Response to Human Familiarity in Ovis aries

Jackie Barnes, Megan Fritz, and Lilly Lo

BI/PSY 441






            Sheep are social animals that are able to recognize others based on individual characteristics. In our experiment, we examined whether domestic sheep were able to differentiate and show preference to humans based on familiarity. We hypothesized that the sheep would respond more to a familiar person than an unfamiliar person. Using seventy-six sheep, we had a familiar and unfamiliar person enter an enclosed area and counted the number of sheep that responded to each person. We found that overall, the sheep tended to prefer the familiar person, however, the data were not significant (p=0.317).



Key words: sheep, Ovis aries, facial recognition, familiarity preference

            Ovis aries, better known as domestic sheep, are very common livestock animals.  They are known to have flocking instincts and are considered to be social animals.  Female sheep are identified as ewes, male as rams, and babies as lambs. Ovis aries are herbivores and can survive on a simple diet consisting of grass and hay.  When it comes to feeding, grazing plays a very important role in this behavior (Sevi, et. al, 1998). A good majority of their time grazing is spent eating as opposed to standing, walking, or other activities. 

            Group size also plays an important role in grazing behavior. In a study done on group size and its effect on grazing behavior, it was found that ewes in smaller groups had significantly smaller herbage and nutrient intake than others in a medium to large group (Sevi, et. al, 1998).  Also, ewes in small groups gained less weight than those in medium to large groups, which suggests that they had spent less time grazing and had shorter meals (Sevi, et. al, 1998).   Large groups of sheep, therefore, are more likely to consume a higher quantity of food and to be more dependent on and responsive to food sources because grazing is a social behavior.

            Ovis aries have good visual sharpness and have excellent hearing (Kendrick, et. al, 1995).  They are able to distinguish faces of members of their own breed from those of different species and breeds and can distinguish between males and females of the same breed.  Individuals within the flock are distinguished based on traits such as body shape, posture, and specific movements. In addition to being able to visually discriminate sheep from humans, they can also make a distinction between a sheep and a human voice (Kendrick, et. al, 1995). 

            Because of their flocking instincts, being able to recognize other individual sheep based on facial characteristics is very important because these characteristics can convey not only the breed of that sheep, but also their individual identity (Kendrick & Baldwin, 1987). This allows them to recognize relatives and determine between familiar companions and sheep with which they have had no previous contact. Facial recognition occurs mainly in the neural circuits of the right temporal cortex of the sheep’s brain, but also in the hippocampus and the basal amygdala. Because of the location in the brain that this processing occurs, faces occurring in their left visual field are more recognizable. When such characteristics are presented in the left visual field, sheep are able to distinguish between them and determine their familiarity with that individual sheep (Broad et al., 2000).

            This experiment was designed to determine whether Ovis aries are able to distinguish between familiar and unfamiliar humans. We know from previous research that they are able to both distinguish sheep from humans and to identify individual sheep based on facial characteristics, so we expect that they may be able to individually recognize humans. Our hypothesis was that the sheep would respond more to a familiar person than an unfamiliar person.




            To test our hypothesis, we used seventy-six adult Ovis aries (five males and seventy-one females) as a part of our experiment. The experiment was conducted on March 21, 2009 at the A-Z Farm in Oregon, Wisconsin. A large enclosure of 300 square meters was divided off from the rest of the barn with various metal, wood, and wall components. All seventy-six sheep were continuously enclosed within this area (before, during, and after our experimental procedure), which was constantly bedded with hay. To begin our procedure, we poured the sheep’s normal corn/feed mixture into two identical 22.7 liter white buckets in an area on the opposite side of the barn where they could not see, but could still hear the corn being poured into the buckets. Equal amount of feed was poured into each bucket, filling the buckets two-thirds of the way full. The area, which consisted of a 10 m by 20 m area aligned with a 10 m by 10 m area, was visually divided in half by the two researchers that were observing the sheep’s responses.







            Figure 1. Diagram of testing area.


            For our experiment, we had two people enter the same gate on one side of the enclosure, one with whom the sheep were familiar, and one with whom they were not. The familiar person was the owner of the farm who fed the sheep daily, and the unfamiliar person was the third researcher whom the sheep had no previous contact with. Both people entered the enclosure from the same gate at the bottom right side of the enclosure, each carrying one of the buckets, and headed to stationary wooden feeders on the opposite ends of the large enclosure. Each person then poured the corn/feed into the feeder and the other two researchers, who were standing outside of the enclosure, counted the number of sheep that were responding to each person. The area was visually divided in half, by the two researchers, with endpoints being they edge of a hay bail and the hinge of the gate. With the two areas approximately equal at 150 m­2, the two researchers counted the number of sheep on each half, respectfully responding to each person with feed.

            This experiment was conducted twice, at both 0800 and at 1200. The exact same experimental procedure was used during both times the experiment was conducted. The important difference between these two times was that the AM test occurred during the normal feeding time for the sheep, whereas the sheep are not regularly fed at 1200.   Also, the familiar and unfamiliar person went to opposite feeders the second time the experiment was conducted.  The only other difference was that 76 sheep were used in the AM test and only 75 were used in the PM test, as one ewe had a lamb between the two experimental periods and was removed from the enclosure.  

            Statistical analyses were conducted using chi-square analysis on Microsoft Office Excel to determine the differences between the expected results and the actual results of our experiment. Analysis was conducted for the overall effect, the effect in the AM test, and the effect in the PM test.



            Results of our experiment indicate that overall, the data were not significant (p=.0317). A trend of the data shows that the sheep are more likely to respond to the familiar person, yet results were not statistically significant.

In our first experiment, we would expect that 36 sheep would respond to each person if this response occurred by chance. We observed that 48 sheep went towards the familiar person and 28 to the unfamiliar person, so we could conclude that more responded to the familiar person.

In the second experiment, we again expected that half (n=35.5) would respond to each person. What we observed was that 40 sheep went to the familiar person and 34 went to the unfamiliar person, again meaning that more responded to the familiar person. By comparing the overall response to familiar and unfamiliar people to what we would expect, we found our p-value of .317.


Figure 2. Observed number of sheep responding to familiar and unfamiliar people in AM & PM hours.






            Overall, we found that our results were not significant, but displayed a trend of the sheep responding more to the familiar person.

            The time at which we conducted this test was the normal feeding time for the sheep and the familiar person was the person that fed them on a daily basis, so the sheep may have perceived an even stronger relationship between the person and the food stimulus at that time. The sheep are fed in both places.  Also because the sheep are not regularly fed in the afternoon, they were likely full from grazing throughout the morning and not as responsive to the food stimulus. It is also possible that the sheep associated the unfamiliar person with the food stimulus from their exposure in the morning test to the afternoon test, creating a lesser response in the afternoon.

            Within our experiment when the unfamiliar and familiar person switched places, it did not have an affect on the sheep’s response.  This is likely due to the sheep’s ability to recognize other sheep based on individual characteristics and to be able to distinguish sheep from humans (Kendrick et al., 1995). Combining these two abilities, it is possible that sheep can distinguish between the individual characteristics of humans based on their familiarity with them. This is what we expect is occurring in our experiment that has led to our results.

            It is also possible that the sheep are connecting the familiar person used in our experiment to the reinforcing stimulus of food. Because this person is the one who feeds them on a daily basis, the connection between his identity and the stimulus lead to an even stronger level of response.  

            Because two tests are not enough, we would need to conduct this experiment with more trials in order to obtain significant results. First, we believe it would be a more ideal situation to conduct multiple tests on various different days so the sheep cannot form specific associations between the morning and afternoon tests. Another possibility would be to eliminate the variable of time of day completely and focus on only one variable. The season of the year which we conducted the experiment may have created an additional variable that we did not account for, so it is important to realize that the results obtained may only hold true for the typical birthing season.

            In expansion of this study, it would be interesting to balance the number of males and females used as subjects. Because this ratio was not equal in our experiment, the overwhelming proportion of females to males is not representative of the entire sheep population. Also, because different species of sheep have different feeding and grazing behaviors; domestic sheep have evolved from various combinations of these behaviors (Steinheim et al., 2005). It would be interesting to test which specific behaviors derive from each individual species. This would provide more insight to the evolution of feeding behaviors and how they relate to human recognition.









Broad, K. D., Mimmack, M. L., Kendrick, K. M. (2000). Is right hemisphere specialization for face discrimination specific to humans? European Journal of Neuroscience, 12: 731-741. Retrieved from Academic Search Elite database.

Kendrick, K. M., Atkins, K., Hinton, M. R., Broad, K. D., Fabre-Nys, C., & Keverne, B. (1995). Facial and vocal discrimination in sheep. Animal Behaviour, 49(6): 1665-1676.

Kendrick, K. M. & Baldwin, B. A. (1987). Cells in temporal cortex of conscious sheep can respond preferentially to the sight of faces. American Association for the Advancement of Science, 236(4800): 448-450. Retrieved from JSTOR database.

Sevi, A., Casamassima, D., Muscio, A. (1999). Group size effects on grazing behaviour and efficiency in sheep. Journal of Range Management, 52: 327-331. Retrieved from JSTOR database.

Steinhein, G., Nordheim, L. A., Weladji, R. B., Gordon, I. J., Adnoy, T., & Holand, O. (2005). Differences in choice of diet between sheep breeds grazing mountain pastures in Norway. Acta Agriculturae Scand Section A, 55: 16-20. Retrieved from Academic OneFile database.