Efficiency of infrared pyrometer and infrared thermography for assessing body surface temperature in hair sheep

The objectives of this study were: (i) to verify and compare the coat surface temperature (CST) of regions of interest (ROI) measured through infrared pyrometer (IRP) and thermography (IRT) under two environmental conditions, to evaluate (ii) reliability and (iii) sensitivity of the two infrared equipments to the environmental effect. Twenty hair ewes (Ovis aries), multiparous, non-lactating, non-pregnant, with body weight 59.38 ± 6.23 kg and body score condition ranging from 2.5 to 4.5 were evaluated. All CST, in both infrared methods, were higher (P < 0.001) in the afternoon (2 to 3 p.m.), with the exception of eyes surface temperature (TSEY) from IRT, which was higher (P < 0.001) in the morning (8 to 9 a.m.). Regarding the infrared methods at different times (twice a day, morning and afternoon), of the day, CST values for all ROIs were significantly higher (P < 0.001) in thermography, regardless of the time shift, except for ear pinna surface temperature (TSE) and rectum surface temperature (TSR), which showed no significant difference (P > 0.05) between the infrared methods during the morning shift. The thermal environment significantly influences (P < 0.001) the surface temperature of various anatomical regions in both infrared methods. The correlation coefficients between IRT and IRP were moderate. The IRT is more sensitive to the thermal environment, as it presented higher canonical coefficients in the environmental variables. Among surface temperatures, eye surface temperature is the most influenced by variables in the thermal environment, especially THI. All three methods (digital thermometer-DT, IRT and IRP) showed >80 % of the data variation in the first two components. Using the DT, we observed an association between vaginal temperature (VT) and RT with air temperature (AT) and temperature humidity index (THI), which was similar to the infrared thermography pattern, in which all ROIs were highly correlated with AT and THI. The opposite was observed in the pyrometer, in which an association between eye temperature and RH and lower factor loadings of the other ROIs with AT and THI in the first component was observed. In conclusion, while IRP and IRT showed similar performance in the morning, IRT proved to be more accurate in the afternoon, demonstrating a higher success rate in classifying the group of origin. Therefore, it is concluded that IRT is more reliable than IRP in detecting surface temperature changes under high temperatures.