Evaluating the Reliability of Non-Specialist Observers in the Behavioural Assessment of Semi-Captive Asian Elephant Welfare

The Royal Zoological Society of Scotland's Approach to Assessing and Promoting Animal Welfare in Collaboration with Universities

Good zoos have four aims-to conserve species, educate the public, engage in research, and provide recreation-all of which can only be achieved when underpinned by high animal welfare standards. In this paper, we share the approach that The Royal Zoological Society of Scotland's (RZSS) Edinburgh Zoo and Highland Wildlife Park take to animal welfare. We highlight the role that animal welfare research, in collaboration with universities, has had in enabling the zoo to take an evidence-based approach to welfare and to put findings into practice. We share the collaborative process through which we developed and piloted the current animal welfare assessment tools, how they were validated, and how they were tested for reliability as part of a long-term collaboration between the Royal Zoological Society of Scotland and the University of Stirling: (1) the RZSS Welfare Assessment Tool, a 50-question animal welfare assessment adapted from the British and Irish Association of Zoos and Aquariums (BIAZA) Toolkit; and (2) the Stirling Toolkit, a package of evidence-based resources for behavioural-data collection. Our aim is to facilitate standardised, evidence-based approaches to assessing animal welfare which, when finalised, can be used collaboratively across zoos.

Qualitative Behavioural Assessment of bonobo emotional expressivity across observer groups and zoo housing environments

Human evaluation of animal emotional expressivity can inform animal welfare. Qualitative Behavioural Assessment (QBA) has been applied to domesticated and some non-domesticated animals, but its use in primates is limited despite their emotional expressivity. We aimed to develop and apply a QBA for bonobos (Pan paniscus) through two consecutive studies. We applied Free Choice Profiling (FCP) and the Fixed List methodology, respectively, in Study 1 and 2, and invited students and bonobo experts to rate video clips of zoo-living bonobos of different sexes and age classes, and before and after moving to a new enclosure. In Study 1, students described dimension 1 as ranging from 'quiet/calm' to 'angry/active' and dimension 2 from 'sad/anxious' to 'happy/loving'. Experts described dimension 1 ranging from 'quiet/relaxed' to 'nervous/alert' and dimension 2 from 'nervous/bored' to 'playful/happy'. Using a fixed list of descriptors, informed by findings from Study 1, students in Study 2 described dimension 1 as ranging from 'quiet/calm' to 'agitated/frustrated', and dimension 2 from 'sad/stressed' to 'happy/positively engaged'. Experts described dimension 1 as ranging from 'quiet/calm' to 'active/excited', and dimension 2 from 'sad/bored' to 'happy/positively engaged'. Students scored adults as more 'calm/quiet' and experts scored subadults as more 'happy/positively engaged'. Additionally, experts in Study 2 rated bonobos as more 'active/excited' in their new enclosure. Reliability was moderate to good for the dimensions. Additionally, animal-directed empathy of observers influenced QBA scores. This is the first time, FCP has been successfully used as a method to study primate emotional expressivity. Our findings show the promise of employing QBA in primate studies and in industry, with validation of additional metrics to enable its use for welfare-monitoring purposes.

A fast technique to induce and measure anticipatory behavior in bottlenose dolphins (Tursiops truncatus)

Anticipatory behavior describes the actions taken to prepare for an upcoming event. It is considered a promising tool for welfare assessment, but it is not fully applied in zoos and other facilities. This is probably due to the need to train animals, which can take a long time and make its use impractical. In this work, we tested, on a group of six dolphins (Tursiops truncatus), the effectiveness of a new technique to induce and measure anticipatory behavior without training animals. Our results show how this new methodology, which we called The Fast Technique (TFT), has been able to induce and measure anticipatory behavior toward the event studied, i.e., the play sessions with objects (PSO). Furthermore, it has produced the same result obtained with The Classical Technique (TCT), which requires a training phase. Therefore, our finding demonstrates the efficacy of a new technique that could facilitate the use of anticipatory behavior for both research and animal welfare assessment protocols.