Validating behavioral sampling techniques for lame sows administered flunixin meglumine and meloxicam

Lameness poses a welfare challenge for pigs as it is associated with pain. Monitoring changes in behavior is a useful tool for recognizing illnesses in animals, including lameness. Lame sows spend more time lying down compared to non-lame animals, but there is currently no practical way of recording these changes in behavior. The objectives of this study were to determine if scan sampling could detect behavioral changes in sows induced lame, and detect changes in behavior associated with pain mitigation in sows induced lame. Lameness was induced in 12 multiparous, crossbred sows using a chemical synovitis model. Standing, sitting and lying behaviors were evaluated using continuous sampling, as well as 5, 10 and 15 min scans ('scan sampling') for 12 h/d (0600-1800) over a 5-day period. Three pain control treatments were evaluated: flunixin meglumine, meloxicam, and sterile saline administered approximately 28 and 52 h after lameness induction. No differences were found in lying and standing behavior in saline treated sows for all sampling methods (P =0.99). Regardless of sampling method, standing time decreased in the days following lameness induction compared to the day before induction (P <0.01). After treatment with flunixin meglumine, sows increased their standing behavior, but this change was only detected using the continuous sampling method (P <0.04). All sampling methods were able to detect standing behavior changes among sows treated with meloxicam (P <0.01). Results suggest that scan samples of 15 min or less are effective in detecting most behavioral deviations in sows treated with meloxicam or not treated at all.