Rudolf Magnus Institute for Neurosciences, Utrecht University, Utrecht, The Netherlands
Introduction. We are interested in the neurobiological properties of attachment behavior along the lines of John Bowlby's attachment theory. Bowlby's theory is the most prominent in describing socio-emotional development in man. It would have practical advantages if Rattus norvegicus could serve as an animal model, for instance because the existing physiological, behavioral and neuroanatomical knowledge can be used. To examine the usefulness of the rat in this sense we did a series of experiments to see if rat pups are able to distinguish their home nest from a nest of the same postnatal age and if they show a preference for it. Our experimental setup involved stimulation of as many senses as possible, without the rat mother being able to retrieve the pup. The first experiments were done with video playback in the fast-forward mode and manual scoring with the aid of the computer package The Observer (Noldus Information Technology). These experiments showed development of a preference for the home side from beginning of testing on day 8 until a peak was reached on day 17. Then a sharp decline followed resulting in a significant preference for the non-home side on day 23. A control experiment showed that the peak on day 17 was not the result of repeated testing. The experiment we describe here was meant to control for a possible influence of repeated testing on the preference for the non-home side on day 23.
Materials. A perspex cage was made consisting of three compartments: a central test arena with two compartments containing a nest on both sides of it. The nest compartments were separated from the central part by wire mesh, providing optimal contact without the possibility of retrieval. The floor of the nest compartments was covered with a layer of approximately 3 cm of sawdust. The floor of the central test area was covered with a removable rubber mat. Two of these cages were placed side by side to allow for registration of the movements of two pups at a time.
Subjects. Four pregnant Wistar rats were purchased from the Animal Facility of the University of Utrecht. Tree days before the expected delivery date they were placed in the test cage. Thirty-nine pups were born within 24 hours. They were left with their biological mother throughout testing.
Method. Testing consisted of placing a pup in the central test area. This area was divided in two equal parts, and the time was measured that the pup spent on the home side during 15 minutes. One group of pups was tested from postnatal day 8 until postnatal day 25, the other group was planned to be tested once the preference for the non home side had reached significance. Measurement was done full automatically: a Panasonic camera was placed above the test cages, and was connected to a computer, on which the program EthoVision (Noldus Information Technology) was installed. Two arenas were programmed, one for each test cage. Each arena was divided into two zones and time was registered that was spent in the zones.
Results. As in previous experiments, there was an increase in preference for the home side from day 8 on until a peak was reached at day 17. Then a decline followed but, in disagreement with previous experiments, a significant preference for the non-home side did not develop. For practical reasons we decided on day 25 not to wait any longer and test the once only group. This group did indeed show a significant preference for the non-home side.
Conclusion. Rat pups distinguish their home nest from and prefer it to a nest of the same postnatal age in the period when it is most necessary (around day 16-17 there is a combination of good locomotor development and dependence on the mother for nurturance). This preference is followed by a preference for the non-home side, although not always significant. We propose that the preference for home can be regarded as a - albeit primitive and temporary - form of attachment behavior along the lines of Bowlby's attachment theory. Our next experiments will be focused on influencing the biphasic development of the home preference through administration of oxytocin and an antagonist. Oxytocin has been shown to influence various other forms of affiliative' behaviors. The scoring will be done automatically like in this experiment because of two advantages: saving of time through scoring of two subjects at the same time and automatically, and blinding procedures are not necessary.