The ingestive behavior patterns were evaluated using nine sheep 1/2 Texel+1/2 Ile-de-France over a six-month period. Sheep were fed the diet at 9 a.m. and 4 p.m., which was composed by 250 g of concentrate (15.45% CP, 36.54% NDF) and ad libitum grass hay (6.69 % CP, 69.10 % NDF). The ingestive behavior was continuously measured by five days per period, and its activities or states were classified as eating, ruminating or idling, and sequentially registered. The time series was necessarily transformed in discrete variable, in the probability estimates, which incorporated the probabilities of being in a given state, of staying in that state and of changing from a state to another one. The probabilities were estimated at different times without assuming the stationary presumption. The 5 minutes time sampling interval was considered the most adequate as compared to 7 and 10 minutes, with less information lost especially in relation to the changes of states. The probability values of being or staying in a given state were higher than transition state probabilities. Measurement periods showed significant differences in overall mean probabilities values partially related to changes in voluntary intake due to photoperiod effect.
animal behavior; changes of state; time scale
