Extending the Shelf-Life of Vacuum-Packaged Pork Liver Pate

881

ROBERT H. MADDEN12

'Food and Agricultural Microbiology Research Division, Department of Agriculture for Northern Ireland, Newforge Lane, Belfast BT9 5PX, Northern Ireland; department of Food and Agricultural Microbiology, The Queen's University of Belfast, Newforge Lane, Belfast B'1'9 5PX, Northern Ireland.

(Received for publication February 29, 1988) Downloaded from http://meridian.allenpress.com/jfp/article-pdf/52/12/881/1654176/0362-028x-52_12_881.pdf by guest on 23 September 2021

ABSTRACT (Oxoid CM275) and yeasts and molds (YAM) on malt extract agar (Oxoid CM59) plus 50 mg/1 chloramphenicol (Sigma 000378). Processing conditions to achieve a minimum shelf life of 1 Aerobic plate counts (APC) were made on nutrient agar (Oxoid month at 4°C for a vacuum packaged pork liver pate were de CM3) and total anaerobe counts (TAC) on brain heart infusion fined. Cooking to a core temperature of 70°C and maintaining agar (Oxoid 375) plus 3 g/1 yeast extract (Oxoid L21) (I). YAM, less than 25% added water prevented spoilage by lactic acid APC, and TAC plates were incubated at 22°C for 3 d, TAC bacteria. Higher added water levels caused spoilage due to greening plates in a Forma Scientific Model 1024 (Marietta, Ohio, USA) resulting from the proliferation of Lactobacillus viridescens. anaerobic cabinet with an atmosphere of 85%> N2, 10% H2, 5% Glycerol at levels up to 1% (w/w) accelerated sour spoilage but CO, (all v/v). All other media were incubated according to glucono-delta-lactone (GDL) at 0.5% (w/w) markedly reduced manufacturers or authors instructions. the growth of lactic acid bacteria. However, with GDL yeasts Preliminary trials were conducted to determine which tended to proliferate. Overall shelf life was best maintained by microorganisms would be significant during spoilage, and two control of the a^ of the product through regulation of added distinct types of spoilage were noted. Pate stored for an ex water or the use of GDL. tended period produced sour odors when the pack was opened, while packs opened after shorter periods subsequently developed visible slime and discoloration if stored in air for several days. The keeping quality of edible offals depends on the Sour spoiled samples had high TAC counts and elevated levels microbial load which they acquire during evisceration and of lactic acid. Flora analyses showed that the dominant popula on their subsequent storage conditions (26). The initial tion was lactic acid bacteria. Packs opened and left in air, however, had high APC counts level is comparable with that of carcass meat (17) but and flora analyses showed Bacillus, Micrococcus, and Coryne- good temperature control is essential (8) to control bacte bacterium spp. to be dominant. It was noted that lactic acid rial activity. The safety of pork liver has been questioned bacteria could produce pinpoint colonies on nutrient agar, as had due to frequent isolations of Salmonella (22). However, also been found in previous studies (16). To differentiate be other workers have found a low incidence of Salmonella tween the causes of the spoilage conditions found, one being in pork liver (4,15). distinctly aerobic and the other anaerobic, it was necessary to Processing of offals can produce a product with a ensure the lactic acid bacteria were not enumerated as part of the substantially longer shelf-life than the original material APC. Hence, nutrient agar plates were checked using hydrogen and a higher profit margin (26). Pate, which is primarily peroxide (30% v/v) and very small catalase negative colonies a perishable comminuted meat product containing nitrite discounted. This ensured the APC was a measure of the micro and optionally nitrate and-or ascorbate, is an example of organisms capable of causing aerobic spoilage. Colonies from TAC plates were also checked with hydro such a processed product (25). gen peroxide to ensure the anaerobic cabinet was oxygen-free. With the establishment of local companies producing Catalase positive colonies were inoculated into Hugh and Leifsons pork liver pate guidelines were developed (14), and a medium (23) to ensure fermentation was taking place. Growth cooperative study undertaken with one manufacturer to of obligate aerobes would indicate leakage, and the work would define the processing conditions required to ensure a be repeated after repairs. minimum shelf life of one month for vacuum-packaged Samples (10 g) for microbial analysis were blended (2 min), tubs (220 g) of pork liver pate held at 4°C. using a stomacher (Colworth 400), in 90 ml peptone saline diluent (Oxoid CM9). All subsequent dilutions used 9 ml portions of MATERIALS AND METHODS this diluent. All sampling for microbial analyses was in dupli cate. Enterobacteriaceae counts (coliforms) and Salmonella Results were statistically analysed by analysis of variance detection were performed as previously described (15). Bro- (10). Presumptive Lactobacillus spp. were initially characterized chothrix thermosphacta was enumerated on STAA agar (5). by standard methods (23) then identified using the API 50CHL Staphylococcus aureus was enumerated on Baird-Parker agar system (API System, S.A., La Balme les Grottes, France).

JOURNAL OF FOOD PROTECTION. VOL. 52, DECEMBER 1989 882 MADDEN

Lactic acid was determined enzymatically (Sigma kit 726- TABLE 1. Microbial counts of raw ingredients, mean of 8 samples UV, Saint Louis, Missouri, USA) and water activity (a ) was taken at the processing factory. determined using a Protimeter DP 680 (Marlow, Buchi, UK) Ingredient L°g,o (APC g>) L°g,o (yeasts and dew-point meter. In both cases triplicate samples were analysed. molds g_1) Pate was prepared in one commercial premises using ingre liver 6.91 0.31 <1 dients routinely available: pork liver 2.3 kg, bacon trim 3.2 kg, cured pork 5.03 0.52 3.60 0.56 rusk (white, superfine) 0.9 kg, water 1.4 kg, dried mixed herbs rusk 3.34 0.32 1.11 0.10 170 g, mixed spices 110 g. Herbs, spices, and rusk (T. Lucas spices 4.52 0.22 2.94 0.21 Ltd, Kingswood, Bristol, UK) were standard butchers wholesale herbs 5.79 0.26 4.52 0.30 supplies. Sodium nitrite was added as a solution made with part of the water to give a final level of 150 mg kg 1. The liver and bacon were blanched in boiling water for approximately 5 min before blending with the other ingredients (2 min) in a Robot Coupe R15 blender. 46.3

A Franke U.L. 2IE fan assisted oven (120°C) was used for Downloaded from http://meridian.allenpress.com/jfp/article-pdf/52/12/881/1654176/0362-028x-52_12_881.pdf by guest on 23 September 2021 all cooking. Temperatures were measured using Thermocouple Instruments model 55 electronic thermometer (Penewyn, Wales). ^ -* 66.9 The probe was marked to allow the sensor to be accurately placed at the center of a tub. The temperature measured is referred 74.1 to subsequently as the core temperature. Six samples were measured to provide mean temperatures. Plastic tubs (approximately 12 X 9 cm oval, 3 cm deep) were filled with 220 g of pate and after cooking placed in vacuum pouches (Intervac, Hemel Hempstead, GB) with transmission 2 rates (20°C, 100% relative humidity) of 1.69 g/m /d (H20), 40 2 ml/nr/d (02), and 120 ml/m /d (C02). After cooling overnight at 2"C, the pouches were evacuated and sealed. Whenever possible, a single large batch of pate was pre pared, subdivided, then individual treatments carried out. Five TIME (DAYS) treatments, including a control, were usually studied at one time, Figure 1. Effect of core temperature (°C) on total anaerobic i.e. 120 tubs which were loaded in a randomized fashion to count (TAC) during storage at 1°C. allow for temperature variations in the oven. Treatments were studied in duplicate. Raw ingredients were sampled at the premises used for cooking, eight times over a period of four weeks. Sampling of age at 1°C, the TAC showed no lag phase after cooking pate during storage trials, in duplicate, took place twice a week to 46.3°C (Fig. 1). Longer cooking gave a marked reduc for up to 34 d. tion in numbers and slower subsequent multiplication. Analy To determine the lowest core cooking temperature, a thermo sis of the TAC at 34 d showed a linear relationship be couple was fixed at the center of a tub in the center of the oven tween the core temperature after cooking and log1() TAC. and 22 tubs were removed at 45, 55, 65, 75, and 80°C. This Since the growth of anaerobes leading to souring was the study used 220 g packs as models for wholesale 2.2 kg tubs, but main problem it was decided to keep numbers below 1% the smaller packs were adopted by the manufacturer as retail of the potential spoilage level of about 107/g- Using the packs, hence subsequent experiments used 4°C storage to simu late good display chills rather than ideal wholesale storage at regression equation the log]() TAC for 65 and 70°C were 1°C. 4.86 and 4.45 respectively. To allow for the variation in temperatures seen in the samples, a core temperature of 70°C was recommended and adopted in the studies de RESULTS scribed below. The effect of altering the quantity of water used was The initial work aimed to define the lowest core cooking then studied (Table 2). The upper limit was determined temperature allowing a storage life of one month at 1°C. experimentally as that which gave a product judged by the Water activity fell linearly with increasing temperature — pate manufacturers to be slightly too soft for the con sumer. The effects of the change in a are shown in Fig. 3 aw = 0.969 - 1.06 x 10 T, p<0.05 (T = temperature in °Q 2. Lowering the aw reduces the rate of exponential growth (p<0.01) and the TAC after 28 d was directly proportional

From the microbial analyses of the raw ingredients (Table to the aw (P<0.05). Again the APC was well controlled 1), the raw pate would have had a mean APC of 2.37 x and the maximum numbers seen were 3.72 x 104/g. 6 10 /g- Cooking decreased the APC by 2 to 4 log cycles. Analysis of variance showed that acid production did

Little increase in numbers was detected in any samples not significantly affect the pH at any aw (P<0.05). How during the first 22 d of storage. ever, at the highest a the concentration of lactic acid rose While the APC was controlled by even the shortest from 4 to 510 u.g/g during 28 d storage. In contrast in the cooking process followed by vacuum packaging and stor- lowest aw sample it rose to 25 |4g/g. Sour spoilage odors

JOURNAL OF FOOD PROTECTION, VOL. 52. DECEMBER 1989 EXTENDING SHELF-LIFE OF PORK LIVER PATE 883

TABLE 2. Additions of water to pate mix and consequent effects TREATMENT on a , drip and microbial counts. Values after 28 d at 4°C 1% GLYCEROL Water added a drip APC TAC (g/kg) (g/pack) (Log,(/g) (Log](/g) 0 0.938 0 2.14 5.96 CONTROL 215 0.958 0.30 2.69 6.83 430 0.960 1.56 3.84 6.89 645 0.964 2.14 3.93 7.53 0.5% GDL 860 0.982 3.49 4.57 7.74 Downloaded from http://meridian.allenpress.com/jfp/article-pdf/52/12/881/1654176/0362-028x-52_12_881.pdf by guest on 23 September 2021

TIME (DAYS) Figure 3. Effect of delta-gluconolactone (0.5% w/v) and glycerol (1% w/v) on the total anaerobic count (TAC) of pate stored at 4°C.

duction in pH of 0.6 units, to a value of 5.77 ± 0.04, meaned over the 30 d storage period. The a , 0.96 in the control, fell to 0.94 and 0.93 with 0.1 and 1.0% glycerol respectively, while DGL (0.1 % and 0.5%) caused respective reductions to 0.95 and 0.93. TIME (DAYS) Pates prepared with glycerol gave sour odors on open ing the packs after 28 d and rapidly turned green. Again Figure 2. Growth of TAC at different vlaues of aw during stor Lactobacillus viridescens was isolated. No yeasts and molds, age at 4 °C. Anaerobic bacteria had highest survival at an aw of 0.96. S. aureus, coliforms, or B. thermosphacta were detected in the control or glycerol containing pates. Pates containing DGL did not exhibit greening or souring. With the lower level of DGL no yeasts and molds, S. aureus, coliforms, or B. thermosphacta were detected. were detected after 28 d in samples with an aw of 0.982. Samples at the two highest a values turned green on With the higher level of DGL none of the organisms noted 2 exposure to air. Of 20 anaerobes selected at random from above were found but after 16 d, 10 yeasts/g were found, 4 the TAC plates of these samples 16 were identified as increasing to 2.51 X 10 yeasts/g at 28 d. The mean gen Lactobacillus viridescens. No coliforms, S. aureus, or yeasts eration time of the yeasts was 36 h and growth was in the 6 and molds were detected. exponential phase at 28 d. APC were below 10 in all To give the required shelf life, 215 ml/kg of added cases. water was chosen. This value also showed low drip loss on storage (Table 2) and drip in packs was reported as un DISCUSSION sightly by retailers test marketing the pate. The effect of humectants on extending the shelf life Prior to cooking, pork liver carried the most contami were then studied. Glycerol (0.1 and 1% w/w) and delta- nation (Table 1), representing about 95% of the APC. Since gluconolactone (DGL) (0.1 and 0.5% w/w) were selected the APC of liver rapidly increases unless properly cooled and the upper concentrations determined as the maximum (8,15) good manufacturing practice could reduce this load acceptable to the sensory assessment panel. DGL also acts significantly. Local livers were reported with an APC (Log10 as an acidulant and caused a sour taint at 1% (w/w). CFU/g) of 4.21 (15) and maintaining these levels would Only DGL at 0.5% (w/w) increased the shelf life have a marked effect on the microbial quality of the pate. (P<0.05) as assessed by TAC (Fig. 3). It also caused a re- However, since the microbial contamination of liver

JOURNAL OF FOOD PROTECTION. VOL. 52, DECEMBER 1989 884 MADDEN is largely confined to its surface (6) the blanching proce yeasts was seen for the first time during this study. The dure should significantly reduce numbers. reduced a and pH, coupled with the reduction in growth The combination of low temperatures and low oxygen of the competing bacteria, may have allowed the yeasts to tension in the product kept the APC low during all of the proliferate. Therefore, care in deciding the final aw would 6 studies. Since the APC was always below 10 /g no prob be required to ensure bacterial spoilage was not simply lems with aerobic bacteria occurred, and hence, aerobic replaced by spoilage by yeasts. spoilage can be discounted during storage. Overall, the major factors affecting the shelf life of Spoilage always occurred due to souring or greening vacuum-packaged pork liver pate were studied. The prod and both of these activities are attributable to lactic acid uct usually spoiled due to souring or greening caused by bacteria, enumerated in the TAC. lactic acid bacteria. Microbial control was best exercised The pate studied was similar in nature to luncheon by manipulation of the aw, either by regulation of the water meat, being a cured, comminuted meat product given a included in the mix or by the addition of a humectant and mild heat treatment resulting in a perishable product. Stud it also reduced the pH with consequent benefits in shelf- ies with raw luncheon meat (2) showed a mean APC of life extension. Downloaded from http://meridian.allenpress.com/jfp/article-pdf/52/12/881/1654176/0362-028x-52_12_881.pdf by guest on 23 September 2021 6 6 1.11 x 10 /g compared with an estimated 2.37 x 10 /g for GDL had no adverse effects on flavor or texture at the the pate studied. Counts of Lactobacillus (Log1()/g) after level used (0.5%), as assessed by a sensory analysis panel. cooking the luncheon meat (2) were 2.78 while a TAC of 2.59 + 0.72 (n=16) was found in our study. Hence the REFERENCES numbers after cooking are similar, despite the luncheon meat spending 89 min in a water bath at 68.5°C, and the 1. Baird, K. J., and J. T. Patterson. 1980. An evaluation of media for the core temperature chosen appears appropriate. cultivation or selective enumeration of lactic acid bacteria from vac uum-packaged beef. Record Agric. Res. 28:55-61. Lowering aw values in such products can reduce the 2. Bell, R. G. 1983. The effect of variation of thermal processing on the rate of microbial growth and the maximum number of microbial spoilage of chub-packed luncheon meat. J. Appl. Bact. bacteria (11). Although 3-fold changes in added water caused 54:249-255. 3. Chirife. J., O. C. Scorza, M. S. Vigo. M. H. Bertoni. and P, Callaneo. a change of only 0.006 in the aw (Table 4) reducing the aw 1979. Preliminary studies on the storage stability of intermediate from 0.967 to 0.961 prevented the spoilage of liver sausage moisture beef formulated with various water binding agents. 3. Food (12). Technol. 14:421-428. 4. D'Aoust, J-Y. U. T. Purvis, R. M. Coulter, and K. Weiss. 1985. Sal The greening seen in the high aw pates was due to monella in fresh pork liver and chicken liver in Canada: a 1979-1980 hydrogen peroxide production by L. viridescens when the survey. Can. Inst. Food Sci. Technol. J. 18:323-325. pate was exposed to air (18). It is therefore essential that 5. Gardner, G. A. 1966. A selective medium for the enumeration of a values below 0.964 are maintained. Microbacterium thermosphactum in meat and meat products. J. Appl. w Bact. 29:455-60. Glycerol has been used to increase the microbiologi 6. Gardner, G. A. 1971. A note on the aerobic microflora of fresh and cal stability of foods (3,9,11) but generally at higher levels frozen porcine liver stored at 4°C. i. Food Technol. 6:225-231. than in this study — about 12-30%. Since 1% was used, 7. Gardner, G. A. 1982. Microbiology of processing: bacon and ham pp. 129-178. In. M. H. Brown (ed.). Meat Microbiology. Applied Science. only minor improvements were expected. However, pates London. with glycerol soured more quickly than the controls. B. th- 8. Gill, C. O. 1984. Prevention of early spoilage of livers, pp. 240-241. ermosphacta can utilize glycerol as a carbon source (13) In Proceedings of the 30th European meeting of meat research work ers. Meat Research Institute, Bristol, UK. and can cause sour spoilage of meat products (7,8). But it 9. Karel. M. 1976. Technology and application of new intermediate was not detected in any pates during this study. moisture foods, pp. 4-31. In R. Davis. G. G. Birch, and K. J. Parker Studies on the taxonomy of lactic acid bacteria from (ed.). Intermediate moisture foods. Applied Science, London. vacuum-packed meats showed that 94% of cluster I (non- 10. Lawes Agricultural Trust. 1980. Genstat Numerical Algorithms Group, Oxford, UK. aciduric streptobacteria) isolates could ferment glycerol 11. Leistncr, L., and W. Rodel. 1976. Inhibition of microorganisms in (24). This property is rare in strains of lactobacilli (20). It food by water activity pp. 219-237. In F. A. Skinner and W. B. Hugo was also noted (21) that the group 1 organisms were more (ed.). Inhibition and inactivation of vegetative microbes. Academic resistant to high temperatures than the other lactobacilli Press. London. 12. Leistner, L.. and S. Karan-Djurdjic. 1970. Beeiflussung dcr Stabilitat studied. Therefore, it is probably that the use of glycerol von Fleishckonserven durch Stcuerung der Wasseraktivitat. Die Fleis- as a humectant simply provided a fermentation substrate chwirtschaft. 50:1547-1549. for the more thermotolerant lactobacilli, leading to rapid 13. Macaskic, L. E., A. G. Sheard, R. H. Dainty, and P. J. F. Henderson. souring. 1984. Glycerol utilisation by Brochothrix thermosphacta J. Appl. Bact. 56:137-143. GDL was much more suitable as a preservative. By 14. Madden, R. H. 1984. Pork liver pate — microbial considerations for combining the effects of an acidulant and a humectant, small scale production. Food Trade Review. 54:601-614. effective microbial control was seen (Fig. 3). The reduc 15. Madden, R. H., B. Hough, and C. W. Gillespie. 1986. Occurrence of tion in a to 0.93 moved the pate from the "highly perish Salmonella in porcine liver in Northern Ireland. J. Food. Prot. 49:893- w 894. able" (aw>0.95 and pH>5.2) category to "perishable" 16. Madden. R. H., and B. Moss. 1987. Extension of shelf life of minced (aw>0.95*or pH>5.2) (19). In Australia, fresh beef to be beef by storage in vacuum packages with carbon dioxide. J. Food vacuum packed must be below pH 5.8 to prevent rapid Prot. 50:229-233. microbial spoilage (27). 17. Patterson, J. T.. and P. A. Gibbs. 1979. Vacuum-packaging of bovine edible offal. Meat Science. 3:209-222. However in the pate with 0.5% GDL the growth of

JOURNAL OF FOOD PROTECTION, VOL. 52, DECEMBER 1989 EXTENDING SHELF-LIFE OF PORK LIVER PATE 885

18. Pivnick, H. 1980. Curing salts and related materials, pp. 136-159. In 23. Smibert, R. M., and N. R. Krieg. 1981. General Charactisation. pp. Microbial ecology of foods. Vol. 1. Factors affecting the life and 409-443. In P. Gerhardt (ed.). Manual of methods for general bacte death of microorganisms. ICMSF. Academic Press, London. riology. ASM. Washington, DC. 19. Rodel, W., H. Ponert, and L. Leistner. 1975. Einstufung von Fleisch- 24. Stanley, G., K. J. Shaw, and A. F. Egan. 1981. Volatile compounds erzeugnissen in leicht verderbliche, verderbliche und lagafahige Pro- associated with spoilage of vacuum-packaged sliced luncheon meat dukte. pp. 77-78. Proceedings of the 21st European Meeting of Meat by Brochothrix thermosphacta. Appl. Environ. Microbiol. 41:816- Research Workers, Bern, Switzerland. 818. 20. Rogosa, M. 1974. Genus Lactobacillus pp. 576-593. In R. E. 25. Sutherland, J. P., and A. Varnan. 1982. Fresh meat processing, pp. Buchanan and N. E. Gibbons (ed.). Bergeys Manual of Determinative 103-128. In M. H. Brown (ed.). Meat microbiology. Applied Science, Bacteriology (8th edn.). Williams and Wilkins, Baltimore. London. 21. Shaw, B. G., and C. D. Harding. 1984. A numerical taxonomic study 26. Swingler, G. R. 1982. Microbiology of meat industry by-products, pp. of lactic acid bacteria from vacuum-packed beef, pork, lamb, and 179-224. In M. H. Brown (ed.). Meat microbiology. Applied Science, bacon. J. Appl. Bact. 56:25-40. London. 22. Sinnell, H.-J., H. Klingbeil, and M. Benner. 1984. Microbiology of 27. Tarrant, P. V. 1981. The occurrence, causes and economic conse edible offal with particular reference to Salmonella. J. Food. Prot. quences of dark cutting in beef — a survey of current information, pp. 47:481-484. 3-36. In D. E. Hood and P. V. Tarrant (ed.). The problem of dark-

cutting in beef. Martinus Ninjhoff Publishers, The Hague. Downloaded from http://meridian.allenpress.com/jfp/article-pdf/52/12/881/1654176/0362-028x-52_12_881.pdf by guest on 23 September 2021

JOURNAL OF FOOD PROTECTION, VOL. 52, DECEMBER 1989