i I I I t I I I I 1 I I I

Weak -«-Intensity-Strong

Weak -«-Intensity-Strong

Figure 1. Composite (flavor/texture) sensory profile for weakfish (Cynoscion regalis) and cusk (Brosme brosme).

Figure 2. Tree diagram depicting the results of the cluster analysis of combined (flavor/texture) sensory data for 17 species of North Atlantic fish.

together because of their strong sensory similarities. Haddock, appearing opposite step 2, was next most similar to those paired in step 1 whereas white hake and whiting, grouped together, were next in line of sensory similarity (step 3). Hake and whiting were shown together because their profiles were more alike than either tilefish, pollock, or haddock. The amalgamation of all the species was continued until the cluster diagram was completed. The last few species (halibut, grouper, and swordfish) represent fish with the least similar sensory attributes and are weak linkages in the branching system. Figure 2 was divided into two major groups of fish. The first group consists of fish with low fat content, flavor intensity, and white flesh (tilefish, pollock, haddock, wolffish, cod, etc). The second major group includes species such as bluefish, mackerel, weakfish, and striped bass. This group is characterized by high fat content, high flavor intensity, and dark flesh. The third group was represented by swordfish. Within the two major groups of white flesh fish are two subgroups. These are represented by tilefish, pollock, haddock, wolffish, cod, cusk, monkfish, and flounder on one branch and hake and whiting on the adjacent branch.

After the 17 fish species were arranged by cluster analysis according to their characteristic sensory attributes, the original sensory data matrix was used to determine the "reasons" that the various species offish were grouped together through the -use of multidimensional unfolding

(148). With this technique, it was possible to visualize the relationships among stimulus objects (individual fish species) and sensory variables (the various flavor or texture attributes) by constructing a statistical map. The resulting map (Figure 3) has embedded within it both the sensory profile attributes as well as the 17 species of fish that were evaluated. Fish species in close proximity to one another are perceptually similar in texture, flavor, and appearance. The proximity of any fish species to any attribute point is an index of the perceived amount of that attribute in that fish species. For a more complete description of these methods and interpretation of the results, the reader is directed to the final report by NLABS summarizing the nomenclature project (140). For a more detailed explanation of the application and uses of cluster analysis, the text by Romesburg (147) is recommended.

After the initial phase of the work was completed, personnel at NLABS trained teams from three NMFS laboratories (Gloucester, Charleston, and Seattle) in their profiling methods. The NMFS laboratories began profiling commercially harvested fish species from their respective regions. Although the NMFS nomenclature program called for a comprehensive study of edibility characteristics, including chemical composition, instrumental color and texture measurements, and evaluation of sensory texture characteristics, we will deal with only the flavor attributes.

For the purposes of edibility profiling, the best-quality fish were used, usually 1 to 3 days postharvest; however, in some cases frozen fish fillets were used because they represented typical commercial product. Nevertheless, all samples used were carefully identified as to species, location, and time (date) of catch. Open consensus sensory panels (with 8-12 members) were run as described by Kapsalis (140) and Prell et al. (110). After scoring each sample, scores were tabulated in common and the results discussed to clarify individual profiles. Where there was disagreement, panelists were encouraged by the panel leader to reach a consensus. Sometimes this was difficult to achieve because of real organoleptic differences between samples of the same species. Each fish sample was replicated a minimum of 3 times during the course of this work. Therefore, the sensory profile of each species described here is a composite of data determined by the taste panel after discussion and agreement. All the combined flavor data describing each species of fish evaluated by NMFS laboratories were subjected to cluster analysis using the hierarchical method employing the BMDP-P2M "Cluster Analysis of Cases" computer program (149).

In all, 74 combined species of fish were analyzed. Results of the initial sensory evaluations are presented in the cluster analysis diagram in Figure 4 and were based on raw, unweighted data. Numbers appearing at the bottom of the diagram indicate the strength of relationships (amalgamations) based on similarities. Low numbers, or early amalgamation, indicate strong similarities between fish species and their flavor characteristics, whereas larger numbers represent fish that were quite dissimilar. From this cluster diagram, it is clear that complex relationships between fish species and their respective flavor profiles were found. Because of the large number of fish species represented in Figure 4 and the complex nature of their

Figure 3. Multidimensional unfolding of analysis (two-dimensional) of relationships among 17 species of North Atlantic fish and estimates of combined sensory (flavor/texture) attributes.

Sleeping Sanctuary

Sleeping Sanctuary

Salvation For The Sleep Deprived The Ultimate Guide To Sleeping, Napping, Resting And  Restoring Your Energy. Of the many things that we do just instinctively and do not give much  of a thought to, sleep is probably the most prominent one. Most of us sleep only because we have to. We sleep because we cannot stay awake all 24 hours in the day.

Get My Free Ebook

Post a comment