Emotional States of African Elephants (Loxodonta africana) Kept for Animal–Visitor Interactions, as Perceived by People Differing in Age and Knowledge of the Species

Abstract

Simple Summary We investigated the use of the qualitative behavior assessment (QBA) to evaluate the emotional state of African elephants managed in captive and semi-captive environments by three groups of people with differing ages and levels of knowledge of the species. We also examined whether their assessments correlated with behaviors exhibited by the animals. Fifteen video-clips of a total of 18 African elephants recorded in three different situations (release from the night boma; interactions with visitors; return to the night boma) were used. The result of the performed analysis supported that the consensus found was not due to chance. This notwithstanding, all the adjectives used by the three observer groups were not strong descriptors of the consensus variables resulting from statistical analysis. All three groups showed a degree of separation between captive and semi-captive management, with semi-captive animals rated as being in a more positive emotional state. For all three groups of observers, stereotypic “trunk swirling” behavior correlated with negative emotional descriptors. Although definitive evaluation of animal welfare requires the services of experts, more studies are needed to investigate the perception of elephants’ emotional states amongst visitors of different ages and background. Abstract This study aimed to investigate how three groups of people of differing ages, and with differing knowledge of the species, perceived the emotional state of African elephants (Loxodonta africana) managed in captive and semi-captive environments. Fifteen video-clips of 18 elephants, observed during three different daily routines (release from and return to the night boma; interactions with visitors), were used for a free choice profiling assessment (FCP) and then analyzed with quantitative methods. A general Procrustes analysis identified two main descriptive dimensions of elephant behavioral expression explaining 27% and 19% of the variability in the children group, 19% and 23.7% in adults, and 21.8% and 17% in the expert group. All the descriptors the observers came up with showed a low level of correlation on the identified dimensions. All three observers’ groups showed a degree of separation between captive and semi-captive management. Spearman analyses showed that stereotypic “trunk swirling” behavior correlated negatively with first dimension (free/friendly versus sad/bored) in the children’s group; second dimension (agitated/confident versus angry/bored) amongst the adults; and first dimension (active/excited versus agitated/bored) amongst the experts. More studies are needed to investigate other potential differences in assessing elephants’ emotional states by visitors of different ages and backgrounds.