The Indian rhinoceros or greater one-horned rhino (Rhinoceros unicornis) is listed as vulnerable by the IUCN Red List of Threatened Species and, therefore, captive individuals have been subject to the European Endangered Species Programme since 1990. Enhancement of welfare is key in ensuring the breeding success of this species in captivity. Salivary cortisol has been recently used to assess welfare of captive and free-ranging animals. Nevertheless, rhythms of cortisol secretion may fluctuate throughout the year and therefore, knowledge of the circannual pattern of cortisol secretion is essential to evaluate the physiological significance of seasonal variations of cortisol levels as an indicator of stress in animals. Here, we analyse monthly differences in cortisol secretion in two Indian rhinoceros. Saliva samples of two rhinoceros were collected and analysed by radioimmunoassay for the period of one year to determine cortisol concentrations. We found a seasonal pattern of salivary cortisol secretion. The highest cortisol concentrations were found in August and decreased until reaching a nadir in January. Cortisol concentrations in these two animals showed a correlation with temperature and visitor numbers but it is not possible to draw conclusions from this study as to whether the variation in cortisol was due to these or other factors.

Mortality rates are often used in population-level animal welfare assessments because they are assumed to reflect rates of disease or injury and other problems likely to cause poor welfare. High mortality is thus assumed to correlate with factors likely to cause negative affective states. Here, we argue that negative affective states are also related to mortality rates more directly, via causal rather than merely correlational routes. In humans, negative affective states predict elevated morbidity and mortality rates as well as decreased longevity, while self-reported happiness does the opposite. This review investigates whether mortality rates and longevity can thus be used to make inferences regarding past affective states in animals. The proposed mechanism is that chronic stressors cause negative affective states and thence harmful physiological consequences through continual activation of the hypothalamic-pituitary-adrenal and sympathetic-adreno-medullary axes, which in turn can lead to increased mortality. The convergent validity of mortality as an indicator of past negative affect is demonstrated via examples of how stressors such as bereavement in humans and social isolation in social non-human species cause negative affective states, and then increase the morbidity of potentially lethal conditions such as cardiovascular disease, cancer and HIV/AIDS, ultimately leading to increased mortality and reduced longevity in both humans and animals. The potential drawbacks of using mortality rates to infer animal welfare are also discussed, including its low sensitivity and the multiplicity of factors unrelated to affective states that can influence mortality rates. However, providing these issues are accounted for, it is suggested that this indicator has value in welfare assessment, and is especially well suited for cases where animals are allowed to live out their natural lifespans, such as tends to be the case with zoo and companion animals.

The welfare of zoo animals depends on a combination of physical, social, dietary and other ecological characteristics of the captive setting. We analysed the influence of the transfer of an adult couple of hamadryas baboons (Papio hamadryas) from a small and non-enriched cage that was closed to the public to a larger and enriched exhibit at the visitation area of the Sapucaia do Sul Zoological Park, RS, Brazil. A total of 350 h divided into four stages were spent observing the baboons: (i) in the non-enriched enclosure before the transfer; (ii) soon after the transfer to the enriched enclosure; (iii) six months; and (iv) 12 months after the transfer. The occurrence of stereotypic, social aggressive and social affiliative behaviours was recorded by ‘all occurrences’ sampling. The female showed a decrease in frequency of a stereotypic behaviour (spinning) and an increase in grooming in the enriched enclosure. The male showed a decrease in the frequency of certain stress-related or stereotypic behaviours after transfer, but other stress-related behaviours either increased or remained constant. The male behaviour of throwing faeces was affected by the presence of visitors. We suggest that the well-being of the female was more positively influenced by the new enclosure than that of the male, although gender differences may reflect the normal behavioural repertoire of this sexually dimorphic species.