The maintenance of high standards of animal welfare is critical issue facing all livestock industries. This is considered both an ethical and financial issue, with a clear need for both production under morally acceptable grounds as well as a growing impact of perceived welfare standards on consumer buying behaviour. Furthermore, this concern is expected to grow in the coming years, with a push for increased productivity for food security reasons resulting in intensification of existing systems. The application of digital technologies has been heralded as the next revolution in farming, allowing for autonomous remote monitoring of livestock and increased productivity. The use of remote sensors for welfare monitoring has also been indicated as a potential benefit, however thorough research of the practical applications is lacking. The aim of this project was to identify the potential for on-animal location and motion sensing technologies to identify lambing behaviour and birth events. From here, the extension of the technology to identify potentially adverse events was studied. Lambing was chosen as a case study due to the obvious welfare implications for both the ewe and lamb.

A field trial was conducted in the Canterbury region of New Zealand over a two week period from September to October 2017. Forty mixed-aged ewes were fitted with GPS tracking collars and accelerometer ear tags, set to record at 3 minute intervals and 12.5Hz, respectively. In addition, the animals were observed for approximately 8 hours daily to record general flock behaviour and individual birth events. Of the 40 ewes, 25 lambed during the experimental period (13 twin bearing and 12 singles), producing 33 live lambs. One of the twin-bearing ewes prolapsed during labour and was unable to give birth prior to being humanely euthanized. A further ewe had scanned as twins though only produced a single lamb. Two lambs, both from separate twin-bearing ewes, died shortly after birth.

Preliminary data analysis suggests there is a change in movement behaviour of ewes on the day of lambing which continues for some time after birth. This is evident through mapping of the hourly distance travelled by ewes and movement metrics recorded by the accelerometer. Whilst it was expected that birth events would be evident by the ewe separating herself from the flock, the visual observations did not support this. However, this will be further explored through examination of distance to closest peer during the time of birth. The behaviour of the animals that lost lambs after birth and the ewe that prolapsed will be examined closely to identify potential data signatures that may indicate this adverse experience. These results could then be used to assist in the development of welfare measurement systems.