Promoting contest skill to reduce the welfare costs of animal agonistic interactions

Summary
Aggression is a major and routine challenge to welfare in many species of managed animals despite decades of research aimed at minimising it. Aggression also affects health, survival and fitness in wild populations. Aggressive contests accrue physical (e.g. injuries), emotional and energetic costs. Individuals vary greatly in the costs paid during aggressive contests, irrespective of whether they win or lose. Some individuals can resolve contests quickly with minimal costs. The mechanism by which they achieve this is unknown and, until we understand it, reducing the welfare penalty of aggression will be difficult. Our aims are to determine a) whether skill facilitates contest resolution and thus reduces contest costs, and (b), if so, whether husbandry practices can be manipulated to promote skilful fighting, reduce contest costs and enhance welfare and production in a commercial setting. The project uses pigs as they offer an ideal model system and because most of the one billion pigs slaughtered annually experience significant aggression.

One large experiment will address all objectives below:
1. Characterise and quantify skill in contest aggression and defence
2. Determine the effect of the social developmental environment and contest experience on skill
3. Assess the contribution of socio-cognitive ability and assessment strategy to skill
4. Quantify the relationship between skill and conventional measures of fighting ability and aggressiveness
5. Determine whether contest skill reduces the welfare costs of aggression in a commercially realistic scenario

Contests demand the use of complex behaviours. Skill is influential in human sporting contests, yet remarkably its role in animal aggressive contests is almost entirely unstudied. A recent framework has proposed that contest skill comprises efficient, accurate, precise and appropriate behavioural execution. We will test this framework and quantify skill by its effects on contest success and costs (Obj. 1). Variation in commercially relevant early-life opportunities for play fighting experience and minimally-damaging contest experience will test their effects on later contest skill (Obj. 2). Behaviour must be underpinned by rapid and well-informed decisions, which requires that animals assimilate and process complex information and turn it into knowledge (termed cognitive ability). Whether cognitive ability improves the behavioural execution of skill is unknown. A specific cognitive challenge during contests is to gather information and use this to decide when to give up. Several classes of information-gathering model have been developed. In the simplest, termed self assessment, animals make fight decisions based purely on their own fighting ability and stamina, without reference to the ability of an opponent. After a threshold amount of energy has been spent on fighting, the individual will give up. In a second class, termed mutual assessment, animals self assess but also assess the fighting ability of an opponent. Although more complex, it allows an animal to quickly withdraw from a fight it is likely to lose and substantially reduces the costs paid. Hence, Obj. 3 will test whether cognitive ability and individual use of assessment strategy promote skill as reflected in appropriate and efficient behaviour. In doing so we provide the first quantification of assessment strategies at the individual animal level. Obj. 4 will also quantify whether superiority in other determinants of fighting ability (e.g. weight) and an aggressive personality enhance or suppress skill. Crucially the project also has applied relevance. Obj. 2 will inform how management can promote skill in dyadic contests. Importantly, in Obj. 5 pigs will be regrouped in a commercially-realistic way (groups of 12 animals) to test whether skill in paired contests improves welfare in a real-world scenario.

Technical summary
Aggression reduces managed animal welfare and wild animal fitness. Individuals differ greatly in their ability to rapidly resolve aggressive contests with minimal costs, the causes of which are poorly understood, limiting our ability to reduce negative welfare impacts. Contest skill is hypothesised to affect contest costs and success but this is almost completely untested. A recent framework predicts that skill requires efficient, accurate, precise and appropriate behaviour. We test this using pigs to characterise skill by the effect of its putative components on contest costs and outcome (Obj.1). Obj.2 examines the effect of early social experience on skill development. We impose commercially feasible interventions (early contest experience; early socialisation to enhance play) and predict that they promote skill in later contests. Obj.3 studies the role of socio-cognitive ability and assessment strategy in skill. Assessment strategies consist of two broad types: In self assessment models each contestant appraises its own fighting ability (resource holding potential; RHP) but not that of the opponent. In mutual assessment models individuals compare opponent RHP against their own, markedly reducing contest costs. Obj.3 predicts that cognitive ability and greater use of mutual assessment enhance the appropriateness aspect of skill. This employs the first test of a recent theory proposing the existence of individual variation in assessment strategies. Obj.4 explores how an aggressive personality and conventional measures of RHP (e.g. weight) impair or enhance skill. Importantly, Obj.5 studies how skill in dyadic contests translates to commercially relevant group mixing in which many unfamiliar individuals meet simultaneously. We predict that dyadic contest skill minimises welfare costs in this scenario and, by promoting contest skill by the early life manipulations of Obj.2, the welfare costs of aggression in commercial production can be reduced.