Group housing of gestating sows is a development following an alteration of the animal protection law in August 1986 and this became even more popular after the introduction of the pig-housing decree (SchhVo) in May 1988. So, yet again, the group housing system has become an alternative to the existing single housing system. The technique of individual, automatic control of feeding and housing using electro-technology and identification of individual animals at centralized transponder feeding stations Ieads, however, to serious problems for the sows, as shown by investigations on the effects of some on animal behavior. This is because centralized feeding areas working against the natural feeding habits and social behavior of these animals.
This paper reports on the project set up to analyze the social behavior of gestating sows in group housing with transponder feeding. Investigations were carried out to see what influence the transponder feeding station has on the animal behavior and social rank-order in a group of 18 sows. In particular, beasing in mind the pig protection law, it was tested whether strewing of straw in the berth area instead of a Ievel cement area influenced animal behavior at the transponder feeding station. The quantitative evaluation of observed behaviors was carried out according to the following criteria: "Parallel/inverse-parallel pressing", Head-Body blow", " Head-Head blow", "Biting activities", "Nose-Nose contact", "Nose-Body contact", "Nosegenital contact", lowering head", and "Withdrawal" as well as occupations of station with and without access to food.
The results of the trail are summarized as follows:
Video-recording technique proved to be an ideal method of observation. Behavioral data such as frequency and duration of the criteria required can be constantly recorded in this way. After the ranking order in the group of sows has been established, an observation period of 24 hours is enough to obtain sufficient, exact results.
In the experiment with straw there was a daily pattern of a phase in main activity between 17.00 and 0.00. This progressed after the beginning of the feeding cycle and reached a maximum at 21.50 with 34 social, repulsive conflicts pro ten minutes. Thereby, each visit to the station was associated with a drastic rise in the frequency of social, repulsive confrontations. The change to straw less housing lengthened the phase of main activity by an hour and resulted in an increase in the frequency of social, repulsive conflicts - 64 pro ten minutes 88,2 % during this time. The total period of time in wich sows stood in contact with one another in housing with straw was 152 minutes. This increased to 266 minutes after changing to housing without straw. During this time, each sow was involved, on average, in 82,8 and 98,6 social, repulsive conflicts daily. Most conflicts lasted not longer than a second. The duration and frequency of conflicts significantly increased after removal of straw.
By far the most number of daily social confrontations (312 and 248) took place in both experiments at the entrance area to the station. The most aggressive fights took place there - 95 and 117 times a day. The aggression index in the berth area had a value of I = 0,554, and rose at the station to I = 0,730.
In housing with straw, the number of social activities declined from 624 to 351 after all sows had eaten at the station till the end of the feeding cycle. On changing to housing without straw there was only a reduction from 658 to 495 social, repulsive conflicts, whereby in the berth area an increase from 63 to 165 social, repulsive conflicts took place over both periods of comparison. In housing with straw, there were no more apparent differences observed.
The presence of straw can significantly reduce the frequency of occasions of false occupation from 1,8 ± 0,6 to 1,1 ± 0,4 pro animal and day. Thereby, the duration of the feeding cycle is reduced by three hours. The sows required, on average, 13,8 ± 0,8 minutes to eat. False occupation of the station lasted 3,7 ± 0,8 and 2,9 ± 0,4 minutes. This means that the feeding station was occupied for about 42% of the access time. On average, each sow needed 2,45 and 1 ,84 hours to reach the station. Sows at the lower end of the ranking order needed longer and, thence, were more often involved in social confrontations. Lower-ranked sows accepted their position in the eating time-table only to an limited extent. The sows eat in succession according to their social order of rank (rs = 0,654) whereas the dependence on order of rank and eating row in housing without straw was more strongly pronounced (rs = 0,725). The sows only showed aggressive behavior if they were unable to satisfy their need in time. Presence of straw reduced the number of waiting sows in front of the station at each animal change-over, wich led to a prevention of an unnecessary high concentration of animals at the station entrance area.
The setting of the social ranking order took place mainly at the transponder feeding station, where social conflicts happened most often. However, there were individual preferences in the berth and station area. Sows which were inferior at the feeding area were quite able to be superior in the berth area. Considering the animal-specific factors: Weight, Frequency of littering and length of gestation - only weight had an influence on the social position of the sows in the hierarchy (rs = 0,227). This was of significant importance in the housing without straw (rs = 0,448).
Following the ethological studies of behavior of sows at a transponder feeding station, it was noticed that, under different housing conditions, the station can have a negative technological effect on animal behaviors. For these reasons, recommendations for the improvement of the technical construction as well as housing conditions are to be aimed for. These improvements, when taken into practice, can contribute to further success of production in sow holding. However it is conceivable that in the future, for reasons of the serious disadvantages of this form of group housing that other housing systems ag that of "biological fixing" (HÖGES 1990, HURNIK 1990) in gestating and suckling pigs will become an object of investigation «