Locomotor activity of dairy cows in relation to season and lactation

• The results represent the activity of cows from a commercial dairy herd.
• Parity, stage of lactation and season had a significant impact on cows activity.
• It is recommended that behaviour pattern is put against these factors.

All systems recording activity of cattle measure it independently of environmental and animal factors. Such recordings may hide differences in activity caused by extraneous interferences or may be inaccurately interpreted. The information produced by sensors could be more accurate, if behaviour pattern is put against other factors affecting activity. The objective of the study was to verify, if information obtained from activity sensors could be improved, by including additional factors such as parity, stage of lactation and season of year. Activity of 132 Polish Holstein-Friesian cows was continuously monitored between April 2011 and August 2012 using IceQube Sensors. Generalized linear models were used to analyse activity parameters of dairy cows. The results showed that parity, stage of lactation and season had a significant impact (P < 0.01) on cows activity. Total lying time decreased with consecutive lactations, whereas older cows preferred longer and less frequent lying bouts. Decreased lying time, higher number of steps and less number of lying bouts during the first 30 days after calving were observed. Including this fact of restlessness behaviour after calving in data from sensors would prevent false conclusions. In winter cows spent more time lying down than in other seasons. Therefore lying time increased over 650 min in summer provided different information (for example may be a sign of a lameness) than in winter (which is the normal occurrence). The results of the study represent the activity of cows from commercial dairy herd. Activity parameters were highly affected by the stage of lactation, parity and season of year, therefore data obtained from activity sensors should be put against these factors. It is highly recommended to exclude these biasing factors from software analysing activity parameters, to improve the information obtained from sensors used in herd management.