Fig 1.
Photograph of two-chamber CPP apparatus.
The CPP apparatus used consisted of two distinctly decorated conditioning chambers (e.g., blue with white stripes on the left, yellow with white dots on the right) separated by a removable partition.
Fig 2.
Visualization of the nucleus accumbens location used for immunolabeling.
Location of the nucleus accumbens (NAc) on a coronal section (A) with corresponding mu opioid receptor label location on a confocal image (B). Abbreviations: MSt = medial striatum.
Table 1.
Descriptive statistics for behavioral data, separated by non-singers (n = 3) and singers (n = 9).
Behavioral data are represented in behavior rate, apart from time spent singing which is represented in seconds. Asterisk denotes statistical significance (p<0.05).
Fig 3.
Singers developed a strong place preference for a place associated with flock mates, and non-singers preferred separation.
Starlings that sang full songs (singers, left) spent more time on the flock-conditioned side of the CPP apparatus compared to the separation-conditioned side. Birds that did not sing full song bouts (non-singers, right) spent more time on the separation-conditioned side than the flock-conditioned side. The Y axis shows the time spent in each chamber (flock, separation) in seconds, with the dotted horizontal line at 450 seconds denoting equal time in each chamber. The X axis shows the conditioned chamber (flock, separation) for each group (singers, non-singers). Numbers denote significant p-values (p<0.05). Singers are shown in blue, non-singers in green, females by circles, and males by squares.
Fig 4.
Starlings removed their foot from hot water more quickly after being separated from the flock overnight (separation), compared to when they were housed in the stable flock (flock). The Y axis shows the latency to withdraw foot from a hot water bath. The X axis shows repeated tests in flock (1, 2) and separation (1, 2) conditions following an ABAB design, as well as the mean from the flock and separation conditions. Analysis was performed on the mean of each test (flock avg, separation avg; see text for details). Numbers denote significant p-values (p<0.05). Thick black lines show the mean latency to withdraw across all birds for each measurement and error bars represent the SEM. Singers are shown in blue, non-singers in green, females by circles, and males by squares.
Table 2.
Principal component analysis eigengvalues and variance for each component.
Table 3.
Principal component analysis results with the loadings >0.60 for each selected behavior in bold.
Fig 5.
Song and social spacing principal components positively correlated with the number of MOR-positive cells in the NAc for singers.
Starlings that sang full songs (singers) had significant positive correlations between the number of mu opioid receptor labeled cells in the NAc and principal components (PCs) for both Song (A) and Social Spacing (B) behaviors. The Y axes show the loading value for each PC. The X axis shows the number of MOR-positive cells counted within the bounds of the NAc. Singers are shown in blue, non-singers in green, females by circles, and males by squares. Differences in MOR immunolabeled cells are shown in representative photomicrographs of the NAc for a non-singer (C) and singer (D). Images were taken using a confocal (Zeiss 710). Arrows indicate example counts of MOR-positive cells.