A statistical evaluation of associations between reproductive performance and milk composition and animal factors on grazing dairy cows in two New Zealand dairy farms

To determine if milk composition, milk yield, live weight, live weight change, breed and heterosis are associated with reproductive performance in dairy cows from two dairy farms under New Zealand grazing conditions. Milk composition was determined in herd tests from 205 Holstein-Friesian (F), 77 Jersey (J) and 351 F × J crossbred cows from two Massey University dairy herds in the 2016 and 2017 production seasons. Mating occurred from October to December in each production season. The start of breeding to first service (SBFS), start of breeding to conception (SBCO), submission rate at 21 days (SR21), pregnancy rate at 21 (PR21) and 42 days (PR42) were calculated for each cow. These traits were analysed using mixed linear models that included fixed effects for herd, production season, regression coefficients of deviation from median calving date, lactation number, proportion of F, F × J heterosis, energy-corrected milk yield (ECMY), percentages of fat, protein and lactose, milk urea nitrogen (MUN), live weight and change in live weight, with random effects for cow and residual error. The variables with binomial distribution were analysed using logistic regression. Deviation from the herd’s median calving date had a significant effect (p < 0.05) on all reproductive traits. Proportion of F was significant (p = 0.022) on PR21, and F × J heterosis effects were significant on PR21 (p = 0.049) and PR42 (p = 0.046). F cows had 17.8% higher PR21 than J cows, and F × J cows had higher reproductive performance than the mean of the two purebreeds. ECMY was negatively associated with SBFS (p = 0.001) and SBCO (p = 0.001) and positively associated with PR21 (p = 0.002) and PR42 (p = 0.001). Protein percentage was positively associated (p < 0.05) with PR21 and PR42, whereas lactose percentage was negatively associated (p < 0.05) with PR21 and PR42. Cows gaining live weight were more likely to become pregnant within 21 days of the start of breeding (p = 0.020). Milk urea nitrogen was negatively associated (p = 0.042) with SR21. This study confirms that breed, heterosis, ECMY, protein and lactose percentages, live weight change and calving date are associated with the reproductive performance of grazing cows. Results from this study contrast with the historical antagonism between milk production and reproductive performance in dairy cattle, demonstrating that well managed cows can achieve high levels of production and good reproductive performance. MUN was not associated with reproductive performance traits, except with SR21.