PSVI-16 Conditioning Dairy Cows to a Virtual Fencing System Is Compatible with Animal Welfare

Abstract

Abstract Virtual fencing (VF) enables livestock grazing without physical fences by conditioning animals to an audio tone (AT) followed by an electric pulse (EP) that is up to 25 times weaker than conventional electric fencing (EF). However, the operating principle of a virtual fencing system raises concerns about animal welfare. Therefore, the present study investigated the learning process of dairy cows in a virtual fencing environment and its effects on animal behavior and welfare. Twenty lactating cows were divided into four groups of five individuals each (2x VF; 2x EF), balanced by lactation stage (203 ± 62 days) and age (3.7 ± 1.5 lactations). All cows were accustomed to daily grazing with EF but were naïve to VF. Each group grazed in a separate paddock for 3 days of acclimation (P0) and 21, 14, 14, and 7 days of experimental treatment (P1-4). All paddocks were electrically fenced, and of similar vegetation structure and botanical composition. During P1-P4, a virtual boundary was activated within each paddock of the VF groups, and a second electric fence was set up within each paddock of the EF groups to maintain comparable paddock sizes. All cows were fitted with a VF collar (Nofence AS, Batnfjordsøra, Norway) and an IceQube pedometer (Peacock Technology Ltd, Stirling, UK). The sensors continuously tracked cow positions, the daily numbers of AT and EP, and activity behavior, respectively. Throughout P0-P4, daily milk yield, feed intake, and body weight were recorded. A total of 26 milk samples were collected per cow to determine milk cortisol concentrations. Behavioral observations were conducted for 2 h on 23 days to record agonistic behaviors, vocalizations, and excretions. During P1-P4, each cow received an average of 1.9 ± 3.3 AT and 0.1 ± 0.7 EP per day. The total number of stimuli ranged from 37 to 225 AT and 3 to 11 EP per cow throughout the experimental period, indicating individual variations in learning (P < 0.003). Within the first three days after virtual fence activation, most cows learned the concept of the VF system. The conditioning of the cows was successful at herd level after three repetitions at a new virtual boundary, reflected by a decreasing ratio of mean EP/AT from 20 to 3, 2, and 0 % with each paddock change. Milk yield and cortisol, feed intake, body weight, and cow activity did not differ between VF and EF groups. Overall, there were a mean of 11.2 vocalizations and 5.9 displacements more per cow in the VF groups (P < 0.05), but similar across P0-P4. Our results suggest that learning the virtual fencing system did not lead to lasting adverse effects on cow welfare.