2.5. Statistical analyses
Two animals (one each from ALC and UPWCW) were
removed from the experiment for reasons unrelated to the
dietary treatments, resulting in data from 68 animals being
used. The data were subjected to analysis of variance using
the general linear model procedures of SAS (SAS Institute,
2005). The model had terms for block (d.f.=13) and treatment
(d.f.=4). Data from the animal behaviour component of the
study were analysed using Chi-square analysis and treatment
comparisons of LSMEANS made using the proc glm procedure
in SAS. In vivo digestibility data were analysed separately
for un-supplemented and supplemented cattle using the
general linear model procedures of SAS with each latin square
model containing terms for period (d.f.=3) animal (d.f.=3)
and treatment (d.f.=3). In all cases, the sums of squares due to
treatment were analysed using the ‘pdiff’ statement in SAS.
The data are presented as the means for the experimental
treatments with the appropriate S.E.M.