Acoustic Target Strength According to Different Growth Stages of Japanese Anchovy (Engraulis japonicus): A Comparison of Juvenile and Adult Fish

Abstract

Japanese anchovy (Engraulis japonicus) is an economically important species found in the coastal waters of the northwestern Pacific Ocean. To conduct effective acoustic surveys on the spawning or fishing grounds of this species, the target strength (TS, dB re 1 m²) during its growth stages must be understood. However, knowledge regarding the TS of juvenile Japanese anchovies is limited. To address this knowledge gap, we measured the TS and proposed equation models for juvenile Japanese anchovies while describing the TS differences between young and adult anchovies. We evaluated 21 live anchovies using 38, 120, and 200 kHz transducers. TS models were fitted for the juvenile stage of anchovies, considering different frequencies. The mean TS values of young and adult anchovies were similar to those reported in previous TS studies, whereas the TS values of juvenile anchovies were significantly lower. Thus, applying the same TS equation based on the acoustic characteristics of anchovies at different growth stages, especially for juvenile anchovies, may not be appropriate. These findings underscore the importance of adopting a suitable TS equation specific to the growth stages of Japanese anchovies for accurate data analysis in acoustic surveys conducted on spawning and fishing grounds.

1. Introduction

Anchovies are globally important commercial fish species that are widely distributed in shallow coastal oceans [1]. They can be of various types, including European, Californian, Peruvian, and Japanese. Among these, the Japanese anchovy (Engraulis japonicus) is distributed throughout the northwestern Pacific Ocean [2,3]. In 2020, the total global annual anchovy catch was approximately 6,731,661 MT, with Korea accounting for approximately 216,748 MT of the annual Japanese anchovy catch [4]. These small anchovies form large schools and primarily feed on zooplankton and small fry, making them an essential part of the diet of larger anchovies [5]. Owing to this ecological role, anchovies are important as coastal fishery resources, playing a vital role in marine ecosystems [6,7,8].

Hydroacoustic methods have been used to estimate fish distribution and abundance in marine environments [9]. Currently, the accuracy of acoustic target strength (TS, dB re 1 m²) information is a key parameter that determines the accuracy of biomass estimates [9,10]. Previous studies have reported that TS information can cause up to 50% errors in biomass calculations [10] and that TS is generally expressed as a function of the total length (L) and wet weight (W) of the fish [11]. The TS characteristics of fish can be obtained by identifying the spatiotemporal patterns of marine living organisms using hydroacoustic techniques [11,12,13]. Thus, detailed information regarding the TS–L and L–W relationships is required to improve fish biomass estimates [14,15].

Until now, research on the TS of Japanese anchovy has been limited. Previous studies have explored TS through in situ measurements in a natural environment [15,16], ex situ measurements using the tethering method in water tanks [14,17], and numerical acoustic scattering models [18]. Zhao et al. (2008) conducted in situ TS measurements of anchovies at various depths in the Yellow Sea [15]. Sawada et al. (2009) investigated in situ TS according to the tilt angle using an acoustic–optical system developed for anchovies [16]. In our previous study, we measured ex situ TS using anchovies from the South Sea of Korea and estimated the TS equation based on the L and W of the anchovies [14]. However, most previous studies have focused on measuring and calculating TS for medium-to-large anchovies. Therefore, in this study, we expanded upon these principles and investigated TS in juvenile anchovies. By focusing on this specific growth stage, we can fill the gap in our understanding of TS in different growth stages of anchovies, leading to a more comprehensive understanding of their acoustic characteristics.

Anchovies are classified into three growth stages based on total length: juvenile (L = 3–6 cm), young (L = 6–8 cm), and adult I and II (L > 8 cm) [19]. As fish size varies significantly across these growth stages, it is expected that the TS equation will also differ between juvenile and adult fish. Therefore, when considering the TS of juvenile anchovies, it is crucial to use an appropriate TS equation tailored for this growth stage. Using the same TS equation designed for adult anchovies for juvenile anchovies may not provide accurate results, given the differences in their sizes and acoustic characteristics.

This study aimed to measure the TS of live Japanese anchovies across the juvenile-to-adult growth stages using a scientific echo-sounder with 38, 120, and 200 kHz frequencies. In accordance with the various growth stages of anchovies, we conducted least-squares linear regression analyses to fit models for TS versus L and compared these models with those obtained from a previous study [14]. The key outcome of this research was the development of a new TS equation specifically tailored for juvenile anchovies. By considering the various growth stages of anchovies and suggesting new TS equations for juvenile anchovies, this study aimed to enhance the accuracy of estimating the biomass of small anchovies in spawning grounds. The TS equations developed here can be applied to different growth stages of anchovies to improve the accuracy of estimating small-anchovy biomass in spawning grounds.

2. Materials and Methods

2.1. Biological Sampling and Experimental Apparatus

The TS of the Japanese anchovies was measured ex situ using the tethered method. This measurement was performed at a coastal fish farm situated at the Tongyeong Megacosm Test Station of the Korea Institute of Ocean Science and Technology (KIOST) in the South Sea of Korea. The TS measurement experiment was conducted within a net cage measuring 4 m in length, 4 m in width, and 4 m in height.

The anchovies used in the TS experiment were captured during the nighttime using a wild thin net. A total of 21 anchovies were captured, with L ranging from 3.13 to 9.85 cm and W ranging from 0.099 to 5.404 g. The relationship between L and W was determined as W = 0.003175$·L^{3.255}$ (Figure 1). Specifically, the L value of the anchovies used in the TS experiment was approximately 3.13–5.56 cm (n = 15) for the juvenile stage, 6.97–7.32 cm (n = 4) for the young stage, and 8.73–9.85 cm (n = 2) for the adult stages of Japanese anchovy.

For the TS measurement, each live anchovy was first anesthetized using FA100 (4-allyl-2-methoxyphenol) and then tethered to a small monofilament line that was attached to its mouth. The monofilament line, along with the anchovy, was secured to a vertical line set at a depth of approximately 3 m. The end of the vertical line was linked to a weight, and the distance between the tethered anchovy and the weight was maintained at approximately 1.0 m. This distance ensured the separation of the anchovy echo signal from the weight and lower net (Figure 2). To ensure accurate measurements, the TS was simultaneously recorded from three frequency transducers when the anchovy remained stable and appropriately positioned within the beam axis. Each anchovy’s L and W values were determined after the TS measurement. Subsequently, the anchovies were promptly shock-frozen to examine their swim bladders and verify their size and conditions.

2.2. TS Measurement

The TS measurements were carried out using a scientific echo-sounder (DT-X; BioSonics, Inc., Seattle, WA, USA). The three split-beam transducers with frequencies of 38, 120, and 200 kHz were installed 0.5 m below the sea surface within a net cage, facing downward in the vertical direction. The experiments were conducted under specific environmental conditions, including a seawater temperature of 16.17 °C, salinity of 33.25 psu, and a sound speed of 1508.38 m/s [11].

The echo-sounder utilized specific source levels for different frequencies, namely, 214.2 dB for 38 kHz, 220.3 dB for 120 kHz, and 221.8 dB for 200 kHz. The corresponding beam widths were 9.2°, 7.8°, and 6.7°, whereas the absorption coefficients were 0.00842, 0.04331, and 0.06527 dB/m, respectively. The pulse length and ping rate were set at 0.2 ms and 2 pings/s, respectively, for all frequencies [20]. Before TS measurements, the transducers underwent calibration using tungsten carbide calibration spheres with diameters of 38 mm, 33 mm, and 36 mm for the respective frequencies.

Table 1. Echo-sounder parameters used to measure the target strength (TS) of Japanese anchovies.

Parameter	38 kHz	120 kHz	200 kHz
Beam type	Split-beam
Source level (dB re 1$\mathsf{\mu }$Pa @ 1 m)	214.2	220.3	221.8
Pulse length (ms)	0.2
Ping rate (pings/s)	2
Beam width (°)	9.2	7.8	6.7
Absorption coefficient (dB/m)	0.00842	0.04331	0.06527
Sound velocity (m/s)	1508.38
Water temperature (°C)	16.17
Salinity (psu)	33.25

presents a complete list of the scientific echo-sounder settings for comprehensive reference.

To observe the behavior, tilt angle, and conditions of the anchovies during the TS measurements, an underwater video camera (Water-7000DX; Tsukamoto Co., Ltd., Suzuka, Japan) was installed in the horizontal direction. The video camera recorded the anchovies’ activities while TS measurements were being conducted. In the data analysis process, the video recordings and acoustic data were synchronized to ensure an accurate correlation between the two sets of information. The acquisition of TS and underwater video data extended for approximately 20 to 60 min, providing valuable insights into the characteristics and behaviors of the anchovies during the experimental period.

2.3. Data Processing and Analysis

The measured TS values were analyzed following the “Target collection and TS distribution” procedure, using Visual Analyzer (v4.1, BioSonics, Inc.), the “Single target detection—split beam” procedure, Echoview (v9.0, Echoview Software Pty Ltd., Hobart, Australia), and Matlab (R2016a, MathWorks, Inc., Natick, MA, USA). To prevent unwanted echo signals, reliable TS values were selected for analysis based on the target depth (2.5–3.5 m) and off-axis position ($\pm$4.6, $\pm$3.9, and $\pm$3.35°). Different threshold levels were set to reduce TS errors related to the anchovy tilt angle: −70 dB for juveniles and −60 dB for young and adult anchovies. Furthermore, the TS values were cross-referenced with time-synchronized underwater camera images to validate the results. In total, the TS values were analyzed for 16 to 21 live anchovies at 38 kHz (16 anchovies), 120 kHz (20 anchovies), and 200 kHz (21 anchovies).

The mean TS (dB re 1 m²) was calculated using the backscattering cross section (${\sigma }_{bs}={10}^{TS/10}$) and logarithmic transformation in Equations (1) and (2). The definition of the mean TS is as follows [10,15,21]:

$${\overline{\sigma }}_{bs}=\frac{1}{n}\sum _i^n{\sigma }_{bsn},$$

(1)

$$\mathrm{m}\mathrm{e}\mathrm{a}\mathrm{n} TS=10·{\log }_{10}\left({\overline{\sigma }}_{bs}\right).$$

(2)

The TS equations were fitted to establish the relationship between the mean TS and logarithm of L through least-squares linear regressions using Equations (1) and (2). Subsequently, the mean TS data were subjected to 2 linear fitting models of the TS equation, utilizing least-square regressions with L. Equations (3) and (4) of the TS equations were defined as follows [10,11,22,23]:

$$TS=a·{\log }_{10}(L,\mathrm{c}\mathrm{m})+b,$$

(3)

$$TS=20·{\log }_{10}(L,\mathrm{c}\mathrm{m})+b_{20},$$

(4)

where a represents the slope (20 is the standard slope) and b and $b_{20}$ are intercepts of the line. The parameters were estimated using the linear mean TS and L.

3. Results

Figure 3 shows a sample echogram obtained from TS measurements of juvenile (L = 4.34 cm) and young anchovies (L = 7.32 cm) at a frequency of 200 kHz. The echo signals from the anchovies, weights, and lower net cages were clearly separated in the echogram. The relatively low TS value observed at the beginning of the measurement can be attributed to the unstable condition and tilt angle of the juvenile and young anchovies. However, as the anchovies stabilized at the appropriate tilt angle, the mean TS was calculated using the TS data collected over two time intervals: 11:28−11:32 (Figure 3a) and 14:46–14:50 (Figure 3b).

Figure 4 shows the TS histograms corresponding to each growth stage at 200 kHz. All TS values exhibited a Gaussian distribution. However, the shapes and distributions of the TS histograms varied based on the anchovy L. The numbers of valid TS measurements were 191 and 360 for juvenile anchovies (L = 3.47 and 5.4 cm) and 351 for the young anchovies (L = 7.03 cm). The corresponding TS values ranged from −76.3 to −60.8 dB for juveniles (Figure 4a), from −76.7 to −46.6 dB for young anchovies (Figure 4b), and from −76.4 to −45.2 dB for the young (Figure 4c). The mean TS values were found to be −66.4 dB for juveniles, −52.8 dB for young anchovies, and −51.4 dB for the young. In general, the TS values increased as the L of the anchovies increased.

Figure 5 illustrates the relationship between the L and mean TS values at different frequencies. The mean TS values ranged from −66.87 to −48.60 dB for 38 kHz (Figure 5a), −68.65 to −50.22 dB for 120 kHz (Figure 5b), and −68.09 to −49.74 dB for 200 kHz (Figure 5c). The TS values depended on the frequency, with varying ranges observed across the growth stages. At a sampling frequency of 200 kHz, the TS values ranged from −68.09 to −52.04 dB for juvenile anchovies, −54.24 to −51.36 dB for young anchovies, and −51.04 to −49.74 dB for adult anchovies. This demonstrates that the TS values differ based on the growth stages of the anchovies at 200 kHz.

Based on the ex situ TS measurements, the mean TS values of young and adult anchovies showed a good fit with our previously reported TS equation [14]. However, the mean TS values of juvenile anchovies did not align well with the standard equation (Equation (4)). Specifically, the mean TS values of juvenile anchovies were consistently lower by 2–13 dB at all frequencies (Figure 5a–c) compared to the values predicted by the previous TS equation [14]. Therefore, a new TS equation was proposed for different growth stages, specifically for juvenile anchovies. The TS equation for juvenile anchovies is represented as Equations (5)–(7), indicated by the solid blue lines in Figure 5. The specific form of the new TS equation is as follows:

$$\begin{gathered} {TS}_{38 \mathrm{k}\mathrm{H}\mathrm{z}\left(\mathrm{J}\mathrm{u}\mathrm{v}\mathrm{e}\mathrm{n}\mathrm{i}\mathrm{l}\mathrm{e}\right)}=56.1·{\log }_{10}\left(L\right)-94.7 \\ (95\% CI: 37.3 \mathrm{to} 74.9 \mathrm{and}-106.5 \mathrm{to}-82.9; R^2 = 0.86), \end{gathered}$$

(5)

$$\begin{gathered} {TS}_{120 \mathrm{k}\mathrm{H}\mathrm{z}\left(\mathrm{J}\mathrm{u}\mathrm{v}\mathrm{e}\mathrm{n}\mathrm{i}\mathrm{l}\mathrm{e}\right)}=59.7·{\log }_{10}\left(L\right)-98.5 \\ (95\% CI: 46.9 \mathrm{to} 72.5 \mathrm{and}-106.6 \mathrm{to}-90.4; R^2 = 0.90), \end{gathered}$$

(6)

$$\begin{gathered} {TS}_{200 \mathrm{k}\mathrm{H}\mathrm{z}\left(\mathrm{J}\mathrm{u}\mathrm{v}\mathrm{e}\mathrm{n}\mathrm{i}\mathrm{l}\mathrm{e}\right)}=64.2·{\log }_{10}\left(L\right)-100.0 \\ (95\% CI: 49.6 \mathrm{to} 78.7 \mathrm{and}-109.3 \mathrm{to}-90.7; R^2 = 0.87), \end{gathered}$$

(7)

where CI indicates the confidence interval for a and b, and $R^2$ is the coefficient of determination. The overall growth stages, L, and mean TS values (dB re 1 m²) at all frequencies are presented in

Table 2. Summary of the growth stages, total length (L), wet weight (W), and mean target strength (TS) values at 38, 120, and 200 kHz for 21 live anchovies.

Fish Number	Growth Stage	Total Length (cm)	Wet Weight (g)	Mean TS (dB re 1 m2)
				38 kHz	120 kHz	200 kHz
1	Juvenile	3.13	0.990	−66.87	−67.88	−65.53
2	Juvenile	3.19	0.114	−65.23	−68.65	−66.21
3	Juvenile	3.21	0.850	-	−66.33	−68.09
4	Juvenile	3.47	0.129	−66.27	−68.18	−67.43
5	Juvenile	4.18	0.305	−60.29	−61.17	−63.00
6	Juvenile	4.25	0.290	-	−63.80	−64.15
7	Juvenile	4.34	0.333	-	−59.09	−57.92
8	Juvenile	4.45	0.345	−55.82	−60.49	−56.49
9	Juvenile	4.47	0.379	−55.63	−59.17	−57.15
10	Juvenile	4.70	0.386	−60.20	−61.50	−59.10
11	Juvenile	4.84	0.457	−58.10	−55.95	−55.59
12	Juvenile	5.08	0.655	−54.44	−57.60	−53.00
13	Juvenile	5.27	0.643	−53.47	−53.89	−52.04
14	Juvenile	5.40	0.818	-	-	−52.84
15	Juvenile	5.56	0.946	-	−52.62	−52.59
16	Young	6.97	1.866	−52.58	−51.29	−51.39
17	Young	7.03	1.881	−52.47	−50.68	−51.36
18	Young	7.11	1.944	−52.52	−50.68	−52.92
19	Young	7.32	2.028	−52.48	−50.33	−54.24
20	Adult	8.73	3.664	−49.07	−50.86	−51.04
21	Adult	9.85	5.404	−48.60	−50.22	−49.74

. A change in TS of about 2–4 dB per unit of frequency appeared, which was judged to have an effect when applying the dB-difference method.

4. Discussion

For the accurate estimation of fish biomass, the TS–L and L–W relationships must be considered along with precise acoustic surveys [24]. Typically, the TS equation is determined based on the growth stage or L and is expressed with a standard value (Equation (4)), where a = 20. However, we encountered discrepancies when calculating the TS of juvenile anchovies using the standard TS equation, as the standard value (slope) did not align with the observed data. Our attempts to calculate a total TS equation using both the present and previous TS values resulted in lower $R^2$ values.

Previous studies have not specifically focused on accurately measuring the TS of juvenile anchovies. Zhao et al. (2008), Kang et al. (2009), Sawada et al. (2009), and Madirolas et al. (2017) mainly measured the TS of adult anchovies [14,15,16,25]. In contrast, Miyashita (2003) and Ito et al. (2011) only performed theoretical TS modeling, excluding the swim bladder size [18,23]. Because of the challenges in accurately measuring TS in small-sized anchovies and the need for a sufficient number of samples, the TS of juvenile anchovies has remained largely unknown. However, in our previous study [14], we conducted TS measurements for both juvenile and adult anchovies, covering a range of L values from 4.8 cm to 12.2 cm. When comparing the TS values obtained in the previous study with the present mean TS values (black circles in Figure 5), we observed that the latter were consistently lower by less than 6.0 cm in L value. Furthermore, upon analyzing the present and previous TS equations, inflection points appeared based on L values of approximately 5.5–6.0 cm for each frequency, indicating the juvenile stage. Consequently, both TS equations can be utilized to include juvenile anchovies within the range of L = 4.8–6.0 cm. These findings provide valuable insights into the TS characteristics of juvenile anchovies and support accurate TS equations for various growth stages.

We aimed to understand how changes in our proposed TS equation for juvenile anchovies can affect biomass estimation. In Korea, juvenile anchovies are primarily present in spawning grounds during the breeding seasons in spring and autumn, whereas adult anchovies are more widely distributed in summer and winter [19]. If the juvenile or adult TS equations are used to estimate anchovy biomass, the biomass when juveniles or adults dominate the population is underestimated or overestimated, respectively. To simulate and evaluate the biomass estimation, we used sample acoustic data collected from the western coast of Korea in September (Figure 6a). The net sampling data revealed that juvenile anchovies dominated, constituting more than 99% of the sample population (Figure 6b). The L values of the 66 sampled anchovies ranged from 1.5 cm to 6.8 cm, with an average L of 2.84 cm (Figure 6c), indicating that the anchovies were predominantly in the juvenile stage during this period. Therefore, we utilized the TS equations for adult anchovies from our previous study [14] and the newly proposed TS equation for juvenile anchovies (this study) to calculate the biomass. The estimated anchovy biomass exhibited a significant difference of approximately 20 times or more in the sample echogram (NASC, nautical area scattering coefficient = 794.36 m²/n·mile²; refer to [21] for NASC calculation). Thus, it is imperative to apply distinct TS equations for both juvenile and adult growth stages when conducting fishing or studying spawning grounds. For accurate estimation of anchovy biomass, it is crucial to employ different TS equations corresponding to the specific growth stages of the anchovies. Additionally, further investigation is necessary to understand the impact of other species on the growth stages of both juveniles and adults.

More than 90% of the TS was found to be associated with the swim bladder size [11]. Similar to the growth curve of human lungs, the growth curve of fish swim bladders also varies with age and size. To verify the presence of swim bladders in anchovies at different L values, we utilized soft X-ray radiography on additional anchovy samples. The results revealed that swim bladders were observed in anchovies with L values greater than 5.2 cm, but they were not observed in anchovies with L values less than 4.6 cm (Figure 7). This indicates that the presence and development of swim bladders are closely related to the size of the anchovies. The observed low TS value in juvenile anchovies can be attributed to the small or immature size of their swim bladders. Previously, information on swim bladders in juvenile anchovies with L < 5.0 cm has been scarce [26]. However, the TS values and images obtained in our study demonstrated that the TS is intricately linked to the growth of anchovies and undergoes changes based on their developmental stage and swim bladder size. The correlation between swim bladder development and TS values reinforces the importance of considering the growth stages and swim bladder size when estimating the TS of anchovies. Such insights will lead to more accurate biomass estimations and improve our understanding of the dynamics of anchovy populations in different stages of development.

5. Conclusions

In our study, we investigated the acoustic TS of anchovies at various growth stages, including juvenile, young, and adult. To achieve this, we examined a total of 21 live Japanese anchovies using 38, 120, and 200 kHz split-beam transducers, which served as the primary parameters for estimating biomass. During the ex situ TS measurements, we obtained TS values for anchovies with different lengths, ranging from 3.13 to 9.85 cm. However, the TS values of juvenile anchovies did not align with those derived from the previous TS equation [14]. This discrepancy indicated that applying the same TS equation based on the acoustic characteristics of anchovies at different growth stages, particularly for juvenile anchovies, may not be appropriate. Therefore, we proposed and fitted a new least-squares regression for a TS equation specifically tailored to estimate the TS of juvenile anchovies. The TS equations were determined for each frequency as follows: at 38 kHz, “a” was 56.1, and “b” was 94.7; at 120 kHz, “a” was 59.7, and “b” was 98.5; and at 200 kHz, “a” was 64.2, and “b” was 100.0 (Equation (3)). It is crucial to apply the TS equation specific to each growth stage when estimating biomass using the L value of the anchovies. These findings have significant implications, as they could improve the accuracy of biomass estimations in spawning grounds. By considering the specific TS equations for different growth stages, we can obtain more reliable and precise estimations of anchovy biomass, contributing to a better understanding of the dynamics of anchovy populations at various growth stages.