• Rhiannon Handcock

Virtual Fencing for dairy cows

Virtual fencing is a relatively new advancement that could be used in rotational grazing systems. This study used the eShepherd virtual fencing system (their website can be found here) which is a device attached to a neckband which uses "audio cues" to create a "virtual fence" in place of an electric fence to keep animals within a designated area.


The study was done to investigate how well the technology works for a herd of 30 cows, including cow behaviour and welfare compared with using electric fences.

Key Points

  1. Virtual fencing uses GPS technology to communicate virtual fencelines to cows through audio and electrical pulses.

  2. Pasture utilisation was better during the electric fence period compared with the virtual fence period.

  3. The authors concluded that based on their short study that there was no evidence of negative behavioural or welfare effects of using the virtual fencing system for dairy cows.


What is virtual fencing?

Virtual fencing is advertised as a time, labour and materials saver compared with electric fencing.


Global positioning system (GPS) technology is used to assign a virtual fence location and is communicated to the cows using the fitted neckband devices. The neckbands let out an "audio cue" when a cow nears the virtual fence, if ignored then a short, sharp electrical pulse sequence occurs along with the audio cue to train cows to stop moving or turn away when the audio cue starts. The electrical pulse is lower in energy than an electric fence however the exact values are confidential to the creators of eShepherd.


Activation of the virtual fence was controlled by a cloud-based web interface that signalled the fence location to the cow's neckband devices through a wireless radio frequency. If a cow became unresponsive to the audio and electrical stimuli (again these values were confidential) the neckband stopped sending all stimuli.


How was this study done?

This study was completed in mid-spring at a Tasmanian Research Farm (Australia). Cows were moved to a new break after the afternoon milking. For the first 10 days, breaks were set up using an electric fence followed by 3 days of virtual fence "learning" then another 10 days using the virtual fence.


On average, cows needed 3 of the audio cues before creating an association between an audio cue and an electrical pulse. At the beginning of the learning period 65% of cows received the electrical stimulus after an audio cue. By the end of the learning period this had dropped to 32% of cows received the electrical stimulus after an audio cue.


Cows were never observed on the wrong side of the electric tape fence over the 10 days they were observed. As expected, cows received fewer electrical stimuli when the electric fence was used compared with the virtual fence, likely because cows were on average 5 years of age and had multiple years of experience with electric fencing.


When the virtual fencing was used, the whole paddock was split into an inclusion and exclusion zone. Cows entered the exclusion zone on average 3.3 times per cow per 24-hour period, however, the average duration of time spent in the exclusion zone was only 5 min per day.


Cows received far more audio cues than electrical cues, which indicates the cows were generally responding to the audio cues to avoid the electrical cue. This is similar to cows going near the electric tape but avoiding touching it to avoid a shock.


Pasture Utilisation

Pasture utilisation was better during the electric fence period at 95.8% compared with the virtual fence period with 71.2% utilisation.


Average pre-grazing pasture mass was 2,864 kg of DM/ha during both periods, therefore during the virtual fence treatment cows ate less per day than they did during the electric fence period.


During both periods cows had access to concentrates fed in the milking shed.


It turned out that the total energy intakes between the two periods were similar due to differences in energy content of the pasture (as they were grazing different paddocks a few weeks after each other) and cows ended up consuming more concentrates during the virtual fencing period than they did in the electric fence period.


From this study, it's hard to draw conclusions as to what this could mean for pasture utilisation for systems using virtual fencing. It is not clear if the virtual fencing caused a drop in pasture intake that was substituted with increased concentrate intake, or if the higher energy of the pasture during the virtual fencing period caused a drop in pasture intakes.


We definitely need to see more research in this area!


Cow Performance

There were no differences in milk production between the two periods, cows produced 25.9 L of milk per day during the electric fence period and 26.2 L during the virtual fence period.


Likewise there were no differences in cow live weight between the two periods.



Cow behaviour

During the early virtual fence period the cows behaviour was similar to that of the electric fence period. Between days 4 to 6 of the virtual fencing period, cows spent less time grazing and more time ruminating than they did during the electric fence period.


The most frequently observed response to the audio cue was for cows to turn around and continue grazing followed second by continuing to graze (not the desired behaviour!). The third most common response was to turn and walk away.


Following an electrical pulse in the virtual fencing period, 73% of the responses were to turn and walk or trot away. The next most common response (18%) was to turn and continue grazing in the other direction.


The authors concluded that based on their short study that there was no evidence of negative behavioural or welfare effects of using the virtual fencing system for dairy cows. They did suggest that longer studies be conducted before making firm suggestions on the effects beyond the short 10-day period.


It might be important to note that after 23 days of wearing the eShepherd devices, the decision was made to stop the experiment due some cows developing abrasions on their lower jaws. Other studies of beef and nonlactating dairy cows have not reported these abrasions on the skin. This might be something to be aware of if you are considering using a neckband type system for your cows.



What did you think about this week's post? Would you consider using a virtual fencing technology on your farm? Let us know your thoughts on our facebook page (www.facebook.com/agrisciencer) or in the comments below.

Full Papers:

Langworthy, A. D., Verdon, M., Freeman, M. J., Corkrey, R., Hills, J. L., & Rawnsley, R. P. (2021). Virtual fencing technology to intensively graze lactating dairy cattle I: Technology efficacy and pasture utilization. Journal of Dairy Science.


Verdon, M., Langworthy, A., & Rawnsley, R. (2021). Virtual fencing technology to intensively graze lactating dairy cattle II: Effects on cow welfare and behavior. Journal of Dairy Science.