“…For each scan, the observer selected the first observable bird passing nearest the window camera from which to record its behaviour, with the provisos that the chamber was full of birds and the conveyor was moving to maintain condition consistency. These provisos occasionally meant the scan was delayed by up to a maximum of 2 min, but on average only 5 s. The behaviour categories recorded were similar to those used by Abeyesinghe et al (2007) in the pilot scale system with the addition of some new postural categories. The bird's posture was described as: i) lie bird lying (ventrally, dorsally or laterally) with body in contact with the conveyor and supine head; ii) sit -bird sitting with sternum in contact with the conveyor with an erect head; and iii) stand -bird standing upright on two legs.…”