UCLA

The integration of spatial maps in pigeons was investigated using a spatial analog to sensory preconditioning. The pigeons were tested in an open-field arena in which they had to locate hidden food among a 4x4 grid of gravel-filled cups. In phase 1, the pigeons were exposed to a consistent spatial relationship (vector) between landmark L (a red L-shaped block of wood), landmark T (a blue T-shaped block of wood) and the hidden food goal. In phase 2, the pigeons were then exposed to landmark T with a different spatial vector to the hidden food goal. Following phase 2, pigeons were tested with trials on which they were presented with only landmark L to examine the potential integration of the phase 1 and 2 vectors via their shared common elements. When these test trials were preceded by phase I and phase 2 reminder trials, pigeons searched for the goal most often at a location consistent with their integration of the L --> T phase 1 and T --> phase 2 goal vectors. This result indicates that integration of spatial vectors acquired during phases I and 2 allowed the pigeons to compute a novel L --> goal vector. This suggests that spatial maps may be enlarged by successively integrating additional spatial information through the linkage of common elements.