Nutritional traits of dorsal and ventral fillets from farmed European sea bass, gilthead sea bream and rainbow trout.

Quality differences due to fillet region have been quite extensively researched in farmed Atlantic salmon, but much less so in other important farmed species, namely European sea bass, gilthead sea bream and rainbow trout. The paucity of information on this issue is most likely due to the relatively small fillet size of these latter species, yet it deserves attention as a potential problem in sampling techniques in view of nutritional, storage and sensory analyses. Therefore this preliminary study was conducted to determine if and to what extent compositional differences exist between dorsal and ventral fillets from farmed European sea bass (Dicentrarchus labrax, ESB in the following), gilthead sea bream (Sparus aurata, GSB) and rainbow trout (Oncorhynchus mykiss, RT). Five specimens for each species were randomly selected from stocks of fish intensively reared in Italian commercial farms producing for the Italian market and ready for sale. Fish were weighed, eviscerated and filleted, each fillet being weighed with skin, then cut along the insertion line of the ribs to obtained a dorsal and a ventral fillet. After skinning, the two dorsal fillets obtained from each fish were joined, their sum being named “dorsal portion” (DP), and weighed. The same was made with the two ventral fillets, thus giving a “ventral portion” (VP). Both DP and VP obtained from each fish were analysed in duplicate for proximate and fatty acid composition, expressed as g/100 g flesh. Despite wide differences in body weight (mean  standard error: ESB = 226  20 g; GSB = 273  18 g; RT = 519  35 g), the three species did not differ as to the yield of DP (range: 22.11-22.80%) or VP (range: 16.08-18.20%), both calculated on body weight. ESB proximate composition proved to be the most affected by fillet location, followed by GSB, and then RT. Lipid content (g/100 g flesh) was by far the most variable trait (4.45 vs 8.58 vs 4.00 in DP, and 12.99 ab vs 14.43 a vs 6.62 b in VP, for ESB, GSB and RT, respectively, the ratio DP:VP being 1:2.92, 1:1.68 and 1:1.66 in that order). Species effect within location was examined to compare ESB, GSB and RT flesh lipid for both their content of health-related polyunsaturated fatty acids (PUFA) and their susceptibility to oxidation, evaluated through the Peroxidisability Index (PI). DP of GSB was significantly (and predictably, given its higher lipid content) richer in n-3 PUFA than ESB and RT (1.83 a vs 1.05 b vs 1.12 b, resp., as g/100 g flesh); GSB superiority was maintained, though only marginally, in VP (2.82 vs 2.68 vs 1.80, in the same order). The ratio between eicosapentaenoic (EPA, C20:5 n-3) and docosahexaenoic (DHA, C22:6 n-3) acids was higher in ESB than GSB and RT, both in DP (0.49 a vs 0.44 a vs 0.33 b, resp.) and in VP (0.59 a vs 0.49 b vs 0.34 c, resp.). Within each species n-6/n-3 was lower, i.e. more favourable, in DP than in VP, RT oil emerging as the most healthful in that respect. The higher n-3 PUFA (most notably DHA) percentage in RT oil was also responsible for its higher PI, both in DP (192 ab vs 170 b vs 227 a, for ESB, GSB and RT, resp.) and in VP (163 b vs 155 b vs 220 a, in the same order). On the grounds of the present results, though obtained on few specimens for each species, there seems to be ample reason to further investigate the effect of location on fillet compositional traits.