2.5.2. Experiment 2
2.5.2.1. Effect of main factors (wheat-bran litter and PSA pad). The
effect of wheat-bran litter distribution was analysed on data from
T1 to T4 treatments. DB bout duration and number of hens in
PSA, in the middle area, and in nest area were analysed with an
ANOVA model where fixed effects were wheat-bran litter and nest
pad. Interactions between wheat-bran litter and nest pad were
systematically tested. DB behaviour from continuous observations
(number of DBs in the cage and in the PSA, proportion of DB performed
in PSA, in the middle area, and in nest area, proportion of
DBs by the trough in PSA and in the others areas (all together))
were analysed with a stratified Wilcoxon–Mann–Whitney test (litter
stratified by nest pad type) as data were summed at the cage
level.
The effect of PSA pad type was analysed on data from T0 and
T3 treatments. DB bout duration and number of hens in PSA, in the
middle area, and in nest area were analysed with ANOVA model
where fixed effect was PSA pad. DB behaviour from continuous
observations (number of DBs in the cage, in the PSA, proportion
of DB performed in PSA, in the middle area, and in nest area, proportion
of DBs by the trough in PSA and in the others areas (all
together)) were analysed with a Mann–Whitney U test.
2.5.2.2. Distribution of behaviours and hens between each area of the
cage. Similarly to the Experiment 1, number of DB and DB bout
durations were compared between areas of the cages thanks to
an ANOVA model where fixed effect was area. Also the number of
hens observed in each area (data from scan) was compared to a
theoretical allocation by a Z-test.
All data analysed with ANOVA are expressed in tables as
least squares means ± standard errors. Data analysed with nonparametric
tests are expressed in tables as medians [Quartile 1,
Quartile 3]. Data analysed with Z-test are expressed as mean [CI
95%].