Linking changes to breeding objectives and genetic evaluation to genetic gain in New Zealand

• Published in: New forests Vol. 45; no. 3; pp. 439 - 448
• Main Author: Apiolaza, Luis A.
• Format: Journal Article
• Language: English
• Published: Dordrecht Springer Netherlands 01-05-2014; Springer Nature B.V
• Subjects: Biomedical and Life Sciences; Forestry; Forests; Genetic improvement; Genotype & phenotype; Life Sciences; Pinus radiata; Selective breeding; Statistical models; Trees; Genetic evaluation; Genotype by environment interaction; Early screening; Wood properties
• ISSN: 0169-4286; 1573-5095
• DOI: 10.1007/s11056-014-9425-9

This article uses the breeder’s equation, which predicts genetic gain in breeding programs, to frame a general discussion on breeding objectives, new phenotypic techniques for selection criteria and statistical models as applied to short rotation species. Short-rotation breeding programs are increasingly working on wood quality traits; however, we keep on treating them as if they were growth traits. Understanding tree-level patterns of variation can lead to alternative strategies for evaluation, analysis and inclusion in breeding objectives; which I describe in a pilot application in
Pinus radiata
in New Zealand. Finally I discuss the relationship between the breeder’s equation and formulations of linear mixed models, using genotype by environment interaction as example, to show the interplay between genetic evaluation and breeding strategies. There is tension between increasing complexity (and the implicitly promised flexibility), information recovery (as more parameters are poorly estimated) and computational demands. The latter can be tackled through much more computer power (a never-ending endeavor), exploiting features of the problem or moving back to a lower complexity level.