annually

1

Nottobecitedwithoutpriorreferencetotheauthors

InternationalCouncilforthe

ExplorationoftheSea

CM2000/N:13Paper

SpatialandTemporalPatterns

inRecruitmentProcess

SCALEANDPATTERNINRECRUITMENTPROCESSESOFBAYANCHOVY

INCHESAPEAKEBAY

UniversityofMarylandCenterforEnvironmentalScience,

ChesapeakeBiologicalLaboratory,

1WilliamsStreet,Solomons,MD20688,USA

ABSTRACT

Recruitmentofbayanchovy(Anchoamitchilli)variesannuallyinChesapeake

Bay,andlevelsandpatternsarerelatedtovariabilityinhydrologicalconditions

er-trawlsurveys,

conductedthreetimesannuallyfrom1995-1999,overtheentire320-kmlengthof

theBay,providedinformationonannualandregionalpatternsofrecruitment,and

iomassof

anchovywithintheBayatthebeginningofspawningasonsin1995-1999

variedsix-fold,butitalonewasnotdirectlyrelatedtotheyoung-of-the-year

(YOY)ofrecruitmentinOctoberwerelowin1995and

1996(6to7X109)buthigherin1997-1999(19to52X109).Animportant

featureoftherecruitmentprocessisanontogeneticmigrationinwhichYOYbay

anchovytendedtomoveupbayuntiltheyareapproximately45mmTL,after

ongsalinitygradientmayactasa

partialbarriertoupbayordownbaymigration,theeffectbeingmorepronounced

forsmall(<60mmTL)r,asonalwatertemperaturewasmore

importantindeterminingthelatitudinaldistributionofspawningstockbiomass.

Late-summerrecruitmentofYOYanchovywashighwhenwatertemperature

waslowinApril-May,inhibitingupbaymigrationofadultsattheontofthe

spawningason,andinsuringthatmostspawningoccurredinthelowerand

iedRickerstock-recruitmentmodelthat

includedthelatitudinalrangeofadultmigrationbetweenAprilandJuly,explained

98%ofrecruitmentvariability.

2

-77.00W-76.00W

37.00N

38.00N

39.00N

0

25

50

100

200

300

400

500

750

1,000

1,500

2,000

2,500

Upper

Middle

Lower

INTRODUCTION

BayanchovyAnchoamitchilli(Engraulididae)isacoastalspecies

distributedinthewesternAtlanticfromMainetotheGulfofMexico,andisthe

mostabundantandubiquitousfishinChesapeakeBay,thelargestestuaryon

theeastcoastofNorthAmerica(Houde&Zastrow,1991;Able&Fahay,1998).

sonzooplankton,primarily

copepodsandothersmallcrustacea,andispredatedbypiscivoresincluding

veraleconomicallyimportantfishspecies(Baird&Ulanowicz,1989;Luo&

Brandt,1993;Hartman&Brandt,1995).Malesandfemalescanmatureat40-45

mmFL(44-50mmTL)atca10monthsposthatch,andpeakspawningoccursin

July(Zastrowetal.1991).Mosteggsareproducedbyage1individuals(Luo&

Musick,1991;Zastrowetal,1991).Bayanchovycansurvivetoage3+and

tallengthsofage-specificbayanchovywere

reportedtobeage1=55mm,age2=74mm,andage3=88mm(Newberger

&Houde,1995).Temporalandspatialvariabilityinabundance,growth,and

mortalityrateswerereportedforChesapeakeBay(Wang&Houde,1995;Dory

etal.,1996,MacGregor&Houde,1996,Rilling&Houde,1999ab).Dovel(1971)

andLoos&Perry(1991)hypothesizedontogeneticmigrationtoexplainregional

variabilityinabundance.

Newberger&Houde(1995)notedthelargedifferenceinabundancesof

bayanchovyamongyears,apparentlyaresultofvariabilityinannual

r,therewaslittleknowledgeaboutfactorsthatcontrol

levelsofrecruitment,thesizethat

recruitmentofbayanchovyisatleastpartiallydeterminedbythespatial

distributionofspawners,andnotonlybyvariablegrowthandmortalityduring

early-lifestages.

ChesapeakeBayisthelargestestuaryintheUnitedStates.

Geomorphically,itisatypicalcoastalplainestuary(Dayetal.,1989).Its

mainstemis320km

long,varyinginwidth

fromabout6.4kmto50

km(Fig.1).

ChesapeakeBayis

an10%

oftheBayareais>18

mdeep,and

approximately50%is<

tributarienterthe

ChesapeakeBay.

Physically,theBayisa

partially-mixedestuary

(Dayetal.,1989).

Eightyto90%ofthe

eakeBay,andmeanannually-aggregatedbay

anchovycatch-per-unit-effort(wetweightingram/20

mintow)ntallinesindicate

boundariesofthethreeregions.

3

freshwaterenteringtheBayisfromtributariesonitsnorthernandwesternsides.

TheBayreceivesabouthalfitswatervolumefromtheAtlanticOcean,andthe

remainderdrainsintotheBayfroma166,000-km2drainagebasin(Chesapeake

BayProgram,2000).TheBay’ssalinitygradesfromnear-fullawateratits

mouthtofreshwateratitshead,

temperaturesmaybeashighas28oto30oCinmidsummer,andmayfallto1oto

4oCinlatewinter(Murdyetal.,1997).Despiteshallowdepth,theBay’s

mainstemusuallyhasastronglydevelopedpycnoclinewithsharpgradientsin

verticaltemperatureandsalinityprofiles.

Objectivesofthispaperare1)toevaluateeffectsofhydrological

conditionsonstage-specificdistribution,ontogeneticmigration,andrecruitment,

and2)todelineatepatternsinthebayanchovyrecruitmentprocess.

METHODS

Cruis

RearchcruisfortheLandMarginEcosystemRearch(LMER)

program,“TrophicInteractionsinEstuarineSystems(TIES)”

(/ties/)surveyedtheentireBayandwereconducted

threetimesannually(April-May,June-AugustandOctober)from1995to1999

(Fig.1).Planktonandfishcollections,inadditiontoobrvationsof

environmentalvariables,lections(midwatertrawl)and

hydroacousticsurveysinTIEScruiswereconductedatthreeorfourstations

pertranctinthelower-Bay(37o55’N-37o05’N),mid-Bay(38o45’N-37o55’N),

andupper-Bay(39o25’N-38o45’N).Thelowerbayincludes51%oftotal

baywidewatervolume,themiddlebay32%,andtheupperbay17%(Fig.1).

Numbersofmidwater-trawlstationsperTIEScruirangedfrom24to52.

Additionalcruisprovidedinformationinsomeperiodswhentherewereno

TIESsurveys;forexample,inJune1997,August1997,1998,andSeptember

1998.

An18-m2mouth-openingmidwatertrawl(MWT),with6-mmcodend

wasfishedfromthesternA-frame

wasfished

in2-minsteppedintervalsfromsurfacetobottomtofishtheentirewatercolumn.

Fishcatches(orsamples)werecounted,measuredandweighedondeck

immediatelyafteratow.

TheMWTwaffectiveincatchingbayanchovy>30mmtotallength.

Theamountofwaterfishedbya20-minMWTtowwasapproximately5,000m3.

Badonthisinformation,weexpandedtheCPUEvaluestoestimateregional

anchovyabundancesandwetweightsbymultiplyingwatervolumesofeach

region(Cronin,1971).

4

Environmentalfactors

DepthprofilesoftemperatureandsalinityweredeterminedfromCTD

etemperatureandsalinityinthispartiallymixedestuaryshow

distinctfeaturesbetweensurfaceandbottom,theywereintegratedoverthe

entiredepthafterdividingthewatercolumnintoasurfaceandbottomlayer

lowingstepswereappliedto

derivehomogenous(randomlydistributed)parametervaluesforenvironmental

conditions.

Foreachstation,thepycnoclinedepthwastimatedfromthederivative

ofthethird-orderpolynomialregressionfittedtosalinityondepth:

s=ax3+bx2+cx+d,

Wheres:salinity;x:depth(m);a<0,b>0.

Afterdefiningpycnoclinedepthateachstation,themeanwater-column

temperatureandsalinityineachregion(lower,middle,andupperBay)were

ferencesoftemperatureandsalinitybetween

cruiperiods(spring,summer,andfall),anddifferencesbetweentwoadjacent

regions(gradients)alsoweresummarized.

Exploratorystatisticalanalysis

Spatialdatausuallyshowstrongcorrelationsbetweensitessimilarto

auto-correlationintime-riesdata(Cressie,1993).Avariogramcanbederived

tointerpolatevalues(abundancesandbiomass)forgridsorstationswhereno

ropic(bothlatitudinalandlongitudinal)linear

variogramfunctionswithout‘nugget’effectwereudtoestimatebayanchovy

abundancesandbiomassbyinterpolationofvaluesforunsampledgridsof1x

iogramfunctionswere

derivedbyprocvariogramofSAS6.11,andthegriddatafilesweregeneratedby

prockrige2dtoproducedistributionmapsofbayanchovynumberandbiomass,

andhydrologicalvariablessuchastemperatureandsalinity(SAS,1998).

Modallengthsofbayanchovy

cohortsweredeterminedfrom

length-frequencydistributions.

Baywiderelativeabundanceofeach

cohortwastimatedby

Bhattacharyaplots(Bhattacharya,

1967;King,1995).Criteriaof

maximumtotallengthofYOYbay

anchovyareshowninTable1.

Weproducedlength

frequenciesbylatitudetodelineate

possibleontogeneticmigrationof

parameterizethedistributionofYOY

mtotallengthofyoung-of-

the-yearbayanchovy(mm).

YearDateLength(mm)

199523-Jul57

28-Oct78

199617-Jul57

22-Oct77

199711-Jul-

02-Aug57

29-Oct69

199804-Aug57

07-Sep62

19-Oct69

199926-Jun-

23-Oct75

5

andadultbiomass,weestimatedbiomass-weightedandabundance-weighted

meanlatitudesofoccurrenceforeachlengthbythefollowingformula.

L

bl

=∑B

nl

V

n

L

n

/∑V

n

B

nl

L

al

=∑A

nl

V

n

L

n

/∑V

n

A

nl

where,L

bl

:biomass-weightedmeanlatitudeofalength,l;L

al

:abundance-

weightedmeanlatitudeofalength,l;B:Biomass(ginwetweight)per20-min

tow;A:Abundance(number)per20-mintow;n:station;V

n

:watervolume(m3)of

theregionwherestationnislocated,

∑V=V

lowerbay

+V

middlebay

+V

upperbay

=26.668+16.840+8.664=52.112(x109

m3).

Inanexploratorystep,correlationanalysiswasudtoexamine

relationshipsofbayanchovyspawningandmigrationpatternswithregional-and

depth-layer-specificmeantemperature,meansalinity,theirgradientsor

differencesfromthepreviouscrui,andmonthlymeanstreamflowfromthe

ametersoftemperatureandsalinitywererelatedto

meanlatitudesofoccurrenceandbaywideabundance/biomassforYOYand

adultbayanchovy.

Ontogeneticmigration

AGeneralLinearModel(procGLMinSAS)wasudtoevaluatethe

importanceofenvironmentalfactorsonsize-dependentmigration:

L

ljk

=C

j

+τ

lj

+β

j

x+ε

ljk

where,L:meanlatitudeofoccurrenceforbayanchovyweightedbyabundance

orbiomass;l:length(mm)ofbayanchovy;j:ason(=1,2,3);k:year(=95,96,

97,98,99);C:constantterm;β:slopecoefficient;τ:effectofbody-size;x:values

foreachenvironmentalfactor(salinity,temperature,theirgradient,freshwater

input,etc.);ε:errorterm.

AfterdeterminingthemostimportantfactorfordistributionofYOYand

adultbayanchovy,regressionanalysisbadonquentialsumsofsquares

(typeISS)wasappliedtofittherelationshipbetweenthefactorandlength-

specificmeanlatitudeofdistribution,afterfilteringoutthesize-dependencyby

polynomialequations(procREGandGLMinSAS6.11:SAS,1990;Littelletal.,

1991).Thestatisticalmodelis:

L

jk

(l)=β

0

+β

1

l+β

2

l2+β

3

l3+β

4

x+β

5

(xl)+ε

ljk

(cubic)

where,β

0

:intercept;β

1

,β

2

,β

3

:slopecoefficenitsforthepolynomialequation;l:

totallength(mm)ofbayanchovy;x:valuesfortheenvironmentalfactor;(xl):

interactionterm;β

5

:slopecoefficientfortheinteractionterm;ε:errorterm.

Ifthehighestordertermortheinteractiontermwasnotsignificantatα=

0.05,wefitagainafterremovingthotermsfromthemodelequation,untilall

termsweresignificant.

6

Recruitmentmodels

Totestwhetherthereisasignificantrelationshipbetweenbayanchovy

spawningstockbiomass(SSB)inApril-MayandYOYrecruitmentinOctober,we

ationshipbetweentheasonal

changesinthespatialdistributionpatternofadultbayanchovyandYOY

recruitmentlevelsinOctoberwastestedbyregressionanalysis,afterexamining

thedistributionmapsofYOYandadultbayanchovydistributionforeachcrui.

Finally,twodifferentrecruitmentmodelsweretestedbadonthe

estimatedSSBinApril-Mayandthedifferencesinbiomass-weightedmean

,the

log-transformedYOYrecruitmentlevelinOctoberwasfittedbyastatistical

model:

Log(R

y

)=β

0

+β

1

S+β

2

∆L+ε

y

whereR

y

:recruitmentlevel=OctoberYOYabundanceintheyear,y;Β

0

:

intercept;β

1

andβ

2

:slopecoefficients;∆L:differenceinbiomass-weightedmean

latitudeofSSBindecimalunitsbetweenApril-MayandJune-August;S:baywide

spawningstockbiomass(male+female)intonsforApril-May,orforJune-

August;ε

y

:errorterm.

Thestatisticalmodelfittheobrvedrecruitmentlevelsquitewell,butwas

biologicallyunacceptable,

includeadensity-dependentterm,amodifiedRickermodelwasapplied(Ricker,

1954,1975):

R

y

=aSexp(–β

1

S-β

2

∆L+ε

y

)

log(R

y

)-log(S)=log(a)–β

1

S-β

2

∆L+ε

y

Forthismultivariaterecruitmentmodel,collinearitydiagnostictools

availableinprocREGofSAS,suchasvarianceinflationfactor(VIF)and

conditionindex(CI),nruleofthumbwasadopted:ifVIF>

10orCI>15,collinearitywasjudgedtobeproblematical.

RESULTS

Environmentalfactors

Monthlystreamflow

fromtheSusquehannaRiver

(Table2)showedhighannual

andasonalvariability.

Streamflowswerehighin

1996and1998,andlowin

1995,1997and1999.

almeanstreamflow(m3/s)

PeriodYear719981999

Jan.-Mar.1,2892,4951,4742,5631,325

Apr.-Jun.7281,7029201,625791

Jul.-Sep.238768239334294

Oct.-Dec.9232,23

Annualmean7951,7998451,179763

7

Watertemperatureandsalinityvariedannually,asonally,andregionally

(Table3).Annually,temperaturewashighestin1995,andlowestin1997.

ally,salinitywasmore

ally,temperaturewasmorevariablethan

aturewashighestintheJune-August,thespawningasonof

bayanchovy,tyincread

tywaslowestinApril-Mayinall

fficientofvariation(CV)forannualmeansalinitieswasabout2

timeshigherthanthatfortemperatures.

Trendsinabundanceandrecruitment

of

YOYrecruitmentinOctoberwerelowin1995and1996(6to7X109)buthigher

in1997-1999(19to52X

109).Baywideestimatesof

bayanchovybiomassfor

fish>30mmTLincread

fromApriltoOctober(Fig.2).

Octoberbaywidebiomass

washighestin1998(mean

±SE=38,000±4,100tons),

andlowestin1996

(5,000±1,100tons).

Spawningstock

biomass(SSB),estimated

asthebaywidebiomassin

April-May,waslowestin

1995(663tons),and

highestin1997(4,010tons).

TheSSB,atfirstglance,did

notshowanyapparent

relationshipwithOctober

YOYabundance(Fig.3-a).

TheSSBinJune-August,

estimatedfromage1+bay

anchovybiomass,alsodid

notshowanyobvious

relationshipwithOctober

YOYabundance(Fig.3-b).

mperatures(oC)andsalinities(psu)

integratedoversurfacetobottom,withpooled

standarderrors

PeriodTemperatureSESalinitySE

Cruidate

28-Apr-9514.080.0917.600.41

23-Jul-9528.040.1917.770.46

28-Oct-9517.710.0920.430.38

28-Apr-9613.240.1313.010.39

17-Jul-9624.570.1514.550.56

22-Oct-9616.610.1514.130.42

20-Apr-9710.960.0813.520.64

11-Jul-9725.460.1815.570.47

29-Oct-9715.000.0620.590.35

11-Apr-9812.200.0811.450.46

04-Aug-9825.860.0715.650.40

19-Oct-9819.080.0818.800.42

19-Apr-9912.090.1115.720.45

26-Jun-9923.230.1418.220.48

23-Oct-9916.310.1019.590.46

Year

199519.950.0718.600.24

199618.140.0913.900.26

199717.140.0616.560.28

199819.050.0515.300.25

199917.210.0717.840.27

CV6.6%11.5%

Season

April-May12.520.0514.260.21

June-August25.430.0716.350.22

October16.940.0518.710.19

Region

Lower18.401.4420.940.74

Middle18.291.5214.090.77

Upper17.991.577.170.78

8

Correlations

Correlationcoefficientsof

biomass-weightedmeanlatitudeofYOY

andadultbayanchovydistributionwith

salinity,temperature,andSusquehanna

RiverstreamflowareprovidedinTable

tygradientsdidnotshow

significantcorrelationswiththemean

latitudeofYOYandadultbayanchovy

difficulttointerpretthehighcorrelationcoefficientsbetweenYOYandadultbay

anchovyabundancesinOctoberwithsalinitychangeinthelowerBayfromApril-

MaytoJune-August,butitsuggestedthatOctoberrecruitmentmightbefixedin

theApriltoAugustperiod(Fig.4).Althoughsomevariablesweresignificantly

correlated,wewereunabletodetectanyconsistentpatternswithrespectto

ason,,SSBinApril-Maywasnotsignificantly

correlatedwiththepreviousyear’sYOYrecruitmentlevel,probablybecauof

plecorrelationsalonewerenot

sufficienttotestthehypothesis,becautherewereonly5yearsofobrvations,

andsignificantcorrelationscouldhaveoccurredbychance.

ruitmentofbayanchovyinOctober

plottedonsalinitychangeinthelowerBayfrom

April-MaytoJune-August.

ng

stockbiomassofbayanchovyin

ChesapeakeBay.

eestimateof>30mmTLbayanchovy

biomass(YOY+adult)intons.

0

10,000

20,000

30,000

40,000

50,000

719981999

APR

JUL

OCT

APR

JUL

OCT

APR

JUL

OCT

APR

AUG

OCT

APR

JUN

OCT

0

10000

20000

30000

40000

50000

60000

R

e

c

r

ui

t

s

(

mi

l

l

i

o

n

s

)

i

n

O

c

t

o

b

e

r

5

SSB(tons)inApril-May

95

96

97

98

99

0

10000

20000

30000

40000

50000

60000

R

e

c

r

ui

t

s

(

mi

l

l

i

o

n

s

)

i

n

O

c

t

o

b

e

r

040005000

SSB(tons)inJune-August

95

96

97

98

99

(a)

(b)

0

10

20

30

40

50

60

N

u

m

b

e

r

of

Y

O

Y

i

n

O

c

t

o

b

e

r

(

1

0

9

)

12345

Salinitydifference(psu)

1995

1996

1997

1998

1999

9

ationcoefficientsforbayanchovydistributionandabundancewithtemperature,salinity

iablesthatdidnotshowanysignificantcorrelationatα=

0.05arenotshownhere.

Explanationoftheabbreviationsforvariablenames:

1)Columnvariablenames

Thefirstcharacter:L=meanlatitudeofoccurrenceweightedbybiomass,N=baywideabundance,

B=biomass

Thethirdcharacter:Y=young-of-the-yearbayanchovy.A=age1+bayanchovy

Thefourthandfifthdigit:04=April-May,07=June-August,10=October

2)Rowvariablenames

Thefirstthreecharacters:SAL=salinity,TEM=watertemperature,

D_S=salinitydifferencefromthepreviouscrui,D_T=temperaturedifference

G_S=salinitygradientbetweentworegions,G_T=temperaturegradient

Thefourthandfifthdigit:04=April-May,07=June-August,10=October

Thesixthcharacter:L=lowerBay,M=middleBay,U=upperBay

Thelastcharacter:S=surface,B=bottomlayer

FLOW:monthlymeanstreamflowfromtheSusquehannaRiverfromJanuary(01)toOctober(10)

3)*=Significantatalpha=0.05,**=significantatalpha=0.01

VariableL_Y07L_Y10N_Y07N_Y10L_A04L_A07L_A10B_A04B_A07B_A10

SAL04LS0.860.430.50-0.760.580.720.60-0.07-0.18

-0.88*

SAL04US0.32-0.310.37-0.17

0.91*

0.090.020.120.41-0.36

SAL04UB0.27-0.320.28-0.13

0.93*

0.040.040.150.42-0.33

SAL10MB

-0.89*

0.47-0.710.38

SAL10UB

-0.94**

0.60-0.710.51

TEM04LS0.790.240.86-0.190.160.700.07

-0.93*

-0.66-0.22

TEM04MS0.910.680.65-0.53-0.03

0.93*

0.52-0.78

-0.89*

-0.54

TEM04US0.480.740.16-0.30-0.600.620.44-0.59

-0.92*

-0.19

TEM04LB0.930.35

0.90*

-0.480.350.830.27-0.73-0.55-0.54

TEM04MB

0.98*

0.510.88-0.630.15

0.91*

0.35-0.65-0.60-0.64

TEM07MS0.61-0.13

0.91*

-0.060.33-0.45-0.140.02

TEM07US0.49-0.29

0.92*

-0.040.24-0.490.08-0.02

D_S07LS-0.82-0.71-0.28

0.91*

-0.79-0.820.33

0.99**

D_S07MS-0.93-0.86-0.55

0.88*-0.95*

-0.620.660.81

D_S07LB-0.49-0.59-0.270.14-0.59-0.34

-0.95*

0.08

D_S07MB

-0.98*

-0.48

-0.90*

0.64-0.87-0.260.520.62

D_S10UB

-0.90*

0.62-0.700.58

D_T07MS-0.22-0.650.500.33-0.34

-0.88*

0.520.43

D_T07LB

-0.96*

-0.140.170.13-0.21-0.600.140.38

G_S04US0.800.870.41-0.62-0.35

0.89*

0.62-0.59

-0.93*

-0.55

G_T04UB0.02-0.580.260.14

0.92*

-0.19-0.220.130.56-0.05

G_S07UB

-0.96*

-0.36-0.670.37-0.510.130.400.18

FLOW010.170.53-0.03-0.02-0.730.320.21-0.52-0.790.12

FLOW02-0.65-0.06-0.600.55-0.71-0.47-0.230.00-0.150.67

FLOW03-0.88-0.42-0.470.72-0.62-0.71-0.620.100.210.86

FLOW04-0.450.07-0.560.51-0.51-0.280.00-0.21-0.400.56

FLOW05-0.270.30-0.310.20-0.06-0.03-0.470.38

FLOW06-0.440.150.02-0.02-0.07-0.33-0.080.24

FLOW070.180.420.12-0.010.27-0.02-0.640.18

FLOW080.73-0.360.41-0.17

FLOW090.84-0.460.83-0.43

FLOW10

0.97**

-0.680.79-0.61

10

Ontogeneticmigration

Thelength-specificmeanlatitudesofoccurrenceofbayanchovy,

weightedbytheirabundance,

anchovytendedtomoveupbayandwerelocatedprimarilyupbayuntiltheywere

approximately45mmTL,afterwhichtheybegantomovedownbay(Fig.5).In

April-May,smallage1bayanchovies<60mmTL,apparentlyrecruitedfromthe

previousyear,variedannuallyintheirmeanlatitudeofoccurrenceintheBay,

whereaslarge(>age1)bayanchovyhadrelativelystablelatitudinallocations

nearthelower-middleBayboundary(Fig.5-a).ComparedtoApril-May,age1+

bayanchovyinJune-Augustweremorevariableintheirannualmeanlatitudesof

occurrence,butbothYOYandadultbayanchovytendedtomoveupbay(Fig.5-

b).In1997and1999,whenannualmeantemperatureswerelowest(Table3),

YOYbayanchovyweretoosmalltobesampledbytheMWTinJune-August.

Themeanlatitudesfor30-60mmTLbayanchovyinApril-Mayshowednotonly

additive,butalsointeractive(sizexlatitude)rmean

latitudes(Fig.5-c)indicatedconsistentontogeneticmigrationpatterns,although

thereweresignificantadditive,annualdifferences,butwithoutinteractive

tprobableexplanationisthatYOYbayanchoviestendedto

moveupbay,andbegantomovedownbayatca.45mmTL.

TheSSB,i.e.,spawningstockbiomassofbayanchovy(excludesYOY),

from1995to1999wascenteredbetween37.30’N–38.00’NinApril-July(Fig.6-

a).However,inJune-August1995-1996,theSSBshiftedtowardtheupperBay,

whereasitsmeanlatitudinalpositionshardlydifferedbetweenApril-Mayand

e-AugustlatitudeofSSBwasstronglyand

significantlyrelatedtosurfacetemperatureinthemiddleBayinApril-May(r2=

86%,p=0.0233;Fig.6-b).Theregressionequationis:

Latitude=33.93+0.35T

where,T=Temperaturefor10.5-14.5oC.

Contourplotsofadultabundancewithrespecttosurfacetemperatureand

salinityshowedthatspawnersweremostlydistributedat10-12oCinApril-May

(Fig.7-a).InJuly1995,whenadultsweredistributedovertheentireBay,the

surfacetemperatureexceeded26oCthroughouttheBay,andadultsweremostly

er-leftconcentrationinthe

1996-1999,thespawnerdistributioninJune-Augustcoincidedwiththe24-26oC

temperaturearea(Fig.7-a).In1996,thelowestrecruitmentyear,adultsmoved

farupbayinJuly(Fig.8),probablybecauthe25-26oCtemperaturerange

occurredthere(Fig.7-b).Thus,thelatitudinalrangeofadultswasbroadinJuly-

August1996(Fig.8).Incontrast,inJuly-August1998,the25-26oCtemperatures

occurredonlyinthelowerBay(Fig.7-b),andmostadultsdidnotmovefar

upbaybetweenAprilandJuly-August(Fig.8).Thus,watertemperature,rather

thanthesalinitygradient,maybemoreimportantvariabletodefinelatitudinal

-MaytemperatureinthemiddleBaywashighin

11

1995-1996,butlowin1997-1999,andthelowAprilmid-Baytemperaturewas

associatedwithhigherYOYrecruitmentinOctober.

nce-weightedmeanlatitudeof

occurrenceofbayanchovy,1995-1999.

37.00

38.00

39.00

M

e

a

n

l

at

i

t

u

d

e

3095

TL(mm)

37.00

38.00

39.00

M

e

a

n

l

at

i

t

u

d

e

30

TL(mm)

37.00

38.00

39.00

M

e

a

n

l

at

i

t

u

d

e

306570

TL(mm)

9596979899

(a)April-May

(b)June-August

(c)October

37

38

39

M

e

a

n

l

at

i

t

u

d

e

1

95

96

97

98

99

37.00

38.00

39.00

40.00

9596979899

April-May

June-August

(a)Meanlatitudeofadults

(b)MeanlatitudeofadultsinJune-August

Surfacetemperature(oC)ofthemiddleBayinApril

titudeofoccurrenceofadultbayanchovy.

Theupperverticalbarreprentsmean+STDofJune-

August,andthelowerverticalbarreprentsmean–

STDofApril-May.

12

Forsmall(<60mmTL)orYOYbayanchovy,correlationandregression

analysrevealedthatsalinitygradientswereconsistentlyrelatedtothelength-

specificmeanlatitudesofoccurrenceofbayanchovy(Fig.9).InApril-May,the

length-specificmeanlatitudeweightedbyabundancemovedupbayasthe

salinitygradient(thedifferencebetweenthemiddleandtheupperBaymean

salinity)decread(Fig.9-a).Asbayanchovylengthincread,themean

latitudesofoccurrencedecread,e.g.,largerbayanchoviesgenerallywere

distributedtowardthelowerBay.

Becautherewasstrong

interactionbetweenthesize(x-

axis)andsalinitygradient(y-axis)

inApril-May,lengthdifferences

mostlydisappeared,ifthesalinity

gradientislargerthan

llest

bayanchovyexperiencedthemost

pronouncedsalinitygradienteffect

inApril-May.

InJune-AugustandOctober,

thelength-specificmeanlatitudes

ofoccurrenceofYOYbayanchovy

didnotshowasignificant

interactioneffect(lengthxsalinity

gradient),andthesalinity-gradient

effectwasadditiveforallsize

class(Fig.9-bandc).The

directionofthesalinity-gradient

effectwasoppositeduringthetwo

owsalinity

gradientsinJune-August,YOYbay

anchovydistributionswere

displacedupbay,whereasin

Octoberalowsalinitygradientwas

associatedwithdownbay

ultindicates

thatsteepsalinitygradients

discourageupbayYOYbay

anchovymigrationfromAprilto

August,anddownbaymigration

salinitygradientapparentlywould

inhibitontogeneticmigrationof

smallorYOYbayanchovyin

eitheranupbayordownbay

direction.

ialspawningareasandconditionsin

ChesapeakeBay.(a)adultabundancewithsurface

salinityandtemperature(1995-1999),(b)potential

spawningareabadonsurfacetemperature(24-26

o

C).

10

12

14

16

18

0

50

100

200

300

400

500

600

700

800April-May

SurfaceSalinity(psu)

22

24

26

28

30

S

u

r

f

a

c

e

T

e

m

p

e

r

at

u

r

e

(

C)

0

50

100

200

400

600

800

1,000

1,200

1,400

1,600June-August

July1996August1998

-77.00-76.50-76.00-75.50

37.00

37.50

38.00

38.50

39.00

39.50

July1996

-77.00-76.50-76.00-75.50

37.00

37.50

38.00

38.50

39.00

39.50

August1998

(a)

(b)

13

1996

1998

AdultBiomass

AprilJuly-August

YOYabundance

July-SeptemberOctober

1996

1998

-77.00-76.50-76.00-75.50

37.00

37.50

38.00

38.50

39.00

39.50

-77.00-76.50-76.00-75.50

37.00

37.50

38.00

38.50

39.00

39.50

-77.00-76.50-76.00-75.50

37.00

37.50

38.00

38.50

39.00

39.50

-77.00-76.50-76.00-75.50

37.00

37.50

38.00

38.50

39.00

39.50

-77.00-76.50-76.00-75.50

37.00

37.50

38.00

38.50

39.00

39.50

-77.00-76.50-76.00-75.50

37.00

37.50

38.00

38.50

39.00

39.50

-77.00-76.50-76.00-75.50

37.00

37.50

38.00

38.50

39.00

39.50

0

100

150

200

300

500

1,000

2,500

5,000

7,500

10,000

13,000

-77.00-76.50-76.00-75.50

37.00

37.50

38.00

38.50

39.00

39.50

0

10

50

100

250

500

750

1,000

2,000

4,000

5,000

butionofbayanchovyspawningstockbiomassandYOYrecruitment

in1996and1998,ChesapeakeBay.

14

RecruitmentModels

AlthoughSSBalonewasnotcorrelatedwithrecruitmentlevel,theshiftin

spatialdistributionofadultspawnersbetweenApril-MayandJune-Augustwas

egression

analysisforthestatisticalmodel,thelatitudinalshiftaloneexplained85%of

recruitmentvariabilityfrom1995to1999(Fig.10-a).Theregressionequationis:

Log(R)=10.55–1.49∆L

Where,R:numberofrecruitsinmillions;∆L:differenceinmeanlocationlatitude

ofSSB,[(Julylatitude)–(Aprillatitude)].

WhenaSSBtermwasincluded,theregressionimprovedandexplained

99.5%ofrecruitmentvariability(Fig.10-b).Theequationis:

Log(R)=11.29-0.000285S–1.83∆L

where,S:SSB(male+female)intons.

ThelatitudinalshiftsinSSBstillexplainedmostofrecruitmentvariability(p

<0.0032).AlthoughSSBalonewasnotsignificantlycorrelatedwithrecruitment,

SSBwasstatisticallysignificantinthemodel,iflatitudinalshiftswereheld

constantforthefiveyears(p=0.0180).Thisstatisticalmodelindicatesthat

recruitmentlevelismaximizedwhenSSB=0,anobviouslyunrealisticoutcome

(Fig.11-a).Theunrealisticresultapparentlywasgeneratedbecauthenegative

slopeofSSBimpliesadensity-dependentrecruitmentprocessthatisnot

accountedforbythestatisticalmodel.

Despiteasmalldecreainr2andthepossibilityofcollinearityamong

SSB,latitudinalshift,andrecruitment,themodifiedRickermodelprovidedagood

fittotheYOYrecruitmentdata(r2=98.2%,p=0.0176),withabiologicallyrealistic

result(Fig.11-b).Theequationis:

R=128Sexp(-0.000785S-1.805∆L)

where,R:recruitmentlevels,millionsofYOYinOctober;S:SSB(male+

female,tons)inApril-May.

Thismodelmaximizesrecruitmentlevelwhenbaywiderelativebiomassof

adultsinApril(SSB)isapproximately1,300tons,ifthelatitudinalshift(∆L)is

sheldconstant,themodelmaximizedrecruitmentlevel

when∆L=0,i.e.,iftheSSBdistributionixactlythesamebetweenApril-May

1998,thehighestrecruitmentyear,theestimatedbaywide

adultbiomasswas1,heVIFwasonly1.28<<10

andallCIswerefarsmallerthan15,weassumedthatcollinearitywasnota

includedtheinteractionterm(Sx∆L)andfitthemodelagain,

butwithoutimprovementbecautheinteractiontermwasnotsignificant(p=

0.06).

15

titudeofoccurrenceofsmall

(<60mmTL)orYOYbayanchovy(1995-

1999)withrespecttosalinitygradient

betweentheupperandthemiddleBay.(a)

for<60mmTLbayanchovy,(b)and(c)

forYOYbananchovy.

sionanalysofYOYbayanchovyrecruitment

levelforfiveyearsinChesapeakeBay.R:YOY

recruitmentlevel;∆L:differenceinbiomass-weighted

meanlatitudeofSSBindecimalunitsbetweenApril-May

andJune-August;S:baywidespawningstockbiomass

(male+female)intonsforApril-May.(a)∆Lalone

includedaxplanatoryvariable,(b)∆LandSincluded

axplanatoryvariables.

(a)April-May

(b)June-August

(c)October

8.5

9.0

9.5

10.0

10.5

11.0

l

o

g

(

R

e

c

r

ui

t

m

e

nt

)

i

n

O

c

t

o

b

e

r

0.00.20.40.60.81.01.2

95

96

97

98

99

R-square=85%

(p=0.026)

ΔL

0.0

10.0

20.0

30.0

40.0

50.0

60.0

R

e

c

r

ui

t

s

i

n

O

c

t

o

b

e

r

(

bi

l

l

i

o

n

s

)

9596979899

Year

Obrved

Fit

R-square=99.5%

(p=0.0054)

(a)Log(R)=10.55–1.49ΔL

(b)Log(R)=11.29-0.000285S

–1.83ΔL

R

e

c

r

u

i

t

s

i

n

O

c

t

o

b

e

r

(

1

0

9

)

16

DISCUSSION

Notonlybiologicalinteractions,butalsocomplexoceanographic

ultsdemonstratedthatthere

isastrongspatialcomponentintherecruitmentdynamicsofbayanchovyin

linityandtemperaturedistributionsareimportanttobay

tion,spawningstocksizealsowas

relatedtorecruitmentlevel,ghfish

recruitmentprocesshavebeendifficulttounderstand,thespatially-extensive

TIESprogramhasprovidedvaluableinformationonprocessrelatedtobay

anchovyrecruitment.

Itispossiblethatthe

apparentontogenetic

migrationpatternmighthave

beencaudbyregional

differencesingrowthand

r,itis

difficulttothinkthatthe

smoothconsistencyfrom

1995to1999,and

significantrelationships

betweenmeanlatitudesof

bayanchovydistributionand

salinitygradientandtotal

length(Fig.5and9)could

havebeengeneratedby

spatialdifferencesingrowth

s

possiblethatsmall

variancesinthelocationsof

bayanchovycouldbe

explainedbyregional

variabilityingrowthand

sprobable

thatontogeneticmigrationis

thedominantfeature

determiningspatialand

temporalpatternsin

abundance,biomass,and

productionofbayanchovy

atthemesoscale.

Itisuncertainwhy

YOYbayanchovymigrate

upbaywhilegrowingin

summerandthenbeginto

movetowardthelowerBay

inlatesummerandfall.A

ck-recruitmentmodels.∆L:difference

inbiomass-weightedmeanlatitudeofSSBin

decimalunitsbetweenApril-MayandJune-

August;SSB:baywidespawningstockbiomass

(male+female)intonsforApril-May.∆LandSSB

includedaxplanatoryvariables.

R=Sx128exp(-0.000785S-1.805L)

R-square=98.2%(p=0.0176)

R-square=99.5%(p=0.0054)

(a)Statisticalmodel

(b)Rickermodel

Recruits(109)

Recruits(109)

ΔL

ΔL

17

steepsalinitygradientemstodeterordiminishtheontogeneticmigrationof

etal.(2000),badonotolithmicrochemicalanalysis,

supportedthehypothesisofanupbayontogeneticmigrationbysmallYOY

anchovy(latelarvaeandsmalljuveniles).Adultbayanchovyprobablyisnot

influencedsignificantlybysalinitybarriersintheBay(Houde&Zastrow,1991),

butmayarchforsuitablespawningconditionsinwhichwatertemperatures,

andprobablypreylevels,tbebeneficialforjuvenilebay

anchovytodisperwidelytoexploitfoodresourcesbaywideiftheyarenot

r,whenthey

approachtheadultstage,theymustreturntosuitablespawningareasthatare

definedbysuitabletemperature,salinity,orotherhydrologicalandbiological

encesinphysiologicalandecological

requirementsamonglifestagesofbayanchovymaycontrolthedistinctive

ontogeneticmigrationpatternthathasbeenobrved,althoughlittleinformation

isavailabletotestforsuchdifferences.

AlthoughthemodifiedRickermodelforbayanchovyrecruitmentfitthe

datawell,thenumberofyearsisonlyfive,andthesignificantrelationshipcould

r,webelievethatpossiblestatisticalerrors

areminimalbecaudiagnosticsdidnotindicatesignificantcollinearity.

Moreover,consideringthenumberofstationsavailabletoparameterizethe

valuesofeachvariableintherecruitmentmodels,thebaywideestimatesofbay

anchovyabundanceandbiomassarequiteaccurateandpossiblemeasurement

berofCTDstationswasnear100foreach

crui,andthesurfaceandbottomtemperatureandsalinitywerenotaveraged

badonanarbitraryfixeddepthofpycnocline,butwerecarefullyparatedby

hery-

independentbaywideestimateofabundanceforeachsizeclassofbayanchovy

alsoisbelievedtobeunbiadandaccurate.

Thespawningareaofadultanchovyin1995-1999wascriticalin

hrecruitmentyearshadsimilar

g&Houde

(1999a,b)reportedhighbiomassofadultanchovypredominantlyinthelower

BayinJuneandJuly1993,ahighrecruitmentyear,inagreementwiththe

wetal.(1991)

periodforthelatitudinalshift,∆L,April-MaytoJune-August,mostlycoincided

r,Itisuncertainwhyrecruitmentlevelswere

higherwhenthelatitudinalshiftswerenarrower.

Recentindividual-badmodelssuggestedthatdensity-dependent

processduringearly-lifestagescouldstabilizebayanchovyrecruitments

(Wangetal.,1997;Roetal.,1999;Cowanetal.,1999).Rilling&Houde(1999

a,b)reportedthatmeandensityofeggsandlarvaeinJuly1993washighestin

thelowerBayandlowestintheupperBay,whereascumulativemortalityfrom

eggto18-d-oldlarvalstagewashighestinthelowerBay,andlowestinthe

tion,otherrecruitmentmodelsthatwefit,butforwhich

resultsarenotgiven,consistentlyindicateddensity-dependence(significantβ

1

18

values).Atsmallscalesofveralmeters,consideredbyWangetal(1997)and

Cowanetal.(1999),feedingprocessareimportantandhighanchovy

spawningstockbiomasscouldinducedensity-dependentfoodcompetitionon

abundantfirst-feedinglarvaebecauplanktonicpreyofbayanchovyarefar

smallerthananchovy,andcertainlysubjecttocontrolbysmall-scaleprocess.

Peebletal.(1996)hypothesizedthatbayanchovy’ssize-specificfecundityis

relatedtopreyavailabilityinTampaBay,Florida,andRoetal.(1999)

suggestedthatdensity-dependentgrowthoflarvaeandjuvenileswouldleadto

&Kimbrell(1980)andAlheit

(1987)propodthatcannibalismduringeggandlarvalstagesisresponsiblefor

ifiedRickermodelaccountedforthe

possibledensity-dependence,althoughwedonotknowatwhichstagessuch

density-dependentprocessaremostimportant.

Atfirstglance,theobrvedrelationshipbetweenlatitudinalshiftofbay

anchovySSBdistributionandYOYrecruitmentconflictswiththedensity-

dependenthypothesis,becauexpecteddensity-dependencewouldbemore

vereandrecruitmentwoulddecreaastherangeofthelatitudinalshift

,weobrvedthatrecruitmentincreadastheshiftdecread.

Thiseminglyprecludespreydistributionasaprimaryfactorexplainingthe

relationshipbetweenlatitudinalshiftsinSSBdistributionandYOYrecruitment.

Theconflictmaybeexplainedbyconsideringthespatialscaleof

ionhasbeensuggestedasanimportant

controlleroffishrecruitmentprocessinearly-lifestages(Sisnwine,1984;

Bailey&Houde,1989).Atthemesoscaleof10-100Km,whichcorrespondsto

rangesofthelatitudinalshiftsofadultbayanchovy,distributionofpredatorsmay

beimportantbecaumostpredatorsarelargerthanbayanchovyandcanmove

speculatethatanabundantandspatiallyconcentratedsupplyoflarvaeor

juvenileanchovy,especiallyinthemiddleBay,couldpromoteearly-lifesurvival,

butmoreanalysofanchovy,

theirpredators,andtheirpreyare

required.

TheabundanceofYOY

weakfish(Cynoscionregalis),a

majorpredatorofbayanchovyin

ChesapeakeBay(Hartman&

Brandt,1995),showedan

exponentialsaturation

relationshipwithrespecttoYOY

bayanchovynumber(Fig.12),

andsupported,atleast

circumstantially,thepredator-

spawningareaisnarrow,the

supplyoflarvaeandjuvenilefish

topotentialpredatorsmaybe

0

20

40

60

80

100

120

N

u

m

b

e

r

of

Y

O

Y

W

e

a

k

f

i

s

h

0

NumberofYOYBayAnchovy

95

96

97

98

99

Y=129(1-exp(-0.046X))

ntialsaturationrelationshipbetween

numberofYOYweakfish(x10

6

)andYOYbay

anchovy(x10

9

)inOctober.

19

chanismforsurvivalhasbeensuggestedforinctsand

plants(Gould,1977),inwhichtemporalquence,i.e.,periodsincyclic

spawningactivity,poralquencingalsohasbeen

emphasizedinfishrecruitmentprocess,e.g.,thematch-mismatchhypothesis

(Cushing,1974).Inthecaofbayanchovyrecruitment,however,we

hypothesizethatitisnotthetemporalquence,butthespatialdistributionof

spawnersthatmaybemoreimportantincontrollingrecruitmentthroughpredator

satiation.

Aconceptualmodelforhypothesizedmechanismsofbayanchovy

arize:

1)LowtemperaturesatabaywidescaleinChesapeakeBayduringApril-

Maywillrestrictdistributionofmostadultsduringthependingspawningason

tivelylowtemperaturespersistuntilthe

June-Augustspawningason,temperature-criticalspawningactivitywillbe

confinedtoarelativelysmallareaofthelowerandlower-midBay.

2)Concentratedspawningactivitythatproduceggsandlarvaeinthe

confinedareacouldsatiatepredatorsofbayanchovylarvaeandYOY(spatial

predator-satiationhypothesis).

ptualmodelofthebayanchovyrecruitmentprocessinChesapeake

latitudinalshiftofspawnersixplainedbyverticalpositionsoftheblackboxes.

Horizontalarrowsindicatedensity-compensatoryanddensity-depensatory

lackarrowsindicate

ontogeneticmigrationofYOYbayanchovy.

20

3)Density-dependencesuggestedbythespawningstock-recruitment

relationshipmightbearesultoffeedingconstraints,cannibalism,and

competitionintheearly-larvalstages.

Ultimately,otherhypothesmayexplainbettertherelationshipsbetween

lati

latitudinalshiftsofadultsmighthavebeencontrolledbythesamecommonfactor

astherecruitmentprocess,andtherelationshipmightbetheresultofthe

tillinvestigatingotherpossiblehypothestoexplain

ing

resultsofouranalysandconsideringscalingissues,thedensity-compensatory

processemscertaintoactduringearly-larvalstagesatsmallspatialscales,

whereasthepossibledensity-depensatoryprocesxplainingtheeffectof

latitudinalshiftmightoccurduringthepost-larvalandjuvenilestageatthe

parentthatspatially-explicitprocessoperatingatthe

mesoscale,andaffectingbothadultandYOY,haveimportantimplicationsfor

ureofdensity-dependent

mechanismsandphysicalprocessoperatingatfine-andmesoscalesarestill

poorlyunderstood.

ACKNOWLEDGEMENTS

g,and,,,,D.O’Brien,n,

,,nlopenforassistanceinfieldsurveys.

alScienceFoundationGrantDEB94-12113.

REFERENCES

Able,K.W.;Fahay,M.P.(1998):ThefirstyearinthelifeofestuarinefishesintheMiddleAtlanticBight.

RutgersUniversityPress,NewBrunswick,NewJery.342pages.

Alheit,J.(1987):Eggcannibalismversuggpredation:.5,

467-470.

Bailey,K.M.;Houde,E.D.(1989):Predationoneggsandlarvaeofmarinefishesandtherecruitment

.25,1-83.

Baird,D.;Ulanowicz,R.E.(1989):.

59(4),329-364.

Bhattachara,C.G.(1967):AsimplemethodofresolutionofadistributionintoGaussiancomponents.

Biometrics23,115-135.

ChesapeakeBayProgram(2000):ChesapeakeBay:Introductiontoanecosystem.30p.

Cowan,J.H.;Ro,K.A.;Houde,E.D.;Wang,S.B.;Young,J.(1999):Modelingeffectsofincreadlarval

mortalityonbayanchovypopulationdynamicsinthemesohalineChesapeakeBay:evidencefor

.185,133-146.

Cressie,N.A.C.(1993):-Intersciencepublication,NewYork.

900pages.

Cronin,W.B.(Ed.)(1971):Volumetric,areal,andtidalstatisticsoftheChesapeakebayestuaryandits

ore.15pages.

Cushing,D.H.(1974):Thepossi:

Theearlylifehistoryoffish.(Ed:Blaxter,J.H.)CambridgeUniversityPress,Cambridge,103-111.

Day,J.W.;Hall,C.A.S.;Kemp,W.M.;Yáñez-Arancibia,A.(1989)::Estuarine

Ecology.(Eds:Day,J.W.;Hall,C.A.S.;Yáñez-Arancibia,A.)JohnWiley&Sons,Inc,NewYork,79-

143.

Dory,S.E.;Houde,E.D.;Gamble,J.C.(1996):Cohortabundancesanddailyvariabilityinmortalityofeggs

andyolk-saclarvaeofbayanchovy,Anchoamitchilli,.94,257-267.

Gould,S.J.(1977):,NewYork.

21

Hartman,K.J.;Brandt,S.B.(1995):Predatorydemandandimpactofstripedbass,bluefish,andweakfishin

theChesapeakeBay:.52,1667-1687.

Houde,E.D.;Zastrow,C.E.(1991)::HabitatrequirementsforChesapeakeBayliving

resources.(Eds:Funderburk,S.L.;Mihursky,J.A.;Jordan,S.J.;Riley,D.)LivingResources

Subcommittee,CheapeakeBayProgram,Annapolis,8.1-8.14.

Hunter,J.R.;Kimbrell,C.A.(1980):Eggcannibalisminthenorthernanchovy,.

78,811-816.

Kimura,R.;Secor,D.H.;Houde,E.D.;Piccoli,P.M.(Inpress):Up-estuarydispersalofyoung-of-the-year

bayanchovyAnchoamitchilliintheChesapeakeBay:InferencesfrommicroprobeanalysisofSrin

.

Kimura,R.;Secor,D.H.;Houde,E.D.;Piccoli,P.M.(2000):Migration/dispersalpatternsofYOYbayanchovy

AnchoamitchilliintheChesapeakeBay:Sr:ript

submittedtoICES2000AnnualScienceConference,SessionN.

King,M.(1995):Fisheriesbiology,gNewsBooks,Oxford.341pages.

Littell,R.C.;Freund,R.J.;Spector,P.C.(Eds.)(1991):titute

Inc.,Cary,NorthCarolina.329pages.

Loos,J.J.;Perry,E.S.(1991)::

fishrecruitmentandrearchintheAmerican:

proceedingsofthe13thannualfishconference;21-26May1989,Merida,Mexico.(Ed:Hoyt,R.D.),

65-76.

Luo,J.;Brandt,S.B.(1993):BayanchovyAnchoamitchilliproductionandconsumptioninmid-Chesapeake

Bay.

.98(3),223-236.

Luo,J.;Musick,J.(1991):.

Soc.120,701-710.

MacGregor,J.M.;Houde,E.D.(1996):Onshore-offshorepatternandvariabilityindistributionand

a.

Ser.138,15-25.

Murdy,E.O.;Birdsong,R.S.;Musick,J.A.(1997):onianInstitutePress,

Washington,don.324pages.

Newberger,T.A.;Houde,E.D.(1995):PopulationbiologyofbayanchovyAnchoamitchilliinthemid

.116(1-3),25-37.

Peebles,E.B.;Hall,J.R.;Tolley,S.G.(1996):EggproductionbythebayanchovyAnchoamitchilliinrelation

.131,61-73.

Ricker,W.E.(1954):11,559-623.

Ricker,W.E.(1975):Com.

.191,1-382.

Rilling,G.C.;Houde,E.D.(1999a):Regionalandtemporalvariabilityindistributionandabundanceofbay

anchovy(Anchoamitchilli)eggs,larvae,ies22,

1096-1109.

Rilling,G.C.;Houde,E.D.(1999b):Regionalandtemporalvariabilityingrowthandmortalityofbayanchovy,

Anchoamitchilli,.97,555-569.

Ro,K.A.;Cowan,J.H.;Clark,M.E.;Houde,E.D.;Wang,S.B.(1999):Anindividual-badmodelofbay

.

185,113-132.

SAS(1990):SAS/STATUr'titudeInc.,Cary,NC.

SAS(1998):SAS/STATTechnicalreport:lishing.80

pages.

Sisnwine,M.P.(1984):Whydofishpopulationsvary?In:Exploitationofmarinecommunities.(Ed:May,

R.M.)Springer-Verlad,Berlin,59-94.

Wang,S.B.;Cowan,J.H.;Ro,K.A.;Houde,E.D.(1997):Individual-badmodellingofrecruitment

ol.51

(SupplementA),121-134.

Wang,S.B.;Houde,E.D.(1995):Distribution,relativeabundance,biomassandproductionofbayanchovy

.121,27-38.

Zastrow,C.E.;Houde,E.D.;Morin,L.G.(1991):Spawning,fecundity,hatch-datefrequencyandyoung-of-

.73,

161-171.