Using behavioral observations in freestalls and at milking to improve pain detection in dairy cows after lipopolysaccharide-induced clinical mastitis.

This study aimed to determine the effect of lipopolysaccharide (LPS)-induced mastitis with or without nonsteroidal anti-inflammatory drug (NSAID) on dairy cows' clinical, physiological, and behavioral responses in the milking parlor and freestalls as well as the specificity (Sp) and sensitivity (Se) of behavioral responses in detecting cows with LPS-induced mastitis. Twenty-seven cows received an intramammary infusion of 25 µg of Escherichia coli LPS in 1 healthy quarter. Following LPS infusion, 14 cows received a placebo (LPS cows), and 13 cows received 3 mg/kg of body weight of ketoprofen i.m. (LPS+NSAID cows). Cow response to the challenge was monitored at regular intervals from 24 h before to 48 h postinfusion (hpi) through direct clinical observations, markers of inflammation in milk, and via point-in-time direct behavioral observations in the barn and at milking. In LPS cows, infusion induced a significant increase of plasma cortisol levels at 3 and 8 hpi, milk cortisol levels at 8 hpi, somatic cell counts from 8 to 48 hpi, IL-6 and IL-8 at 8 hpi, milk amyloid A (mAA) and haptoglobin at 8 and 24 hpi, rectal temperature at 8 hpi, and respiratory rate at 8 hpi. Their rumen motility rate decreased at 8 and 32 hpi. Compared with before the challenge, significantly more LPS cows stopped feeding/ruminating and pressed their tail between their legs at 3 and 5 hpi, increased feeding/ruminating at 24 hpi, and had the tendency to be less responsive, dropping their head, and dropping their ears at 5 hpi. At milking, compared with before challenge, significantly more LPS cows lifted their hooves at forestripping at 8 hpi. The 2 groups showed similar patterns of response for milk cortisol, somatic cell count, respiratory rate, mAA, haptoglobin, and IL-6, IL-1ß, and IL-8. Compared with LPS cows, LPS+NSAID cows had significantly lower plasma cortisol levels at 3 hpi, their rectal temperature decreased at 8 hpi, their rumen motility rate increased at 8 and 32 hpi, and their heart rate increased at 32 hpi. Compared with LPS cows, a significantly larger proportion of LPS+NSAID cows were feeding/ruminating, a lower proportion had ears down at 5 hpi, and a larger proportion lied down at 24 hpi. At milking, whatever the phase of milking, for "hoof to belly," 9 out of 14 cows did not show this behavior before infusion (Sp = 64%) and 14/14 did not kick during pre-infusion milking (Sp = 100%). Regarding sensitivity, at maximum, 5 cows out of 14 (Se = 36%) displayed "hoof to belly" after infusion. For "lifting hoof," 14/14 did not show hoof-lifting before infusion (Sp = 100%) and 6/14 displayed it after infusion (Se = 43%) at forestripping only. In the freestall barn, 9 behaviors had a Sp >75% (at minimum, 10/14 did not show the behavior) whatever the time point but Se < 60% (at maximum, 8/14 displayed the behavior). Finally, "absence of feeding and ruminating" had Sp of 86% (12/14 ate/ruminated) and Se of 71% (10/14 did not eat/ruminate) at 5 hpi. This study shows that feeding/ruminating, tail position, and reactivity at forestripping could be used as behavioral indictors for early detection of mastitis-related pain in dairy cows.

Large-size fattening calves' lots fed with automatic milk feeders may have an increased risk for Mycoplasma bovis infection spread and for antibiotic use.

Bovine respiratory disease is the leading user of antibiotics (AB) in calf production. Mycoplasma (M.) bovis could lead to greater use of AB as it is a persistent and AB resistant causative agent for respiratory diseases. Two cross-sectional studies were set up to assess the effects of lot size and feeding system on M. bovis infection and the effects of M. bovis seroconversion, lot size and feeding system on AB use in calves' feedlots. Twenty-six lots in 22 fattening farms were monitored for 41-81 days, from all-in entry of calves until three consecutive weeks without using any collective antibiotics. M. bovis spread was estimated by measuring seroconversion at entry and at the end of study period in 10-15 calves randomly sampled in each lot. All AB treatments used in the meanwhile were recorded. The lots were selected according to feeding system, i.e. individual bucket (n = 7) vs. automated milk feeder (AMF, n = 19), and lot size (30-519 calves), less than 50 calves (n = 9) vs. more than 50 calves (n = 17). Statistical analysis was performed using multivariable generalised linear models with fattening farms as random effect. M. bovis spread increased with lot size (odds ratio (OR) 2.9[1.4; 5.8] per two-fold increase in lot size). This proportion of seroconverted calves was lower in bucket-fed lots compared to lots fed with the AMF using a shared nipple (OR = 0.03[0.003; 0.41]). The main risk factor for AB use was the lot size, with an increase of 1.5[0.94; 1.98] treatments per two-fold increase in lot size. For same size lots, the use of bucket can decrease AB consumption by up to 1.03[-2.18; 0.14] treatments per calf compared to AMF. Analysis of the association between seroconversion to M. bovis and AB use was inconclusive. We found that bucket feeding in small-size lots, i.e. up to a maximum of 50 calves in the same space, limits seroconversion to M. bovis and enables lower use of AB in veal calf production.