Validation of a system for monitoring feeding behavior in beef cattle1

The objective of this study was to evaluate the use of an electronic radio-frequency-identification-based system (GrowSafe System Ltd., Airdrie, Alberta, Canada) to measure feeding behavior traits in beef cattle fed a high-grain diet. Feeding behavior data were recorded by the GrowSafe system and time-lapse video using 10 heifers over a 6-d period. Observed bunk visit (BV) and meal event data (frequency and duration) were compared with electronic feeding behavior data generated by the GrowSafe system at 5 parameter settings (MPS; 30, 60, 100, 150, and 300 s), which are used to define the maximum duration between consecutive electronic identification recordings to initiate a subsequent BV event. A random coefficient model was used to compare video and electronic data using orthogonal contrasts. Video data were regressed on the electronic feeding behavior data to obtain an estimate of precision (r(2)) and other statistical estimates, including mean square error of prediction and concordance correlation coefficient, to access the adequacy of the electronic system predictions. The variation in MPS values affected BV data, but not meal event data. Electronic meal frequency and duration data were not different (P > 0.50) from observed values, and were not affected by electronic MPS values. The optimal MPS value for prediction of BV and meal event frequency and duration traits was 100 s. Our evaluation indicated the GrowSafe system 4000E was able to predict BV and meal event data when the 100-s MPS was used to analyze the feeding behavior data.