Behavior Genetics of the Horse (Equus caballus) - Genetics and the Behavior of Domestic Animals - page 252

experience and subjective judgment of handlers or caretakers are not always
reliable. A handler's prior expectations of how horses should react may
affect a horse's behavior, even if exact instructions are given on how to
perform a test. Finally, standardized questionnaire surveys completed by
knowledgeable handlers may also have limitations. However, comparing
questionnaires with measures from standardized behavioral test assessments
produce more valid and reliable results.
Finding easy and reliable ways to identify temperamental differences has
two main advantages: the ability to predict behavior, and the means to select
horses with suitable work-related characteristics. The following review is not
exhaustive, but is intended to show methods used by researchers to assess
temperament traits. Some of these tests are practical and can be conducted
by a lay person, but others are less practical. Suspending objects from a ceil-
ing would be difficult to apply in typical riding centers or schools, where the
ceiling is usually high. However, methods using novel objects such as an
umbrella to test reaction to sudden novelty, or an unfamiliar arena to test
emotional reactions to social isolation are more practical. The aim of temper-
ament tests for horses is that interested people can perform the tests them-
selves and make use of the results.
TEMPERAMENT TESTS
Behavioral Measures
Visser et al. (2001) measured variation and consistency of horses of the same
breed and age, reared under controlled housing and management conditions.
Forty-one Dutch Warmblood horses were tested in two behavioral tests (novel
object and bridge test) at 9, 10, 21, and 22 months of age. The study identified
two temperament traits (flightiness and sensitiveness), but did not show long-
term consistency of traits. Seaman et al. (2002) recorded the behavior of
thirty-three young horses in three behavioral tests (novel arena, human
approach, novel object). Tests were repeated three times with an average of
9 days between trials. The responses of horses in a human-approach test and
novel object test were similar, but inconsistent: only the open-field test was
consistent over time. The authors suggest responses in the open field may
reflect a core factor of temperament. A questionnaire completed by the farm
manager confirmed the results. Hausberger and Muller (2002) used posture in
224 adult horses to evaluate factors involved in reactions to humans. The
results showed breed differences but, more importantly, caretaker involvement
revealed differences between the groups. One person cared for every horse
in one school, suggesting the important influence of the caretaker on the well-
being of the horses and their reactions to an unknown person. Sondergaard
and Halekoh (2003) housed 40 young Danish Warmblood horses either alone
or in groups of three after weaning. One half of each group received handling