This paper not to be cited without prior reference to the authors

Transcription

1 , This paper not to be cited without prior reference to the authors International Council for Exploration of the Sea C.M. 1983/J:9 altic Fish Committee Acoustic estimates of the herrinq.and sprat stocks in the Baltic proter, the Bothnian Sea and the Gulf of Bothnia pre iminary results by Olle Hagsträm, Lars-Erik Palmen Paul Degnbol otto Rechlin Inst. of Marine Research, Lysekil, Sweden North Sea Center, Hirtshals, Denmark Inst. for Deep Sea Fishery, and Fish Processing, Rostock, GOR Abstract This paper presents preliminary results of the fifth acoustic.survey carried out.as a cooperativ research between Sweden, Denmark, GOR and Polen in the Baltic proper during Septernber October The estimates are based on a length dependent conversion factor and the methods are described. Performance data and calibration data are given. The biomass and the number per age group of herring and sprat are calculated on a subdivision basis. F-values are calculated from reportea catch data and acoustic estimates. Enclosed figures show the course tracks and positions of the hauls. August, 1933

2 .. 1. Introduction In September-october 1982, as in October the Baltic proper was covered by a combined acoustic and pelagic trawl survey. As in 1981 the survey was extended to the Bothnian Sea and in addition the Gulf of Bothnia was covered for the first time. The survey was carried out in co- operation between the Institute for Deep Sea Fishery and Fish Processing, Rostock-Marienehe, German Democratic Republic, Morski Instytut Rybacki W Gdyni, Gdynia, Polen, Danmarks Fiskeri og Havunders~gelser, Hirtshals, Denmark and the Institute of Marine Research, Lysekil, Sweden. The participating vessels were R/V "Argos", R/V "Danall and M/T "Hel-l00". The authors thank A. Orlowski and O. Wresinski from Poland, N. Schulz and H. Bremer Rothbarth from DDR who participated in the survey on board R/V "Argos ll, and the crews on board the participating vessels Materials and methods 2.1. Acoustic equipment and calibration The integration wascarried out by R/V IIArgos", equipped with a Simrad EK-S 120 khz echo sounder and a Simrad QM MK 11 echo integrator. The technical data and settings of the equipment are shown in Table 1. The equipment was calibrated against a standard copper sphere in the 3rd of September before the Baltic survey. The results of the calibration of"argos"equipment is given below as the sum of Source level (SL) and Voltage responce (VR)..e Date SL + VR Method hydrophone (116.4) copper sphere (117.7) copper sphere The values in.brackets are corrected for temperature and are thus direct comparable with the hydrophone measurement. All values presented above are measured with Simrad standard TVG function. In the Baltic surveys a new TVG amplifier has been used since The difference between the two TVG amplifier was measured twice, showing a constant higher amplification of 1.8 db with the Baltic TVG function. As a planned calibration during the survey in the Baltic. failed, the most recent calibration was corrected for the difference in amplification giving the SUffi of SL and VR of db. The intercalibration between "Argos" and "Dana" gave results with good linear relation. The run was carried out during excellent condition on scattering layers of varying intensities. The regression equation: MM (Dana) = MM (Argos) were r=o.94 describes the relation

3 3.. However, this 4.8 db lower level of performance for "Dana" do not fit with the most reeent ealibrations, whieh do imply a differenee of 9.2 db between the ships. "Dana" has only performcd one ealibration against the standard coppar sphere target which gave a quite different result compared with the initial hydrophone measurement. Furthermore the 120 khz transducer on "Dana" has displayed a drift in impedance. For this reason ealibration data from "Dana" ean not be eonsidered realible. Due to the uneertanties of transdueer stability on "Dana" only the aeoustic data from "Argos" has been used in the ealeulations Correetion for TVG-deviation The hydrographie eonditions in the Baltie differ from those used by Simrad to ealeulate and establish the time varied gain (TVG). Aglen etal. (1981) have pointed out that the integrated values are overeompensated by the equipment and that the deviation between eorreet and assumed values inereased with fish depth. In order to have more relevant TVG, a new TVG-amplifier was installed during 1981 (sound absorption= db/m and sound velocity 1445 m/s).'fuis amplifier has also been used during the 1982 survey. Hydrographie data eolleeted during the survey was used to ealeulate averag~ salinity and temperature for each lces sub-division. Mean values was used to ealeulate sound absorption, sound velocity in order to make eomparison with the values used in the new TVG-amplifiere This minor eorrection, in the order of 0.1 db, has not been applied to 1981 and 1982 estimate due to computer problems Conversion factor and TS-values The integrated values were converted to biomass using the same length dependent eonvernion faetor as previous years. Details of this method are given in Anon. (1981a), seetions 2.4. and 2.5. TS-values reported by Aglen et al., 1981 were used. TSkg -10 log L db for herring TSkg = -10 log L db for cod Sprat, salmon and 3-spined sticklebaek were assumed to have the same aeoustie properties as herring and all gadoids the same as eod Trawling Trawling with single boat pelagic trawl was performed to identify echo traees. A bar length of 11 mm was. used in the rear parts of trawls. The intention was to carry out at least one haul in eaeh lces rectangle. In ease of se-

4 4. parate echo traces at different depths, hauls were made through each appropriate depth layer. The fishing depth and gear performance were checked by net sonde.. In all 103 hauls were made, 61 by "Argos", 36 by "HEL-100" and 6 by "Dana" Data analysis ICES statistical rectangles were used as strata and the estimated biomass in a rectangle was split on the different species according to the weight proportion in hauls. All valid hauls were treated as representative sampies. In case of two or more hauls in a rectangel catch data were averaged using arithmetic means. If data were lacking in a rectangle an aritmetic mean of the data from neighbouring rectangles were used. Due to lack of data on mean weight at age in the Polish trawl sampies a length-weight equation was established from "Argos" trawl 'samples (correlation coefficient 0.96)and used to calculate mean weights for the Polish sampies 3. Results 3.1. Coverage The 1982 survey was carried out to get comparable coverage with the earlier surveys. The over all covered area of the surveys are shown below. Year Baltic prop. Area (NM 2 ) Bothnian Sea Area (NH 2 ) ( NH 2 in subdivision 31) Course tracks and positions of pelagic haul are found in Figure 1 a-d. The slightly modified lces' Figure 2. sub-divisions are found in 3.2. Herring Biomass and distribution The herring biomass in the covered part of Baltic proper (sub-divisions 24-29) was estimated to x 103 tonnes, 91.7 x 103 tonnes in sub-division 30 and 55.6 x 103 tonnes in sub-division 31 (Table 3). The highest biomass was found in sub-division 28 were also the most herring by number was found (Table 4). The herring densities in tonnes NM2 are presented in Figure 3.

5 Age distributions The compositions in sub-divisions are shown in Table 4 and the herring spawning stock (~3-group) in Table 5. The juvenile herring, here defined as 2-group and younger, was most abundant in a band in the western part of the Baltic from SW to NE and in the southern part of Bothnian Sea. The adult herring, defined as 3-group and older, was dispersed over the central area with lower concentrations in the SW, Zatoka Gdanska and at the mouth of the Gulf of Finland high numbers of adults was found in the Gulf of Bothnia (Figure 5).. Figure 6 shows the mean length at age group for herring. The mean length are weighed according to the numbers in each age group in the samples and an unweighted mean length of all samples by sub-divisions were calculated Sprat Biomass and distribution The sprat biomass in the covered part of the Baltic proper was estimated to x 103 tonnes and 43.9 x 10 3 tonnes in lees sub-division 30 (Table 6). The highest biomass as well as the highest number of sprat were found in subdivision 26 (Table 7). Sprat was not found in sub-division 31 (Gulf of Bothnia) Age distribution The age-compositions in sub-divisions are shown in Table 7 and the sprat spawning stock (~3-group) in Table 8. The juvenile sprat, defined as 2-group and younger, showed high densities in the SW, SE and NE Baltic proper (Figure 8). The densities were also high in sub-division 29N. The adult sprat, defined as 3-group and older, was more dispersed with the highest densities in sub-division 29N and in parts of sub-division 30 (Figure 9). 4. Discussion 4.1. Coverage The over all coverage was almost the same aso during previous years. Aglen (1982) has defined the "degree of coverage" as N/~ were N is the distance (NM) sailed within the survey area

6 6. SD(t)/t and A is the surveyed area (NM 2 ). If this is applied on the Baltic proper the results are These figures are fairly high compared to coverage used by Aglen (1982). He has also defined a "coefficient of variation" as the ratio between the standard deviation of the abundance indexes from different coveragesof an area with the same fish and the mean value of the abundance indexes. Aglen's (1982) results show a decreasing coefficient of variation with increasing degree of coverage and that the differences between different survey areas are small. With our figures on degree of coverage, the coefficient of variation should be less than 0.2 and the difference between the years small. This level of precision does however only apply to the estimates of totals. Degrees of coverage for subdivisions are presented in Table 2. The expected coefficient of variation according to Aglen's (1982) results is for subdivision data in the range With an average track of 41.4 NM per 30 x 30 NM2 the degree of coverage on a square basis becomes 1.4. This yields an expected coefficient of variation of for data on a square basis Target strength The choice of target strength for the integrated fish species is still a critical part in the acoustic estimates of fish stock size. Since no new and clearly verified estimates of target strength of the integrated species in situat 120 khz have been presented, the authors have chosen to use the same values as during previous year (Aglen et al.,1981, Hagström et al., 1981) Revision of 1981 estimate The data from 1981 has been rerun taking into account a minor correction 0.1 db of the parameter V and about the same correction of fish depth. The revised e~timate i tons or 4.3 % higher than reported in Hagström et al., lierring in the Baltic proper The bioma5s of herring in the Baltic proper is in the same level as 1981 which is a slight increase since 1980, while the stock size in numbers has continued to decreas (Table 3 and 4). The decreasing stock in numbers has been compensated by increasing mean weights in the stock.

7 The spawning stock (~3-group) in numbers has continued to decline and is only half of the 1978 value (Table 5). Though the figures from 1979 are considered doubtful, this trend is obvious. The reduction of the spawning stock in 1982 is restricted to sub-division 29 N and Sand 27. In sub-division 24 the estimate imply a 6 times large spawning stock compared to 1981 and twice the 1978 estimate. However, the spawning stock estimate is still lovler than the VPA estimate. It is not clear if the a~oustic estimate suggests a. real increase in the spawning stock or the increase is due to a change in migration pattern of th~'~dult stock. It was impl..ied in Hagström et al., 1982 that two main nursery areas,.',one in SW and another in NE of the Baltic proper could be recognized both in 1980 and The hypothesis was supported by different mean length at age of the young herring in the areas. The distribution of juvenile herring in 1982 shows high concentrations in the same areas, but the areas were connected by a band of high concentration in the western part of the Baltic. The hypothesis is still supported by mean length at age as shown in Figure 6. The herring in sub-division 24 and 25 grew faster than the stock from other sub-divisions. 'The adult herring in sub-division 24 shows higher mean length at age, which has been consistent during the trawl data serie. The growth pattern of the herring in sub-division 24 is more similar to the growth of the Kattegatt Spring Spawner component (Petursson and Rosenberg, 1982} for the same period of the year. This supports the hypotesis discussed in Hagström et al., 1982 that at least ~ part of the adult stock migrates to the Kattegatt/Skagerrak area during the feeding period Sprat in the Baltic proper The sprat bioma~s shows a small increase since 1981 but here the stock size in numbers has also increased (Table 6 and 7). However, the spawning stock (~3-group) in numbers has decreased by more than 75 % since 1978 (Table 8). The trend in the decline of the spawning stock is even stronger than for herring. In the southern part, in sub-division 24, 25 and sub-division 2~ a substantial increase of the spawning stock has occurred. The decline of the spawning stock has continued in subdivisions 29 N and S Herring and serat in the Bothnian Sea and the Gulf of Bothnia Thc total biomass of herring in sub-division 30 has decreased about 35 % since 1981 (Table 3). This is due to the decline in numbers of adult herring, about 60 % (Table 5). The biomass in sub-division 31 was about 60 % of the biomass in sub-division 30 (Table 3) and the numbers of adult herring was in fact higher than in sub-division 30. The survey in sub-division 31 was disturbed by bad weather conditions and the analyses is based on 3 hauls only ~~-

8 8. The juvenile herring was concentrated to both sides of the Bothnian Sea and the Gulf of Bothnia (Fig. 4). The densities of adult herring was higher in the western part of the Bothnian Sea (Fig. 5) and in the southern part of the Gulf of Bothnia. The decline of the total biomass of sprat in sub-division 30 was about 20 %. The juveni~sprat do not show any dense concentrations in the Bothnian Sea except to some extent in the NE (Fig. 8). The adult sprat was registrated in high concentrations in the eastern part of the Bothnian Sea Estimates of mortality Mortalities, Fand M, have been calculated from acoustic stock estimates and catch in numbers. The results are given in Tables 9 to 15 for the herring and sprat stocks. In case A is F derived from catches in year i and the stock estimate from Oct.year i-1, whereas in case B the catch and the stock estimate from Oct.the same year are used. The following formula was applied: C = N exp (oe.. x(f+m» x (1-exp(-F-M» xf/(f+m) 0<. = in case A ando<.= 0.75 in case B. In case C has total mortality Z = M+F been estimated from two successive acoustic surveys, F is estirnated as in case Band M as Z-F B

9 9. 5. References Aglen, A., Random errors of acoustic fish abundance estimates in relation to the survey grid density applied. Symposium on Fisheries Acoustics, Bergen, Norway, June No. 33 (Mimeo.). Aglen, A., Hagström, O. and Hlkansson, N., Target strength measurements and C-Value determinations of live Skagerrak herring and cod. lces Doc. C M. "1 981 /B : 12 (Himeo.). Anon., 1981a. Final report of the joint Danish-Swedish acoustic survey in the Skagerrak-Kattegat, 25 August 12 September lces Doc. C.M. 1981/H:30 (Mimeo.) Falk, U., "Kästner, 0., Hagström, 0., Häkansson, N. and Lindquist, A., Hydroacoustic observations in the Baltic proper in October Preliminary results, lees Doc. C.M. 1980/J:16, Hagström, 0., Häkansson, N., Lindquist, A., Kästner, D. and Rechlin, 0., Acoustic estirnates of the herring and thc sprat stocks in the Baltic proper October Preliminary results. lces Doc. C.M. 1981/J:14 (Mirneo.). Hagström, 0., Häkansson, N., Lindquist, A., Falk, U. and Kästner, 0., Acoustic estimates of herring and sprat stocks of Baltic proper in October Medd. fr. Havsfiskelab., Lysekil, nr Hagström, 0., Häkansson, N., Kästner, 0., and Schultz, N., Acoustic estimates of the herring and the sprat stocks in thc Baltic proper and the Bothnian Sea Preliminary results. lces Doc. C.M. 1982/J:6 (Mirneo.) Hlkansson, N., Kollberg, S., Falk, U., Goetze, E. and Rechlin, 0., A hydroacoustic and trawl survey of herring and sprat cf the Baltic proper in October Fischerei-Forschung 17(2) :7-23. Petursson, P.M., Rosenberg, R., Maturity and growth of the spring spawning herring in the Kattegat Skagerrak. Ophelia, 21 (2): ). (Decernber,

10 Table 1. Technical data and settings of the acoustic equipment on board R/V "Argos". Echosounder 5irnrad Ek Frequence Output power Transducer 10 log'f Band width/pulse length TVG and Gain Basic range Discriminator Recorder Gain SL + VR khz 1/1 10 cm ~, ceramic -18 db 3 khz, 0.6 ms 20 log R, 0 db attenuation m 3-8 (varying) db (according to sect. 2.1.) Echointegrator QM Hk II Gain Treshold Interval Bottom stop Channel A db 0 Varying Off Channel B db Varying 0- Depending on Gain Varying On

11 Table 2. Surveyed area, NM 2, by subdiv1sion~, during October A N(NM) NI A N Tot Table 3. Estimated biomass of herring, tonnes x } N Tot

12 Tab14 e. I Numbers by age'gfoup'of herring, ~tober 1981 and 1982, numbers x N Tot. I ~ ' EG ' Tot Table 4 Cont. Table 5. Herring spawning stock (?3 group) in numb~rs (millions) in ICE5'sub-div1s1ona, October" ~he estimate s considered doubt~ul due to calibrat10n uncerta1nty Tot N } } } Tot

13 Table 7. Numbers by age group of sprat, October 1981 and 1982, numbers x N. Tot , A Tot ,- Table 7 Cont ' Tot Table 6~ Eatimated biomasa of aprat, tannes x 10~ N 49.5 } Tot

14 Table 8. Sprat spawning stock ( ;> 3 group) in numbers (milliona) in ICES'sub-divisions,October The estimate 1979 is considered doubtful due to calibration uncertainty , \ N } } Tot

15

16 Table 11. HERRING sub1ltv. 29N, ==0.2 Year- stock Catch stock Catch B A B C class Oct Oct F 81 F 82 F 82 F 82 M Table 12. HERRING sub-div. 31- M=0.15 Year- Stock Catch B class Oct,; F

17 Year Stock Catch Stock Catch B A B C class Oct Oct F 81 F 82 F 82 F 82 M 82 Table 13. SPRAT SUb-ditt 22, 24 and 25. M= Table 14. SPRAT sub-diva 26 and 28. I~=O.5 Year- Stock Catch Stock Catch B A B C class Oct Oct F 81 F 82 F 82 F 82 M

18 Table 15. SPRAT sub-diva 27, M=0.3 Year- Stock Catch Stock Catch B A B C class Oct Oct F 81 F 82 F 82 F 82 M

19 Fig. 1 a. Cruise track of R/V "Argos" in the southern Baltic October 1982 and trawl positions of: R/V "Argos ll A R/V IIDana" = D M/T "HEL-100" = H 1.: 1:r

20 Fig.1 b. Cruise track of R/V "Argos" in the northern Baltic October 1982 and trawl positions of: R/V "Argos" = A HIT "HEL-100" = H, SYENSKA BJffiN _--.JBl2-':c----~~_---.L.-cA A150 oh38 H36

21 Fig. 1 c. Cruise track of R/V "Argos' in the Bothnian Sea October 1982 and trawl positions of: R/V "Argos" = A 3 e SUNDSVAl Al66 Al6S Äl64 62" A1S8 ~]:A.15,::r J,. A1S3 1T 19 20" 22" 23

22 i I Fig.! 1 d. Cruise track of R/V "Argos" in the Gulf of Bothnia October 1982 and trawl positions of: R/V "Argos" ::: A I"

23 _._--_. - Fig. 2. IeES' sub-divisions 'd--fli:::--:-2~~~~fx-~~--j!l---+--f

24 Fig HERRING densities in tonnes per NM October 1982 (rectangles without 5 59 numbers or minus sign are not covered~4---~-+~~ by the survey) o = density < 0.5 tonnes/nm - = no herring found S

25 .. Fig. 4. HERRING, 2'group October 1982 Densit~es as ind. x per NM

26 213 0 Fig. 5. HERRING ~ 3 group October 1982 Densities as ind. x 10-3 per NM

27 MEAN LENGTH (CM) v' x N- 30 -* 1 Ag. 6. o AGE GROUP HERRING mean lengths at age by ICES sub-divisions October 1982

28 Fig SPRAT densities in tonnes per NM October 1982 (rectangles without n~rnbers or minus sign are not covere~~--r--+~~ by the survey) 2 o = density < 0.5 tonnes/nm - = no sprat found

29 ' 16' 20 24' 213,. 6ft Fig SPRAT ~ 2 group October 1982 x Densities as ind. per NM2,, e , 6L

30 I ßo 54 Fig. 9. SPRAT ~ 3 group Densities as ind. x 10-3 per NM '