Reproductive performance with automated activity monitoring or a timed insemination program for first insemination in dairy cows

Abstract Text: The objective of this study was to compare reproductive performance in lactating cows inseminated exclusively with timed artificial insemination (TAI) or with maximal use of an automated activity monitoring (AAM) system for the first insemination postpartum. From April 2014 to December 2015, a total of 998 cows in two herds in Ontario were randomly assigned to be inseminated at 85 ±3 days in milk (DIM) following a Double Ovsynch protocol (DO), or be inseminated following detection of estrus by the AAM system between 50 and 75 DIM. In the AAM group, if estrus had not been signaled by 75 DIM, cows received the Ovsynch protocol and were inseminated at 85 ±3 DIM. After first insemination, cows were managed according to routine herd management programs (combination of AAM and timed AI). The odds of pregnancy at first insemination and by 88 DIM were used as the outcome for logistic regression models. Models were adjusted for herd and parity as fixed effects, and interactions between treatment and covariates were tested. Analyses were done on cows that completed the protocol as assigned (completed protocol basis, n = 719) and on all cows that were not culled before first insemination (intention to treat basis, n = 849). The odds of being pregnant to first insemination were higher for cows in the DO group than in the AAM group in the intention-to-treat analysis (0.56 vs 0.42, P = 0.05), but were not statistically significant (0.58 vs 0.45, P = 0.12) for the completed protocol analysis.  The odds of being pregnant by 88 DIM tended to be higher for cows in the AAM group than in the DO group, but was not statistically different for the completed protocol (0.74 vs 0.59, P = 0.13) or the intention–to-treat analyses (0.70 vs 0.56, P = 0.11).There was an interaction of treatment with herd in both models, such that more cows in the AAM group were pregnant by 88 DIM in one herd, but there was no difference in the other. In this study, the exclusive use of Double Ovsynch had a higher probability of pregnancy at first AI than AAM, but earlier insemination in the AAM group and the possibility of re-insemination resulted in no statistical difference in the proportion of cows pregnant by 88 DIM. There were differences in the relative performance of TAI and AAM between herds.

Keywords: automated activity monitoring, Double Ovsynch, reproductive performance