The AMS project - New approaches for management and breeding of dairy cows in automatic milking systems (AMS)

Large amounts of heterogeneous data has been generated in the AMS project. To establish reliable predictions of metabolic function and ovarian activity, data from different sources has been combined. This part of the project is strongly affiliated with the activities of The Biospectroscopy and Data Modeling Group at The Department of Matematical sciences and Technology at NMBU

Energy balance, health, and reproduction in AMS

Identification of cows with increased risk of subclinical ketosis or impaired ovarian function due to negative energy balance enable farmers to initiate feed corrections to avoid compromised animal health and reproduction. We have used information from infrared spectra in milk to determine energy balance in dairy cattle during the first 100 days in milk. We are also using automated body condition scoring (camera), activity and rumination, together with real time AMS data for early detection of subclinical ketosis and delayed onset of ovarian activity.

We have also developed methods for detection and monitoring of cows with subclinical intramammary infection by the use of online cell count (OCC) (http://www.delaval.com). OCC deliver information on a cow’s milk somatic cell count from every milking, but thresholds between pathology and normal processes have not been sufficiently established prior to the AMS project. We have validated the discriminatory capacity of this device and used OCC to build a deterministic forecasting model of the herd udder health status.

Milking time testing

We have found relationships between teat-end condition and milking time tests such as; milk flow, length of overmilking period, vacuum levels in the teatcups, total milking time and vacuum in the mouth piece chamber. We conclude that quarter milk flow rate provide useful information when evaluating teat-end condition in AMS herds. We have also found that vacuum level can be expressed by the milk flow variables in an AMS system. We have compared milk time testing in parlor, tie stalls and AMS to determine the impact on SCC between milking systems. We have also studied how much of the variability of SCC that can be explained by different traits monitored by the AMS system.

Breeding

We have successfully used milking time as a proxy for milkability in heritability studies, and also found kick off and incomplete milkings to be useful for the assessment of temperanmet in cows milked by AMS.

Technology adaptation, quality of life and new management methods for farmers with AMS herds

We have studied farmer’s perceptions of quality of life, health environment and safety and economic revenue on AMS farms in Norway. Data from both NDHRS and the financial dairy herd recording system has been used.

Feeding

We have expressed the cows energy balance by Fourier Transform Infrared Spectroscopi. We have studied the We have feed a group of cows using a standard lactation curve and compared this approach with a dynamic feeding regime. We have aslo assessed the association between chewing time and fiber content in feed.