Correlative evidence suggests that high problem-solving and foraging abilities in a mate are associated with direct fitness advantages, so it would benefit females to prefer problem-solving males. Recent work has also shown that females of several bird species who directly observe males prefer those that can solve a novel foraging task over those that cannot. In addition to or instead of direct observation of cognitive skills, many species utilize assessment signals when choosing a mate. Here we test whether females can select a problem-solving male over a non-solving male when presented only with a signal known to be used in mate assessment: song. Using an operant conditioning assay, we compared female zebra finch (Taeniopygia guttata) preference for the songs of males that could quickly solve a novel foraging task to the songs of males that could not solve the task. Females were never housed with the test subject males whose song they heard, and the only information provided about the males was their song. We found that females elicited more songs of problem-solving males than of non-solvers, indicating that song may contain information about a male’s ability to solve a novel foraging task and that naïve females prefer the songs of problem-solving males.

Cognition data: zebra finch males were given a novel foraging task in which they learned to remove plastic lids from baited wells in a plastic block. The overall task was broken down into five stages: (1) just the block, (2) lids placed adjacent to wells, (3) lids half-covering wells, (4) lids tipped into wells, and (5) lids completely covering wells. To pass a trial, birds had to eat from at least two baited wells out of four in under two minutes, and to pass a stage, birds had to pass three out of four consecutive trials. The data presented are the number of trials it took each male to pass each stage, three being the minimum. If birds did not complete a stage within sixty trials, they were pulled from testing and designated Non-Solvers. 

Stimulus data: songs were recorded from six of the fastest Solvers and six Non-Solvers. These songs were analyzed for length, number of phrases, average length per phrase, average elements per phrase, and number of unique elements per song. Data on male cognitive performance and song analysis is presented here. 

Female preference: female zebra finches were presented with different pairs of songs in operant conditioning chambers. Data is presented here from the initial training period, in which females were presented with pairs of conspecific and heterospecific songs, and from the testing period, in which females were presented with Solver and Non-Solver song. Within the operant conditioning chamber, a hop on one designated perch would trigger one song, and a hop on the other perch would trigger the other song. The number of hops that triggered each song type is presented here. 

For more detailed methods, please see the main article.