Lactic acid fermentation of food waste for swine feed

Bioresource Technology 97 (2006) 1858–1864

Lactic acid fermentation of food waste for swine feed S.Y. Yang a, K.S. Ji b, Y.H. Baik b, W.S. Kwak

b,¤

, T.A. McCaskey

c

a

b

Life Sciences Research Center, Dan Biotech, Chunan, Chung-Nam 330-834, Republic of Korea Animal Science, School of Life Resource and Environmental Sciences, College of Natural Sciences, Konkuk University, Danwol-dong 322, Chung-Ju, Chung-Buk 380-701, Republic of Korea c Department of Animal Sciences, Auburn University, Auburn, AL 36830, USA Received 4 May 2005; received in revised form 13 August 2005; accepted 26 August 2005 Available online 27 October 2005

Abstract This study was conducted to determine the eVects of lactic acid bacteria (LAB, Lactobacillus salivarius) inoculation on the microbial, physical and chemical properties of food waste mixture (FWM) stored at ambient temperature (25 °C) for 10 and 30 days. A complete pig diet including restaurant food waste, bakery by-product, barley and wheat bran, and broiler poultry litter was amended with LAB at the levels of 0.1%, 0.2%, 0.5% and 1.0% and fermented anaerobically. These treatments were compared with intact FWM before storage and non-anaerobically stored FWM. Non-anaerobic storage of FWM showed microbial putrefaction with the loss (P < 0.05) of water and water soluble carbohydrate (WSC) and increases (P < 0.005) in protein and Wber. Anaerobic fermentation of FWM with or without LAB seemed eVective in both 10and 30-day-storage. The addition of LAB inoculants to FWM showed a linear trend (P < 0.05) toward an increase in the number of total and lactic acid bacteria and toward the nutritional improvement with WSC increased and Wber decreased. Long-term (30 days) storage resulted in consistent reduction (P < 0.05) in numbers of total and lactic acid bacteria and pH and showed little change in chemical components, compared with short-term (10 days) storage. On the basis of these results, LAB inoculation improved fermentative characteristics of FWM. Among anaerobic treatments, further WSC increase and NDF reduction did not occur (P > 0.05) when LAB-added levels were over 0.2%. Based on these observations the optimum level of LAB addition to FWM was 0.2%. © 2005 Elsevier Ltd. All rights reserved. Abbreviations: FWM, food waste mixture; LAB, lactic acid bacteria; DM, dry matter; CP, crude protein; NDF, neutral detergent Wber; ADF, acid detergent Wber; WSC, water soluble carbohydrate Keywords: Food waste; Lactic acid bacteria; Fermentation; Pig; Feed

1. Introduction Food wastes derived from restaurant and households are principally putricible wastes that must be managed to control vermin, odor, disease and other sanitary issues associated with modern cities. In Korea, some food wastes are used as fertilizer or as feed ingredients for animal production. The huge production (about 4.5 million metric tons annually) of food wastes has led to intensive research in the Weld of waste recycling in Korea and the exploration *

Corresponding author. Tel.: +82 043 840 3521; fax: +82 043 851 8675. E-mail address: [email protected] (W.S. Kwak).

0960-8524/$ - see front matter © 2005 Elsevier Ltd. All rights reserved. doi:10.1016/j.biortech.2005.08.020

for bioengineering to transform food wastes into useful products. Bioconversion processes involving microbial metabolic processes oVer opportunities to transform food wastes into recycled products. Not only can the putrefaction of the wastes be stopped by processing the waste, but also the waste can be preserved and transformed into economically useful products. Lactic acid fermentation process, typical of the silage technique applied to forages, has been investigated by several workers (Jalil et al., 2001; Kherrati et al., 1998; Shirai et al., 2001). Food wastes have not been studied extensively to determine what type of processes might be applied to convert the wastes to useful products. Recently we developed

S.Y. Yang et al. / Bioresource Technology 97 (2006) 1858–1864

a method based on a biotechnological transformation/ preservation process to convert food wastes into a feed ingredient for swine. Lactic acid starter bacteria are commonly used as inoculants by food fermentation industries to produce a variety of fermented foods. The fermentative process not only creates a diVerent food product but also adds other attributes to the food such as preservation of the food and amends the food with desirable microbes. Also, the addition of microbial inoculants with or without enzymes has been reported to improve ensiling characteristics of alfalfa (Kung et al., 1991; Shockey and Borger, 1991; Sheperd et al., 1995), wheat forage (Froetschel et al., 1991), grass–legume forage (Stokes, 1992), corn forage (Kung et al., 1993) and wet brewers grains (Schneider et al., 1995). Because food wastes are typically wet and contain high levels of fermentable carbohydrate, the lactic fermentation process appears to be the method of choice for processing food wastes into animal feed. A study was conducted to determine the eVects of a lactic acid bacterium (Lactobacillus salivarius) inoculant to FWM on the microbial, physical and chemical properties of the FWM anaerobically fermented at ambient temperature (25 °C) for 10 and 30 days. 2. Methods 2.1. Manufacture of food waste mixture (FWM) Fresh food waste (79% moisture) collected from several oriental restaurants in Seoul, Korea was used in this experiment. The food waste was composed primarily of cooked rice, meat and vegetables. For pre-treatment, the food waste was ground and screened for foreign material removal with a screening grinder (Deplus Engineering Inc., Korea). Food waste (225 kg lots) was mixed 112.5 kg of bakery by-product, 87.5 kg of barley bran, 50 kg of wheat bran and 25 kg of broiler poultry litter which had been deep-stack processed as described in Kwak and Park (2003). The ratio of the ingredients was 16.1:34.8:26.8: 15.2:7.1 on a DM basis. An aerobic microbial culture was added at the level of 0.3% on a DM basis to hydrolyze starch to simple carbohydrates in the FWM. The aerobic microbial culture developed by Sakai and Kubota (1989) was composed primarily of Bacillus spp. Bakery by-product, barley bran and wheat bran were used as sources of energy and as water absorbents and broiler poultry litter as source of readily available N and minerals for fermentative microbes. Bakery by-product was a dried mixture of various bakery and bread by-products. The chemical composition of the Wnal mixture was as follows: DM 53.9%, crude protein 16.1%, neutral detergent Wber 26.5%, and acid detergent Wber 13.7%. The mixture was heated at 80 °C for 30 min using a 1 ton capacity vacuum drier (Deplus Engineering Inc., Korea) to meet the processing regulation of food waste as pig feed in Korea (KFMR, 2001). Following the heat process the mixture was cured for 18 h inside a plywood box (90 cm

1859

long £ 90 cm wide £ 130 cm high) with the upper surface open and the metal bottom perforated to facilitate free air movement. In order to determine the eVect of lactic acid bacteria addition on the storage of FWM, the manufactured FWM was treated as follows; intact FWM before storage (BS), non-anaerobic storage (NAN), anaerobic storage (AN), and anaerobic storage with lactic acid bacteria culture added at the level of 0.1% (ANL0.1), 0.2% (ANL0.2), 0.5% (ANL0.5), and 1.0% (ANL1.0) on a wet basis. A lactic acid bacteria (LAB) was inoculated at levels of 0.1%, 0.2%, 0.5%, and 1.0% (wet basis) into 15 kg aliquots of the FWM and mixed in a small scale mixer (Atika, Italia). The inoculated mixtures were sealed in double-lined plastic bags within 2 kg capacity plastic containers and fermented for 10 and 30 days at ambient temperature (25 °C) under anaerobic condition. For the non-anaerobic storage, the intact mixture was left over on the laboratory table at ambient temperature (25 °C) for 10 and 30 days. 2.2. Preparation of LAB culture An acid and bile tolerant Lactobacillus culture that was isolated from fresh piglet feces was used as the inoculant for the food waste. This was done to improve the opportunity that the lactic culture might serve as a probiotic culture in the gut of the pigs fed the lactic acid-fermented food waste. Selection for acid and bile tolerance also might help ensure that the culture would adapt to the gut environment of the pig. To screen for acid and bile tolerance, MRS (Difco) agar acidiWed to pH 2.0, amended with 1% oxgall was used to isolate Lactobacillus from piglet feces. The predominant lactic acid bacteria isolated from the feces was L. salivarius. The culture was identiWed based on morphological and biochemical characteristics provided by Bergey’s Manual (Krieg, 1984). Carbohydrate utilization patterns of the isolate were determined with the API CH50 kit (BioMerieux, France). The L. salivarius culture was maintained on MRS agar held at 4 °C. Prior to inoculation of the food waste, the culture was grown on MRS broth at 37 °C for 48 h to produce cell mass which was harvested by centrifugation at 5000g. The centrifugate had a viable lactobacilli count of 1.2 £ 108 cfu/ml, and this was used as starter (inoculum) for the FWM. The starter was added to the FWM at 0%, 0.1%, 0.2%, 0.5%, and 1.0% (v/w). The inoculated FWM was held at 25 °C and analyzed at 0, 10 and 30 days after fermentation. The lactobacillus bacterial count was determined by the pour plate procedure on MRS agar inoculated at 37 °C for 48–72 h. 2.3. Physical and microbial analysis Physical properties of the FWM fermented 10 and 30 days under diVerent conditions were observed for fungal growth, for putrefactive appearance and for alcoholic and acidic odors. Three trained evaluators observed the treated

1860

S.Y. Yang et al. / Bioresource Technology 97 (2006) 1858–1864

FWM subjectively by a casual observation method used in our laboratory. Five grams of each sample were homogenized with 495 ml of sterile physiological saline (0.85% NaCl) to achieve a 1:100 dilution of the FWM. Additional dilutions were achieved using 9 ml aliquots of sterile saline. The total bacterial count (aerobes and facultative anaerobes) was determined on plate count agar incubated at 30 °C for 48 h. LAB were determined on MRS agar incubated at 37 °C for 24 h. Yeasts were determined on yeast-malt extract agar (Difco Laboratories, USA) incubated at 37 °C for 48 h. Colony counts of yeast and Bacillus spp. were determined by selective counting characteristic colonies on the agar plates. 2.4. Chemical analysis Dry matter was determined by drying samples at 65 °C for 48 h to constant weight. Crude protein was analyzed by the Kjeltec System (Tecator, Denmark) using AOAC methods (1990). Neutral detergent Wber (NDF) and acid detergent Wber (ADF) were determined according to the method of Van Soest et al. (1991). pH was measured using a pH meter (HI9321, Hanna Instrument, Portugal). Water soluble carbohydrate was determined by the method of Dubois et al. (1956). 2.5. Statistical analysis All data were analyzed by the general linear model procedure of SAS (SAS Institute Inc., 1990). Means were compared by general contrast including BS vs NAN; BS vs AN treatments; and NAN vs AN treatments (SAS Institute Inc., 1990). Among AN treatments, linear, quadratic and cubic trends were tested by polynomial contrasts (SAS

Institute Inc., 1990). Comparison of means among AN treatments was made using LSD test (SAS Institute Inc., 1990). Comparison of means between 10- and 30-day treatments was made using student-t test (SAS Institute Inc., 1990). 3. Results and discussion 3.1. Physical parameters of fermented FWM Physical parameters of the fermented FWM for swine feed were analyzed subjectively by trained evaluators (Table not presented). Both the 10- and 30-day observation patterns for fungal growth and for alcoholic and acidic odors were the same. The non-anaerobically (NAN) stored FWM was putreWed and had signiWcant fungal growth. The anaerobically (AN) stored waste did not putrefy and fermentation was evident by an acidic odor of the FWM. Yeast fermentation was evident by an alcoholic odor of the FWM, and this appeared to diminish as the lactic starter inoculum addition to the FWM was increased. 3.2. Microbial parameters Microbial population and pH changes in the FWM held for 10 and 30 days are shown in Table 1, Fig. 1. The pH of anaerobically stored FWM dropped from pH 5.98 to pH 4.5 at the 10-day storage period and to pH 4.4 at the 30-day storage period, whereas the pH of aerobically stored FWM increased from 5.98 to 6.51 for a 0.5 pH unit increase. The lower (P < 0.005) pH of the fermented FWM is attributed to the anaerobic storage of the FWM which encourages the growth of acid-producing bacteria that convert fermentable FWM carbohydrate to lactic acid. A fast pH decline by

Table 1 Microbial population (Log10 cfu/g1) and pH of diVerently treated food waste mixtures for swine feed depending upon the 10- and 30-day storage perioda,b Item

BS

NANc

AN

ANL0.1

ANL0.2

ANL0.5

ANL1.0

SE

0-day vs 10-day storage Total bacteriad,e,f,g Lactic acid bacteriad,e,f,g Bacillus spp. Yeastd,f,h pHd,e,f

7.49 7.40 6.39 6.29 5.98

UC UC 7.01 8.91 6.41

9.17 9.16 6.36 6.15 4.58

9.13 9.12 5.86 5.83 4.55

9.42 9.41 6.81 6.08 4.60

9.15 9.14 6.01 6.18 4.55

9.64 9.63 6.65 5.97 4.54

0.20 0.26 0.35 0.13 0.02

0-day vs 30-day storage Total bacteriad,e,g Lactic acid bacteriad,e,g Bacillus spp.d,f,h Yeastd,f,h pHd,e,f

7.49 7.40 6.39 6.29 5.98

9.08 9.02 8.41 8.59 6.51

8.38 8.35 6.32 6.02 4.45

8.67 8.65 6.03 6.22 4.43

8.47 8.45 6.32 6.13 4.43

8.58 8.57 5.92 5.82 4.42

8.99 8.89 6.52 6.25 4.43

0.21 0.27 0.36 0.19 0.03

a

Colony-forming unit per gram of wet samples. BS D before storage; NAN D non-anaerobic storage; AN D anaerobic storage; ANL D anaerobic storage with lactic acid bacteria added at the level of 0.1% for ANL0.1, 0.2% for ANL0.2, 0.5% for ANL0.5, and 1.0% for ANL1.0. c UC D uncountable due to the over-growth of molds. d BS diVers from NAN (P < 0.005). e BS diVers from AN, ANL0.1, ANL0.2 ANL0.5 and ANL1.0 (P < 0.005). f NAN diVers from AN, ANL0.1, ANL0.2 ANL0.5 and ANL1.0 (P < 0.005). g Among AN treatments, polynomial contrasts showed a linear trend (P < 0.05). h Among AN treatments, polynomial contrasts showed a cubic trend (P < 0.05). b

S.Y. Yang et al. / Bioresource Technology 97 (2006) 1858–1864 10

a

a 9.5

a

a a

9

a b

b b

b

b

Lactic acid bacteria (Log10cfu/g)

Total bacteria (Log10cfu/g)

10

8.5

8

7.5

9.5

a

a

a

a

9

b

8.5

8

ANL0.1

ANL0.2

ANL0.5

AN

ANL1.0

ANL0.1

10d storage

30d storage

10d storage

ANL0.5

ANL1.0

30d storage

10

ab

Bacillus sp. (Log10cfu/g)

9

ANL0.2

Treatments

Treatments

10

b

b

b

b

7.5

AN

Yeast (Log10cfu/g)

1861

8 7 6 5 4 3 2 1 0

9 8 7 6 5 4 3 2 1 0

NAN

AN

ANL0.1

ANL0.2

ANL0.5

NAN

ANL1.0

AN

10d storage

ANL0.1

ANL0.2

ANL0.5

ANL1.0

Treatments

Treatments

10d storage

30d storage

30d storage

7 6.5

a

b

pH

6 5.5 5

a 4.5

b

a

b

a

b

a

b

ab

4 NAN

AN

ANL0.1

ANL0.2

ANL0.5

ANL1.0

Treatments 10d storage

30d storage

Fig. 1. EVects of the 10-day vs 30-day storage on chemical composition (% of dry matter) of diVerently processed food waste mixture for swine feed [NAN D non-anaerobic treatment; AN D anaerobic fermentation; ANL D anaerobic fermentation with lactic acid bacteria added at the level of 0.1% for ANL0.1, 0.2% for ANL0.2, 0.5% for ANL0.5, and 1.0% for ANL1.0; Means with diVerent letters within the same treatment diVer (P < 0.05)].

LAB inoculation to silages was observed in other research (Schneider et al., 1995; Sheperd et al., 1995). A pH of 4–5 is desired for fermented feed ingredients because below pH 4 feed intake is decreased and over pH 5 microbial spoilage is likely to occur (Lee et al., 2004). Total bacteria and lactic acid bacteria counts increased (P < 0.05) during anaerobic storage of the FWM. The addition of lactic acid bacteria inoculum to FWM showed a trend toward an increase in the number of total and lactic acid bacteria relating to the inoculum level. Compared with the AN treatment (LAB not added), LAB counts increased (P < 0.05) when LAB was added at 0.2% and 1.0% levels for 10 days storage and when LAB was added at the 1.0% level for 30 days storage.

Yeast and Bacillus spp. counts were higher (P < 0.005) for aerobically stored FWM than for anaerobically stored FWM. In addition, the concentration of water-soluble carbohydrates was increased (P < 0.005) by additional levels of LAB (Table 2). According to a general theoretical scheme suggested for lactic acid fermentation, insoluble complex organic polymer components (such as protein, fats and carbohydrates) are Wrst hydrolyzed into smaller soluble compounds by extracellular enzymes and the low pH. In the next stage, the soluble sugars are converted into lactic acid by LAB (Wang et al., 2003). Lactobacilli generally have a higher tolerance to low pH than lactococci and other lactic acid bacteria (Ohmomo et al., 2002). The minimum pH for the growth of most lactobacilli is about pH 4

1862

S.Y. Yang et al. / Bioresource Technology 97 (2006) 1858–1864

Table 2 Chemical composition (% of dry matter) of diVerently treated food waste mixtures for swine feed depending upon the 10- and 30-day storage perioda Item

BS

NAN

AN

ANL0.1

ANL0.2

ANL0.5

ANL1.0

SE

0-day vs 10-day storage Dry matterb,c,d Water soluble carbohydrateb,c,d,e Crude proteinb,d Neutral detergent Wberb,d,e,f Acid detergent Wberb,c,d,e

61.8 5.14 15.9 25.5 14.7

73.5 0.78 19.8 33.4 19.2

60.8 7.43 16.1 28.1 19.0

60.7 7.86 16.1 26.2 18.6

60.9 9.37 16.1 26.8 17.0

60.8 9.57 16.2 26.3 17.5

60.5 10.30 16.1 27.2 15.9

0.21 0.57 0.21 0.52 0.36

0-day vs 30-day storage Dry matterb,c,d Water soluble carbohydrateb,c,d,e,f Crude proteinb,d Neutral detergent Wberb,d,e,f Acid detergent Wberb,c

61.8 5.14 15.9 25.5 14.7

80.4 1.39 20.3 37.9 17.9

59.9 7.07 16.1 30.2 17.4

59.9 8.37 16.0 28.6 17.8

60.6 10.44 16.1 26.9 17.2

60.5 10.88 16.0 27.6 16.8

60.1 10.37 16.2 28.2 16.7

0.35 0.49 0.16 0.94 0.60

a BS D before storage; NAN D non-anaerobic storage; AN D anaerobic storage; ANL D anaerobic storage with lactic acid bacteria added at the level of 0.1% for ANL0.1, 0.2% for ANL0.2, 0.5% for ANL0.5, and 1.0% for ANL1.0. b BS diVers from NAN (P < 0.005). c BS diVers from AN, ANL0.1, ANL0.2 ANL0.5 and ANL1.0 (P < 0.005). d NAN diVers from AN, ANL0.1, ANL0.2 ANL0.5 and ANL1.0 (P < 0.005). e Among AN treatments, polynomial contrasts showed a linear trend (P < 0.01). f Among AN treatments, polynomial contrasts showed a quadratic trend (P < 0.05).

(Wang et al., 2001). If a LAB strain could survive at about pH 3.5 and maintain a high concentration of living cells, it would competitively inhibit the growth of other microbes and thereby contribute to the preservation of the FWM. L. salivarius, which was used as the inoculum for the FWM, showed high acid tolerance. The culture tolerated pH 2.5 for 3 h without appreciable loss of viability (Shin et al., 2002). Yeast populations were higher (P < 0.005) in aerobically stored FWM than in anaerobically stored FWM and the populations were not aVected by storage time. Among AN treatments, polynomial contrasts of yeast counts showed a cubic trend (P < 0.05) at both of the storage periods. Yeast can attain more energy from soluble carbohydrates during aerobic growth (respiration) than during anaerobic growth (fermentation). Yeast can metabolize the lactic acid and tolerate the low pH of the fermented FWM. During aerobic storage, an increase in DM and CP and a decrease in watersoluble carbohydrates was the result of proliWc yeast growth (Table 2, Fig. 2). 3.3. Chemical parameters Chemical composition of the treated FWM held for 10 and 30 days are presented in Table 2. The FWM held for 10 days under non-anaerobic storage (NAN) showed microbial putrefaction with the loss (P < 0.005) of water and WSC and an increase (P < 0.005) in protein and Wber, both NDF and ADF Wber. Compared to FWM before storage (BS), anaerobic storage of FWM with or without LAB addition resulted in a decrease (P < 0.005) in DM level, an increase (P < 0.005) in WSC (45%–100%) and ADF levels, with little change (P > 0.05) in CP and NDF levels. The increased level of WSC indicated that the rate of carbohydrate breakdown

into WSC exceeded that of WSC conversion into microbial biomass or other metabolites. Among anaerobic treatments, as the added levels of LAB increased, WSC was increased linearly (P < 0.01), whereas NDF and ADF were linearly (P < 0.01) and quadratically (P < 0.05) decreased, and DM and CP were not aVected (P > 0.05). Compared with the AN treatment (LAB not added), WSC in the FWM increased (P < 0.05) when LAB was added at the levels between 0.2% and 1.0% for both of the storage periods and NDF decreased (P < 0.05) when LAB was added at 0.2% and 0.5% levels. Increased WSC content was also observed when wet brewers grains inoculated with LAB was fermented for 6 days (Schneider et al., 1995). Although microbial inoculation to forage and by-product silage did not aVect NDF and ADF contents in other studies (Kung et al., 1991, 1993; Stokes, 1992; Schneider et al., 1995), its inoculation to FWM in our study showed a slight but signiWcant decline in Wber contents. Anaerobic storage of FWM with or without LAB addition resulted in lower (P < 0.005) levels of DM, CP and Wber (NDF and ADF) and higher (P < 0.005) levels of WSC compared to the levels in non-anaerobically treated FWM. When the FWM treatments were stored for long-term (30 days), the pattern of change in chemical composition depending upon treatments was identical to the short-term (10 days) storage with the exception of ADF, in which there was no diVerence detected between non-anaerobic and anaerobic treatments and also among anaerobic treatments. Long-term (30 days) non-anaerobic storage induced higher (P < 0.05) DM, CP and NDF contents and lower (P < 0.05) ADF content in FWM, however, the anaerobic treatments showed little change in chemical components, compared with short-term (10 days) storage. In general, among AN treatments, further WSC increase and NDF reduction did not occur (P > 0.05) when LAB-added levels were over 0.2%.

S.Y. Yang et al. / Bioresource Technology 97 (2006) 1858–1864 100

14

b 80

1863

b

12

a

a

WSC (%)

DM (%)

10 60

40

8 6 4

20 2 0

0 NAN

AN

ANL0.1

ANL0.2

ANL0.5

ANL1.0

NAN

AN

Treatments 10d storage

ANL0.2

ANL0.5

ANL1.0

Treatments

30d storage

10d storage 45

25

ab

40

20

30d storage

b a

35

NDF (%)

CP (%)

ANL0.1

15

10

30 25 20 15 10

5

5 0

0 NAN

AN

ANL0.1

ANL0.2

ANL0.5

NAN

ANL1.0

AN

Treatments 10d storage

ANL0.1

ANL0.2

ANL0.5

ANL1.0

Treatments

30d storage

10d storage

30d storage

25

ADF (%)

20

a b

15 10 5 0 NAN

AN

ANL0.1

ANL0.2

ANL0.5

ANL1.0

Treatments 10d storage

30d storage

Fig. 2. EVects of the 10-day vs 30-day storage on microbial population and pH of diVerently treated food waste mixture for swine feed [NAN D non-anaerobic treatment; AN D anaerobic fermentation; ANL D anaerobic fermentation with lactic acid bacteria added at the level of 0.1% for ANL0.1, 0.2% for ANL0.2, 0.5% for ANL0.5, and 1.0% for ANL1.0; Means with diVerent letters within the same treatment diVer (P < 0.05)].

4. Conclusion Our results indicated that non-anaerobic storage of FWM resulted in microbial putrefaction. Anaerobic treatment of FWM was an eVective storage method. The microbial inoculant was beneWcial in stimulating microbial fermentation and in improving fermentative characteristics, that is, consistently higher WSC content and lower NDF content in FWM. These results indicated that LAB inoculants induced signiWcant breakdown of Wber (NDF) into soluble carbohydrate. Food waste mixture could be more eVectively utilized for short-term (10 days) anaerobic storage showing more microbial population, compared with long-term (30 days) storage. No further WSC increase and NDF decrease at LAB-added levels

over 0.2% indicated that the optimum level of LAB addition to FWM was 0.2% on a wet basis. Improved microbial and chemical characteristics of LAB addition to FWM needs to be further proved by an in vivo animal feeding study. References AOAC, 1990. OYcial methods of analysis, 15th ed. Association of OYcial Analytical Chemists. Washington, DC, USA. Dubois, M., Gilles, K.A., Hamilton, J.K., Rebers, P.A., Smith, F., 1956. Colorimetric method for determination of sugars and related substances. Anal. Chem. 28, 350. Froetschel, M.A., Ely, L.O., Amos, H.E., 1991. EVects of additives and growth environment on preservation and digestibility of wheat silage fed to Holstein heifers. J. Dairy Sci. 74, 546–556.

1864

S.Y. Yang et al. / Bioresource Technology 97 (2006) 1858–1864

Jalil, M.H.E., Faid, M., Elyachioui, M., 2001. A biotechnological process for treatment and recycling poultry wastes manure as a feed ingredient. Biomass Bioenerg. 21, 301–309. Kherrati, B., Faid, M., Elyachioui, M., Wahmane, A., 1998. Process for recycling slaughterhouses wastes and by-products by fermentation. Bioresour. Technol. 63, 75–79. Korean Feed Management Regulation, 2001. Range and Standard of Hazardous Feed. Ministry of Agriculture and Forestry Notice No. 2001-61, 2001.10.5. Korea. Kwak, W.S., Park, J.M., 2003. EVects of molasses addition and (or) pelleting on nutritional characteristics of broiler litter processed by ensiling or deepstacking and palatability improvement by ‘Hanwoo’ during the adjustment period. J. Anim. Sci. Technol. (Kor.) 45 (1), 87–100. Krieg, N.R., 1984. Bergey’s Manual of Systematic Bacteriology (vol. 1). Williams & Wilkins, Baltimore, USA. Kung Jr., L., Chen, J.H., Kreck, E.M., Knutsen, K., 1993. EVect of microbial inoculants on the nutritive value of corn silage for lactating dairy cows. J. Dairy Sci. 76, 3763–3770. Kung Jr., L., Tung, R.S., Maciorowski, K.G., BuVum, K., Knutsen, K., 1991. EVects of plant cell-wall-degrading enzymes and lactic acid bacteria on silage fermentation and composition. J. Dairy Sci. 74, 4284– 4296. Lee, K.S., Lee, K.Y., Oh, C.S., Lee, D.G., Kim, Y.J., 2004. EVect of aeration for the probiotic feed production from food wastes by Lactobacillus acidophilus and Saccharomyces cerevisiae. J. KOWREC. 11 (4), 114– 119. Ohmomo, S., Tanaka, O., Kitamoto, H.K., Cai, Y., 2002. Silage and microbial performance, old story but new problems. JARQ 36, 59–71. Sakai, S., Kubota, Y., 1989. EYcient utilization of unused organic resources in rural district (1)—on fermentation treatment of various organic residues. Shinshu University, Ann. Environ. Sci. 11, 27–32.

SAS Institute Inc., 1990. SAS User’s Guide, version 6.08, Fourth ed. SAS Inst. Inc., Cary, NC, USA. Schneider, R.M., Harrison, J.H., Loney, K.A., 1995. The eVects of bacterial inoculants, beet pulp, and propionic acid on ensiled wet brewers grains. J. Dairy Sci. 78, 1096–1105. Sheperd, A.C., Maslanka, M., Quinn, D., Kung Jr., L., 1995. Additives containing bacteria and enzymes for alfalfa silage. J. Dairy Sci. 78, 565–572. Shin, J.Y., Yang, S.Y., Song, M.D., 2002. Isolation and identiWcation of lactic acid bacteria for direct-fed microbials. Konkuk J. Nat. Sci. (Kor.) 13, 37–43. Shockey, W.L., Borger, D.C., 1991. EVect of salt on fermentation of alfalfa. 2. Treatment with sodium chloride, Clostridium butyricum, and lactic acid bacteria. J. Dairy Sci. 74, 160–166. Shirai, K., Guerrero, I., Huerta, S., Saucedo, G., Castillo, A., Gonzalez, R.O., Hall, G.M., 2001. EVect of initial glucose concentration and inoculation level of lactic acid bacteria in shrimp waste ensilation. Enzyme Microb. Technol. 28, 446–452. Stokes, M.R., 1992. EVects of an enzyme mixture, an inoculant, and their interaction on silage fermentation and dairy production. J. Dairy Sci. 75, 764–773. Van Soest, P.J., Robertson, J.B., Lewis, B.A., 1991. Methods of dietary Wber, neutral detergent Wber, and nonstarch polysaccharides in relation to animal nutrition. J. Dairy Sci. 74, 3583–3597. Wang, Q., Narita, J., Ren, N., Fukushima, T., Ohsumi, Y., Kusano, K., Shirai, Y., Ogawa, H.I., 2003. EVect of pH adjustment on preservation of kitchen waste used for producing lactic acid. Water Air Soil Pollut. 144, 405–418. Wang, Q., Yamabe, K., Narita, J., Morishita, M., Ohsumi, Y., Kusano, K., Shirai, Y., Ogawa, H.I., 2001. Suppression of growth of putrefactive and food poisoning bacteria by lactic acid fermentation of kitchen waste. Process Biochem. 37, 351–357.