Records of feed intake in ducks are usually reported as means over a large number of birds and extended periods. By this procedure information on the variability between the individual birds and the temporary structure of feed intake is lost. So far, experiments on individual feed intake in ducks (as in other poultry species) have been carried out in a few birds and over a few days only. Also the birds had to be kept individually. Since ducks are usually kept in large flocks, the results of these experiments may not be applicable for practical husbandry conditions. Methods to record feed intake in group reared individuals were only available for larger domestic species, such as pigs and cattle. Similar equipment for ducks using an antennareceiver system had to be developed for the present study. The antenna was fitted nearby the feeder and the transponder was attached to the wing mark. The feeders were connected to an electronic scale and the following characteristics were recorded by a PC continuously: number of the bird, day, time of day, time of visits to the feeder, feed consumption. The system was suitable for duckling from 14 days old onwards. Two experiments were carried out. In experiment 1 a total of 50 male Pekintype hybrids from 14 to 49 days of age were used. In experiment 2 records were taken from 480 male and female ducks of an experimental line from day 21 to 45. The birds were raised in groups on deep litter (straw). Commercial pelleted. feed and water were provided ad libitum. The distance between feeders and drinkers was I m in experiment 1 and 5 m in experiment 2. The following traits were extracted from the records: - Number of meals per day (defined as visits to the feeder with feed intake) - Duration of meals (sec.) - Size of meals (grams) - Intensity of feeding (feed intake in grams per min): feed intake per meal in relation to duration of meal - Duration of pauses (sec. or min.) - Duration of feed intake per day (min.) - Feed intake per day (grams) - Live weight (grams) Data of a total of 530 ducks and containing 260.000 visits to the feeder were recorded and analysed. On the basis of the duration of the frequency of meals per day two distinct groups, birds with large number of short meals (HM), and birds with a small number of big meals (LM), were identified. Comparisons were made between the two mealtype groups for all traits and the circadian rhythms of feed intake. HM ducks were characterised. by meals of short duration, small size, and short pauses between meals. There was only a small difference between the mealtypes for the total duration of feeding and feeding intensity. HM birds were lighter than LM birds. With increasing age, number of meals per day and duration of feeding per day decreased while daily feed consumption increased. Duration and size of meals, and intensity of feeding increased with age. Changes in feeding activities were most important between 3 and 4 weeks of age. An effect of sex was only noted for the number of meals at seven weeks of age. Differences in the occurrence of pauses were found in response to the experiment, mealtype and age. With regard to the presentation and analysis of pauses the usual procedures using log survivorship functions were discussed. In the present study, however, the differences in the distribution of pauses have been demonstrated using the relative frequency for the pauses of long duration and the cumulative frequency for the pauses of shorter duration. With regard to the high variation of the duration of pauses among individuals and with age the use of a single meal criterion, which separates pauses within and between meals, for all individuals and over all ages was not considered helpful. On the basis of the distribution of pauses of the different mealtypes it seems that the meal criterion is less than one minute in the HM types, and over 35 minutes in the LM types. The circadian rhythm of feed intake was similar in both mealtype groups and shows a biphasic shape with maxima at the beginning and the end of the light period. There was also feed intake during the dark period, the consumption, however, was lower than in the light period. The power spectra of the feed intake showed main components in the 24; 12; 6 and 4.8 hours rhythms and, in addition, various different rhythms in the area of 2 - 4 hours. The rhythm pattern was more pronounced in the HM types as compared to the LM types. The power of the 24 hours rhythms decreased with age, while a 6 hours rhythm developed at the end of the test period. In general it can be said, that there are distinct types of temporal structures in feeding behaviour in ducks. Agedepending changes appear mainly in the early fattening phase up to 5 weeks of age. There is no general meal criterion in ducks which may be applied to all individuals and at all ages. Feed intake underlies rhythmic processes which are modified with age.
