Laying hens selected for low mortality: Behaviour in tests of fearfulness, anxiety and cognition
Publication Type: |
Journal Article |
Year of Publication: |
2011 |
Authors: |
Rebecca E. Nordquist, Jasper L. T. Heerkens, T. Bas Rodenburg, Sabine Boks, Esther D. Ellen, F. Josef van der Staay |
Publication/Journal: |
Applied Animal Behaviour Science |
Keywords: |
cannibalism, chicken, feather pecking, genetic selection, holeboard, open-field test, voluntary approach test |
ISBN: |
01681591 |
Abstract:
Feather pecking and cannibalism in chickens can lead to injury or to death of the victims, and is thus a serious welfare and economic issue in modern poultry farming. A sib selection scheme has been initiated to genetically select a low mortality line (LML), which shows decreased mortality due to cannibalism compared to a control line (CL). To determine whether undesired behavioural effects such as impaired cognition, increased fearfulness, or increased anxiety might have inadvertently been co-selected, we tested hens from the fourth generation of selection (n=9 per selection line) in several behavioural tasks. When tested in a 122cm×122cm open field at age 7 days, the lines showed no differences in locomotion or vocalization. In a T-maze test in which the chickens could navigate to find conspecifics, testing sociality and fearfulness, 12 to 16-day-old CL chickens showed a lack of exploratory behaviour; they did not leave the start box. In contrast, most LML chickens negotiated the maze, and approximately half of them found and stayed close to their conspecifics. This difference points toward higher levels of fearfulness in the CL than LML. In a voluntary approach test assessing fearfulness for humans, conducted when the chicks were 26 days old, the LML approached a familiar human faster, thus displaying lower levels of fearfulness. The same birds were tested in a holeboard test at an age of 25–65 days, the first time this test of spatial memory has been used in an avian species. Our results demonstrated high levels of working memory performance and low levels of reference memory performance in both lines, with no differences between the lines. Overall, the present results indicate that unwanted behavioural effects were not co-selected with selection on low mortality, and support the feasibility of the use of the LML in farming practice.