Acoustic Characteristics of Spawning Biological Sounds of Brown Croaker (Miichthys miiuy)

Abstract

Marine organisms make sounds for various reasons, including spawning and avoidance, which are species-specific. Traditionally, Korean fishermen in the northwest Pacific Ocean have listened for spawning sounds to locate spawning grounds of Brown croaker (Miichthys miiuy), one of the important commercial fish species. We measured the spawning sounds recorded in October, the croakers’ spawning season (August to October). The mean signal duration with standard deviation was 0.184 ± 0.027 s, and the mean pulse interval was 0.022 ± 0.001 s. The zero-to-peak mean sound pressure level was 165.2 ± 0.7 dB. The peak frequency was 459.2 ± 93.8 Hz, with maximum and minimum frequencies observed at 863.0 ± 225.9 Hz and 231.2 ± 67.9 Hz. The spawning sounds occurred around sunset (16:00–21:00, local time) and only when the water temperature was above 22 °C. These findings help interpret the environmental ecology and manage the fishery resources of the Brown croaker spawning grounds.

1. Introduction

Many fish species in aquatic environments produce biological sounds associated with feeding, spawning, social interactions, and other behaviors [1,2,3,4]. In certain situations, many individuals in a specific area produce these sounds simultaneously, leading to notable, temporary increases in ambient noise levels that alter the soundscape [5,6]. The sound characteristics have evolved based on their specific functions and the soundscape they create [7,8].

At least 178 fish families, accounting for two-thirds of fish species, communicate using sound. They use sound primarily to court mates, protect food sources and territories, and announce their presence to others [9,10]. Fish produce sound through various mechanisms, particularly the stridulation of rubbing bones or vibration of the swimbladder [11,12]. In particular, during the spawning season, male croakers produce the sounds to announce their locations [13,14].

The Brown croaker (Miichthys miiuy, Basilewsky 1855) is a fish species of considerable commercial importance in Korea with a total body length (TL) of 60–90 cm, a grayish-blue dorsal color, and a light white ventral side (Figure 1) [15,16]. It is found in the northwest Pacific Ocean and the Yellow/South and East/South China Seas [17,18]. The Brown croaker migrates southward and spends the winter in the Yellow and East China Seas. After the wintering period, it begins its northward migration in spring. The Brown croaker inhabits coastal oceans at a depth of 15–100 m, with mud to sandy mud bottoms. It feeds on benthic crustaceans and fishes. The spawning season is locally from late August to October for fishermen [18,19,20]. No study has been conducted on the acoustic characteristics of the Brown croaker.

Our study focuses on the species-specific acoustic characteristics of the sound production, particularly the Brown croaker. We measured and analyzed the sounds of the Brown croaker, an important fish species in the northwest Pacific Oceans. Correlations among the daily variation, water temperature, and fish sounds were analyzed. This study investigated the practical applications of these characteristics for monitoring aquatic fish species’ presence, abundance, and activity patterns.

2. Materials and Methods

2.1. Measurement of Biological Fish Sounds

The Brown croakers were recorded in Korea in an indoor seawater tank in Tongyeong (34°49.604′ N, 128°20.037′ E). The seawater tank was made of concrete and octagonal-shaped with a side length of 2 m and depth of 1.5 m. The 6 adult Brown croakers (4 female and 2 male specimens) freely swum and produced the sounds for the experiments. A self-recording hydrophone (SM3M, Wildlife, Maynard, MA, USA [21]) was installed in the water tank (distance to the wall: 0.1 m; distance to the bottom: 0.7 m) to record the fish sounds (Figure 1). The acoustic data were recorded continuously over 18 days, from 11 to 29 October 2022. The hydrophone had a sampling frequency of 48 kHz, gain of 0 dB, and sensitivity of −164.6 dB V/µPa.

The indoor tank housed in the room had windows on all sides to allow natural light into the tank. We referenced the sunset and sunrise times from the Korea Meteorological Administration (KMA, https://www.weather.go.kr/w/index.do, (accessed on 1 January 2024) [22]) to analyze the daily variation in the sounds. Water temperature data were collected using a commercial temperature sensor installed in the tank to examine correlations between the water temperature and the fish sounds.

2.2. Signal Processing and Acoustic Characteristic Analysis Methods

We analyzed a dataset comprising 18 days (25,920 min) of the fish sounds using MATLAB R2023a (Mathworks, Natick, MA, USA). The periodic events in the measured data were identified using power spectral density (PSD) calculations via the Welch method [23] with a 50% overlap for 1 s each and confirmed with a spectrogram.

The primary objective was quantifying and characterizing fish sounds, focusing on signals with a high signal-to-noise ratio (SNR) that did not clip the recording system. Time series analyses were employed to estimate several key parameters of the fish sounds. These parameters included signal duration, pulses per call, pulse duration, and pulse period. To further investigate the characteristics of the pulses, we isolated them from the fish sounds and calculated the sound pressure level (SPL), peak frequency, and 3 and 10 dB bandwidths across the entire dataset (Figure 2).

Signal length was determined based on the percentage of energy signal duration, following the ISO 18405 guidelines [24]. Specifically, the energy percentage was defined as 95%, corresponding to the 95% cumulative energy points. Pulse interval is the time between the peaks of each pulse. SPL was measured from zero to the signal’s peak level.

PSD was calculated for each independent signal length for frequency analysis. We obtained the peak frequency and the 3 and 10 dB bandwidths from the PSD spectrograms. The maximum and minimum frequencies were identified as the upper and lower limits of the ±10 dB band, respectively.

3. Results

Figure 2b shows an example of fish time-series data of the fish chorus, which contains eight sounds of Brown croakers. The signal consists of pulses repeated several times at almost regular intervals (Figure 2c) and has high energy at frequencies below 2 kHz (Figure 2d).

Among the acoustic data, 411 representative sounds of the Brown croaker that did not clip acoustic signals were selected and analyzed. The mean signal duration with standard deviation (S.D) was 0.184 ± 0.027 s, the mean pulse interval was 0.022 ± 0.001 s, and the pulse per call repeated between 2 and 11 times throughout the call duration. The zero-to-peak mean SPL was 165.2 ± 0.7 dB. The peak frequency was 459.2 ± 93.8 Hz, and the croaker’s frequencies ranged from 129 to ~1589 Hz, respectively. The spectral analysis revealed that the 3 dB and 10 dB bandwidths were 79.1 ± 47.4 Hz and 632.8 ± 237.6 Hz, respectively. Figure 3 and

Table 1. Characterization of the acoustic parameters from sounds of Brown croaker.

Acoustic Parameters	n	Mean ± S.D	Minimum	Maximum
Pulse duration (ms)	411	183.8 ± 27.4	23.2	240.4
Pulse interval (ms)		21.9 ± 1.0	16.6	26.1
Pulse per call		8.9 ± 1.36	2	12
Sound pressure level (0-peak) (dB re 1 µPa)		165.2 ± 0.7	163.4	167.4
Peak frequency (Hz)		459.2 ± 93.8	265	768
Max frequency (Hz)		863.0 ± 225.9	447	1589
Min frequency (Hz)		231.2 ± 67.9	129	475
3 dB bandwidth (Hz)		79.1 ± 47.4	30	381
10 dB bandwidth (Hz)		632.8 ± 237.6	108	1370

present the acoustic characteristics of a Brown croaker’s calls.

All observed signals were accumulated over 24 h to confirm the daily variation pattern and are displayed as a spectrogram up to a frequency of 5 kHz in Figure 4a. It was confirmed that the Brown croaker’s sounds were generated from approximately 16:30 to 20:30 (KST). The sounds of the Brown croaker predominantly occurred near sunset, similar to previous studies [25,26]. The spectrogram was acquired throughout the observation period. The sounds were confirmed to have high energy at frequencies below 2 kHz. The overall background noise increased, which was confirmed to be the sound of the pump for circulating seawater in the tank (Figure 4b). The mean SPL readings were taken every 10 min to measure its correlation with the water temperature. From October 25 onward, the croaker’s sounds did not occur, coinciding with the water temperature dropping below 22 °C (Figure 4c).

4. Discussion

The acoustic characteristics of the Brown croaker provide valuable insights into their behavioral and reproductive ecology. The pronounced increase in the sounds during spawning suggests that acoustic communication is crucial in coordinating reproductive activities and social interactions among individuals.

No acoustic characteristics are known for the Brown croaker, although severddal studies have been conducted on the acoustic properties of other and similar fish species [27,28,29]. Erisman and Rowell [27] recorded fish calls and choruses at a point where millions of Gulf corvina (Cynoscion othonopterus) of croaker species arrive during the spawning season every spring in the bay delta near the Colorado River. From the analysis of acoustic characteristics, the mean pulse duration was 0.389 s, the pulse interval was 0.033 s, and the pulse count per call was 11.1. On the other hand, Mok et al. (2020) [28] recorded the fish calls of Asian fresh-water species of Boeseman croaker in the concrete tanks and rivers. From the analysis of acoustic characteristics in the measuring tank results, the pulse count per call was 15.1, the mean pulse duration was 0.007, the mean signal duration was 0.139 s, and the main dominant frequency was 1263 Hz. In the evening advertisement calls, a sequence of multicycle tonal pulses was observed; however, the fundamental frequency and initial harmonics were either absent or weakened, while the peak frequencies were notably high and exhibited variation. Also, Javier et al. (2010) measured the advertisement calls for two male whitemouth croakers (28.0–30.5 cm) in the laboratory [29]. As a result, the pulse duration was 0.018~0.02 s, the interpulse intervals were 0.496~718 s, and the dominant frequency was 280~316 Hz.

These acoustic characteristics differed (longer call duration, more pulses per call, longer pulse interval, and high dominant frequency) from the Brown croaker measured and analyzed in this study. In particular, the peak (main) frequency is expected to appear in the low-frequency band because the body and swimbladder lengths of the Gulf corvina, Boeseman, and whitemouth croakers are different.

The correlation between the sounds and daily variations underscores the importance of understanding acoustic communication in the context of reproductive strategies and population dynamics in the Brown croaker. We recorded the sounds throughout the day, but fish sounds occurred only around sunset. Like many sciaenids, fishes produce the fish calls in the late afternoon into evening time (sunset) [25,26,28]. Studies have been conducted comparing fish sounds with daily variation patterns of other fish species [25,26,28]. Using passive acoustic monitoring (PAM) in the Condor seamount, Carrico et al. (2020) reported that the fish sounds exhibited temporal dynamics in diversity patterns [25]. Analyzing annual, seasonal, and daily patterns revealed that a higher abundance of sound sequences was observed at dusk and night. Boyle et al. (2022) analyzed the spatial and temporal variation in fish sounds that are predicted to vary in association with species composition and abundance, as well as daily and seasonal influences [26]. In our study, the fish sounds occurred mainly around sunset in an environment with windows on all sides of the indoor tank. Since this pattern is expected to differ in ocean areas, a comparative analysis of the daily variation patterns in the field is needed.

The sound production of the Brown croaker ceased as the water temperature decreased (Figure 4c). In general, the temporal pattern of pulses within the sound is almost always correlated with temperature if it is directly based on the contraction of the sonic muscles [30,31]. Some fish species, such as Silver perch, Oyster toadfish, Black drum, Spotted seatrout, and Red drum, have specific water temperature ranges in which they produce sounds [32]. The courtship calls reveal that each species has a specific water temperature preference for breeding, and analysis of the soundscape reveals seasonal and spatial patterns of spawning in fish species.

In the future, research on the acoustic characteristics of the sounds will be applied to soundscape management to protect fish species, croaker habitats, and maintenance resources. In addition, research on the acoustic characteristics of the sounds can be applied to study the ecological characteristics, such as fishing and spawning grounds of the Korean oceans of the Brown croaker, through observation of passive acoustic systems such as hydrophones and vector sensors.