Using Fourier Transform Infrared (FTIR) Spectra to Investigate GXE Interactions in Danish Holstein Cattle
77th Annual Meeting of the European Federation of Animal Science (EAAP), Hamburg, Almanya, 6 - 11 Eylül 2026, ss.197, (Özet Bildiri)
- Yayın Türü: Bildiri / Özet Bildiri
- Basıldığı Şehir: Hamburg
- Basıldığı Ülke: Almanya
- Sayfa Sayıları: ss.197
- Ankara Üniversitesi Adresli: Evet
Özet
Genotype‑environment interaction (G×E) plays a significant role in milk production systems because animals can respond differently to changing climatic conditions. In dairy cattle, it is often observed that current practic es overlook this aspect, assuming that the same genotype will be effective across all environments. However, G×E interactions can often be a significant source of variability in productivity and reproductive performance. Understanding G×E interactions is important, especially for changing dairy cattle breeding practices. In recent years, Fourier‑transform infrared (FTIR) spectra have begun to provide a new perspective not only for routine analysis of milk composition but also for the analysis of complex phenotypes with low predictability, such as disease incidence, by integrating genomic data with high‑throughput phenomic information. This can serve as an indirect indicator to characterize the environmental variation that animals are exposed to. The aim of this study is to evaluate the integration of FTIR spectral data into genomic prediction models and to examine its impact on prediction accuracy across herds, considering potential genotype‑environment interactions. Mastitis was used as a model trait to evaluate this approach. The dataset includes test day records (39 000) for milk yield (MY), clinical mastitis cases, and somatic cell count of 10 000 Holstein animals as well as infrared spectral data in the period of October 2023 to December 2024. Spectral data were obtained using MilkoScan FT2 (FOSS Analytical A/S), which captures infrared absorbance in the range of 5008 to 925 cm⁻¹, encompassing a total of 1060 individual wavelengths. Genotypes (50K SNP chip) and pedigree data were available for all animals. In this study, environ mental information was extracted from phenotypic and spectral data. This study employed a kernel‑based genom ic prediction approach integrating milk FTIR spectral data. Preliminary findings for SCC showed a heritability of 0.08 (SE 0.02) when including a GRM only in the model. The variance explained by the IRM for SCC was 0.63, indicating that the infra‑red spectra can explain some of the variance of SCC. Including both the GRM and IRM in the model showed that the GRM took most of the variance while the IRM variance went to zero. Further analyses are ongoing to evaluate model performance and investigate the possibility to use milk infrared spectra data in relation to the prediction of genotype‑environment interaction.