Swimming angle and target strength of larval Japanese ......using the dorsal camera. The swimming...
Transcript of Swimming angle and target strength of larval Japanese ......using the dorsal camera. The swimming...
Instructions for use
Title Swimming angle and target strength of larval Japanese anchovy (Engraulis japonicus)
Author(s) Ito, Yusuke; Yasuma, Hiroki; Masuda, Reiji; Minami, Kenji; Matsukura, Ryuichi; Morioka, Saho; Miyashita, Kazushi
Citation Fisheries Science, 77(2), 161-167https://doi.org/10.1007/s12562-011-0323-1
Issue Date 2011-03
Doc URL http://hdl.handle.net/2115/45136
Rights © 2011 公益社団法人日本水産学会; The final publication is available at www.springerlink.com; © 2011 TheJapanese Society of Fisheries Science
Type article (author version)
File Information FS77-2_161-167.pdf
Hokkaido University Collection of Scholarly and Academic Papers : HUSCAP
1
T i t l e :
Swimming angle and target s t rength of l a rval J apanese anchovy
(Engraul i s japonicus )
Running Ti t le : Swimming angle and target s t rength of l a rval J apanese
anchovy
Authors and the i r Aff i l i a t ions : Yusuke Ito , Hi roki Yasuma, Rei j i Masuda,
Kenj i Minami , Ryuichi Matsukura , Saho Morioka , Kazushi Miyashi ta
Y. Ito
Labora tory o f Mar ine Ecosys tem Change Analys i s , Graduate School of
Envi ronmenta l Science , Hokkaido Univers i t y, 3 -1-1 Minato-cho , Hakodate
041-8611, JAPAN
Tel : 81-138-40-8817
Fax: 81-138-40-8856
e-mai l : i -you@ees .hokudai .ac . jp
H. Yasuma, K. Miyashi ta
Labora tory of Mar ine Ecos ys tem Change Anal ys i s , Fie ld Science Center fo r
Nor thern Biosphere , Hokkaido Univers i t y, 3 -1-1 Minato-cho , Hakodate ,
Hokkaido 041-8611, JAPAN
R. Masuda, K. Minami
Fisher ies Research Sta t ion Kyoto Univers i t y, Nagahama, Maizuru , Kyoto
2
625-0086, JAPAN
R. Matsukura
Nat ional Research Inst i tute of Fi sher ies Engineer ing, FRA, 7620-7 Hasak ,
Kamisu , Ibaraki 314-0408, JAPAN
S. Morioka
Fisher ies Research Inst i tute , Tokushima Agr icu l ture , Fores t ry and Fisher ies
Technology Suppor t Center , 1 -3 , Hiwasa , Minami , Ka i fu , Tokushima,
779-2304, JAPAN
3
Abs t rac t
The swimming angle of l a rval J apanese anchovy (Engraul i s japonicus )
was measured in a t ank , and target s t rength (TS) was ca l cu la ted us ing a
theore t ica l sca t ter ing model . The mean swimming angle was 12 .8° (s .d . =
±22.1) . Increased speeds of f low led to increased mean swimming angles .
The mean swimming angle a t f low of 5 cm s - 1 was h igher than those a t other
speeds . TS values were es t imated us ing a d i s tor ted-wave Born
approx imat ion model for two cases . Average values were 1–3 cm s - 1 (Case
1 : 11 .5°±22.1) and 5 cm s - 1 (Case 2 : 16 .6°±21.7) fo r cases 1 and 2 ,
respect ive ly. Fo r case 1 , TS ranged f rom -92 .0 to -74 .7 dB with a mean of
-79 .4 dB a t 120 kHz. For case 2 , TS ranged f rom -92 .2 to -75 .2 dB wi th a
mean of -79 .9 dB. The mean TS in case 2 was lower than tha t in case 1 , wi th
the max imum di f ference being 1 .0 dB a t 120 kHz [ s tandard length : 22 .0
mm] . However , there were no s igni f ican t d i f ferences between the regress ion
l ines of cases 1 and 2 . Thus , changes in f low speed a l tered the swimming
angle of l a rval J apanese anchovy, bu t had l i t t l e in f luence on TS.
Key words : l a rval J apanese anchov y, swimming angle , t a rget s t rength
4
T i t l e :
カタクチイワシ (Engraul i s Japonicus )シラスの遊泳姿勢とターゲットス
トレングス
Authors and Aff i l i a t ions :
伊藤祐介 (北大院環 )
安間洋樹 (北大フィールド科セ )
益田玲爾 (京大フィールド研セ )
南憲吏 (京大フィールド研セ )
松倉隆一 (水研セ水工研 )
守岡佐保 (徳島農水総技セ )
宮下和士 (北大フィールド研セ )
Abs t rac t :
カタクチイワシシラスのような無鰾魚を対象に音響計測を実施する
場合 ,わずかな姿勢角変化がターゲットストレングスに大きく影響する .
本研究では ,頭部を上向きにした状態で遊泳する傾向が強いこと (平均
12.8° ,標準偏差±22.1) ,流速が上昇するとこの傾向が強くなることを明
らかにした .また ,分散分析により姿勢角結果を 2 ケースに分け TS を平
均化したが ,両者にほとんど差はみられなかった . したがって ,物理環境
により遊泳姿勢にわずかな違いが生じても ,標準偏差が高い場合は平均
化 TS への影響は小さいと考えられた .
5
Introduct ion
The Japanese anchovy (Engraul i s japonicus ) i s one of the mos t impor tan t
coas ta l f i sher ies species in J apan . The larvae are an especia l l y impor tan t
resource in near-shore f i sher ies . Accord ing to annual f i shery s ta t i s t i cs ,
commercia l tow-net harves t s were valued a t 30 b i l l ion yen in 2007 [1] .
Therefore , da ta on the d is t r ibu t ion and abundance of l a rvae are impor tan t
for near -shore f i sher ies . Accura te quant i t a t ive data on larvae a re
par t icu lar ly impor tan t , no t on ly for appropr ia te ly managing larvae f i sher ies
bu t a l so for pred ic t ing the rec ru i tment and management of adul t s .
Var ious s tudies have used commerc ia l ca tch data to es t imate the
abundance and d i s t r ibu t ion of l a rval anchovy, of ten as l inked to
envi ronmenta l parameters such as curren t , sa l in i t y, and prec ip i ta t ion [2 , 3 ,
4] . However , such analyses mus t use data co l lec ted over a long per iod
(months or years ) , and i t i s thus d i f f icu l t to ob ta in quant i ta t ive es t imates
wi th in a s ingle f i shery season . Converse ly, acous t ic observat ions us ing
quant i ta t ive echo-sounders can p rovide quant i ta t ive data in a shor t t e rm and
have been used for s tock assessments of many species [5 , 6 ] .
In acous t ic surve ys , a quant i t a t ive echo sounder provides ref lec t ions
f rom a f i sh school a t var ious echo in tens i t i es . This acous t ic ref lec t ion i s
conver ted to quant i t a t ive data (e .g . , number of ind ividuals , biomass) us ing
the target s t rength (TS) . Recent ly, the TSs of var ious species has been
repor ted for use in es t imat ing abundance in the f ie ld [7 , 8] . A spl i t -beam
echo sounder can be used to measure TS, i f a t a rget i s no t too smal l in
compar i son with wave length . In larval anchovy, however , f i e ld
6
measurement of TS i s d i f f icu l t . Acous t ic ref lec t ions f rom larval anchovy
are weak because of the smal l s ize of the f i sh and they form school dur ing
the day inf luence the wa y the target de tec t s them. Therefore , the TS o f
l a rval J apanese anchovy should be es t imated us ing a theore t ica l
sound-scat ter ing model .
Many theore t ica l models have been developed for ca lcu la t ing the TS of
f i sh [9] . Most of these models use an approx imate geomet r ic conf igura t ion
to represent the swim-bladder ( in swim-bladdered f i sh) or body ( in
swim-bladder less f i sh) of the target species , which can be obta ined f rom the
re la t ionship between TS and swimming angle (p i tch or ya w) . The Japanese
anchovy i s a physostomous f ish . The swim-bladders of the i r l a rvae are
f i l l ed wi th gas dur ing the night , which helps to reduce energy consumpt ion .
However , dur ing the day, they d i scharge gas f rom thei r swim-bladders and
form schools [10] . Miyashi ta [11] repor ted tha t acous t ic surveys of l a rval
anchovy should be conducted on h igh-dens i ty s chools dur ing the da y.
Moreover , l a rval anchovy TSs of 50 and 200 kHz were es t imated us ing the
DWBA-based deformed cyl inder mode l (DWBA model ) [12 , 13] tha t was
developed for swim-bladder less spec ies . This s tudy sugges ted tha t the
dorsa l average TS i s very sens i t ive to changes in t i l t angle . Therefore , the
TS should be determined us ing the t i l t angle d i s t r ibut ion af ter observat ion
of swimming behavior .
In th i s s tudy, we observed the swimming angles of l a rval J apanese
anchovy us ing a v ideo camera and es t imated the TS as a funct ion of t i l t
angle ( avgTS ) for use in acous t ic surveys . We used the swimming angle
observat ion resu l ts to ca lcu la te the avgTS a t 38 and 120 kHz (which are the
7
main f requencies used by coas ta l research vesse l s in Japan) us ing the
DWBA model . We a l so cons idered the avai lab i l i t y o f avgTS as a sca le fac to r
in acous t ic surve ys , us ing the re la t ionship between avgTS and the body
lengths of l a rval anchovies .
Mater ia l s and Methods
Observat ion of swimming behavior
In J anuary and December 2008, l ive la rval anchovy were provided f rom
f ixed shore net s in Wakasa Ba y. Dur ing both exper imenta l per iods , the
larvae were t ransfe r red as soon as poss ib le to a b lack f iberglass c yl indr ica l
t ank (500 L, 116 cm diameter , 77 cm deep) a t the Fisher ies Research S ta t ion
of Kyoto Univers i ty. One da y la ter , we chose l ive larval anchovy and
t ranspor ted them to an exper imenta l t ank which was main ta ined a t ambient
t empera ture (10℃ ) in an envi ronment tha t blocked natura l l igh t . A 12-h
photoper iod was main ta ined wi thout a dawn or dusk t r ans i t ion in l igh t
intens i t y. Dur ing the day, inc ident l igh t a t the surface measured
approx imate ly 350 lx ; no l igh t was provided dur ing the n ight .
We conducted the swimming exper iment dur ing the da y; to s imula te the
envi ronmenta l condi t ions in the bay, and to ad just the f low in the
exper imenta l t ank seawater was added as needed . The speed of f low was
measured before and af ter v ideo recod ing us ing the Elec t romagnet ic f low
veloci ty (AEM1-D; ALEC Elect ronics Corp . , Tokyo) wi thin the camera
f ie ld angle . Video record ing began 10 minutes af ter the speed of f low was
8
s table and las ted fo r 1 hour a t each speed .
Video record ings of the swimming behavior of 40 ind ividuals were
co l lec ted us ing two underwater cameras (T-WATER-2200c ; HERO, Tokyo)
(TL: 39 .0±3.4 mm) in J anuar y. In December , two Digi ta l HD video camera
recorders (HDR-SR12; SONY Corp . , Tokyo) were used to give h igh qual i t y
image resu l t s for 12 ind ividuals ( to ta l l ength : 39 .2±1.8 mm). Record ing
sys tems were p laced a t a depth of 45 cm (Fig. 1 ) , and two addi t ional
underwater cameras were used to check the r e la t ionships between
indiv iduals and the camera lenses . In thi s envi ronment , we observed the
swimming behavior of l a rval anchovies . The obta ined v ideo record ings were
conver ted to photographs a t 1 -minute in tervals for each exper imenta l per iod .
Swimming angles were ca lcu la ted f rom these photographs us ing image
ed i t ing sof tware (SCM Measure ; Mori tex Corp . , Tokyo) (F ig. 2 ) . There were
two requi rements for measurements : the center l ine of the ind ividual be ing
measured could not be bending, and the indiv idual had to be hor izonta l f rom
the poin t of v iew of the camera lens . The second parameter was conf i rmed
using the dorsa l camera . The swimming angle was computed accord ing to
the above parameter s . The angle was tha t be tween the center l ine of the f i sh ,
an imaginar y l ine running f rom the root of the ta i l to the t ip , and the t rue
hor izonta l . We def ined pos i t ive angles as those for which the f i sh was
head-up and negat ive angles as those fo r which the f i sh was head-down.
Theore t ica l model
A tota l o f 200 larval anchovy ranging f rom 18 .0 to 35 .7 mm SL (SL =
9
1 .22*TL-2 .41 , R 2 =0.99) were used for the TS ca lcu la t ion. The sound
sca t ter ing f rom specimens was es t imated using the DWBA model . A
modi f ica t ion of the Mat lab codes descr ibed in McGehee e t a l . [14] ver . 6 .1
(MathWorks , Nat ick , MA) was used to es t imate the TS of the deformed
cyl inder , given as
( )pos
tilt
tiltcanimallr2ik
2
2
animal
c2SW
rbs
rdcosβ
cosβa2kJe
2gh
h14k
akf
posanimal
pos
⋅×
−
+= ∫
where ( )bsf i s the complex backscat ter ing ampl i tude , the re la t ion to
backscat ter ing cros s -sec t ion bsσ i s given by 2bsbs fσ = , posr i s the pos i t ion
a long the ax is of the deformed cyl inder , and k i s the acous t ic wave
number given b y λ2πk = , where λ i s the acous t ic wavelength . The
subscr ipt sw re fe rs to the surrounding seawater , the subscr ip t an imal
refers to l a rval anchovy, lJ i s a Besse l funct ion of the f i r s t k ind of order l ,
and ca i s the cross -sec t ion rad ius of the cyl inder and inc ident wave. We
d igi t ized and obta ined 200 se t s of posr and ca f rom the dorsa l images o f
specimens . We t r ied to measure the dens i ty cont ras t ( g ) and sound-speed
cont ras t ( h ) o f the larval anchovy us ing the dens i ty-bot t l e [16] and
t ime-of- f l i gh t [17] methods . These values were appl ied to the DWBA model
as parameters ( g = 1 .068 , h = 1 .037) [ Ito et a l . , unpubl ished] . TS (dB) i s
def ined as ( )bs10 σ10logTS = , and the average TS was def ined as the mean TS,
ranging f rom -90° to 90° a t 1° s teps ( t i l t angle : head-down, head-up
posi t ion) .
The t i l t -averaged TS ( avgTS ) was ca lcu la ted using the probabi l i ty dens i ty
(1 )
10
funct ion (PDF) of f i sh t i l t angle ( )bsf in Equat ions (2) and (3) [15] :
( ) ( )∫−
=π/2
π/2Avg dθθfθσσ
AvgAvg σ10logTS =
where θ represents swimming angle ; ( )θf was assumed to be a t runcated
normal d i s t r ibu t ion funct ion . The t runcat ions were made a t θ3Sθ − and
θ3Sθ + , where θ and θS denote the mean and s tandard devia t ion of the t i l t
angle , respect ive ly. Since the number of samples was d i f feren t in each
photograph . S t r ic t ly, we should ex t rac t the same number of samples f rom
each photograph . However , we were not ab le to ex t rac t i t because the
exper iment was l imi ted . Therefore , mean swimming angle and s tandard
devia t ion were ca lcu la ted using those l imi ted samples in th is s tudy.
Resul ts
Theore t ica l TS
Typical examples of the re la t ionship between var ia t ions in TS and body
t i l t angle a t 38 and 120 kHz obta ined by the DWBA model are shown in Fig.
3 . The var ia t ion in TS versus t i l t angle showed peaks a t about 0° (dorsa l ) a t
bo th f requencies , and the maximum TS a t 120 kHz was h igher than those a t
38 kHz in a l l specimens . Moreover, these peaks were re la t ive ly sharp ,
especia l l y for the h igher f requency, sugges t ing tha t s l igh t changes in f ish
t i l t angles have a major effec t on TS. The max imum TSs a t the 120 kHz were
-85 .6 dB and -68 .1 dB in the smal les t and the larges t indiv iduals ,
(2
(3
11
respect ive ly. In ind iv iduals wi th larger s tandard length , the t i l t angle had a
cons iderable impact on TS.
Rela t ionship between swimming angle and f low speed
The f low speeds in an exper imenta l t ank were 1 , 2 , 3 , 5 cm s - 1 ,
respect ive ly. We ext rac ted 2637 angles (J anuar y: 2102, December : 535)
f rom photograph . F ig. 4 shows box p lo ts of the swimming angles a t each
f low speeds . In th i s s tudy, the mean and s tandard devia t ion of the swimming
angle of l a rval anchovy tended to r i se as the f low speed increased . However,
the s tandard devia t ion was not re la ted to the f low speed . Larval anchovy
were a ffec ted by f low var ia t ions (P< 0 .001 , ANOVA), and the mean
swimming angle a t 5 cm s - 1 was h igher those under the o ther f low speeds
(P<0.01 , Tuke y-Kramer tes t) . Therefo re , we separa ted the ca lcu la t ions of
TS in to two cases . One case combined the data f rom f low speeds of 1 to 3
cm s - 1 (case 1) , and o ther used data f rom f low speeds of 5 cm s - 1 (case 2) .
Fig. 5 shows hi s tograms of the swimming angles in cases 1 and 2 . There
were no di ffe rences in e i ther the ranges of va lues between cases 1 and 2 or
the i r s tandard devia t ions (22 .1 and 21 .7 respect ive ly) . In cont ras t , the mean
swimming angles d i ffered b y more than 5° (case 1 : 11 .5° , case 2 : 16 .5°) .
Ti l t -averaged TS
The values of TS were ca lcu la ted wi th respect to f i sh t i l t -angle
d is t r ibu t ion (PDF) , which was used for two values (case 1 and case 2) a t two
12
f requencies (38 and 120 kHz) . TS values are p lo t ted in Fig. 6as funct ions of
f i sh SL on a logar i thmic sca le . Ranges of avgTS , and the equa t ions of the
regress ion l ines shown in Fig. 6 ( the TS- length equat ion) , a re shown in
Table 1 . In genera l , the resu l t s of TS can be expressed in terms of the body
length L us ing the fo l lowing equat ion:
bLlogmTS 10 +=
Where m and L a re constan ts for a given species . This equat ion has been
a genera l l y accepted descr ip t ion of the way in which mean TS depends on
f i sh length [18] . The s lope m and in tercept b can be es t imated by
l inear l y regress ing the TS on log L . In thi s s tudy, the regress ion l ine was
f i t t ed to the es t imated data us ing cases 1 and 2 . The resu l ts were
107.4Llog60.9TS 10avg −= (R2 =0.93) for case 1 , and 107.5Llog60.1TS 10avg −=
(R2 =0.92) for case 2 a t 120 kHz . The mean avgTS was -79 .4 for case 1 and
-79 .9 dB for case 2 . The d i fference in each case was 0 .98 dB (SL: 22 .0 mm)
a t the maximum. Analys i s of covar iance (ANCOVA) was used to tes t for
d i fferences in the regress ion l ines . The s lopes and intercepts of the
re la t ionships d id not d i ffer among cases (p < 0 .05) . avgTS a t 38 kHz was
weak because L was too smal l for a wavelength .
Discuss ion
In th i s s tudy, we used indiv iduals wi th to ta l l engths of 28 .3 to 45 .4 mm
for the observat ion exper iment . The swimming angles of these indiv iduals
t ended to be head-up for each speed of f low. However, the re la t ionship
between swimming angle and bod y length was not c lea r. We conjec ture tha t
(4
13
the swimming angle of the larvae d i ffers b y bod y length . Hunter [19]
showed that the swimming speed of nor thern anchovy (Engraul i s mordax )
l a rvae obvious ly changed wi th increases in body length . Addi t ional ly, Bat t y
[20] noted tha t the speed of l a rval her r ing increased wi th length . Thus , the
swimming ab i l i t i es and swimming angles of l a rvae would a l so change wi th
increases in body length . In juveni le s tage , the swimming angle was a lmost
0° and there was no inf luence of f low speed on the mean swimming angle
(1–5 cm s - 1 four s teps , one wa y ANOVA, P>0.05) [ Ito et a l . , unpubl i shed] .
Thus , the d i fferences in swimming angle between larvae and juveni les are
l ike ly caused b y body growth . Blax ter e t a l . [21] , for example , repor ted tha t
metamorphos is en ta i l s the development of f ins , resu l t ing in improved
swimming ab i l i t y. Therefore , as the body grows , the swimming angle wi l l be
c lose to 0° . However, the mean swimming angle of ind iv iduals in th i s s tudy
would not be changed b y bod y growth because larval anchovy complete
metamorphos is a t about 40 mm in length . Ind iv iduals measur ing 20 mm or
less in l ength may swim at var ious swimming angles . Therefore , an
exper iment examining the swimming angles of smal ler f i sh i s necessar y to
improve the prec i s ion of the TS. However, our resu l t s sugges t tha t the
inf luence of the TS of smal l ind iv iduals i s low because the TS of the larges t
indiv idual was s igni f icant compared wi th tha t o f the smal les t ind ividual due
to changes in f i sh t i l t angles
We sugges t tha t swimming angle gives larval anchovy the ab i l i t y to swim
agains t the f low and i s de termined by body length . In thi s s tudy, bod y
lengths ranged f rom 28 .3 to 45 .4 mm and f rom 36 .4 to 41 .3 mm in January
and December, r espect ive ly. The mean swimming angles were 13 .6° and
14
9 .6° for each per iod ( t - t es t P < 0 .05) . Thus , the swimming angle obvious ly
changed wi th changes in the range of body lengths . However, wi th the
except ion of the data ga the red a t 5 cm s - 1 , these values were not
s ign i f icant ly d i fferen t be tween per iods . In ind iv iduals wi th smal l body
lengths , the swimming angle re l a tes to the f low. In th is s tudy, the f low a t 5
cm s - 1 a ffec ted the mean swimming angle , and the avgTS was ca lcu la ted
using the PDF for two cases (1–3 cm s - 1 , 5 cm s - 1 ) . However, the regress ion
l ines of the ca lcu la ted avgTS fo r the two cases d id not d i ffer s ign i f icant ly
(Table 1) . The Japanese anchov y i s widely d i s t r ibu ted in the Nor thwes t
Paci f ic under var ious envi ronmenta l condi t ions , and swimming angles under
these condi t ions have not ye t been c lar i f i ed . Therefore , in the fu ture ,
swimming behavio r must be observed under var ying speeds of f low.
Addi t ional ly, i t i s necessar y to cont inuous measur ing the f low speed dur ing
the exper iment per iod .
In th i s s tudy, the TS of l a rval anchov y c lear ly depended on changes in the
swimming angle , which was in turn determined by body length . In a
prev ious s tudy, the swimming angle of pe lagic f i sh was f ixed a t the TS of
l a rvae because the swimming angle of t he larvae was not known. As a resu l t ,
the avgTS wi l l become the fac tor of the er ror when the abundance was
es t imated . Therefore , re l iab le swimming data mus t be used to ca lcu la te
avgTS fo r use in acous t ic surveys . Addi t ional ly, th i s s tudy shows tha t when
the swimming angle has a l a rge s tandard devia t ion i t has l i t t l e in f luence on
avgTS (Fig.6) .
There a re a l so some addi t ional parameters tha t a ffec t TS . Mikami et a l .
[22] t r i ed to measu re the densi ty ( g ) and sound-speed cont ras t ( h ) o f E.
15
paci f i ca , which has no swim-bladder, and thus no a i r bubb les in the body,
be tween spr ing and autumn. Thei r resu l t s showed that the maximum TS
values for E. paci f i ca , ca lcu la ted f rom a theore t ica l s ca t ter ing model ,
changed by about 5 dB f rom spr ing to au tumn. Fur thermore , Matsukura et a l .
[23] sugges ted tha t g and h a re a ffec ted by seasonal changes in l ip id
prof i les . In th i s s tudy, the larval anchovy g and h va lues were f ixed a t
1 .068 and 1 .037 , respect ive ly. However, these values mus t be changed under
varying condi t ions ( i . e . , season , area) , and we must fur ther examine such
parameters in the fu ture .
As ment ioned above, these fac tors are impor tan t when a t tempt ing to
es t imate the target s t rength (TS) of l a rval J apanese anchovy. In pa r t icu lar,
the inf luence of swimming angle was h igh . This i s the f i r s t paper to provide
swimming angle dat a . The avgTS g iven in Fig. 6 are recommended for use in
acous t ic surve ys of f i shing ground areas . Addi t ional ly, schools of the larvae
of th i s species can be di s t inguished f rom those of other species using the
volume back-scat ter ing s t rength d i ffe rence method.
Acknowledgements
We thank the capt a ins and crew of Kanagasaki Maru No. 8 for the i r
coopera t ion in co l lec t ing specimens . We a l so thank Yukio Ueta , Keisuke
Mori Fi sher ies Research Ins t i tute , Tokushima Agr icu l ture and the Fores t r y
and Fisher ies Technology Suppor t Center for the i r suppor t in co l lec t ing
specimens . This s tudy was suppor ted in par t by the Fisher ies Agenc y o f
Japan under the pro jec t “Research and development pro jec t s for appl ica t ion
16
in promot ing new pol ic y of Agr icu l ture Fores t ry and Fisher ies .” We thank
thi s ins t i tut ion for the i r support .
References
1 . Annual S ta t i s t i cs on Fisher y and Aquacul ture Product ion 2006,
Sta t i s t i cs Depar tment , Minis t ry of Agr icu l ture 2006, Fores t r y and
Fisher ies , Tokyo. 2008.
2 . Saiura K, Takeda Y (2001) Spr ing f i sh ing ground fo rmat ion of anchov y,
Engraul i s Japonica l a rvae in 1999 and 2000 in the Ki i Channel .
Fi sher ies Bio logy and Oceanograph y in the Kuroshio 2 : 109-118 ( in
Japanese) .
3 . Morioka S (2006) Tr ia l o f spr ing sh i rasu (Engraul i s japonicus l a rvae)
f i shery forecas t in Ki i Channel b y us ing mul t ip le regress ion analys i s .
Fi sher ies Bio logy and Oceanograph y in the Kuroshio 7 : 21-27 ( in
Japanese) .
4 . Mitani I (1988) The b iologica l s tud ies on the larvae of J apanese anchov y,
Engraul i s japonica HOTTUYN, In Sagami Bay. PhD Thes i s . Univers i t y
of Hokkaido , Hokkaido ( in J apanese) .
5 . Honda S (2004) Abundance es t imat ion of the young cohor t s of the
Japanese Paci f ic popula t ion of wal le ye pol lock (Theragra
chalcogramma ) by acous t ic surve ys . PhD thes is . Univers i t y of Hokkaido ,
Hokkaido ( In J apanese)
17
6 . Yasuma H (2004) S tudies on the acous t ica l b iomass es t imat ion of
myctophid f ishes . PhD thes i s . Univers i ty o f Tokyo, Tokyo ( In J apanese)
7 . Zhao X, Wang Y, Dai F (2008) Depth-dependent t a rget s t rength of
anchovy (Engraul i s japonicus ) measured in s i tu . ICES J Mar Sci 65 :
882-888.
8 . Foote KG, Tra ynor JJ (1988) Compar i son of wal leye pol lock
target -s t rength es t imates de termined f rom in s i tu measurements and
ca lcu la t ions based on swimbladder form. J Acous t Soc Am 83(1) : 9 -17 .
9 . Simmonds J , MacKennan D (2005) Fisher ies Acous t ics 2nd edn .
Blackwel l Publ i sh ing, Oxford .
10 . Uotani I (1973) Diurnal changes of gas b ladder and behavior of
post larval anchov y and o ther r e la ted species . Bul l J ap Soc Sci Fi sh 39:
867-876 ( in J apanese wi th Engl i sh abst rac t ) .
11 . Miyashi ta K (2003) . Diurnal changes in the acous t ic- f requenc y
characte r i s t i cs of Japanese anchov y (Engraul i s japonicus ) pos t - larvae
“sh i rasu” infer red f rom theore t ica l sca t ter ing models . ICES J Mar Sci
60 : 532-537.
12 . Stanton TK, Chu D, Wiebe PH (1993) Sound sca t ter ing b y severa l
zooplankton groups . Ⅱ . Scat ter ing models . J Acous t Soc Am
103:236-253.
13 . Chu D, Foote KG, S tan ton TK (1993) Fur ther analys i s of t a rget s t rength
measurements of Antarc t ic kr i l l a t 38 and 120 kHz: Compar i son wi th
deformed cyl inder model and inference of or ien ta t ion . J Acous t Soc Am
93: 2985-2988.
14 . McGehee D E, O’Drisco l l RL Mart in Traykovdk y LV (1998) Effec t s o f
18
or ien ta t ion on acoust ic sca t ter ing f rom Antarc t ic kr i l l a t 120 kHz.
Deep-Sea Res 45: 1273-1394.
15 . Foote KG (1980a) . Averaging of f i sh ta rget s t r ength funct ions . J Acous t
Soc Am 67(2) : 504-515.
16 . Foote KG (1990) . Speed of sound in Euphaus ia superba . J Acous t Soc
Am 87: 1405-1408.
17 . Greenlaw CF (1977) . Backscat ter ing spect ra of prese rved zooplankton . J
Acous t Soc Am 62: 44-52 .
18 . Foote KG (1979a) . On represent ing the length dependence of acous t ic
t a rget s t rengths of f i sh . J Fi sh Res Board Can 36(12) : 1490-1496.
19 . Hunter J R (1972) Swimming and feed ing behavior of l a rval anchovy,
Engrauis mordax . F i sh Bul l 70 : 821-838.
20 . Bat ty R S (1984) . Development of swimming movements and
muscula ture of l a rval her r ing (Clupea harengus ) . J exp Bio l 110:
217-229.
21 . Blaxter J H S , Sta ines M E (1971) . Food searching poten t ia l in mar ine
f i sh larvae . Four th European mar ine b io logica l sympos ium. Cambridge
Univers i t y Press , Cambridge , England .
22 . Mikami H, Mukai T , Iida K (2000) . Measurements of dens i ty and sound
speed cont ras t s for es t imat ing kr i l l t a rget s t rength us ing theore t ica l
sca t ter ing models . Nippon Suisan Gakkaishi 66(4) 682-689 ( in J apanese
wi th Engl i sh abs t rac t ) .
23 . Matsukura R , Yasuma H, Murase H, Yonezak S , Funamoto T, Honda S ,
Miyashi ta K (2009) Measurement of dens i ty cont ras t and sound-speed
cont ras t for t a rge t s t rength es t imat ion of Neocalanus copepods
19
(Neocalanus cr i s ta tus and Neocalanus p lumchrus ) in the Nor th Paci f ic
Ocean . Fi sh Sci 75 : 1377-1387.
20
Table 1 Equat ions for the l inear regress ion in Fig. 5 and mean target
s t rength a t each f requency; y rep resent s TS (dB) , and x represents the log of
s tandard length in cent imeters .
21
Fig. 1 Schemat ic d iagram of the exper imenta l t ank used for the penned
Japanese anchov y larvae .
F ig. 2 T ypical example of a photograph obta ined f rom the observat ion
exper iment . The upper panel shows a dorsa l image and the lower shows a
la tera l image; θ ind ica tes the swimming angle .
F ig. 3 T ypical var i a t ions in the target s t rength (TS) of minimum (a) and
max imum larval l ength (b) as a funct ion of f i sh p i tch angle , es t imated by
the d is tor ted-wave Born approx imat ion (DWBA) model a t 38 (dot ted l ines )
and 120 (bold l ines ) kHz . Pos i t ive angles are head-up and negat ive angles
are head-down.
Fig. 4 Box p lot s of the swimming angles of l a rval J apanese anchovy. Middle
squares ind ica te the mean; whiskers show the s tandard devia t ion . Numbers
in parentheses ind ica te the max imum and minimum values .
F ig. 5 Frequenc y d i s t r ibu t ions of the swimming angles o f l a rval J apanese
anchovy under f low ra tes of 1–3 cm s - 1 (a) and 5 cm s - 1 (b) .
F ig. 6 Rela t ionship between TS and log of s tandard length (L in mm) each
case . TS i s 38 kHz for the upper panel and 120 kHz for the lower panel . The
c i rc le and cross ind ica te ca lcu la ted TSs for cases 1 and 2 , respect ive ly.
Equat ions for the regress ion l ines in each panel are given in Table 1 .
Table 1 Equations for the linear regression in Fig. 5 and mean target strength at each frequency; y represents TS (dB), and x represents the log of standard length in centimeters.
Frequency (kHz)
Mean TSave (dB)
y = p x + q n p q r2 S.E.
Case 1 (1-3cm/sec) 1955
38 -94.2 67.3 -125.3 0.93 1.5 120 -79.4 60.1 -107.4 0.92 1.4
Case 2 (5cm/sec) 682
38 -94.5 66.2 -125.1 0.92 1.5 120 -79.9 60.1 -107.5 0.92 1.4
Fig.1
116cm
77cm
97.5cm
Video camera
Side strut
45cm10cm
70cm
13cm
116cm
77cm
97.5cm
Video camera
Side strut
45cm10cm
70cm
13cm
Fig.2
Dorsal image Target
Lateral image
Target
Non-target 0° +θ -θ
Fig.3
20
-140
-120
-100
-80
-60
-90 -60 -30 0 30 60 90
Tilt angle (degree)
Targ
et
stre
ngt
h (
dB
)
-120
-100
-80
-60
-90 -60 -30 0 30 60 90
Tilt angle (degree)
Targ
et
stre
ngt
h (
dB
)
SL: 18.0mm
SL: 35.7mm
:38 kHz
:120 kHz
:38 kHz
:120 kHz
(a)
(b)
Fig.4
(78.0,-53.1) (69.4,-77.0) (73.4, -70.5) (69.8, -62.1)
16.5±21.7 12.3±26.2 11.9±19.6 9.4±22.9
n = 423 n = 1080 n = 452 n = 682
Fig.5
0
10
20
30
-90 -60 -30 0 30 60 90
Tilt angle (degree)Fre
quency (
%)
0
10
20
30
-90 -60 -30 0 30 60 90
Tilt angle (degree)
Fre
quency (
%)
1-3cm/sec n = 1955 Mean: 11.5 S.D.: 22.1
(a) (b)
5cm/sec n = 682 Mean: 16.5 S.D.: 22.0
Fig.6
-110
-100
-90
-80
0.2 0.3 0.4 0.5 0.6
Log SL (cm)
Tar
get
stre
ngt
h (
dB)
-100
-90
-80
-70
0.2 0.3 0.4 0.5 0.6
Log SL (cm)
Tar
get
stre
ngt
h (
dB)
120 kHz
38 kHz
: Case 1 : Case 2
: Case 1 : Case 2