VARIATION IN CHANNEL CATFISH ICTALURUS-PUNCTATUS FLAVOR QUALITY AND ITS QUALITY-CONTROL IMPLICATIONS

Producers may capture two to three channel catfish Ictalurus punctatus L. for flavor analysis from ponds scheduled for harvest, If off-flavo rs are not present in several consecutive fish samples, the population may be considered acceptable for harvest. However, instrumental analy sis of the muddy/earthy off-flavor metabolites 2-methylisoborneol o-1, 2,7,7-tetramethyl-bicyclo-[2,2,1]-heptan-2-ol) (MIB) and geosmin (1 al pha, 10 beta-dimethyl-9 alpha-decalol) concentrations in approximately 80 catfish from each of three commercial ponds found both acceptable ton-flavor) and unacceptable (off-flavor) individuals within a single pond (i.e., mixed-flavor populations). Ascertaining the frequency of m ixed-flavor populations by instrumentally determining the off-flavor m etabolite concentration in muscle tissues from a large number of fish sampled from many ponds at several locations ol er time Is not current ly feasible. However, analysis of 12,725 commercial processor flavor a ssessments collected in 1994 and 1995 indicated 120 instances of indiv idual ponds yielding off-flavor fish follow ed by on-flavor fish in on e day or less. Reports indicate that fish require approximately four d ays to depurate off-flavors fully, suggesting that a proportion of the se rapidly changing flavor assessments may reflect the stochastic sele ction of fish from mixed-flavor populations rather than a complete and rapid conversion of the flavor quality of entire populations. Factors that contribute to the occurrence of mixed-flavor-populations have no t been identified fully. However, increased proportion of fat in catfi sh fillet tissues has been reported to he correlated with a greater re tention of MIB by fish, In this investigation, fish fat contents range d front 4.45% to 30.45%, and were positively correlated (P < 0.0001) w ith MIB concentrations, Additionally, the spatial distributions of alg al populations within certain commercial catfish ponds were not unifor m, and the sensory analysis of the flavor intensity of MIB and geosmin in catfish was more variable than the assessment of the intensity of chickeny and nutty flavors. Probability analysis indicated that in cer tain mixed-flavor populations there was a 10 to 25% probability of a s hipment of fish being rejected due to the random collection of only on -flavor fish prior to harvest followed by the post harvest capture of at least one off-flavor fish. In addition, a chance of no off-flavor f ish being sampled from a population containing a proportion of unaccep table fish was indicated, At current sample sizes, replacing sensory a nalysis with instrumental analysis would not completely avoid problems associated with sampling mixed-flavor-populations. Until effective me ans to reduce/avoid off-flavor metabolite accumulations in fish are wi dely available, careful attention to proper sensory evaluation protoco ls, an enhanced attention to pond conditions that affect flavor qualit y, the production of leaner more uniform populations of fish, and opti mizations of sampling strategies offer the most practical near-term au gmentations of current practice.