Drawing the line: comparing zone-specific spatial behaviour of heifers on pasture with virtual and physical fences

Virtual fencing facilitates dynamic and adaptive grazing management yet research on the influence of proximity to invisible boundaries on movement behaviour of cattle remains limited. This study tested zone-specific behaviour of 31 beef heifers on paddocks with virtual fences (VF) and electric wire fences (EF). All animals wore collars providing approximately 1-min global navigation satellite system fixes and grazed paddocks of 0.5 ha in groups of eight for 3 consecutive days each in a 2021 rotational grazing trial in Germany. For analyses, each paddock was divided into a perimeter zone and a centre zone (> 10 m from the fence). Fixes were used to derive the zone-specific time-in-zone, behavioural time budgets (lying vs active), inverse speed, and spatial evenness (Camargo's index) overall and within behaviours. Generalised linear mixed-effects models were fitted to evaluate the fixed effects of zone, fencing system, and day (and their interaction), with random intercepts for animal and for the group-paddock structure. Across fencing systems, heifers used the perimeter less than the centre (P < 0.001). This contrast persisted on all days under VF but attenuated under EF by the last day on a paddock. Lying concentrated in the centre, with a stronger zone effect under VF (17.5 ± 1.5% in the perimeter vs 45.7 ± 2.0% in the centre, P < 0.0001) than EF (35.1 ± 1.9% in the perimeter vs 44.4 ± 2.0% in the centre, P < 0.0001). More lying bouts occurred in the perimeter, indicating shorter, less consolidated resting near boundaries. Movement was slower near boundaries irrespective of fencing system (P < 0.0001). Overall spatial dispersion depended on fencing system: under VF, evenness for lying and active time was higher in the perimeter (0.507 ± 0.013) compared to the centre (0.346 ± 0.013, P < 0.001) on all days; whereas for EF, the difference (0.391 ± 0.013 in the perimeter and 0.364 ± 0.013 in the centre) was only significant on day 1. These findings indicate that boundary proximity, rather than fencing system, affects cattle movement and shifts behavioural time budgets. Virtual fencing maintains boundary-related behavioural patterns comparable to physical electric fencing. Overall, the results highlight that any fence, whether virtual or physical, represents an imposed boundary shaping animal behaviour.