An Assessment of Monil virtual fence system on sheep
ALURES NTS Database ID: NTS-NO-707771
Etterevaluering
The Norwegian Food Safety Authority will retrospectively assess the current experiment. You are therefore required to submit necessary information making it possible for us to evaluate 1) whether the objectives of the project were achieved, 2) the harm inflicted on the animals, and 3) any elements that may contribute to further implementation of 3 R in future studies. This information must be submitted via FOTS.
Begrunnelse for etterevalueringen
The experiment successfully met its objectives, which included evaluating learning ability and welfare while comparing two audio cues and the effects of wool treatment on the Monil virtual fence system used for sheep. The study demonstrated that ewes effectively learned to avoid restricted zones, relying primarily on auditory cues rather than electrical cues. The acute stress induced by the system was brief, and there were no indications of chronic behavioral stress. For the audio cue treatment, the results indicated no significant effect, meaning that whether the audio frequency changed over the time played or remained constant with variations in tone length did not influence the ewes' ability to learn the system. Additionally, the wool treatment showed that shaved ewes were more compliant due to less insulation.
The functionality of the collars was crucial, if the collars malfunctioned, it led to the animals ignoring auditory cues and breaching boundaries.
Recommendations include improving collar reliability and using the system on shaved sheep to ensure effectiveness and welfare. With these adjustments, the Monil virtual fence system can be effectively used for sheep without compromising animal welfare. These findings provide valuable insights for advancing virtual fencing technologies.
In the context of using virtual fences for managing livestock, potential welfare issues may arise if the animals are unable to gain control over the system. In the study, the animals learned to navigate the virtual fence system effectively and relied more on auditory signals through the progression of the experiment (controllability and predictability).
The stress caused by the virtual fence cues was brief, with both behavioral and physiological metrics returning to baseline levels within five minutes. Additionally, auditory cues imposed significantly less stress on the animals compared to electric cues. Importantly, the study did not identify any signs of extreme fear or chronic stress in the animals.
The refinement measures implemented during the experiment focused on minimizing stress and severity for the animals. Key points included conducting a training phase near the farm unit to allow for immediate intervention and providing a designated safe area within the arena. Training of the animals to the virtual fence system should be conducted in groups. Assisted introduction to the virtual fence boundary and pressure mark assessments are important during the initial days to monitor animal welfare.
Severity was already very low. Ensuring that the animals effectively learn to navigate the fence and rely more on auditory cues, paired with good and constant GPS coverage, are, however, points that we discussed during the analysis of our results. Shearing the animals' wool to allow freer collar movement and reduce pressure marks was also highlighted as beneficial.
Video surveillance and pressure mark assessments were identified as more informative for monitoring welfare compared to direct observation time budget methods and heart rate monitors.
To improve the training phase, the study recommended conducting the assessment on an open landscape without dense canopy cover.
The functionality of the collars was crucial, if the collars malfunctioned, it led to the animals ignoring auditory cues and breaching boundaries.
Recommendations include improving collar reliability and using the system on shaved sheep to ensure effectiveness and welfare. With these adjustments, the Monil virtual fence system can be effectively used for sheep without compromising animal welfare. These findings provide valuable insights for advancing virtual fencing technologies.
In the context of using virtual fences for managing livestock, potential welfare issues may arise if the animals are unable to gain control over the system. In the study, the animals learned to navigate the virtual fence system effectively and relied more on auditory signals through the progression of the experiment (controllability and predictability).
The stress caused by the virtual fence cues was brief, with both behavioral and physiological metrics returning to baseline levels within five minutes. Additionally, auditory cues imposed significantly less stress on the animals compared to electric cues. Importantly, the study did not identify any signs of extreme fear or chronic stress in the animals.
The refinement measures implemented during the experiment focused on minimizing stress and severity for the animals. Key points included conducting a training phase near the farm unit to allow for immediate intervention and providing a designated safe area within the arena. Training of the animals to the virtual fence system should be conducted in groups. Assisted introduction to the virtual fence boundary and pressure mark assessments are important during the initial days to monitor animal welfare.
Severity was already very low. Ensuring that the animals effectively learn to navigate the fence and rely more on auditory cues, paired with good and constant GPS coverage, are, however, points that we discussed during the analysis of our results. Shearing the animals' wool to allow freer collar movement and reduce pressure marks was also highlighted as beneficial.
Video surveillance and pressure mark assessments were identified as more informative for monitoring welfare compared to direct observation time budget methods and heart rate monitors.
To improve the training phase, the study recommended conducting the assessment on an open landscape without dense canopy cover.