Effect of 6.0% polyphosphate treatment on microbial growth and composition of microbial flora of three representative seafoods were investigated. Polyphosphate treatment washed off microorganisms by 1 log in peeled shrimp (10⁷ to 10⁶ and 10⁶ to 10⁵) and cooked crab meat (10⁶ to 10⁵) and by 2 logs in red snapper fillet (10⁶ to 10⁴). Phosphate treatment more readily washed off Pseudomonas sp. and Moraxella sp. and Lactobacillus sp., Corynebacterium sp. and Acinetobacter sp. were the most difficult to wash off. Polyphosphate also raised pH from 7.1 to 7.5 in cooked shrimp, from 7.7 to 8.2 in cooked crab meat and from 6.6 to 7.5 in red snapper fillet. The alkaline pH of up to 9.0, however, had no effect on the growth rates of Pseudomonas I, II and III sp., A. putrifacien sp., Micrococcus sp., Moraxella sp., Flavobacterium- Cytophaga sp., Acinetobacter sp., and Lactobacillus sp. Microbial growth was inhibited immediately following phosphate treatment. The phosphate induced lag period was almost 4 days at 1.1°C for shrimp and crab meat but was not observed in red snapper fillet. The degree of inhibition by phosphate depended on the initial number and kind of bacteria, especially those belonging to genera Pseudomonas, Moraxella, Lactobacillus and Corynebacterium. If still present in phosphate-treated seafoods, Pseudomonas sp. and Moraxella sp. quickly outgrew all others at 1.1°C. Lactobacillus sp. and Corynebacterium sp. became the predominant microorganisms when Pseudomonas sp. and Moraxella sp. were absent. / Graduation date: 1983

Pacific oysters (Crassostrea gigas), Alaska pink salmon (Oncorynchus gorbuscha) roe and chum salmon (Oncorynchus keta) fillets were treated with aqueous ozone in both pilot plant and commercial settings to determine its effect on shelf-life and microbial changes. The microbial quality was analyzed by conducting pyschrotrophic and coliform plate counts on 3M petrifilm. Oxidative rancidity, pH and moisture were measured during the shelf-life study to determine the effects of ozone on quality. Concentrations of 0.5-1.3 ppm of ozone were applied for periods of 30 s, 1, 2 and 4 min at 5°C, 9°C and 15°C to determine an optimum ozone concentration and contact time that would decrease the seafood microbial load and increase shelf-life in the pilot plant studies. Oysters and roe were treated at 15°C and 5°C and fillets were treated at 9°C. There where only minor microbial differences between ozonated and non-ozonated samples. Treatment temperatures rather than treatment types affected the microbial load. The pilot plant experiments at 15°C and 5°C showed 1 log decrease in oysters and roe treated with aqueous ozone at variable concentrations. No increase in shelf-life was observed when salmon fillets were treated with aqueous ozone (1.3 ppm) for 2 min. Bacteria strains were isolated from treated (ozone and water) and control salmon fillet groups at 0, 5 and 10 days of storage and identified using the API20 NE system. The microbial change in the fillet flora did not differ between ozonated and non-ozonated treatment groups. Gram-positive bacteria were predominant in all groups at day 0, Pseudomonas flunrescens and P. putida were the predominant bacterial species found from all groups at days 5 and 10. Aqueous ozone (0.5-1.7 ppm) was applied in a commercial Ikura roe processing facility. A decrease in the microbial load was seen in the pre-processed samples which were ozonated with eggs in the skein. There were no differences in microbial loads from the non-ozonated and ozonated post-processed samples of individual eggs removed from the skein. / Graduation date: 2003

The validation and potential use of mathematical models to estimate the shelf-life of refrigerated food exposed to temperature abuse and basing such estimations on microbial growth was analyzed. Combined heat transfer, microbial growth models, and non-parametric statistical procedures formed a computer-based predictive tool to assess shelf-life and estimate the accuracy of the prediction. Experiments were carried out to assess the precision of the combined model parameters. The different situations analyzed considered stepwise fluctuations in environmental temperature and a change in package characteristic (size and packaging material). Computer simulations showed that even when the temperature abuse period constitutes a small fraction of the total exposure time (2%-3%), shelf-life can be highly affected (20%-30%). To analyze the precision of the combined model response, two sources of variation were considered, microbial growth and heat transfer parameters. First order, pseudo-zero order kinetics and Arrhenius model formed the basis for the microbial model. The accuracy of lag and exponential phase of microbial growth for a mixture of three microorganisms (P. fluorescens, S. aureus, and A. Iwoffi) was assessed using a nonparametric statistical procedure based on the bootstrap method. The activation energy (E [subscript a]) and the logarithm of the frequency factor (InK₀) were found to be 109±3.4 J/mole and 48.3±1.5 for the exponential phase of this microbial mixture. The values for the exponential phase were 152±4 J/mole and 64.0±1.7, respectively. These parameters together with experimental values for the overall heat transfer coefficient were used to analyze the precision of the model response. This precision was not affected by a change in environmental temperature and packaging characteristics and remained constant at ±1 day. Two different temperature abuse situations yield estimated shelf-life of 4.8±1 and 8.9±1 day, respectively. This result can not be generalized as it depends on the particular examples analyzed. / Graduation date: 1993