Training for a cognitive judgement bias task does not affect fear or telomere shortening in laying hens

@deFor behavioural tests such as the cognitive judgement bias task (JBT), animals usually undergo weeks of training involving various elements, such as handling, social isolation and learning the task. These may influence their affective state and other welfare indicators. Here, we investigated the effects of training for a JBT on fear-related behaviour in laying hens and a potential welfare indicator - telomere shortening. Hens were trained for a JBT (N = 16, across 2 batches), or kept as controls (N = 16, across 2 batches) that received no training or trainingrelated handling. Hens that reached the learning criterion (N = 14) were tested in six JBT test sessions. Before and after the trained hens received JBT training, blood was collected from all hens, and they were assessed for fear-related behaviour by being subjected to open field, human approach, and tonic immobility tests. Relative telomere lengths (RTL; i.e., ratio of telomeric repeats versus a single copy control) were obtained from blood samples with a quantitative real-time polymerase chain reaction (PCR) method. As anticipation of the JBT sessions could affect the JBT outcome, we also recorded behavioural indicators of anticipation prior to each JBT test session. Except for an increased latency to first step in the open field test after training in the trained hens (t = 3.1, df=28, P = 0.012; conditional R2=0.59), no other fear-related behaviour significantly differed in either group (P > 0.05) and RTL was not affected. In addition, none of the anticipation indicators predicted the JBT test outcome. Contra to our hypotheses, welfare-related aspects may not be strongly affected by the JBT training. However, the enriched housing of the hens from both treatments might have minimised the effect of training on fear and telomere shortening, or training may have affected indicators other than the ones we investigated. We conclude that because JBT training and testing does not affect fear assessment and telomere shortening outcomes in hens housed in enriched housing conditions, the same individuals can be used in welfare assessments involving similar assessments. Our results also suggest that the JBT itself may be robust against the temporary affective states associated with anticipation. @enFor behavioural tests such as the cognitive judgement bias task (JBT), animals usually undergo weeks of training involving various elements, such as handling, social isolation and learning the task. These may influence their affective state and other welfare indicators. Here, we investigated the effects of training for a JBT on fear-related behaviour in laying hens and a potential welfare indicator - telomere shortening. Hens were trained for a JBT (N = 16, across 2 batches), or kept as controls (N = 16, across 2 batches) that received no training or trainingrelated handling. Hens that reached the learning criterion (N = 14) were tested in six JBT test sessions. Before and after the trained hens received JBT training, blood was collected from all hens, and they were assessed for fear-related behaviour by being subjected to open field, human approach, and tonic immobility tests. Relative telomere lengths (RTL; i.e., ratio of telomeric repeats versus a single copy control) were obtained from blood samples with a quantitative real-time polymerase chain reaction (PCR) method. As anticipation of the JBT sessions could affect the JBT outcome, we also recorded behavioural indicators of anticipation prior to each JBT test session. Except for an increased latency to first step in the open field test after training in the trained hens (t = 3.1, df=28, P = 0.012; conditional R2=0.59), no other fear-related behaviour significantly differed in either group (P > 0.05) and RTL was not affected. In addition, none of the anticipation indicators predicted the JBT test outcome. Contra to our hypotheses, welfare-related aspects may not be strongly affected by the JBT training. However, the enriched housing of the hens from both treatments might have minimised the effect of training on fear and telomere shortening, or training may have affected indicators other than the ones we investigated. We conclude that because JBT training and testing does not affect fear assessment and telomere shortening outcomes in hens housed in enriched housing conditions, the same individuals can be used in welfare assessments involving similar assessments. Our results also suggest that the JBT itself may be robust against the temporary affective states associated with anticipation.

Veterinärmedizinische Universität Wien
Universität für Bodenkultur