 17 g,34 g, 51 g or 68 g FWH. All the mineral salts and glucosewere from Wako Pure Chemical Industries, Ltd.

2.3. Culture conditions

Fermentations were performed in a 2 L jar–fermentor(working volume: 1 L) equipped with temperature, agita-

tion and pH controllers. Nitrogen of small amount wassparged into the fermentor to remove dissolved oxygen infermentation broth. The temperature of the fermentationbroth was maintained at 42 �C and the rotation speed at100 rpm. The value of pH was controlled at 6.0 by theaddition of 10% aqueous ammonia.

2.4. Fish wastes

Fish heads and bony parts obtained from a supermarketwere used in this study. The fish wastes were ground andhomogenized immediately before use or were kept frozenuntil use.

2.5. Acid-hydrolysis

The flowsheets of acid-hydrolysis processes are shown inFig. 1. In process A, firstly the wastes were minced by agrinder (Ohmichi OMS-220, Japan). Secondly the mincedwastes were mixed with water to make fish wastes (wetweight)/water a ratio of 1:1. Thirdly the initial pH of thewaste slurry was set at 1 by the addition of 6 M HCl.And lastly the slurry was hydrolyzed at 121 �C for 20 minand were further centrifuged at 2706g for 20 min. Thesupernatant was used as a nutrient source for the produc-tion of lactic acid and the residue would be used as fertilizerafter neutralized. In process B, firstly the wastes afterminced by the grinder were mixed with water of equalweight. Secondly the mixer was pretreated at 121 �C for20 min without addition of acid. Thirdly supernatant Aand residue A were obtained after centrifugal separationof the pretreated fish wastes and residue A was furtheracid-hydrolyzed by the same procedure as process A.And lastly FWH was prepared by mixing supernatants Aand B. In addition, above 90% of fish wastes was hydro-lyzed by process B for 40 min.

2.6. Analyses

The concentrations of glucose and lactic acid were mea-sured by a HPLC system (Jasco CO-965, Japan). Sampleswere eluted with 0.75 mM H2SO4 in the column of TosohTSKgel Oapak-P at 40 �C. After heated with a dry-bathfor 10 min, samples were centrifuged and, then, microfil-trated with cellulose acetate filter of 0.45 lm. Glucose con-centration was detected with a refractive index detector(Jasco RI-930, Japan) and lactic acid concentration wasdetected with a UV/VIS detector (Jasco UV-975, Japan)at 210 nm. Total nitrogen was measured with an elemen-tary analyzer (Perkin–Elmer, Japan). Cell density was mea-sured with a spectrophotometer (UV1200/2000, Shimadzu,Japan) at 660 nm. Dry weight of FWH was measured afterdrying samples in an oven at 104 �C till the weight was sta-ble. In this study, FWH concentration was defined as dryweight of FWH divided by the volume of fermentationbroth that was 1 L, FWH volume included. Lactic acidyield was defined as amount of lactic acid produced divided

Residue

Acid

Fish Wastes

Grinding

Hydrolysis(121 ºC for 20 min)

Process A

pH Adjustment(pH 1)

FWH-A

Separation

Residue B

Acid

Fish Wastes

Grinding

Hydrolysis(121 ºC for 20 min)

pH Adjustment(pH 1)

Supernatant B

Separation

Pretreatment (Extraction by water at 121 ºC for 20 min)

Separation

Supernatant A Residue A

FWH-B

Process B

Fig. 1. Flowsheet of the production of FWH.

00

10 20 30 40 50 60

1

2

3

Hydrolysis Time [min]

Prod

uctiv

ity [

g/L

h]

Process A; Process B

Fig. 2. Effect of hydrolysis time on lactic acid production when FWH wasused as a nutrient source.

2416 M.-T. Gao et al. / Bioresource Technology 97 (2006) 2414–2420

by total amount of glucose and productivity as lactic acidconcentration divided by fermentation time. Moreover, Afreeze dry system (Asahi Life Science FZ-2.5, Japan) wasused to produce FWH powder.

3. Results and discussion

3.1. Effect of FWH on lactic acid production

In the preliminary study, the effect of hydrolysis time onlactic acid production was investigated. To each of the fer-mentations 200 g fish wastes (wet weight) were applied. Asshown in Fig. 2, the unhydrolyzed fish wastes were verypoor in lactic acid production. The fermentation with theunhydrolyzed fish wastes proceeded slowly and, finally, alarge amount of glucose was left in the fermentation broth,

0 10 20 30 400

20

40

60

80

100

Time [h]

Am

ount

s of

Glu

cose

and

L

acti

c A

cid

[g]

Supernatant A; Supernatant B

0

10

20

30

O.D

. [-]

Fig. 4. Effects of supernatants A and B on lactic acid production: (openkey) glucose; (closed key) lactic acid.

M.-T. Gao et al. / Bioresource Technology 97 (2006) 2414–2420 2417

leading to a low yield (data not shown). The result indi-cated that bacteria could not be capable of using the unhy-drolyzed protein in the fish wastes as a nutrient source. Onthe contrary, a positive effect was seen on productivity withthe supplementation of the hydrolyzed fish wastes. Theproductivity increased with hydrolysis time increasingand reached the maximum when the fish wastes werehydrolyzed for 40 min. The further hydrolysis caused theproductivity to decrease. The hydrolysis time was, there-fore, set at 40 min in the experiments that followed.Fig. 3 shows the detailed results of the fermentation withthe supplementation of FWH hydrolyzed by process A(FWH-A) or B (FWH-B) for 40 min. The result from thefermentation with YE supplementation was also shownin Fig. 3 to evaluate FWH performance by contrast withYE because YE was known to be the best nutrient for cellgrowth and lactic acid production. Since 15–20 g/L YE hadthe highest performance in lactic acid production andhigher YE concentration stimulated cell growth only, YEconcentration was set at 20 g/L in this study. As shownin the figure, both FWH-A and FWH-B had significanteffect on lactic acid production and glucose consumption,that is, high productivities of lactic acid and rapid con-sumption of glucose. Especially, FWH-B was more excel-lent than 20 g/L YE, which was significant for loweringthe production cost of lactic acid because YE had been rec-ognized as the best nutrient for the production of lacticacid (Arasaratnam et al., 1996; Goksungur and Guvenc,1997; Nancib et al., 2005). Process B can, therefore, be con-sidered to be a suitable method to produce high perfor-mance nutrients from organic wastes.

In process B, the mixture of supernatants A and B wasused as a nutrient source. To clarify their effects on lacticacid production, the fermentations were carried out withthe supplementation of supernatants A and B, respectively.As shown in Fig. 4, when using supernatant A as a nutrientsource, fermentation proceeded very slowly and only 17%glucose was converted to lactic acid, demonstrating a nitro-

0 10 20 300

50

100

Time [h]

Am

ount

s of

Glu

cose

and

L

acitc

Aci

d [g

]

FWH-A; FWH-B; YE 20 g/L

0

10

20

30

O.D

. [-]

Fig. 3. Time-courses of the fermentations with the supplementation ofFWH-A and FWH-B (hydrolysis time: 40 min): (open key) glucose;(closed key) lactic acid.

gen limitation in supernatant A. Supernatant B had highperformance in lactic acid production and the data fromsupernatant B were similar to those from FWH-A. Theresults indicated that the pretreatment of fish wastes in pro-cess B had no significant effect on acid-hydrolysis and thehigher performance of FWH-B than that of FWH-A canbe attributed to supernatant A obtained from the pretreat-ment. According to Arasaratnam et al. (1996) and Amraneand Prigent (1997), the nutrients to ensure the growth ofbacteria were considered to be peptides and growth factorssuch as vitamins and oligonucleotides. The growth factorswere liable to be destroyed by acid. Process B, however,overcame the drawback of acid-hydrolysis in the destruc-tion of growth factors. We supposed that some growth fac-tors were, during the pretreatment, extracted from fishwastes into supernatant A, avoiding the destruction inthe step of acid-hydrolysis. When supernatant B was com-bined with supernatant A, a performance higher than thatof 20 g/L YE was obtained though only supernatant A hadpoor performance in lactic acid production. Throughimproving process A to process B, the performance ofFWH increased by 22% in lactic acid production. In addi-tion, there was no need to centrifuge the pretreated fishwastes for obtaining supernatant A since the fish wasteswere denatured and congealed by autoclaving. Indeed,the procedure was not complicated.

3.2. Comparison with the other low-cost nutrients

To evaluate FWH performance in lactic acid produc-tion, it was compared with some low-cost nutrients interms of productivity, yield and lactic acid concentration.There were two standards for the nutrient concentrations.In Table 1 superscript �a� indicates the standard that wasbased on fermentation efficiency, in which case the opti-mum concentrations of the nutrients were chosen for thecomparison with YE. In Table 1 superscript �b� indicatesthe standard that was based on the nitrogen concentrationof the nutrients, in which case the nitrogen concentrations

Table 1Comparison of various low-cost nutrients in terms of productivity, yield and lactic acid production

Organism Nutrient Substrate Time (h) LA (g/L) P (g/Lh) Y (%) References

Lactobacillus rhamnosus

NBRC 3863YE 20 g/L Glucose 100 g/L 36 87 2.4 97 This work*FWH-A 6.8% 37 80 2.2 96*FWH-B 6.8% 29 79 2.7 100*FWH-B 5.1% 36 80 2.3 99*FWH-B 3.4% 45 80 1.8 97*FWH-B 1.7% 63 75 1.1 81

Lactobacillus casei

ATCC 10863Peptone 10 g/L + YE 5 g/L Glucose 50 g/L 30 41 1.4 100 Kurbanoglu and

Kurbanoglu (2003)*RHH 6% + YE 5 g/La 26 44 1.7 82

Lactobacillus amylovorus

NRRL B-4542

**N–Z–Soy 3.0%a Starch 5% – 53 1.9 – Hsieh et al. (1999)*Amisoy 3.0%a – 17 0.3 –

***A mixture of fiveLactobacillus strains

YE 15 g/L Glucose 120 g/L 40 118 2.9 – Lee (2005)CSL 7.5%a 68 111 1.6 –YE 3.1 g/L + CSL 5% 48 115 2.4 –

Lactobacillus helveticus

strain 02

**WPH 4%a Lactose 50 g/L 29 – – 96 Fitzpatrick andO�Keeffe (2001)

Lactobacillus casei

strain 01YE 5 g/L Lactose 55 g/L 34 42 1.2 – Pauli and Fitzpatrick (2002)Malt combing nuts 5%a 55 44 0.8

Lactobacillus casei subsp.rhamnosus ATCC 10863

YE 10 g/L Glucose 100 g/L 48 88 – – Yoo et al. (1997)*Casamino acid 10.8 g/Lb 48 22.1 – –

Lactobacillus casei

NRCC B-441YE 22 g/L Glucose 100 g/L 48 100 2.1 – Hujanen and Linko (1996)Peptone 15 g/Lb 66 79 1.2 –Malt sprouts 5.4%b 66 88 1.3 –

Lactobacillus rhamnosus

ATCC 10863YE 15 g/L Glucose 150 g/L 72 86 – – Kwon et al. (2000)*Amisoy 12.3 g/Lb 72 0 – –

Lactobacillus rhamnosus

NRRL-B455YE 10 g/L Glucose 50 g/L 40 24.8 0.62 – Nancib et al. (2005)**Tryptic soy 12.6 g/Lb 40 18 0.45 –

LA: lactic acid; P: productivity; Y: yield.* Hydrolysis with acids.** Hydrolysis with enzymes.*** L. delbrueckii subsp. lactis ATCC 12315, L. casei NRRLB1445, L. delbrueckii NRRL-B445, L. helveticus NRRL-B1937, and L. casei NRRL-B1922.

a The optimum concentrations of the nutrients that were chosen for the comparison with YE.b The concentrations of the nutrients were set to the value at which a nitrogen concentration of YE equivalent could be yielded.

2418 M.-T. Gao et al. / Bioresource Technology 97 (2006) 2414–2420

of the nutrients were set to the same as that of YE. Asshown in Table 1, none of the nutrients under the latterstandard brought lactic acid concentrations and productiv-ities as high as YE. To raise fermentation efficiency, nutri-ents had to be used in large amount. Indeed, none of thenutrients except FWH was, even at their optimum concen-trations, advantageous over YE in the production of lacticacid. Compared with the nutrients hydrolyzed withenzymes, the nutrients hydrolyzed with acids were muchlower in lactic acid production. That may be due to thedestruction of some important nutrients (Kwon et al.,2000) or excessive hydrolysis of protein (Hsieh et al.,1999; Kristinsson and Rasco, 2000) in the condition ofsevere acid-hydrolysis. In this study, hydrolysis withdiluted acid was coupled with extraction by water to reducethe destruction of growth factors by acid. By this way,FWH performance was much higher than the others. Theproductivity increased with more FWH supplementationand the 6.8% FWH-B was more excellent than 20 g/LYE. To make a comparison, the nitrogen concentrationsof FWH and YE were measured with an elementary ana-

lyzer and they were 75.1 g/kg (dry weight) and 105.2 g/kg, respectively. In the case of 6.8% FWH, 5.1 g nitrogenwas supplemented into 1 L fermentation broth, whereas2.1 g nitrogen was supplemented in the case of 20 g/LYE. The more nitrogen showed a positive effect on lacticacid production. However, amount of nitrogen utilizedby bacteria was less than 20% of the total amount in thecase of 6.8% FWH while about 35% in the case of 20 g/LYE. It indicated that the more supplementation of FWHmeant higher separation cost. Nevertheless, FWH was con-sidered as a substitute nutrient for YE because of its lowproduction cost. For instance, preparing 1 kg FWH (dryweight) in the experimental condition needed 4.27 kW h(2.35 kW h for autoclaving and 1.92 kW h for centrifuga-tion). Taking only energy consumption into account, theproduction cost was 38.4 cent/kg as electricity rate was9 cent/kW h. When 6.8% FWH was used, the nutrient costwas only 2.61 cent/L fermentation broth, too much lowerthan that of YE (6.04$/20 g YE) used in this study. As tothe price of YE, it was taken from Chemicals Catalog,33rd Edition, 2004 (Wako Pure Chemical Industries, Ltd.).

0 10 20 300

20

40

60

80

100

Time [h]

Am

ount

s of

Glu

cose

and

Lac

tic A

cid

[g]

FWH Liquid; FWH Powder

0102030

O.D

. [-]

Fig. 5. Comparison between FWH liquid and FWH powder at concen-tration of 5.1%: (open key) glucose; (closed key) lactic acid.

M.-T. Gao et al. / Bioresource Technology 97 (2006) 2414–2420 2419

3.3. Preparation of FWH powder

Because it is better for FWH to be applied in powderform, FWH was, in this study, dried with a freeze dry sys-tem into powder and the powder was tested for its perfor-mance in lactic acid production. Fish wastes werehydrolyzed by process B first and, then, dried by the freezedry system until its weight became stable. The producedpowder was stored in a refrigerator until use. To make acomparison, FWH in liquid and FWH in powder, bothprepared from the same amount of the fish wastes, wereused in the following fermentations. On account of the sep-aration cost, 5.1% FWH was chosen because the productiv-ity did not decrease much relative to 6.8% FWH and thenutrient cost could be reduced to 1.95 cent/L fermentationbroth. As shown in Fig. 5, similar to the fermentation withFWH liquid, the fermentation with FWH powder came toan end in 36 h with glucose completely consumed. Theresult indicated that the performance of FWH was notdeclined by drying and the powder was capable to be uti-lized as a substitute for YE as FWH liquid was. However,because drying by this way caused the production cost toincrease to 16.26$/kg, direct utilization of FWH liquid,when possible, is more economical and recommendable inlactic acid production.

4. Conclusions

Enzymatic hydrolysis of nutrients is generally used,however, acid-hydrolysis of nutrients was investigated inthis study for producing lactic acid in low-cost. Here, twoprocesses of acid-hydrolysis, process A and process B, wereproposed for the purpose of reducing the destruction ofnutrients by acid. Process A was a direct way to hydrolyzeprotein with diluted acid while process B was process Aplus fish wastes pretreatment (an extraction by water).Hydrolysis with diluted acid (process A) was an effectivemethod to hydrolyze nutrients and, when combined withthe extraction by water, became very capable of increasingnutrient performance in lactic acid production.

Taking fish wastes for example, the unhydrolyzed fishwastes were very poor in lactic acid production while thefish wastes hydrolyzed by the proposed processes were verygood and, especially, 6.8% FWH-B had greater productiv-ity than 20 g/L YE. Compared with the low-cost nutrientsreported, the FWH-B had the highest performance in lacticacid production. Considering its low-cost, FWH-B couldbe a potential nutrient and an substitute for YE, with anenvironmental solution in addition.

References

Amrane, A., Prigent, Y., 1997. Growth and lactic acid productioncoupling for Lactobacillus helveticus cultivated on supplemented whey:influence of peptidic nitrogen deficiency. J. Biotechnol. 55, 1–8.

Arasaratnam, V., Senthuram, A., Balasubramaniam, K., 1996. Supple-mentation of whey with glucose and different nitrogen sources forlactic acid production by Lactobacillus delbrueckii. Enzyme Microb.Technol. 19, 482–486.

Evers, D.J., Carroll, D.J., 1998. Ensiling salt-preserved shrimp waste withgrass straw and molasses. Anim. Feed Sci. Technol. 71, 241–249.

Ferrer, J., Paez, G., Marmol, Z., Ramones, E., Garcia, H., Forster, C.F.,1996. Acid hydrolysis of shrimp-shell wastes and the production ofsingle cell protein from the hydrolysate. Bioresour. Technol. 57, 55–60.

Fitzpatrick, J.J., O�Keeffe, U., 2001. Influence of whey protein hydrolysateaddition to whey permeate batch fermentations for producing lacticacid. Process Biochem. 37, 183–186.

Gao, M.-T., Hirata, M., Koide, M., Hano, T., Takanashi, H., 2004.Production of L-lactic acid by electrodialysis fermentation (EDF).Process Biochem. 39, 1903–1907.

Gao, M.-T., Koide, M., Gotou, R., Takanashi, H., Hirata, M., Hano, T.,2005. Development of continuous electrodialysis fermentation systemfor Lactobacillus rhamnosus based production of lactic acid. ProcessBiochem. 40, 1033–1036.

Goksungur, Y., Guvenc, U., 1997. Batch and continuous production oflactic acid from beet molasses by Lactobacillus delbrueckii IFO 3202. J.Chem. Technol. Biotechnol. 69, 399–404.

Guerard, F., Guimas, L., Binet, A., 2002. Production of tuna wastehydrolysates by a commercial neutral protease preparation. J. Mol.Catal. B: Enzym. 19–20, 489–498.

Hirata, M., Gao, M.-T., Toorisaka, E., Takanashi, H., Hano, T., 2005.Production of lactic acid by electrodialysis fermentation with a glucoseconcentration controller. Biochem. Eng. J. 25, 159–163.

Hofvendahl, K., Hagerdal, H.B., 2000. Factors affecting the fermentativelactic acid production from renewable resources. Enzyme Microb.Technol. 26, 87–107.

Hsieh, C.M., Yang, F.-C., Iannotti, E.L., 1999. The effect of soy proteinhydrolyzates on fermentation by Lactobacillus amylovorus. ProcessBiochem. 34, 173–179.

Hugenholtz, J., Kleerebezem, M., 1999. Metabolic engineering of lacticacid bacteria: overview of the approaches and results of pathwayrerouting involved in food fermentations. Curr. Opin. Biotechnol. 10,492–497.

Hujanen, M., Linko, Y.-Y., 1996. Effect of temperature and variousnitrogen sources on L(+)-lactic acid production by Lactobacillus casei.Appl. Microbiol. Biotechnol. 45, 307–313.

Kristinsson, H.G., Rasco, B.A., 2000. Fish protein hydrolysates: Produc-tion, biochemical, and functional properties. Crit. Rev. Food Sci.Nutr. 40, 43–81.

Kurbanoglu, E.B., Kurbanoglu, N.I., 2003. Utilization for lactic acidproduction with a new acid hydrolysis of ram horn waste. FEMSMicrobiol. Lett. 225, 29–34.

Kwon, S., Lee, P.-C., Lee, E.G., Chang, Y.-K., Chang, N., 2000.Production of lactic acid by Lactobacillus rhamnosus with vitamin-supplemented soybean hydrolysate. Enzyme Microb. Technol. 26,209–215.

2420 M.-T. Gao et al. / Bioresource Technology 97 (2006) 2414–2420

Lee, K., 2005. A media design program for lactic acid production coupledwith extraction by electrodialysis. Bioresour. Technol. 96, 1505–1510.

Martin, A.M., 1999. A low-energy process for the conversion of fisherieswaste biomass. Renew. Energ. 16, 1102–1105.

Martone, C.B., Borla, O.P., Sanchez, J.J., 2005. Fishery by-product as anutrient source for bacteria and archaea growth media. Bioresour.Technol. 96, 383–387.

Nancib, A., Nancib, N., Djalal, M.-C., Boubendir, A., Fick, M.,Boudrant, J., 2005. Joint effect of nitrogen sources and B vitaminsupplementation of date juice on lactic acid production by Lactoba-

cillus casei subsp. rhamnosus. Bioresour. Technol. 96, 63–67.Pauli, T., Fitzpatrick, J.J., 2002. Malt combing nuts as a nutrient

supplement to whey permeate for producing lactic acid by fermenta-tion with Lactobacillus casei. Process Biochem. 38, 1–6.

Shahidi, F., Han, X.-Q., Synowiecki, J., 1995. Production and character-istics of protein hydrolysates from capelin. Food Chem. 53, 285–293.

Shirai, K., Guerrero, I., Huerta, S., Saucedo, G., Castillo, A., Gonzalez,R.O., Hall, G.M., 2001. Effect of initial glucose concentration andinoculation level of lactic acid bacteria in shrimp waste ensilation.Enzyme Microb. Technol. 28, 446–452.

Vazquez, J.A., Gonzalez, M.P., Murado, M.A., 2004. Peptones fromautohydrolysed fish viscera for nisin and pediocin production. J.Biotechnol. 112, 299–311.

Wang, Q., Narita, J.-Y., Xie, W., Ohsumi, Y., Kusano, K., Shirai, Y.,Ogawa, H., 2002. Effects of anaerobic/aerobic incubation and storagetemperature on preservation and deodorization of kitchen garbage.Bioresour. Technol. 84, 213–220.

Yoo, I.-K., Chang, H.N., Lee, E.G., Chang, Y.K., Moon, S.-H., 1997.Effect of B vitamin supplementation on lactic acid production byLactobacillus casei. J. Ferment. Bioeng. 84, 172–175.