Sensitivity to reward change: a novel cognitive approach to understanding and measuring affective state in animals

The question as to what feelings animals might or might not experience underlies not just the assessment of animal welfare, but public concern for how captive animals are treated. As the number of animals used in agriculture, research and kept as pets continues to increase, for example over 700million chickens are farmed and killed annually in the UK for meat production alone, the belief that many animal species have the capacity to suffer drives our desire to improve the housing conditions and husbandry procedures of those animals. Yet, because such experiences are private, they cannot be accessed directly. Instead, welfare scientists use indirect measures thought to reflect animals' emotional states. These indicators are often behavioural (e.g. approaching food) or physiological (e.g. a change in heart rate). However, despite the benefits of these measures, they have critical limitations. For example, an animal's heart rate will increase in the presence of something fearful (e.g. a predator), but will also increase during the experience of something pleasurable (e.g. food), such that reliable interpretation of this measure becomes impossible. For this reason, research into developing new approaches to measuring animal emotion and mood is urgently required to allow a more accurate and targeted assessment of animal welfare. Research that underpins and advances in animal welfare science in this way will have a major impact on a captive animal population that consists of large numbers of individuals. For example, there are an estimated 10 million owned dogs in the UK (2006).

The aim of this proposed research programme is therefore to develop a completely new approach to understanding and measuring emotions and moods in animals that will significantly advance our assessment of animal welfare, and this approach is based on reward evaluation.

How we feel about a reward that we have been given depends not just upon its absolute value, but its value compared to what was expected. So, although you might normally be happy to receive a pay rise, you will be disappointed if you were expecting to be offered a promotion as well. This sensitivity to the contrast between what you expected and what you actually received is common to many animal species and is thought to help individuals to make appropriate decisions. It has recently been demonstrated for the first time that the mood an individual is currently experiencing can influence its sensitivity to a change in reward, with those animals in a positive mood recovering sooner from an unexpected loss of reward. This therefore provides an exciting opportunity to use sensitivity to reward change as a novel measure of animal emotion and mood. Initial research indicates that sensitivity to reward change is a quick, effective and non-invasive way of measuring moods in animals. However, this approach needs to be studied in more detail in order to determine the full extent of its potential.

In this project we will be studying emotions and moods in domestic dogs, both owned (pet) dogs and those housed at rescue centres. In contrast to laboratory and farm animals, the study of emotions, moods and associated behaviour in dogs, because they share a close relationship and environment with people, can reveal valuable insights into the interaction between cognition (e.g. decision-making) and emotion that applies not just to dogs and other animal species, but also to humans themselves. Whilst the outcomes of the study will clearly have a direct impact upon the welfare assessment of dogs, resulting in ways to improve the welfare of dogs housed in rescue centres and rehoming success, it will have a far wider impact, delivering techniques that can be applied across captive animal species.

The aim of this proposed research programme is to develop a completely novel, theoretically driven, approach to understanding and measuring affective state (i.e. emotions and moods) in animals that will significantly advance our ability to assess animal welfare. There is an urgent requirement for the development and refinement of additional indicators of affective state that may be less susceptible to the limitations identified in many existing measures. Physiological measures (e.g. heart rate) reveal affective arousal, but may not reflect valence (i.e. positive or negative), and behavioural measures can be hard to interpret. Affective states in both humans and animals incorporate cognitive, behavioural and physiological components in addition to the associated conscious subjective experience, and it is the cognitive component that is the focus of this research programme - specifically, how animals respond to unexpected changes in the quality/quantity of rewards. It has recently been demonstrated that the way in which animals respond to unexpected changes in reward (i.e. their reward sensitivity) can be influenced by their affective state, with animals in a positive affective state being faster to recover from surprising decreases in reward quantity. The objectives of this programme are therefore to further validate the use of sensitivity to reward change as a novel measure of affective state, by investigating the impact of differently induced affective states on this measure and by revealing the links between sensitivity to reward change and other more established cognitive, behavioural and physiological indicators. We will also extend current understanding of this paradigm to encompass different reward categories, by investigating whether animals show sensitivity to an unexpected change of social reward, opening up an entirely new way of assessing what an animal wants, likes and needs.

The impact of this research concerns the development of a more accurate and targeted assessment of animal welfare, an output likely to have wide ranging benefits. There will be immediate and direct benefits with regard to the welfare of companion animals, but also longer-term benefits as a result of the development of a general theoretical framework and approach that can be applied to all captive animals. This will have significant benefits: contributing to the debate about the use of animals in agriculture, research and as companion animals, and the way that they are treated; and informing policy makers, charitable organisations, owners and carers about animal welfare requirements. These benefits are likely to impact both nationally and globally as animal welfare is a topic of worldwide relevance, not just in the developed world, but in the developing world where the relationship between animal and human welfare is often inter-dependent.
The development of a more accurate and targeted assessment of welfare in dogs in particular, will directly and rapidly benefit those associated with companion animals, including: the pet industry sector - who focus on improving the health and welfare of companion animals and make a significant contribution to the global economy; pet owners - educating the general public as to how to best care for their pets will enhance quality of life and have a significant impact (e.g. there were an estimated 10million owned dogs in the UK (2006), an increase from 6.7million in 4.8million households in 2002); charitable organisations - they will benefit from the refinement of housing of kennelled dogs at rescue shelters (the Humane Society of the United States estimates that c.15 million dogs are relinquished to animal welfare charities each year). A topic of considerable public interest is the effect of artificial selection and domestication on dog welfare, with major health and welfare problems associated with pedigree dogs. Refinement of the welfare assessment of dogs will therefore have an input into the process of addressing this problem by investigating the link between genotypic and phenotypic abnormalities in dogs and their relation to affective state.
The development of a more accurate and targeted approach to assessing animal welfare and greater understanding of animal effect in general, will have a longer term benefit for all those associated with the management of captive animals, including: the agricultural sector - improving animal welfare assessment will make a significant contribution to the interface between food production and public concern for animal welfare with a potentially massive impact (e.g. over 700million chickens are farmed and killed annually in the UK for meat production alone, and livestock product consumption results in an estimated UK farm-gate value of livestock products of £7.6billion), as well as contributing scientific evidence to the debate about intensive farming and the impact on animal welfare (a topic that is increasingly back on the public agenda with, for example, recent proposals for the creation of 'super-dairies'). The outputs of the proposed research programme will also directly address the stated interest in promoting positive affective states to improve welfare (e.g. FAWC, Defra). Research animals - improved welfare assessment will allow a valuable evidence-based contribution to the debate on the use of animals in research (numbers of which have increased in recent years) and to the cost-benefit ethical consideration of invasive scientific procedures. Government and NGO bodies concerned with the policy advancement and the formation of recommendations/guidelines for the regulation of laboratory and farm animal welfare (e.g. EU, NC3Rs, HO, Defra, UFAW, RSPCA) will benefit from improved animal welfare assessment and a better understanding of animal affective state, using this knowledge to inform refinements of animal welfare, and to monitor effectiveness.