Thermographic variation of the udder of dairy ewes in early lactation and following an Escherichia coli endotoxin intramammary challenge in late lactation

A total of 83 lactating dairy ewes (Manchega, n = 48; Lacaune, n = 35) were used in 2 consecutive experiments for assessing the ability of infrared thermography (IRT) to detect intramammary infections (IMI) by measuring udder skin temperatures (UST). In experiment 1, ewes were milked twice daily and IRT pictures of the udder were taken before and after milking at 46 and 56 d in milk (DIM). Milk yield was 1.46 ± 0.04 L/d, on average. Detection of IMI was done using standard bacterial culture by udder half at 15, 34, and 64 DIM. Twenty-two ewes were classified as having IMI in at least one udder half, the others being healthy (142 healthy and 24 IMI halves, respectively). Four IMI halves had clinical mastitis. No UST differences were detected by IMI and udder side, being 32.94 ± 0.04°C on average. Nevertheless, differences in UST were detected for breed (Lacaune – Manchega = 0.35 ± 0.08°C), milking process moment (after – before = 0.13 ± 0.11°C), and milking schedule (p.m. – a.m. = 0.79 ± 0.07°C). The UST increased linearly with ambient temperature (r = 0.88). In experiment 2, the UST response to an Escherichia coli O55:B5 endotoxin challenge (5 μg/udder half) was studied in 9 healthy Lacaune ewes milked once daily in late lactation (0.58 ± 0.03 L/d; 155 ± 26 DIM). Ewes were allocated into 3 balanced groups of 3 ewes to which treatments were applied by udder half after milking. Treatments were (1) control (C00, both udder halves untreated), (2) half udder treated (T10 and C01, one udder half infused with endotoxin and the other untreated, respectively), and (3) treated udder halves (T11, both udder halves infused with endotoxin). Body (vaginal) temperature and UST, milk yield, and milk composition changes were monitored by udder half at different time intervals (2 to 72 h). First local and systemic signs of IMI were observed at 4 and 6 h postchallenge, respectively. For all treatments, UST increased after the challenge, peaking at 6 h in T 0055 (which differed from that in C00, C01, and T10), and decreased thereafter without differences by treatment. Vaginal temperature and milk somatic cell count increased by 6 h postchallenge, whereas lactose content decreased, in the endotoxin-infused udder halves. Effects of endotoxin on lactose and somatic cell count values were detectable in the infused udder halves until 72 h. In conclusion, despite the accuracy of the camera (±0.15°C) and the moderate standard errors of the mean obtained for UST measures (±0.05 to 0.24°C), we were unable to discriminate between healthy and infected (subclinically or clinically) udder halves in dairy ewes.