Investigation of morphological predictors of fillet and carcass yield in European sea bass (Dicentrarchus labrax) for application in selective breeding
|Author(s)||Vandeputte Marc1, 2, Puledda Antonio3, 4, Tyran Anne Sophie5, Bestin Anastasia5, Coulombet Celine6, Bajek Aline6, Baldit Gwenaelle1, 2, Vergnet Alain2, Allal Francois3, Bugeon Jerome7, Haffray Pierrick5|
|Affiliation(s)||1 : Univ Paris Saclay, INRA, AgroParisTech, GABI, F-78350 Jouy En Josas, France.
2 : IFREMER, Chemin Maguelone, F-34250 Palavas Les Flots, France.
3 : IFREMER, MARBEC, UMR9190, Chemin Maguelone, F-34250 Palavas Les Flots, France.
4 : Univ Sassari, Dept Agr, Sect Anim Sci, Sassari, Italy.
5 : SYSAAF Sect Aquacole, Campus Beaulieu, F-35000 Rennes, France.
6 : Ecloserie Marine Gravelines Ichtus, F-59820 Gravelines, France.
7 : INRA, LPGP, F-35000 Rennes, France.
|Source||Aquaculture (0044-8486) (Elsevier Science Bv), 2017-03 , Vol. 470 , P. 40-49|
|WOS© Times Cited||17|
|Keyword(s)||Ultrasound tomography, Genetics, Heritability, Indirect selection, Processing yields|
|Abstract||Genetic parameters for carcass and fillet percentage were estimated in 760 European sea bass reared under commercial conditions and slaughtered at 573 days post fertilization (395 g mean body weight). Phenotyped fish were the offspring of 45 sires and 20 dams crossed in a factorial mating design. Pedigrees were re-constructed with 90.7% success using 12 microsatellites. The heritability of fillet yield was moderately low (0.21), while it was high for carcass yield (0.57). Both traits were poorly correlated (− 0.01 to 0.28) making space for their combined improvement. We investigated different predictors derived from measurement of surfaces on digital pictures and ultrasound measurements at several points of the body. The accuracy of the phenotypic prediction was rather low for fillet yield (r2 = 0.02–0.18), but higher for carcass yield (r2 = 0.27–0.41). However, genetic correlations of predictors with the traits to predict were reasonably high (up to 0.67 for fillet yield and 0.95 for carcass yield), thus allowing to consider them for performing indirect individual selection instead of sib selection. However, it was difficult to design a predictor that would simultaneously increase fillet yield and carcass yield because of contradicting effects of relative head size, an important component of the predictors which was positively correlated to carcass yield but not to fillet yield.
Statement of relevanceWe estimated phenotypic predictors for processing yields in the European sea basslane estimated their genetic variation and correlations with the traits to predict. This is important to be able to apply indirect selection for processing yields in this species. This showed that although the traits of interest were hardly correlated, it was not possible to find external predictors having a significant positive impacts on both traits (carcass and fillet yield) simultaneously. This highlights the need to study specifically these issues in different species and conditions, as the picture here is very different to the well studied case of rainbow trout for example.