Win-shift and win-stay learning in the rainbow lorikeet

The tendency to win-shift (to better learn to avoid, rather than return to, recently rewarded locations) has been demonstrated in a variety of nectarivorous birds and in honeybees. It is hypothesised to be a cognitive adaptation to the depleting nature of nectar. In the present study, we report the first attempt to test for a win-shift bias in a nectarivorous parrot, the rainbow lorikeet (Trichoglossus haematodus). This species differs from others tested for a win-shift bias in that it is a facultative, rather thanan obligate nectarivore. We tested a captive-reared population of the birds on a shift/stay task at long and short retention intervals. The data show no evidence of either a win-shift or a win-stay bias. The birds demonstrated efficient spatial search ability and above chance performance for both shoft and stay contingencies at long and short delays. These data suggest that an innate tendency to win-shift may not be present in all avian nectarivores, or that the role experience plays in shaping such behaviours is different for different species.