Use of Laminated High and Low Density Polyethylene Flexible Packaging to Store Trout (Salmo Gairdneri) in a Modified Atmosphere

Transcription

1 645 Journal of Food Protection, Vol. 50, No. 8, Pages (August 1987) Copyright 11 International Association of Milk, Food and Environmental Sanitarians Use of Laminated High and Low Density Polyethylene Flexible Packaging to Store Trout (Salmo Gairdneri) in a Modified Atmosphere HAROLD J. BARNETT*, JIM W. CONRAD, and RICHARD W. NELSON U.S. Department of Commerce, NOAA, National Marine Fisheries Service, Northwest and Alaska Fisheries Center, Utilization Research Division, 2725 Montlake Boulevard East, Seattle, Washington (Received for publication August 8, 1986) ABSTRACT The effect on extension of refrigerated shelf life of boned trout, treated or not treated with a 2.3% potassium dip and packaged in laminated high/low density "semi-permeable" polyethylene bags in the presence of a C0 2 -enriched modified atmosphere (), was studied. The combination of packaging, refrigeration, and was effective in doubling the fresh shelf life of the trout product. Potassium limited bacterial growth but did not extend shelf life beyond what was obtained with alone. Changes in ph and ammonia concentration in the fish samples were not good quality indicators in this study. Although not detected by sensory evaluations, TBA analyses indicated the possible development of oxidative rancidity in the -held fish. Carbon dioxide content in the flesh of the trout appeared to be a function of C0 2 in the headspace of the bags which was reduced by about 50% during the first 20 d of storage. Based on changes in organoleptic characteristics, the quality of the raw trout, including the potassium -treated trout, was only marginally acceptable after 25 d of storage in the. However, the cooked product was rated favorably by a consumer type taste panel. The maximum post- refrigerated shelf life of the trout held under the experimental conditions of this study, including those treated with potassium, was about 1 week. Because of improved technology, over-production of farm-raised trout has become a real problem for domestic growers in the U.S. resulting in decreased profits and sometimes business failures (]). It has been suggested that extending the fresh shelf-life of commercial trout products beyond the 5 to 7 d generally obtained by the industry today would help growers avoid these potential problems by providing better service to existing markets and stimulating new markets. In recently published studies (2,3,17,20,21), it was re- 'Use of trade names in this publication does not imply endorsement by the National Marine Fisheries Service. ported that the fresh shelf-life of fishery products can be significantly extended when held in modified atmospheres (s). These results and other research on poultry and meat (19) suggest that s could also be used to preserve the fresh quality of farm-raised trout. Also, modified atmospheres have been successfully used in conjunction with "semi-permeable" type packaging by the poultry and pork industries (24). A recent development, "semi-permeable" films permit the controlled transport of certain gas components from s across the walls of the packaging material. Theoretically, this technique should discourage users from holding fresh fish products for excessively long periods in refrigerated storage because they would eventually spoil or develop off-odors rendering them unacceptable as the s in the packaging became increasingly atmospheric. In addition to s and improved pckaging materials, s (anti-microbial agents used extensively to preserve meat and poultry products) have also been used successfully in extending the shelf-life of fresh fish (4,9). Besides their antimicrobial properties, potassium and sorbic acid also delay growth and toxin production by Clostridium botulinum in red meats and poultry sausage (18). The experimental results reported here were based on a study to determine the effects of packaging, modified atmospheres, and treatment with potassium on extension of the fresh shelf-life of commercially grown trout and also to determine the cumulative effect of the treatments on their post- storage shelf-life, i.e., subsequent refrigerated storage in air after removal from the modified atmosphere package (P). TERIALS AND METHODS Samples and packaging Trout (Salmo gairdneri) samples used in this experiment were obtained from a local Northwest grower/processor. The trout (65 lb., each weighing between 8 to 10 oz) were cleaned, deboned, and chilled in ice before use.

2 646 BARNETT, CONRAD AND NELSON The trout were divided into three groups. The first group (control) was subdivided into two lots of approximately equal weight which were packaged in 1.5-mil polyethylene bags for storage. The second group was dipped for 1 min in a solution containing 2.3% (w/w) potassium (ph 6.03), drained for 2 min, and packed in 8 x 12 x 3-in. aluminum trays (about 7-8 fish/tray) covered with lids that had been perforated with six 3/16-in. holes to allow passage of air or gas. Each tray of fish was placed in a 14-in. by 20-in., 4-mil Fresh Vac 1 "semi-permeable-type" laminated high/low density polyethylene bag with an 0 2 and C0 2 transmission rate of 82 and 317 g/100 in. 2 /24 h at 25 C, respectively. The bags were evacuated to 25 in. of Hg and back-flushed once with a premix gas containing 80% C0 2 and 20% N 2 at 14 psi using CVP Systems (CVP Systems, Inc., 2518 Wisconsin Ave., Downers Grove, IL) Model 200 vacuum sealer. The third group of trout was packed in a similar manner as the second group but was not treated with potassium. All of the samples were held at 35 F. At each examination period, a tray of trout from the treatment groups and a bag containing the control were removed from storage and the fish evaluated subjectively for raw quality. Three fish were then randomly selected from each test group and analyzed, in duplicate, for microbiological and chemical changes. After obtaining microbial samples, the three fish from each test group were filleted. One-half of each fish was used in making the chemical analysis and the other half (plus additional fish as needed) was used for taste panel evaluations. Excess trout from each test group not used in the above examinations were returned to refrigerated storage. Trout held in s were replaced in their original packaging, but left unsealed. They were re-evaluated organoleptically after an additional week in storage. Chemical methods The composition of the gaseous components of the was measured by gas chromatography using a Carle GL8700 chromatograph. At each sampling, 1 ml of gas was withdrawn from each bag examined using a pressure lock sampling syringe and needle and injected into the gas chromatograph. Separation of the gas components was made by a Series/Bypass technique over dual Poropak-Molecular Sieve columns using helium carrier gas. The thermal conductivity detector as well as the column were operated at 96 C. Resolution of the chromatograph's output was made with a Sargent-Welch XKR recorder/integrator. The C0 2 content of the edible muscle of the fish was determined by blending 40 g of muscle in a ph 10.4 Tris buffer (0.05 M) and liberating the C0 2 with an acid buffer. Concentrations were read from a calibration curve prepared from millivolt readings obtained with an Orion carbon dioxide electrode attached to an Orion Model 801 ph meter (5). Ammonia (NH 3 ) was determined by the method of Ward et al. (23). Ten ml of a 0.1% peptone Tris buffered sample was brought to 50 ml with distilled water and millivolt readings obtained with an Orion ammonia electrode connected to an Orion 801 ph meter after adding 1 ml of 10TV NaOH to the buffered sample. Ammonia concentrations were determined from a standard curve. The trout were chemically analyzed for oxidative rancidity (TBA number) using the method described by Lemon (13). After extracting a sample of fish in a solution containing trichloroacetic acid, propyl gallate, and EDTA, the filtrate was reacted with TBA reagent by boiling. The absorbance of the cooled samples was read at 535 nm with a Hitachi model spectrophotometer. The ph of the trout flesh was measured by placing the tip of a Corning combination-type electrode directly on the surface of the exposed muscle and making the appropriate reading. Microbiological method Total aerobic bacterial plate counts were made by the method described by Tretsven (22). Using a sterilized metal template, swab samples were taken from a 2-cm 2 area of exposed belly wall sampled from approximately the same area of each fish examined. The tip of the swab was broken off into 100 ml of 0.1% cold peptone diluent and mixed thoroughly. Further serial dilutions in 0.1% peptone water were made before plating. Plate counts were made with a TPY agar (1.5% trypticase, 0.5% peptone, 0.5% yeast extract, 0.2% glucose, 0.5% NaCl, and 1.5% agar) and incubation was at 22 C (72 F) for 5 d. Sensory evaluations The trout were evaluated raw and cooked. Raw samples were evaluated for appearance, texture (determined tactily), odor, and color of the flesh. The cooked samples, baked in covered aluminum containers for 15 min at 350 F, were evaluated by a 5-member consumer-type taste panel for flavor, texture, and rancidity using a 5-point hedonic scale (10). A score of 2 or below indicated a product of unacceptable quality. Sensory tests were made on the trout both during and after their removal from the (post- storage). Statistical analysis Sensory test data were averaged to obtain mean scores and standard deviations and chemical test data to give mean scores. Data with unequal replications were analyzed for significance of differences by the Kruskal-Wallis rank sum test (25). The t-statistic (25) was used to test for significance of differences between paired samples. Correlations between flavor and sensory rancidity and ph and C0 2 dissolved in the fish muscle were determined by analysis of the data using the computer program, Statistical Package for Social Sciences (15). RESULTS AND DISCUSSION Microbiological measurements Results of the total aerobic bacterial plate counts (APC) made on the control and treated samples are presented in Fig. 1. The initial plate counts for the samples of 9.5 X 10 4 /2 cm 2 were somewhat higher than expected for such fresh fish. The results were attributed to the swab method used for estimating bacterial numbers on the surface of the fish in this experiment. If the method of choice for estimating bacterial numbers had been the analysis of excised flesh from the fish, the initial counts would likely have been lower (22j. The presence of a heavy layer of mucous on the skin of the fish, a normal condition for post-harvest fresh-water raised trout, also prevented efficient removal of the bacteria from the fish when rinsed before packaging and storing. Incipient spoilage, manifested in the uncooked fish by a slight sour odor, was evident in the belly cavity and on the skin of some of the control samples after 10 d of storage and, after 12 d, all of the control samples JOURNAL OF FOOD PROTECTION. VOL. 50, AUGUST 1987

3 STORING TROUT IN MODIFIED ATMOSPHERE " i p«- / i Storage time (days) I Modified / atmosphere Modified atmosphere K- i i i Figure 1. Total aerobic plate counts (APC) of control and packaged trout during storage at 35 F. APC for control, APC for -held trout, ± APC for -held trout treated with K-. Point marked with arrow represents a count greater than 4.5 x examined were considered organoleptically unacceptable because of overt spoilage odors. This was confirmed by the significant (P<0.05) increase in APC for the control samples between the 6th and 12th days of storage. Unlike the APC for the controls, which increased slightly from the initial counts during the first 6 d of the experiment, total plate counts made on the P trout were observed to decrease during the same period. This was followed by a general increase in bacterial numbers that remained slightly in excess of the initial counts until the 18th day of the experiment. At this time, the lag phase of the growth curve had increased 3-fold beyond that of the control. Between the 18th and the final days of storage, the APC for the P trout not treated with potassium increased significantly (P<0.05). Total aerobic plate counts for the potassium -treated trout remained essentially unchanged during this period indicating that in addition to the C0 2 -enriched, potassium also had an inhibitory effect on microbial growth. Chemical measurements Results of the chemical measurements are shown in Table 1. PH Although surface ph of the control samples increased slightly during storage and the surface ph of the P trout remained at or slightly depressed from the initial TABLE 1. Mean" chemical data for control and potassium -treated and untreated boned trout held in a C0 2 -modified atmosphere. Storage time (days) a Each number b Means in the =Significantly d Significantly "Means in the Sample treatment 0 3 -potassium 6 -potassium 10 -potassium potassium 18 -potasium 20 -potassium 25 -potassium ph 6.54 b Not tested Not tested 6.61 b 6.48 b 6.53 b 6.62 b 6.40" 6.26 b 6.44 b 6.51 b 6.52 b 6.52 b 6.36 b 6.40 b co 2 cone. (ppm) = 988= = 551= 751= 848= 481= 566= 487= 566= 177= d 170= d Chemical data NH 2 cone. (ppm) 287 b 381 b 278 b 168 b 210 b 276 b 256 b 254 b (Spoiled) 196 b 159 b 236 b 234 b 236 b 235 b 230 b 227 b represents the mean of 3 samples analyzed in duplicate. same column with a common superscript letter are not significantly different (P>0.05). different from the control at the 5% level of significance. different from the P trout at 20 d at the 5% level of significance. same column for the same day with a common superscript letter are not significantly different (P>0.05). TBA (xmoles Malonaldehyde/ _ 1.56= e 1.29= e 1.56= e 1.29= e 1.29=* 1.09= e

4 648 BARNETT, CONRAD AND NELSON 80 -S 70 o CO, WmmmZ mmssm *m Storage time (days) Figure 2. C0 2 and <9 2 concentration of headspace gases in semi-permeable bags as a function of storage time. ph (attributed to the effect of dissolved absorbed C0 2 on the surface of the flesh of the trout), the differences were not significant (P>0.05). Similar observations on ph changes in fish stored in C0 2 -enriched were reported previously by Coyne (6), Makashev (14), Fey and Regenstein (9), Lannelongue et al. (12), and Strasdine et al. (21). C0 2 concentration The concentration of C0 2 in the flesh of the P trout samples increased significantly (P<0.05) between the first (0 time) and sixth days of storage. The increase was attributed to C0 2 tension in the, as C0 2 is absorbed into the flesh of the fish, and to other chemical/ physical relationships with muscle protein, e.g., complex - ing of amino (NH 2 ) groups with C0 2 (19). The solubility of C0 2 in muscle proteins is dependent on several variables including temperature, moisture content, and concentration of C0 2. Regardless of the increase in absorbed C0 2, there was no correlation (r = 0.14) between ph and C0 2 concentration. The decrease (P<0.05) in the concentration of C0 2 in the flesh of the trout between storage days 20 and 25 is the result of reduced PC0 2 in the "semi-permeable" bags and the fish's own ability to buffer. It would seem that the lower concentration of C0 2 was not effective in controlling bacterial growth as evidenced by the increased APC on the trout samples not treated with potassium. These findings are supported by the results of the research by Brown et al. (3). Earlier work reported by King and Nagel (11) to study the inhibitory effect of C0 2 on growth of Pseudomonas aeurgniosa suggested that the inhibitory effect of C0 2 had to do less with ph than to the mass action effect of C0 2 on the metabolism of the organism. TBA analysis Storing the trout samples in semi-permeable bags with s (with or without potassium treatment) did not appear to suppress development of oxidative rancidity as measured by the TBA test (Table 1). The gradual increase in 0 2 in the headspace gas (Fig. 2) of the bags was seen as contributing to the apparent rancidity condition in the samples as manifested by the increase in TBA values. Brown et al. (3) observed similar changes in TBA values in salmon held in modified atmospheres. Although significant (at the 5% level) increases in TBA values occurred in the P trout between the 10 and 18 d of storage, indicating at least the potential for rancidity, taste panel evaluations (discussed later) did not detect the changes. TBA values were not determined on the control samples after the 10th day of storage because of spoilage. Differences in TBA numbers between the P trout from 18 through 25 d of storage were not significant (P>0.05). NH 3 concentration Concentrations of ammonia (NH 3 ) in the trout, as reported here, remained essentially unchanged in both the control and P trout samples throughout the experiment (Table 1). Because differences in NH 3 concentration between the sample treatments (control and P trout) were not significant (P>0.05), there was no correlation between NH 3 concentration and bacteria on the samples. The relatively low ph of the trout samples and low storage temperature probably combined to discourage growth of ammonia-forming organisms associated with the spoilage of fish, although spoilage odors other than ammonia were noted in the control samples near the end of the second week of storage. Headspace composition As expected, the composition of the headspace gas in the Fresh Vac bags changed during the experiment (Fig. 2). Of particular interest was the rate of reduction in C0 2, the active bacteriostatic component of the, and the simultaneous increase in 0 2. It is assumed that most of the reduction in C0 2 between the initial sampling and the 20th day of storage was the result of its transmission across the walls of the plastic bags and to a lesser degree due to the absorption of C0 2 into the flesh of the trout. There was no further reduction in C0 2 concentration after the 20th day of storage. In modified atmosphere preservation studies using barrier films, Lannelongue et al. (12) observed a reduction in C0 2 concentration of the headspace gases followed by a gradual increase in the C0 2 tension. The increase was thought to be caused by microbial respiration. Sensory evaluations Fresh raw trout. Sensory quality of the fresh, raw trout in all three sample groups was excellent through the first 6 d of the experiment (Table 2). Between the 6th and

5 STORING TROUT IN MODIFIED ATMOSPHERE 649 TABLE 2. Sensory" evaluations of potassium -treated and untreated boned trout held in a C0 2 -modified atmosphere at 35 F. Storage time (days) Sample treatment -potassium -potassium -potassium -potassium -potassium Odor 4.5 ± ± ± ± ± ± ± ± ± Mean b sensory scores for Flavor 4.5± ± ± ± ±0.19 spoiled 4.2± ± ± ± ±0.30 c 3.0±0.26 c baked trout samples Texture 4.5± ± ± ± ± ± ± ± ± ±0.30 Rancidity ± ± Subjective evaluation of raw trout Excellent quality " " Very good quality Excellent quality Slight sour odor on skin and in visceral cavity; cloudy eyes Very good quality, normal color, no off odor Overt spoilage odors present Very good quality, no off odors " Fair quality but with some loss of characteristic fresh trout odor Acceptable but poor quality, some offodors not associated with normal spoilage patterns Acceptable but poor quality, normal flesh color, loss of fresh trout odor, dull skin color "Sensory scale of 1-5 with 5 indicating a fish with very good flavor and odor, normal (firm) texture, and no rancidity. A score of 2.0 indicates a product of unacceptable quality. b Each number represents the mean of 5 samples ± the standard deviation. c Significantly different from the control at the 5% level of significance. 10th days of storage, some of the control samples examined showed evidence of incipient spoilage as manifested by slight but perceptible sour odors on the skin and in the belly cavities and accompanied by other sensory changes such as softening of the flesh and cloudy eyes. The control samples were overtly spoiled by the 12th day of storage. This was determined by the presence of strong spoilage odors rendering the samples unfit for tasting and verified by the APC shown in Fig. 1. By comparison, the fresh quality of the P trout was very good after 14 d of storage. However, by the 20th day, the -held trout began to show a definite loss of quality. These changes were mainly seen as the disappearance of characteristic grassy or seaweedy-like odors often associated with fresh fish and bright skin color. Otherwise the P trout were of acceptable (fair) quality. After 25 d, the fresh quality of these samples was considered marginal. By this time, their normal fresh, grassy odor had been replaced with a predominantly stale odor and they were uniformly dull in appearance. Subjectively, quality differences between the potassium- treated and untreated -stored trout were negligible. Cooked trout. Results of the taste panel evaluations of the cooked trout in Table 2 show that the control and -held samples were of good quality for the first 10 d of the experiment. However, as previously mentioned, evaluations of the uncooked control samples held in storage for 10 d revealed a very slight sour odor present on

6 650 BARNETT, CONRAD AND NELSON TABLE 3. Sensory" evaluations of -treated and untreated boned trout held post C0 2 -modified atmosphere storage for one week at 35 F. Storage time in (days) Sample treatment -potassium -potassium -potassium -potassium -potassium Odor 4.0 ±0 C 4.2±0.16 c 4.4±0.28 c 4.2±0.31 c 3.7±0.16 c 4.0 ±0 C 2.0±0.26 d 3.2±0.16 d Mean b sensory scores for baked trout samples Flavor 3.8±0.03 c 4.2±0.30 c 3.8±0.26 c 3.2±1.26 c 3.5±0.18 c 3.0±0.30 c 1.0±0 d 2.2±0.30 d Texture 4.0 ±0 C 4.5±0.24 c 4.6±0.17 c c 4.0±0.26 c 4.2±0.16 c 4.2±0.30 c 4.0±0.26 c Spoiled after one week of refrigerated storage at 35 F Spoiled after one week of refrigerated storage at 35 F Rancidity 4.2±0.16 c C C 3.7±1.5 C 4.2±0.30 c 3.0±1.8 C 1.7±1.15 d 2.0±1.15 d Subjective evaluation of raw trout Very good product, normal appearance and odor Good quality, normal appearance and odor Good quality, normal odor Acceptable product, normal appearance, loss of fresh odor Slight to moderate yellowing of skin and meat; slight sour odor on skin and in visceral cavity Stale odor, normal appearance "Sensory scale of 1-5 with 5 indicating a fish with very good flavor and odor, normal (firm) texture, and no rancidity. A sensory score of 2.0 indicates a product of unacceptable quality. b Each number represents the mean of 5 samples ± the standard deviation. c Means in the same column with a common superscript letter are not significantly different (P>0.05). d Means in the same column for the same day with a common superscript letter are not significantly different (P>0.05). products but not to the point of rejection. Differences in the skin and in the belly cavities of some of the fish indicative of incipient spoilage. This defect was not reflected in the mean odor and flavor scores for the cooked control samples. The apparent contradiction can be explained by the fact that fresh fish do not always undergo chemical and bacterial degradation at the same rate, especially in early spoilage when bacteria are usually confined mainly to surface areas. As a result, in any given lot of fish some will show subjective signs of spoilage and others will not, depending on the degree of spoilage. Often unpleasant odors can be washed from these fish rendering them temporarily acceptable. Thus, when these fish are cooked, bacterially developed odors are often volatilized on heating resulting in acceptable products. For this study, therefore, overt spoilage odors in which all of the fish in each lot were involved determined unacceptable spoilage. Because of the presence of overt spoilage odors on the 12th day of storage, the control samples were not taste tested. Between 10 and 20 d, the status of the cooked quality of the P trout remained "good-to-fair." Essentially, there were no differences in the sensory parameters evaluated between the two P groups during this period. By the 25th and last day of the experiment, losses (P<0.05) in cooked flavor characteristics were observed in both P groups. A condition of time, the changes diminished the acceptability of the the sensory parameters of cooked fish measured between the two P groups were not significant (P>0.05). Though there was an increase in TBA values (Table 1) with time in the P trout and a corresponding correlation (r = 0.96) between flavor and sensory rancidity, rancidity per se did not appear to be a factor in the taste panel analyses. It is possible that products of oxidation were not sufficiently strong to affect flavor. The correlation between flavor and bacterial counts (r = 0.96) for both treatment groups suggest chemical changes brought about by the bacteria produced negative flavor changes with time that complicated sensory analyses. These negative flavor changes could also have masked detectable rancidity flavors. Post- storage in air In addition to making sensory tests on the trout during storage, their raw and cooked sensory properties were also determined after they were removed from storage. Using the original Fresh Vac packaging, the trout were held unsealed for 1 week at 35 F and evaluated for sensory changes as previously described. The results in Table 3 show that trout can be stored for a maximum of 2 weeks in P followed by 1 week in refrigerated storage without seriously affecting their

7 STORING TROUT IN MODIFIED ATMOSPHERE 651 fresh quality characteristics. When cooked, however, the -treated samples had a slight "fishy" flavor. A high correlation (r = 0.99 for the potassium treated trout and r = 0.97 for the trout not treated with potassium ) between flavor and sensory rancidity scores suggests that oxidative rancidity was a major contributor to off flavors in the cooked products. Trout samples held for 20 d in the were of acceptable quality 3 to 4 d post- storage. However, after a full week of post- storage only the trout treated with were judged to be of acceptable, but only marginal, quality after cooking. By this time noticeable (P>0.05) changes were observed in the sensory parameters evaluated for each P treatment group. Trout held for 25 d in the, whether or not treated with potassium, were judged unfit for tasting because of obvious spoilage odors after 1 week of post- storage. These results suggest that the post-treatment storage life of fresh trout in air is greatly reduced when they are held in longer than 2 weeks. CONCLUSIONS The results of the work described here show that the refrigerated shelf-life of fresh trout can be effectively doubled, i.e., extended from 10 to 20 d, when held in "semi-permeable" type packaging in the presence of a C0 2 -enriched. Microbial growth was controlled and chemical and organoleptic changes were minimized by the improved packaging method. Treatment with before storage increased control of microbial growth but did not extend the fresh shelf-life of the trout beyond what was obtained with the alone. Limiting storage of the trout to 2 weeks in the permitted them to be held for an additional week outside the in a refrigerated environment with minimal loss of quality. Storing the trout in the beyond 2 weeks significantly reduced their post-modified atmosphere shelf-life. Because of the public health concern for the increase in the potential for botulism in fresh fish subjected to temperature abuse or prolonged refrigerated storage above 38 F when held in a, the commercial application of this packaging method in the United States has been delayed pending further research (7). REFERENCES 1. Anonymous A forecast: Aquaculture in the 1980's. Aquaculture Buyers Guide, pp Barnett, H. J., R. W. Nelson, P. J. Hunter, F. E. Stone, G. C. Roberts, and J. Kwok A study on the use of a high concentration of CO2 in a modified atmosphere to preserve fresh salmon. Mar. Fish. Rev. 44(3):7-ll. 3. Brown, W. D., M. Albright, D. A. Watts, B. Heyer, B. Spruce, and R. J. Price Modified atmosphere storage of rockfish (Sebastes minialus) and silver salmon (Oncorhynchus kisulch). J. Food Sci. 45: Chung, Y.-M., and J. S. Lee Potassium inhibition of microorganisms isolated from seafood. J. Food Prot. 45: Conrad, J. W., G. C. Roberts, and H. J. Barnett Manometric and electrode-probe determinations of CO2 in fish flesh. J. Food Sci. 45: Coyne, F. P The effect of carbon dioxide on bacterial growth with special reference to the preservation of fish. Part I. J. Soc. Chem. Ind. (Lond.) 51: Eklund, M. W Effect of C0 2 -modified atmospheres and vacuum packaging on Clostridium botulinum and spoilage organisms of fishery products, pp In R. Martin (ed.), Proceedings of First National Conference on Seafood Packaging and Shipping, National Fisheries Institute, Washington, DC. 8. Erickson, J. D American trout farming marks 100 years plus Still growing. Aquaculture 7(3): Fey, M. S., and J. M. Regenstein Extending shelf-life of fresh wet red hake and salmon using C02/0 2 -modified atmosphere and potassium ice at 1 C. J. Food Sci. 47: Hamilton, M., and R. Bennett An investigation into consumer preferences for nine fresh white fish species and sensory attributes which determine acceptability. J. Food Technol. 18: King, Jr., D. A., and C. W. Nagel Growth inhibition of a Pseudomonas by carbon dioxide. J. Food Sci. 32: Lannelongue, M., G. Finney, M. O. Hanna, R. Nickelson, and G. Vanderzant Storage characteristics of brown shrimp (Panaeus aztecus) stored in retail packages containing C02-enriched atmospheres. J. Food Sci. 47: , Lemon, D. W An improved TBA test for rancidity. Environment Canada. New Series Circular #51. Fisheries and Marine Service. New Ser. Circ. 51, 4 pp. 14. Makshev, A. P Carbon dioxide as a means of prolonging the storage life of chilled fish products. Trudy Vsesoyuz, Nauch- Issledovatel Inst. Morsk. Rybn. Khozzi Abeanogr. 37, 138 p. 15. Nie, N. H., C. H. Hull, J. G. Jenkins, K. Steinbrenner, and D. H. Bent Statistical package for the social sciences, 2nd ed. McGraw Hill, Inc., New York. 675 pp. 16. Ogrydziak, D. M., and Duane W. Brown Temperature effects in modified atmosphere storage of seafoods. Food Technol. 47(3): Parkin, K., M. J. Wells, and W. D. Brown Modified atmosphere storage of rockfish fillets. J. Food Sci. 47: Robach, M. C, and J. N. Sofos Use of s in meat products, fresh poultry, and poultry products: A review. J. Food Prot. 45: Silliker, J. H The efficacy of controlled atmospheres in extending the shelf life of meat, poultry, and fish. pp In R. Martin (ed.), Proceedings of the First National Conference on Modified and led Atmosphere Packaging of Seafood Products. National Fisheries Institute, Washington, DC. 20. Stier, R. F., L. Bell, K. A. Ito, B. D. Shafer, L. A. Browm, M. L. Seeger, B. H. Allen, M. N. Porcuna, and P. A. Lerke Effect of modified atmosphere storage on C. botulinum toxigenesis and the spoilage microflora of salmon fillets. J. Food Sci. 46: Strasdine, G. A., G. Mah, and D. Moteith Modified atmosphere storage Microbiology of cod fillets. Technical Report #7. B. C. Research, Division of Fisheries Technology, 3650 Wesbrook Mall, Vancouver, Canada V6S 2L2, 27 p. 22. Tretsven, W. I Bacteriological survey of filleting processes in the Pacific Northwest. II. Swab techniques for bacteriological sampling. J. Milk Food Technol. 26: Ward, D. R., G. Finne, and R. Nickelson III Use of a specific ion electrode (ammonia) in determining the quality of shrimp. J. Food Sci. 44: Wolfe, S. K Use of CO- and C0 2 -enriched atmospheres for meats, fish, and produce. Food Technol. 34(3):55-58, Zar, J. H Biostatistical analysis. Prentice-Hall, Inc. Englewood Cliffs, NJ.