Evaluation of oestrous detection in dairy cattle comparing an automated activity monitoring system to visual observation.

The objective of the study was to evaluate an automated activity monitoring (AAM) system for oestrous detection in dairy cows. Specifically, we set out to determine the oestrous detection efficiency and accuracy of the AAM system and to compare reproductive performance of cows detected either by AAM, by visual observation (VO) or by at least 1 of the 2 methods (EOM). A total of 1004 potential cow-periods from 348 cows were analysed. Oestrous detection rates (EDR) were calculated for 21 day (VWP+21), 42 day (VWP+42), and 63 day (VWP+63) after voluntary waiting period (VWP). Oestrous detection rate did not differ between AAM (42.1%) and VO (37.3%) during VWP+21 (p > 0.05) but was significantly higher in EOM (56.3%, p < 0.05). Oestrous detection by AAM achieved an efficiency (number of correctly detected oestruses/total number of cow-periods*100) of 35.6% and an accuracy (number of correctly detected oestruses/(number of true+ false oestruses)*100) of 83.8%. Visual observation resulted in an efficiency of 34.3% and accuracy of 75.1%. Pregnancy rate at 200 DIM was higher in AAM (66.8%) and EOM (68.8%) than in VO (57.1%, p < 0.05). Cows detected by AAM (p < 0.05) showed a 1.37 times greater risk to conceive than cows detected by VO as the reference (p < 0.05). The results show that an automated activity monitoring system can support oestrous detection in dairy cattle. The exclusive use of such a system, however, cannot be recommended.