Stags locking antlers, bull elephant seals body-slamming, lions circling. Creatures engaged in combat is one of the most iconic spectacles of the animal kingdom – and the often-bloody highlight of many a nature documentary. We watch in awe as animals fight for access to mates or territory, to establish position within a hierarchy or, in the case of the hermit crab, simply for a rather nice shell to inhabit. But the career of behavioural biologist Dr Gareth Arnott, Reader in Biological Sciences, has been spent revealing how much more is going on in such combats – and in the minds of animals – than meets the eye.

“When I started out, the largely accepted hypothesis was ‘mutual assessment’ – that a fight is always about you and the opponent, with animals sizing up the opposition,” Arnott says. It may appear reasonable that animals assess whether their opponent is likely to beat them, he suggests, a judgement that will influence how long they stay in the fight, or whether they engage at all. But that’s to look at the matter through an anthropomorphising lens. “There’s good evidence in humans about our ability to measure dominance and aggression,” Arnott explains. “We are very capable at these comparative assessments. But scientifically that has clouded our judgement – we shouldn’t assume it is the case for animals.”

Arnott’s journey to unravelling the intricacies of animal contests began when he teamed up with Queen’s now-Emeritus Professor of Animal Behaviour, Robert Elwood, who was working on a simpler alternative to the widely held idea of mutual assessment – namely ‘self-assessment’, the idea that animals were aware only of their own ability in a fight. 

The two collaborated in 2009 on a paper that proved seminal in the field of animal behaviour. It laid out a framework for identifying and analysing self and mutual assessment in contest participants. “We put forward the idea that in terms of animal cognition, the processing, use and storage of information is key. We set out our hypothesis that self-assessment requires fewer cognitive abilities. It is a simpler psychological process than mutual assessment.”

Arnott says that we can never know exactly what animals are thinking. “But we can know about the information that they are gathering, and their decision-making.” One species that perfectly illustrates the complexities of animal cognition in contest situations is the humble hermit crab. 

“The life of the hermit crab can be summed up as the quest for the perfect shell,” Arnott says. Shells may be discovered empty on the shore, but it’s a lucky crab that finds its new home this way. Most often, they fight another crab and take theirs. “Hermit crabs gather lots of information about the quality of the desired shell, including likely improvements, and change fight tactics accordingly,” says Arnott. “For significantly increased shell gain, they’ll fight more aggressively, or resume a fight more quickly.” But this information-gathering about the desirability of the targeted shell on the part of the attacker does not equate to an assessment of the fighting prowess of its opponent. The attacker appears to fight according to self-assessment only.

Not so the defender. “Defending hermit crabs seem to gather information about the attacker,” Arnott explains. The attacker will deliver bouts of “shell rapping”, causing the defendant to withdraw inside. Victory is achieved when the defender is removed from their shell and thrown aside, while the victor engages in a prolonged session of switching between both shells to decide whether to trade its current abode for the one it has just won. “If their opponent is really strong, the defender will give up quickly, but if it’s a weaker opponent, the defender is likely to persist for much longer. This suggests mutual assessment on the part of the defender,” says Arnott.

So complex is hermit crab behaviour, it is opening up even larger – and more controversial – questions about animal sentience. “The working definition of sentience is the capacity to experience feelings and this is associated with levels of awareness – for example, of our own future behavioural patterns,” says Arnott.“So even with a simple animal like a hermit crab, we’re now questioning whether invertebrates know more than we think.”

It’s a subject to which behavioural scientists are making significant contributions. Elwood has just published an article in the journal Animal Cognition exploring sentience, while one of Arnott’s former PhD students, Dr Andrew Crump, was an author on a recent report outlining evidence of sentience in decapods and cephalopods. Animal sentience is now recognised and enshrined in the Animal  Welfare (Sentience) Act which passed into law in the UK in 2022 and, as well as including all vertebrates, it also includes decapods and cephalopods based on the evidence presented in the report. 

HIGH ON EMOTION
Animal welfare is a key factor in the acknowledgement of animal sentience. Standard scientific terminology speaks of negative or positive “affective states”, and Arnott’s current work explores whether consideration of animal “emotions” can improve 
welfare standards, from licensed, large-scale dog breeding to dairy farming. And one major current investigation, involving commercially reared pigs, marries Arnott’s long-term contest work with the emerging exploration of animal “emotion” and its 
importance for welfare. 

“Pig aggression is a big welfare issue when animals are mixed together at various stages of the production cycle,” he explains. “They fight to establish dominance, which is a major stressor.” Initial work found that pigs that are socialised earlier in life, by mixing litters for play, for example, would be capable of resolving contests faster and with less aggression when they were mature.

The team has gone on to explore how animals’ “emotions” are affected by contest outcome, using an elegant experiment. Pigs are trained to recognise that if a bucket is put in one corner of their pen it will contain a food reward, while if the bucket is in another corner it will be empty. The bucket is then placed in an ambiguous position, between the two locations. The hypothesis is that a pig that has recently lost a contest – and will therefore be in a negative affective state – will interpret the ambiguous location as signalling an empty bucket and not approach to feed, whereas a winner, in a positive affective state, will be more likely to view it as a reward and go to it.

Initial phases of Arnott’s research, published in 2020 in Proceedings of the Royal Society, also float the question of how recent contest outcomes induce long-term moods. “So we’ve been on a journey from basic questions of self or mutual assessment in contests to marrying that with this exploratory research on emotion,” says Arnott.

As academic lead of the Animal Welfare Research Network, a coalition of more than 1,000 scholars, Arnott has recently been appointed to the UK Government’s Animal Welfare Panel, which he hopes will provide an opportunity to influence policy guidelines. Providing robust, cutting-edge scientific advice to Defra is crucial, he says, because researchers’ ability to explore what animals might be thinking is growing all the time, with consequences for how we think about and treat non-human life.

“We now accept aspects of animal sentience that we would previously have assumed them incapable of,” says Arnott. “All this research has shown that animals have a richer behavioural repertoire and thinking capacity than was accepted even just a few years ago, and we’re now at the point where we have the tools to access that. It’s incredibly exciting: we’re learning more and more about the inner lives of animals – and what they’re really thinking.”