Quantitative data analysis
Lesion size (cm) was analysed using a linear mixed model, with fixed
effects for Treatment, Day, Treatment × Day, Age, Breed and Sex, and a
random effect for Animal ID. Model fitting was via the lmer function in
the lme4, and model-based means obtained using the emmeans package.
Event duration data (number of days that cattle are immobile, walk with
difficulty, walk normally, on pasture, and salivating) were analysed
using linear models with fixed effects for Treatment, Age, Breed and
Sex. Due to the positive skew, a loge (y +
1) transformation was applied. However, with the large number of ‘zero’
durations, hypothesis testing was conducted with permutation tests
rather than F -tests, using the aovperm function in the permuco
package (Frossard & Renaud, 2019) in R.