Figure 7.

Predictive ability for DMY All Cuts in each population within TP1 was estimated for genomic prediction models using genotypic and phenotypic data from four populations as a training set and the remaining fifth population as a test set. This was repeated five times with a different population as test set each time. Predictive ability was estimated using single trait (ST-CV1) and multi-trait (MT-CV2) genomic prediction models, with the MT-CV2 model further assessed by comparing use of a low correlated secondary trait (GS Spring) and a high correlated secondary trait (GS All Cuts).