Epsilon Open Archive

Abstract

This thesis comprises the results from three separate studies performed in the experimental automatic milking system at Kungsängen Research Centre, Swedish University of Agricultural Sciences, Uppsala, Sweden. In the first study, 30 high-yielding cows in early lactation were subjected to two different degrees of controlled cow traffic, and the effects on milk yield, dry matter intake, feeding patterns and voluntary visits to the milking unit and the control gates were measured. A model of mixed distributions for estimations of biologically relevant meal criteria from registrations in roughage stations was also evaluated. In the second study, the behaviour of 24 cows after they had been redirected in control gates was observed, and the cause of long redirection times from gates until they showed up in the milking unit was examined. In the third study, 9 cows were subjected to three different cow traffic systems in a carry-over design and the effects on cortisol concentrations in milk and ruminating patterns were studied. The studies showed that milking frequency and thereby milk production can be altered by different time settings in the control gates without limiting the daily feed intake of the cows. A high degree of guidance provokes social effects in the queue in front of the milking unit and in the feeding areas. It also makes it difficult for the cows to follow their natural feeding patterns. Judging from measurements of milk cortisol concentrations, controlled cow traffic was not stressful for the cows. Cows initiated meals with short intervals, which offered many opportunities to milk them. But the queue in front of the milking unit caused long redirection times, and the control gates failed to guide cows to high milking frequencies. Individual differences in feeding patterns and how cows respond to redirections in the control gates suggest that the control gates should be making decisions on an individual level.