The purpose of the study was to attempt to identify personality traits in domestic rabbits
(Oryctolagus cuniculus) and to evaluate a range of tools, suitable for use in a shelter
setting, that can be used to measure personality traits. A literature review highlighted
limited evaluation of reliability and validity in rabbit personality research published to
date. Additionally, there is a lack of clarity on what is being measured by some
behaviour tests that are currently employed in animal personality research and there
are limited tools available to measure domestic rabbit responses to humans.
Chapter three highlights several uses of rabbit behaviour and personality data in
United Kingdom (UK) shelters. Shelter staff reported uses for understanding the
behaviour of an individual rabbit to support the management of the individual while at
the shelter and to match the rabbit to the most suitable future home. Challenges facing
shelter staff to collect behavioural data for their rabbits centred around a lack of
resources, specifically time available for collecting behavioural data. An additional
challenge reported by shelter staff was inaccurate information being reported by the
person handing the rabbit into the shelter. To ensure any personality assessment tool
could be integrated into shelter routines, the tools would need to be relatively quick to
complete and should ideally include a range of data collection methods so that a full
picture can be available.
In Chapter four, the results of a behaviour rating survey that was distributed to a selfselected pool of rabbit owners or those that worked with rabbits, using social media
are reported. The survey was also completed by animal care technicians for rabbits
taking part in direct behavioural observations, including a suite of behaviour tests and
observations within the home cage. The use of an online survey enabled a large
number of participants to take part. Following examination of the reliability of the data (interrater) and dimension reduction statistics, three components were retained that
included 15 of the initial 47 items and accounted for 60.6% of the variance in the data
(n=1,234). However, sufficient thresholds for inter-rater reliability were not achieved.
As intended in the selection of survey items, the retained components accounted for
intraspecific social behaviour, human-rabbit interactions (avoidance of humans) and
boldness in relation to the environment. However, only the human-rabbit interaction
component had sufficient distribution of scores across the sample population to
consider this a personality trait.
Behavioural tests are commonly used as measures of an individual animal’s
personality; however, several tests have conflicting interpretations of the underlying
traits that may drive behaviour in these tests. In Chapter 5, a suite of tests were used,
reflecting three commonly used test paradigms for domestic rabbits; the open field test,
novel object test and a new human interaction test. Five human-interaction items
measured were reliable between raters and between tests and two items, location
during subtest 3 where the handler was sat inside the door of the enclosure and a
combined outcome score for subtest 3, 4 (stroke rabbit) and 5 (pick up rabbit) were
retained to create component 2 on the final solution of the principal component
analysis. From two variations of both the open field and novel object tests, two
components were also derived, reflecting exploration and curiosity in rabbits. These
three components were reliable between raters and between tests and accounted for
75.2% of the cumulative variance in the data. The component labelled ‘exploration’
comprising variables of activity in the open field tests were found to negatively correlate
with component 2 from the behaviour rating scale, reflecting avoidance of humans.
This is similar to past research in young rabbits where resistance to handling was
correlated with activity in the open field.
The use of behavioural observations in the home cage environment is rarely performed
for personality assessment in domestic animals due to how time consuming such
observations can be. As a requirement for the tools was to be able to be utilised by
shelter staff, where time constraints are an important factor, home cage behavioural
observations were designed to be quick to complete. Following a pilot test including
three hours of observations over the day, it was possible to determine the behaviours
that could be observed using video cameras positioned adjacent to or above rabbit
enclosures. Additionally, this pilot test revealed that within the times of day available
for testing, none were preferable over any other in terms of the range of behaviours
observed in 12 rabbits. The main study therefore utilised three five-minute sampling
points across the day with the refined ethogram and 30 second focal sampling. It was
not possible to complete dimension reductive statistics on the sample of 16 rabbits
used for this part of the study, although the behaviours observed in the relatively short
time frame did represent activity patterns observed in past research.
Two tools, the behaviour rating survey and suite of behaviour tests, are proposed to
be retained for future examination of the utility of these tests in a shelter setting to
measure rabbit behaviour and personality. These retained tests would provide
information on an individual rabbit’s social behaviour (intraspecific), response to
humans, boldness in relation to the environment, exploration and curiosity. Future
research is recommended to determine the suitability of these tests for use in shelters,
and to understand the predictive validity of these tools. That is to understand the
usefulness of rabbit personality assessments to identify aspects of behaviour that are
stable between different environmental contexts, such as between a shelter setting
and within a home following being rehomed.
http://nectar.northampton.ac.uk/13599/