This paper describes the method of random regressions applied to the study of reaction norms for description of genotype by environment interaction (G * E). Currently, the main problem is how to define the environmental scale. Using the phenotypic herd average of the trait itself may lead to biased estimates, and underestimation of G * E, with non-random use of sire and small herd sizes. Using climate, weather and herd management variables would be preferable, but such data are often lacking. Extension of reaction norm models to multiple traits, multiple environments and heterogeneous residual variance is possible. Even though linear reaction norms are the first approximation, non-linear norms have been found. Linear reaction norms are limiting in that they can only result in quadratic variance functions. It was pointed out that scaling type of G * E for individual traits may lead to re-ranking G * E for the total merit index.

E. Strandberg

Proceedings of the World Congress on Genetics Applied to Livestock Production, Volume , , 25.05, 2006
Download Full PDF BibTEX Citation Endnote Citation Search the Proceedings

Creative Commons License
This work is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License.