Abstract #495

# 495
Characterization of the non-genetic causes of variation of bovine milk calcium concentrations on French farms.
P. Gaignon*1,2, M. Gele3, C. Hurtaud1, A. Boudon1, 1PEGASE, INRA, Agrocampus Ouest, Saint-Gilles, France, 2CMI, 18 avenue F. Roosevelt, Saint-Malo, France, 3Institut de l'élevage, Angers, France.

Calcium concentration (CaC) in bovine milk has often been described as independent of feeding strategy and mainly dependent on cow genetics and lactation stage. However, isolated experiments showed that variations in milk CaC could be linked to the diet of cows. Our objective was to identify and quantify non-genetic factors of variation in CaC in milk samples collected from about a thousand French dairy farms with contrasting feeding strategies and cow breeds. This study was based on the PhénoFinlait program that consisted of a survey performed between 2009 and 2010 in 924 dairy farms located in the major French milk production areas. The breeds used in the investigated farms were Holstein, Normande and Montbeliarde. Each farm was visited on average 4 to 6 times during the year. Each time, information about cow diets and production were gathered and individual milk samples were collected to extract their mid-infrared (MIR) spectra. More than 200,000 MIR spectra were measured. Nearly 10,000 milk samples were also frozen and stored in a bank for further analyses. We estimated CaC in milk samples from their MIR spectra using a predictive equation. This equation was established from 300 milk samples extracted from the bank and chosen to represent the diversity of investigated dairy systems. From the composition of the cow diets collected at each survey, we characterized 7 feeding strategies using multiple factorial analyses across 3 periods: winter, early and late summer. For each breed, the variations in milk CaC were quantified by ANOVA with a model including the effects of feeding strategies, stage of lactation, parity, and calendar month as fixed effects and the cow as random effect. The feeding strategy affected milk CaC with the constant fact that the diets based on fresh or conserved grass induced lower milk CaC whatever the month of the year (P < 0.05). The difference in CaC can be up to 100 mg/kg between 2 extreme diets at a given month, which is as important as the drop in CaC observed at the beginning of lactation. This study reinforces the idea that the diet of cows has an influence on milk CaC.

Key Words: calcium, milk, feeding strategies