Table 1 Subject’s number, gender, birthdate and weaning date
Subject number Gender
Birthdate Weaning date
06F Male
May 1, 1996 September 27, 1996
07F Female
April 20, 1996 September 20, 1996
12F Female
April 6, 1996 September 11, 1996
18F Female
April 11, 1996 August 24, 1996
24F Female
April 2, 1996 August 26, 1996
30F Male
April 11, 1996 September 11, 1996
34F Male
May 28, 1996 August 26, 1996
35F Male
March 18, 1996 August 19, 1996
45F Male
April 6, 1996 September 27, 1996
53F Female
April 30, 1996 September 11, 1996
57F Male
April 22, 1996 September 27, 1996
00F Female
April 27, 1996 September 27, 1996
37F Male
April 22, 1996 September 20, 1996
100F Female
March 16, 1996 August 2, 1996
measured. The small paddock consisted of a mostly dirt substrate. Mares and foals were brought from the large field to the small paddock approximately five times a week until
early July. After this time, they were brought to the small paddock one to two times a week. Mares and foals would spend several hours in the paddock each time they were
housed there.
Foals were weaned at approximately 4 months of age. Therefore, foals were not weaned at the same time, though they were sometimes weaned in groups of two or three.
Throughout the weaning process, weaned foals remained in the pasture with unweaned foals and their mothers. The dams of weaned foals were moved to another pasture on the
premises. The practice of weaning foals at 4 months of age is a standard practice at the Smithonia Breeding Center.
Ž The caretakers distributed grain twice daily between 0830 and 0900 and between
. 1700 and 1730 in buckets approximately 3 m apart. Throughout the study there were
Ž .
Ž .
between 11 and 18 buckets from which 14 post-weaning to 28 pre-weaning animals fed. Distribution of food was not under the researcher’s control.
3. Methods
Observations were made on mares and foals from the middle of June until the last foal was weaned in September. Data were recorded for foals post-weaning until
approximately 6 months of age. 3.1. Pre-weaning data collection
Ž .
Multiple 30-min random focal samples Altman, 1974 were conducted on each Ž
. mare–foal pair weekly. All occurrence Altman, 1974 observations were also made
during times when supplemental feed was given. Only agonistic data were recorded during supplemental feedings. Within each 30-min sample session the following types of
data were recorded: agonistic encounters, non-aggressive encounters and spatial relation- ship.
3.1.1. Agonistic encounters: All agonistic encounters involving the focal mare and foal were recorded. For each
encounter the initiator of the aggression and the recipient of the aggression were recorded. A description of the action of the aggressor and the reaction of the recipient
were also recorded. The following behavioral patterns were scored as aggressive:
1. Head threat: The extension of the aggressor’s head and neck towards another individual while laying the ears against its head.
2. Bite threat: The aggressor’s ears were laid back. Its mouth was opened and a biting motion was made while it moved its head or whole body toward another animal, but
no contact was made. 3. Bite: The aggressor’s ears were laid back and its teeth were closed on some body part
of another animal. 4. Kick threat: Any of the three following behavioral sequences were scored as a threat
to kick: a. The aggressor, with its ears laid back, made a rapid movement so as to place its
hindquarters near another animal. b. The aggressor, with its ears laid back, raised one hind limb and held it in the air
while in a position so as to potentially strike another. c. The aggressor, with its ears laid back, rapidly struck with one or both hind limbs
towards another animal, but no contact was made.
5. Kick: With its ears laid back, one or both hindlimbs of the aggressor were projected outward rapidly and struck another animal.
6. Strike threat: The aggressor’s ears were laid back and its head and shoulders were oriented toward another individual. One or both forelimbs moved outward and
forward toward the other animal but no contact was made. 7. Strike: The aggressor’s ears were laid back, and one or both forelimbs moved
outward and struck the body of another animal. 8. Chase: With its ears laid back, the aggressor pursued an individual, for at least one
full stride. The following behavioral patterns were scored as being submissive responses to one
of the previously mentioned aggressive behaviors: 1. Head turn: The recipient of the aggression turned its head away from the aggressor.
The recipient’s limbs did not move. 2. AÕoid: The recipient of the aggression moved away from the aggressor.
3.2. Non-aggressiÕe behaÕior The following behavioral pattern was recorded as a ‘‘win or lose’’ even though no
aggression occurred: 1. Approach and retreat: This pattern was scored whenever one individual immedi-
ately moved away from an animal that had just moved to within 2 m of it and had failed to exhibit an aggressive behavior.
Only interactions in which there was a clear winner and loser were used to calculate the dominance hierarchy. In these instances, the animal that approached, threatened
aggression, or aggressed was scored as a winner. The animal that responded either with a head turn, avoidance, or retreat was scored as a loser. Approach and aggressive
behavior that did not elicit a submissive response was not used in the calculation of the dominance hierarchy. All aggressive behavior was used to calculate rates of aggression.
An animal was considered subordinate to another animal if it lost five encounters and won none. If an animal won one or more encounters with a particular foal, then, in
addition to the required five losses, that foal had to lose another two encounters for each win in order to be scored as the subordinate foal. If a dyad existed that did not meet the
above-mentioned criteria, the foal that lost the greater number of encounters was scored subordinate.
In order to determine the rank order among mares and foals, these dyadic relation- ships were evaluated, such that the highest ranking mare or foal submitted to the least
number of mares or foals. The foal rank order was compared to the rank order of the Ž
. dams, using Spearman’s rank order correlation coefficient rho
SPSS to test for
maternal influence on the rank of the foals. A correlation test was also performed on foal Ž
. rank and foal birth order using Spearman’s rho. The wither height in inches and weight
Ž .
in grams of the foals were also subject to a Spearman’s rho correlation test. Body size Ž
. measurements were taken 1 week prior to weaning approximately 4 months of age and
Ž .
during the last week of the study approximately 6 months of age . A Mann–Whitney test was used to test for any differences in rank among colts and fillies.
Rates of aggression were calculated for individual mares and foals by dividing the focal animal’s total number of aggressive interactions by the total number of hours of
observation in which agonistic behaviors were recorded. These aggression rates were broken down into rate of aggression given, rate of aggression received, rate of received
aggression that was ignored and the rate of aggression given by the focal animal that was ignored by the recipient. All of the specific rates of aggression were calculated in
the same manner as the overall rate of aggression.
Aggression was also categorized as a percentage of aggression administered up the dominance hierarchy for each individual as well as the entire herd. The rates of
aggression given, aggression received, and aggression ignored were all tested for correlation with the rank of the individual using a Spearman’s rho. Mares and foals were
also ranked in accordance to their respective rates of giving and receiving aggression. These ranks were tested for correlation using the Spearman rank order correlation.
Aggression rates per subordinate were calculated for mares and foals post-weaning. Total aggression to subordinates was divided by the total number of subordinates. This
was then divided by the total number of hours of observation for each animal. An
ANOVA was performed to test for any effects of gender on the type of aggression most often displayed. All correlations were performed one-tailed as the direction of the
correlation was predicted prior to collection of data.
3.3. Spatial relationships The third category of data involved the focal animal’s spatial relationships with other
animals. Every 2 min during each 30-min focal session the following data were recorded:
1. The identities of all animals within 1 m of the focal animal. 2. The identities of all animals greater than 1 m away but within 5 m of the focal
animal. 3. The identities of all animals greater than 5 m away but within 10 m of the focal
animal. These spatial relations were used to determine each mare and foal’s most preferred
associate. The most preferred associate was defined as the animal that was observed within 5 m of the focal animal the greatest percentage of 2-min samples. A t-test for
paired samples was used to test for an effect of the foal’s gender on the gender of its most preferred associate. Only fillies and colts were used for this test as the small
number of early pre-pubertal castrates in the study group precluded their use in these analyses.
In order to test whether or not a foal preferentially associated with the foal of its dam’s most preferred associate, the probability of the focal foal associating with the foal
of the dam’s most preferred associate if association was random was determined. This was tested using a t-test for paired samples. Physical proximity rates were calculated for
mares and foals by taking the total number of animals that were observed within 10 m of the focal animal and dividing by the total number of hours in which spatial data was
recorded. These rates were then correlated with rank in mares and foals. A correlation test was also run between the sociability rates of mares and the sociability rates of their
foals to test for any maternal influence on sociability. Aggression rates to preferred associates were calculated and a t-test for paired samples was used to test for any
differences in aggression towards other foals and preferred associates.
4. Results