Abstract #97
Section: Production, Management and the Environment (orals)
Session: Production, Management, and Environment I
Format: Oral
Day/Time: Monday 11:30 AM–11:45 AM
Location: Ballroom C
Session: Production, Management, and Environment I
Format: Oral
Day/Time: Monday 11:30 AM–11:45 AM
Location: Ballroom C
# 97
Estrus detection with an activity and rumination monitoring system in an organic grazing and in a low-input conventional herd.
Glenda M. Pereira*1, Bradley J. Heins1, Marcia Endres1, Kota Minegishi1, 1University of Minnesota, St. Paul, MN.
Key Words: automated estrus detection, grazing, low-input dairy
Estrus detection with an activity and rumination monitoring system in an organic grazing and in a low-input conventional herd.
Glenda M. Pereira*1, Bradley J. Heins1, Marcia Endres1, Kota Minegishi1, 1University of Minnesota, St. Paul, MN.
Estrus detection using an activity and rumination system (ARS) was evaluated in an organic grazing and in a low-input conventional seasonally calving herd. Estrus prediction models were created using raw data provided by the ARS. The study was conducted from June 2014 to August 2017 at the University of Minnesota West Central Research and Outreach Center, Morris, MN. All cows were fitted with an ARS tag (HR-LD tags; SCR Engineers Ltd., Netanya, Israel) at calving, and the tag was removed at dry off. Cows calved in the spring and autumn and were bred in the summer and winter. During each breeding season (4 summer and 3 winter breeding seasons), activity and rumination (daily and 2-h periods) were monitored electronically using the ARS tags. Activity was reported in activity units, and rumination was reported in min per 2-h block and min per d from SCR DataFlow II software. Estrus alerts of individual cows provided by the SCR DataFlow II software were used to determine if the alert coincided with the breeding date of a cow. The gold standard for this study were breeding dates of cows that were determined by breeder evaluation of an Estrotect patch placed on the rump of a cow. The study included 1,463 breeding dates across the 4-yr period. The ARS had a sensitivity of 56.7%, a specificity of 99.3% and a positive predictive value of 59.8% for the organic herd, and a sensitivity of 70.1%, a specificity of 99.2% and a positive predictive value of 66.3% for the low-input herd across breeding seasons. Higher sensitivity indicates that estrus alerts provided by the ARS coincide with true estrus events, defined by the current gold standard. Custom models illustrated the potential tradeoffs that can be achieved with ARS raw data, and the models had a sensitivity from 73.0% to 84.2% and a specificity from 94.3% to 97.7%. Adjusting the threshold of estrus detection may provide producers more control of ARS generated estrus alerts depending on the breeding season. The ARS evaluated in this study showed potential for estrus detection in grazing and low-input dairy herds.
Key Words: automated estrus detection, grazing, low-input dairy