| dc.description.abstract | 1. This study is the first to quantitatively compare the structure of feeding behaviour of broilers, ducks and turkeys as recorded by electronic feeders. It tests the hypothesis that this structure is so similar that the same models would be suitable to group the feeding behaviour of these species into meals. Visits to electronic feeders were recorded from 3470 broilers, 3314 turkeys and 480 ducks. The frequency distributions of the length of short intervals between visits to feeders varied between species as a result of differences in the number of visits within a feeding bout, the frequency of re-visits to the same feeder and probably in the likelihood of birds drinking within meals. The lengths of longer day-time intervals between visits to feeders were all log-normally distributed. Disaggregation of these intervals by feeding strategy (meal frequency) showed that the probability of birds starting to feed increased with time since feeding last in all species, which is consistent with the satiety concept. Two methods, one based on fitting a truncated log-normal, function, the other on observed changes in the probability of birds starting to feed with time since last feeding, gave very similar meal criteria estimates. These ranged from 1050 to 1200 s in broilers, 1650 to 1725s in ducks and 1250 to 1320 s in turkeys. There were large between-species differences in the average number of daily meals, intake per meal, and feeding rate. Despite this variation, the overall structure of feeding behaviour of broilers, ducks and turkeys was so similar that the same models were suitable for application in all three species. This would allow for standardised analyses of feeding behaviour of different avian species kept in different husbandry systems. © 2010 British Poultry Science Ltd. | en |
