Behavioral observations and analysis of two subspecies of equids, Tarpan Konik (E. ferus caballus) and Przewalski (E. ferus przewalskii), in an extensive controlled environment, with a focus on social interactions

Several comparative behavioural studies were made in the past on Przewalski horses and feral domestic horses. However, behavioural data were mostly collected in not similar environmental situations among the subspecies. To investigate if behavioural differences exist also in presence of reduced environmental confounders, a consistent data collection was performed in a semi-captive zoological garden in Belgium, by observing, recording and comparing data of the two subspecies living in an extensive controlled environment with similar treatments and similar habitat, without individual recognition. Data were collected between June and August 2024, with 2 weeks of ad libitum observations and 6 following weeks of data recording. The subspecies were observed 2 times a day, for 30 minutes per observation each, during the opening hours of the park, for a total of 30 hours of observations each. Instantaneous scan method with 1-minute sample interval was used for the first 15 minutes, recording general “Maintenance” behaviours plus “Social affiliative”, “Social agonistic” and “Social other” categories. Continuous behavioural method was used for the last 10 minutes, recording specific affiliative, agonistic and other behaviours. The average of the scores (i.e. the number of individuals performing a behaviour) of each behaviour was calculated for instantaneous scan, and the total episodes of each behaviour and the values of reciprocated/not reciprocated episodes for the continuous behavioural. We investigated which factors could affect the likelihood of seeing at least one animal expressing a certain behaviour during the observation, such as temperature, weather, humidity, wind speed, part of the day (morning/afternoon) and membership of individuals to two different subspecies. We found that most of maintenance behaviours were influenced by the subspecies' factor, probably due to subtle habitat differences, and that Przewalskis performed slightly more social agonistic behaviours, while Tarpan Koniks were more involved in social affiliative ones. The differences encountered seem to be more associated with the harem composition than with real differences between wild and feral horses, but there is no certainty of that.