PSXI-20 Milking collar activity data is associated with health events and feed intake in lactating Holstein cattle. (30th November 2020)

• Record Type: Journal Article
• Authors: Jantzi, Abigail E; Siberski, Cori J; Goetz, Brady M; Healey, Mary; Hayman, Kristen; Gorden, Patrick J; Baumgard, Lance H; Koltes, James E
• Is Part Of: Journal of animal science. Volume 98(2020)Supplement 4
• Page Start: 392
• Page End: 393
• Publication Date: 2020-11-30
• Subjects: activity sensors -- feed intake -- milking collars; Livestock -- Periodicals; Livestock; Electronic journals; Periodicals; 636.005
• DOI: 10.1093/jas/skaa278.690
• Languages: English
• ISSNs: 0021-8812

Abstract

Abstract: Feed is the largest expense for dairy farms, thus feed efficiency is essential to the sustainability and future of the industry. Our objective was to evaluate the association of milking collar activity with feed intake and health status in lactating cows. Health status was classified for impact of three durations of time (overall, current, or post diagnosis) and as: healthy, mastitis, lame, multiple, or other. Activity data for 155 lactating cows with feed intake records were averaged across two-hour windows to obtain a daily two-hour average. A larger population (n > 1, 600) was used to filter out sensor failures and normalize data. Sensor data were adjusted for parity and contemporary group creating adjusted sensor measure (ASM). Dry matter intake (DMI) was adjusted (aDMI) for metabolic body weight, days in milk, and energy sinks used to calculate residual feed intake. Associations between ASM and aDMI, DMI, or health were conducted in SAS9.4. An association of ASM with aDMI was identified (estimate = 0.1635 kg/log count of average activity in a 2-hour period; P < 0.0029). ASM was also associated with DMI (0.2329 kg/log count of average activity, P < 0.0007). ASM was associated with current and overall health timeframes (P < 0.0008 and P < 0.0001, respectively). When health, ASM, and their interaction were included in a model with the response variable aDMI, significant associations were found in the models, including current and overall health (current health: ASM and health: P < 0.0001, interaction: P < 0.0009; overall health: ASM, health, and interaction: P < 0.0001). These results indicate that milking collar data may be useful as a predictor of feed intake either directly or indirectly through detection of health events. Additional studies are needed to determine the predictive ability of collar activity data and the relationship between collar data and health, and to assess if collar activity is an environmental proxy or heritable trait.