Estimating feed intakes of individual cows in a loose-housed system is difficult and unreliable. It is known that estimating intakes from the number of bites taken at grazing is unreliable. Feeding from a total mixed ration (TMR) is likely to have fewer confounding variables. All cows were over 30 months old, of three breeds: Holstein Friesian, Estonian Red and Estonian Native. There were 30 feeding bins and 66 cows. Cows were observed over the whole 24 hour period, from 1st July to October 22nd 2011, in Märja experimental farm, of the Estonian University of Life Sciences, and data from 37 cows were collected. Measures taken included: time spent at the feed bin, number of bites taken at each visit, weight lost from the bin, and evidence of stealing. Mean time spent at each visit to the feed bin with standard error was 4.4 + 0.68 minutes. The mean number of bites per visit was 7.1 + 1.01. There was a significant correlation between the number of bites and total time cows spent feeding (p is less than 0.001), with a regression R2 value of 79.8%. No correlation was found between the number of bites and feed removed at each visit (p=0.851) nor between the time per visit and amount of feed removed (p=0.681). Therefore, there is no evidence that either the number of bites taken per visit or the time spent at bin per visit is related to intakes. There was no effect of age. Stealing was evident and widespread.

The research was conducted with the aim of studying the possibility of using automatic feeding stations for individual sheep (Ovis aries) feeding. The research was organized in 3 periods, in which 10–13 months old sheep – female only (1st research period), 6–10 months old sheep – female only (2nd research period) and 8–12 months old sheep – female only (3rd research period) were used. Part of the sheep did not visit the feeding station at all. Sequentially, over the research periods the percentage was 13%, 37% and 49%. Upon setting the concentrate daily ration 400 g per sheep, on the average the number of visits to the particular feeding station was 6.1 and 4.7 (sequentially in the 1st and 2nd research period). Upon increasing the concentrate ration amount up to 700 g per day, the average number of feeding station visits was 6.9 (in the 3rd research period). The number of visits to the automatic feeding station was varied (V is greater than 10%). Over the research period live weight of the sheep increased significantly (p ≤ 0.05) for the subgroups of animals which made visits to the feeding station regularly during the entire research period. In all research periods, the average daily concentrate intake (373.5 g in the 1st research period, 333.1 g in the 2nd research period and 581.2 g in the 3rd research period) was 93%, 83% and 83% respectively, of the maximum established feeding standard.

In automatic milking systems, where the concentrate and forage components of the ration are offered to the cows separately, lack of control over intakes can result in difficulties balancing the forage and concentrate portions of the diet, leading to problems associated with high concentrate intakes and concomitant low forage intakes. In order to check this as a problem on a dairy unit, the feeding behaviour of a sample of cows was observed by video recording. As a pilot study, four Holstein Friesian cows (two at the highest yield and two at the lowest yield of the milk production range) were selected from sixty lactating cows on the Estonian University of Life Sciences’ farm near Tartu, Estonia. The study took place from May 18th till November 4th 2014. The cows were robot-milked and fed a ration comprising, separately, concentrate feed from a robot and a feeder, and a grass/clover silage mix forage at the feed barrier. With the low number of samples the results are indicative and descriptive, but it appears from the raw data that individual variation in visiting times and times spent at the feed barrier are greater than the effect of level of production. Cows spent a significant portion of their time idling at the feed barrier, not actively feeding. It is concluded that care should be taken to presume behaviour from positional data, and there is no evidence that cows with higher and lower milk yields are differentially motivated to feed from a forage source.