Female song in birds is more widespread than previously thought, and studies across different species are critical for better understanding the entire evolutionary process of bird song. In this study, we recorded the songs of males and females in a duetting, montane bird species, the Elliot’s laughingthrush Trochalopteron elliotii, across consecutive breeding seasons. We specifically focused on identifying the number and structure of different song types by males and females, and compared these acoustic structures between the sexes. Our findings revealed that both males and females sang sex-specific solos. More specifically, females sang a single type of solo that varied significantly in the number of notes, whereas males produced three different solos composed of 2, 3, or 4 notes, respectively. Female solos exhibited significantly more notes and longer song duration compared to males. Male solos typically had a significantly higher maximum frequency for the entire song. No significant differences were observed in the duration of the first note, song rate, and other frequency characteristics between male and female solos. Furthermore, paired males and females coordinated their sex-specific solos to form duets when challenged by conspecific territory intruders, both within and outside the breeding season. Sex-specific solos suggested that male and female songs play different roles and may be subject to different selective pressures. Further research is necessary for elucidating the functions of male song, female song, and duets in this montane bird species.

Study site and species

The fieldwork was conducted in the Lianhuashan Nature Reserve (34°58′16″ N, 103°45′30″ E; 2040 m a.s.l.), located on the eastern edge of the Qinghai-Tibet Plateau in central China. The study area is mainly composed of cultivated farmlands, open woodlands, and orchards, as well as abandoned fields dominated by scrub vegetation. Since 2013, we have been studying the ecology of the resident Elliot’s laughingthrush population at this site and each year, catch and colour-band males and females with a unique combination of colour rings for sex and individual identification. The sex of all marked individuals was determined by the swelling of the male gonads and differences in morphological size (Liu and Sun 2016).

Field observation and vocalization recording

From March 5 to May 20, 2021, and April 5 to June 13, 2022, we observed male and female singing behaviour every three days during suitable weather conditions. All observations were conducted between 08:00 and 11:00 am. Within each territory, we conducted one-hour long observations, during which we recorded the date, sex of the singer, time, singing contexts (agonistic interactions, spontaneous singing, chorus singing), presence of intruders, and whether the focal pair divorced. For each singing individual, we collected solos and duets using a TASCAMHD-P2 sound recorder (44.1 kHz, 16 bits) and a Sennheiser MKH416 P48 shotgun microphone at a distance of 5-10 meters from the singing birds. We also recorded solos from unpaired males, which were determined through direct observations of male-female interactions, and from observations made at nests being attended by these same males.

Acoustic structure

Avisoft-SASLab Pro was used to analyse the song structure of all recordings. The sampling rate was 22.05 kHz, with a sample size of 16 bits, the FFT length was 512 points, the frequency resolution was 43 Hz, the temporal resolution was 2.9 ms, and the window used was FlatTop, with an 87.5% overlap. The simplest elements of the song on the sonogram are referred to as “notes”. We defined a discrete song as being made up of two or more consecutive notes, typically separated by a brief pause (Brenowitz et al. 1997). Solo songs refer to the distinctive individual songs produced by a single bird (either male or female). All definitions of songs and notes are illustrated in Figure 1A.

For each individual, the following song characteristics were recorded: the total number of notes within a song, duration of song, first note duration of a song, the speed of a song (the number of notes within a song per second), maximum frequency, minimum frequency, and frequency range of the song, the first and last note. To determine these parameters, we analysed two solos from each female and male. In order to include a diverse range of variation in vocal characteristics within individuals, we only selected recordings that were both high-quality and representative of the individual’s vocal repertoire.

Spectrograms creation

Spectrograms were extracted using R package 'Seewave' with an FFT-window length of 512 and an overlap of 87.5% (Sueur et al. 2008). The 'tuneR' package was used to load sound files in WAV format as an object (Ligges et al. 2013). In a spectrogram, time is displayed on the x-axis, frequency on the y-axis, and amplitude is represented by colour intensity.

Statistical analysis

Wilcoxon rank sum tests were used to determine if each of the vocal characteristics (the number of notes, speed, duration, first note duration) and frequency of the song differed between males and females. Bonferroni correction was used to adjust the p-values in the analysis. All statistical analyses were conducted using R version 4.1.3 (R Core Team, 2022). Values are given as Mean ± SE, with Bonferroni corrected P value < 0.004.