Network characteristics and development of social structure of agonistic behaviour in pigs across three repeated rehousing and mixing events
Publication Type: |
Journal Article |
Year of Publication: |
2015 |
Authors: |
Kathrin Büttner, Katharina Scheffler, Irena Czycholl, Joachim Krieter |
Publication/Journal: |
Applied Animal Behaviour Science |
Keywords: |
agonistic behaviour, group level, pigs, social network analysis |
ISBN: |
0168-1591 |
Abstract:
In commercial pig production, the mixing of unacquainted pigs is a standard procedure which leads to agonistic interactions with a wide range of individual pig behaviour. A reduction in these agonistic interactions positively affects animal health, welfare aspects and production parameters. With the help of social network analysis, insight into animal societies can be gained and structures of agonistic behaviour can be detected. The aim of the present study was to analyse how the general network properties (density, clustering coefficient, weakly and strongly connected components and centralisation) develop over three different rehousing and mixing events (weaned pigs, growing pigs and gilts). Therefore, the agonistic behaviour of the animals in the flatdeck pens, the growing and the breeding stables were recorded directly after mixing (in total 17 h of video observation for each pen). In total, 65 groups in the rearing area (flatdeck pens), 24 groups in the growing stable and 12 groups in the breeding stable were recorded. In contrast to the second and the third rehousing and mixing events, the results for the density and the clustering coefficient showed that in the agonistic interaction networks of the first mixing event the animals were very densely connected and therefore formed highly connected clusters, indicating that nearly all animals in a pen fought against each other during the observation period. The results for the weakly and strongly connected components suggest that agonistic interactions have a group-wide structure which is not immediately obvious in a dyadic analysis. Furthermore, the in-degree centralisation showed significantly smaller values than the out-degree centralisation at all three observed age levels, which suggests that the distribution of individuals initiating agonistic interactions was more skewed than the distribution of individuals receiving agonistic interactions. By quantifying the important aspects of group structure in different production stages, information based on social network analysis could help identify and understand the formation of behavioural patterns, e.g. agonistic interactions, tail biting and access to resources.