Sensitivity of heat detection attentions and their relation with the moment of LH surge in dairy cows

The sensitivity of an estrus detection system and the consistency of alarms relative to ovulation determine its value for a farmer. The objective of this study was to compare four different heat detection systems for their ability to detect heat and predict the moment of the LH surge in a single herd during the same study period. This comparison, in which the moment of the LH surge was used as an indicator for ovulation, allowed for the objective evaluation of each system and the potential for optimizing the fertility management on farm. The four systems were traditional visual observation, an activity-based system and two progesterone-based methods: (1) Herd NavigatorTM and (2) a novel algorithm which combines a mathematical model and synergistic control. The latter algorithm also allowed to test whether using the mathematical model could improve the relation with the LH surge. First, the systems were compared in terms of sensitivity and positive predictive value for heat detection. Then, the time interval between the attentions and the LH surge was investigated and compared based on their range and SD. Heat attentions based on visual observations had the lowest sensitivity to detect heat (40%), and were noted from 4 hours before until 5 hours after the LH surge (range 9 hours, SD 4 hours), indicating a strong relation. Activity-attentions proved more sensitive (80%). They had the least accurate relation with the moment of the LH surge and were observed from 39 hours before until 8 hours after it (range 47 hours, SD 16 hours). Attentions based on milk progesterone measurements correctly identified all estrous periods. Herd NavigatorTM attentions system were followed by the LH surge after 22 to 66 hours (range of 45 hours, SD 11 hours). The model-based approach generated attentions 49 to 81 hours (range 33 hours, SD 11 hours) before the LH surge. As detection of the LH surge was very labor-intensive, only a limited number of potential heat periods could be studied. For some of the methods (e.g. visual observations), the sensitivity restricted the number of cases even more. Nevertheless, the approach ensured an objective comparison between relevant heat detection systems in a commercially representative setting. Accordingly, this study helps to place larger studies linking heat detection and ovulation into perspective and shows the potential of new P4-interpreting algorithms, thereby highlighting the need for further research.