Probability of Transition in Behaviour of Grazing Dairy Cows

Abstract

Improving our knowledge of the pasture-animal interface will help farmers make better decisions on the timing of allocation of new grazing areas and the type and amount of supplements offered to grazing dairy cows. Understanding why and what motivates grazing dairy cows to change their grazing behaviour will also help improve these decisions. Hence, improving feed allocation systems to maximise intake of pasture and supplements more efficiently. Dutilleul et al. (2000) calculated the time-dependent transitions for chewing behaviour in sheep using transition probabilities at each sampling time point. To determine the time-dependent probabilities (TDP) of grazing dairy cows, data collected in an experiment designed to investigate the effects of sward height (SH, 10 v 13cm) and grazing duration (GD, 1, 2, 4, 8 and 15h) on grazing behaviour (Dobos et al. 2009) was used. Briefly, 6 dairy cows were randomly allocated to one of 2 SH x 5 GD treatments on kikuyu pastures over a 15h grazing period (1600h to 0700h) for 3 days (replicates). Cows were observed at 20-minute intervals for the first 2 hrs after entering their experimental paddocks at 1600h and thereafter at 30 minute intervals until 0700h the next day. Within 'time-dependent transition probabilities' are incorporated three types of probabilities: the probabilities of (1) being in a given state, (2) staying in the same state, and (3) changing of state, although being in a given state is not a transition per se. These probabilities are calculated at multiples of a 'sampling interval' (Δt) in discrete time, from the raw data collected in continuous time. The grazing TDP of cows in the 10cm SH treatment was similar to their ruminating and idling/resting TDP (0.37 grazing v. 0.31 ruminating v. 0.32 idling). However, the grazing TDP of cows in the 13cm treatment was almost twice that of them either ruminating or idling/resting (0.49 grazing v. 0.25 ruminating v. 0.26 idling). Therefore, 13cm SH treatment cows are more likely to be grazing than either ruminating or idling. The total time spent grazing for these cows was found to be 45min longer (P<0.001) than the 10cm SH cows and they had stabilised their grazing behaviour within the first 2h of grazing, while the 10cm SH treatment cows stabilised within the first 4h (Dobos et al. 2009). Cows in the 10cm SH treatment were twice as likely to move from ruminating to grazing as those cows in the 13cm SH treatment (0.12 v. 0.05). The probability of the herd moving from idling to grazing and its converse for both SH treatments was very unlikely as the length of the idling session progressed. An interesting result was the relatively low probability of transition from ruminating to grazing for both SH treatments. This could have implications on what mechanism initiates grazing.