Selection Methods in Forage Breeding: A Quantitative Appraisal

Abstract

ABSTRACT Forage breeding can be extraordinarily complex because of the number of species, perenniality, mode of reproduction, mating system, and a variable genetic correlation between spaced plants and sward plots. Aiming to compare eight forage breeding methods for direct selection gain and correlated response, different scenarios of trait heritability and genetic correlation have been evaluated based on deterministic equations of expected (theoretical) breeding accuracy applied to half‐sib progenies evaluated in spaced‐plant trails, sward‐plot trials, or both. Relative efficiency for each method is given in relation to individual selection. Methods differ most when heritability is lower than 0.3, which coincides with the majority of the situations met by forage breeders. Genetic gain of best linear unbiased prediction (BLUP)‐based methods is superior to all other methods for trait heritability lower than 0.3, independent of field‐trail plot methods, except for parental selection. Methods based on BLUP have also shown higher efficiency when the genetic correlation between spaced‐plant and sward‐plot trials evaluations of a trait is lower than 0.7 and indirect‐trait heritability is lower than 0.3. The choice of forage breeding method should include consideration of the mode of reproduction and the target‐trait heritability. The benefits of BLUP‐based selection methods should receive more serious consideration by forage breeders.