Efficiency of a Tactical Phenotyping Strategy for Multi-Stage Selection

Abstract

Accurate genetic evaluation relies on measurements, which can be difficult to achieve for some economical important traits (hard and/or costly to measure). We developed a strategy that can select an optimised subset of animals to phenotype based on pedigree relationship, prior information (previously phenotyped animals) and diversity to maximise genetic gain under inbreeding and cost constraints. We simulated a two-stage two-trait selection scenario for a small population of 10 paternal half-sib families of size IO (pilot study). One trait was phenotyped for all animals (parents and selection candidates) and the second trait was phenotyped on only a selected set of 20 selection candidates based on a prior decision on phenotyping (stage 1). Phenotyping decisions were made either based on maximizing diversity of the set chosen to be phenotyped (DIVERSITY) or simply based on breeding values at stage 1 (MERIT). After phenotyping, the second stage selection of animals as parents for the next generation was based on optimum contributions. The DIVERSITY strategy was most useful when there was limited prior information about the Mendelian sampling term of predicted breeding value. When parents of selection candidates have not been phenotyped, DIVERSITY does not provide any advantage over truncation selection (MERIT). However, when sires or both parents have been previously phenotyped, DIVERSITY resulted in higher genetic gain for similar level of inbreeding. From this study, we conclude that an optimized phenotyping strategy can have potential long term benefits in breeding programs but more work is needed to investigate under which conditions benefits are largest.