working paper series - pdf.usaid.govpdf.usaid.gov/pdf_docs/pnabk256.pdf · the university ef...

WORKING PAPER SERIES

FISHERIES STOCK ASSESSMENT TITLE XII

Collaborative Research Support Program

Fisheries Stock Assessment CRSP Management Office International Programs College of Agriculture

The University of Maryland College Park Maryland 20742

In cooperation with the United States Agency for Intemutlonai Development (Grant No DAN-4146-G-SS507100) the Fisheries Stock Assessment CRSP involves the following participating institutions

The University ef Maryland-Center for Enttronmental aid Estuarine Studies The University of Rhode Island-International Center for Marine Resource DevelopmentThe University ot Washington-Center for Quantitative Sciences The University of Costa Rica-Centro de Investigaci6n en Cienclas del Mar y Limnologfa The Univerrilty of the Philippines-Marine Science Institute (Di- nan)-College of Fishshy

eries (Vie ayas)

In collaboration with The University of Delaware The University of Maryland-Collegeof Business and Management The University of Miami and The International Centew for Living Aquatic Resources Management (ICLARM)


Working Paper 83

Simulation of the Effects of Spakning and Recruitment Pattcrns plusmnn Tropical

and Subtropical Fish Stccks on Traditional Management Assessments

by

Jerald S Ault and William W Fox Jr

June 1991

Fisheries Stock Assessment Title XII


The Fisheries Stock Assessient CRSP (sponsored in part by USAID Grant No DAN-4146-G-SS-5071-00) is intented to supportcollaborative research between the US and developing countries universities and research institutions on fisheries stock assessment and management strategies

This Working Paper was produced by the Univrbity oi Maryland-Center for Environmental and Estuarine Studies and the Universityof Costa Rica-Centro de Investigacion en Ciencias del Mar yLimnologia (CIMAR) in association with the University of Delaware and the University of Miami Additional copies are avdilable from the CRSP Management Office and from

Chesapeake Biological Laboratory Center for Environmental and Estuarine Studies

The University of Maryland PO Box 38

Solomons Maryland 20688

Simulation of (lie Effects of Spawning and Recruitment Patterns in Tropical and Subtropical Fish Stocks

on Tradiliional IIimagenment Assessmens

JERALD S AULT AND WILLIAM W FOX JR Rosersdel School of Marine and Atmosplheric Science

University of Alianii 4600 Rickenbacker Causeway

Aliami hlorida33149

ABSTRACT Many subtropical and tropical fish strcks are believed to spawn rccruit and

grow quasi-continuously Typically triditional iceid models derived forseasonally (discrete) spawning and growing temperate zone fishes are utilizedo assess fish stocks and provide management advice in subtropical and tropicalimarine environments Poptration processes specific to fish s in these regionsar( bclieved to impair conclusions derived from traditional fishery modeling No one has investigatcd the effects of continuous spawning and recruitment incxploield pxopulalions ina gcncralized and comprehensive maimer In this studyclasses of spawning and recruitment were letrmnined A continuous-tinuc computer simulaion niodel which accomodates variable recruiimcnt strategicsand creates tie respective population structures with facility for seasonallyvariable growth is described Resuhs of a simulation study are utilied to

1Compare discrete-vcrsus conltiotms-rccruiting cases2 Determine the efficacy of traditional stock assessinent techniques under

both cases 3 Identify and evarite scurces anti effects of bias

The effects on management advice produced by standard yield models are discussed

INTRODUCTION Fisheries in tropical seas are under pressure for rigorous management policy

due to competing interests for available finite resources Concurrently with heLaw of dhe Sea and the advent of nationtalized 20() mile limits for marine resources increased attention is being given to fishery management particularlyin pro(uctive tro)ical seas such as those bountded and fished )y developingcountries

Combinations of species richness and the abundance of individuals intropical seas has conlributed to a myriad of problems in understanding fisherydemographics Due primarily to the paucity of adequate and apl)ropriate fisherydala simple single-species equilibriun models with minimal dala require entsare touted as the supporting meclnisns for the development of fisherymargement policy in tropical region (Pauly 1982) These methods areoperated inlight of the fact ihat virtuallv no exploited fish population is in asleady sate yet as a first approximation iii mixleling a fish population a steadystate is usially assumed (Gullamd 1983)

radilional exploited fish stock assessment techtmiques developedspecifically for advanccd fisheries it len Iplliate waters rely primarily ottiuformation concerning the distribution of the ages of fish in die catch

Aul JS and WWFox Jr 1989 Gulf amp CaribFihnst 39361-388 361

Proceedings of the 39th

Temperate region fisheries are characterized by1 Single species2 Single or few $ear types3Heavy industrialization (with good statistical reporting)4 Well-defined seasonal growth5 Homogeneous stock distribution for a given area 6 Fishing effort affecting all age groups equivalently

Note that management in temperate regions has not r cessarily served aas paradigm

In contrast because of subtle seasonal variation in tropical waters and the remoteness of many productive fishing areas large-scale sampling programs for age distribution are usually not feasible (Jones 1981 1984) This i becauseaging of these fishes is quite difficult due to multiple cohort production and thecontinuous growth conditions occurring intra-annually Typically the catchesfrom fish stocks are aged by an average-length key Staistical averages are more meaningful when they refer to homogeneous groups (Keyfitz 1977)Protracted spawning which is distributed asymmetrically induces considerable error in the basic catch-at-age data the magnitude of which is not known Thusage-based stock assessment techniques may be of limited utility in tropicalfisheries

Tropical multigear multispecies fisheries are hard to understand Tropicalfisheries are combinations of large and diffuse artsanal operations andorrecreational fisheries that are compounded with developed or developingfisheries They attempt to operate on selected species from multiple assemblageswhich are frequently estuarine or coral reef based Catch is landed at multipleports (with species lumping ia statistics) and catch and effort data whenexisting are incomplete by areas Tropical species life history characteristics include

1Continuous or protracted spawning activity2 Alternative reproductive strategies (ie hermaphroditism) 3 Continuous growth4 High predation 5Intense competition among and between cohorts 6 High species richness and diversityFisheries associated with tropical reef systems tend to center on stocks

which are top predators in the system Gear is more size-specific than speciesspecific The differentia fishing mortality strategies by sex age and spawningaggregations can damage the stocks reproductive potential Additionalysimultaneous harvesting of a large nunber of species in multispecies systemswill often manifest complex catastrophic behavior whereby the system isdiscontinuously transformed to different equilibrium states (bifurcating multiplestable or chaotic points) as the harves rates increase or environmental circumstances alter

Traditional single-species stock assessments applied to tropical situations most likely fail because the species under study are embedded in complexcommunities The dynamics as well as the equilibrium densities of the speciesof interest undoubteaiy reflect the embedding A paradox of the tropics resultswhen assessment methods are tacitly applied to these seemingly simple fisherieswhich are embedded within complex frameworks If studies of subtropical andtropical fish communities are to have the adequate theoretical background

362

Gulf and Caribbean Fisheries Institute

essential for a proper understanding of the complex phenomena involved then afcasible alternative is to develop models according to the properties of theparticular systems rather than attempt to apply any general theory Modelsshould emerge as syntheses of knowledge of a range of actual cases Simulationmodels are the current vogue in practically all branches of science fisheriesscience is no exception A simulation model is constructed in the hope that itwill successfully mimic a real world system The model may as a consequencebecome very complicated and involved Numerical modeis provide iheroalistic representation mostof complex problems of apy quantitative techniquebccause they allow characterization of large sets of differential equations withfeedback mechanisms from both deterministic and stochastic perspectives

STATEMENT OF THE PROBLEMFishery management programs require estimates of stock size a function ofrecruitment growth and mortality These variables are governed by timedependent and spatial mechanisms such as competition predation andmesoscale environmental processes The stock and recruitment problem hasbuen identified as the most serious scientific problem facing those concernedwith fisiery management (Gulland 1983) Many subtropical and tropical fishand invertebrate stocks are believed to spawn recruit and growquasi-continuously Notable among these families are the economicallyimportant serranids scombrids engraulids clupeids lutjanids and penaeidslrimps These types of life history strategies can have profound effects on ourperceptions of stock response to exploitation Current stock assessmenttechniques developed specifically for temperate discrete spawningsingle-species fisheries are basically elaborate methods of fitting curves to dataiose models simplifying assumptions may have an insidious nature whenapplied to tropical situations To date no one has adequately quantified orevaluated continuous recruitment nor its effects on traditonal fishery modelsFurthermore Ihere is a paucity of competent technology to deal with thesespccific conditions Population structure and the exploitation potential ofmarine fish stocks are dependent upon intrinsic recruitment patterns Failure toadcquately address indigenous recruitment patterns may have deleterious effectson specific analysis procedures and impair management adviceWe believe that tropical and subtropical fishery systems can be moreeffectively managed by systematically evaluatingspecific to the fish stocks in the population processesthese regions Therefore the objectives of this

study were1To determine and describe classes of spawning and recruitment2To develop a quasi-continuous time simulation model which mimicsobserved tropical and subtropical fish recruitment patterns andsatisfactorily models the respective population structures3 To compare parochial discrete-time recruiting populations with thedynamic quasi-continuous cases4 Perform sensitivity analyses to determine the efficacy of traditional stockassessments techniques under b)th the quasi-continuous and discrete

cases 5To determine and identify sources of biasOur overall research goal is the construction of widely applicablemathematical algorithms with the ability to model the variability and scale of the

363


continuous spawning and multiple cohort recruitment processes which presently are believed to impact resource assessments of tropical and subtropical marine species and which alleviate the problem of having to investigate the full factorial design numerically

SPAWNING AND RECRUITMENT PATTERNS IN TROPICAL MARINE WATERS

A supposed universal reproductive strategy in fish stocks is to maximize the production of surviving progeny in relation to available energy and parental life expectancy (Ware 1984) However the vaiiety of reproductive strategies andtactics in fishes clearly indicate that no single reproductive pattern is universallyadaptive even within the same ecological community Rather there is some combination of life history characteristics where each specific strategy confers the highest fitness Natality processes are critically implicated in establishingfish stock structure Some remaining questions are

1How is recruitment disturbed temporally2 What is the magnitude of its variation relative to transitional fishing

strategiesTo date most theoretical studies of alternative reproduction and survil

patterns have concentrated on the evolutionary implications of the equilibrium state and have largely ignored transient dynamics which are an important concern of management agencies (Ware 1984)

For fishes inhabiting an environment with pronounced seasonal climaticvariability the breeding season is almost invariably confined to a brief and specific period of the year At high latitudes many fishes produce a singlespawning batch per year (Qasim 1956) This is reflected in a discrete unimodal pulse of recruits entering the population annually as a single population cohorl (Figure IA) A cohort here is defined as a set of fish from the same speciesentering the population at the same instant Conversely the less variable environmental fluctuations in tropical latitudes may produce a relativeyconstant environment in which breeding spawning and hence recruitment cail theoretically occur throughout the year This can result in multiple cohort production within any one calendar year (Figure 113) By our presentnomenclature these multi-cohort yearclass members are considered of the same annual cohort Furthermore in tropical regions fish populations are suspected tolose individuals principally through predation Since predation occuis unpredictably both temporally and spatially it may be beneficial for a certain species to have a permanent pool of recruits ready to buoy population levels Reproductive strategies should therefore have evolved to maintain a continual source of recruits to buffer losses (Luckhurst and Luckhurst 1977)

SpawningMany tropical and subtropical marine fish have protracted breeding seasons

with different individuals breeding at different times throughout the year (Sale1980 1982 Munro et al 1973 McKaye 1977 Nzioka 1979 Bayliff 1980Munro 1982 Bye 1984 Grimes 1986) Most of these fishes are believed to grow quasi-continuously At least 27 families of economically importanttropical and subtropical marine fishes and invertebrates spawn year-round(Table 1) Whereas intense spawning activity is periodic in tropical latitudes (Bye 1984) it is clear that small population fractions of many species spawn in

364


(A) agcO

FREQUENCY age I

ae III-

LEOTH

Figure 1 Hypothetical population length frequency distributions for the (A)Retracted disshycrete spawning temperate zone and (B) Protracted quasi-continuous spawning tropicalzone marine fish life histories

a continuum over much longer portions of the year Qasim (1956) suggested thatthe breeding cycles of marine fishes are adapted to provide optimum conditionsof food and temperature for the larvae The lower the latitude the Ion ger thescason when temperature and food conditions favor the survival of juveniles(Johannes 1978) During the relatively wann breeding seasons of mostsouthern fish stocks sufficient food is available for adults to support thematuration of successive baches of eggs Cushing (1970) proposed that in anenvironment where variability in the timing of the annual production cyclercsults from a combination of factors a fixed spawning time on the samespawning ground each year is most likely to ensure a match between larvae andfood In the subtropics and tropics the production cycle is relatively continuousand of low amplitude so that precisely timed spawning is apparently not asimportant and breeding occurs througLout much of the year (Cushing 1975)Perhaps this is because there is no feeding advantage to be gained by restrictingthe spawning season however predation can be avoided by spawning over anextended period (Grimes 1986)Serial spawning fishes are characteristic of subtropical and tropical seas(Nikolsky 1963) Repeat spawning withia a season is reflected in

365

Table 1 List of some families of subtropical and tropical marine fishes and invertebrates that are known or suspected to spawn andor recruit

continuously or quasi-continuously on the basis of empirical data

Family

Labridae 1 (Wrasses) 2

3

Scaridae 1 (Parrotfishes) 2

3 4

Serranidae 1Hamlets 2 Groupers 3 4 5

Lutjanidae 1 L apodus 2 3 L campechnus 4 Lane Snapper 5 L vivanus 6 L kasmira 7 L filamentosus8 L multidens

Spawns

Daily-year-round year-round semi lunar

Oct-May

year-round (daily) year-round (peaks Jan Oct)

Jan-June Oct-May

year-round (daily) Jan-July (peak Feb-Mar)

July-Feb (peaks Jan Oct) Nov-Jan

year-round (peaks spring amp fall) July-Feb (peaks Jan Oct)

Apr-Dec year-round year-round

Mar-Dec Mar-Dec

year-round

Recruits

Concentrated July-Dec year-iound

year-round common winter ampqpring

LocatlonSource

Panama Victor (83) Hawaii Ross (83)

Australia Russell et al (77)

Panama Robertson ampWarner (78) East Africa Nz~oka (7S)

Jamaica Munro etal (73) Australia Russell et al (77)

Panama Fischer (81) Jamaica Munro (82)

East Africa Nzioka (79)Australia Russell et al (77)

Curagao Luckhurst amp Luckhurst (77)

Jamaica Munro et al (73) Munro (82) East Africa Nzioka (79)

NW Gulf of Mexico Bradley ampBryan (75) 0 Trinidad Manickchand-Dass (87) 0

Puerto Rico Boardman 8- Weiler (80) Western Samoa Mizenico (84) Hawaiian Islands Ralston (81) o

New Hebrides Brouard amp Grandperrin (84)

CA

Table 1(continued)

Family Scombridae 1Yeilowfin

T albacares

2 Albacore T alalunga

3 Skipjack

4 So Bluefin 5 Bigeye

6 No Bluefin 7 Pacific Mackerel

Scomber japonicus

Sciaenidae 1 C Nothus

2 3

4

Engraulidae

Spawns

year-round

multiple spawning

Mar-Sep (peak Mar-May)

year-round multiple spawning

Sep-Mar (peak Nov-Dee) year-round (peak Apr-Sep)

Apr-July Mar-Oct (peak Apr-Aug)

multiple spawnings (batches)

May-Nov (bimodal)

8 months

year-round (20 batchesyear)

Recruit

larvae year round seasonal

density year-round(2 peaks) 2 cohorts

period becomes with increased distance

from equator

larvae more numerous in W ampE Pac (30N-20S)

bimodel

bimodal

year-round (peak July-Nov)

LocationSource

Mexico Central America Equatorial 5 waters of W ampCentral Pacific Cr

Bayliff (80) Otsu amp Uchida (59) Knudsen (77) Hawaii Yoshida (68)

Foreman in Bayliff (80)

Pacific near 0 Forsbergy in Bayliff (80)

Australia Olson in Bayliff (80) Equator - 12N

Calkins in Bayliff (80)Betw Japan amp Philippines Bayliff (80)Baja California Schaefer in Bayliff (80)

N Gulf of Mexico DeVries amp Chittenden (82)Curaqao Luckhurst ampLuckhurst (77)

California De Martini amp Fountain (81) S Africa Wallace (75)

California Lasker ampSmith (77) Hunter ampLeong (81)

Table I (continued)

Family

Clupeidae

Haemulidae

Pomacentridae 1P wacoi (Damselfishes) Fflavicauda

2 Chromis 3 4

Lethrinidae 1(Emperor fish) 2

Gobidae 1 2

Apogonidae (Carcrinalfishes)

Pomacanthidae I (Angelfishes) 2

3

Spawns

Aug-Feb (2 peaks

Aug-Sep ampDec-Jan)

year-round

year-round

year-round

year-round (2 peaks

Sep-Oct ampJan-Feb)

Oct-Apr

year-round

summer-fall year-round

OctAay

Recruits

Nov-Dec Larvae

year-round dominant 15 day period

semi-unarsporadic pulses-wave fronts

Oct-Mar semi-lunar year-round

bimodalSep ampApr

mid-wintrune-Aug

year-round

bimodal-mulL modes

year-troundpeaks spring amp fall

LocationJSource

Baja California Barret amp Howard (61)

Virgin Islnnds McFaland et a(85)

Australia Dougherty (83)

Australia Sale (82)Hawaii Lobel (78)


Australia Waker (75)

East Africa Nzioka (79)

Australia Russell et al (77)Curagao Luckhurst amp Luckhurst (177) 0

i

Curacao Luckhurst amp Luckhurst (57) 0

Gulf of California Thresher (84)Florida Thresher (84)

AutiIi Russell etal (77)

Table 1 (continued)

Family Grammilae

Spawns Recuts LocationSource C

(ramaets) (Fairy Bassiets) year-round

bimodal Sep ampMay Curaqa Luckhurst amp Luckhurst (77)

=L

Sr

Canthigaster U (Puffer) year-round bimodal Curaao Luckhurst amp Luckhurst (77) Carangidae

Pomadasyidae

Holocantridae

Mullidae 1

2 Chaetodontjdae

Acanthundae

Balistidae

Centropomidae 1 Common Snook

Penaeid 1Penaeus subtilis Shrimps Brasiliansis

year-round (largest in coolest months)

year-round (peak Feb-Apr)

July-May

year-round year-round (2 peaks Mar Nov)

Oct Feb Apr-Aug

year-round

July-Mar

Apr-Dec (2 peaks June-July Aug-Oct

year-round year-round (2 peaks Feb-June Oct)

Jamaica Munro eta (-73)Munro (82)

Jamaica Munro et al(73) Munro (82) Jamaica Munro et al (73)

Jamaica Munro eta (873) East Africa Nzioka (73)

Jamaica Munro et a (1713)

Jamaica Munro eta (73)

Jamaica Munro et al(73)

S Florida Gilmore eta (83) Garcia ampLeRasla (81)

Suriname Willmann amp Garcia (85)Garcia ampLeReste (81)

2

0

Panuluridae 1Spiny Lobster Bermuda Ward (1989)


characteristic size-frequency distribution of oocytes within mature ovaries where as well as a batch completing viteflogenesis or already ripened there is also a comparably numeroits group at mid-vitellogenesis which will provide the next spawning (Hunter and Macewicz 1985) This pattern has been documented inter alia for queenfish (DeMartini and Fountain 1981) anchovies (Hunter and Leong 1981) and snappers (Grimes 1986) Protracted spawning is characteristic of serial spawners (Nikolsky 1963 DeMartini and Fountain1981) Conversely northern forms spawn at a season which is not veryfavorable for feeding so that the Oemales put all their reserves into a singlebatch For example skipjack tuna bigeye tuna and yellowfin tuna are believed to spawn year-round in the equatorial Pacific Ocean Tuna larvae are found all over the northern subtropical Pacific at nearly all seasons (Matsumoto 1966)Skipjack tuna however are believed to have a spawning period which decreases with increased distance fiom the equator (Forsberg 1980) Grimes(1986) identifies two patterns of reproductive sea-onality for the lutjanids

1Continental populations have a spawning season which is typicallycentered around summer

2 Insular populations associated with oceanic islands reproduce year-roundwith pulses of activity in the spring and fall

Within the breeding seasoq of vaous reef fihes productiop of multiplebroods is usual Daily weekly biweekly and monthly spawning cycles are all common (Robertson and Warner 1978 Fischer 1981 Warner and Robertson 1978 Sale 1982 Victor 1983 Dougherty 1983) Daily spawning isa common phenomenon among coral reef fishes (Victor 1983) particularly in wrasses (Labridae) parrotishes (Scaridae) and the basses (Serranidae) Other families are reported to spawn once or twice every two weeks (Ross 1978 McFarland et al 1985) and others spawn monthly with perhaps several spawnings occurring over two or three days each month (Lobel 1978) The queenfish a sciaenid is a serial spawner with a protracted spawning season Spawning of the queenfish is asynchronous among females but has monthly peaks in intensity during the Frst quarter of the moon (DeMartini and Fountain 1981) The queenfish closelyresembles many or most small planktonic spawners of warm temperate regionsLunar spawning intervals are also known for several littoral fishes that spawndemersai eggs (Hines et al 1985)

Recruitment The term recruitment for commercially exploited fish populations has

been defined by Beverton and Holt (1957) as the process in which younz fish enter the exploited area and become liable to contact with fishing gear Parrish (1978) distinguishes two life history stages to the recruitment process

1 Prerecruit - recruitment to the fishable stock 2 Recruits - recruitment to the fishery

Cushing recognizes a third stage3 Reproductive recruit - recruitment to the mature stock Survivorship to the recruit phase is apparently very low The numbers of

recruits are substantially less than the number of eggs produced each season Several recent studies have measured recruitment to monitored sites Althoughreproductive activities of marine fishes in these areas often extends over considerable periods seasonal patterns in reproductive activity of adults and larval abundance and recruitment of these fishes have been well documcnted

370

Gulf and Caribbean Fisheries Insitute

(Johannes 1978 Munro et al 1973 Nzioka 1979 Luckhurst and Luckhurst1977 Bayliff 1980 Murphy 1982) The most detailed information on both spatial and temporal variation in -ecruitment concerns the more abundant reef species particularly pomacentrids in the Indo-Pacific and the Caribbean (Eckert1984 Victor 1983 Luckhurst and Luckhmst 1977) and haemulids in the Caribbean (McFarland et al 1985) The relative paucity of data for other species groups reflects to a great extent

1The low densities in which other species settle 2 The cryptic behavior of some juveniles3 The inability to sample larval habitats properly (Lasker 1981)Recruitment during seasonal peaks is apparently not uniform but tends to

occur in pulses Seasonal peaks in larval abundance and recruitment have been observed on the Great Barrier Reef (Dougherty 1983 Williams 1983Sale1980) in the Philippines (Pauly and Navaluna 1983) Japan (Yamamoto1976) Guam (Kami and Ikehara 1976) Hawaii (Watson and Leis 1974) the eastern Pacific Ocean (Bayliff 1980) the Gulf of California (Molles 1978) and the Caribbean sea (Luckhurst and Luckhurst 1977) Spring and fall peakswhich seem to be confirmed for the Caribbean (Munro et al 1973 Luckhurst and Luckhurst 1977) might converge in more northern waters and be replacedby a single summer peak Presently there is no eviuence that the degree of annual variation in recruitment is any greater or less in some species thanothers (Sale 1982) For many reef species 3 to 4 ftl differences between years in given months have been demonstrated and up to 10 fold differences between integer years in the number of recruits arriving (Sale 1982) In some locations (eg Philippines Hawaii Jamaica Panama) seasonal recruitment peaks are multimodel In other areas there appears to be just one seasonal peak(eg Great Barrier Reef eastern Pacific Ocean western Atlantic Ocean) Most species in an area tend to recruit during the same season(s) but exceptions do occur

Predictable recruitment patterns of coral reef fishes closely parallelspawning patterns (Victor 1983 McFarland et al 1985) The occurrence of ripe fishes or the presence of very young individuals has been considered as evidence of spawning activity Corresponding recruitment peaks could be expected to be lagged to some degree after peaks in larval abundance (seeHunter and Leong 1981) For example spawning in Jamaican waters shows two main periods for most of the larger species of reef fishes with maxima around March to April and September to October (Munro et al 1973) Correspondingrecruitment peaks occur after peak larval abundance (Parrishs prerecruitment)(Luckhurst and Luckhurst 1977) Allen (1975) states that periodic blooms may occur in some coral reef fishes in seasons when favorable environmental conditions enhance recruit survival Nzioka (1979) suggested that peakspawning occurs among continental East African lutjanids at the time of highestproductivity during the southeast monsoon which speeds up the East African current lowers the thermocline and promotes vertical mixing Wallace (1975)noted that the dominant species in South African estuaries had extended spawning seasons of up to eight months Prolonged periods of post-larval and juvenile recruitment to South African estuaries was thought to provide a buffer against recruitment failures resulting from droughts or unseasonal floods (Wallace 1975) Lunar settlement cycles in subtropical and tropical fishes are reported (Shulman et al 1983 Hines et al 1985 McFarland et al 1985) The

371


timing of settlement of potential prey species relative to their predators can be acrucial factor in successful settlement into a habitat (Shulman et al 1983) Alsoeddies and other larval fish transport mechanisms are believed to influencetemporal and spatial fluctuations in recruitment Eddies appear to be commonfeatures of tropical islands and landfronts (Olson and Backus 1985 Lobel andRobinson 1986) Reproductive strategies of marine animals may reflect tosome extent these transport mechanisms There may be an important link to theperiodicity of the spawning and recruitment of species taking advantage of such features (Lobel and Robinson 1986)

CONCEPTUAL MODELThe reproductive strategy of a species of fish is that complex ofreproductive and behavioral traits that fish manifest so as to lave some

offspring In population biology this usage has led to the development of theconcept of evolutionary stable strategies (ESS) that is a mixture of strategiesadopted by a population that is not susceptible to invasion by a hypotheticalalternative(Maynard Smith 1976) A strategy designates a plan so completethat it cannot be upset by enemy action or nature The terms strategy and tacticimply rational planning or natural selection but note fiiat it does not have to be a optimal strategy Thus a major problem for populationr biologists is to showhow particular reproductive strategies and tacics are adaptive i particularenvironmental circumstances

The spawning and recruitment patterns exhibited by temperate and tropicalmarine fishes have been classified within the context of a 3 x 3 matrixarrangement (Table 2) Three clases were assigned to each of the followingthiee categories

1The spawningrecruitment patterns2 The distribution of modes observed annually3 The affinity to well known statistical distributions Obviously overlapped and interconnected breeding and recruiting strategiesexist Our approach allows a basis for conceptual framework from which wewill subsequently numerically mode these fish populations The continuum

and protracted spawning and recruitment states are characteristic of tropicalenvirons The continuum recruitment condition refers to an infinite set ofrecruit particle emissions such that between any two of them there is a thirdpulse of recruits This condition is by definition continuousspawningrecruitment The protracted condition refers to an intermediate process where spawningrecruitment has an extended duration in time and thusis quasi-continuous Most subtropical and tropical fishes are iteroparous (ie

Table 2 Classes of heterochronal annual spawingrecruitment patterns for tropical andsubtropical iteroparous marine fishes

Class UI Ill

SpawningRecruitment Retracted Protracted ContinuumForms Uniform Unimodal MultimodalDistribution Normal Gamma Other

372

Culf and Caribbean Fisheries Institute

reproduction occurs on more than one occasion during the life-span) Annualfecundity is seasonally indeterminate in many of these fishes (frequently calledmultiple partial serial or heterochronal spawners) (Huiuer el al 1985) Insuch fishes the standing stock of yolked eggs regardless oi maturity state givesno indication of annual fecundity because these fishes continuously mature newspawning batches throughout the protracted spawning season Such fishesusually spawn many times during a season and several through a continuumAnnual fecundity is a function of batch fecundity and the number of spawningsper year The most conservative assumption is that seasonal fecundity isdeterminate for muitiple spawning fishes and that estimates of batch fecundityd spawning frequency are required The documented cases of determinate ficundity appear to be restricted to boreal or cold temperate climates wherespawning seasons are short Thus in most of the worlds oceans indeterminatefccundity and multiple spawning are the rule for epipelagic spawners (Hunterand Macewicz 1985) Retracted spawningrecruitment is typical ofdiscrete-pulsed temperate zone fishes These fishes spawn and recruit over a very short interval of the year (gt3 months) In most boreal and temperate speciesall the eggs to be released in a season develop synchronously prior to spawningand spawning typically takes place over a retracted period In such speciesthe standing stock of oocytes within certain classes isa range of maturityconsidered to represent the annual fecundity of the spawner (Hunter et al1985) Although some of these fishes may spawn repeatedly during the seasontne standing stock of yolked eggs is considered representative of the annualfccundity (Hislop et al 1978) The spawningrecruitment conditions delineatedabove have various modal forms associated with them Statistically cohortpropagation may be modeled by either a uniform distribution (ie time andmagnitude constant) or single or multiple normal gamma or some otherprobability distribution Any equation representing a real biological systemshould be valid irrespective of the units in which we measure the quantitiesinvolved in the system In all cases the magnitude of the recruitment is notparticularly revealing othcr than to allow calculation of potential fishery yieldsThe form and periodicity of recruitment are critical in establishing theprobability of a particular sizeage population distribution and the subsequertimpact those particular distributions may have in influencing recruitment

A spectrum of selected examples of both the spawning and recruitment patterns aie shown in Figures 2 and 3 respectively In no case have weaccounted for any more than two years of interannual variability due to abiotic(ie climatological) characteristics Rather we are more interested in theapparent patterns and forms of spawning and recruitment in these fishes Ourapproach is to model specific observed spawning and recruitment patterns in ourinvestigations and examine the influence these conditions have on ourinterpretations when standard stock assessment methodology is applied Thetime dependence of these event- suggests that variations in the amplitudes ofsinusoidal rccruitmn-nt functions may be important in understanding whether attractor spaces exist in continuous population cycles

MATHEMATICAL SIMULATION MODEL amp EXPERIMENTAL APPROACH

In addition to the numbers of recruits produced within a given time interval (eg a year) an important aspect of recruitment is its structure within that

373


(A)WKWEMIN11 - I[ RICA

11TA(B) LUTJAMIDAE LAFRICA

PERCENT OF THE I POPULATION (C)SPAWNING20

SPAWNNG IA1LETINIDAE -IN

20shy

3 (D) MUUIDA[ - AFRICA

20shy

(E)

30jCARWDAE- jM ICA

ONE YEAR

Figure 2 Selected circum-tropical protracted spawning patterns for five families of marine fishes Sources of data Nzioka 1979 (ABD) Qasim 1973 (C) Munro 1982 (E)

interval viz whether recruitment isdiscrete or continuous Very little work has been done on stock-spawner-recruitment relationships for tropical fishes(Pauly1982) Tropical fishes have been rported by many authors to haveprotractedspawning seasons (op ciL) In natural fish populations the number of recruits entering the system will vary substantially Thus an understanding of the

374

_ _ __

Gulf and Caribbean Flsherls Institute

so (A)40- Iamulllsib A5AI (e)reasi kinSl VlEU SI U)S

30shy

10i __ __ _ I __ _

CUNACAG km CUACAS

jGaM~idld

p I I

(E) F)

TOTAL 30-j FRCRUITMENT PERCENT OP M

C X NIA20

I) l(J)

2- PHIUPPINKS PseekUASkmUIINUZ~

One Year One Year

Figure 3Selected circum-tropicaj probacted andor contiuous rmcitment patterns for tenfamilies of marine fishes and invertebrates Sources of data Victor 1983 (A) McFarland atal 1985 (6) Luckhurst ampLuckhurst 1977 (CDEF) Cushing 1977 (GH) Pauly 1982 (I)Willman amp Garcia 1985 (J)

growth and decay properties of population cohorts is fundamental to manyfishery management decisions as well as to population modeling (Alverson and Carney 1975)

One technique 7-sir analyzing the probable adaptive significance ofreproductive strategies or tactics is the of mathematical models Ifuse continuously recruiting populations do respond differently to exploitation thandiscrete-pulsed recruiting populations then use of temperate-based modelingand decision-making to determine optimal management strategy will likely beinfluenced If the relationships are simple enough then the application of closed

375

Prowedings of the 39th

solution analytical models might be possible However real systems aregenerally far too complex for smoothed-over analytical evaluation Thereforethe system must be studied utilizing numerical models Because of the strongdifferences between population reproductive behavior in the temperate zones versus the tropical zones witi regard to discrete versus coatinuous response analternative approach utilizing continuous mathematical pzocesses wasdeveloped Functions of continuous variables are the principal means of lookingmore deeply into demographic matters (Keyfitz 1977) A simulation modelCORECS (COntinuous-time RECruitment Simulation) was developed (Figure4) Specialized model features include the ability to divide the time stream upinto as many finite intervals as desired and allow the introduction of cohorts among any andor every finite time step Since cohorts in tropical latitudes are produced more or less in a continuum relative to the disciete once annualrecruitment of temperate based fishes population structure is simulatednumerically by multiple production of cohorts by species group within any one year and allowed to be constant or variable among years Because of thisscaling feature then probability density functions peculiar to the variousaforementioned classes of multiple-annual recruitment growth and mortalitymay be introduced Thus simulated populations with varying lengths ofprotracted spawning and variable density-dependent [ecruitment rates but withknown catches-at-ages (Lc sizes) can be parametrized in the algorithm Thisallows simulations to fie conducted which contrast disrete-event versuscontinuous-event approa ches We have investigated the outcomes of the woclasses of results on predictions rendered by traditional yield models (ieanalytical and dynamic pool) with special reference to whether the advicederived from these predictions is conservative or nonconservative relative to risk to the resource

The Simulation The mathematical description of continuous recruitment in marine fishpopulations has an affinity to periodic funcliens The magnitudes Uf these

functions are generally not known but iferred The magnitude can simply beviewed as a scaling feature The pure form of the function is desired in order toestablish its effects on population structure at size under various recruitmenthypotheses Three such plausible annual recruitment structures are presentedhere

1A uniform time-step-constant recruitment 2 A unimodal-annual recruitment 3 A bimodal-annual recruitment (Figure 5)For a simulated Gruperoid (=grouper-like) population characterized bylow natural mortality slow growth and thirteen year-classes the population

length-frequency distribution is shown for the deterministic uniform andunimodal recruitment hypotheses (Figure 6) Note the fundamental differencesin expectation of individuals at size (or age) under the two hypotheses Thepooled length frequency distributions for two recruitment hypotheses and twodifferent pooling schemes one by 20 mm increments and two by 40 mmincrements are shown (Figure 7) Note the roughly constant exponentialdecline for the constant uniform hypothesis (Panel A) with no apparentdiscernment of age stncture Clearly recruitment even in tropical fishes mustoscillate seasonally since year-round constant recruitment would generate size

376

Gulf and Caribbean Fisheries Inslftute

CORECS - Continuous-Time Recruitment Simulation Model

I 2 4 N

J1 21 31 41 bull bull bull NJ RIj

2 12 22 32 bull bull N-12

3 13 23 bull bull bull N-33

4 14 bull N-45

5

N

IN

where N (oldest age in the cach) z (division of time stream)I a time stream j a length (or time) class

and Rij a recruitment in time stream i for length clss j a I

- 00 i

Flgure 4General stucture of the continuous-time simulation model CORECS showing nshydimensonal upper angular structure

frequency distributions lacking peaks and troughs Note that for the unimodalannual recruitment models (Figure 7 Panels B and C) modes are obvious atsmaller sizes but these become obliterated with increased size (=age) andorincreased pooling This suggests that estimation procedures can be biased byimproper pooling andor dispropo-tional temporal or areal sampling Potentialyield surfaces for the grouperoids under

377


100 tft RECRi 20I N (C

1000 t

01I1 11 t

0 9

2~~ ~ Dest-eedn rerimn ar hwaFgr )(A) (8 ihe ind nfon mdaad()

FIureT reuimentrte etment remain sttaicconn s

regardless of the level of population abundance (Figure 8 Panel A) The density-dependent recruitment on the other hand suggests that yields areconstrained by a plane that cuts the surface obliquely at high revels of fishing

mortality andor size of capture Additionally the global maximu potend~alyield shifts to lower levels of flashing mortality and higher levels of age of

378

Gulf and Carbbean Flerles Inatitute

90shy

-I 7-

I- NUMBERS IN IjPOPULATION I I

(1deg2) I

-igiIit e Ig I

a- II it Ig

10 20 30 408 s 0 0 0 140 3 LENGTH (cm)

Figure S Simulated populadon length frequency struicture for a Grouperoid stock with an instantaneous natural mortality rate of M - 02 von Bertalanfly growth coeffiient of K shy

0122 and the oldest ago in the catch of t = 13 years for two hypothetical recruitment patshyterns (1)uniform constant disbibuted monthly and (2)unimodai with an annual period

capture When we cut through a plane parallel to the age of first capturecorresponding to tp=7 under the density-dependem hypothesis thestationary and the dynamic trends in yield and profit are shown (Pigure 8 PanelB) Note that trends in increased fishing mortality cause the yield and profit to exceed the equilibrium surface and subsequently excessive levels of fishingmortality cause both yield and profit to fall below the stationary surfaceStatistical fits of the dynamic trends have potential for overestimatingsustainable yields profits and fishing effort This condition can be exaggeratedby continuous but seasonally periodic recruitment If the integrated area underthe three recruitment functions is identical the fishing mortality operates at a constant level on the stationary population throughout the year and the samplingof the catch is absolute then the annual yield per annual recruits is identicalunder all three cases However if fishing mortality growth andor natural mortality is seasonal or varies through the year then the resulting estimates can be significantly biased

If we examine the potential yield surface of a Pelagoid population (highnatural mortality high growth coefficient ten year-classes) then we note quite adifferent somewhat pinnacle-shaped surface under a constant recruitmenthypothesis (Figure 9 Panel A) This suggests a more critical balance between the optimization of yield through size of capture and the respective fishingmortality disregarding abiotic influences For a simulated population structure

379

ProceedIngs of the 39th

(A)

f

80I-

C(B)

Is 100

ILENGTH (ema)

a (C)H (m

Figure 7 Grouperoid population length frequency distributional structure under several reshycruitment and pooling hypotheses (A) uniform constant monthly recruitments pooled by 20mm size class Increments (B)unimodal annual recruitment pooled by 20 mm size class inshycrements and (C) unimodal annual recruitment pooled by 40 mm size class increments

under the three annual recruitment hypotheses we note that the rate of decline ofpopulation numbers is much greater than that in the grouperoids and that themodal distribution corresponding to annual lumped cohorts are stretched outalong the abscissa their discernment diminishes rapidly with increased size orage (Figure 9 Panel B) This suggests that seasonal mortality and sub-optimal

380

Gulf and Canrlbban Fisheries Inutltute

IA) Ie)

Dtsa- Ne1

i S--0 t L be i

aim

Figur dSiuaerueodpplto()Tredmninlptnilyedsrlc

Figure uSimulated Grouperoid population (A)Three dmenslonaI potential yield surfaces using the (1) constant and (2) density-dependent recruitment hypotheses and (B)Therelationships between (1) stationary (eq) and dynamic yields and fishing mortality and (2)stationary (eq) and transitional profit and fishing mortality

381


catch sampling can induce significant bias into the estimation of stock dynamicsand potential Additionally there may be a critical balance of stock numbers inthe older size groups to ensure sufficient reproductive input to the stock

Several estimators have been proposed for total instantaneous mortalityestimation from information on average length in the catch (Beverton and Holt1956 Ssentongo and Larkin 1973) Both Beverton and Holt (1956) and theSsentongo and Larkin (1973) estimators assume an infinite life-span and both are biased even at equilibrium (Figure 10 Panel A) We introduce a modifiedformulation which utilizes a distribution truncated to the oldest age observcd inthe catch and which significantly reduces bias at low to intermediate levels of Zin a stationary population

Our formulation solves as

Ls-- LX Z(L - L) + K(L- L)

where Z = total instantaneous mortality rate

= smallest size in the catch L = average size in the catch LX = largest size in the catchand L K are parameters of the von Bertalanffy growth equation

If various rates of increased fishing mortality are applied to thesepopulations initially in equilibrium the induced perturbations cause a significantpropagation of bias (Figure 10 Panel B) This is particularly exacerbated whenthe dimension of continuous or protracted recruitment is operated in a seasonallybased fishery Thus attention should be focused on methodology that accountsfor the particular features of the life history in question and for transitionaldynamics in fishing effort and life history phenomena

SUMMARYPresently it is not possible to predict population fluctuations in single ormultispecies assemblages of fishes Without such an understanding it isgenerally impossible for us to tell whether a decline in a fish population resultsfrom intrinsic population pressure environmental change fishing or somecombination of these events The study of recruitment patterns such as brieflysketched here seems very promising Most probably recruitment patterns willamong other things allow for a quantitative evaluation of the relative impact ofvarious environmental factors affecting the recruitment of commercial andrecreationally important tropical marine fish and the optimization of fishingactivities in these regions

LITERATURE CITEDAllen GR 1975 Damselfshes of the South Seas Tropical Fish Hobbyist

Neptune New JerseyAlverson DL and MJ Carney 1975 A graphic review of the growth and

decay of population cohorts J ConsInt ExplorMer 36(2) 133-143Bayliff WH 1980 Synopses of biological data on eight species of scrombids Inter-Amer Trop Tuna Comm Spec RpL No 2 530p

V 382

Gulf and CarIlbean Flsherles Institute

(A) $I- I

S

-305lL

t

I i

So

I

4 -- 251

t0

-31018 -- - u

Figure 10 Performance of total Instantaneous mortality rate estimators from average size of fish in the catch under (A) stationary and (B)transitional conditions

Barret I and GV Howard 1961 Studies of the age growth sexual maturityand spawning of populations of anchoveta (Centengraulismysticetus)off the coast of the eastern tropical Pacific Ocean IATTC Bull V (2)113-216

Beverton RJH and SJ Holt 1956 A review of methods for estimating mortality rates in exploited fish populations with special reference to sources of bias in catch sampling RappP-V R CIEM 14067-83

383


RHJ------and SJ Holt 1957 On the dynamics of exploited fishpopulations FishInvest MinistAgric FishFood UK (Series 2) No19533pBoardman C and D Weiler 1980 Aspects of the life history of threedeepwater snappers around Puerto Rico Proc Gulf CaribFish Inst32158-172Bradley E and CE Bryan 1975 Life history and fishery of the red snapper(Lutjanus campechanus) in the northeastern Gulf of Mexico ProcGulf CaribFishInst 2777-106Brouard F and R Grandperrin 1984 Les poissons profonds de la pintarecifale externe a Vanatu ORSTOM Notes Doc DOceanogr

1171-79Bye VJ 1984 The role of environmental factors in the timing of reproductivecycles In Fish Reproduction Strategies and Tactics GW Potts ampRJ Wooten (eds) pp 187-205 Academic Press New YorkCushing DH 1970 The regularity of the spawning season of some fishesConsInt Explor Mer 33(1)81-97

J 1975 DH ------ Marine Ecology amp Fisheries Cambridge Press

CambridgeDeMartini EE and RK Fountain 1981 Ovarian cycling frequency and batchfecundity of the queenfish Seriphus politus Attributes representativeof serial spawning fishes FishBull US 79547-560DeVries DA and ME Chittenden Jr 1982 Spawning age determinationlongevity and mortality of the silver seatroLut Cynoscion nothus in theGulf of Mexico FishBull 80(3)487-500Dougherty P 1983 Diel lunar and seasonal rhythms in the reproduction oftwo tropical damselfishes Pomacentrusflavicaudaand P wardi MarBio 75215-224Eckert CJ 19914 Annual and spatial variation in recruitment of labroid lishesamong seven reefs in the CapricornB unker Group Great Barrier RecfMarBio 78123-127Fischer EA 1981 Sexual allocation in a simultaneously hermaphroditic coral

reef fish Amer Nat 117(1)64-82Forsberg ED 1980 Synopsis of biological data on the skipjack tunaKatsuwonus pelamis in the Pacific Ocean Inter-Amer Trop TunaComm Spec Rpt 2295-360Garcia S and L LeReste 1981 Life cycles dynamics exploitation andmanagement of coastal penaeid shrimp stocks FAO Fish Tech Pap203215 ppGilmore RC CJ Donohue and DW Cooke 1983 Observations on thedistribution and biology of east-central Florida populations of thecommon snook Centropomus undecimalis (Bloch) Fla Sci SpecSuppl Issue 45(4) Part 2Grimes CB 1986 Reproductive biology of the Lutjanidae a reviewSnapper-grouper workshop S Ralston and J Polovina (eds)Honolulu May 1985 p 239-294Guliand JA 1983 Fish Stock Assessment A Manual of Basic MethodsWiley-Interscience Publ 223 p

384

- --------

F Gulf and Caribbean Fisheries Insitute

Hislop JRG AP Robb and JA Guild 1978 Observations on the effects offeeding level on growth Ind reproduction in haddock Melanogrammusaeglefinus in captivity JFishBiol 1385-98

Hines AH KE Osgood and JJ Miklas 1985 Semilunar reproductivecycles in Fundus heteroclitus (Pices Cyprinodontidae) in an area without lunar cycles FishBull 83(3)467-472

Hunter JR and R Leong 1981 The spawning energetics of female northern anchovy Engraulis mordax FishBull 79(2)215-230

JR and BJ facewicz 1985 Measurement of spawning frequencyin multiple spawning fishes In R Lasker (ed) An egg productionmethod for estimating spawning biomass of pelagic fish Applicationto the northern anchovy Engraulis mordax pp 79-94 US DepCommer NOAA Tech Rep NMFS 36 JR NCH Lo and RJH Leong 1985 Batch fecundity in multiplespawning fishes In R Lasker (ed) An egg production method forestimating spawning biomass of pelagic fish Application to thenorthern anchovy Engraulismordax pp 67-77 US Dep CommerNOAA Tech Rep NMFS 36Johannes RE 1978 Reproductive strategies of coastal marine fishes in the tropics Env Biol Fish3(1)65-84

Jones R 1981 The use of length composition data in fish stock assessments(with notes on VPA and cohort analysis) FAO Fish Circ No 734 55 PP

- ------R 1984 Assessing the effects of changes in exploitation patternusing length composition data (with notes on VPA and cohort analysis)FAO Fish Tech Pap 256 118 pp

Kaini HT and II Ikehara 1976 Notes on the annual siganid harvest in Guam Microneia 12323-325

Keyfitz N 1977 Introduction to the Mathematics of PopulationsAddison-Wesley Publ Co Menlo Park California 490 ppKnudsen PK 1977 Spawning of yellowfin tuna and the discrimination of subpopulations IA7TCBull 17(2)

Lasker R (ed) 1981 Marine FishLarvae Univ Wash Press Seattle WA 131 pp

------- R and PE Smith 1977 Estimation of the effects of environmentalvariations on the eggs and larvae of the northern anchovy CalifCoopOceanicFishInvest Rep 178128-137

Lobel PS 1978 Diel lunar and seasonal periodicity in the reproductivebehavior of the Pomacanthid fish Centropygepotteri and some otherreef fishes in Hawaii PacSci 32(2) 193-207

PS and A R Robinson 1986 Transport and entrapment of fishlarvae by ocean mesoscale eddies and currents in Hawaiian waters J Deep-SeaRes 33(4)483-500

Luckhurst BE and K LuckhursL 1977 Recruitment patterns of coral reeffishes on the fringing reef of Curagao Netherlands Antilles Can JZool 55681-689

Manickchand-Dass S 1987 Reproduction age and growth of the lane snapperLutjanus synagris in Trinidad West Indies Bull Mar Sci (1)234-240

38S

40


Matsumoto WM 1966 Distribution and abundance of tuna larvae in thePacific Ocean Proc Gov Conf CenirPacFish Res 221-230 State of Hawaii 1966266 pp

Maynard Smith J 1976 Evolution and the theory of games Amer Sci 6441shy45

McFarland WN EB Brothers JC Ogden MJ Shulman EL Berminghamand NM Kotchian-Prentiss 1985 Recruitment patterns in youngFrench grunts Haemulon flavolineatum (Family Haeniulidae) at StCroix Virgin Islands FishBull 83(3)413-426

McKaye KR 1977 Competition for breeding sites between the cichlid fishesof Lake Jilob Nicaragua Ecology 58291-302

Mizenko D 1984 The biology of the western Samoan reef-slope snapperpopulations MS Thesis Univ Rhode IslandMolles MC 1978 Fish species diversity on model and natural reef patchesexperimental insular biogeography EcolMonogr 48289-305Munro JL 1982 Some advances and developments in coral reel fisheriesresearch 1973-1982 ProcGulf CaribFishInst 35 161-178JL -VC Gaut R Thompson and PH Reeson 1973 The spawning

seasons of Caribbean reef fishes JFishB~iol 569-84Murphy GI 1982 Recruitment of tropical fishes p 141-148 In Pauly D

and GI Murphy (cLs) Theory and management of tropical fisheriesICLARM Conference Proceedings 9 360 pp Internatiojl Center forLiving Aquatic Resources Management Manila Phifippines andDivision of Fisheries Research Commonwealth Scientific andIndustrial Research Organization Cronulla Australia

Nzioka RM 1979 Observations on the spawning seasons of east African reeffishes JFishBio 14329-342

Nikolsky GV 1963 The ecology of fishes Academic Press NY 352 ppOlson DB and RH Backus 1985 The concentrating of organisms at fronts a coldwater fish and a warm-core Gulf Stream ring J Mar Res4313-137

Otsu T and R Uchida 1959 Sexual maturity and spawning of albacore in thePacific Ocean U S FishBull 59287-305

Parrish BB editor 1978 Fish stocks and recruitment Rapp P-V Reun Cons lnt ExplorMer 164Pauly D 1982 Studying single-species dynamics in a tropical multispeciescontext p 33-70 In Paul D and GI Murphy (eds) Theory and management of tropical fisheries ICLARM Conference Proceedings 9360 pp International Center for Living Aquatic ResourcesManagement Manila Philippines and Division of Fisheries ResearchCommonwealth Scientific and Industrial Research OrganizationCronulla Australia

D and------NA Navaluna 1983 Monsoon-induced seasonality in therecruitment of Philippine fishes FAO Fi-heries Rept 291(3)823-833

Qasim SZ 1956 Time and duration of the spawning season in some marineteleosts in relation to their distribution J Cons Int Explor Mer 21144-154

$2 ---i973 An appraisal of the studies on maturation and spawning inmarine teleos from the Indian waters IndianJ Fish20(l) 166-181

386


Ralston S 1981 Aspects of the reproductive biology and feeding ecology ofChaetodonmiliarisa Hawaiian endemznX buterflyfish Env Biol Fish 6167-176

Robertson DR and RR Warner 1978 Sexual patterns in the labroid fishesof the western Caribbean II The parrotfishes (Scaridae) Smithsonian Contributions to Zoology No 255 26 ppRoss RM 1978 Reproductive behavior of the anemonefish Amphiprionmelanopuson Guam Copeia 1978103-107

- RM 1983 Annual semilunar and diel reproductive rhythms in the Hawaiian labrid Thalassomaduperrey MarBiol 72311-318Russell BC GRV Anderson and FH Talbot 1977 Seasonality andrecruitment of coral reef fishes Aust J Mar Freshwater Res 28521-528

Sale PF 1980 The ecology of fishes on coral reels OceangrMarBid Ann Rev 18367-421

- PF 1982 The structure and dynamics of coral reef fish communities p 241-253 In Paul D and GI Murphy (eds) Theory and management of tropical fisheries ICLARM Conference Proceedings 9360 pp International Center for Living Aquatic ResourcesManagement Manila Philippines and Division of Fisheries ResearchCommonwealth Scientific and Industrial Research OrganizationCronulla Australia

Schulman MJ JC Ogden JP Ebersole WN McFarland SL Miller andNG Wolf 1983 Priority effects in the recruitment of juvenile coral reef fishes Ecology 64(6)1508-1513

Ssentongo GW and PA Larkin 1973 Some simple methods of estimatingmortality rates of exploited fish populations J FishRes Bd Can 30 (5)695-698

Thresher RE 1984 Reproduction in reef fishes TFH PublicationsNeptune City New Jersey 399 ppVictor BC 1983 Recruitment and population dynamics of a coral reef fish Science 219419-210

Wallace JH 1975 Investi Rep Oceanogr Res Inst Durban 411-51Ware DM 1984 Fitness of ditferent reproductive strategies in teleost fishes

In Fish Reproduction Strategies amp Tactics GW Potts amp RJWooten (eds) pp 349-366 Academic Press New York

Warner RR and DR Robertson 1978 Sexual patterns in the labroid fishesof the western Caribbean I The wrasses (Labridae) Smithsonian Contributionsto Zoology No 25427 pp

Watson W and JM Leis 1974 Icthyoplankton in Kaneohe Bay Hawaii A one year study of eggs and larvae Univ Hawaii Sea Grant Tech ReptTR-75-01 Honolulu 178 pp

Williams DMcB 1983 Daily monthly and yearly variability in recruitmentof a guild of coral reef fishes MarEcolProgSer 10231-237

Willman R and SM Garcia 1985 A bio-economic model for the analysis ofsequential artisanal and industrial fisheries for tropical shrimp (with a case study of Surinam shrimp fisheries) FAO Fish Tech Pap(270)49 pp

387

Yamamoto T 1976 Seasonal variations in abundance size compositions anddistributional patterns of residing damselfishes in Sesoko IslandOkinawa Sesoko Mar Sci Lab Tech Rept 419-41

Yoshida HO 196amp Early life history and spawning of the albacore Thunnus alalungain Hawaiian waters US FishBull 69205-211

In cooperation with the United States Agency for Intemutlonai Development (Grant No DAN-4146-G-SS507100) the Fisheries Stock Assessment CRSP involves the following participating institutions

The University ef Maryland-Center for Enttronmental aid Estuarine Studies The University of Rhode Island-International Center for Marine Resource DevelopmentThe University ot Washington-Center for Quantitative Sciences The University of Costa Rica-Centro de Investigaci6n en Cienclas del Mar y Limnologfa The Univerrilty of the Philippines-Marine Science Institute (Di- nan)-College of Fishshy

eries (Vie ayas)

In collaboration with The University of Delaware The University of Maryland-Collegeof Business and Management The University of Miami and The International Centew for Living Aquatic Resources Management (ICLARM)


Working Paper 83



by


June 1991












Aliami hlorida33149



















362






363











364


(A) agcO

FREQUENCY age I

ae III-

LEOTH



365



Family


3


3 4



Spawns


Oct-May


Jan-June Oct-May





Mar-Dec Mar-Dec

year-round

Recruits



LocatlonSource












CA

Table 1(continued)


T albacares


3 Skipjack



Scomber japonicus


2 3

4

Engraulidae

Spawns

year-round

multiple spawning






May-Nov (bimodal)

8 months


Recruit




from equator


bimodel

bimodal


LocationSource










Table I (continued)

Family

Clupeidae

Haemulidae


2 Chromis 3 4


Gobidae 1 2



3

Spawns

Aug-Feb (2 peaks

Aug-Sep ampDec-Jan)

year-round

year-round

year-round

year-round (2 peaks

Sep-Oct ampJan-Feb)

Oct-Apr

year-round


OctAay

Recruits

Nov-Dec Larvae




bimodalSep ampApr

mid-wintrune-Aug

year-round

bimodal-mulL modes


LocationJSource









i




Table 1 (continued)

Family Grammilae




=L

Sr


Pomadasyidae

Holocantridae

Mullidae 1

2 Chaetodontjdae

Acanthundae

Balistidae





July-May


Oct Feb Apr-Aug

year-round

July-Mar











2

0












370






371









Class UI Ill


372






373






20shy


20shy

(E)

30jCARWDAE- jM ICA

ONE YEAR



374

_ _ __



30shy

10i __ __ _ I __ _

CUNACAG km CUACAS

jGaM~idld

p I I

(E) F)


C X NIA20

I) l(J)


One Year One Year




375







376



I 2 4 N


2 12 22 32 bull bull N-12


4 14 bull N-45

5

N

IN



- 00 i



377



1000 t

01I1 11 t

0 9





378


90shy

-I 7-


(1deg2) I

-igiIit e Ig I

a- II it Ig

10 20 30 408 s 0 0 0 140 3 LENGTH (cm)





379


(A)

f

80I-

C(B)

Is 100

ILENGTH (ema)

a (C)H (m



380


IA) Ie)

Dtsa- Ne1

i S--0 t L be i

aim



381





Ls-- LX Z(L - L) + K(L- L)








V 382


(A) $I- I

S

-305lL

t

I i

So

I

4 -- 251

t0

-31018 -- - u




383







384

- --------
















38S

40
















386
















387




Working Paper 83



by


June 1991












Aliami hlorida33149



















362






363











364


(A) agcO

FREQUENCY age I

ae III-

LEOTH



365



Family


3


3 4



Spawns


Oct-May


Jan-June Oct-May





Mar-Dec Mar-Dec

year-round

Recruits



LocatlonSource












CA

Table 1(continued)


T albacares


3 Skipjack



Scomber japonicus


2 3

4

Engraulidae

Spawns

year-round

multiple spawning






May-Nov (bimodal)

8 months


Recruit




from equator


bimodel

bimodal


LocationSource










Table I (continued)

Family

Clupeidae

Haemulidae


2 Chromis 3 4


Gobidae 1 2



3

Spawns

Aug-Feb (2 peaks

Aug-Sep ampDec-Jan)

year-round

year-round

year-round

year-round (2 peaks

Sep-Oct ampJan-Feb)

Oct-Apr

year-round


OctAay

Recruits

Nov-Dec Larvae




bimodalSep ampApr

mid-wintrune-Aug

year-round

bimodal-mulL modes


LocationJSource









i




Table 1 (continued)

Family Grammilae




=L

Sr


Pomadasyidae

Holocantridae

Mullidae 1

2 Chaetodontjdae

Acanthundae

Balistidae





July-May


Oct Feb Apr-Aug

year-round

July-Mar











2

0












370






371









Class UI Ill


372






373






20shy


20shy

(E)

30jCARWDAE- jM ICA

ONE YEAR



374

_ _ __



30shy

10i __ __ _ I __ _

CUNACAG km CUACAS

jGaM~idld

p I I

(E) F)


C X NIA20

I) l(J)


One Year One Year




375







376



I 2 4 N


2 12 22 32 bull bull N-12


4 14 bull N-45

5

N

IN



- 00 i



377



1000 t

01I1 11 t

0 9





378


90shy

-I 7-


(1deg2) I

-igiIit e Ig I

a- II it Ig

10 20 30 408 s 0 0 0 140 3 LENGTH (cm)





379


(A)

f

80I-

C(B)

Is 100

ILENGTH (ema)

a (C)H (m



380


IA) Ie)

Dtsa- Ne1

i S--0 t L be i

aim



381





Ls-- LX Z(L - L) + K(L- L)








V 382


(A) $I- I

S

-305lL

t

I i

So

I

4 -- 251

t0

-31018 -- - u




383







384

- --------
















38S

40
















386
















387







Aliami hlorida33149



















362






363











364


(A) agcO

FREQUENCY age I

ae III-

LEOTH



365



Family


3


3 4



Spawns


Oct-May


Jan-June Oct-May





Mar-Dec Mar-Dec

year-round

Recruits



LocatlonSource












CA

Table 1(continued)


T albacares


3 Skipjack



Scomber japonicus


2 3

4

Engraulidae

Spawns

year-round

multiple spawning






May-Nov (bimodal)

8 months


Recruit




from equator


bimodel

bimodal


LocationSource










Table I (continued)

Family

Clupeidae

Haemulidae


2 Chromis 3 4


Gobidae 1 2



3

Spawns

Aug-Feb (2 peaks

Aug-Sep ampDec-Jan)

year-round

year-round

year-round

year-round (2 peaks

Sep-Oct ampJan-Feb)

Oct-Apr

year-round


OctAay

Recruits

Nov-Dec Larvae




bimodalSep ampApr

mid-wintrune-Aug

year-round

bimodal-mulL modes


LocationJSource









i




Table 1 (continued)

Family Grammilae




=L

Sr


Pomadasyidae

Holocantridae

Mullidae 1

2 Chaetodontjdae

Acanthundae

Balistidae





July-May


Oct Feb Apr-Aug

year-round

July-Mar











2

0












370






371









Class UI Ill


372






373






20shy


20shy

(E)

30jCARWDAE- jM ICA

ONE YEAR



374

_ _ __



30shy

10i __ __ _ I __ _

CUNACAG km CUACAS

jGaM~idld

p I I

(E) F)


C X NIA20

I) l(J)


One Year One Year




375







376



I 2 4 N


2 12 22 32 bull bull N-12


4 14 bull N-45

5

N

IN



- 00 i



377



1000 t

01I1 11 t

0 9





378


90shy

-I 7-


(1deg2) I

-igiIit e Ig I

a- II it Ig

10 20 30 408 s 0 0 0 140 3 LENGTH (cm)





379


(A)

f

80I-

C(B)

Is 100

ILENGTH (ema)

a (C)H (m



380


IA) Ie)

Dtsa- Ne1

i S--0 t L be i

aim



381





Ls-- LX Z(L - L) + K(L- L)








V 382


(A) $I- I

S

-305lL

t

I i

So

I

4 -- 251

t0

-31018 -- - u




383







384

- --------
















38S

40
















386
















387











362






363











364


(A) agcO

FREQUENCY age I

ae III-

LEOTH



365



Family


3


3 4



Spawns


Oct-May


Jan-June Oct-May





Mar-Dec Mar-Dec

year-round

Recruits



LocatlonSource












CA

Table 1(continued)


T albacares


3 Skipjack



Scomber japonicus


2 3

4

Engraulidae

Spawns

year-round

multiple spawning






May-Nov (bimodal)

8 months


Recruit




from equator


bimodel

bimodal


LocationSource










Table I (continued)

Family

Clupeidae

Haemulidae


2 Chromis 3 4


Gobidae 1 2



3

Spawns

Aug-Feb (2 peaks

Aug-Sep ampDec-Jan)

year-round

year-round

year-round

year-round (2 peaks

Sep-Oct ampJan-Feb)

Oct-Apr

year-round


OctAay

Recruits

Nov-Dec Larvae




bimodalSep ampApr

mid-wintrune-Aug

year-round

bimodal-mulL modes


LocationJSource









i




Table 1 (continued)

Family Grammilae




=L

Sr


Pomadasyidae

Holocantridae

Mullidae 1

2 Chaetodontjdae

Acanthundae

Balistidae





July-May


Oct Feb Apr-Aug

year-round

July-Mar











2

0












370






371









Class UI Ill


372






373






20shy


20shy

(E)

30jCARWDAE- jM ICA

ONE YEAR



374

_ _ __



30shy

10i __ __ _ I __ _

CUNACAG km CUACAS

jGaM~idld

p I I

(E) F)


C X NIA20

I) l(J)


One Year One Year




375







376



I 2 4 N


2 12 22 32 bull bull N-12


4 14 bull N-45

5

N

IN



- 00 i



377



1000 t

01I1 11 t

0 9





378


90shy

-I 7-


(1deg2) I

-igiIit e Ig I

a- II it Ig

10 20 30 408 s 0 0 0 140 3 LENGTH (cm)





379


(A)

f

80I-

C(B)

Is 100

ILENGTH (ema)

a (C)H (m



380


IA) Ie)

Dtsa- Ne1

i S--0 t L be i

aim



381





Ls-- LX Z(L - L) + K(L- L)








V 382


(A) $I- I

S

-305lL

t

I i

So

I

4 -- 251

t0

-31018 -- - u




383







384

- --------
















38S

40
















386
















387








363











364


(A) agcO

FREQUENCY age I

ae III-

LEOTH



365



Family


3


3 4



Spawns


Oct-May


Jan-June Oct-May





Mar-Dec Mar-Dec

year-round

Recruits



LocatlonSource












CA

Table 1(continued)


T albacares


3 Skipjack



Scomber japonicus


2 3

4

Engraulidae

Spawns

year-round

multiple spawning






May-Nov (bimodal)

8 months


Recruit




from equator


bimodel

bimodal


LocationSource










Table I (continued)

Family

Clupeidae

Haemulidae


2 Chromis 3 4


Gobidae 1 2



3

Spawns

Aug-Feb (2 peaks

Aug-Sep ampDec-Jan)

year-round

year-round

year-round

year-round (2 peaks

Sep-Oct ampJan-Feb)

Oct-Apr

year-round


OctAay

Recruits

Nov-Dec Larvae




bimodalSep ampApr

mid-wintrune-Aug

year-round

bimodal-mulL modes


LocationJSource









i




Table 1 (continued)

Family Grammilae




=L

Sr


Pomadasyidae

Holocantridae

Mullidae 1

2 Chaetodontjdae

Acanthundae

Balistidae





July-May


Oct Feb Apr-Aug

year-round

July-Mar











2

0












370






371









Class UI Ill


372






373






20shy


20shy

(E)

30jCARWDAE- jM ICA

ONE YEAR



374

_ _ __



30shy

10i __ __ _ I __ _

CUNACAG km CUACAS

jGaM~idld

p I I

(E) F)


C X NIA20

I) l(J)


One Year One Year




375







376



I 2 4 N


2 12 22 32 bull bull N-12


4 14 bull N-45

5

N

IN



- 00 i



377



1000 t

01I1 11 t

0 9





378


90shy

-I 7-


(1deg2) I

-igiIit e Ig I

a- II it Ig

10 20 30 408 s 0 0 0 140 3 LENGTH (cm)





379


(A)

f

80I-

C(B)

Is 100

ILENGTH (ema)

a (C)H (m



380


IA) Ie)

Dtsa- Ne1

i S--0 t L be i

aim



381





Ls-- LX Z(L - L) + K(L- L)








V 382


(A) $I- I

S

-305lL

t

I i

So

I

4 -- 251

t0

-31018 -- - u




383







384

- --------
















38S

40
















386
















387













364


(A) agcO

FREQUENCY age I

ae III-

LEOTH



365



Family


3


3 4



Spawns


Oct-May


Jan-June Oct-May





Mar-Dec Mar-Dec

year-round

Recruits



LocatlonSource












CA

Table 1(continued)


T albacares


3 Skipjack



Scomber japonicus


2 3

4

Engraulidae

Spawns

year-round

multiple spawning






May-Nov (bimodal)

8 months


Recruit




from equator


bimodel

bimodal


LocationSource










Table I (continued)

Family

Clupeidae

Haemulidae


2 Chromis 3 4


Gobidae 1 2



3

Spawns

Aug-Feb (2 peaks

Aug-Sep ampDec-Jan)

year-round

year-round

year-round

year-round (2 peaks

Sep-Oct ampJan-Feb)

Oct-Apr

year-round


OctAay

Recruits

Nov-Dec Larvae




bimodalSep ampApr

mid-wintrune-Aug

year-round

bimodal-mulL modes


LocationJSource









i




Table 1 (continued)

Family Grammilae




=L

Sr


Pomadasyidae

Holocantridae

Mullidae 1

2 Chaetodontjdae

Acanthundae

Balistidae





July-May


Oct Feb Apr-Aug

year-round

July-Mar











2

0












370






371









Class UI Ill


372






373






20shy


20shy

(E)

30jCARWDAE- jM ICA

ONE YEAR



374

_ _ __



30shy

10i __ __ _ I __ _

CUNACAG km CUACAS

jGaM~idld

p I I

(E) F)


C X NIA20

I) l(J)


One Year One Year




375







376



I 2 4 N


2 12 22 32 bull bull N-12


4 14 bull N-45

5

N

IN



- 00 i



377



1000 t

01I1 11 t

0 9





378


90shy

-I 7-


(1deg2) I

-igiIit e Ig I

a- II it Ig

10 20 30 408 s 0 0 0 140 3 LENGTH (cm)





379


(A)

f

80I-

C(B)

Is 100

ILENGTH (ema)

a (C)H (m



380


IA) Ie)

Dtsa- Ne1

i S--0 t L be i

aim



381





Ls-- LX Z(L - L) + K(L- L)








V 382


(A) $I- I

S

-305lL

t

I i

So

I

4 -- 251

t0

-31018 -- - u




383







384

- --------
















38S

40
















386
















387




(A) agcO

FREQUENCY age I

ae III-

LEOTH



365



Family


3


3 4



Spawns


Oct-May


Jan-June Oct-May





Mar-Dec Mar-Dec

year-round

Recruits



LocatlonSource












CA

Table 1(continued)


T albacares


3 Skipjack



Scomber japonicus


2 3

4

Engraulidae

Spawns

year-round

multiple spawning






May-Nov (bimodal)

8 months


Recruit




from equator


bimodel

bimodal


LocationSource










Table I (continued)

Family

Clupeidae

Haemulidae


2 Chromis 3 4


Gobidae 1 2



3

Spawns

Aug-Feb (2 peaks

Aug-Sep ampDec-Jan)

year-round

year-round

year-round

year-round (2 peaks

Sep-Oct ampJan-Feb)

Oct-Apr

year-round


OctAay

Recruits

Nov-Dec Larvae




bimodalSep ampApr

mid-wintrune-Aug

year-round

bimodal-mulL modes


LocationJSource









i




Table 1 (continued)

Family Grammilae




=L

Sr


Pomadasyidae

Holocantridae

Mullidae 1

2 Chaetodontjdae

Acanthundae

Balistidae





July-May


Oct Feb Apr-Aug

year-round

July-Mar











2

0












370






371









Class UI Ill


372






373






20shy


20shy

(E)

30jCARWDAE- jM ICA

ONE YEAR



374

_ _ __



30shy

10i __ __ _ I __ _

CUNACAG km CUACAS

jGaM~idld

p I I

(E) F)


C X NIA20

I) l(J)


One Year One Year




375







376



I 2 4 N


2 12 22 32 bull bull N-12


4 14 bull N-45

5

N

IN



- 00 i



377



1000 t

01I1 11 t

0 9





378


90shy

-I 7-


(1deg2) I

-igiIit e Ig I

a- II it Ig

10 20 30 408 s 0 0 0 140 3 LENGTH (cm)





379


(A)

f

80I-

C(B)

Is 100

ILENGTH (ema)

a (C)H (m



380


IA) Ie)

Dtsa- Ne1

i S--0 t L be i

aim



381





Ls-- LX Z(L - L) + K(L- L)








V 382


(A) $I- I

S

-305lL

t

I i

So

I

4 -- 251

t0

-31018 -- - u




383







384

- --------
















38S

40
















386
















387





Family


3


3 4



Spawns


Oct-May


Jan-June Oct-May





Mar-Dec Mar-Dec

year-round

Recruits



LocatlonSource












CA

Table 1(continued)


T albacares


3 Skipjack



Scomber japonicus


2 3

4

Engraulidae

Spawns

year-round

multiple spawning






May-Nov (bimodal)

8 months


Recruit




from equator


bimodel

bimodal


LocationSource










Table I (continued)

Family

Clupeidae

Haemulidae


2 Chromis 3 4


Gobidae 1 2



3

Spawns

Aug-Feb (2 peaks

Aug-Sep ampDec-Jan)

year-round

year-round

year-round

year-round (2 peaks

Sep-Oct ampJan-Feb)

Oct-Apr

year-round


OctAay

Recruits

Nov-Dec Larvae




bimodalSep ampApr

mid-wintrune-Aug

year-round

bimodal-mulL modes


LocationJSource









i




Table 1 (continued)

Family Grammilae




=L

Sr


Pomadasyidae

Holocantridae

Mullidae 1

2 Chaetodontjdae

Acanthundae

Balistidae





July-May


Oct Feb Apr-Aug

year-round

July-Mar











2

0












370






371









Class UI Ill


372






373






20shy


20shy

(E)

30jCARWDAE- jM ICA

ONE YEAR



374

_ _ __



30shy

10i __ __ _ I __ _

CUNACAG km CUACAS

jGaM~idld

p I I

(E) F)


C X NIA20

I) l(J)


One Year One Year




375







376



I 2 4 N


2 12 22 32 bull bull N-12


4 14 bull N-45

5

N

IN



- 00 i



377



1000 t

01I1 11 t

0 9





378


90shy

-I 7-


(1deg2) I

-igiIit e Ig I

a- II it Ig

10 20 30 408 s 0 0 0 140 3 LENGTH (cm)





379


(A)

f

80I-

C(B)

Is 100

ILENGTH (ema)

a (C)H (m



380


IA) Ie)

Dtsa- Ne1

i S--0 t L be i

aim



381





Ls-- LX Z(L - L) + K(L- L)








V 382


(A) $I- I

S

-305lL

t

I i

So

I

4 -- 251

t0

-31018 -- - u




383







384

- --------
















38S

40
















386
















387



Table 1(continued)


T albacares


3 Skipjack



Scomber japonicus


2 3

4

Engraulidae

Spawns

year-round

multiple spawning






May-Nov (bimodal)

8 months


Recruit




from equator


bimodel

bimodal


LocationSource










Table I (continued)

Family

Clupeidae

Haemulidae


2 Chromis 3 4


Gobidae 1 2



3

Spawns

Aug-Feb (2 peaks

Aug-Sep ampDec-Jan)

year-round

year-round

year-round

year-round (2 peaks

Sep-Oct ampJan-Feb)

Oct-Apr

year-round


OctAay

Recruits

Nov-Dec Larvae




bimodalSep ampApr

mid-wintrune-Aug

year-round

bimodal-mulL modes


LocationJSource









i




Table 1 (continued)

Family Grammilae




=L

Sr


Pomadasyidae

Holocantridae

Mullidae 1

2 Chaetodontjdae

Acanthundae

Balistidae





July-May


Oct Feb Apr-Aug

year-round

July-Mar











2

0












370






371









Class UI Ill


372






373






20shy


20shy

(E)

30jCARWDAE- jM ICA

ONE YEAR



374

_ _ __



30shy

10i __ __ _ I __ _

CUNACAG km CUACAS

jGaM~idld

p I I

(E) F)


C X NIA20

I) l(J)


One Year One Year




375







376



I 2 4 N


2 12 22 32 bull bull N-12


4 14 bull N-45

5

N

IN



- 00 i



377



1000 t

01I1 11 t

0 9





378


90shy

-I 7-


(1deg2) I

-igiIit e Ig I

a- II it Ig

10 20 30 408 s 0 0 0 140 3 LENGTH (cm)





379


(A)

f

80I-

C(B)

Is 100

ILENGTH (ema)

a (C)H (m



380


IA) Ie)

Dtsa- Ne1

i S--0 t L be i

aim



381





Ls-- LX Z(L - L) + K(L- L)








V 382


(A) $I- I

S

-305lL

t

I i

So

I

4 -- 251

t0

-31018 -- - u




383







384

- --------
















38S

40
















386
















387



Table I (continued)

Family

Clupeidae

Haemulidae


2 Chromis 3 4


Gobidae 1 2



3

Spawns

Aug-Feb (2 peaks

Aug-Sep ampDec-Jan)

year-round

year-round

year-round

year-round (2 peaks

Sep-Oct ampJan-Feb)

Oct-Apr

year-round


OctAay

Recruits

Nov-Dec Larvae




bimodalSep ampApr

mid-wintrune-Aug

year-round

bimodal-mulL modes


LocationJSource









i




Table 1 (continued)

Family Grammilae




=L

Sr


Pomadasyidae

Holocantridae

Mullidae 1

2 Chaetodontjdae

Acanthundae

Balistidae





July-May


Oct Feb Apr-Aug

year-round

July-Mar











2

0












370






371









Class UI Ill


372






373






20shy


20shy

(E)

30jCARWDAE- jM ICA

ONE YEAR



374

_ _ __



30shy

10i __ __ _ I __ _

CUNACAG km CUACAS

jGaM~idld

p I I

(E) F)


C X NIA20

I) l(J)


One Year One Year




375







376



I 2 4 N


2 12 22 32 bull bull N-12


4 14 bull N-45

5

N

IN



- 00 i



377



1000 t

01I1 11 t

0 9





378


90shy

-I 7-


(1deg2) I

-igiIit e Ig I

a- II it Ig

10 20 30 408 s 0 0 0 140 3 LENGTH (cm)





379


(A)

f

80I-

C(B)

Is 100

ILENGTH (ema)

a (C)H (m



380


IA) Ie)

Dtsa- Ne1

i S--0 t L be i

aim



381





Ls-- LX Z(L - L) + K(L- L)








V 382


(A) $I- I

S

-305lL

t

I i

So

I

4 -- 251

t0

-31018 -- - u




383







384

- --------
















38S

40
















386
















387



Table 1 (continued)

Family Grammilae




=L

Sr


Pomadasyidae

Holocantridae

Mullidae 1

2 Chaetodontjdae

Acanthundae

Balistidae





July-May


Oct Feb Apr-Aug

year-round

July-Mar











2

0












370






371









Class UI Ill


372






373






20shy


20shy

(E)

30jCARWDAE- jM ICA

ONE YEAR



374

_ _ __



30shy

10i __ __ _ I __ _

CUNACAG km CUACAS

jGaM~idld

p I I

(E) F)


C X NIA20

I) l(J)


One Year One Year




375







376



I 2 4 N


2 12 22 32 bull bull N-12


4 14 bull N-45

5

N

IN



- 00 i



377



1000 t

01I1 11 t

0 9





378


90shy

-I 7-


(1deg2) I

-igiIit e Ig I

a- II it Ig

10 20 30 408 s 0 0 0 140 3 LENGTH (cm)





379


(A)

f

80I-

C(B)

Is 100

ILENGTH (ema)

a (C)H (m



380


IA) Ie)

Dtsa- Ne1

i S--0 t L be i

aim



381





Ls-- LX Z(L - L) + K(L- L)








V 382


(A) $I- I

S

-305lL

t

I i

So

I

4 -- 251

t0

-31018 -- - u




383







384

- --------
















38S

40
















386
















387













370






371









Class UI Ill


372






373






20shy


20shy

(E)

30jCARWDAE- jM ICA

ONE YEAR



374

_ _ __



30shy

10i __ __ _ I __ _

CUNACAG km CUACAS

jGaM~idld

p I I

(E) F)


C X NIA20

I) l(J)


One Year One Year




375







376



I 2 4 N


2 12 22 32 bull bull N-12


4 14 bull N-45

5

N

IN



- 00 i



377



1000 t

01I1 11 t

0 9





378


90shy

-I 7-


(1deg2) I

-igiIit e Ig I

a- II it Ig

10 20 30 408 s 0 0 0 140 3 LENGTH (cm)





379


(A)

f

80I-

C(B)

Is 100

ILENGTH (ema)

a (C)H (m



380


IA) Ie)

Dtsa- Ne1

i S--0 t L be i

aim



381





Ls-- LX Z(L - L) + K(L- L)








V 382


(A) $I- I

S

-305lL

t

I i

So

I

4 -- 251

t0

-31018 -- - u




383







384

- --------
















38S

40
















386
















387








371









Class UI Ill


372






373






20shy


20shy

(E)

30jCARWDAE- jM ICA

ONE YEAR



374

_ _ __



30shy

10i __ __ _ I __ _

CUNACAG km CUACAS

jGaM~idld

p I I

(E) F)


C X NIA20

I) l(J)


One Year One Year




375







376



I 2 4 N


2 12 22 32 bull bull N-12


4 14 bull N-45

5

N

IN



- 00 i



377



1000 t

01I1 11 t

0 9





378


90shy

-I 7-


(1deg2) I

-igiIit e Ig I

a- II it Ig

10 20 30 408 s 0 0 0 140 3 LENGTH (cm)





379


(A)

f

80I-

C(B)

Is 100

ILENGTH (ema)

a (C)H (m



380


IA) Ie)

Dtsa- Ne1

i S--0 t L be i

aim



381





Ls-- LX Z(L - L) + K(L- L)








V 382


(A) $I- I

S

-305lL

t

I i

So

I

4 -- 251

t0

-31018 -- - u




383







384

- --------
















38S

40
















386
















387











Class UI Ill


372






373






20shy


20shy

(E)

30jCARWDAE- jM ICA

ONE YEAR



374

_ _ __



30shy

10i __ __ _ I __ _

CUNACAG km CUACAS

jGaM~idld

p I I

(E) F)


C X NIA20

I) l(J)


One Year One Year




375







376



I 2 4 N


2 12 22 32 bull bull N-12


4 14 bull N-45

5

N

IN



- 00 i



377



1000 t

01I1 11 t

0 9





378


90shy

-I 7-


(1deg2) I

-igiIit e Ig I

a- II it Ig

10 20 30 408 s 0 0 0 140 3 LENGTH (cm)





379


(A)

f

80I-

C(B)

Is 100

ILENGTH (ema)

a (C)H (m



380


IA) Ie)

Dtsa- Ne1

i S--0 t L be i

aim



381





Ls-- LX Z(L - L) + K(L- L)








V 382


(A) $I- I

S

-305lL

t

I i

So

I

4 -- 251

t0

-31018 -- - u




383







384

- --------
















38S

40
















386
















387








373






20shy


20shy

(E)

30jCARWDAE- jM ICA

ONE YEAR



374

_ _ __



30shy

10i __ __ _ I __ _

CUNACAG km CUACAS

jGaM~idld

p I I

(E) F)


C X NIA20

I) l(J)


One Year One Year




375







376



I 2 4 N


2 12 22 32 bull bull N-12


4 14 bull N-45

5

N

IN



- 00 i



377



1000 t

01I1 11 t

0 9





378


90shy

-I 7-


(1deg2) I

-igiIit e Ig I

a- II it Ig

10 20 30 408 s 0 0 0 140 3 LENGTH (cm)





379


(A)

f

80I-

C(B)

Is 100

ILENGTH (ema)

a (C)H (m



380


IA) Ie)

Dtsa- Ne1

i S--0 t L be i

aim



381





Ls-- LX Z(L - L) + K(L- L)








V 382


(A) $I- I

S

-305lL

t

I i

So

I

4 -- 251

t0

-31018 -- - u




383







384

- --------
















38S

40
















386
















387








20shy


20shy

(E)

30jCARWDAE- jM ICA

ONE YEAR



374

_ _ __



30shy

10i __ __ _ I __ _

CUNACAG km CUACAS

jGaM~idld

p I I

(E) F)


C X NIA20

I) l(J)


One Year One Year




375







376



I 2 4 N


2 12 22 32 bull bull N-12


4 14 bull N-45

5

N

IN



- 00 i



377



1000 t

01I1 11 t

0 9





378


90shy

-I 7-


(1deg2) I

-igiIit e Ig I

a- II it Ig

10 20 30 408 s 0 0 0 140 3 LENGTH (cm)





379


(A)

f

80I-

C(B)

Is 100

ILENGTH (ema)

a (C)H (m



380


IA) Ie)

Dtsa- Ne1

i S--0 t L be i

aim



381





Ls-- LX Z(L - L) + K(L- L)








V 382


(A) $I- I

S

-305lL

t

I i

So

I

4 -- 251

t0

-31018 -- - u




383







384

- --------
















38S

40
















386
















387



_ _ __



30shy

10i __ __ _ I __ _

CUNACAG km CUACAS

jGaM~idld

p I I

(E) F)


C X NIA20

I) l(J)


One Year One Year




375







376



I 2 4 N


2 12 22 32 bull bull N-12


4 14 bull N-45

5

N

IN



- 00 i



377



1000 t

01I1 11 t

0 9





378


90shy

-I 7-


(1deg2) I

-igiIit e Ig I

a- II it Ig

10 20 30 408 s 0 0 0 140 3 LENGTH (cm)





379


(A)

f

80I-

C(B)

Is 100

ILENGTH (ema)

a (C)H (m



380


IA) Ie)

Dtsa- Ne1

i S--0 t L be i

aim



381





Ls-- LX Z(L - L) + K(L- L)








V 382


(A) $I- I

S

-305lL

t

I i

So

I

4 -- 251

t0

-31018 -- - u




383







384

- --------
















38S

40
















386
















387









376



I 2 4 N


2 12 22 32 bull bull N-12


4 14 bull N-45

5

N

IN



- 00 i



377



1000 t

01I1 11 t

0 9





378


90shy

-I 7-


(1deg2) I

-igiIit e Ig I

a- II it Ig

10 20 30 408 s 0 0 0 140 3 LENGTH (cm)





379


(A)

f

80I-

C(B)

Is 100

ILENGTH (ema)

a (C)H (m



380


IA) Ie)

Dtsa- Ne1

i S--0 t L be i

aim



381





Ls-- LX Z(L - L) + K(L- L)








V 382


(A) $I- I

S

-305lL

t

I i

So

I

4 -- 251

t0

-31018 -- - u




383







384

- --------
















38S

40
















386
















387





I 2 4 N


2 12 22 32 bull bull N-12


4 14 bull N-45

5

N

IN



- 00 i



377



1000 t

01I1 11 t

0 9





378


90shy

-I 7-


(1deg2) I

-igiIit e Ig I

a- II it Ig

10 20 30 408 s 0 0 0 140 3 LENGTH (cm)





379


(A)

f

80I-

C(B)

Is 100

ILENGTH (ema)

a (C)H (m



380


IA) Ie)

Dtsa- Ne1

i S--0 t L be i

aim



381





Ls-- LX Z(L - L) + K(L- L)








V 382


(A) $I- I

S

-305lL

t

I i

So

I

4 -- 251

t0

-31018 -- - u




383







384

- --------
















38S

40
















386
















387





1000 t

01I1 11 t

0 9





378


90shy

-I 7-


(1deg2) I

-igiIit e Ig I

a- II it Ig

10 20 30 408 s 0 0 0 140 3 LENGTH (cm)





379


(A)

f

80I-

C(B)

Is 100

ILENGTH (ema)

a (C)H (m



380


IA) Ie)

Dtsa- Ne1

i S--0 t L be i

aim



381





Ls-- LX Z(L - L) + K(L- L)








V 382


(A) $I- I

S

-305lL

t

I i

So

I

4 -- 251

t0

-31018 -- - u




383







384

- --------
















38S

40
















386
















387




90shy

-I 7-


(1deg2) I

-igiIit e Ig I

a- II it Ig

10 20 30 408 s 0 0 0 140 3 LENGTH (cm)





379


(A)

f

80I-

C(B)

Is 100

ILENGTH (ema)

a (C)H (m



380


IA) Ie)

Dtsa- Ne1

i S--0 t L be i

aim



381





Ls-- LX Z(L - L) + K(L- L)








V 382


(A) $I- I

S

-305lL

t

I i

So

I

4 -- 251

t0

-31018 -- - u




383







384

- --------
















38S

40
















386
















387




(A)

f

80I-

C(B)

Is 100

ILENGTH (ema)

a (C)H (m



380


IA) Ie)

Dtsa- Ne1

i S--0 t L be i

aim



381





Ls-- LX Z(L - L) + K(L- L)








V 382


(A) $I- I

S

-305lL

t

I i

So

I

4 -- 251

t0

-31018 -- - u




383







384

- --------
















38S

40
















386
















387




IA) Ie)

Dtsa- Ne1

i S--0 t L be i

aim



381





Ls-- LX Z(L - L) + K(L- L)








V 382


(A) $I- I

S

-305lL

t

I i

So

I

4 -- 251

t0

-31018 -- - u




383







384

- --------
















38S

40
















386
















387







Ls-- LX Z(L - L) + K(L- L)








V 382


(A) $I- I

S

-305lL

t

I i

So

I

4 -- 251

t0

-31018 -- - u




383







384

- --------
















38S

40
















386
















387




(A) $I- I

S

-305lL

t

I i

So

I

4 -- 251

t0

-31018 -- - u




383







384

- --------
















38S

40
















386
















387









384

- --------
















38S

40
















386
















387



- --------
















38S

40
















386
















387


















386
















387


















387



working paper series - pdf.usaid.govpdf.usaid.gov/pdf_docs/pnabk256.pdf · the university ef...

Documents