Genotype-by-environment interactions may limit the selection efficiency of chickens intended for free-range systems.

In response to the growing demand for alternative poultry systems, such as free-range or organic farming, slow-growing chicken lines are preferred for their robustness and meat quality. However, genetic selection is usually conducted under indoor conditions, which may not reflect outdoor challenges and may decrease the effectiveness of selection if genotype × environment (G × E) interactions exist. This study assessed G × E interactions in a slow-growing meat chicken line, focusing on feed-use efficiency, growth, carcass composition, and meat quality. The study also compared direct selection under production conditions to indirect selection in indoor environments. Chickens were reared over two summer generations under two rearing conditions: completely indoors (IN) or with outdoor access from 28 days of age to slaughter (OUT). Most heritability estimates were higher in OUT, particularly for traits such as the feed conversion ratio (FCR) (0.42 ± 0.09, vs. 0.04 ± 0.03 in IN) and residual feed intake (RFI) (0.41 ± 0.08, vs. 0.08 ± 0.04 in IN). In addition to differences in heritability, lower genetic correlations between IN and OUT (0.77 ± 0.15 for FCR and 0.74 ± 0.13 for RFI) indicated significant G × E interactions, with a loss in genetic gain up to 76% for FCR and 67% for RFI when selecting under IN conditions for OUT performance. In contrast, traits such as breast yield showed higher genetic correlations (0.94 ± 0.04) and less loss in genetic gain (17%), indicating robustness to rearing conditions, with smaller differences in heritability between environments (0.66 ± 0.09 in OUT vs. 0.51 ± 0.07 in IN). These results highlight the strong influence of G × E interactions on selection in slow-growing chickens, particularly for feed-use-efficiency traits, which have different responses to selection depending on the selection environment. When direct selection in outdoor environments is impractical, strategies such as selecting collateral relatives tested outdoors, applying genomic selection with outdoor phenotyping, or modeling G × E interactions could optimize genetic progress for outdoor systems.