The effects of parity, litter size, physiological state and milking frequency on the metabolic profile of Lacaune dairy ewes

Effects of parity (primiparous, PRIM vs multiparous, MULT) and litter size (singletons, SING vs twins, TWIN) on metabolic profiles from 1 wk before lambing to the end of lactation were studied in 48 Lacaune dairy ewes reared in confinement during most of the year and grazed on improved pastures at the end of lactation (summer). Another group of 48 ewes was incorporated during the milking period (ie, from 1 wk after weaning), to measure the effects of milking frequency (1 vs 2 milkings per day) on intake, milk production and composition, and body energy usage. Thus, in a 2 × 2 × 2 factorial design, ewes (n = 96) were allocated to homogeneous groups according to body weight (BW) and body condition score (BCS) and were monitored from late pregnancy to late lactation as a function of parity (PRIM, n = 48; MULT, n = 48), litter size (LSi) (SING, n = 40; TWIN, n = 56) and daily milking frequency (FREQ; milked once, ONE; n = 48; or twice, TWO; n = 48). Individual BW, BCS, plasma metabolites, and metabolic hormones were measured regularly (ie, 9 consecutive sampling dates). The BW was higher in MULT but no differences because of LSi or FREQ were detected at the intra-parity group level. The BCS was higher in MULT and in ewes with SING throughout the experiment. The latter was related to the demands for body reserves mobilization, as expressed by higher nonesterified fatty acids and β-hydroxybutyrate concentrations in ewes with TWIN from late pregnancy to weaning (35 d postpartum) in both PRIM and MULT ewes. This was consistent with higher insulin in MULT and higher triiodothyronine, leptin and insulin-like growth factor 1 in ewes with SING during this period. Differences in energy balance because of FREQ were evident after interpretation of plasma nonesterified fatty acids, glucose, insulin, and leptin concentration during the milking period. At similar feed intakes, ewes in ONE were in positive balance with regard to TWO. Overall, clear effects of parity, LSi, physiological states, and FREQ on metabolic profiles were found because of differences in nutrient partitioning when combining these experimental factors. Without considering FREQ, changes in metabolic measures in milking period were marginal compared with the periparturient adjustments performed until weaning to compensate energy deficiencies