Same-sex mouse dyads produce calls of lower frequency when separated by a perforated divider.

(A) Schematic of experimental setup during separation phase. The arena surface area measured 930 cm2 divided into two equally sized compartments. The walls were 30 cm high. Hole diameters in divider measured 0.5 cm. (B) Schematic of experimental setup during unification phase when the perforated divider was removed. (C) Histogram of peak frequencies during separation (blue) and unification (red) phases. Black arrows highlight troughs used to differentiate frequency bands. LFVs = low frequency vocalizations, MFVs = middle frequency vocalizations, USVs = ultrasonic vocalizations. (D) Quantification of call rates for each call class during separation (blue) and unification (red). Each point represents one same-sex mouse dyad (n = 21 dyads). Black lines represent group mean and error bars represent SD. For statistical analysis a two-way RM ANOVA with Sidak’s multiple comparisons was employed. Call class x context interaction F(3, 80) = 14.80, P < 0.0001. *p < 0.05; **p < 0.01; ***p < 0.001, ****p < 0.0001. (E) Pie charts displaying total call repertoire during separation, and during unification (F). Call classes are color coded: Noisy = blue, LFV = green, MFV = purple, USV = orange. The total number of recorded calls is displayed at the bottom of each pie chart. (G) Spectrograms of exemplary vocalizations during separation phase; first row: Noisy calls, second row: lowfrequency vocalizations (LFV < 32 kHz), third row: middle-frequency vocalizations (MFV 32 – 50 kHz), fourth row: ultrasonic vocalizations (USV > 50 kHz).

Spectro-temporal properties of call classes differ significantly between call repertoires used by separated and united same-sex dyad.

(A) UMAP of call classes recorded during both physical separation (blue shading) and unification (red shading) (total calls n = 3757). Each dot represents one call of either Noisy (blue), LFV (green), MFV (purple), and USV (orange) class. (B – E) Quantification of relative call type distribution (B), average peak frequencies (C), average bandwidths (D), and call lengths (E) during separation (blue) and unification (red). Each dot represents the average of one same-sex mouse dyad. Black lines represent group mean and error bars represent SD. Data was analyzed using two-way RM ANOVA or mixed effects model with Sidak’s multiple comparisons was employed. Relative distribution: call class x context: F(3,80) = 25.29, P < 0.0001, peak frequency: call class: F(1.747, 34.95) = 183.6, P < 0.0001; context: F(1.000, 20.00) = 5.994, P = 0.0237, bandwidth: call class x context interaction: F(3, 20) = 7.600, P = 0.0014, call length: F(3, 120) = 9.635, P < 0.0001. *p < 0.05; **p < 0.01; ***p < 0.001, ****p < 0.0001. LFV = low frequency vocalizations, MFV = middle frequency vocalizations, and USV = ultrasonic vocalizations. A total of 21 same-sex mouse dyads was used.

Temporal organization of separation-induced calls.

(A–D) Exemplary spectrograms of Noisy (A), LFV (B), USV (C), and mixed (D) calls occurring in groups (bouts) with an inter-call interval (ICI) < 140.6 ms (i.e. within one bout). (E–H) Histograms display the distribution of ICI of Noisy (E), LFV (F), MFV (G), and USV (H) calls during both separation (blue) and unification (red). The grey bar in the histograms marks ICIs shorter than 140.6 ms and thus, the number of calls occurring in bouts.

Female same-sex dyads produce more calls with greater bandwidth and longer call length compared to male same-sex dyads.

(A–D) Quantification of the relative distribution of Noisy calls (A), low frequency vocalizations (LFV, B), middle frequency vocalizations (MFV, C) and ultrasonic vocalizations (USV, D) between male (orange) and female (purple) same-sex dyads. Each dot corresponds to one same-sex dyad. (E – H) Quantification of spectrotemporal call parameters: call rate (E), peak frequency (F), bandwidth (G), and call length (H) between male and female same-sex dyads. Black lines represent group mean and error bars represent SD. Each dot represents the average of the respective parameter of one same-sex dyad. Differences between male and female same-sex dyads were analyzed using unpaired t-test. Call rate: t(19) = 1.907, P = 0.0718, bandwidth: t(19) = 2.931, P = 0.0086, call length: t(19) = 2.548, P = 0.0197. *p < 0.05; **p < 0.01. Male same-sex dyads: n = 11, female same-sex dyads: n = 10.

Physically separated same-sex mouse dyads face each other when vocalizing in close proximity.

(A)Schematic depiction of the calculation of both snout-to-snout distance and angle. Dashed black line between mouse heads represents the divider. Purple dots represent snout of left (SL) and right (SR) mouse. Blue and teal dots represent the left and right ear of each mouse, respectively. The middle of the distance between left and right ear is represented as a red cross and named between-ears point for both the left (BL) and the right (BR) mouse. The snout-to-snout distance is calculated as (purple arrow). The snout-to-snout angle (α) is calculated between the vectors (brown arrow) and (green arrow). (B) Schematic depiction of the calculation of both snout-to-tail base distance and angle. Snout-to-tail base distance and angle (α) are calculated analogously to snout-to-snout distance and angle, instead of two snouts, one is replaced by the tail base, i.e. snout-to-tail base distance is calculated of the vector between the snout of the right mouse (SR, purple dot) and the tail base of the left mouse (AL, orange dot), (purple arrow). The angle is calculated between and . (C1–D2) Line plots show the probabilities of Noisy calls (blue), low frequency vocalizations (LFV, green), middle frequency vocalizations (MFV, purple), and ultrasonic vocalizations (USV, orange) for all mouse-to-mouse distances found throughout the experiments for male dyads during both separation (C1) and unification (C2) and female dyads during both separation (D1) and unification (D2). The grey bar represents one mouse body length in pixels (px). (E1–F2) Scatter plots show the relationship between snout-to-snout distance and snout-to-snout angle during separation in males (E1) and females (F1), as well as during unification in males (E2) and females (F2). Inset graphs in E2 and F2 show the relationship between snout-to-tail base distance and snout-to-tail base angle. Call types are color coded: Noisy = blue, LFV = green, MFV = purple, USV = yellow. Male same-sex dyads: n = 11, female same-sex dyads: n = 10.

Spearman correlation between snout-to-snout distance (px) and snout-to-snout angle (°) for separated same-sex mouse dyads.

Genetic background influences call rates, call repertoire and acoustic call properties.

(A)Schematic depiction of the experimental design for the separation context. (B–D) Exemplary spectrograms of call bouts of FVB (B), B6J (C), and CBA (D) mice. (E–H) Quantification of the relative distribution of Noisy (E), LFV (F), MFV (G), and USV (H) calls during separation of FVB (blue), CBA (green) and B6J (purple) samesex mouse dyads. Each dot represents one same-sex dyad. (I–L) Quantification of spectro-temporal call properties: call rate (I), peak frequency (J), bandwidth (K), and call length (L) of FVB, CBA and BL6 same-sex mouse dyads. Each dot represents the average parameter for one same-sex dyad. Black lines represent group mean and error bars represent SD. FVB: n = 21, CBA: n = 5, B6J: n = 5. Data was analyzed using one-way ANOVA with Tukey’s multiple comparisons or Kruskal-Wallis test with Dunn’s multiple comparisons. Noisy distribution: F(2, 28) = 10.06, P = 0.0005, USV distribution: H(2) = 7.523, P = 0.0233, call rate: F(2, 28) = 86.74, P < 0.0001, bandwidth: F(2, 28) = 10.56, P = 0.0004, call length: F(2,28) = 10.23, P = 0.0005. *p < 0.05; **p < 0.01; ***p < 0.001, ****p < 0.0001.

Influence of hole-diameter in the divider on call repertoire and acoustic call features.

(A) Schematic depiction of the experimental design and the different types of dividers used: hole diameter = 0.5 cm (small), hole diameter = 1 cm (large), and no holes (none). (B–D) Pie charts displaying total call repertoire during separation of same-sex pairs utilizing a divider with small holes (B), large holes (C), and no holes (D). Total number of recorded calls is displayed at the bottom of each pie chart. Call classes are color coded: Noisy (blue), low frequency vocalizations (LFV, green), middle frequency vocalizations (MFV, purple), and ultrasonic vocalizations (USV, orange). (E–H) Quantification of the relative distribution of Noisy (E), LFV (F), MFV (G), and USV (H) calls during separation with dividers carrying different diameter bore holes. Small diameter holes = small black dots, large diameter holes = large black dots, no holes = white dots with black border. Each dot represents one same-sex dyad. Noisy, LFV, and MFV data was analyzed using one-way ANOVA with Tukey’s multiple comparisons. USV data was analyzed using Kruskal-Wallis test with Dunn’s multiple comparisons. Noisy ratio: F(2, 34) = 9.758, P = 0.0004, LFV ratio: F(2, 34) = 25.39, P < 0.0001, MFV ratio: F(2, 34) = 12.53, P < 0.0001, USV ratio: H(2) = 14.92, P = 0.0006. (I–L) Quantification of spectro-temporal properties: call rate (I), peak frequency (J), bandwidth (K), and call length (L) of same-sex mouse dyads during separation with either small-diameter hole carrying divider (small), large diameter hole carrying divider (large), or no hole carrying divider (none). Each dot represents the average parameter for one same-sex dyad. Small: n = 21, large: n = 8, none: n = 8. Call rate was analyzed using Kruskal-Wallis test, while peak frequency, bandwidth, and call length was analyzed using one-way ANOVA with Tukey’s multiple comparisons. Call rate: H(2) = 6.785, P = 0.0336, peak frequency: F(2, 34) = 25.86, P < 0.0001, bandwidth: F(2, 34) = 25.92, P < 0.0001, call length: F(2, 34) = 54.62, P < 0.0001. *p < 0.05; **p < 0.01; ***p < 0.001, ****p < 0.0001.

Treatment with the anxiolytic buspirone had no effect on both separation-induced call repertoire and spectro-temporal call properties.

(A) Schematic depiction of the separation context of experiment design. (B – D) Pie charts display total call repertoire during physical separation of male same-sex mouse dyads aftertreatment with buspirone (B) or vehicle (physiological saline; C) and untreated male same-sex dyads (D). The total number of recorded calls is displayed at the bottom of the pie charts. Call classes are color coded: Noisy = blue, low frequency vocalizations (LFV) = green, middle frequency vocalizations (MFV) = purple, ultrasonic vocalizations (USV) = orange. (E–H) Quantification of the relative distribution of Noisy (E), LFV (F), MFV (G), and USV (H) calls during the physical separation of same-sex mouse dyads after treatment with buspirone (n = 5) or vehicle (n = 5), or of untreated dyads (n = 11). Each dot represents the proportion of the respective call class of one same-sex dyad. Noisy distribution: t(14) = 2.064, P = 0.0580, MFV distribution: t(14) = 2.368, P = 0.0328, USV distribution: U = 9, n1 = 5, n2 = 11, P = 0.0380. (I–L) Quantification of spectro-temporal call properties: call rate (I), peak frequency (J), bandwidth (K), and call length (L). Each dot represents the average of one samesex dyad of the respective parameter. Treatments are color coded, buspirone = black dots, vehicle = white dots, and untreated = grey dots. Paired data points are connected by a dashed line. The differences between buspirone and vehicle treatment were analyzed using either paired t-test or Wilcoxon signed rank test. Comparisons between buspirone-treated and not treated mice, as well as between vehicle-treated and not treated mice were made using either unpaired t-test or Mann-Whitney-U test. Call rate: t(14) = 2.637, P = 0.0195. *p < 0.05

Call repertoires are adapted when a mixed-sex dyad is separated by a divider.

Pie chart shows the call repertoire of separated mice of opposite sex. Call classes are color coded: Noisy (blue), low frequency vocalizations (LFV, green), middle frequency vocalizations (MFV, purple), and ultrasonic vocalizations (USV, orange). The total number of calls recorded under the respective condition is displayed at the bottom of each pie chart. Opposite-sex dyads: n = 5.

Repeated exposure to the experimental context affects the call repertoire used by separated female same-sex dyads

(A) Schematic depiction of the separation context of experiment design. (B–D) Pie charts display the total call repertoire of separated female same-sex dyads treated with either vehicle or buspirone recorded during their first exposure to the experimental context (B), during their second exposure to the experimental context (C), and of untreated female same-sex dyads’ first and only exposure (D). The total number of recorded calls is displayed at the bottom of each pie chart. Noisy calls (blue), low frequency vocalizations (LFV, green), middle frequency vocalizations (MFV, purple), and ultrasonic vocalizations (USV, orange). (E–H) Quantification of call class distribution during first and second exposure to the experimental context for Noisy (E), LFV (F), MFV (G), and USV (H) calls. Female dyads that were treated with vehicle before the first exposure and buspirone before the second exposure are represented by white dots. Female dyads that were treated first with buspirone and with vehicle before the second exposure are represented by black dots. Each dot represents the proportion of the respective call class of one female same-sex dyad. (I – L) Quantification of spectro-temporal call parameters: call rate (I), peak frequency (J), bandwidth (K), and call length (L). Each dot represents the average of the respective spectro-temporal call parameter of one female same-sex dyad. Vehicle  buspirone dyads: n = 2, buspirone → vehicle dyads: n = 2.