In structured livestock populations like cattle herds, movements between groups of animals often depend on age or physiological status of animals, thus on the time spent in a group. In epidemiological models, transitions between groups are usually assumed Markovian (independent on the time spent in the group). Our objective was to study the effect of a transition between groups under a triangular distribution of the transition time on the pathogen spread within a structured livestock population, compared to the effect of a Markovian transition under an exponential distribution. A SIR stochastic model was defined for a population separated into youngstock and adults. Due to the separation into groups, an heterogeneity of the pathogen transmission was assumed: direct transmission within each group and indirect transmission between groups The pathogen spread after an introduction of a newborn infected animal was simulated both with an exponential (Markov) and a triangular distribution for the youngstock to adults transition time. These two distributions may lead to significant differences in the epidemic size both in the youngstock group and in the adult group. But, at the population level, the epidemic sizes were not significantly different. The persistence of the pathogen was not significantly different.