The present work aims to estimate the methods of repeatability and of a new non-parametric approach based on typifying individuals into classes and quantifying (%) the pigs in a group that show similar feeding behaviour habits (FBHs) in consecutive periods ("maintenance"). Both methods were estimated over six consecutive 14-day periods in two trials of group-housed growing-finishing pigs (n = 60 each). The first trial started in summer and ended in autumn, and pigs were fed a pelleted diet (HT-P). The second trial started in spring and ended in summer, and the same diet was fed mash (TH-M). The average daily feed intake obtained the lowest repeatability and maintenance values, and it progressively decreased as pigs grew, independent of environmental conditions or physical feed form, whereas the maintenance and repeatability of the number of feeder visits and the visit size decreased when environmental conditions changed from temperate to hot, and mash-fed pigs had higher maintenance and repeatability values for the time spent eating than pellet-fed pigs. In conclusion, the new approach (maintenance) is a tool that is complementary to the classic repeatability concept and is useful for analysing the evolution of FBHs across periods of time at the individual level.
Keywords: environmental conditions; feeding behaviour habits repeatability; growing-finishing pigs; physical feed form.