543
Effect of an automated estrous detection system during a timed AI program on first postpartum AI
The aim of this study was to compare two reproductive programs for first AI based on activity monitors and timed AI. Seven hundred and seventy-four lactating Holstein cows from two commercial farms (Farm A, n=322; Farm B, n=452) were enrolled. Animals were presynchronized with two injections of PGF followed by an ovsynch protocol. Treatments were 1) TAI: all first inseminations performed by timed AI and 2) ACT: first insemination based upon estrous detection by activity monitors (Heatime, SCR Engineering, Israel) after the presynchronization, whereas the remaining non-inseminated cows were enrolled onto the ovsynch protocol. All animals had their body condition score (BCS; scale 1-5), hock score (HS; scale: 1-4), gait score (GS; scale:1-4) and corpus luteum presence by ovarian ultrasonography recorded twice during presynchronization. Continuous data was analysed using ANOVA by proc glm whereas binomial data was analysed by logistic regression using proc logistic of SAS. The ACT treatment had 52.3% of cows inseminated by estrous detection, whereas 90.3% of cows were bred by timed AI in the TAI treatment. Pregnancy per AI did not differ (P=0.12) and was 30.0% and 34.5% for ACT and TAI, respectively. Overall, pregnancy per AI was similar in cows bred by estrous detection compared with the ovsynch protocol (29.5% vs 31.8%; P=0.38). Median days open was similar between treatments (P=0.92). From all cows inseminated upon estrus only 30.9% and 36.3% were considered lame (GS>2) and with a swollen hock (HS>2), respectively. Fertility was affected by foot and leg conditions as animals that were classified as being lame, with a swollen hock or both had decreased pregnancy per AI compared with healthy animals (P<0.01; 25.6% vs 39.9%). The BCS did not affect the number of animals inseminated by estrous detection, but low BCS decreased pregnancy per AI (P<0.01; 23.9% vs 38.9%). The proportion of non-cyclic cows during the presynchronization that were inseminated upon estrus was 38.7% compared with 54.9% in cyclic animals (P=0.05). In conclusion, pregnancy per AI did not differ between treatments; however, gait score, hock score, BCS and cyclic status highly impacted fertility and the number of animals detected in estrus. The use of activity monitors for automated estrous detection can be used strategically in a traditional reproductive program for first AI after calving.
Keywords: Activity monitor, dairy cow, timed AI