This study aimed to investigate if there were any attachment-like behaviour expressed by the horse, such as seeking proximity to its owner, considering its’ owner as a safe haven, and secure base, when performing a simplified Stranger Situation Procedure. I further wanted to test if the owner’s attachment style and training style may have an impact on the relationship and behaviour of the horse.

The results of the present study are not consentaneous regarding attachment-like behaviour, for example the horses explored more when being left alone rather than in the presence of either the owner or the stranger which is the opposite to what characterises a secure base. However, other behavioural differences found when comparing separation episode to reunion episode indicates that human presence, regardless of familiarity, has an impact on the horse behaviour. In the present study the horses showed an elevated heart rate when they were left alone followed by a decreased heart rate when the human returned, which suggests that the horses regarded both the owner and stranger as a safe haven. In line with this, the horses spent more time in door proximity when being alone followed by a significant decrease of time spent in door proximity when human returned. One could argue that the elevated heart rate could be due to arousal of being alone, especially since exploration behaviour was more prevalent during separation. However, horses have a strong motivation for social contact with conspecifics (Søndergaard et al., 2011) so the probability that they would have been aroused or excited during the separation episode is very low. Another cause to the elevated heart rate could simply be locomotion (i.e. that the horse walks around), but there were no significant differences in time spent walking between the separation episode and the reunion episode to indicate that locomotion would be the reason for the increased heart rate during separation.  It could further be discussed that the decrease in heart rate at the reunion episode was not a result of the return of the owner/stranger but simply due to habituation of being alone. Although, approximately half of the 26 horses approached the returning human (owner or stranger) and sought human proximity after separation which is in line with attachment-like behaviour. The significant decrease in time the horses spent in door proximity between separation and reunion also indicates that when the horses were left alone, they wanted to leave the arena but not to the same extent when a human was present. Górecka et al., (2007) showed that horses approached a novel object with shorter latency when led by a handler compared to not being led by a handler, but the authors did not specify if the handler was familiar to the horse or not. The results from the present study together with previous studies indicates that human presence per se do have an impact on horse behaviour in such way that the horses displayed lesser interest in leaving the arena with a human present and sought proximity to human at reunion. The present study further indicate that human presence has an impact on horse physiology as in decreasing horse heart rate, which is partly the definition of a safe haven (Ainsworth et al., 2015, Payne et al., 2015).

Comparing the training styles used by the owners revealed some significant differences in the horse behaviour. Several studies show that positive reinforcement training results in a positive attitude in animals towards humans, regardless of it being a familiar human or a stranger (Innes & McBride, 2008, Sankey et al., 2010a, Sankey et al., 2010b, Perlman et al., 2012, DelDalle & Gaunet, 2014, Payne et al., 2015). Studies have also shown that horses seem to remember the positive experience of positive reinforcement training up to eight months after testing (Sankey et al., 2010a, Sankey et al., 2010b). The horses trained mainly with positive reinforcement in the present study spent more time in proximity to the stranger compared to the owner, which could indicate that these horses had a more curious mind-set and investigated the chance to receive a treat when tested with the stranger. If considering that the owners were instructed to stand passively and relaxed which could be a signal for the horse that it would not receive any treat, as well as they had not received any treat during the previous experimental episodes. Whilst the stranger could have been perceived by the horses as a novel person that the horses did not have any reference behaviour for when a treat can be expected and therefore investigating the stranger more. Further results from the present study, which emphasises an increased curiosity in horses trained with positive reinforcement, is the significantly more time they spent in contact with the stranger, whilst the horses trained with negative reinforcement does not have any contact at all with the stranger. When being separated from either the owner or the stranger, the horses in the positive reinforcement group spent significantly more time in proximity to the door. Again, this might suggest that these horses even more readily wanted to leave the arena compared to the horses in the other two training groups. Although, it is difficult to draw any conclusions by this since we cannot know if it is because it wants to be with conspecifics or if it is the lack of treats in the absence of a human and therefore the horse wants to be with the human.

The owner’s attachment style revealed a tendency of an effect on horse behaviour. The horses in the group with an owner that had an ambivalent attachment style walked significantly more compared to horses with owners that had distanced or secure attachment style. Even if not significant we can still see a trend during the reunion episode where the horses that walked more also spent more time in human proximity, indicating that they walked in and out of proximity at a higher rate which is very similar to an ambivalent behaviour pattern (Håkansson & Tengström, 1996, Ainsworth et al., 2015). The horses with an owner that had a secure attachment style did not spend as much time in human proximity as the horses with an owner that had ambivalent or distanced attachment styles, which do align with the behaviour related to a secure attachment style (Håkansson & Tengström, 1996, Ainsworth et al., 2015). This may indicate that the horses mirror the owner’s attachment style or that the owners chose a horse that have a similar behaviour pattern to their own when they are buying a horse. However, the questionnaire refers to attachment style between adults and not the attachment style between owner and animal, so the relevancy of the attachment style results can be questioned. Although, the present study revealed some similarities between the owner’s attachment style and the horse behaviour which may suggest that the questionnaire can be used when investigating human-animal relationships as well. The small number of owners in the ambivalent and distanced attachment style groups could also be a reason that no significant results were found. Since specific attachment style was not a selection criterium to participate in this study it could be interesting to perform a study that would investigate this matter and select the subjects according to their attachment style.

To conclude, this study suggests that horses are affected by the presence of a human, regardless of it being the owner or a stranger. Hence, in alignment with previous research, the results from the present study showed that horses consider a human as a safe haven (Górecka et al., 2007). In addition, this study revealed that the horses trained with positive reinforcement displayed a more curious attitude towards a stranger than horses trained with negative reinforcement or mixing positive and negative reinforcement, which might be related to the positive association between human and a treat.