Skip Navigation

ICES Journal of Marine Science: Journal du Conseil 2004 61(3):378-382; doi:10.1016/j.icesjms.2003.08.003
© 2004 by ICES/CIEM International Council for the Exploration of the Sea/Conseil International pour l'Exploration de la Mer
This Article
Right arrow Abstract Freely available
Right arrow FREE Full Text (PDF) Freely available
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Alert me to new issues of the journal
Right arrow Add to My Personal Archive
Right arrow Download to citation manager
Right arrowRequest Permissions
Google Scholar
Right arrow Articles by Didrikas, T.
Right arrow Articles by Hansson, S.
Right arrow Search for Related Content
PubMed
Right arrow Articles by Didrikas, T.
Right arrow Articles by Hansson, S.
Social Bookmarking
Add to CiteULike   Add to Connotea   Add to Del.icio.us  
What's this?

In situ target strength of the Baltic Sea herring and sprat

Tomas Didrikasa,b,* and Sture Hanssona

a Department of Systems Ecology, Stockholm University SE-106 91 Stockholm, Sweden
b Institute of Ecology LT-2600 Vilnius, Lithuania

*Correspondence to T. Didrikas: tel: +46 8 161353; fax: +46 8 158417. e-mail: tomas{at}ecology.su.se.

Using single- and split-beam hydroacoustic equipment (70 and 38 kHz), and gillnet and trawl catches, we derived the relationship TS = 25.5 log10 L – 73.6 (r2=0.95) between acoustic target strength (TS) and fish length (L, cm) for Baltic Sea herring (Clupea harengus membras L.) and sprat (Sprattus sprattus balticus (Schneider)). Fixing the slope to 20, which is a standard practice in hydroacoustics, gave TS = 20 log10 L – 67.8 (r2=0.91). Normally, the fisheries agencies around the Baltic use a TS–length relationship that is based mainly on data from the North Sea and the intercept-value in this equation is 3.4 dB lower than that reported in this paper. This difference corresponds to an approximately twofold difference in assessed stock biomass.

Keywords: Baltic Sea, fat content, herring, hydroacoustics, in situ target strength, salinity, sprat

Received 24 February 2003; accepted 8 August 2003.


   1 Introduction
Top
 1 Introduction
2 Materials and methods
 3 Results and discussion
 References
 
Hydroacoustic techniques are frequently used to assess the abundance and biomass of fish. In such studies, the quantitative relationship between the size of a fish and its target strength (TS), i.e. the intensity of the echo returned from the fish, is critical (MacLennan and Simmonds, 1992). In fish, the swimbladder is responsible for most of the reflected sound (90–95% of the back-scattered energy, Foote, 1980) and factors that influence the size and shape of the swimbladder also influence the TS value of the fish. Since the swimbladder is important for the fish to maintain neutral buoyancy, individuals that have a high fat content or live at high salinity are likely to have smaller swimbladders than conspecifics that are lean or live at low salinity (Ona, 1990; Ona et al., 2001).

Because of these factors, we expect Baltic Sea herring (Clupea harengus membras L.) to have a larger swimbladder and, therefore, a higher TS than similarly sized North Sea herring, as the former is both leaner and lives at a lower salinity. Despite this, the fisheries agencies around the Baltic base their acoustic assessments of herring on the same TS-to-size relationship as is used for North Sea herring (ICES, 1997). In this paper, we explore the relationship between in situ TS values and the size of herring and sprat (Sprattus sprattus balticus (Schneider)) in the Baltic Sea.


   2 Materials and methods
Top
 1 Introduction
 2 Materials and methods
3 Results and discussion
 References
 
2.1 Acoustic data
Acoustic data were collected in Swedish coastal areas and the open areas of the Lithuanian Exclusive Economic Zone (LEEZ) in the Baltic Sea. The echosounders used in the coastal studies in 1990–1992 and 2001–2002 were a 70-kHz, single-beam SIMRAD EY/M unit and a 70-kHz, split-beam SIMRAD EY500, respectively. Both units operated at 11.2° half-power beam angle and 0.6-ms pulselength. In two further surveys in 2002 in the LEEZ we used a 38-kHz, SIMRAD EY500 split-beam echosounder (11.5° half-power beam angle, 1-ms pulselength). The echosounders were calibrated with standard copper spheres as recommended by the manufacturer. The Lobe program by SIMRAD was used to calibrate the split-beam units. Calibration differences between years were minor, i.e. less than 1 dB.

Single-beam data were analysed using HADAS software (version 3.99, developed by T. Lindem, Institute of Physics, University of Oslo, Norway, and described in Walline et al., 1992). This programme uses a modified Craig and Forbes (1969) algorithm to derive fish target-strength distributions from the measured distribution of peak voltage response from single-fish echoes (40 log R TVG function). Single-fish echoes are defined as echoes with less than twice the pulselength and with a monotonic increase and decrease in voltage response (Rudstam et al., 1999). Due to sonar-hardware noise and software limitation, we used –56 dB as the smallest target-strength group for the single-beam sonar.

Split-beam data were analysed using the EP500 analysis software (version 5.5, SIMRAD, 2000). This program calculates volume-backscattering strength and target strength from individual fish by simultaneously applying a 20 and 40 log R TVG function. We used –60 dB as the lower threshold for TS values when analysing 70-kHz data and –57 dB for 38 kHz, in accordance with the thresholds suggested for different frequencies by Balk (2001).

Calculation of the mean TS values was done in the linear domain and the results transformed back to dB values (c.f. MacLennan and Simmonds, 1992). TS values from a single-beam unit are biased towards lower values as a consequence of the Craig and Forbes algorithm used in HADAS. To compensate for this, 0.8 has been added to TS values from the single-beam unit (Rudstam et al., 1999).

2.2 Biological sampling
Fish in coastal areas were sampled with vertical gillnets (Hansson, 1988), which were 1.5 m wide and 25 m deep, made from monofilament nylon yarn and with one mesh size per net. Each time, seven gillnets (bar mesh sizes 4, 6.25, 8, 10, 12, 15, and 18.75 mm), covering the water column from the bottom to the surface, were set overnight at 22–25-m depth. The distance between each gillnet at the fishing site was 20–70 m. The total length of all fish caught was measured. Fishing was done within a few hundred metres of the acoustic transect used for TS estimations. These transects were 0.8–2.0 km long and the acoustics recordings were always done at night, at least 1 h after sunset and not later than 1 h before sunrise.

To account for the selectivity of the different mesh sizes, we used the approach described by Hansson and Rudstam (1995). This is a modification of a method originally developed by Wulff (1986). Hansson and Rudstam (1995) assumed skewed normal selectivity curves, with the selectivity of each mesh size scaled to a maximum of 1. For calculations, we used the equations:


Formula 1

(1)


Formula 2

(2)


Formula 3

(3)


Formula 4

(4)
where SL,m is the selectivity for a fish of length L (millimetres) in a net with bar mesh m (millimetres), L0,m is the modal fish length in mesh size m, {sigma}m is the standard deviation for mesh m, and k is the skewness constant. Values of the used parameters a, b, and k were 11, 1.2, and 0.016, respectively. The average fish length in a catch was calculated from the size frequency distribution of herring and sprat combined, after dividing numbers-per-size class with the gillnet selectivity factor summed over all panels for that size class.

In the open-sea area (LEEZ), fish were caught at night with a pelagic trawl with a circular opening diameter of 32.2 m, fitted with a fine mesh (8 mm) codend. The trawl was towed in mid-water (30–50 m) at a speed of three knots for 30–40 min. All fish in a 20–25-kg subsample from the catch were identified to species and their total length measured. Acoustic data were recorded during the trawling.

2.3 Conditions for TS measurement
TS distributions can be biased if echoes from two or more fish are overlapping and misinterpreted as an echo from a single fish. To handle this problem, Sawada et al. (1993) developed an index (Nv) that represents the number of fish per effective reverberation volume and recommended some maximum values for this index. The index is calculated as:


Formula 5

(5)
where c is the sound speed (m s–1), {tau} is the transmit pulselength (s), {psi} is the equivalent beam angle (sr), r is the distance from the transducer (m), and n is volumetric fish density (fish m–3). We calculated the latter value as:


Formula 6

(6)


Formula 7

(7)
where {rho} is fish density (fish ha–1), r1 and r2 are upper and lower depths for the layer of interest, sv is the volume-backscattering coefficient, and {sigma}bs is the backscattering cross-section of the layer.


   3 Results and discussion
Top
 1 Introduction
 2 Materials and methods
 3 Results and discussion
References
 
Only those data sets where clupeids contributed more than 80% of the total catch by number were used in this analysis. In the trawl hauls, clupeids constituted more than 95% of the catch and in the gillnets, they made up to 90% of the catch on eight of 24 occasions. The other species caught in the gillnets were mainly smelt (Osmerus eperlanus L.), roach (Rutilus rutilus L.), and on several occasions fourhorn sculpin (Triglopsis quadricornis L.). Roach were consistently caught very close to the surface, above the depth of the transducer in fact, and therefore were not included in the acoustic estimates. Smelt were often caught close to the bottom and fourhorn sculpin were always caught at the bottom. In our analyses we assume that all the acoustic targets were from clupeids and that herring and sprat of the same length have the same TS value (Nakken and Olsen, 1977).

Clupeids in coastal gillnet catches were smaller (x = 10.5cm, Table 1) than in trawl catches from the open sea (x = 17.1 cm). The length range was wider in coastal areas due to a larger fraction of smaller fish. The fish-length distributions were frequently bi- or even multi-modal and this was sometimes caused by abundant catches of young-of-the-year (YOY) herring. It was generally difficult to fit these modes with modes in the TS values, and to avoid a rather arbitrary matching of modes we correlated the average length of clupeids in the catches with the average TS value from the acoustics.


View this table:
[in this window]
[in a new window]

  		Table 1 The data used for the calculations in this study: TS from single-beam systems (Single b) corrected for the 0.8-dB bias compared with corresponding values from split-beam systems (Split b), as described by Rudstam et al. (1999). N.B. Only those observations in which the Sawada index (Nv) is <0.1 and where clupeids constituted ≥80% of the catch were used.

 
Unlike O'Driscoll and Rose (2001), who worked with juvenile capelin (Mallotus villosus Müller), we found no clear correlation between the Sawada index (Nv, range 0.001–0.54) and the observed average TS value. The relationship between TS (in dB) and fish length (L in cm) was (±s.e. of coefficients):


Formula 8

(8)
or if the slope was fixed to 20:


Formula 9

(9)
However, to reduce the risk that misclassified echoes from multiple fish biased our results, we made a second analysis where we included only samples with Nv<0.1 (Figure 1). This Nv threshold is 2.5 times higher than that recommended by Sawada et al. (1993), but the use of higher Nv values has been discussed by Ona and Barange (1999) and practised in several studies (e.g. Rudstam et al., 1999; O'Driscoll and Rose, 2001; Warner et al., 2002). This subset of data gave:


Formula 10

(10)


Formula 11

(11)
By excluding data sets where Nv>0.1, we expected a decrease in the constant in the equation with a fixed slope. This is exactly what happened; the constant dropped from –67.3 to –67.8 (Equations (9) and (11)). As Equations (10) and (11) have different numbers of fitted parameters, we estimated the ratio of the residuals' sum of squares vs. number of fitted parameters, as described by Hilborn and Mangel (1997). This showed that the stronger relationship of Equation (10) was real and not a result of fitting an extra parameter. Adjusting the slope parameter to the data, instead of using 20 as a default value, is contrary to the "normal" practise in hydroacoustics. However, McClatchie et al. (2003) argue that a slope of 20 is neither better nor worse than any other value, and that the parameter which gives the best fit should be used.


Figure 1
View larger version (5K):
[in this window]
[in a new window]
[Download PowerPoint slide]
  		Figure 1 The relationship between fish length (log L) and mean target strength (TS) for the Baltic Sea clupeids. (1) The best-fit regression equation TS = 25.5 log10 L – 73.6 (n = 15 r2 = 0.95); (2) the standard-form regression equation with the slope set to 20: TS = 20 log10 L – 67.8 (n = 15, r2 = 0.95); and (3) the equations suggested by ICES: TS = 20 log10 L – 71.2.

 
Three of our data points are from a 38-kHz transducer, while the rest are from 70-kHz systems. This may have influenced our findings (c.f. Love, 1971), but excluding the data that were collected at 38 kHz changed the result only very marginally: the slope and intercept in Equation (10) changed to 25.4 and 73.5, respectively (difference 0.1). Furthermore, model analyses predict that the difference in TS values between 38 and 70 kHz for Baltic herring should be only about 1 dB (John Horne, personal communication, 2003; ICES, 2003).

From the same archipelago area studied here and also based on clupeid catches in vertical gillnets and a 70-kHz system, Rudstam et al. (1988) suggested the relationship TS = 20 log10 L – 69.9. After correcting for the systematic difference between split- and single-beam acoustics (see Acoustic data), the calculated TS value from this relationship is 1.3 dB lower than ours. One possible explanation for this difference is that the smallest gillnet mesh size used by Rudstam et al. (1988) was 8-mm bar mesh. As a consequence of this, they had relatively few small fish (<10 cm) despite the fact that YOY herring can be particularly abundant in the archipelago. This kind of bias would result in an underestimated intercept value, which is consistent with the difference between these two studies.

In herring and sprat stock assessments in the Baltic, the fisheries agencies use 38-kHz sounders and are recommended by the International Council for Exploration of the Sea (ICES, 1997) to apply the relationship TS = 20 log10 L – 71.2. This equation is mainly based on in situ studies from the mid-1980s (Degnbol et al., 1985; Lassen and Stæhr, 1985; Foote et al., 1986, Table 2). With the exception of the study by Lassen and Stæhr (1985), these TS estimates were made in the North Sea. Given the much lower salinity in the Baltic than in the North Sea (approximately 7{per thousand} and 35{per thousand}, respectively) and a considerably lower muscle fat content in the Baltic herring (2–5%) compared to 7–19% in the North Sea (Aidos et al., 2002; Bignert, 2002), the swimbladder is likely to be bigger in the Baltic. As a result of this, Baltic herring of a given size are expected to have a higher TS value and the TS-to-length relation to have a less negative intercept as suggested by our results and those of Rudstam et al. (1988).


View this table:
[in this window]
[in a new window]

  		Table 2 The target strength of the Baltic Sea clupeids. The first two equations are calculated from the data in Table 1.

 
Our finding is similar to that of Ona et al. (2001), who found that the TS values of Norwegian spring spawning herring were 4–8 dB higher than used in regular assessments of this stock. This has substantial implications to resource management. If our TS–length relationship is correct, the Baltic Sea herring and sprat biomasses are less than half as large as suggested by the fisheries agencies using the TS–length equation recommended by ICES.


   Acknowledgements
 
This study was supported by The Knut and Alice Wallenberg Foundation, The Swedish Institute (SI), the Swedish Council for Forestry and Agricultural Research (SJFR), the Royal Swedish Academy of Agriculture and Forestry, the Swedish National Board of Fisheries, and the Fisheries Department under the Ministry of Agriculture of The Republic of Lithuania. Ian Winfield gave valuable comments on the manuscript.


   References
Top
 1 Introduction
 2 Materials and methods
 3 Results and discussion
 References
 

    Aidos I., van der Padt A., Luten J.B., Boom R.M. (2002) Seasonal changes in crude and lipid composition of herring fillets, byproducts, and respective produced oils. Journal of Agricultural and Food Chemistry 50:4589–4599.[CrossRef][Web of Science][Medline]

    Balk H. (2001) Development of hydroacoustic methods for fish detection in shallow water. Thesis for the degree of Doctoral Scientiarium. Faculty of Mathematics and Natural Sciences, University of Oslo, Oslo, Norway. 309 pp.

    Bignert A. (2002) Comments concerning the National Swedish contaminant monitoring programme in marine biota. Mimeographed report from the Swedish Museum of Natural History, Stockholm pp. 1–177.

    Craig R.E. and Forbes S.T. (1969) A sonar for fish counting. Fiskedirektoratets Skrifter Serie Havundersøkelser 15:210–219.

    Degnbol P., Lassen H., Stæhr K.J. (1985) In situ determination of target strength of herring and sprat at 38 and 120 kHz. Dana 5:45–54.

    Foote K.G. (1980) Importance of the swimbladder in acoustic scattering by fish: a comparison of gadoid and mackerel target strengths. Journal of the Acoustical Society of America 67:2084–2089.[CrossRef][Web of Science]

    Foote K.G., Aglen A., Nakken O. (1986) Measurement of fish target strength with a split-beam echosounder. Journal of the Acoustical Society of America 80:612–621.[CrossRef][Web of Science]

    Hansson S. (1988) A simple vertical gillnet system for variable current conditions. Hydrobiologia 160:107–110.[CrossRef][Web of Science]

    Hansson S. and Rudstam L.G. (1995) Gillnet catches as an estimate of fish abundance: a comparison between vertical gillnet catches and hydroacoustic abundances of Baltic Sea herring (Clupea harengus) and sprat (Sprattus sprattus). Canadian Journal of Fisheries and Aquatic Sciences 52:75–83.

    Hilborn R. and Mangel M. (1997) Model selection using sum of squares. The Ecological Detective: Confronting Models with Data(Princeton University Press, Princeton, New Jersey) pp. 114–117.

    ICES. (1997) Report of the study group on Baltic acoustic data. ICES CM 1997/J: 3. 47.

    ICES. (2003) Report of the study group on target-strength estimation in the Baltic Sea. ICES CM 2003/B: 05. 14.

    Lassen H. and Stæhr K. J. (1985) Target strength of Baltic herring and sprat measured in situ. ICES CM B: 41. 1–14.

    Love R.H. (1971) Dorsal-aspect target strength of an individual fish. Journal of the Acoustical Society of America 49:816–823.[CrossRef][Web of Science]

    MacLennan D.N. and Simmonds E.J. (1992) Fisheries Acoustic(Chapman & Hall, London) 325 pp.

    McClatchie S., Macaulay G.J., Coombs R.F. (2003) A requiem for the use of 20 log 10 length for acoustic target strength with special reference to deep-sea fishes. ICES Journal of Marine Science 60:419–428.[Abstract/Free Full Text]

    Nakken O. and Olsen K. (1977) Target-strength measurements of fish. Rapports et Procès-Verbaux des Réunions du Conseil International pour l'Exploration de la Mer 170:52–69.

    O'Driscoll R.L. and Rose G.A. (2001) In situ acoustic target strength of juvenile capelin. ICES Journal of Marine Science 58:342–345.[Abstract/Free Full Text]

    Ona E. (1990) Physiological factors causing natural variations in acoustic target strength of fish. Journal of the Marine Biological Association of the United Kingdom 70:107–127.[Web of Science]

    Ona E. and Barange M. (1999) Single-target recognition. In Ona E. (Ed.). Methodology for Target Strength Measurements (With Special Reference to In Situ Techniques for Fish and Micro-Nekton)(ICES, Copenhagen, Denmark) pp. 28–43 ICES Cooperative Research Report No. 235 59 pp.

    Ona E., Zhao X., Svellingen I., Fosseidengen J.E. (2001) In Funk F., Blackburn J., Hay D., Paul A.J., Stephenson R., Toresen R., Witherell D. (Eds.). Seasonal variation in herring target strength. Herring: Expectations for a New Millennium 461–487 University of Alaska Sea Grant, AK-SG-01-04, Fairbanks.

    Rudstam L.G., Hansson S., Lindem T., Einhouse D.W. (1999) Comparison of target-strength distributions and fish densities obtained with split- and single-beam echosounders. Fisheries Research 42:207–214.[CrossRef][Web of Science]

    Rudstam L.G., Lindem T., Hansson S. (1988) Density and in situ target strength of herring and sprat: a comparison between two methods of analyzing single-beam sonar data. Fisheries Research 6:305–315.[CrossRef][Web of Science]

    Sawada K., Furusawa M., Williamson N.J. (1993) Conditions for the precise measurement of fish target strength in situ. The Journal of the Marine Acoustic Society of Japan 20:73–79.

    Walline P.D., Pisanty S., Lindem T. (1992) Acoustic assessment of the number of pelagic fish in Lake Kinnert, Israel. Hydrobiologia 231:153–163.[CrossRef][Web of Science]

    Warner D.M., Rudstam L.G., Klumb R.A. (2002) In situ target strength of alewives in freshwater. Transactions of the American Fisheries Society 131:212–223.[CrossRef]

    Wulff A. (1986) Mathematical model for selectivity of gill nets. Archiv fuer Fischereiwissenschaft 37:101–106.


Add to CiteULike CiteULike   Add to Connotea Connotea   Add to Del.icio.us Del.icio.us    What's this?


This article has been cited by other articles:


Home page
 ICES J. Mar. Sci.Home page
S. M. M. Fassler and N. Gorska
On the target strength of Baltic clupeids
ICES J. Mar. Sci., July 1, 2009; 66(6): 1184 - 1190.
[Abstract] [Full Text] [PDF]

Home page
 ICES J. Mar. Sci.Home page
S. M. M. Fassler, A. S. Brierley, and P. G. Fernandes
A Bayesian approach to estimating target strength
ICES J. Mar. Sci., July 1, 2009; 66(6): 1197 - 1204.
[Abstract] [Full Text] [PDF]

Home page
 ICES J. Mar. Sci.Home page
S. M. Kasatkina
The influence of uncertainty in target strength on abundance indices based on acoustic surveys: examples of the Baltic Sea herring and sprat
ICES J. Mar. Sci., July 1, 2009; 66(6): 1404 - 1409.
[Abstract] [Full Text] [PDF]

Home page
 ICES J. Mar. Sci.Home page
G. Pedersen, N. O. Handegard, and E. Ona
Lateral-aspect, target-strength measurements of in situ herring (Clupea harengus)
ICES J. Mar. Sci., July 1, 2009; 66(6): 1191 - 1196.
[Abstract] [Full Text] [PDF]

Home page
 ICES J. Mar. Sci.Home page
H. Peltonen, M. Luoto, J.-P. Paakkonen, M. Karjalainen, A. Tuomaala, J. Ponni, and M. Viitasalo
Pelagic fish abundance in relation to regional environmental variation in the Gulf of Finland, northern Baltic Sea
ICES J. Mar. Sci., April 1, 2007; 64(3): 487 - 495.
[Abstract] [Full Text] [PDF]

Home page
 ICES J. Mar. Sci.Home page
H. Peltonen and H. Balk
The acoustic target strength of herring (Clupea harengus L.) in the northern Baltic Sea
ICES J. Mar. Sci., January 1, 2005; 62(4): 803 - 808.
[Abstract] [Full Text] [PDF]

This Article
Right arrow Abstract Freely available
Right arrow FREE Full Text (PDF) Freely available
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Alert me to new issues of the journal
Right arrow Add to My Personal Archive
Right arrow Download to citation manager
Right arrowRequest Permissions
Google Scholar
Right arrow Articles by Didrikas, T.
Right arrow Articles by Hansson, S.
Right arrow Search for Related Content
PubMed
Right arrow Articles by Didrikas, T.
Right arrow Articles by Hansson, S.
Social Bookmarking
Add to CiteULike   Add to Connotea   Add to Del.icio.us  
What's this?