Data on the eating, ruminating and idling activity of grazing dairy co
ws recorded to a resolution of 1 min were used to study the choice of
activity bout criteria and the implications of this choice. An inter-b
out length of 5 min was found to suitable for all three activity categ
ories across a range of treatments, at different times of the day and
for all animals. After smoothing using this criterion bouts contained
proportionally up to 0.069 of other activities for eating, 0.076 for r
uminating and 0.054 for idling. The proportion of intra-meal intervals
decreased slightly during the main evening meal and increased during
meals taken at night. It is concluded that the intensity of bouts shou
ld be reported as well as their number and duration and that total act
ivity times should be calculated from unsmoothed data. (C) 1997 Elsevi
er Science B.V.