Evaluation of an ear-attached movement sensor to record cow feeding behaviour and activity.

Worldwide the number of cows per dairy farm has increased and, as a consequence, the time farmers spend on individual cows has been reduced. This trend, driven by reducing workload and increasing efficiency and profitability, has resulted in a rapid increase in the interest of farmers and industry for precision dairy farming technology. Precision dairy farming technology aims to support decision making and to improve farm productivity and profitability. Daily ruminating, eating and activity patterns are widely considered as indicators closely related to health events and productivity for individual cows. For research, easy measurement of these parameters may assist in better understanding of nutritional physiology. Therefore, automatic monitoring of feeding behaviour and activity could provide scientists and farmers a research tool and an early warning tool, respectively. The electronic behaviour sensor called the Cow-Manager SensOor system (Agis Automatisering BV, Harmelen, the Netherlands) provides such an opportunity. This system enables real-time quantification of (multi-point) ear temperature, rumination, eating behaviour and activity of dairy cows, but a scientific evaluation on the accuracy and precision of the model is required. Therefore, the objective of the study by Dr J.P. Bikker and colleagues was to determine the agreement between the classification categories of behaviour of the sensor against direct visual observations. The results were published in the Journal of Dairy Science, Volume 97 of 2014, page 2974 to 2979. The title of their technical note was: Evaluation of an ear-attached movement sensor to record cow feeding behaviour and activity. 

Based on the principle that behaviour can be identified by ear movements, the proprietary model of the sensor classifies the data as “ruminating,”“eating,”“resting,” or “active.”In the experiment a pilot evaluation of agreement between two independent observers was firstly established. They recorded the behaviour of three cows for a period of approximately 9 hours each. Then, in order to evaluate the sensor, the behaviour of 15 cows was monitored both visually (VIS) and with the sensor (SENS), for approximately 20 hours per cow, evenly distributed over a 24-h period, but excluding milking. Cows were chosen from groups of animals in different lactation stages and parities. Each minute of SENS and VIS data was classified into one of 9 categories (8 behaviours and 1 transition behaviour) and summarized into four behavioural groups, namely ruminating, eating, resting, or active, which were statistically analyzed by calculating so-called kappa (κ) values.

For the pilot evaluation, a high level of agreement between observers was obtained, with κ values of equal or more than 0.96 for all behavioural categories, indicating that visual observation provides a good standard. For the second trial, relationships between SENS and VIS were studied by κ values on a minute basis and Pearson correlation and concordance correlation coefficient analysis on behaviour, expressed as percentage of total time. The time spent ruminating, eating, resting and active were 42.6, 15.9, 31.6 and 9.9% for SENS and 42.1, 13.0, 30.0, and 14.9% for VIS, respectively. Overall, theκvalues for the comparison of SENS and VIS was high, with κ values of 0.85, 0.77, 0.86 and 0.47 for “ruminating,”“eating,”“resting,” and “active,” respectively. In support, the Pearson and concordance correlation coefficients between SENS and VIS for“ruminating,”“eating,”“resting” and “active” were 0.93, 0.88, 0.98, and 0.73 and 0.93, 0.75, 0.97, and 0.35, respectively. This suggests that, apart from “active”,  the agreement was high for all other categories. 

In conclusion, the results provide strong evidence that the Cow Manager SensOor  technology can be used to monitor ruminating, eating and resting behaviour of freestall-housed dairy cattle, but more work is required to determine its suitability to monitor the activity of dairy cattle. 

FURTHER RESULTS: In a recent publication (Journal of Dairy Science, Volume 100 of 2017, page 9635 to 9642), author M.V. Byskov and colleagues studied the potential of rumination time as an indicator of feed intake and feed efficiency, by estimating genetic correlations and heritability values, using the data of one research herd and 72 commercial herds. The title of their paper was: genetic parameters of rumination time and feed efficiency traits in primiparous Holstein cows under research and commercial conditions. Both genetic correlation and heritability values were not satisfactory from a selection point of view, and it was concluded that rumination time is not a suitable indicator trait for feed intake, but it is a weak indicator of feed efficiency.