No-contact microchip measurements of body temperature and behavioural changes prior to foaling

estational length is highly variable in horses ranging from 320 to 360 days. Thus, determining parturition time is an important challenge for the horse industry. Body temperature can be used in cows and ewes as an indicator of parturition. Thus, the aim of this study is to determine if temperature can also be used as indicator of foaling. Thirty-nine mares were monitored over two foaling seasons (2018 and 2019). They were housed in 16 m2 stalls with access to pasture in group three times a week from 10:00 to 16:00. Night watch as well as video monitoring was ensured during foaling periods. Body temperature was monitored using an identification and temperature sensor microchip implanted in the neckline. Measurement were taken manually every 2 h from 5 days before to 6 h after parturition by moving a microchip reader close to the mares' neck. Mares were equipped with a tail accelerometer recording tail movements and lateral recumbency 24 h before parturition. In addition, behaviour was monitored by video analysis in the hour preceding expulsion of the foal in 8 individuals in 2019. Relationships between behavioural and temperature data were explored throughout principal component analysis (PCA). All foals were born healthy and no human intervention was required during foaling. Mean daily body temperature decreased significantly by 0.3 °C (95%; range: 0.42 to -0.19 °C) between the day of parturition and the mean temperature of the 5 preceding days. A significant temperature decrease was also detected 12 h before and at the onset of parturition. With a 0.5 °C threshold, foaling could be detected 12 h in advance with 96.6% sensitivity and 95.0% specificity, respectively. Tail movements were more frequent and shorter with impending parturition. Body temperature was positively correlated with increased frequency and duration of specific behaviours (flehmen, looking at their flank and rump scratching against the stall wall). In conclusion, as in other species, body temperature was related to signs usually associated with impeding parturition, with a significant temperature drop observed from 12 h before and at the time of foaling. Providing automated measurements become available, temperature monitoring could become an additional tool to predict parturition in mares.