Relationships between activity monitoring device data and ovarian, uterine, hormonal, and pregnancy variables in beef cows

Implementing accelerometer technologies in beef operations is an alternative to increase precision in estrous detection. We hypothesized that (1) the accelerometer algorithm has similar accuracy in detecting behavioral estrus as does visual observation of pressure-sensitive sensors (estrus patches) in grazing beef cows; (2) variables measured by the accelerometer, such as estrus intensity, are associated with hormonal, ovarian, and uterine variables monitored before, during, and after estrus; and (3) the accelerometer variables are associated with the probability of pregnancy in grazing beef cows submitted to embryo transfer (ET). Fifty cows were fitted with accelerometer and patches to detect estrus after a synchronization protocol in eight subsequent rounds. For each round, only cows that showed estrus (day 0; D0) received ET. Follicular diameter, endometrial thickness, corpus luteum (CL) area, and estradiol (E2) and progesterone (P4) concentrations were measured during proestrus, estrus, and early diestrus. On D7, ET was performed. Pregnancies were diagnosed on D46 and cows recovered for 35D before a new replicate. Patches had a greater accuracy (98 % vs. 91 %) of detection of behavioral estrus than accelerometer algorithm. Cows with lower estrus intensity in the accelerometer had greater follicular diameter on D0 (P = 0.022), CL area on D4 and D7 (P = 0.05), endometrial thickness on D-1 (P = 0.10), and reduced E2 concentrations on D-1 (P = 0.0032). The accelerometer variables did not predict accurately the probability of pregnancy/ET. In conclusion, visual observation of patches was more accurate in detecting estrus than the accelerometer algorithm and most of the associations between accelerometers and physiological variables were for characteristics measured at proestrus.