'.

J. Appl. Ichthyol. 1 (1989) 12-17@ 1989 Verlag Paul Parey, Hamburg und BerlinISSN 0175-8659

Received: February 29, 1988Accepted: August 15,1988

Hook Selectivity ofYellowfin tuna (Thunnus albacares) caught offDarigayos Cove, La Union, Philippines

By E. CORTEZ-ZARAGOZA I, P. DALZELL2 and D. PAULY

I Philippine Council for Aquatic and Marine Research and Development, Los Banos, Laguna, Philip-pines

2 International Center for Living Aquatic, Resources Management, Makati, Metro Manila, Philippines

Summary

Hook selection curve for yellowfin tuna Thunnus albacares caught off Darigayos Cove, NonheroLuzon, Philippines were estimated using the BARANOV/HOLTmethod. Estimated optimum capturelength (Lop,)Increased with hook size, as ilid the selection range. Overall selection was found to be veryasymmetncal, with small yellowfins n cm below Lop,having a much lower probability of capture thanlarger yellowfin n cm above Lop,.

Zusammenfassung

Hakenselektivitiit bei Fang van Gelbf/ossenthunen (Thunnus albacares) vor Darigayos Cove, im Nor-den von Luzon, Philippinen

Selektionskurven wurden mittels der BARANov/HoLT-Methode fUr Haken bestirnmt, die bei derFischerei des Gelbflossenthuns Thunnus albacares bei Darigayos Cove, im Norden von Luzon, Philip-pinen, Verwendung finden. Die geschiitzte optimale Fanglange (Lop,) und die Selektionsspanne nehrnenmit der GroBe der Haken zu. Die Selektionskurven sind asymmetnsch, und Gelbflossenthune, die n cmkleiner sind als Lop''' werden rnit einer vie! geringeren Wahrscheinlichkeit gefangen als Gelbflossen-thune, die n ern groRer sind als Lop,.

Resume

Silectiviti d'hamerons dans la piche d'albacore (Thunnus albacares) au large de Darigayos, Luzon duNord, Philippines

Les courbes de selectivite des ham~ons utilises pour la peche d'albacore Thunnus albacares au large deDarigayos Cove, Luzon du Nord, Philippines, ont ete estimees, a l'aide de la methode de BARANOV/HOLT. La longueur optimale de prise (Lop,) et I'ecan de selectivite augrnentent avec la taille de l'hame-,<on. Les courbes de selectivite sont tres asymetriques et les petitS albacores n cm en d~a de Lop, ont uneprobabilite de prise beaucoup plus faible que les gros albacores n cm au dela de Lop,.

,

Table1. Annual landings of yellowfin tuna inthe Philippines and contribution to the total

marine landings.

Introduction

Recent landings (1980-1985) of yellowfmtuna (Thunnus albacares) in the Philippineshave ranged between 48,000 and 64,000 ton-nes, accounting for 4.5% of total marinelandings (Table 1). Small fish are consumedfresh or canned, whilst fish larger than 1 mare greatly esteemed for sashimi or raw fish.These large tunas are particularly abundantin the northern waters of the Celebes Seaand a large handline fishery was established

0175-8659/89/0501-0012$ 02.50/0U.S. Copyright Clearance Center Code Statement:

..

..

.....

----

Year Yellowftn Percentof totallandings marinelandings(tOnnes)

1980 48023 4.231981 56176 4.661982 51922 4.211983 62036 4.811984 58924 4.521985 64293 4.49

11~. . .~~,.

::

..

'-"

~ ,",:";i:;.'~..::: :::,J...~~::...1::..."S'..<'.."' ,:.:...;..:' c__

Hook Selectivity ofYellowfm tuna

along the southeastern coast of Mindanao (GANADENand ALl 1983).Details on the seasonalabundance and morphometrics of T.albacares from nonhern Luzon have been presented byCORTEZ-ZARAGOZA et al. (1987).

A recent analysis by RALsTON(1982) of a handline fishery in Hawaii for percoid fishessuggests that selectivity is strongest against smaller fish. We present data which suggest thatthis may also be the case for yellowfin tuna caught by handline.

Materials and methods

A detailed description of the Darigayos Cove (Fig. 1) handline fishery and the methods ofdata collection is given in CORTEZ-ZARAGOZA(1983). All handline fishermen operate fromsmall boats or bancas with an average length of 10m, with an average gross tonnage (GT) of0.3 which are powered by 16 hp Briggs and Stratton gasoline engines. Observations oncatch, fishing effon, fork length and weight of catch were made between May andNovember 1981. Fishing effon was expressed as the number of line days per month or theproduct of number of days at sea and the number of handlines per vessel. The hook sizes usedin the fishery were 1.2, 1.3, 1.4, 1.5, 1.6, 2.4, 2.7 and 2.9 cm. Hook size refers specifically tothe gap between hook point and shank (Fig. 2).

eye

\)..~ shonk

point

crook

Fig. 2. Parts of a hook; "size" as defined here isthe gap

Fig. 1. Map of sampling area

The length distribution of each daily yellowfin catch by the fishing fleet was recorded byhook size. GULLAND(1983) suggested that the selection effects of different hook size in fishcatches can be compared by a modification of the BARANovlHoLTmodel for gill net selec-tion (BARANOV1914; HOLT 1963), whose properties and fitting are explained in detail inPAULY(1984). Selection curves of handline catches were thus estimated by comparing thelength-frequency distributions of yellowfin caught by two adjacent hook sizes.

Given the smaller hook size A and the larger hook size B, it is possible to conven thecatch data into a linear regression of catch ratios on length of the form (y = a + bx), where:

. .

13

(1)

14 E. Cortez-Zaragoza, P.Dalzell and D. Pauly

where CAis catch by length class for hook size A and CBis the corresponding catch for hooksize B, and where

x = Lj(classmidlength) (2)

The intercept (a) and slope (b) of this regressionare used to estimateoptimum length forhook sizeA from:

and

-2aA

LA= b (A + B)(3)

-2aB

LB= b(A+B)

The standard deviation of either selection curve (SA=SB = S) is given by:

""

(3b)

S= 2a(A-B)

b2 (A + B)(4)

1.3i1.4cmhooks

4

o

1.lIil.6emhooks

-I 2.4i2.7cmhooks

-2

L I ,..:3L,J-l4 4.8 4.9 11.0 5.1

Loge fork length (cm)

2.7i2.9cmhooks

Fig. 3. Log. catch ratio vs. log.fork length for catches of yellow-fin tuna for different hook sizecombinations

14.9

fork

I I

11.0 11.1

length (cm)

3I/) I rZ =0. 2c...

.cE 0c0

- 1Q)

C> -20...J

-31".,!3

3I/)

t

rl =0 0.8116-a :-211.111C...b : 6.864

.c0

0(;0 - IQ)

C>0 -2...J

-3 L....:3

3I/) rl: 0.891.g a :-101.04

b: 20.290.cE 0c0

-IQ)

C>0 -2...J

"':3""4.8

Loge

3rl = 0.9112

2 I- a :-211.339b : 6.801

"I0

1.2i'.3cm -Ihooks

-2

L--....I -3L4 3

3

I rl:2

,..

. '. ....-........ ..

Hook Selectivity of Yellowfin tuna 15

When LA,LBand 5 have been estimated, the probability of capture (P) at a given length (Lj) isgiven for a hook size A by:

P =exp (_ (Lj- LA)2

)A 252

and for hook size B by:

P =exp ( _ (Lj - LB)2)B 252

In its original fonn, this selection model yields symmetrical curves about the optimumcapture length. However, asymmetrical curves, where selection is less intense at larger sizes,can be fined by replacing length by lo~ (length) in the above equations.

(5)

Results

Plots of the logarithms of catch ratios versus the logarithms of the corresponding midlengthsfor the different hook size combination are shown in Fig. 3, based on the data in Table 2.Apart from one instance, the data were well described by linear functions. The exception wasthe comparison of 1.4 and 1.5 cm hooks, which could not be perfonned. The optimumcapture lengths estimated from each pair of hook sizes are given in Table 3.

A curve of the fonn y =axbwas fined to the scaner of optimum lengths versus hook size(Fig. 4). A similar curve was also fitted to the scatter of the selection range (defined here asone standard deviation on either side of the optimum length) versus optimum length (Fig. 5).

Table 2. Catch by Length of Different Hook Sizes to Estimate their Selectivity for Thunnus albaca-res, off Darigayos Cove, La Union

Midlength of Hook size(em)siz,egroup 1.2 1.3 1.4 1.5 1.6 2.4 2.7 2.9mem

22 124 1026 145 1730 170 82 8 96 734 60 91 20 160 2938 50 143 67 125 8842 15 60 91 190 45546 1 8 24 45 16050 1 6 23 3 2554 - 1 8 5 2558 - 1 5 4 1562 - - 1 1

I ....... . -", ." ...."-.

126 - - - - - 26 3...... "' .

130 - - - - - 53 11134 - - - - - 31 18 2138 - - - - - 5 8 3142 - - - - - 3 7 4146 - - - - - 2 10 15150 - - - - - 1 7 25154 - - - - - - 4 17158 - - - - - - 1 6162 - - - - - - 1 4166

'.

. ...........-

16 E. Cortez-Zaragoza, P.Dalzell and D. Pauly

200

CD~Ea. u 100o .-u...JI.LE~::>.§iio

3

,2. 0.952. .20.94b' 1.935

o2

Hook size (em)

Fig. 4. Optimum capture length vs. hook size foryelIowfin tUna

Optimum length (FL; cn\)

Fig. 5. Selection range vs. optimum length for yel-lowfin tuna

Discussion

The resultS presented here document the selective effects of different hook sizes on yellowfintuna. Similar observations of hook selectivity on skipjack were made by TANDOG-EDRALINetal. (1987). Larger hooks catch larger fish, but the standard deviations of the selection curves

increase as the hook size increases, sugge-sting that bigger hooks capture a wider rangeof lengths. RALSTON(1982) suggested thathook selection conformed to a flat-toppedsigmoidal curve similar to that observed fortrawls. In RALSTON'Sanalysis, the largesthook used was 71% greater than the smallest,whilst in this study, the difference was 240%. .

KOIKE et al. (1968) and KANDA et a1. (1978)

used series of hooks in which the largest sizeswere 215% and 115%, respectively, largerthan the smallest of sizes. Both studies repor-ted shifts in the size composition on thecatch. SAETERSDAL(1963) reported a changein the selective characteristics of fishinghooks which differed in size by 76%.

The schematic forms of different proposed selection curves are shown in Fig. 6. Curves Aand B correspond to the BARANOV/HoLTmodel and the flat-topped sigmoidal curve, respec-tively. The latter may include, as suggested by S. RALSTON(pers. comm.), a descendinglimb, albeit at very large sizes (dotted in Fig. 6B).

An appreciation of hook selectivity (or lack of it) is important for two reasons. First,where length data are used for stock assessment purposes in a fishery employing a wide rangeof hook sizes, adjustmentS to the data based on probabilities of capture may be necessary.Simply pooling all data in the hope that selectivity effectS may be negligible or will cancelout, may introduce biases into estimates of groWth and mortality parameters. Secondly,hook size may be used to regulate minimum capture sizes from a fishery in the same way thatminimum mesh sizes are employed in net fisheries. We suggest that further investigation ofthis type should be undertaken, especially on the large handline fisheries for tunas aroundthe southern coast of Mindanao.

Table 3. Hook size and predicted optimumlength of yelIowfintuna.

Hook size(em)

aOptimum capture

length (em)

1.21.31.41.51.62.42.72.9

29.437.647.636.546.4

102.6152.7173.8

a Where there was more than one optimumlength for a hook size means are given.

. . . . . ~ . ".,".-..

.. ,.

..

60

40 l r2. 0.950Q . 0.975b . 0.754

00c

c20

uCDIiIf)

OV ,50 100 150 200

Hook Selectivity ofYe//owfin tuna 17

;>.Q)-...._::J=.....c c-oo.cuo

G:'Q

Lop,

o

.... ........, '..

Length Length

Fig. 6. Different types of possible selection curves for fish hooks. Type A is the BARANOV/HoLT model,Type B was proposed by RALSTON(1982)

References

BARANOV, F. I., 1914: The CaptUre of Fish by GilInets. Mater. Poznsniyu Russ. Rybolovsta 3 (6):56-99. In Russian; translated version in 1976 Selected works on fishing gear. Vol. I. CommercialFishing Techniques. Israel Program for Scientific Translations, Jerusalem.

CORTEZ-ZARAGOZA, E. c., 1983: Morphometries and relative abundance of tUnas (Perciformes: Scom-bridae) caught in Darigayos Cove, La Union. M. Sc. thesis, Univ. of the Philippines: 85 p.

CORTEZ-ZARAGOZA, E.; DALZELL, P.; PAULYD., 1987: Seasonal abundance, morphometries and hookselectivity of yellowfin (Thunnus albacares) caught off Darigayos Cove, La Union, Philippines.Presented at the FAO/IPTP Second Meeting of Tuna Research Groups in Southeast Asia, Manila,25-28 August 1987. mimeo.

GANADEN, R.A.; Au, S.M.P. Jr., 1983: Observations on the handline fisheries of South Cotabato.Fisheries Newsletter (Philippines). 12 (1): 24-27.

GULLAND, J. A., 1983: Fish stock assessment: a manual of basic method. FAO/Wiley Series on Foodand AgricultUre, Vol. 1. Wiley Interscience, Chichester and New York: 233 p.

HOLT, S.]., 1963: A method for determining gear selectivity and its application. Spec. Publ. Int.Comm. Northwest Ad. 5: 106-115.

ICLARM Contribution No. 489; based on a report presented at the Second Meeting of Tuna ResearchGroups in the Southeast Asian Region, Manila, 25-28 August 1987.

KANDA, K.A.; KOIKE, A.; TAKEUCHI, S.; OGURA, M., 1978: Selectivity of the hook for mackerel,Scomber japonicus HouttUyn, pole fishing. Tokyo Univ. Fish. 64: 109-114.

KOIKE, A,; TAKEUCHI, S.; OGURA, M.; KANDA, K. and ARIHARA, c., 1968: Selection curve of the hookof long line. J. Tokyo Univ. Fish. 55, 77-82.

PAULY, D., 1984: Fish population dynamics in tropical waters: a manual for use with programmablecalculators. ICLARM StUdies and Reviews 8, 325 p. International Center for Living Aquatic

Resources Management, Manila, Philippines.RALSTON, S., 1982: Influence of hook size in the Hawaiian deepsea handline fishery. Can J. Fish.

Aquat. Sci. 39 (9): 1297-1302.SAETERSDAL,G., 1963: Selectivity of long lines. Spec. Publ. Int. Comm. Northwest. At!. 5: 106-115.TANDOG-EDRAUN, D. D.; CORTEZ-ZARAGOZA, E. C.; DALZELL, P.; PAULY, D., 1987: Some aspects of

the biology and population dynamics of skipjack (Katsuwonus pe/amis) in Philippine waters. Pre-sented at the FAO/IPTP meeting of Tuna Research Groups in the Southeast Asian Region, Manila,25-28 August 1987. mimeo.

Authors' address: E. CORTEZ-ZARAGOZA, Philippine Council for Aquatic and Marine Research andDevelopment, Los Banos, Laguna, PhilippinesP. DALZELL and D. PAULY, International Center for Living Aquatic, ResourcesManagement, MC P.O. Box 1501, Makati, Metro Manila, Philippines(Send reprint requests to D. Pauly)

--

,.