Genomic prediction of bone strength in laying hens using different sources of information

Bone damage in laying hens remains a significant welfare concern in the egg industry. Breeding companies rely on selective cross-breeding of purebred birds to produce commercial hybrids, which farmers raise for table-egg production. Genomic prediction is a potential tool to improve bone quality in laying hens. Because commercial layers are crossbred and kept in different environments than pure lines, the question arises whether to use within-line purebred selection or whether to use crossbred data. While selection based on pure line data is common, achieving optimal bone strength in hybrids may require incorporating hybrid data to account for heterosis and housing-specific effects. This study aims to evaluate how combining pure line and hybrid data could affect the accuracy of breeding values for bone strength. Genotypes and phenotypes were available from two types of white hybrids (Bovans White and Lohmann Selected Leghorn Classic) housed in two housing systems (furnished cages and floor housing). This resulted in four hybrid-housing combinations (n ∼ 220 for each). Tibia strength and genotypes for pure breeding lines of White Leghorn (WL, n = 947) and Rhode Island Red (RIR, n = 924) were also included. Each of the hybrid-housing combinations and pure lines was fitted separately into (1) single-trait Genomic Best Linear Unbiased Prediction (GBLUP), then simultaneously via multitrait GBLUP, (2) within hybrids across housing, (3) across hybrids within housing, (4) across hybrids and housing, (5) the latter in combination with WL and/or RIR data. Including hybrid data slightly increased the accuracy of the genomic estimated breeding value (GEBV) of other hybrids, but not that of pure lines. Pure line data increased the GEBV accuracy of hybrids over and above that of combining hybrid information. Combining data from two pure lines improved the GEBV accuracy of both. In comparison to the combination of data across lines and/or houses, combining tibia strength and BW within-lines increased tibia strength GEBV accuracy. The maximum GEBV accuracy obtained for tibia strength ranged from 0.42 to 0.65 for hybrids and from 0.63 to 0.78 for pure lines. Further study is required to test whether modelling the interactions of genotype by environment could help to breed hybrids for specific housing systems.