Potential for Genetic Improvement of the Main Slaughter Yields in Common Carp With in vivo Morphological Predictors

Abstract

Common carp is a major aquaculture species worldwide, commonly sold alive but also as processed headless carcass or filets. However, recording of processing yields is impossible on live breeding candidates, and alternatives for genetic improvement are either sib selection based on slaughtered fish, or indirect selection on correlated traits recorded in vivo. Morphological predictors that can be measured on live fish and that correlate with real slaughter yields hence remain a possible alternative. To quantify the power of morphological predictors for genetic improvement of yields, we estimated genetic parameters of slaughter yields and various predictors in 3-year-old common carp reared communally under semi-intensive pond conditions. The experimental stock was established by a partial factorial design of 20 dams and 40 sires, and 1553 progenies were assigned to their parents using 12 microsatellites. Slaughter yields were highly heritable (h² = 0.46 for headless carcass yield, 0.50 for filet yield) and strongly genetically correlated with each other (r_g = 0.96). To create morphological predictors, external (phenotypes, 2D digitization) and internal measurements (ultrasound imagery) were recorded and combined by multiple linear regression to predict slaughter yields. The accuracy of the phenotypic prediction was high for headless carcass yield (R² = 0.63) and intermediate for filet yield (R² = 0.49). Interestingly, heritability of predicted slaughter yields (0.48-0.63) was higher than that of the real yields to predict, and had high genetic correlations with the real yields (r_g = 0.84-0.88). In addition, both predicted yields were highly phenotypically and genetically correlated with each other (0.95 for both), suggesting that using predicted headless carcass yield in a breeding program would be a good way to also improve filet yield. Besides, two individual predictors (P₁ and P₂) included in the prediction models and two simple internal measurements (E4 and E23) exhibited intermediate to high heritability estimates (h² = 0.34 - 0.72) and significant genetic correlations to the slaughter yields (r_g = |0.39 - 0.83|). The results show that there is a solid potential for genetic improvement of slaughter yields by selecting for predictor traits recorded on live breeding candidates of common carp.

Keywords: genetic correlations; heritability estimates; indirect selection; morphological landmarks; slaughter yields; ultrasound imagery.

FIGURE 1

Landmarks placed on each common carp photo. (1) Head extremity; (2) end of the head beginning of the filet on the back; (3) intersection between the back and the vertical of point 4; (4) intersection between opercula and lateral line; (5) opercula at the maximum length from the landmark 1; (6) end of the head beginning of the filet on the ventral part; (7) beginning of the dorsal fin; (8) intersection between the lateral line and the vertical of landmark 6; (9) intersection of the ventral part and the vertical of point; (10) beginning of the anal fin; (11) intersection between lateral line and vertical of point 9 toward the carp back; (12) vertical of point 10 on the back; (13) end of anal fin; (14) intersection of lateral line and vertical of 12; (15) vertical of point 12 on the carp back; (16) narrowest point on the caudal peduncle on the back; (17) intersection of the lateral line and vertical of 15; (18) vertical of point 16 on the ventral part (normally the narrowest point of the caudal peduncle; (19) end of the filet (on the skin) on the lateral line; (20) end of the caudal fin at the fork.

FIGURE 2

Graphical representation of body morphology for highest values (red lines) and lowest values (blue lines) of Logr_hl-Carss (A) and Logr_Fil (B).