Detection of estrus by three methods.

Estrus was detected by three methods in 88 postpartum Holstein cows housed in tie-stalls and allowed free activity 1 to 2 h daily in separate exercise lots. Cows in each of three groups were observed visually for standing estrus by herdsmen. Cows (31) in Group I were observed only visually and served as controls, cows (33) in Group II were fitted with MateMaster rump-mounted detectors, and cows (24) in Group III were exposed to a testosterone-treated marking heifer. Cows with less than 1 ng/mg progesterone in serum on the day of detection and greater than 1 ng/ml 10 days later were designated as accurately detected in estrus. Accuracies of detections for cows in Groups I, II, and III were 68, 66, and 79%. Efficiencies of detections were the total observed estrous periods divided by the total expected estrous periods and were 51, 51, and 52%. In addition, overall accuracy, a mathematical product of individual accuracy and efficiency, represented the best measure of a detection method. Overall accuracies in Groups I, II, and III were 35, 34, and 41%.