., ..

NAFO ScL Coun. Slu(hes. 9: 143-154

p'

Population Dynamics of Short-lived Species,with Emphasis on Squids*

Daniel Pauly .

International Center for Living Aquatic Resources Management (ICLARM)MCC P. O. Box 1501, Makati, Metro Manila, Philippines

Abstract

The leatures 01 short-lived Iishes and invertebrates relevant to their population dynamics and exploitation ere reviewed. withemphasis on methodological approaches. which could lead to improved management 01 squid stocks. Several models are presentedwhich offer the possibility 01 reducing what presently appear to be qualitative differences In ttie population dynamics olllshes andsquids to quantitative differences. and, hence. to render squids amenable to the kind of comparative studies which. for fish, haverendered preliminary stock assessments possible even when data are scarce.

Introduction

This invited paper which deals with the populationdynamics of short-lived species touches upon a sub-ject which has not received much attention in thefisheries literature.The majorreason for this is that, asfar as fishes are concerned, the commercially-Important short-lived species occur predominantly Intropical waters, where relatively little research In popu-lation dynamics is conducted. Thus, many stock-assessment specialists. who rely predominantly on'virtual population analysis (VPA). found themselvesconceptually and methodologically unprepared whenthey'were confronted In the mid-1970's with having to .assess the newly-explolted squid stocks and to derivemanagement measures for them. Indeed. investiga-tions of short-Jived species. such as the squids. requirea return to the basic concept of Russell (1931) whichstates that

B2=B, + (Ro+G")- (V+MO) (1)

or th'at future biomass (B2)of a stock with no emigra-tion or Immigration is a f~nction of its original biomass(B,). plus weight of recruits (A") and growth of therecruited animals (GO). minus the catch (V) and theweight of animals dying of natural causes (M").Equa-tion (1) Is an obvious basis for this contribution, whichIs structured as a reviewof methods to measure orpredict each of the parameters Inpopulationsofshort-lived animals,'particularlysquids.

Growth

What appear to be successful attempts to agesquids by means of daily ringsintheir statoliths havebeen reported by Spratt (1978)for Loligo opalescensand by Kristensen(1980)for Gonatuslabrlcil. Atheoryexists (Lutzand Rhoads, 1977) which seeks to explain

. ICLARM Contribution No, 207.---

the origin and nature of these rings, with the Implica-tion that they should be as ubiquitous In molluscs asdally rings are in the otoliths of fishes. Indeed, thistheory could, with suitable generalization. also explainthe occurrence of daily rings in fish otoliths, since thebasic mechanisms In both fishes and molluscs Involveshort-term anaerobiosis duri~g periods of activity(Pauly, 1981, 1982a). .

Ageing of squids by means of dally rings in stato-liths appears to Involve too much work, even in short-lived forms, to become a routinely-applied method, .especially if the generalization that squid groy.'thrateisextremelyvariablewithin populationsprovestobetrue(Caddy, 1983).This obviously leads, for the lack of.abetter alternative, .to often murky methodologies thatare used to Infer growth rates and growth curves fromsize (especially length) frequencies. Manyresearchers, InvolvedIn studying fish,squid andotherInvertebrates, unfortunately believe that Inferringgrowth from length frequency data simplyconsists ofvisually identifying "broods" from polymodal length-frequency distributions, tracing by eye whatever linesone feels like tracing, and representing the results as"growth" rates .(Fig. 1).

As far as research on squid growth Is concerned,the chief result of this belief has been growth parame-ters that are often mutually Inconsistent and representan array of contradictory statements about the growthof squids, which is perceived to be "linear", "exponen-tial", "asymptotic" or "oscillating" (Fig. 2) (see ~Isotables 6 and 7 of Lange and Sissenwine, 1983). Untilrecently, similar confusion was characteristic ofshrimps, but, upon close examination, many of theseseemingly conflicting growth patterns could beresolved In a basic growth pattern of the asymptotictype (as in fish), with superimposed seasonal growthoscillations (Garcia and LeReste, 1981). This can be

144' Sci. Council Studies, No.9, 1985

40

l20>.uc..:)a~ 40~ n =18

, , , , ,Winter 1977

20

Io 5 10 15 20

Manlle tength (cm)

Ag. 1. Original tegend to thi, graph, redrawn from Hlxon.t a/ (19S1)reads as foUows: NSize frequency distribution of _ females ofLDligo pealel obtained from six seasonal collections In 1976

and 1977. Mean lengths of well defined modes designated bya solid anow. Dashed arrows indicate less cenain meanmodal lengths est/mated by Ihe probability paper method.Unes drawn between modes depict Increases In manUelength between successive seasons. Solid lines Indlcat.

growth between well defined modes: dashed lines designategrowth based on less certain modes.- Note lack of explicilcriteria In seleelion of modes !Inked by assumed growth (seealso Fig. 6), and erroneous tracing 01 lines, which should linkthe bases of the modal classes considered rather than theirpeakS.

expressed by a modified von Bertalanfly model of theform _

where L._is length (cm) at age t, Leo Is asymptoticlength, Kand toare the same constants as Inthe normalvon Bertalanfty model, C is a constant expressing theamplitude of the growth oscillations, and t. is the onset

Time

Fig. 2. Growth curves lor five loliglnid species (alter Hixon. MS1980).A,lInear growth of Loligoopa/ncens (Fields, 19651:B,asymptotic growth 01L.pea/el (Verrill,1881): C, cycllcgrowttl01 L. vulgaris (T/nbergen end Verwey, 1945): D, exponentialgrowth 01 L. vulgar/if (Mengold, 1963) .nd L. pea/el(Summers, 1971); E, sigmoid growth of L. pealal, L. plei andLo/liguncula brevis (Hixon, MS 1980).

of sinusoid oscillations with respect to t = o.Thus,thekey hypothesis here Is that, within populations, whatappear to be differences In growth characteristics areartefacts due to changes In birth date (Fig. 3). SuchInterpretation. should be considered Ifonly because itoffers the possibility of explaining what appear to bespecific (and hence hard to study) characteristics ofsquids as a general feature of aquatic polkllothermsthat are explosed to seasonal temperature oscillations(Fig. 4). .

Another advantage of using a standard growthmodel, such as the above-mentioned seasonallze~ van

.Bertalanffy growth equation. to express the results ofgrowth studies Is that it offers the possibility of con-ducting comparative studies among various popula-tions of the same squid species, among various squid

. species (see Table 1) and between squids and fishes(Fig. 5), and hence to use the Information available forsome species within a given genus orfamllyto Ihferthegrowth performance of species for which data are notavailable.

(2)

One approach, which works rather well for:fishesand shrimps. is to rely on the within-group constancyof the parameter t/J,which is defined in base 10 loga-rithms as

t/J =log K + 2/3 log Weo (3)

where Woois the asymptotic whole weight (g) corres-ponding to Leoin equation (2), and the growth coeffi-cient (K) is expressed on an annual basis (Pauly,

,40'- n =366

, ' Spring-summer1977, \, \

20'- L_, ,,

\ ,I, ", .-

-II:z::

401. n=44 \. \. Autumn 1977

20

PAULY: Dynamics of Short-lived Species with Emphasis on Squids 145

20

_ 15E.!:!.~ 10a.c:....J

All curves with:

loo = 20 em. K = 1C = 1. 10= 0

Fig.3.

18 24 30 36Absolute age (months)

Four growth curves drawn Irom a seasonally oscillallng ver-sion 01 the von Benalanffy growth lunction with the same set01 parameters fL K. C. \0) and dillerent values 01 I.. (Notethat these curves closely duplicate a variety 01 shapes as-sumed representative of squid growth. SUbjective length-frequency analysis. whicl1 must rely on the well-separated.younger age-groups. will necessarily overemphasize the ap-parent dillerences In growth and miss the underlying overallgrowth panern modified by temperature-induced oscilla-tions.)

42

1980b; Munro and Pauly, 1983; Pauly and Munro,1984).

For practical purposes, the use of equation (3)consIsts of estimating a mean value of t/J for a givengroup of species from available Wooand K pairs ofvaluesfor the group,and relyingon t/Jto obtain prelimi-nary estimates of K for species (or stocks), for whichgrowth data are not available, from the formula

log K = t/J - 213 log (Woo) (4)

where (Woo)is an estimated value of the asymptoticweight which is derived from maxImum weight (Wmu)for a given stock (e.g. Roper at al., 1984) by therelationship

Woo= WmulO.86 (5)

This Is based on an assumed cubic length-weight rela-tionship and the observation that organisms withasymptotic growth often reach about 95°A.of theirasymptotic length (Taylor, 1958, 1959; 1960; Pauly,1984).

These simple rules, and the systematic applicationof consistent, objective methods for length-frequencyanalysis (Fig. 6), could allevIate rapidly the presentdearth of knowledge on the growth of squids and helpresearchers obtaIn a "feel" forthe growth performanceof squids similar to that which biologists now have forfishes.

Estimation of Mortality and Biomass.

Estimating mortalities of short-lived animals isusually problematic, especially when individual anim-als are difficult to age. One approach, which is becom-ing increasingly popular in studies of fishes and

1.5

1.4,.

.

48

1.3

£ 1.2..

.§ t.1OJ .:: 1.0u..00.9.r:"i 0.8oC.0.7'0III0.6"0

~o.sQ.E 0.4<

0.3

0.2

0.1

.

. .

.J:x

"" Loligopea/~'Gulf of Mexico

Fig.4.

2 3 4 5 6 7 8 9 10 " 12 13 14 15AT (OC)

Relationship between amplllude 01aeasonal growth oscilla-tions for Iishes and squids. aa expressed by parameter C 01equation (2) and the difference between highest and lowestmonthly water temperature In the course 011 year (AT). asadapted from Pauly el al. (1984). (Note that C =0.65 lorL. pea/ai. derived ill Fig. 6. matches with the appropriate valueof AT =7.2.C (from Rivas. 1968) and fits Into the general pat-tern for lishes and shrimps.)

shrimps, is the use of "length-converted catch curves".which are based on length-frequency data wherebylengths are converted to their corresponding (relative)ages by means of a set of growth parameters fLooandK,with a simple correction for non-linearity of growth).Total mortality (Z) is estimated from the descendingright limb .of the curve (Rg. 7A). Length-convertedcatch curves. as In the case of standard age-basedcatch curves, are based on the assumption that allcohorts in the available data experience the same total'mortality (Ricker. 1975; Pauly, 1984). This assumptionis probably not fully applicable In the majority of cases.However. estimating mortality for each cohort separ-ately will often be too demanding of data from thesquid stocks, and the pooling of cohorts will usually bethe only approach toward obtaining reasonable esti-mates of mortality.

The steps toward the construction of length-converted catch curves, such as those in F.ig.7, are ~sfollows:

1. Obtain periodic (e.g. monthly) length-frequencysamples, preferably over a period of 1 year, suchthat a cumulative sample can be derived. This sam-ple should, as far as possible, be representative ofthe average length composition of the stock, suchthat seasonal pulses are evened out. .

------

---------- ---

146 Sci. Council Studies, No.9, 1985

..1\10;

-:L 0.50-'E1\1

;g'iiou.r;ie 0.10C)

2

Log Woo

Fig. 5. Comparison of growth performance of squids (No. HI from Table 1) wllh that 01 a family offast-growlng fishes (Scombridae,tunas and mackerels), a family of short-lived IIshes (Cyprinodontidae, guppies). along-llved redfish (Sabas/as mar/nus). anda family 01 shrimps (Penaeidae) by means of an auxlmetrlc grid. (The fish growth data were documented by Pauly (1980b)and the shrimp data by Pauly eI al. (1984). The grid Indicates that squids are as fast-growing es their pelagic competitors. but'he groWth parameters for data sets 5 and 6 In Table 1 are probably erroneous.)

2. Estimate the "relative" age (tt?that corresponds tothe lower class limit of each length class (Li)fromthe relationship

. 11=-In(1-LJLoo)/K (6)

where In refers to natural logrithms. Here, tt' Iscalled "relative", because, with Z being estimatedfrom the slope of the catch curve, to can beassumed to be zero. Also, seasonal growlh oscilla-tions need not be considered because an "aver-age" sample is analyzed.

Plot the cumulated number of squid Ineach lengthclass (N;)against the corresponding t.'value (seeFig. 7, A-C) and estimate (when possible) Z fromthe descending (linear) 'right limp of the catchcurve, i.e. from the slope of .

In Nj'= a + bt.' (7)

which can be turned into an estimate ot Z by therelationship .

3.

Ibf+ K=Z

When parameter K of the von Bertalanfty growthequation is not known, relative age can be definedas tt'= (age-to) K, i.e. by setting K= 1 in equatlon(6). In this case, the slope of the catch curve yieldsan estimate of (Z/K)-1 (see Fig. 7B, and Pauly,1984). .

Equations (7) and (8) provide estimates of Z whichare extremely close (within less than 1%) to the truevalues of Z, as was ascertained from analysis of con-structed data (Pauly, 1984: P. Sparre, Danish Institute

. of Fisheries, Charlottenlund, pers. comm.). Themethod has been shown to yield remarkably straightcatch curves in a large number of fish and shrimpstocks (Pauly and Ingles, 1981; Pauly, 1982a: Inglesand Pauly, 1984; Pauly at al., 1984).

(8)

This approach was applied to a number of squidsamples during the preparation of this paper, and itled, in many cases, to catch curves that were not linear(e.g. Fig. 7C), thus apparently confirming the viewotJuanico (1983) that "squids are not fish and that few

PAUL Y: Dynamics 01 Short-lived Species with Emphasis on Squids 147

ASample too smalllor ELEFAN I

co....m...

J

5 10 15

Length (em)

n=6 8

I equivalent to 20% frequency

20 0 5 10 .15

Length (em)

20

Reanalyzad length-frequency data of Fig. 1 (Lol/go pea/a; females from Gull of Mexico). with (A) regrouping ofdata into smaller number of length classes; and (8) restructuring of samples. as performed by ELEFAN I program.to identify seasonally-osdllaling growth curves that best explain the data. (The resultant growth curve has theparameters Loo.. 21.5. K .. 0.285 and C .. 0.65.)

prlnctples of fish biology apply to these molluscs".More interestingly, however, these catch curves, sim-ple as they are, may, if applied widely. help to resolvewhether the high postspawning mortalities that arereported for squids are ubiquitous In.the group, as isapparently assumed by those .workers who let theirexperimental animals starve after they have spawned.or are a peculiar feature of the few species that havebeen investigated to date. This laUer viewpoint seemsto be favored by Voss (1963).

During the application of catch-curve methods.especially good care must be taken to ensure therepresentativeness of the samples that are used. II)theparticular case of squids. this implies that the samplesmust not be biased by massive Immigration from thesampling area, e.g. as reported for /liex /IJecebrosus byHurley and Waldron (MS 1978) and Caddy (1'983),orbyhighly selective gears. Rigorous experiments withjiggers and other gears (except trawls) that are used tocatch squids seem not to have been performed. thus

----- --- --

,... IJ,...en... 'J

A

5

0

N

D

'l0

Fig.6.

fig. 7.

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

148 Scl. Council Studies, No.9, 1985

A Loligo peale;Gull 01 Mexico.

o.

20.1 1

Relative age Iyr-t.)

7Ommastrephes bar/rami011 Japan

B6

00 00o

1.5

z 5r 0.at.34

3

20.$ 0.7 0.8 1.1 '1.3

Relallve age ((yr-t.) K)

Dosld/cu$ gigasGull of California5

c00 0

o 0

o000000000°

o. 0

o °1

o.J... _O.SO 0.15

Relative age Iyr-I.)

Examples of length-converted catch curves for squids: A.LeI/go pea/e/, based on data In fig. 1, with successive sam-plea first transformed 10 percentages and Ihen weighted bytheaquare root of sample size prior to addilion by length classto obtain an annual average sample, more or less representa-tive of the population. with conversion to relative ages by theLOGand K values In fig. 6: B, Ommas/rephes bar/rami, basedon length-frequency data in fig. 9 of Araya (1983), with L..,=45 em as suggested by Roper et al. (1984): and C, Dos;d;cusgigas, based on length-frequency daia of Ehrhardt e/ al.(1983) and on Looand K values In Table 1. (Note lack of avi-dence lor posts pawning mortality In B and marked increaseIn this parameter In C.)

0.25 1.00

leaving open, for the time being, the interpretation ofestimates of apparent total mortality which are derivedfrom leng1h-converted catch curves.

Pauly (1984) reported a simple-method which canbe used to estimate probabilities of capture and meanlength at first capture (Lsoor lc) directly from a length-converted catch curve without the benefit of selectionexperiments. The method, which has been field-tested(i.e. shown to provide estimates similar to thoseobtained from selection experim'ents) with fishes, hasyet to be applied to squids. If it is assumed that thecentral part of a length-converted catch curve for squidcan generally'be used to obtain reasonable estimatesof Z, the question still remains of how to separate total

mortality into its constituent parts (Mand F).Two basicapproaches are available for ~stimating (F), namely (i)from the ratio of catch and biomass when the latter canbe estimated, and (ii) from the difference between Zand M.when Z can be estimated from a catch curve anda plausible estimate of Mcan be derived from compara-tive life-history studies.

The first method, which can be applied profitablywhen the sampling gear Is a trawl, involves estimatingthe biomass (S) by the swept-area method (Gulland,1969) and then estimating F from the ratio

F= VIS (9)

where V is the catch in weight from the stock. Analternative to this is to relate the area swept by all gearsof the commercial fleet (a') during a certain period oftime and the total area of occurrence of the stock (A),whereby

F = a'IA (10)

Applications of these methods to squid were reportedby Lange and Sissenwine (1980) and by Sato andHatanaka (1983).

The potential pitfalls in the use of equations (9)and (10) are numer.ous. They range from the generallyunknown and probably changing catchabilities ofsquids to the seasonal contraction and expansion oftheir area of occurrence (A). together with the prob-lems associated with conducting a suitably-weightedstratified-random survey to quantify a biomass that ischanging rapidly d,uring the course of the survey.These and associated problems cannot be discussedhere. and there seems to be no alternative but to applyjudiciously the best standard approaches (Gulland,1983). However. the data that are needed to obtainreasonable estimates of F from equations (9) and (10)will often not be available. In such cases, it will benecessary to rely on the opposite approach, namely,that of substracting an approximate value of Mfrom Z,when both are obtained independently.

Methods to obtain independent estimates of Mfrom comparative studies have been derived by anumber of researchers, starting wilh Beverton and Holt(1959). One such approach, which is based on growthparameters and independent estimates of M for 175fish stocks is that of Pauly (1980a). who derived thepredictive relationship

log M=- 0.2107-0.0824 log Woo+0.6757 log K+0.4627 log T

(11)

where K (annual rate) and We>:>(whole weight in grams)are defined as above and T is the mean annual watertemperature (0C). The equivalent relationship basedon length (Pauly, 1980a) has been widely applied andfound by several researchers .to provide reasonable

5Z0. 10.3 4

3

PAULY: Dynamics of Short-lived Species with Emphasis on Squids 149

..

TABLE 1. Some eslimates 01growth parameters forsquids, wllhcorresponding values of 4Jand M computed from equalions (3) and (10) respective-ly. and appriximate mean environmentaltemperalures. (L is manlle length (cm). Wac.is whole weight (g). and K is expressed on an annuafbasis.)

Species (sex)

1. Lollgo pealei

Area (temperature)'

011 New England (14.S' C)

M

0.B7

Remarks

Loa and K from Ikeda and

Nagasaki (MS 1975); conver-sion of Loa 10 Woa based on

Lango and Johnson (1981).

See Fig. 6; conversion of Loo10 Woo basad on Longo endJohnson (1981).

Loa and K from Lange andSissenwine (1983). cilingAmaratunga (1980) which Ihave not seen. with Loocor-rected from 2.94 lor femalesand 2.39 for males to thoseshown here.

Looand K from Au (MS 1975);conversion of Loo to Wbased on Lango and Johnson(1981).

Loa.W_snd Kcompuled fromdata of Ehrhardt el a/. (1983).

3B.3

K

0.59 1.66680

5. /IIer iIIecebrosus Northwest Allanlic (12" C)

6. Dosidicus gigas Gulf of Call1ornia (21.5' C)

32.0 2.5 2.13

1.26

L_ Wooand Kcomputed Iromdata 01 Okutanl and Murata

11983). using the method 01Gaschu\% 'el III. IMS 19BO).which also provided esti-mates 01 C (equallon 2) 011.40 for females and 4.3 formales. both Indicating a longperiod (winter) 01 no growth.

·Sea-surface temperatures from Anon. (1944) except for Gulf~' Mexico v.alue Irom Rivas (1968).· Values o' (/)appear to be mutually consistent for data sets 1 and 2, sets 3 and 4. and sets 7 and 8. whereas the lack of consistency In data sets 5 and 6

indicate erroneous growth parameters (see also Ag. 5). which may be attributed largely to the subjectlvlly 01 methods used In tracing growth curvasand to tha ellect 01 saasonal growUi oscillations.·Genus Is Onycholheulhls.

7. O. boreaU/apon/ca (~)c

B. O. borealijaponlca (d' )C

East of Hokkaldo (8" C)

East 01 Hokkaldo (B' C)

estimates of M for fish as diverse as 'Pacific skipjacktuna (Kleiber at al.. 1983). Arabian myctophids (Sand-ers a~d Bouhlel, 1982), and West African demersalfishes (Longhurst. 1983).

The predictive relationship (equation 11) shouldalso provide reasonable first estimates of Mfor squids,because fish and squid usually share the same habitat,resources and predators and they are not Jikelyto differwidely in Interrelationships between their vital parame-ters (see Fig. 5). This reasoning, which was also app-lied to shrimps by Pauly at al. (1964)was justified posthoc by the fact that the resultant M-values were wellwithin the range of values considered to be reasonable(S. Garcia, FAD,pers. comm.). The M-values in Table 1were derived from equation (11). 'When growthparameters (LcoorWoo.and K)an~ estimates of Fano Mare available, yield-per-recruit analyses can be con-ducted quite straightforwardly (Beverton and Holt,1966; Ehrhardt al al., 1983), and first steps toward fish-ing management can be proposed.

600 2.25

96 4.4222.000 1.2

40 0.531.575

1.600

0.47 1.80

0.48 1.82 0.5335

Recruitment

The real problem with squids and other short-livedexploited organisms Is not the estimation of growth,mortality and biomass but rather the prediction ofrecruitment, which Is of minor Importance to stocks of

. long-lived fishes because annual recruitment usuallycontributes a small part to the overall stock biomass.

Predicting recruitment In stocks that have beenexploited for only' a few years is doubly difficult whenthe conventional approach is relied upon (i.e. accumu-lation of one point per year until enough pointsbecome available, from which a single "stock-recruitment curve" of some sort can be derived). Forshort-lived fishes and shrimps, detailed examination oflength-frequency data reveals that recruitment ispulsed and that two pulses (of unequal strength) aregenerated every year (Fig. 8A) (Pauly and Navaluna,1963; Pauly at al., 1964). This may also be the generalcase in squid stocks (see Fig. 8B), although relatively

-- - - - - -

2. Loligo poalol Gulf 01 Mexico (24' C) 23 207 0.95 1.52 0.64

3. Iller iIIecabrosus () 011 Eastern Canada (9'C) 29.4 500 0.65 1.61 0.76

4..lIIer I/Iocabrosus (e() Off Eastom Canada 19' C) 23.9 290 1.08 1.68 1.12

150 Sci. Council Studies, No.9, 1985

o

Fig. 8. Recruilmenl pallerns obla;ned by projecling length-frequency dilla onto !he lime axis wi!h growth paramelers t-and K and the ELEFAN II program: A, a very short-lived(1m..-1 yr) anchovy from Manila Bay In the Philippines (Paulyand Navaluna. 1983); and B. Loligo pea/at from the Gull ofMexico based on dala in Fig. 1 and growth parameters esli-maled In Fig. 6. (Nole Indication oflwo recruitmenl pulses peryear In both cases, allhough little importance can be placedon this point in the case of L. peale; due to small sample size.)

few sets of squid length-frequency data have beenexamined_ Moreover, a generalized bimodal recruit-ment would be at variance with some of the accounts Inthe literature, notably that of Ehrhardtet al. (1983),whoclaim to have detected five distinct pulses of recruit-ment per year in the giant squid (Dosidicus gigas) butrelied on a subjective method of analysis (I.e. tracing

growth curves by eye). An objective method, whichutilized all of the available data at one time, was used toderive recruitment patterns of the type illustrated inFig.8.

More elaborate approaches to obtain estimates ofrecruitment in short-lived fishes and invertebrates arelength-structured cohort analyses (Jones and van Zal-jnge, 1981), two forms of which are incorporated in theELEFANIII program (written in BASIC and availablefrom the author) and which were applied to the Peru-vian anchoveta stock by Pauly and Tsukayama (1983).These methods, which require fairly detailed catchcomposition data (e.g. monthly) over a period of atleast 4-5 years offer .real possibility of increasingmarkedly the number of data "points" for stock-recruitment studies. The statistical problems that mustbe considered when such approaches are used cannotbe discussed here, but all of them can be solved withthe use of appropriate statistical techniques, e.g. timeseries analysis (Chatfield, 1975). .

Another aspect of the population dynamics ofshort-lived animals, which should be discussed explic-Itly, is estimation of their annual absolute recruitment.This approach Is straightforward because annualcatch is the product of yield-per-recrult and thenumber of recruits. After performing a yleld-per-recruit analysis, absolute recruitment (R) is estimatedfrom

R =annual catch/yield-per-recrult (12)

Such estimates of recruitment can then be plottedagainst any variable or combination of variables whichare thought to Impact on recruitment, e.g. parent stocksize, predator biomass, and the like. This approach isillustrated here by using data from the Inshore watersof the Gulf of Thailand to a depth of 50 m. This rela-tively well-monitored fishery, which accounts forapproximately 800,000 tons .of fish and invertebrates

TABLE2. Vitalstalislics for a generalized loliginid squid thaI were used in the yield-par-recruitanalysis of the cephalopod stock in the Gull of Thailand.

. 201- A ", , -.

Slo/sphores zolllngsr/

_ 151- Philippines

C...§ 10

"2u..5

0I',

B

301

.Lolipo pea/el

c Gull of Mexico..§ 20"2uOJ

10

Parameter Symbol Value Remarks

Asymptolic weight Woo 150 9 See Table 1 and Fig. 5.

Annual growth rale K 1 Yearly

Age at zero lenglh I. oyr AssumedSelecllonlaclor - SF 2.1 From Amaratunga el a/. (1979)

Mean Jenglh atfirsl capture l.c 5.25 em SF X mesh size

Mean weighl alfirSI caplure We 2.1 g Based on lenghl-weighl relation-ship 01 Lange and Johnson (1981)

Mean age alfirsl cap lure Ie 0.28 yr Equal mean age al recruilmenltofishery (I,)

Mean age al first maturity 1 yr

Mean rei alive fecundily .coo/g Number 01 eggs per gram basedon loliginid fecundity data ofRoper el a/. (198<1)

PAULY: Dynamics of Shari-lived Species with Emphasis on Squids

TABLE 3. Selected inlormalion for Ihe demersal stock and fishery inIhe Gulf of Thailand. 1961-80. (Dala from Booniubol andPramokchulima. MS 1982. complemented with inlorma-lion by Sakurai. 1974.)

.Based on the swepl-area melhod.

· Basl!d on annual esUmates of lolal calch and equation (91: esti-ma'es are assumed 10 apply '0 cephalopod component of mulli-species Slock.

per year, also catches 60,000-70,000 tons of cephalop-ods, most of which (80%) consists of loliginids (Loligochlnensls = L. formosana,L. duvaucelia~d L.uyii=L.lagol) (Chotlyaputta, 1982: Roper at a/., 1984, for syn-onyms). Tha exercise was performed by using vitalstatistics for a "generalized loliglnld squl~" (Table 2).11must be realized that those values are not "averages" or"best estimates" of any kind, but are rather "possible"values for use In the Illustrative example. They alsorepresent the minimum set of data that must be availa-ble for any stock to enable application of the method.Table 3 presents the catch data. for cephalopods andthe corresponding fishing mortality, as well as the esti-mated biomasses of all fish and invertebrates In theGulf of Thailand, representing the biomass' of all pre-dators (and possibly competitors) of cephalopodpre recruits. The estimated standing stock values forcephalopods in Table 3 were reduced by a fraction (m),corresponding to the fraction of cephalopods olderthan 1 year In the stock. This fraction, which is a func-tion of F. was obtained from

~ 1_ 3e-K'2 3e -2K'2_ e-~K'21m = e Z Z+K + Z+2K Z+3K {1 3e-K'1 3e-2Kt, e-:lKt,Z ""' Z+K + Z+2K .- Z+3K

where r, = tc- io, r2 = t~ - 10, and r~ = tm - tc.

(13)

Results of the computations, including derivationof the prerecruil mortality index (-In R/S) are summar-

151

)(IU

"C

.:

8

7

6 .

I I

. .... ..

..~....

01 I . I I0123456

Total standing stock (Ions X 105)

Fig. 9. Relationship between esllma'od prcrocrult mortality 01cophalopods In the Gull 01Thailand end total fish biomass inIhe Gulf's inshore waters. 1981-80.

ized in Table 4. There was a marked decline in prere-cruit mortality with decrease In total demersal standingstock (Fig. 9). This decline, which corresponds to achange of one order of magnitude (Irom 7% to about0.7%) in the daily prerecrult mortality, confirms that the"squid outburst" in the Gulf of Thailand was due to arelease of predatory control on cephalopod eggs andprerecruits, as suggested earlier by Pauly (1979) andby some runs of the "Gulf of Thailand model" of Larkinand Gazey (1982).

Discussion and Conclusions

Juanlco (1983) stated that squids differ markedlyfrom fishes with regard to the major features of theirpopulation dynamics. This view should not remain as amatter of opinion but be seen as a hypothesis whichcan be tested by deliberately applying to squids thosemethods and models which have been applied suc-cessfully to fishes and shrimps.

To model the growth of squids, a seasonally-oscillating version of the von Bertalanfty growth equa-tion is proposed, along with an objective method(ELEFAN I program) for fitting it to length-frequencydata, which should (i) allow for comparative studies ofgrowth performance in different species of squids (andthus for Inferring growth parameters in little-knownspecies), and (iI)help resolve whether growth patterns.differ markedly between fishes and squids whenaccount is taken of seasonal growth oscillations. Thislatter point seems to be particlliarly Important In viewof statements by various researchers who reportedhigh variation in growth of individual squids (and squidcohorts) and expressed doubts about the applicabilityof a single growth model (Caddy, 1983, and referencestherein). However, studies of feeding and food conver-sion in I. iIIecebrosus (O'Dor et a/., 1980) do not indi-cate results that are vastly different .from those insimilar studies on fishes (e.g. Pandian, 1970). It may

TOlal slanding Fishing CephalopodSlock' mortalily" catch

Year (000 lonsl (FI (000 tonsl1961 624 0.17 3.331962 621 0.21 4.031963 618 0.32 6.001964 604 0.53 9.121965 471 0.72 12.21966 353 1.03 15.41967 307 1.42 18.51968 204 2.52 23.81969 275 1.89 29.11970 258 2.05 34.41971 176 3.46 39.71972 168 4.39 45.01973 139 5.98 51.01974 155 3.90 48.41975 125 6.02 53.51976 152 5.18 58.21977 125 6.78 76.91978 139 5.86 71.31979 137 6.08 66.41980 127 6.28 61.4

'.

. 152, . Sci. Council Studies, No.9, 1985

TASLE 4. Step-by-step approach to computalion of prerecruil monelily index (-In RJS ) lorcephalopods in the Gull of Thailand. following the method ot Pauly (1980b. 1982b)except for the feature that recruitment Is divided by egg production olthe precedingyear. as suggested by Murpny (1982). Garcia (1983) and Saytey (1984).

also be mentioned that the cannibalism 011.iIIacabro-sus and other squids, at least under conditions of loodscarcity (O'Oor at al., 1980) should have the tendencyto reduce the variability in growth rates of survivingsquids and, at the same time, to increase the variabilityIn mortality (Mohn, 1982). The study of squid mortali-ties could be greatly accelerated by the systematicapplication of length-converted catch curves. How-ever, rigorous selection experiments, notably for Jig-ging gear, will have to be conducted if definitiveconclusions (e.g. on postspawning mortality) are to bedrawn from the shape of the catch curves.

Natural mortality in squids (prior to postspawning)should be similar to those of fishes with matchinggrowth p~rameters. It so, it should be possible toobtain reasonable estimates of M from the empiricalequations given above, and this will help to test and/orsubstantiate estimates of M based on even simplermethods, e.g. from the assumed life expectancy of thesquids under consideration (Au, MS~1975).

With regard t6 recruitment studies,. the short lile-span of squids can be quite readily turned from a liabil-ity into an asset. Two lineS'01 thought were noted toillustrate this: one makes use 01 within-year recruit-ment fluctuations to provide inferences on between-year fluctuations, and the other relies on the lack ofoverlap between cohorts to provide mutually inde-

pendent estimates of recruitment from annual catchesand a few ancillary data. Such independent estimatesof recruitment can be used to test hypotheses on thetrophic and other interactions that link squids to theirbiotic and abiotic environments. This approach helpsto bridge the gap between "population dynamics" andthe "ecological" studies that Caddy (1983) wroteabouL . .

Finally, it is suggested that nothing can be gainedby emphasizing the differences between squids andfishes and by inferring that methods which have beendeveloped to study fishes will not work when they'areapplied to squids. Rather, "fish" models of knownproperties should be applied systematically to squidsand the "deviations" from such model studied in detail,because it is these deviations that will indicate howsquids are different from fishes.

Acknowledgements

I thank Lourdes "Deng" Palomares for her assist-ance in preparing this manuscript and Chan Eng Hengfor uselul references on Southeast Asian squids. Mythanks also go to Anne Lange, Northeast FisheriesCenter, Woods Hole, Mass., for presenting (in myabsence) an earlier version of this paper at the NAFOSpecial Session on Squids in September 1984.

-.

Cephalopod S '" eggs Prerecrullbiomass. Fraction spawned" Y/R. Recruits. monality

Year (000 tons) mature. (10') (9) (10') index

1961 19.6 0.575 2.250 1.86 1.791962 19.2 0.566 .2.170 2.22 1.82 7.121963 18.2 0.542 2.040 3.05 1.97 7.001964 17.2 0.497 1.710 4.23 2.16 .s.851965 16.9 0.458 1.550 4.94 2.43 6.561966 14.9 0.398 1.190 5.64 2.73 6.341967 13.0 0.330 858 6.04 3.06 5.961968 9.44 0.187 353 6.01 3.96 5.381969 15.4 0.261 804 6.15 4.73 4.311970 16.8 0.240 806 6.13 5.61 4.971971 11.5 0.110 253 5.83 7.05 4.7<41972 10.2 0.063 128 5.2<4 8.59 3.381973' 8.53 0.024 40.9 <4.68 10.9 2.461974 12.4 0.085 211 5.<4<4 8.90 1.531975 8.89 0.023 40.9 <4.68 11.5 2.911976 11.2 0.039 87.4 <4.9<4 11.8 1.2<41977 11.3 0.014 31.6 4.<45 17.3 1.621978 12.2 0.026 63.<4 4.71 15.1 0.741979 10.9 0.222 48.0 4.65 14.3 1.491980 9.78 0.020 39.1 4.59 13.4 1.28

.Obtained by using catch. F-values In Table 3. and the relationship S .. Y/F.·Derived from equation (13) with vital statistics In Table 2 and F-values in Table 3.

" S = biomass x fraction mature x relative fecundity x 0.5.·Conventional YIR analysis. using vital statistics In Table 2 and F-values in Table 3..Obtained by applying equalion (11).

. .

PAULY: Dynamics of Short-lived Species with Emphasis on Squids 153

References

AMARATUNGA, T. MS 1980. Growth and maturation patterns of theshorl-finned squid (l1Iet' iIIecebrosus) on the Scotian Shell. NAFOSCR Doc.. No. 30, Serial No. N062. 17 p.

AMARATUNGA, T" S. KAWAHARA, and H. KONO. MS 1979. Meshseleclion 01 the short-finned squid, l1Iex I/Iecebrosus, on the Sco-tian Shell using a bottom trawl: a joint Canada-Japan 1978research program./CNAF Res. Doc.. No. 35, Serial No. 5361, 29 p.

ANON. 1944. World alias of sea surface temperalures. U.S. NavyHydrographic Ollice, Washington, H. O. No. 235.

ARAYA, H. 1983. Fishery. biology and stock assessment 01 Ommas-trephes barlramiin the Nonh Pacilic Ocean. Mern. Nal. Mus. ViCI..44: 269-283.

AU. D. W. K. MS 1975. Considerations 01 squid (lotigo and /IIex)poputation dynamics and recommendations lor rational exploita-tion. ICNAF Res. Doc., No. 61. Serial No. 3543, 13 p.

BAYLEY. P. 1984. Comment: prerecruil monality In Gull of Thailandshrimp stocks. Trans. Amer. Fish. Soc.. 113: 403-404.

BEVERTON. R. J. H.. and S. J. HOLT. 1959. A review of the lilespansand monality rates of !Ish in nature and their relation togrowlh andand other physiological characteristics. CIBA Colloquia on Age-Ing. 5: 142-180.

1966. Manual 01 methods for fish stock assessment. Part II.

Tables of yield functions. FAO Fish. Tech. Pap., 38(Rev. 1): 1-67.BOONJUBOL, M.. and S. PRAMOKCHUTIMA. 1982. Demersal trawl-

Ing in the Gull of Thailand. Tllai Fish. Rep. Bangkok, 9: 23 p. (inThai.), (Available In English as ICLARM Trans'. No.6.)

CADDY. J. F. 1983. The cephalopods: factors relevant 10their popula-tion dynamics and to the assessment and managemenl of stocks'In Advances In assessment 01 world cephalopo.d resources(p. 416-4521.J. F. Caddy (ed.). FAO Fish: Tech. Pap.,231: 452 p.

. CHATFIELD. C. 1975. The analysis 01 time series: theory and practice.Chapman and Hall, London. England, 263 p.

CHOTIYAPUTTA.C. 1982. Squid fisheries of Thailand. FAO Fish.'Rep.. 275: 124-134.

EHRHARDT, N. M.. P. S. JACQUEMIN, F. GARCIA. G. CONZALEZ.J.M. LOPEZ. B. ORTIZ, and A. SOLIS. 1983. On the fishery andbiology 01 the giant squid Dosidicus glpas In the Gull 01Call1ornia,Mexico. In Advances In assessment 01 world cephalopod resour-ces (p. 306-340), J. F. Caddy (ed.), FAO Fish. Tech. Pap.. 231:452p.

FIELDS. W. G. 1965. Tha slructure, development. food relations. repro-duction and lifehislory ollhe squld.Loligo opa/escens Berry.Fish.BulL Calif. DepL Fish Game, 131: 1-108.

GARCIA, 5.1983. The stock-recruitment relationship In shrimps: real-lIy or artefacts and misinterpretations? OCBenogr. Trop., 1B81:25""8.

GARCIA, D~ and L leRESTE. 1981. Life cycles, dynamics, exploita-tion and management of coastal penaeld shrimp stocks.FI\O Fish.Tech. Pep~ 203: 1-215.

GASCHihZ, G., D. PAULY. and N. DAVID.MS 1980. A versalile BASIC

program lor filling weight and seasonally oscillating lengthgroWlh data. ICES C. M. Doc., No. D:6. 14 p.

GULLAND. J. A. 1969. Manual 01 methods for fish stock assessmenl

Part I. Fish population analysis. FAO Man. Fish. Sci., 4: 1-154.1983. Fish slock assessment: a manual 01 basic methods.

FAOlWiley Ser. Food and Agric., Vol. 1, John Wiley and Sons, NewYork. 223 p.

HIXON, R. F. MS 1980. Growth, reproductive biology, dislribuUon and

abundance of three species of 10Uginid squids (Myopsida. Cepha-

lopoda) in the northwestern Gulf 01 Mexico. Ph.D. Thesis, Univ. 01

Miami.Florida,234.p. . .HIXON. R. F.. R. T. HANLON, and W. H. HULET. 1981. Growth'and

maximum size 01the long-linnet! squid loligo peele/In thenorth-western Gull of Mexico. J. Shell/ish Res., 1: 181-185.

HURLEY,G. V.:and D. E. WALDRON. MS 1978.Population eslimateslor squid (II/el<iIIecebrosus) in ICNAF Subarea 4 fromthe interna-tionallishery In 1977. ICNAF Res. Doc., No. 61. Serial No. 5229,10 p.

IKEDA,I.. and F. NAGASAKI.MS 1975. Siock assessment of loUgo inICNAFSubarea 5 and Statistical Area 6.ICNAF Res. Doc.. No. 44,Serial No. 3523. 5 p.

INGLES. J.. and D. PAULY. 1984. An atlas 01 the groYfth. mortality andrecruilment of Philippine fishes. ICLARM Tech. Rep., 13: 127 p.

JONES, R., and N. P. van ZALiNGE. 1981. Esllmele of mortality rateand population size lor shrimps in Kuwall waters. Kuwait Bull. Mar.Sc;.. 2: 273-288.

JUANICO. M. 1983. Squid malurity scales lor population analysis. InAdvances in assessment of world cephalopod resources (p. 341-

_ 378). J. F. Caddy (ed.), FAO Fish. TOCh..Pep.. 231: 452 p.KLEIBER, P.. A. W. ARGUE, and R. E. KEARNEY. 1983. Assessment of

skipjack (Kalsuwonus pe/amis) resources In the Central and West-ern Pacific by estlmallng standing stock and components 01popu-lation turnover Irom tagging data. Tech. Rep.. New Caledonia,Tuna and BiIIl/sh Assess. Propr. S. Pac. Comm., B: 38 p.

KRISTENSEN, T. K. 1980. Periodic groWlh rings In cephalopod stato-liths. Dana. 1: 39-52. .

LANGE, A. M. T.. and K. L. JOHNSON. 1981. Dorsal manlle length-total

weight relallonship 01 squids lol/go pea/el and lIIex Illecebrosusfrom the Allantic coast oflhe Unlled States. NOAA Tech. Rep.NMFS Circ.. 745: 17 p. '

LANGE, A. M. T.. and M. P. SISSENWINE. 1980. Biological considera-tions relevanllo the management of squid (lotigo pealel and lIIexlIIecebrosus) of the Northwest Atlanllc. Mar. Fish. Rev., 42(7-8):23-28.

1983. Squid resources of the Northwest AUanlic./n AdvancesIn assessment of world cephalopod resources (p. 21-54), J. F.Caddy (ed.), FAO Fish. Tech. Pap., 231: 452p.

LARKIN, P. A.. and W. GAZEY. 1982. Appllcallons of ecological slmu-lalion modets to management of tropical mullispecles fisheries. In

Theory and management 01 tropical fisheries (p. 123-140), D.Pauly and G. I. Murphy (ed.),ICLARM Cont. Proc., 9: 360 p.

LONGHURST, A. 1983. Benthic pelagic coupling and export of organiccarbon from a tropical Allanllc continental shelf - Sierra Leone.Esluarine Coasl. Shall Scl., 17: 261-285.

LUTZ. R. A., and D. C. RHOADS. 1977. Anaerobiosis and II theory ofgrowth line formation. Science, 198: 1222-1227.

MANGOLD, K. 1963. Biologle des cephalopodes benlhlquas etnectoniques de la Mer Catalane. Vie Milieu, 13(suppJ.): 1-285.

MOHN, R. K. '1982. Yleld-per-recrult analysis of shorHinned aquld onthe Scollan Shell by slmulallon. NAFO Scl. Coun. Siudles, 3:39""5.

MUNRO.J. l., and D. PAULY.1983. A simple method for comparingthe growth 01 fish and Invertebrates. Flshbyte, 1(1): 5-6.

MURPHY, G. I. 1982. Recruitment 01 tropicaillshes. In Theory andmanagement 01 tropicaillsheries (p. 141-148), D. Pauly and G.I.Murphy(ed.),ICLARMConI.Proc..9: 360 p, .

O'DOR, R. K., R. D. DURWARD, E. VESSEY, and T. AMARATUNGA.1980. Fedding and growth In cepllve squid, ifill!( lIIec.brosus, andthe Influence 01 food availability on growth In the natural popula-lion. ICNAF Se/. Papers, 6: 15-21.

OKUTANI,T.,and M. MURATA. 1983. A review of the biology 01 theoceanic squid, Onycholheulis boreaU/apon/ca. Mem. Nat. Mus.Vicl., 44: 189-195.

PANOIAN. T. J. 1970. Intake and conversion ollood in Ihe IIsh LImanda

limanda exposed to dillerenttemperatures. Mar, 8;101.5: 1-17.PAULY, D. 1979. Theory and management of tropical multlspecies

Slocks: a review, with emphasis on the southeast Asian demersalfisheries. IClARM Slud. Rev., 1: 35 p.

1980a. On the Interrelationships between natural mor1allty.groWlh..parameters and mean environmental temperature In ;75fish stocks. ICES J. Cons., 39: 175-192.

1980b. A new methodology for rapidly acquiring basic Inlor-mallon on tropical fish stocks: growth, mortality and stock-recruitmenl relationships. In Stock assessments for tropicalsmall-scale fisheries (p. 154-172), S.Saila and P. Roedel (ed.),Inlern. Cent Mer. Res. Dev.. Univ. Rhode I.. 198. p.

1981. The rolalionshlp between gill surface area and growth

perlormance In 'ish: 8 generalizalion 01 von Bertalanlly's theoryo(growlh. Ber. DIsch. W;ss. Komm. Meerestorsch.. 28: 251-282.

--

Sci. Council Studies, No.9, 1985

1982a. Studying single-species dynamics in a muliispeciescontext. In Theory and management oltropicallisheries. D. Paulyand G. I. Murphy led.). ICLARM Conf. Proc.. 8: 360 p.

1982b. A method to esllmate the slack-recruitment relallon-snip 01 shrimps. Trans. Amer. Fish. Soc.. 111: 13-20.

J984. Fish population dynamics In tropical waters: a manuallor use with programmable calculatora. ICt.ARM Slud. Rev.. 8:325 p.

PAULY, D.. and J.INGLES. 1981. Aspecta 01 the growth and mortality01 exploited coral reel fishes. Proe. 41h Inl. Coral Reel Symp..Manila. 1: 89-98.

PAULY, D.. and J. L. MUNRO. 1984. Once more on growth comparisonin lish and Inverlebrates. Fishbyl8. 2(1): 21.

PAULY. D.. and N. A. NAVALUNA. 1983. Monsoon-induced seasonal-ily in the recruitment 01 Philippine lishes. FAO Fish. Rep.. 29113):823-833.

PAULY, D.. and I. TSUKAYAMA. 1983. On the leasonal growth.monthly recruitment and monthly biomass 01 the Peruviananchoveta Irom 1961 to 1979. FAO Fish. Rep.. 291(3): 987-1004.

PAULY, D., J. INGLES. and R. NEAL. 1984. Application to shrimpstocks 01 objective melhods lor the esllmation 01 growth. mortalityand recruitment-related parameters Irom length-Irequency dala(ELEfAN I and II). In Penaeld shrimps - their biology and man-agement (p. 220-234), J. A. Gulland and B. I. Rothschild (ed.).Fishing News Books, Farnham, England. 312 p.

RICKER, W. E. 1975. Compulatlon and interprelation 01 biologIcalslatisllcs'olllsh populations. Bull. Fish. Res. Board Can., 191:1-382.

RIVAS, L. R. 1968. Fishermen's atlas 01 monlhly sea surface tempera-ture lor the Gull 01 Mexico. Clre. U.S. Fish. Wildl. Serv.. 300: 33 p.

ROPER. C. F. E.. M. J. SWEENEY. and C. E. NAUEN. 1984. An anno-tated and illustrated catalogue 01 species 01 inlerestlo lisheries.FAO species calalogue, Vol. 3: cephaloPQds 01 the world.

RUSSELL, E. S. 1931.Some theoretical considerations on the overfish-ing problem. ICES J. Cons., 5: 3-20.

SAKURAt, T. 1974. Squid and cullielish lisherlea In Thailand. Prce.IPFC, 15111I):496-511.

SANDERS, M. J.. and M. BOUHLEL. MS 1982. Preliminary esllmales 01some stock assessment parameters lor the meso pelagic speciesBenlhosem, plerotum Inhablllng the western and northern Ara-bIan Sea. FAO/UNDP (Tach. A.tai.,.) Rap. FIsh.. RAB771OO8, 23 p.

SATO. T., and H. HATANAKA. 1983. A reviewal assessment 01 Japa-nase distant_ater lisherles lor cephalopods. fn Advances Inassessment 01 world cephalopod resources (p. 145-180), J. F.Caddy (ed.), FAO Fish. Tech. Rap.. 231: 452 p.

SPRATT. J. D. 1978. Age and growth 01 the market squid. Lolipoopalescans Berry. in Monterey Bay Irom statoliths. Fish. Butl.Calif. Dapl. Fish Game, 169: 35-44.

SUMMERS, W. C. 1971. Age and growth 01 Lol/po pa.f.l:. populationstudy 01 the common AUanlic coastlquid. BloL Sul,. Mar. BiaLL,b. Woods Hole, Mass., 141: 189-201.

TAYLOA. C. C. 1958. Cod growth and temperature. ICES J. Cons., 23:366-370.

1959. T omperature and growth -the Paclllc razor clem. ICESJ. Cons.. 25: 93-101.

1960. Temperature. growth and mortality - the Pacillccockle. ICES J. Cons.. 26: 117-124.

TINBERGEN, L.. and J. VERWEY. 1945. Zur Blologl8 yon Lollgo vulga-ris. Arch. NeerL Zoot.. 3: 333-364.

VERRILL. A. E. 1881. The cephalopods 01 the northeastern coast 01American Part II. The smaller cephalopods Including Ihe squidand the octopus, with other allied lorms. Trans. Conn. Acad. AflsSci.. 5: 260-448.

VOSS. G. L 1983. A review 01 cephalopod lishery biology. Mem. Nat.Mus. ViCI., 44: 229-241.