Use of early lactation milk recording data to predict the calving to conception interval in dairy herds

Cook, J.G.; Green, Martin J.

Abstract

Economic success in dairy herds is heavily reliant on obtaining pregnancies at an early stage of lactation. Our objective in this study was to attempt to predict the likelihood of conception occurring by d 100 and 150 of lactation (days in milk, DIM) by Markov chain Monte Carlo analysis using test day milk recording data and reproductive records gathered retrospectively from 8,750 cows from 33 dairy herds located in the United Kingdom. Overall, 65% of cows recalved with 30, 46, and 65% of cows conceiving by 100 DIM, 150 DIM, and beyond 150 DIM, respectively. Overall conception rate (total cows pregnant/total number of inseminations) was 27.47%. Median and mean calving to conception intervals were 123 and 105 d, respectively. The probability of conception by both 100 DIM and 150 DIM was positively associated with the average daily milk weight produced during the fourth week of lactation (W4MK) and protein percentage for test day samples collected between 0 to 30 and 31 to 60 DIM. Butterfat percentage at 0 to 30 DIM was negatively associated with the probability of conception by 100 DIM but not at 150 DIM. High somatic cell count (SCC) at both 0 to 30 and 31 to 60 DIM was negatively associated with the probability of conception by 100 DIM, whereas high SCC at 31 to 60 DIM was associated with a reduced probability of conception by 150 DIM. Increasing parity was associated with a reduced odds of pregnancy. Posterior predictions of the likelihood of conception for cows categorized as having “good” (W4MK >30 kg and protein percentage at 0 to 30 and 31 to 60 DIM >3.2%) or “poor” (W4MK <25 kg and protein percentage at 0 to 30 and 31 to 60 DIM <3.0%) early lactation attributes with actual observed values indicated model fit was good. The predicted likelihood of a “good” cow conceiving by 100 and 150 DIM was 0.39 and 0.57, respectively (actual observed values 0.40 and 0.59). The corresponding values for a “poor” cow were 0.28 and 0.42 (actual observed values 0.26 and 0.37). Predictions of the future reproductive success of cows may be possible using a limited number of early lactation attributes.

Citation

Cook, J., & Green, M. J. (in press). Use of early lactation milk recording data to predict the calving to conception interval in dairy herds. Journal of Dairy Science, 99(6), https://doi.org/10.3168/jds.2015-10264

• Journal Article Type: Article
• Acceptance Date: Feb 27, 2016
• Online Publication Date: Mar 31, 2016
• Deposit Date: Oct 24, 2016
• Publicly Available Date: Oct 24, 2016
• Journal: Journal of Dairy Science
• Print ISSN: 0022-0302
• Electronic ISSN: 1525-3198
• Publisher: American Dairy Science Association
• Peer Reviewed: Peer Reviewed
• Volume: 99
• Issue: 6
• DOI: https://doi.org/10.3168/jds.2015-10264
• Keywords: dairy cow; pregnancy; early lactation; milk recording data
• Public URL: https://nottingham-repository.worktribe.com/output/778312
• Publisher URL: http://www.sciencedirect.com/science/article/pii/S0022030216301114
• Contract Date: Oct 24, 2016