PSV-16 Validation of an algorithm to assess feedbunk replacement events in beef cattle using an electronic feeding system

Abstract

Abstract Visual observations of social behavior and dominance relationships in cattle have been used to examine associations with productivity and well-being. This method is time consuming limiting the number of animals that can be evaluated. The objective of this study was to validate an algorithm to quantify feedbunk replacement events using data from an electronic feeding system. Crossbred beef steers (n = 20) fed a grower diet were housed in 1 of 2 pens each equipped with 3 electronic feedbunks (GrowSafe Systems) and video recorders. A trained video observer recorded all feedbunk replacement events and other agonistic activities at the feedbunk over a 4-d period (24 h/d). The electronic feeding system recorded the start and end timestamps of bunk visit (BV) events for each animal. An algorithm was developed to determine BV events deemed to be replacement events, defined as a BV event when an actor animal displaced a reactor animal from the feedbunk and occupied the same feeder within a specified period of time (replacement criterion). We calculated the recall and precision corresponding to replacement criterions from 1 to 60 s, and the optimum replacement criterion was determined to be between 18 and 20 s. The recall, precision and F-score of the algorithm using this replacement criterion were high (on average > 0.75). Furthermore, a replacement competition index was computed as a proxy for competitive feedbunk behavior, calculated as the number of actor-initiated replacement events divided by the total number of replacement events for each steer. Using Spearmans rank correlation we found high correlations (r > 0.7; P < 0.05) between the electronic and observed indices. These preliminary results demonstrate the potential of the GrowSafe system to quantify feedbunk replacement events for confined beef cattle, providing opportunities to evaluate associations between competitive feedbunk behavior and economically relevant traits.