From X-Rays to the Palate: Mapping Fat, Lean, and Bone in Kingfish Using CT and Sensory Analysis

Dino Milotic

doi:10.60867/00000113

Aquaculture of fish is expanding rapidly to provide high quality nutrition to the world population and to off-set humanity’s dependence on wild-catch. Not all species are suitable for aquaculture. Yellowtail kingfish (Seriola lalandi, YTK) are growing in popularity as a production species in the Australian, Chilean and Japanese aquaculture industry as this species targets high-value, premium grade sashimi markets. Novel methods to phenotype quality traits are being investigated in the aquaculture industry, particularly non-invasive measures of meat quality traits such as fat content and lean meat yield. While the incorporation of non-invasive chain-speed measures of carcase composition are currently used extensively within the red meat industry, such technologies are in their infancy within the aquaculture industry. Here, a computed tomography scanner (CT) was used to scan YTK at various machine settings, including voltage (100 kV, 120 kV, 135 kV), field of view (FoV) (500 mm and 320 mm) and slice width (0.5 mm, 1 mm, 3 mm and 4 mm). Two tissue allocation methods were also tested: the thresholding approach and a factor analysis. Chemically extracted body fat, protein and fillet fat % was predicted by CT-fat% and CT-protein% from whole, bulleted and skinless fillet scans. A lower voltage, 100 kV, bigger FoV of 500 mm, and smaller slice width, 0.5mm, was optimal when using the thresholding tissue allocation method. The factor analysis outperformed the thresholding approach with an optimal FoV scanning setting of 500 mm, while voltage and slice width had minimal impact. Including whole fish weight into prediction equations greatly improved the thresholding approach to tissue allocation, but did not affect the factor analysis. These findings have implications for accurate YTK phenotyping, which forms the basis for breeding programs. The allometry of body fat, lean and bone tissues, as measured by CT, was analysed through the grow-out period in 259 farmed YTK. The 259 YTK consisted of 4 sampling groups of 60 YTK, sampled at average tank weight of 600g, 1000g, 1500g and 4000g, fed four different diets. Body fat, lean and bone weight was determined through CT scanning whole fish and bulleted fish. Fillet fat was determined by NIR and chemical fat extraction. Lean meat yield traits were also determined, including the fillet weight relative to bullet weight, fillet weight relative to whole carcase weights, and bullet weight relative to whole body weight. Allometric equations were used to describe the development of these traits, following the function of log(y) = log(a) + b * log(x) or y = a + xb, where y is a trait of interest, x is body weight, ‘b’ is the rate of development of y relative x, and ‘a’ describes the scaling coefficient. The effect of sex (male or female) on the phenotyped traits was tested. The ‘b’ coefficients of body fat, lean and bone differed between fish examined in the entire grow-out period and those examined at in sampling group 4 only. In the entire grow-out period (600g to 3000g) body lean and bone developed isometrically (b = 1), while body fat and fillet fat had late maturation (b > 1). However, when examining YTK at commercial weight (3000g, sampling group 4), it was found that body fat was late maturing (b > 1), while lean and bone were slightly early maturing (b < 1). Moreover, at the same whole fish weight, YTK fed on certain diets had up to 30% more fillet fat weight. Male YTK had 12% more fillet fat weight than females. The effect of fillet fat % on consumer sensory evaluation scores of YTK fillets at market size (>3000g) was evaluated. The fillets of 42 YTK were separated into the shoulder and belly portion, which were both further split into 7 cuts, shoulder-front (SF), shouldermiddle (SM), shoulder-back (SB), belly-front (BF), belly-middle (BM), belly-back (BB) and tail. Cuts from the left fillet were used for chemical fat extraction analysis and NIR-derived fat analysis, while cuts from the right fillet underwent consumer sensory testing. From each cut, 10 slices (7 mm) were obtained, which were lightly seared, and tasted by untrained consumers to obtain sensory scores for tenderness, juiciness, flavour liking, smell and overall liking on a scale of 0 (worst) to 100 (best). Examining differences in individual cuts showed that in the shoulder portion the SF and SM had similar scores, but SF tended to score higher than SB. In the belly portion, BF and BB had similar scores (P>0.05), while BM tended to have better scores (P<0.05). Fillet fat % of the SM cut was better at predicting sensory scores of all cuts than fillet fat % of the individual cuts. Overall liking, flavour and juiciness were improved by SM fat % increase. Whole fish weight had a positive effect on all sensory scores. Consumer sensory scores were grouped into Taste, Mouthfeel and Olfactory factors, where whole fish weight had a positive impact on all factors, but SM fat % only enhanced Taste. In conclusion, we found a significant effect of cut, whole fish weight, and fillet fat % on consumer sensory scores, where scores were improved in heavier and fattier fish.

From X-Rays to the Palate: Mapping Fat, Lean, and Bone in Kingfish Using CT and Sensory Analysis

Files and links (1)

Abstract

Details

Metrics