Forty Holstein heifer calves, ranging in age from 4 to 8 wk were housed in the calf nurseries of the Elora Dairy Research Centre (University of Guelph, Guelph, Ontario). Calves were individually housed in 1.3 m × 1.5 m pens, and had visual and tactile contact with the adjacent calves. Calves were fed 2 L of milk twice per day and had ad libitum access to calf starter (20% crude protein with lasalocid; Floradale Feed Mill, Elora, Ontario) and water from birth.
Calves were randomly assigned to 1 of 2 treatments. “Ketoprofen” calves were administered an intramuscular (IM) injection of ketoprofen (Anafen; Merial Canada, Baie d’Urfé, Quebec) at a dose of 3 mg/kg bodyweight (BW). “Control” calves were administered physiological saline at the same dose. All calves were administered 5 mL lidocaine (2% lidocaine HCl with 0.05 mg/mL epinephrine; Bimeda-MTC, Cambridge, Ontario) as a cornual nerve block for each horn bud. Treatments and nerve blocks were administered and the first jugular sample was obtained 10 min prior to dehorning. The order of procedures was cornual nerve block, jugular blood sample, and finally IM injection of ketoprofen or saline. This series of procedures typically took no longer than 30 s to complete. Calves were dehorned using an electric cautery iron (Rhinehart X30; Rhinehart Development Corporation, Spencerville, Indiana, USA) which was pre-heated for 10 min to approximately 600°C. Treatments and dehorning were always performed at 10.00 h (±15 min) in the animal’s home pen by the same trained veterinary technician.
Blood was collected by jugular venipuncture into glass tubes containing no anti-coagulant (Vacutainers; Becton Dickinson, Franklin Lakes, New Jersey, USA) immediately following the cornual nerve block 10 min prior to dehorning (0 h) and again at 3 h and 6 h. Samples were allowed to clot for 30 min at room temperature and then were placed on ice and transported to the University of Guelph where they were centrifuged at 1400 × g. Serum was harvested and frozen at −20°C until all samples were collected. Serum samples were then delivered to the Animal Health Laboratory of the University of Guelph for cortisol analysis using a solid-phase, competitive chemiluminescent enzyme immunoassay kit (Diagnostic Products, Los Angeles, California, USA).
Video cameras (Panasonic, Model WV BP100, http://www.Panasonic.com
) and wide angle camera lenses (F 1.4/6 mm, Cosmicar, HS614GX; Pentax, Englewood, Colorado, USA) were installed on the wall facing the calf pens, providing a panoramic view of each pen. Calf behavior was recorded between 0 to 2, 3 to 5, and 6 to 8 h post-dehorning. Frequency of ear-flicks, head-shakes, and head rubs were recorded using continuous observation for 20-minute intervals immediately following dehorning (0 h) and at 1, 3, 4, 6, and 7 h post-dehorning. The observer was blind to treatment allocation. Scan sampling was conducted at the start of each 1-minute interval for the first 20 min of each observational hour and the posture of the calf (standing, lying, feeding, or grooming) at each of these time points was recorded. A complete ethogram of the behaviors observed is shown in and .
Description of behaviors recorded during 20-minute continuous observation periods
Description of behaviors recorded during 1-minute scan samples
The amount of calf starter (Rumax 20% calf starter; Floradale Feed Mill, Elora, Ontario) consumed on the day of dehorning was measured by weighing the amount of starter offered at time 0 and the amount remaining 24 h later.
Normality was assessed using Proc Univariate in SAS (version 9.1.3; SAS Institute, Cary, North Carolina, USA). Cortisol data were log transformed and analyzed by analysis of covariance (ANCOVA) using Proc Mixed. The 0 h cortisol concentration was used as the covariate. Main effects of treatment, time, age, season, and calf weight were included in the model.
Frequencies of ear-flicks, head-shakes, head-rubs, and total head behavior (ear-flicks, head-shakes, and head-rubs combined) were analyzed by Poisson regression using Proc Genmod and a Poisson distribution, with treatment, time, age, season, and weight included in the model. Repeated measures on calf were also taken into account using an autoregressive correlation structure.
Postural behaviors (standing, lying, grooming) were analyzed by logistic regression using Proc Genmod and a binary distribution, with treatment, time, age, season, and weight being included in the model.
Starter consumption was analyzed by a simple analysis of variance (ANOVA) using Proc Mixed (SAS Institute). Main effects of treatment, time, age, season, and weight were included in the model.