Fig 1.
Example of conifer breeding cycle for conventional forward selection with a generation interval of 17 years.
Fig 2.
Example of conifer breeding cycle for genomic selection with a generation interval of 9 years.
Fig 3.
Two deployment options: Deployment of progeny or clones of the selected trees.
Table 1.
Four selection schemes in breeding population, their combinations with two deployment options, generation interval, numbers of years needed from crossing parents to selection and from selection to deployment in the forest.
Table 2.
Five SNP panels were simulated with 600, 1,200, 2,500, 4,900 and 8,000 SNPs in one chromosome, which was equvalent to 7K, 14K, 30K, 60K and 90K SNPs.
Fig 4.
Plot of the squared correlations of allele frequencies (R2) versus distance in cM between closely located SNPs in a panel with 7K SNPs across 150 cM linkage map.
Fig 5.
Diagram of calculating total genetic gain from deployment of progeny of the selected individuals for 8 rotations for forward selection (FSs) and genomic selection (GSs).
Ri is the start of i rotation with 25 years long, Bi is the time when selection is made in breeding cycle i, Di is the time when genetic material is available for deployment in breeding cycle i, which is 5 years after selection in breeding cycle i. gf and gg are the genetic gain per generation from FSs and GSs, respectively.
Table 3.
Accuracies of GEBVs using a 60K SNP panel and a different number of individuals (800, 1,000, 2,000 and 3,000) in the training population for low (h2 = 0.2) and high (h2 = 0.5) heritability traits, compared with the accuracy of equivalent EBVs.
Table 4.
Accuracies of GEBVs for different numbers of makers in different sizes of SNP panel: 7K, 14K, 30K, 60K or 90K SNPs used in the prediction population for low (h2 = 0.2) and high (h2 = 0.5) heritabilities, compared with the accuracy of equivalent EBVs.
Marker effects were estimated using 1,000 clones in the training population.
Table 5.
Genetic gain per generation obtained in the prediction population from GS with 60K SNP panel for different training population sizes (800, 1,000, 2,000 and 3,000 clones) for traits with low (h2 = 0.2) and high (h2 = 0.5) heritabilities in the breeding population, compared with genetic gains obtained from FS.
Table 6.
Genetic gain per year obtained in the breeding population from genomic selection with (GSTG) or without (GS) top grafting and from forward selection with (FSCA) and without (FS) clonal archive for traits with low (h2 = 0.2) or high (h2 = 0.5) heritabilities for various training population sizes of 800, 1,000, 2,000 and 3,000. GEBVs were estimated using a 60K SNP panel.
Fig 6.
Additional genetic gain per year from FSCA when deployed as seedlings (FSCAs), or clones (FSCAc), when compared with deployment of selected individuals from FS via seedlings (FSs).
Table 7.
Additional genetic gains (percentage) across 8 rotations obtained from deploying genetic material selected from GS over a baseline of deploying progeny of the selected individuals from FS (FSs) for different deployment options: Deploying progeny (seedlings) of the individuals selected from GS (GSs), deploying clones of the individuals selected from GS (GSc), deploying progeny of the individuals selected through GSTG (GSTGs) or deploying clones of the individuals selected through GSTG (GSTGc).