Chemical and Microbiological Analysis of Vacuum-packed, Pasteurized Flaked Imitation Crabmeat THOMAS A. HOLLINGWORTH, JR., CHARLES A. KAYSNER, KAREN G. COLBURN, JOHN J. SULLIVAN, CARLOS ABEYTA, JR., KEVIN D. WALKER, JAMES D. TORKELSON, JR., HAROLD R. THROM, and MARLEEN M. WEKELL ABSTRACT Various procedureswere used to attempt to correlate spoilage of im- itation crabmeat with sensory analysis after storage at 4°C 10°C and 22°C. Total volatile acids, total volatile bases, cadaverine, putrescine, histamine, aerobic plate count, and proteolytic count correlated well with product spoilage at 22°C. Even though 22°C is considered abu- sive, the only pathogenic organism isolated was Bacillus cereus. In contrast, at 4°C and lO”C, neither the chemical nor microbial indi- cators were adequateto assess quality of the product. Thus sensory analysis, despite its limitations, remained the only method currently available to assess product acceptability after prolonged storage at reduced temperatures. INTRODUCTION FLAKED IMITATION CRABMEAT is currently one of the most prevalent surimi-based fabricated seafoodproducts in the marketplace. For this reason, evaluation of chemical and mi- crobiological indicators of decomposition for applicability to such fabricated seafood products is of considerable interest. For example, Elliot (1987) reported a study assessing micro- biological quality of Alaskan pollock surimi. Yoon et al. (1988) studied both microbiological and chemical changesof surimi- based imitation crabmeat during storage and concluded that total volatile acids (TVA) was the most promising chemical indicator of those tested [TVA, total volatile bases (TVB), butanediol, and ethanol]. A time/temperature abuse study on vacuum-packaged, pasteurized flaked imitation crabmeat (commercially repacked) was recently reported from our lab- oratory (Hollingworth et al., 1990). The common chemical indicators of decomposition for seafood products, TVA and TVB, as well as, cadaverine, putrescine and histamine (Mietz and Karmas, 1977, 1978; Staruszkiewicz and Bond, 1981; Farn and Sims, 1987; Taylor and Sumner, 1987); and the microbiological indicators, aerobic plate count (APC), proteo- lytic count (PC), and yeast and mold count (YMC) were eval- uated. The chemical indicators TVA and TVB had the best potential as indicators of decomposition for this type product at the three storage temperatures(4”C, 10°C and 22°C) eval- uated. Commercially prepared, vacuum-packed, pasteurizedflaked imitation crabmeat (not repacked as in the previous study) would be expectedto be more stable (longer shelf life) than repacked product. Consequently, we did not know if chemical and mi- crobiological indicators of decomposition would be applicable to product subjected to time/temperatureabuse without atmos- pheric exposure after initial packaging. The purpose of this study was to evaluate the same chemical and microbiological indicators evaluated in our earlier study (TVA, TVB, cadav- Authors Hollingworth, Kaysnar, Colburn, Abeyta, Walker, and Wekelf are with the Seafood Products Research Center, Food & Drug Administration, Bothell, WA 980413012. Author Sullivan, formerly with the Seafood Products Research Center, is now with Varian Associates, Walnut Creek, CA 94598. Authors Tor- kelson and Throm are with the Seattle District Laboratory, Food & Drug Administration, Bothell, WA 98041-3012. erine, putrescine, histamine, APC, PC, YMC) to assess their applicability to different product conditions in this study. MATERIALS & METHODS Flaked imitation crabmeat samples Vacuum-packed, pasteurized flaked imitation crabmeat, in 907g packages (pouches), was prepared commercially from raw surimi as previously described (Hollingworth et al., 1990). Frozen product was thentransported to the laboratory andstared at - 20°C. Sample treatment Samples were allowed to thaw at room temperature. After about 4 hr the sampleswere no longer frozen, but still cool to the touch. Prior to initiation of variable temperature studies, six pouches randomly selected, were evaluated by sensory analysis to insure they were class 1 (passable, good quality) (Hollingworth et al., 1990). Remaining pouches were-then placed.(designated as time 0) in thermostatically controlled incubators set at 4°C. 10°C and 22°C. Each uouch was I assigned a number indicating position in the incubator. A random number scheme was used to select and remove pouches from incu- bators at each sampling time. Contents of each were mixed by knead- ing by hand in the pouch, aseptically opened and portions analyzed. Sensory classification of sample Approximately 1OOg of samplewere examined for odor, flavor, and texture by a sensory panel of three to four qualified examiners and assigned to one of three sensory classes of decomposition [class 1 (passable), class 2 (slight, but definite decomposition) or class 3 (ad- vanced decomposition)] that are utilized by the Food & Drug Admin- istration (FDA) (Hollingworth et al., 1990). FDA personnelcan only be qualified as certified sensory examiners upon successful completion of the following requirements: partici- pation and excellent evaluation at FDA Sensory Training Workshops; successful and sustained performance on unknown quality assurance samples; frequent examination of a wide variety of seafood samples for a minimum of l-2 years. Chemical analysis Sample extractions. Sample extracts were prepared from eachpouch by combining a 50 g sample and 200 mL of extracting solution in a biender and blending at high speed for 2 min. One 50-g sample was extractedwith 100% methanol (HPLC grade) and a second 50 e. sam- ple was extractedwith 0.02N sulfuric acid (H,SO,). The extraci were stored at -20°C until time of analysis. Volatile acids and bases. Acid extracts were analyzed for total volatile bases(TVB) and total volatile acids (TVA) by flow injection analysis (FIA) following the proceduresof Wekell et al. (1987). Cadaverine, putrescine and histamine. Methanol extracts were analyzed for cadaverine, putrescine and histamine by high perform- ance liquid chromatography (HPLC). The identity of cadaverine and putrescine was confirmed by gas chromatography/mass spectrometry as previously described (Hollingworth et al., 1990). Microbiological analysis Sample preparation. Fifty grams of each sample were weighed aseptically into a sterile blender jar (Oster) and 450 mL sterile 1% NaCl diluent (pH 7.2) were added. Samples were blended at high speed for 2 min and subsequentdecimal dilutions were prepared in 164-JOURNAL OF FOOD SCIENCE-Volume 56, No. 1, 1991