**5. Discussion**

According to Keeling [10], in the absence of ewes, subordinate rams tend to initiate some of the agonistic interactions challenging high-ranking rams. After losing an encounter, the subordinate ram may display submissive behaviour to the winner. In the present study, agonistic interactions involved pushing, horn threat, head butt, chasing and low stretch. However, these behaviours were not quantified. These results were in accordance with the behaviours observed by Keeling [10] where low stretch and horn threat were observed as an agonistic behaviour in rams. The author described the low stretch behaviour as a threat display in which a ram lengthens its neck forward and horizontal to the ground. Pelletier and Festa-Bianchet [9] observed similar agonistic interactions during contests in rams. The observed behaviours included front kick, frontal clash, rubbing, butt, non-contact displacement, and horn threat. Similar to the present study, Roberts et al. [11] observed head butt behaviour as common in rams. This is when rams are slamming their heads together until one ram withdraws from the encounter. Squeezing was another behaviour, which was commonly observed in the present study where a ram would squeeze itself between closely aligned rams. This was similar to the observation by Erhard et al. [12] where a feeding ram would stand almost parallel to the wall holding the feed hopper and eventually other rams squeezed in between the feeding rams.

Roberts et al. [11] described 'win' as a situation where a sheep wins an encounter, either by initiating and displacing another sheep or fending off another ram trying to displace it. The 'loss' is a situation where one sheep loses an encounter, either by starting an encounter and failing to displace the other sheep or by being displaced by another individual initiating an encounter. The insignificant linear relationship between the proportion of wins and time spent at the feeder suggests that rams with a higher number of wins did not necessarily spend more time at the feeder. Some rams with a low proportion of wins were able to spend more time at the feeder compared to rams with higher proportions of wins. Low-ranking rams might have gained more access to the feeder by shifting laterally and squeezing themselves between aligned rams, or they would wait for the dominant ram to turn its head down and they would quickly rush into the available space. Squires and Daws [8] reported results similar observations. Dwyer [13] suggested that when the feeding space is limited there is an increase in displacement at the feeder, and some of the sheep will stop feeding and become non-feeders. Thus, a decrease in time spent at the feeder was due to forceful displacement and disturbances at the feeder of low-ranking rams by high-ranking rams. In dairy cattle, the reduction of feeding space per cow in dairy cattle increases agonistic encounters even if the feed is provided *ad libitum* [14]. The same was concluded in dairy goats by Jørgensen et al. [15], suggesting that there is a decrease in time spent at the feeder if there is limited feeding space.

A significant correlation between the proportion of wins against horn length and chest girth suggests that rams might have used their larger horns and wider girth to fight and gain feeding space in the feeder, thereby increasing the proportions of wins. Body mass and horn length mainly affect social rank [5]. Bergeron

### *Animal Husbandry*

et al. [16] stated that the heaviest males with long horns are generally at the top of the hierarchy. In this study, there was strong positive correlation existed between the proportion of wins and chest girth. Body weight was positively correlated with withers height, scrotal circumference, horn length, and time spent feeding but strongly and positively correlated with chest girth. In accordance with findings in this study, Maksimović et al. [2] also obtained a significant relationship between body weight and chest girth.
