Abstract
Virtual fence is a new technology that could transform the way we manage the distribution of grazing cattle and other ruminant livestock species. Virtual fence delivers sequences of warning cues and aversive electrical stimuli from animal borne devices that modify animal behavior and influence where animals spend their time. Animals learn to respond to and avoid virtual fence boundaries through a combination of classical and operant conditioning. The characteristics of audio or visual cues can influence the animals’ ability to learn the relationship between the warning cues and the aversive stimuli. However, the specific properties of these warning cues that influence their effectiveness have not been well studied. I compared two different delay intervals between audio cues and electrical stimuli, three different audio cue patterns, and two different visual cues in a series of exploratory studies. Animals responded best to a 2-second delay, and there was no substantial benefit to altering the pitch or rapidity of audio cue over a single beep. A comparison of lights and flagging as visual cues suggested that either type of cue was effective. I conducted a comparison of how animals respond to visual, audio, and combined audio-visual warning cues. Visual cues alone or combined audio-visual cues were equally effective as warning cues. However, it appears that including a visual cue accelerates learning and increases compliance to a virtual boundary. Applications of these findings could improve training practices and overall compliance in virtual fence systems.