Testing behavioral flexibility in pigeons using conditional midsession reversal tasks

Abstract

Midsession reversal examines behavioral flexibility by requiring animals to reverse a discrimination midway within a session. This reversal behavior is controlled by changing temporal factors over the session. We tested behavioral flexibility in pigeons by reversing the contingencies within a session from a visual matching-to-sample (MTS) task to a non-matching-to-sample (NMTS) task. To examine how visual context influenced the reversal, the stimuli were assigned according to two different mapping conditions. In the bounded mapping condition, different stimuli exclusively appeared in either the MTS or NMTS for half of the session, and in the unbounded condition, the stimuli appeared across both the MTS and NMTS halves of a session. In the unbounded condition, pigeons showed a modest switching function, from matching to non-matching, at the reversal boundary. In contrast, in the bounded condition, the pigeons learned the contingences faster and to a higher accuracy and exhibited a more precise switching function at the reversal. The latter outcome suggests there was little or no temporal interference, as the different visual contexts allowed the MTS and NMTS tasks to be segregated within a session. These contrasting results show that reversal interference in the midsession reversal task is influenced by context and temporal factors, with corresponding implications for behavioral flexibility.

Keywords: behavioral flexibility; interference; matching; non‐matching; reversal learning.