The use of livestock guardian dogs (LGDs) to resolve human-wildlife conflict is increasing in many parts of the world. Most often, LGDs are used to protect livestock from predators, and support conservation of wild predators by reducing retaliatory killing by aggrieved farmers. They also reduce overlap between livestock and wild herbivores, and can directly protect threatened wildlife from invasive predators (van Bommel, 2010).

One argument for wider use of LGDs is that they cause less animal suffering than alternative approaches to management of troublesome wildlife, which often involve lethal control by trapping, shooting or poisoning. Allen et al. (2019) challenge this view, arguing that "guardian dogs cause considerable lethal and non-lethal animal welfare impacts" to target animals. They suggest these harms are of similar magnitude or worse than those caused by conventional forms of lethal control.

We argue that Allen's et al. analysis is unsound because it is not informed by evidence on the way that LGDs interact with other species. They assume that LGDs, behaving as predators, influence other animals by pursuing, attacking and often killing them, causing extreme pain and distress. They also suggest that LGDs may sometimes lose fights with predators, and thus experience the same suffering that they otherwise inflict.

But, the experience of most managers of LGDs is that their dogs rarely engage in direct aggressive interactions with other species. Instead, the effectiveness of LGDs may be mainly due to avoidance by target species.