Abstract

A calibration equation based on partial least squares using Fourier transform infrared spectroscopy (FTIR) data was derived to predict concentration of lactoferrin (Lf) in bovine milk. The concordance correlation coefficient of the calibration equation was estimated at 0.81 using 246 milk samples collected in the 2003-04 season from second-parity crossbred cows during late lactation. The calibration equation was then used to predict the concentration of Lf (pLf) in 9,742 milk samples from 814 Holstein-Friesian (HF), 955 Jersey (J) and 842 crossbred HF x J cows sampled on average 3.7 times during the 2007-08 season. A repeatability animal model was used to obtain breed effects and estimates of variance components for pLf. Concentrations of pLF (mg/L milk) were significantly (P <0.001) different between breed groups (HF = 163 ± 3.8, HF x J = 175 ± 5.0 and J = 193 ± 6.3). Heritability and repeatability of pLf concentration were 0.16 ± 0.036 and 0.35 ± 0.012, respectively. These results confirm previous studies showing that Lf concentration in milk can be predicted using FTIR spectroscopy and that there are significant breed differences and animal variation that can be exploited in a breeding program.

N, Lopez-Villalobos, SR Davis, EM Beattie, J Melis, S Berry, SE Holroyd, RJ Spelman, and RG Snell

Proceedings of the New Zealand Society of Animal Production, Volume 69, Christchurch, 60-64, 2009