Introduction
The global and regional declines of wild stocks of Atlantic cod (Gadus morhua) prompted for the establishment of The National Cod Breeding program in Norway in 2003. The main selection criterium has been rapid growth, specifically body weight at 2+ age, after approximately 22 months of sea-rearing. It is well established that body weight is highly genetically correlated with gutted weight and filet weight, e.g., in Nile tilapia, rainbow trout and gilthead seabream (Gjerde et al. 2012; Kause et al. 2002; Navarro et al. 2009). This relationship has not been established in Atlantic cod.
Atlantic cod is a lean fish. Skeletal muscles contain only minute amounts of fat, whereas large amounts of fat can accumulate in liver. Enlarged liver has been frequently reported in practical cod aquaculture. This can be a consequence of unlimited access to feed that is not optimal or optimally utilized (Ingebrigtsen et al. 2014), but genetics may also play a role. In addition to the unfavourable distribution of ingested energy, large livers may be a welfare issue in aquaculture cod.
In this study we analysed slaughter data of Atlantic cod to get deeper insight of the genetic variation in body weight and carcass composition traits and assess the magnitude of the genetic and phenotypic interrelationships between these traits.
Material and methods
Data comprised records from 2557 individually tagged Atlantic cod from 218 full-sib families (158 sires and 134 dams) representing the fourth generation of selectively bred cod of National Cod Breeding Program. A priori power calculations (Kettunen and Lillehammer, 2019) were conducted to optimize the experimental design to obtain reliable estimates of heritability and genetic correlations. Fish were slaughtered at 948 days age from hatching and individual records for body weight traits (round RWT, gutted GWT, headed-and-gutted HGWT), body length (LENGTH), liver weight (LIVERWT), intestine weight (INTWT) and head weight (HEADWT) were registered. Hepatosomatic index (HSI), HEAD% and INT% were calculated as a percentage of LIVERWT, HEADWT and INTWT of RWT, respectively.
Estimates of heritability were calculated from variance components estimated with a univariate animal model using ASReml (Gilmour et al. 2015):
,
where, is the vector of the phenotypic observations; is the vector of fixed effect of sex; is the vector of random family effects; is the vector of random additive genetic effects; is the vector of random residual effects. Bivariate animal model was used for estimation of phenotypic and genetic correlations.
We also calculated conditional heritability for LENGTH, carcass weight traits (LIVERWT, INTWT, HEADWT) and HSI to examine the magnitude of genetic variation displayed in these traits independent from the genetic variation in RWT (Kause et al. 2002): ).
Results
Moderate estimates of heritability, ranging from 0.27 to 0.29 (±0.07), were obtained for RWT, GWT and HGWT from a univariate model. Slightly higher heritability was estimated for LIVERWT (0.32±0.08) and HEADWT (0.39±0.08), whereas LENGTH, INTWT, and INT% expressed relatively low heritability: 0.15-0.19 (±0.06-0.07). HSI and HEAD% were shown to be highly heritable traits (0.47±0.09 and 0.43±0.05). Proportion of family variance from phenotypic variance varied between 0.02-0.08. Conditional estimates of heritability for LENGTH, LIVERWT, INTWT and HEADWT (0.06-0.11) indicated that although a significant proportion of the genetic variation is dependent on RWT, this dependency is not complete. The conditional heritability for HSI was identical with the original estimate: 0.46.
Genetic correlations between RWT and GWT/HGWT were close to unity (0.98). RWT was highly genetically correlated with LENGTH, LIVERWT, INTWT and HEADWT (0.75-0.87). LIVERWT, INTWT and HEADWT were strongly genetically intercorrelated (0.54-0.78). Low phenotypic (0.30±0.03) and non-significant genetic correlation (0.15±0.17) was estimated between RWT and HSI. Negative genetic interrelationships, with high standard errors of the estimates, were estimated between RWT and HEAD%
(-0.29±0.10) and INT% (-0.48±0.20).
Discussion
As RWT, GWT and HGWT are genetically same trait, selection for RWT will result in feasible genetic change also in the valuable proportion of carcass in Atlantic cod. The selection for increased body weight at 2+ will increase the absolute weight for intestine, head, and liver. In contrast, selection for RWT will genetically reduce HEAD% and INT%, albeit these changes are expected to be minimal. HSI is not genetically strongly affected by selection for RWT, as indicated by non-significant genetic correlation between RWT and HSI. This observation is also supported by the high conditional heritability for HSI. Given high (moderate) heritability and moderate (high) phenotypic variation of HSI, CV=13%, (LIVERWT, CV=26%) selection for smaller liver is expected to be successful. This said, holistic understanding of the underlying biological reason(s) for frequently reported large liver in cod aquaculture is imperative before pursuing selective breeding against large liver size.
Acknowledgements
The staff of the Centre for Marine Aquaculture is warmly thanked for acquiring the data.
References
Gilmour, A. G., Gogel, B. J., Cullis, B. R., Welham, S. J., Thomson, R. 2015. ASReml User Guide Release 4.1. Structural Specification, VSN International Ltd, Hemel Hempstead, HP1 1ES, UK www.vsni.co.uk
Gjerde, B., Mengistu, S.B., Ødegård, J., Johansen, H., Altamirano, D.S. 2012. Quantitative genetics of body weight, fillet weight and fillet yield in Nile tilapia (Oreochromis niloticus). Aquaculture 342-343: 117-124.
Ingebrigtsen, I.A., Berge, G.M., Ruyter, B., Kjær, M.A., Mørkøre, T., Sørensen, M., Gjøen, T. 2014. Growth and quality of Atlantic Cod (Gadus Morhua) fed with high and low fat diets supplemented with glutamate. Aquaculture 433: 367–76.
Kause, A., Ritola, O., Paananen, T., Mäntysaari, E., Eskelinen, U. 2002. Coupling body weight and its composition: a quantitative genetic analysis in rainbow trout. Aquaculture 211: 65-79.
Kettunen, A., Lillehammer, M., 2019. Power calculations for optimisation of the experimental design to detect GxE: salmon experiments in FutureEUAqua. Oral presentation in Aquaculture Europe 2019, Montpellier, 9th of October, 2019.
Navarro, A., Zamorano, M.J., Hildebrandt, S., Ginés, R., Aguilera, C., Afonso, J.M. 2009. Estimates of heritability and genetic correlations for growth and carcass traits in gilthead seabream (Sparus auratus L.), under industrial conditions. Aquaculture 289: 225-230.