Animals and Grief: An Exploration of the Emotion and its Measurement in Non-Human Animals

Abstract

Concerns for the welfare of animals in part stem from the recognition that animals are able to experience emotions, as such, the emotional experience of animals is a fundamental element of animal welfare. The investigation of emotions in animals is complicated by the difficulty of measuring the internal experience of another being, human or non-human animal, and therefore we are unable to know conclusively whether animals experience emotions. However, scientists engaged in this field of study have been undeterred and in endeavouring to measure animal emotions, a number of methods have been developed and applied with which we might be able to examine the emotional experiences of animals. Comparative to scientific investigations of animal emotions, and despite the fact that public concern for animal welfare ultimately drives improvements to animal welfare legislation, public perceptions surrounding the existence of emotions in animals is a relatively understudied area. This two-phase study focuses on one complex emotion: grief. Behaviour patterns of wild animals in response to dying and deceased individuals are increasingly being reported and interpreted as grief, yet little rigorous scientific investigation of this emotion in animals has been conducted making the exploration of grief in animals overdue. In the first phase public belief in animals’ experience of grief is investigated through quantitative surveys. This is followed in the second phase by experimental studies of the separation of conspecifics. In the first phase of this study, 1000 members of the public were surveyed to investigate the demographic variables that influence the attribution of emotions to animals and beliefs about whether animals can grieve. Ninety % of all respondents believed that animals grieve, with 66% believing an animal’s experience of grief is the same as the human experience. Respondents who currently owned a companion animal were more likely than those that did not to believe that some animals can experience grief (P = 0.002), and those that owned a companion animal during childhood were more likely to believe that animals experience emotions in general (P = 0.008) and that dolphins, elephants, pigs, cows and ants (P < 0.05) grieve. Females were more likely than males to believe that animals experience emotions (P < 0.05). Companion animal ownership and gender therefore play a significant role in the public perception of the emotional experiences of animals and belief in the animals’ ability to grieve. Investigation into the behavioural responses of dogs and cats to the loss of an animal companion, through owner-reported behavioural changes, was also carried out in the first phase of this study. Surveys (n = 279) were distributed in Australia and New Zealand via RSPCA publications and within veterinary clinics. There was consensus among owners that the behaviour of companion animals changes in response to the loss of an animal companion. The most common behavioural changes reported were increased affectionate (75% of dogs and 53% of cats) and territorial behaviours (59% of dogs and 63% of cats). Dogs were reported to reduce food consumption (P < 0.05), increase sleep (P = 0.008), decrease vocalisations (P = 0.006), and cats were reported to increase vocalisations (P = 0.05) and increase aggression (P = 0.05), following the death of a companion. These behavioural changes appear similar to those displayed by companion animals with human-orientated separation anxiety and those displayed in other species as a result of the severance of mother-offspring and conspecific bonds. In the second phase of this study the separation of bonded individuals, hypothesised to result in emotional distress analogous to grief, was investigated in two species, the domestic dog (Canis familiaris) and cattle (Bos taurus). For the purposes of these studies grief was considered to be a biological response to separation, described by a bi-phasic model of ‘active/passive’ behavioural responses, characterised by increased searching activity and vocalisations in the first phase and decreased activity and withdrawal from the environment in the second (Bowlby, 1961). In 12 adult pair-housed dogs, maintained in a private animal shelter, behaviour, faecal S-IgA and cortisol levels were examined before and after separation from a conspecific. There were increases in running (P = 0.02), grooming (P = 0.02), circling (P = 0.006), figure of 8 movement (P = 0.01), posture changes (P = 0.003) and stretching (P = 0.005) and a decrease in play behaviour (P = 0.01) following separation. S-IgA values increased (P = 0.02) after separation (mean = 453.6 ± 182.5 ng/mL) compared to before (mean = 357 ± 108.2 ng/mL), but cortisol levels were not affected by separation. Cognitive bias (the idea that when an animal evaluates a situation with ambiguous stimuli, its emotional valence affects its interpretation of the situation and possible outcomes (Broom, 2010; Mendl et al., 2009) and this is measured through the animals latency to approach the stimuli) was tested and indicated no effect of separation on emotional valence. Therefore the separation of dogs from a conspecific results in increased activity and immunoglobulin concentrations suggesting a negative impact upon welfare that is likely associated with a negative emotional experience, however, further research is required to understand the detailed nature of the emotional experience. In cattle, the response to separation from bonded individuals was investigated in a series of four experiments (E1-E4) with four distinct groups of dairy cattle (E1 = 20 outdoor-housed lactating dairy cows, E2 = 20 lactating loose indoor-housed lactating cows, E3 = 9 dairy calves and E4 = 4 dairy bulls). Each experiment was centred on the removal of a subset of individuals by farm management. Behavioural recordings using direct observation were carried out in all 4 experiments before and after the removal of preferred conspecifics and compared to a control group (preferred conspecifics not removed) in E1 and E2. Preference for conspecifics was established using near-neighbour observations in E1, E2 and E3. Serum samples were analysed for IgA and cortisol concentrations before and after removal occurred in E3 and E4. In E1 an increase in time spent eating (P = 0.003) was observed in remaining cows after conspecific removal. In E2 an increase in time spent eating after conspecific removal was observed in the control cows (P = 0.024), but not the treatment cows (P = 0.99). It is worth considering that social buffering could have played a role in reducing the effect of separation stress, which might explain why the removal of preselected animals did not appear to have a negative impact on remaining individuals in E1 and E2. In E3, a reduction in ruminating behaviour (P = 0.007) was observed and in E4 a decrease in eating (P = 0.037) as well as and an increase in walking (P < 0.0001) was observed. IgA concentrations increased in both experiments following separation (E3: P = 0.011; E4: P = 0.034), however, cortisol concentrations did not differ after removal in either E3 or E4 (P ≥ 0.32). These behavioural observations, taken in combination with the increase in IgA concentrations, suggest that the individuals in E3 and E4 experienced stress and reduced welfare as a result of separation from conspecifics. The small sample size in E4 must be acknowledged, yet despite this, and in line with the findings for the experiment conducted with pair-housed shelter dogs, the negative impact upon welfare observed in E3 and E4 is likely to be associated with a negative emotional experience, but further research is required to categorically conclude this. This research confirms that the public are overwhelmingly willing to attribute complex emotions such as grief to a range of animals, and companion animal owners appear to recognise consistent behavioural changes in their animals following the loss of an animal companion. Experimentation suggested that both dogs and some cattle are adversely affected by separation from conspecifics, however the emotional experiences associated with this separation remain elusive.