Assessing the stability of herd productivity groups across lactation periods in automatic milking systems using multi-algorithm clustering
• Inverse relationship found between milk yield and fat-protein corrected milk. • Higher milk protein content observed in farms that retain cows for more lactations. • Non-continued production farms had more intensive management practices. • Non-continued production farms had earlier, longer first lactations. Automatic Milking Systems (AMSs) generate extensive data at each milking event, potentially offering valuable insights for data-driven herd management and sustainable farming practices. This study investigates an innovative analysis of AMS data aiming to identify and characterise dairy farms that effectively maintain cows with high productivity levels over multiple lactation periods. This analysis represents a new data-driven tool to guide farmers and decision-makers towards more informed herd management. Using Multi-Algorithm Clustering Analysis, we analysed data from 16 AMS-equipped farms to assess the continuity of High Productivity Group (PGs), defined by milk yield and quality, across seven lactation periods. Our findings reveal that farms capable of retaining cows in the High PG, called Continued Productivity farms, exhibit distinctive characteristics, such as slightly lower milk yield but higher milk protein content, compared to farms unable to maintain their High PGs. Notably, the Continued Productivity farms show less intensive milking events, longer milking intervals, and manage lactation cycles to mitigate early-life production pressures, especially in the first lactation. Conversely, Non-Continued Productivity farms, i.e. those unable to retain high PG cows, demonstrate higher milking frequency, shorter intervals, and younger delivery ages, particularly during the first lactation, which may contribute to higher herd turnover. These novel insights support more targeted farm management strategies aimed at sustainability and animal welfare, providing actionable information for decision-makers to optimise herd productivity across lactation periods.
Authors
- Karina Brotto Rebuli
- Elske Vrieze (ORCID: https://orcid.org/0000-0001-7766-9245)
- Laura Ozella (ORCID: https://orcid.org/0000-0001-7371-3309)
- Mario Giacobini (ORCID: https://orcid.org/0000-0002-7647-5649)
- Fernando Masía
Institutions
- Lely (Netherlands) (NL)
- University of Turin (IT)
Publication Details
- Journal
- Computers and Electronics in Agriculture
- Published
- 2025-03-29
- DOI
- https://doi.org/10.1016/j.compag.2025.110295
- Primary Topic
- Milk Quality and Mastitis in Dairy Cows
- Type
- article
- Field-Weighted Citation Impact
- 0.00
Funders
- Compagnia di San Paolo