Application of a Network-Analysis Algorithm for the Definition of Locations in Open Field Behavior: How Rats Establish Behavioral Symmetry in Spatial Asymmetry

The present study applied network analysis to scrutinize spatial behavior of rats tested with symmetrical or asymmetrical layout of 4, 8, or 12 objects along the perimeter of a dark, round arena. We considered locations as the units of a network (nodes), and passes between locations as the connections of the network. In terms of Ethovision parameters, there were only minor differences between rats tested in either symmetrical or asymmetrical object layouts. However, network analysis revealed substantial difference in the behavior between the layouts. For the network analysis, we first defined locations in the environment, where each ‘location’ was a cluster of stopping coordinates (defined as no progression for at least 1 second) extracted from Ethovision. From the set of locations and the passes between them we extracted the network analysis parameters: for each nodedegree,clustering coeficient and shortest mean path were calculate. In addition the average network degree,clustering coeficient and shortest mean path were extracted for each rat. It was found that behavior in either a symmetrical or asymmetrical layout of 4 objects, the key locations coincided with the objects. However, in the asymmetrical layout with 4 objects, additional key locations were spaced at a distance that was identical to the distance between other objects, forming behavioral symmetry among the key locations. In other words, it was as if the rats imposed behavioral symmetry in their spatial behavior in the asymmetrical environment. We suggest that wayfinding was easier in symmetrical environments, and therefore, when the physical attributes of the environment were not symmetrical, rats behaviorally established a symmetric layout of key locations, thereby gaining a more legible representation of the environment despite its more complex physical structure. Altogether, the present study adds a behavioral definition for a location, a term that so far has been mostly discussed according to its physical attributes or neurobiological correlates (place, border and grid neurons). Moreover, the network analysis enabled to identify a location and to assess its importance, even when that location did not have distinctive physical properties.