Most animals possess an ability to recognize behaviorally relevant locations in their environments. By studying the behavioral mechanisms that guide goal recognition, we can gain insight as to how an animal forms memory representations of goal locations. The two sources of information examined in the present study are feature cues (color) associated with the goal and the goal’s location in space. Previous researchers have indicated that species with differing ecological spatial demands may employ different behavioral mechanisms to remember and navigate toward a goal. For example, food-storing birds typically display a preference for spatial information over feature information. Although homing pigeons are not food-storers, they are known for their navigational abilities. Some studies indicated that pigeons may also display a preference for spatial information, yet other studies found no preference between spatial and feature information. The inconsistency of cue preference in the homing pigeon literature may reflect differences in the size of the experimental space or task demands. In the current study, pigeon representational cue preference was examined in a large, outdoor setting with distances up to 100 meters using two tasks that differed in goal stability and presumed memory demands.
Experiment 1, simulating a stable environment, was designed with a food goal that always occurred with the same feature cues and always occupied the same location in space. Experiment 2, which simulated an unstable environment, was designed with a food goal that periodically changed with respect to location and associated feature cue. In the stable environment, the pigeons reliably chose the correct goal location after only a few exposures, and continued to perform at near 100 percent accuracy throughout the experiment. On probe trials, the pigeons showed a strong preference for the color cue when both cues were present but set in conflict. However, in the unstable environment, the pigeons were given only two exposures to each goal configuration before testing, and 12 sessions were required before a stable performance of approximately 70% accuracy was achieved. On probe trials, the pigeons showed no preference for either spatial or feature information when the two cues were set in conflict. In pigeons, the relative stability of goal information likely affects memory representation including how the task is approached and how reliably the task is solved.