Figure 2. Most HVC_RA neurons exhibit peri-song activity.

(a) Normalized calcium transients from 67 simultaneously recorded HVC_RA neurons during production of a song phrase. The red dashed lines delimit five consecutive motifs. (b) Percentages of active neurons during peri-song (54.3 ± 17.8%, SD, purple) and song (59.9 ± 22.5%, gray) are similar (30 phrases, t = −1.1, p=0.29 paired two-sample t test). Box plots show the median, 25^th and 75^th percentiles with whiskers showing ±1.5 interquartile range (IQR). (c) Sample neurons with diverse phrase indices ranging from −1 to 1 and their corresponding calcium traces during 6 motifs over three bouts. Dashed lines indicate bout onsets and offsets. Bars above spectrogram indicate the presence of cage noise related to hopping and wing flapping (yellow) or female calling (FC, orange). (d) Histogram of phrase indices for all 223 neurons from six birds. (e) Undirected song from a different male showing periods of cage noise or hopping behavior (yellow) and feeding behavior (green). Blue boxes indicate the male calling. Red dashed lines indicate onsets and offsets of song bouts.

Figure 2—figure supplement 1. Description of motifs, bouts, and phrases underlying zebra finch courtship song structure.

(a) A spectrogram of a single song phrase composed of 4 bouts and seven motifs (Bout 1: three motifs, Bout 2: two motifs, Bout 3: one motif, Bout 4: one motif). Green bars on top of spectrogram indicate representative silent periods between bouts and numbers above indicate duration in milliseconds. (b) A spectrogram highlighting a trial where the bird sang three different phrases. The magenta bars above the spectrogram indicate silent periods between phrases and the numbers above indicate duration in seconds. Yellow bars below spectrogram indicate cage noise.

Figure 2—figure supplement 2. Proportion of imaged neurons by bird (N = 6 birds, 197 neurons) that exhibited calcium events only before song onset calcium events (Pre-only, 0.44 ± 0.31, SD), only after song offset calcium events (Post-only, 0.26 ± 0.28), or were active before and after song (Pre-Post, 0.29 ± 0.25).

Figure 2—figure supplement 3. 62 events (31 paired pre-song and post-song events) from 31 neurons from six birds.

Showing a small positive correlation (r² = 0.127) between pre-song onset time to post-song onset time in cases where a neuron was active only once before and after singing.

Figure 2—figure supplement 4. Spatial organization of HVC-RA neuron activity in one exemplary bird.

(a) FOV of ROIs identified in one exemplary bird (O262). Peri-song neurons are shown in purple, pan-song neurons are shown in green, and song neurons are shown in red. Scale bar is 100 micrometers. (b) Same FOV as in a, but neurons are color coded by their phrase index. (c) Euclidean distances between neurons shown in figure (a). Distances (in micrometers) were calculated between the three functional neuron pools mentioned in this paper: peri-song neurons, pan-song neurons, and song neurons. Song-song to peri-peri, p=0.02, song-song to peri-pan, p=0.04, Mann-Whitney U Test. 2278 pairwise distances.

Figure 2—figure supplement 5. Pan-song neuron events organized in tripartite groups according to phrase index.

Within each subdivision neurons are organized by Motif onset times. 23 neurons are shown between −1 to −0.33. 56 neurons are shown between −0.33 to 0.33. 53 neurons are shown between 0.33 to 1.

Figure 2—figure supplement 6. Synchronized calcium traces for all available trials for one bird (O248) across directed, undirected, and non-singing behaviors.

Bars above spectrograms indicate cage noise (yellow), female calling (blue), and male calling (red). Each trace under the spectrograms correspond to different neurons but they are not necessarily the same neuron across the trials.

Figure 2—figure supplement 7. Expanded spectrograms from Figure 2, showing female distance calls and tet calls during pre-song and post-song periods.

The labels next to the high-resoultion spectrograms indicate the corresponding audio file name.

Figure 2—figure supplement 8. Expanded spectrograms from two examples in Figure 2—figure supplement 6, showing the spectral structure of cage noise during pre-song behavioral epochs.

The labels next to the high-resolution spectrograms indicae the corresponding audio file name.

Figure 2—figure supplement 9. Event rates for a bird during Directed (561 CEs, 1.67 ± 0.74 std), Undirected (252 CEs, 1.06 ± 0.66), and Non-Singing (53 CEs, 0.19 ± 0.27) periods of behavior (Directed song is significantly different from non-singing, F(2,10) = 4.83, p<0.05).

Each marker on the plot corresponds to a different trial. Event rate was calculated as the ratio of the number of events to the duration of the trial.

Figure 2—figure supplement 10. Comparison between maximum pre-song event rate to the number of introductory notes prior to motif onset across five birds (Mean Event Rate = 0.91 E/s, Mean Introductory notes =~2).

one bird did not have any introductory notes during the imaged trials.

Figure 2—figure supplement 11. Event rate calculated across the full length of 5 trials for one bird (O262).

Event rate was calculated by first binning calcium onsets in 100 ms bins and smoothing with a 1 s moving window. Green dashed line indicates onset of song phrase and red dashed line indicates offset of song phrase. Purple (horizontal) line indicates 2/3 s of the maximum value during the pre-song period. Black arrow with time marks maximum peak song event rate. Magenta arrow indicates when female was introduced. Asterisk indicates onset of short call.