The role of Fourier-transform mid-infrared spectroscopy in improving the prediction of new and existing traits in New Zealand dairy cattle : a thesis presented in partial fulfilment of the requirements for the degree of Doctor of Philosophy in Animal Science at Massey University, AL Rae Centre, Hamilton, New Zealand

Thumbnail Image
Open Access Location
Journal Title
Journal ISSN
Volume Title
Massey University
The Author
Bovine milk is a rich source of dietary nutrients that are important to human health. Market demand for bovine milk is driven by its nutritional value, price, processability, and consumer expectations and perceptions about food production systems. The ability to quantify traits associated with milk quality, processability, animal health and environmental impact is critical for selective breeding and thus highly valuable to the dairy industry. However, obtaining direct measurements of such traits can be difficult and expensive. Estimation of major milk components using Fourier-transform mid-infrared (FT-MIR) spectroscopy is common practice, and spectral-based predictions of these traits are already widely used in milk payment and animal evaluation systems. Applications using FT-MIR spectra to predict other traits have increased in popularity over the last decade, and are attractive alternatives to directly measuring phenotypes because the FT-MIR spectra are readily available as a by-product of routine milk testing. The objectives of this thesis were to improve understanding of the phenotypic and genetic characteristics of FT-MIR spectra, and assess the role that such data can play in predicting new traits or improving the prediction of existing traits in New Zealand dairy cattle. We assessed different strategies for improving the quality of spectral data and demonstrated that there are limitations in predicting traits such as pregnancy status, due to confounding effects such as stage of lactation. From a genetics perspective, we reviewed the evolving role of spectral data in the improvement of dairy cattle by selection and discussed opportunities for consolidating spectral datasets with other genomic and molecular data sources. We conducted GWAS on individual FT-MIR wavenumbers and demonstrated that the individual wavenumbers provided stronger association effects and improved power for identifying candidate causal variants, compared to conducting GWAS on FT-MIR predicted traits. We also demonstrated the potential utility of spectral data for predicting and incorporating fatty acids and protein traits into breeding programs, but showed that even when genetic correlations between directly measured and FT-MIR predicted traits were high, the detectable QTL underpinning these traits were not always the same. Although there are many potential applications for FT-MIR spectral datasets, there are still challenges to developing robust prediction equations and understanding the genetic relationships between traits of interest and their FT-MIR predictions. Addressing these challenges will provide opportunities to improve the prediction of new and existing traits in dairy cattle milk production systems and breeding programs into the future.
Listed in 2022 Dean's List of Exceptional Theses
Dairy cattle, New Zealand, Genetics, Breeding, Fourier transform infrared spectroscopy, Dean's List of Exceptional Theses