One tool to rule them all - Automatic detection and analysis of group dynamics in turkeys (Meleagris gallopavo)

• Applicants: Dr.-Ing. Max Mehltretter; Professorin Dr. Jenny Stracke
• Subject Area: Animal Breeding, Animal Nutrition, Animal Husbandry ; Geodesy, Photogrammetry, Remote Sensing, Geoinformatics, Cartography
• Term: since 2025
• Project identifier: Deutsche Forschungsgemeinschaft (DFG) - Project number 566328535

Welfare of farm animals is a declared goal of sustainable, modern and socially accepted animal husbandry. Social dynamics within a group play a major role regarding the performance of a group, and, with respect to animal husbandry, for individual`s well-being. The composition of different individual personalities might play a crucial role for the overall performance. In turkeys, with group-housing being the standard husbandry system, there is neither scientific evidence on the effect of different personalities on group performance, nor on the indicators describing quality of beforementioned. One reason for this lack of information is, that the analysis of behaviour requires a considerable manual effort, thus, limiting the number of individuals that can be observed as well as the time of observation and the complexity of the behaviour model. With the proposed project we aim to address this issue by developing a methodology to automatically extract information on position, body pose and behaviour of individual birds over time based on stereo image sequences. This method will be used to analyse the effect of different group compositions (based on personality types) on group dynamics, focusing on behavioural synchronicity and the automatic detection of specific qualitative behavioural patterns. The intended goal is to define the optimal composition of individuals in terms of animal welfare for turkey husbandry in order to improve animal welfare.

• DFG Programme: Research Grants