Project details
Status: Completed
At a glance
- This project investigated the potential of virtual fencing to improve equity in pasture access and optimise intake in large, pasture-based dairy herds.
- Specifically, it explored the effects of strategically staggering access to pasture allocations, aligned with the trickle-back of cows from milking, and increasing the proportion of the 24-hour pasture allowance offered at dusk.
- By focusing on cows later in the milking order, the study aimed to assess whether these management strategies can enhance productivity and welfare outcomes in large grazing dairy herds.
- This trial was conducted in September 2025.
Virtual fencing research at TIA
Learn more about TIA's virtual fencing research program and associated research projects.
Project background
Increasing herd sizes in Australian dairy systems are creating challenges with unequal pasture access, as cows later in the milking order spend more time off-paddock and produce less milk. Current grazing systems cannot easily address this due to labour and infrastructure constraints.
This project aimed to use virtual fencing to improve equity of pasture access and optimise intake, productivity, and welfare in large herds.
Research question
This project investigated the use of virtual fencing to stagger pasture allocation with the goal of improving pasture access equity and milk production in large dairy herds.
Project activities
This trial involved one thousand lactating cows, which were allocated to two treatment herds of 500 and balanced for breed, position in the herd, parity, days in milk, milk production, and baseline behaviour. Both herds received two pasture allocations daily: 60% of their total daily allowance following the afternoon (PM) milking and 40% after the morning (AM) milking.
Treatment 1: Cows in treatment 1 received the entire pasture allocation immediately after each milking as they trickle back from the dairy.
Treatment 2: Cows in treatment 2 received the same total pasture allocation as cows in treatment 1, but access was staggered using virtual fencing into two sequential sections. Initially after each milking, the first half of the allocation was made available to the first cows that returned from the dairy. As the remaining cows trickled back, the virtual boundary shifted to grant all cows access to the total allocation.
The timing of the shift was determined during a pre-trial observation period, based on when the first ~20% of cows stop grazing and begin ruminating. Aligning the shift with this point of satiety was expected to improve the productivity of cows later in the milking order, while maintaining milk production in cows at the front of the order.
Importantly, no cues were delivered to cows already present in the paddock at the time of the fence shift in treatment 1. This approach was designed to allow newly arriving cows to naturally discover access to the fresh pasture allocation without stimulating renewed grazing behaviour in cows that have already accessed the pasture in the first part of the allocation. It was anticipated that this approach would reduce competition and enable later arrivals to have equal access fresh pasture.
After two weeks, the treatments were crossed over with herd 1 swapping to be managed using the staggered virtual boundary approach, while herd 2 were managed as a single cohort with full access to each allocation. The trial concluded after a further two weeks of observation under the alternate treatment.
Data recorded
Milk production: The milk yield of each cow was automatically recorded at each milking (morning and afternoon) for each day of the trial using the farm's individual milk meters.
Milk quality: Milk samples were assessed multiple times throughout the trial for fat and protein concentrations, as well as somatic cell count (SCC).
Pasture composition and utilisation: Pasture samples were taken twice weekly for compositional analysis. Measurements of compressed pasture height will be converted into pasture biomass (kg of DM/ha) using a site-specific linear regression equation. Pasture consumption will then be calculated as total pasture consumed (pre-grazing minus post-grazing pasture biomass) divided by total grazeable pasture (kg of DM/ha).
Grazing and ruminating behaviour: The time cows spend grazing and ruminating will be continuously recorded using the Halter devices. These data will used to calculate time cows spend grazing, resting and ruminating per day.
Virtual fencing and herding: Using Halter technology to record the number of sound and vibration cues, shock reinforcements delivered per cow and day, and time off pasture.
Project outcomes
Further information will be added once the data has been analysed.