PSII-15 Profiling behavioral changes during the transition period in clinically healthy grazing dairy cows.

Abstract

Lying behavior may provide predictive information about the timing of calving in grazing dairy cows. We hypothesized that cows would be more active around calving. Standing and lying data from Icetag or Iceqube accelerometers were available from 311 mixed-age and breed grazing dairy cows. Lying behavior during the transition period (experimental days (expday); d -21 to d +34 relative to calving) and on the day of calving (d 0) was examined to calculate mean lying time (LT; h/d) and number of lying bouts (LB; n/d). Relationships between lying parameters and expday were examined using PROC NLIN, then modelled using PROC GLM with cow and expday as fixed effects. The effects of period (i.e., PRE: d -21 to d -3, d 0, or POST: d 3 to d 21) on lying behavior were analyzed using PROC MIXED, including expday and cow as fixed and random effects, respectively. The largest changes occurred on d -3 and d 0 for LT and on d -2 and d -1 for LB. Within cow, there was a decrease (P g< 0.001) in LT between d -3 and d 0 (R² = 0.57: slope = -1.0 ± 0.06; intercept = 8.5 ± 1.07), and the mean LT was less on d 0 (7.0 ± 0.12 h/d) than (P < 0.05) PRE or POST calving (8.3 and 10.3 ± 0.12 h/d, respectively). The number of LB increased (P < 0.001) during the 2 d before calving (R² = 0.62: slope = 4.51 ± 0.42; intercept = 14.8 ± 3.72), and the number of LB was greater on d 0 (12.5 n/d; P < 0.05) than PRE or POST calving (8.4 vs 8.0 n/d, respectively). Changes in behavior between d -3 to d -1 indicate that lying parameters may predict calving time in individual grazing dairy cows.