Phenolab: ultra-wide band tracking shows feather pecking hens spent less time in close proximity compared to controls

Hens which feather peck can cause multiple victims in a group. As a result pen-mates could keep a greater distance from feather peckers. Our aim was to relate time spent in close proximity between birds selected divergently on feather pecking. Automatic ultra-wide band location data was obtained by using an active sending tag (UbisenseR) placed in a backpack on the birds. A group of birds was then placed for 15 minutes in a barren test-room equipped with four receiving beacons. We used 37-week old White Leghorn laying hen lines selected for high (HFP) or low feather pecking (LFP), and an unselected control (CON) line (n=76 in total, 5-7 hens per pen; 4 pens per line). Hens were habituated to the backpacks four weeks prior to observations, and were tested for 5-min to the test-room individually one week prior to group testing. Sample rate of tags was set to twice per second, detecting location of each tag separately and simultaneously. Location data was provided by TrackLab software (Noldus, Wageningen, The Netherlands) and compared between tags by Excel calculations. Percentage of time in close proximity was defined as being less than 25 cm apart from another tag, calculated for each bird and averaged per group. Behavioural sampling of feather pecking of 2×20 min at 28-29 wks of age was included on pen-level and used for identification of peckers (>2 bouts/20 min). Percentage of time in close proximity was correlated to feather pecking (pearson correlation), tested to differ between peckers and average of the group (t-test of groups with peckers) and tested to differ between lines (ANOVA including the number of animals per pen as covariate). Feather pecking on pen-level was not correlated to proximity measures (r=-0.28, P>0.10), but birds spent less time in close proximity of the feather pecker as opposed to time spent on average with other pen-mates (17.6・}1% vs 21.6・}1.2, t16=-1.18, P=0.04).