Table 1.
Sequence of testing conditions for all birds.
Fig 1.
Plasma androgen concentrations in female and male cowbirds between breeding conditions and performance of spatial and colour tasks.
Means are presented with ± SE. Asterisks indicate p ≤ 0.05.
Fig 2.
A) Peak performance on the spatial delayed-matching-to-sample touchscreen task.
Performance was calculated from the last three sessions of the randomized retention intervals phase. Means of raw data are presented with ± SE. Proportion correct expected by chance equals 0.33. Two males were only tested in non-breeding condition and their performance was weighed more heavily in the raw means than in the linear mixed models, for which least squares means of significant interactions are presented in B and C. B) Summary by sex and breeding condition of data shown in A), showing least squares means ± SE of arcsine square root transformed data. Breeding males were significantly better than non-breeding males, but there was no difference between non-breeding and breeding females. In addition, breeding males performed significantly better than breeding females. C) Summary of data shown in A) by retention interval (RI), showing least squares means ± SE of arcsine square root transformed data. Males performed significantly better than females at the 15s RI. Males performed significantly worse at the 60s RI than at the 15s RI. Females are in light grey and males are in dark grey. Asterisks indicate p ≤ 0.05.
Fig 3.
A) Peak performance on the colour delayed-matching-to-sample touchscreen task.
Performance was calculated from the last three sessions of the randomized retention intervals phase. Means of raw data are presented with ± SE. Proportion correct expected by chance equals 0.33. One female and one male were only tested in non-breeding condition and these missing points were corrected for in our linear mixed model, for which the only significant interaction is shown in B. B) Summary of the data shown in A) by sex and breeding condition, showing least squares means ± SE of arcsine transformed square root data. Females performed significantly better in breeding than in non-breeding condition, with no effect of breeding condition for males. Asterisks indicate p ≤ 0.05.
Fig 4.
Comparison of performance between the spatial and colour tasks during the testing sessions (see Fig 1).
Female and male cowbirds performed significantly better on the spatial task than on the colour task, indicated by asterisks. Performance of both females and males was significantly better than chance, indicated by the dashed line. Means are presented with ± SE. Asterisks indicate p ≤ 0.05.
Fig 5.
Transition from spatial to colour delayed-matching-to-sample touchscreen tasks.
The figure shows the proportion correct in the colour task when the correct match was near or far from the spatial location of the sample. A) In the first three sessions on the colour task (retention interval [RI] = 0 s), both males and females were more likely to make the correct colour choice when the matching stimulus was near the location the sample had occupied. Random performance was 0.5 (dashed line) because there was only one distracter presented during the choice phase. B) After two weeks of testing, when the Progressive RI testing phase began (RI = 5 s), the location of the matching colour stimulus had no effect on performance. Random performance was 0.33 (dashed line) because there were two distracters presented during the choice phase. Means are presented with ± SE. Asterisks indicate a significant difference (p ≤ 0.05) between near and far trials.