The Biologically Relevant Unit for the Analysis of Short-term Feeding Behavior of Dairy Cows

To determine whether visits or meals are the most biologically relevant unit of short-term feeding behavior, we analyzed 209,025 records of visits to feeders by 37 cows. Two feeds were used that differed in protein content. Cows were divided into control groups for the low and high protein feeds and a choice group that had access to both. Daily number of visits and intake per visit were very variable. Cows fed low protein feed had lowest daily intakes, but single-visit characteristics were poorly correlated with daily intake. The probability of cows ending a visit did not change greatly with visit length. Log-normal models were used to estimate individual meal criteria (44.7+/-2.1 min), and visits were grouped into meals. Meal duration (36.9+/-1.3 min) and daily number of meals (6.1+/-0.1) were not affected by treatment. Feeding rate and intake per meal were lowest for cows fed low protein feed. Meal size decreased systematically during the day. The probability of cows ending and starting a meal increased with meal length and interval between meals, as predicted by the satiety concept. Meals are, therefore, a biologically relevant unit of short-term feeding behavior and visits are not.