Effects of changing milking and feeding times on the behaviour, body temperature, respiration rate and milk production of dairy cows on pasture

Management strategies to reduce heat stress are needed, especially when there is no or little shade in pasture -based dairy systems. We investigated management practices used to reduce heat load in summer: milking later when it is cooler, feeding later, and milking only in the morning. Fifteen groups (n = 4 pregnant Friesian -cross cows/group) were managed on pasture and milked at 0700 h followed by a new pasture allocation including silage and one of five afternoon/evening treatments (n = 3 groups/treatment): 1) Late milking (1935 h)/early feed (1630 h), 2) Late milking (1935 h)/late feed (2015 h), 3) Early milking (1550 h)/early feed (1630 h, control), 4) Early milking (1550 h)/late feed (2015 h), 5) Once-a-day milking (OAD): cows were milked only in the morning and provided feed at 1630 h. Lying, grazing and ruminating were recorded using validated accel-erometers over 25 d (mean temperature: 19 degrees C, range: 5-32 degrees C). Body temperature (BT) was recorded using vaginal temperature loggers and respiration rate (RR) was recorded manually. Individual milk production and water intake (group level) were recorded daily. Data were analysed using linear mixed models with group as the experimental unit (n = 3/treatment). There was no evidence of an overall treatment effect at the 5% significance level for grazing, ruminating, lying time, BT or RR, however, the diurnal pattern varied among treatments (grazing, ruminating, lying, BT: P < 0.001). Typically, cows spent more time grazing after fresh pasture was provided followed by rumination, however, this may have resulted in an observed numerical reduction in lying time for cows with delayed milking and/or feeding. Control cows had numerically the highest RR in the after-noon and OAD cows and cows with delayed feeding the lowest. Water intake tended to differ between treatments (P = 0.060); cows milked early in the afternoon consumed most water. Cows with delayed milking had the greatest peak in BT in the afternoon/evening, whereas cows with delayed milking and feeding, and OAD cows had the lowest BT in the afternoon/evening (P < 0.001). Milk production declined over time, which is normal in summer, however, this decline was numerically lowest for cows with delayed milking and feeding. In summary, OAD cows may have experienced reduced heat load. Delaying milking and feeding and milking OAD reduced the afternoon peak in BT that is associated with walking into milking. Modifying milking and feeding times can be used to change diurnal patterns of behaviour but more information is needed to understand what this means for cow welfare.