HAL Id: hal-00452257
https://hal.univ-brest.fr/hal-00452257
Submitted on 2 Feb 2010
HAL is a multi-disciplinary open access
archive for the deposit and dissemination of
sci-entific research documents, whether they are
pub-lished or not. The documents may come from
teaching and research institutions in France or
L’archive ouverte pluridisciplinaire HAL, est
destinée au dépôt et à la diffusion de documents
scientifiques de niveau recherche, publiés ou non,
émanant des établissements d’enseignement et de
recherche français ou étrangers, des laboratoires
Ecophysiological dynamic model of individual growth of
Ruditapes philippinarum
Jonathan Flye-Sainte-Marie, Fred Jean, Christine Paillard, Susan Ford, Eric
Powell, Eileen Hofmann, John Klinck
To cite this version:
Jonathan Flye-Sainte-Marie, Fred Jean, Christine Paillard, Susan Ford, Eric Powell, et al..
Ecophysio-logical dynamic model of individual growth of Ruditapes philippinarum. Aquaculture, Elsevier, 2007,
266 (1-4), pp.130-143. �10.1016/j.aquaculture.2007.02.017�. �hal-00452257�
growth of Ruditapes philippinarum
⋆
Jonathan Flye Sainte Marie
a
, Fred Jeana
,
∗
, Christine Paillarda
, Susan Fordb
, Eri Powellb
, Eileen Hofmannc
, John Klin kc
a
IUEM- LEMAR,Pla e Coperni , 29280 Plouzané, Fran e
b
HaskinShellsh Resear h Laboratory, Rutgers University, 6959 Miller Avenue,
PortNorris, NJ08349 USA
c
Center for CoastalPhysi al O eanography, Old DominionUniversity,Crittendon
Hall, Norfolk, VA 23529, USA
Abstra t
A bioenergeti s model of the Manila lam (Ruditapes philippinarum) was built
to simulate growth,reprodu tion and spawning in ulture and sheryeld sites in
MarennesOléron Bay (Fren h Atlanti oast). The model is driven by two
envi-ronmental variables:temperature andfoodsupply.The foodsupplyand the lam's
ltrationratedeterminesofttissue onditionindex,whi hinturndrives lamgrowth
andreprodu tion.Themodelwas alibrated andthenvalidatedusingtwo
indepen-dentdatasets.
Thispaperdis ussesthe di ultyof omparingexperimentaldataandindividual
modeloutputswhenasyn hronousspawningeventso urinthestudiedpopulation.
In spite of this di ulty, the simulations reprodu e the typi al pattern of growth
and reprodu tion ofthe Manila lam.
Simulationsshowedthatwater olumn hlorophylla on entrationisnotaperfe t
estimatorof foodresour es for anear bottom suspensionfeeder su hastheManila
lamandemphasizethe la kofknowledgeaboutRuditapesphilippinarumnutrition.
The individual growth model presented in this paper will be integrated into a
numeri al population model des ribing the host-parasite-environment relationship
inBrown RingDisease, aused bytheba terium Vibrio tapetis.
Key words: model, individualgrowth, energy balan e,reprodu tion, foodinput,
Brown RingDisease
⋆
SupportedbyNSF-CNRSjointprogram
The Manila lam (Ruditapes philippinarum) was introdu ed for aqua ulture
purposes to Fran e between 1972 and 1975 (Flass h and Leborgne, 1992). In
Fran e, this venerid ulture be ame in reasingly widespread, and sin e 1988
natural populations have olonized most embayments along the Fren h
At-lanti oast, resulting ina shery of a1500 tons inthe Gulf of Morbihan at
the end of the 1990s. R. philippinarum is often ae ted by high mortalities
inlate winter, whi h onsiderably limitaqua ulture and sheries produ tion.
These mortalitieswere rst asso iated with unfavourable environmental
fa -torssu haslowtrophi resour es, oldtemperaturesandlowsalinity
(Goullet-quer,1989b;Bower,1992).Later,theywere attributedtoBrownRingDisease
(BRD), a ba terial disease indu ed by the pathogen Vibrio tapetis (Paillard
etal.,1989;PaillardandMaes, 1990),in onjun tionwithlow onditionindex
and lowbio hemi alreserves (Goulletquer,1989a).BRDprogression in lams
andV.tapetis strategiesforinfe tionhavebeendes ribed indepthbyPaillard
etal.(1994)and Paillardand Maes(1994).Infe tions maytheoreti allyo ur
any time of the year, but BRD development may be modulated by
environ-mental onditions (Paillard et al., 2004) and lam's defen e system (Allam
and Paillard,1998;Allametal.,2000).Inadditiontheseauthorsshowed that
BRD developmentis modulatedby the energy balan eof the lam.
Asa rststep indeveloping anumeri almodeldes ribing the
host-pathogen-environmentrelationshipsinBRD,thisstudypresentstheindividual
e ophys-iologi almodel ofgrowth and reprodu tionofthe Manila lam.The aimisto
des ribe the variability in energy balan e for an individual without BRD as
a fun tion of the two main environmental fa tors presumed to ontrol BRD
development:the trophi resour es(Goulletquer,1989a)andthe temperature
(Paillard etal.,2004).
2 Materials and methods
2.1 Basi on ept and state variables
The modelisbased ona thewidely applied(see reviewinBayne,1998)s ope
for growth (
SF G
) on ept. TheSF G
on ept assumesthat energy ormattergained by food a quisition is equal to the energy or matter lost for
mainte-nan e,growth and reprodu tion.
SF G
an be al ulated asfollowsa ordingSF G = Gg + Sg + Sh = Cons − (P p + F p + Resp + U)
where:
Gg
isgonadalgrowth;Sg
issomati growth;Sh
isshellgrowth;Cons
is onsumption(retained organi matter);P p
ispseudofae es produ tion;F p
is fae esprodu tion;Resp
isrespirationandU
isureaandaminoa idex retion.If
SF G < 0
, energy or matter is mobilized, rstly from gonad, then fromsomati tissue, and the lam loses weight while using energy reserves. If
SF G > 0
,energy ormattergainedispartitionedintoshell, soma,and gonad.Partitioning among those three ompartments is ontrolled both by
environ-mental parameters (quantity and quality of food, temperature, salinity, ...)
and by endogenous fa tors su h asgenotype, size and physiologi al ondition
(see e.g. Shafee and Lu as, 1982; Lu as, 1993; Goulletquer, 1989a;Pérez
Ca-ma hoetal., 2003).
The on eptual design is des ribed in Fig. 1. The stru ture and formulation
of this model is based on the Hofmann et al. (2006) Mer enaria mer enaria
growth model.The twostate variables are the shelllength and the total esh
weight in ludingsomati weight and gonadalweight.
Ingestion
Assimilation
Respiration
Reproduction
Filtration
Net Production
Division
Net Production
Trophic
Ressource
Temperature
Condition
Index
Somatic
Growth
Shell
Growth
(length)
LEGEND:
Negligible energy flow
Influence
Spawning
Energy flow
Three data sets (Fig. 2) were extra ted from Goulletquer (1989a), a survey
of Manila lam aqua ulture in the Marennes-Oléron basin (Fren h Atlanti
oast, see Fig. 3):
(1) atimeseriesfromNolestation(Mar h1984toSeptember1985)of
hloro-phyll a (
Chla
) on entration was used for alibration. Temperature (T
;◦
C) was obtained from the same study site duringa 1986 survey;
(2) two time series, one from Nole station and the other from Lilleau
sta-tion,both extendingfromMar h1985 toMay 1986,were used formodel
validation.Both onsistedof
Chla
on entrationandtemperatureseries.Datafromalltimeserieswerelinearlyinterpolatedtoobtaindailyvalueswhen
ne essary.
In order to evaluate trophi resour es,
Chla
on entration was onverted toingestible dry organi matter (trophi resour e
F ood
,g L
−1
) to take into
a ount the pseudofae es produ tionstep, followingthe equation:
F ood = Chla × F oodCoeff
where
F oodCoef f
ing DW µg
−1
Chla
is the oe ient that onverts
Chla
toingestibledry organi matter.
2.3 Physiologi al me hanisms formulations
2.3.1 Net Produ tion
Netprodu tionofdryorgani matter(
P net
;g day
−1
)was al ulatedwiththe
following equation, whi h ree ts energy gain from assimilationand lossdue
torespiration:
P net = Ingest × Assim − Resp
(1)Food a quisition Using data from literature (Goulletquer et al., 1989;
Goulletquer,1989a), theltrationrate ofa 1-
g DW
individualwasestimated asa fun tion of temperature using the followingempiri alequation:F ilt
T
(T ) = −5.62 10
−3
× T
2
0
2
4
6
8
10
12
14
16
18
50
100
150
200
250
300
350
400
450
500
Chlorophyll a (µg.l−1)
Time (Days)
A
0
2
4
6
8
10
12
14
16
18
50
100
150
200
250
300
350
400
450
500
Chlorophyll a (µg.l−1)
Time (Days)
B
0
2
4
6
8
10
12
14
16
18
20
22
24
50
100
150
200
250
300
350
400
450
500
Temperature (°C)
Time (Days)
C
0
2
4
6
8
10
12
14
16
18
50
100
150
200
250
300
350
400
450
500
Chlorophyll a (µg.l−1)
Time (Days)
D
0
2
4
6
8
10
12
14
16
18
20
22
24
50
100
150
200
250
300
350
400
450
500
Temperature (°C)
Time (Days)
E
Fig. 2. Environmental data sets used for alibration and validation, interpolated
from Goulletquer (1989a). Data sets A and B and C were usedfor the alibration
at Nole station. Data sets Band C were used for validation at Nole and datasets
1°25
1°20
1°15
1°10
1°05
46°05
46°00
45°55
45°50
France
Marennes−Oléron
basin
B
A
Fig. 3. Lo ation of the the two stations studied by Goulletquer(1989a) in
Maren-nes-Oléronbasin(Atlanti oast,Fran e).Nole(A),anestuarinestationandLilleau
(B),an o eani station.
F ilt
T
inL h
−1
rea hes itsmaximum (
F ilt
T
(16) = 1.16 L h
−1
) at16
◦
C.
Filtrationratewasestimatedasafun tionofdryweight(
F ilt
W
(W )
inmL min
−1
)
following the empiri alequation:
F ilt
W
(W ) = a
f
× W
bf
(3)with
a
f
= 20.049
andbf = 0.257
.Usingequations(2)and(3), ltrationratewas nallyexpressed as afun tion
of individualweight and temperature(
F ilt
;L day
−1
)followingthe equation:
F ilt(W, T ) = (a
f
× W
bf
×
F ilt
T
(T )
F ilt
T
(16)
) × 1.44
(4)
Ingestion (
Ingest
;g day
−1
) of dry organi matter an be then al ulated
a ording tothe equation:
Ingest = F ood × F ilt
(5)Assimilation Aminoa idand ureaex retionisverylow omparedtoother
parametersandwasthusomittedfromthemodel.Therefore, assimilationwas
onsidered equal to absorption. Assimilation rate (
AE
) was estimated as afun tion of weight a ording to aMi haelisMenten formulation:
AE = AE
0
+
AE
1
× W
K
A
+ W
(6)
where
AE
0
= 0.1
andAE
1
= 0.6
arethe minimumandmaximum additiveas-similationratesrespe tively,and
K
A
= 2.74 g
is the half-saturation onstant. Assimilation(Assim
;g day
−1
) wasthen al ulated following :
Assim = AE × F ood × F ilt
(7)Respiration Respiration was also al ulated, using data from Goulletquer
et al. (1989) and Goulletquer (1989a), as a fun tion of temperature for a
1-g DW
individualasResp
T
(T ) = −4.75 10
−4
× T
2
+ 2.02 10
−2
× T + 2.43 10
−2
(8)Resp
T
(ml O
2
h
−1
) rea hes its maximum
Resp
T
(21) = 0.24 ml O
2
h
−1
at21
◦
C. Respirationforindividual lams wasestimated fromweight (
Resp
W
inCal day
−1
)following the empiri alequation:
Resp
W
(W ) = a
r
× W
br
(9)where
a
r
= 27.88
andbr = 0.85
.Using equations (8) and (9), respiration (
Resp
;g DW day
−1
) was nally
expressed asa fun tionof weight and temperaturefollowingthe equation:
with
2.177 10
−3
being the oe ient to onvert alories to
g DW
.2.3.2 Condition index
In bivalve studies, ondition index is usually the ratio between shell and dry
esh weights(see review inLu asand Beninger, 1985).In our modela
ondi-tionindex,basedonvariationoftotaleshweightinsideinternalshellvolume,
likethatused inoyster studies(Lawren e andS ott, 1982),was al ulated as
follows:
C =
W − W
o
W
max
− W
o
(11)
wheretherelationshipbetweenthemeaneshweight(
W
o
)andtheshelllength (L
)isW
o
= a
o
× L
bo
(12)and
a
o
= 8.52 10
−7
and
bo = 3.728
.The values werea
o
andbo
were evaluatedusing a omposite data set (Paillard et al., unpublished data;
n = 921
,r
2
=
0.904
), andW
max
= a
om
× L
bo
(13) wherea
om
= 1.70 10
−6
.C
is related to the shell weight-based ondition index (CI
) using a linear relationship:C = 0.0124 × CI − 1.0562
(n = 558
,r
2
= 0.95
).2.3.3 Length growth
Theorgani matrixisonly2to3%ofshellweightoftheManila lam
(Goullet-quer and Wolowi z,1989) and generally a ounts for onlya small fra tionof
SF G
inbivalves(Thompson,1984;Dame,1996;Pouvreauetal.,2000);inthe present model,energy allo ationtoshellgrowth wasthusomitted.Therefore,shelllengthgrowthrate
LgrowthRate
(day
−1
)isaMi haelisMentenfun tion
of ondition index (
C
)and is saturated as afun tion of the maximum length (L
):LgrowthRate(L, C) = dl
o
×
L
max
− L
L
max
×
C − C
S
K
l
+ C − C
S
(14)where
dl
o
is maximum daily growth rate,L
is the individual length,L
max
is the maximum lam length,
K
l
is the half saturation onstant andC
S
isthe minimum value of
C
forLgrowthRate > 0
. Parameters of the Mi haelis Mentenequationwereevaluatedusinga ompositedatasetfrom(Goulletquer,1989a)in MarennesOléron Bay.
2.3.4 Resour es allo ation to reprodu tion
As in most venerids, the gonad of R. philippinarum is a diuse organ in the
vis eralmass.Thisisthemainreasonwhyrulesspe ifyingtheprioritiesfor
re-sour epartioningbytheManila lamhavenotbeenextensivelystudied
(Laru-elle, 1999;Laruelleetal.,1994;Calvez, 2003). However, availableinformation
was used to make assumptions on erning energy allo ation. Most of the
in-formationisbasedonvarious ondition index andhistologi alstudies,aswell
as re ent investigations using enzyme-linked immunosorbent assays (ELISA)
toquantifyegg weight(ParkandChoi,2004;NgoandChoi,2004).Thegonad
is almost never developed in lams less than 20 mm, whi h is assumed to be
the minimum size for reprodu tion (Laruelle, 1999, Calvez omm. pers). R.
philippinarum isknown tohavea reprodu tiveperiodextendingfrom
Mar h-April to O tober on western European and Korean oasts. This orresponds
to a period when water temperature rises above 12
◦
C, whi h is onsidered
as the minimum temperature allowing gametogenesis (Laruelle et al., 1994;
Mann,1979; Park andChoi,2004; Ngoand Choi,2004).Conditionindex and
histologi alstudies showed that most of net produ tionis allo atedto
repro-du tion frommid-June to the end of O tober (Laruelle, 1999; Calvez, 2003).
Thisphenomenonisobservedatallstationsmonitoredbythoseauthorsalong
theFren hAtlanti oastando urswhen onditionindexisaboveathreshold
value.
Reprodu tive eort
RepEf f
is the adimensional fra tion of net produ tionallo ated to gonadal tissues.
RepEf f
be omes positive, and gametogenesispossible, when lam length is above a minimum (
LengthRepMin = 20
mm;Laruelle,1999)andtemperatureisaboveaminimum(
T empRepMin = 12
◦
C;Laruelle,1999;ParkandChoi,2004;NgoandChoi,2004;Mann,1979).Below
those thresholds,
RepEf f = 0
. Above the thresholds, reprodu tion eort isal ulated asa fun tion of temperature a ording to
RepEf f = RepEf f
M ax
×
1
2
× (1 + tanh(
T − 13.8
with
RepEf f
M ax
, the maximum fra tion of net produ tion allo ated to the gonad.2.3.5 Gamete release
Fa torstriggeringspawningarealsonotwellknownorintensivelystudiedinR.
philippinarum,forthesamereasonsmentionedforresour eallo ation.Spawns
o ur from May to O tober (Laruelle, 1999; Goulletquer, 1989a; Meneghetti
et al.,2004). Spawnings and rapid rematuration after spawning are observed
during the whole breeding period. Individuals are well syn hronized during
the rst springmaturation ofgametes, whi his orrelatedtotemperature
in- rease,butthis syn hrony may belostduringsubsequentspawnings(Laruelle
et al., 1994; Toba et al., 1993; Meneghetti et al., 2004). Asyn hronous
par-tial spawnings o ur fromend of May until the end of August. During these
partialspawnings events, onlya fra tion of the gonad ontent, orresponding
to the mature ovo ytes, is released. The main spawning is observed in late
AugustorSeptember, when almostallindividualsreleasetheir entire gonadal
ontent. Under good environmental onditions, late spawning events an
o - ur until end of O tober (Goulletquer, 1989a; Laruelle, 1999; Calvez, 2003).
In the model, partialand major spawning events are distinguisheda ording
tothe following riteria:
•
a lam is ready for a partial spawning event if its ondition index andgonado-somati index(gonadtoeshweightratio,GSI)rea hthreshold
val-ues (
P arIC
andP arGSI
, respe tively). After the partial spawning event,the ondition index drops to a value (
P arP ostIC
) orresponding to the weight lostby gameterelease.•
a lam isready foramajorspawningeventif itsGSIis abovethethresholdvalue (
SpawnRatio
) and it has experien ed a minimal number of days ofgametogenesis,whi hisevaluatedas119daysfromthedaythetemperature
denitively rises above 12
◦
C.
Threshold values for spawning events were estimated using data sets from
Calvez (2003) for the Gulf of Morbihan.
2.4 Parameter evaluation and model adjustment to experimental data
Mostof the parameters of the modelwere rst estimated frompublished
val-ues for individualswithout BRD symptoms ( Table 1), however, some of the
parameters ould not be evaluatedthis way and were evaluated by adjusting
On e parameters were evaluated, the model was then validated using one
data set from Nole station and one from the Lilleau station. Chlorophyll a
and temperature time series are shown in Fig. 2. Both data sets are from
Goulletquer(1989a) (see se tion2.2).
Forboth alibrationandvalidation,thetbetweenexperimentalandmodeled
results was as ertained using a linear regression through the origin for ea h
state variable (length, weight and ondition index). The slope of regression
line of modeled vs experimental values was ompared to 1 (t-test) and the
oe ient of determinationwas al ulated.
3 Results
3.1 Model evaluation
Results of a simulation after adjustment of the model to experimental eld
data are shown in Fig. 4. This simulation was obtained using the parameter
values in Table 1. For all state variables, the slopes of the regression lines
between observed and simulatedvalueswere not signi antly dierent from1
(t-test, p-value >0.05).
The simulated lengthtraje toryof a standard individualfollowed that of the
average length observed in the eld (Fig. 4A). Length growth was almost
ompletely explained by the model (
r
2
= 0.98
). Simulated dry weight (Fig.4B) and ondition index (Fig. 4C) were also lose to eld observations, even
though oe ients of determination were lower (
r
2
= 0.77
andr
2
= 0.66
,respe tively).Theselowervalues an beattributed tothe simulatedspawning
pattern: the model simulated partial spawning events at days 140 and 153
(end of May and begining of June respe tively). Spawning at this time was
notnotedbyGoulletquer(1989a),who olle tedthedataset usedto alibrate
and validate the model, but they are in a ordan e with observed spawning
patterns ontheFren hAtlanti oast (Laruelle,1999; Calvez, 2003).Amajor
spawningwassimulatedatday230(lateAugust)atthedatewherethisgamete
are expresedinthe model onditionindex (
C
)experimentalequivalentvalues(
CI
) are shown in parenthesis.Parameter Explanation Value Unit Evaluatedfrom
F ilt
T
(16)
MaximumoftheF T
(T )
fun tion(at16◦
C)
1.16
L h
−1
Goulletqueretal.(1989);Goulletquer(1989a)
a
f
Allometri oe.oftheeq.relatingltrationratetoDW20.049
nounit Goulletqueretal.(1989);Goulletquer(1989a)bf
Allometri exponentoftheeq.relatingltrationratetoDW0.257
nounit Goulletqueretal.(1989);Goulletquer(1989a)AE
0
Minimumadditiveassimilationrate0.1
nounit Goulletqueretal.(1989)AE
1
Maximumadditiveassimilationrate0.6
nounit Goulletqueretal.(1989)K
A
Half-sat. onst.oftheM.-M.eq.relatingassim.ratetoDW2.74
g
thisstudyResp
T
(21)
MaximumoftheRT
(T )
fun tion(at21◦
C)
0.24
ml O
2
h
−1
Goulletqueretal.(1989);Goulletquer(1989a)
a
r
Allometri oe.oftheeq.relatingrespirationratetoDW27.88
nounit PowellandStanton(1985)br
Allometri exponentoftheeq.relatingrespirationratetoDW0.85
nounit PowellandStanton(1985)a
o
Allometri oe.oftheeq.relatingaverageDW(W
0
)tolength851.88 10
−9
nounit Paillard(unpublisheddata)
bo
Allometri exponentoftheeq.relatingaverageDW(W
0
)tolength3.728
nounit Paillard(unpublisheddata)a
om
Allometri oe.oftheeq.relatingmax.DW(W
max
)tolength1703.76 10
−9
nounit Paillard(unpublisheddata)
L
max
Maximum lamlength60
mm
MortensenandStrand(2000)dl
o
Maximumlengthgrowthrate0.0045
day
−1
Goulletquer(1989a)
K
l
Half-sat. onst.oftheM.-M.eq.relatingC
tolengthgrowthrate 0.123 nounit Goulletquer(1989a)C
S
MinimumvalueofC
forR
L,C
>
0
-0.45(49
) nounit Goulletquer(1989a)LengthRepM in
Minimum lamlengthforreprodu tion20
mm
Laruelle(1999),Calvez(unpublisheddata)T empRepM in
Minimumtemperatureforreprodu tion12
◦
C Laruelle(1999)
P arGSI
GSIthresholdvalueforpartialspawning0.30
nounit thisstudyP arIC
Conditionindexthresholdvalueforpartialspawning0.50
(125
) nounit Calvez(2003)P arP ostIC
onditionindexvalueafterapartialspawningevent0.23
(104
) nounit Calvez(2003)SpawnRatio
GSIthresholdvalueforprin ipalspawning0.42
nounit ParkandChoi(2004)C2F
Chla
todryingestibleorg.mat. onversion oe.atNolestation0.2552
g DW µg
−1
thisstudy
C2F
Chla
todryingestibleorg.mat. onversion oe.atLilleaustation0.3432
g DW µg
−1
thisstudy
Nole
Validation simulations at Nole station are shown in Fig. 5. For length and
weight,the slopesof the regressionlines between observed and simulated
val-ues were not signi antlydierent from1 (t-test, p-value >0.05). The
simu-latedlengthtraje tory(Fig.5A)was losetoobserved dataandthe oe ient
of determinationwas very high (
r
2
= 0.95
). Simulated traje tories for weight(Fig.5B)and ondition index(Fig.5C) werealsonearthe observed data;two
partial spawnings were simulated at days 154 and 187. Laruelle (1999) and
Calvez (2003) showed that partial spawnings indu ed asyn hrony of gonadal
state, resulting in in reased variability of esh weight within a ohort. The
two large standard deviations for weight in the eld data between days 154
and day 187 tend to onrmthe o urren e of su h spawnings.
During the whole gamete release period, from day 150 (late May) to day
240(beginingofSeptember),the simulatedtraje torydivergedfromobserved
data be ause of spawnings. The importan eof the prin ipalspawning in the
fall indu ed a divergen e of the model ompared to observed data. These
di-vergen es indu ed lower oe ients ofdeterminationbetween simulationand
elddataforweight(
r
2
= 0.63
)and onditionindex(r
2
= 0.78
).For onditionindex,slopeoftheregressionlinebetweenexperimentalandmodeleddatawas
0.84 and signi antly dierent from 1 (t-test, p-value < 0.05). Nevertheless,
for both weight and length, simulations were within the observed standard
deviationrange.
Lilleau
Validation simulations for the Lilleau site are shown in Fig. 6. For all state
variables, the slopes of the regression lines between observed and simulated
values were not signi antly dierent from 1 (t-test, p-value > 0.05). The
simulated length traje tory followed losely the observed values at Lilleau
(
r
2
= 0.89
) andlaywithin standard deviationrange(Fig. 6A).Conditionin-dexwaswellexplainedbythemodel(
r
2
= 0.72
)andtheweightsimulationwasbetter explained than in the Nole validation simulation(
r
2
= 0.77
). Like theNole simulation, the rst gamete emissions were simulated at the end of the
spring (day 144, late May) and ontinued duringthe summer. The simulated
prin ipal spawning o ured at day 227 (mid August), whi h was onsistent
with the observation of Laruelle (1999) and Calvez (2003) for the oast of
Brittany,Fran e. In earlysummer, many (8) partialspawnings are simulated
whi h are not des ribed in the eld dataset of Goulletquer (1989a), thus
26
28
30
32
34
36
38
40
42
44
50
100
150
200
250
300
350
400
450
500
Length (mm)
Time (Days)
A
0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
50
100
150
200
250
300
350
400
450
500
Weight (g DW)
Time (Days)
Spawning
B
20
30
40
50
60
70
80
90
100
110
120
130
50
100
150
200
250
300
350
400
450
500
Conditon index
Time (Days)
Spawning
C
Fig.4.Results oftheadjustment of themodelto experimentaldataof Goulletquer
(1989a)for(A)length (mm);(B)weight (gDW)and(C) onditionindex.
Continu-ouslines: modeloutput.Crosses(+):observeddatapoints.The arrowindi atesthe
datewhere spawning wasobservedbyGoulletquer(1989a).Time (days) is ounted
fromthe1st ofJanuary 1984.
partial spawnings are in a ordan e with Laruelle (1999) and Calvez (2003)
observations.A simulatedlate spawningevent o uredatday 285(mid
O to-ber). This gameterelease event wasnoted inthe eld dataset ofGoulletquer
(1989a) who emphasized that at Lilleau station the reprodu tion eort was
26
28
30
32
34
36
38
40
50
100
150
200
250
300
350
400
450
500
Lenght (mm)
Time (Days)
A
0
0.1
0.2
0.3
0.4
0.5
0.6
50
100
150
200
250
300
350
400
450
500
Weight (g DW)
Time (Days)
B
20
30
40
50
60
70
80
90
100
110
120
50
100
150
200
250
300
350
400
450
500
Condition index
Time (Days)
C
Fig.5.Resultsofthevalidationofthemodelusingtemperatureand
Chla
timesseriesat theNole Sation for length (A), weight (B) and onditionindex (C).Continuous
lines: model output; rosses (+): observed datapoints; errorbars indi ate standard
deviationforelddata(Goulletquer,1989a).Standarddeviationsfor onditionindex
werenot available. Time(days) is ountedfrom the 1stof January1985.
4 Dis ussion
Shell growth
An important and original aspe t of this model, whi h is derived from the
Hofmannetal.(2006)hard lam growthmodel, omparedtootherindividual
growthmodelsbasedontheS opeForGrowth on ept(e.g.Barilléetal.,1997;
Grantand Ba her, 1998; Pouvreau etal., 2000; Hawkins et al.,2002;
Gangn-eryetal.,2003;Fré hette etal.,2005),isthat ondition indexdrivesbothnet
24
26
28
30
32
34
36
38
40
42
44
46
48
50
100
150
200
250
300
350
400
450
500
Length (mm)
Time (Days)
A
0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
50
100
150
200
250
300
350
400
450
500
Weight (g DW)
Time (Days)
B
20
30
40
50
60
70
80
90
100
110
120
130
50
100
150
200
250
300
350
400
450
500
Condition index
Days (Jours)
C
Fig.6.ResultsofthevalidationofthemodelatLilleauforlength(A),weight(B)and
onditionindex(C).Continuouslinesshowmodeloutput, rosses(+)showobserved
data points and errorbars indi ate standard deviation for eld data (Goulletquer,
1989a). Standard deviations for ondition index were not available. Time (days) is
ountedfromthe1st ofJanuary 1985.
bivalves, length is al ulated from weight through an allometri equation. In
present study weight and length are two state variables independently
al u-lated. Su h a formulation for the Manila lam allowed un oupling of length
growth in rements and weight growth in rements. This is in agreement with
observationsofLewisandCerrato(1997)onMyaarenaria,whi hshowedthat
shell growth ould be un oupled from somati growth. Simulationswere in a
very good agreement with eld data as the oe ient of orrelation values
were very high (
r
2
= 0.89
at Lilleauandr
2
= 0.95
atNole). This formulationalso allowed our model to simulate post-spawning shell growth stops, whi h
For both Nole and Lilleau sites, the simulated reprodu tive pattern
devi-ated somewhat from eld observations. The aim of most studies on growth
of R. philippinarum is to assess the best ultivation strategy for
maximiz-ing growth. Most often, authors monitor the evolution of biologi al variables
(length,weight and ondition index are amongthe most ommon) of aset of
homogeneously sized individuals (i.e. a ohort Maître-Allain,1982; Bodoy
etal.,1980; Goulletqueretal.,1987;Goulletquer, 1989a;Robertetal.,1993).
Onea hsamplingo asion,allindividualsinthesampleare sa riedtoassess
thosevariablesand anaveragelength,weightor onditionindex is al ulated.
Usingsu h amethod,oneobtainsthe traje toryofthe averageweight,length
and ondition index of a ohort through time. It is often said to be the
tra-je toryof an"averageindividual"(i.e. anindividualthat has average
physio-logi al features),although noreal individual traje tories are involved in su h
adata set.
The o urren e of asyn hronous events, likepartial spawnings, may indu e a
dieren ebetweenthetraje toryofan"averageindividual"andthetraje tory
of the average of a ohort.In order toexplain this dieren e, two theoreti al
ases arehypothesized both of whi hlead tothe same average ohortweight
time series (Fig.7).
Weight
Weight
Time
Time
Time
Weight
Time
Weight
Case 1
Case 2
Cohort average trajetory
Individual trajectories
Fig.7.Twohypotheti al asesoftheinter-individualvariabilityofspawningstrategy
ina ohort.Case 1:the ohort is omposedofindividualspresentingsimilar
spawn-ing eorts, with ea h individual spawning four times, but not simultaneously. The
individualtraje tories graphshowstheevolution ofweightsoftwoindividuals
( on-tinuousanddotted lines). Case2 :the ohortis omposedofindividuals presenting
dierentspawningstrategies.One lassofindividualsspawnsfourtimes( ontinuous
line in individual traje tories) and the other doesn't spawn at all (dotted lines in
individual traje tories). In both ases,be ause of theasyn hronous spawnings, the
averageweight ofthe ohort mayremain onstant.
spawn-the individual weight traje tories. If these spawnings don't o ur
simulta-neouslyinalltheindividualsofthe ohort,the average weightofthe ohort
may remain onstant.
•
Case 2 is a situation where the ohort is omposed of two kinds ofindi-viduals: some of them spawn with a high frequen y and the other ones
don't spawn at all. As in ase 1, if spawnings, when they o ur, are not
syn hronous, the average weight of the ohort may alsoremain onstant.
Inboth ases, one mayobtainthe same evolutionof averageweight;however,
theweightvarian ewillbehigherin ase1.Itisimpossible,however, tomake
assumptionsabout thebehaviorof individualsand todistinguishbetween the
two asesjustbyexaminingthemeanandvarian eoftheweightofthe ohort.
Pro esses of reprodu tionin bivalves are mu h more omplex than these two
examples;however, these twotheoreti al ases illustratethat if asyn hronous
events o ur in a ohort, the average traje tory of the ohort doesn't allow
any assumption about individual behavior.
Consequently, simulations of individual variables annot easily be ompared
toobserved data when asyn hronous events o ur within a ohort.Tryingto
ta modelsimulatingthe evolution ofmean weightof the ohortwhen
asyn- hronous partial spawning events o ur would not in lude these events thus
resultingin a large underestimate of the reprodu tiveoutput be auseit does
not take into a ount the peak reprodu tive output of individuals lams. In
su h a ase, a high value of the oe ient of orrelation (i.e. tending toward
1)between simulatedand elddatamaya tuallyrevealapoorrepresentation
of individual behavior. This points out (1) the di ulty of omparing
mod-eled toobserved dataduring the reprodu tiveseason and (2)the di ulty of
evaluatingquantitatively the reprodu tiveoutput of an"average individual".
For allthese reasons, deviations of the modelfrom observed weight and
on-dition index values during the gamete-emission season appear relevant: they
explainthelowervaluesofthe oe ientsofdeterminationobtainedforweight
and ondition index inthe validationsimulationsatNole(
r
2
= 0.64
and0.77
,respe tively) and at Lilleau (
r
2
= 0.77
and0.71
, respe tively); they alsoex-plain the weak value of the slope of regression line between observed and
modeled ondition indexatNole.During the spawning season,goodness oft
of thesimulated weight and ondition index an't bevalidated:poort ould
beattributedtoboth(1)highvariabilityamongindividualweightsand
ondi-tionindexesor(2)poormodelrepresentationoftheindividualenergybalan e.
Thusthemodel shouldbevalidatedoutsideof thereprodu tiveseason.When
ex luding the main spawning season (i.e. from mid May to mid September),
simulationsofboth variableswerevery lose tothe observeddata; oe ients
ofdetermination al ulatedforweightwere higher(
r
2
= 0.86
atNoleand0.84
simu-allstatevariables.Thisindi atesthatourmodelproperlyestimatestheManila
lam energy balan eunder for ing by temperatureand trophi resour e.
Several population dynami s models have been built for bivalves, whi h use
bioenergeti equations to simulate bivalves standing sto k transfers among
size lasses (Powell et al., 1992; Hofmann et al., 1992). They are not
indi-vidually based. This strategy allows a dire t omparison between the model
outputs and the experimental data means. Nevertheless, the present study
learly shows that observation at the ohort (or the population) s ale may
hide some individuals ale pro esses. As individuals an be onsidered the
basi entity of populations (Kooijman, 1995), the individuals ale pro esses
inuen epopulationspro esses.Thisproblemofs aletransferwasemphasized
by Kooijman(1995). The Manila lam model has been developed in order to
buildapopulationdynami s,individual-basedmodelfollowingthatdeveloped
by Hofmann et al. (2006) for Mer enaria mer enaria. This strategy permits
a ounting for the inuen e of individuals ale pro esses at the population
level.
Trophi resour e
The problem of estimating food resour es available to suspension feeding
bi-valves is well known and re ognized as a major problem in modelling
bi-valve energeti s (see e.g. Bayne, 1998; Grant and Ba her, 1998). The Manila
lam uses food resour es available at the sediment-water interfa e, and may
have a very omplex diet : it is known for ltering prey items as diverse as
ba teria, pi o yanoba teria (Nakamura, 2001) and diatoms, as well as
de-triti parti ulate organi matter and small rotifers (Sorokin and Giovanardi,
1995) fromthe water olumnand the sediment surfa e. The Manila lam has
alsobeen des ribed asingesting toxi and non toxi dinoagellatesby Liand
Wang (2001).Moreover, growth ofmi rophytobenthosis very ee tiveonthe
MarennesOléron mudats (Guarini et al., 2000), and was shown to be an
importantfoodsour eforoysters Crassostreagigas (Rieraand Ri hard,1996)
on e resuspended by tidal urrents (Blan hard et al., 1998, 2001). In su h
mudats, we an hypothesize that mi rophytobenthos is an important food
sour e fornear-bottomsuspension-feeding bivalves su has R. philippinarum.
Fegleyetal.(1992)haveshownthatshort-termvariabilityinthequantityand
nutritionalqualityoffooditemsavailabletointertidalnear-bottomsuspension
feeders like R. philippinarum may be of parti ular importan e for assessing
their growth and reprodu tion.
Insu ient information is available on erning these diverse potential food
sour esandthe asso iatedvariabilityoffeedingrateforthemtobe onsidered
a rough estimate of the food input (
Chla × C2F
) for lams exhibiting su h a omplex diet. Following Grant and Ba her (1998) with Mytilus edulis, oursimulations demonstrate that simple formulation of food and feeding may
su ein predi ting Manila lam energeti s. A areful study of the lam diet,
using arbon and nitrogen isotope ratios (Kasai et al., 2004) and pigment
analysis of bottom interfa e water POM, shouldimproveour knowledge.
5 Con lusion
Despite thedi ulty of omparing eldaveraged datasets tomodeloutputs,
simulationofgrowthand reprodu tionforanaverage individual lam under
naturalfoodand temperaturegaverealisti results.The model an beusedto
predi tlengthandweightgrowthaswellasgameteprodu tionofManila lams
in temperate e osystems. Further work will fo us on modelling intera tions
between environment,host and pathogenusing our results,whi hprovidethe
environment lam oupling.
A knowledgements
This study was nan ed by a NSF-CNRS joint program. It was also
sup-portedbythe ConseilRégionalde Bretagnewithinthe MODELMABregional
resear h program. The authors thank Me Fola e and B. Le aunar for their
te hni al help. Thisis IUEM ontributionNo 1023.
Referen es
Allam,B., Paillard, C., 1998. Defense fa tors in lam extrapallialuids. Dis.
Aquat. Org. 33, 123128.
Allam, B., Paillard, C., Howard, A., Le Penne , M., 2000. Isolation of the
pathogen Vibrio tapetis and defense parameters in brown ring diseased
Manila lams Ruditapes philippinarum ultivated in England. Dis. Aquat.
Org.41, 105113.
Barillé, L., Héral, M., Barillé-Boyer, A.-L., 1997. Modélisation de
l'é ophysiologie de l'huitre Crassostrea gigas dans un environnement
es-tuarien.Aquat. LivingResour. 10,3148.
Bayne,B.L., 1998.The physiologyof suspensionfeeding by bivalvemollus s:
anintrodu tiontothePlymouth'TROPHEE' workshop.J.Exp.Mar.Biol.
short-term dynami s of intertidal mi rophytobenthos by the
exondation-submersion y le. C. R. A ad. S i. Paris, S ien es de la vie / Life S ien es
321 (6), 501508.
Blan hard, G. F., Guarini, J. M., Orvain, F., Sauriau, P. G., 2001. Dynami
behaviourofbenthi mi roalgalbiomassinintertidalmudats.J.Exp.Mar.
Biol.E ol. 264, 85100.
Bodoy, A., Maitre-Allain, T., Riva, A., 1980. Croissan e omparée de la
palourde européenne (Ruditapes de ussatus) et de la palourde japonaise
(Ruditapesphilippinarum)dansun é osystèmearti ielmeditérannéen.Vie
Marine2, 3951.
Bower, S. M., 1992. Winter mortalities and histopathology in Manila lams
(Tapes philippinarum) inBritishColumbiadue tofreezing temperature. J.
ShellshRes. 11 (2), 255263.
Calvez,I.,2003.Appro hedelavariabilitéspatialed'unepopulationde
palour-desRuditapesphilippinarum(AdamsetReeve),auxstadeslarvaireset
post-larvaires.Thèse de Do torat, Université de BretagneO identale, Brest.
Dame, R., 1996. E ology of marine bivalves: an e osystem approa h. CRC
Press, Bo a Raton.
Fegley, S.R., Ma Donald,B. A., Ja obsen, T. R.,1992. Short-term variation
inthequantityandqualityofsestonavailabletobenthi suspensionfeeders.
Estuar.Coast. ShelfS i. 34 (4), 393412.
Flass h,J.P.,Leborgne,Y.,1992.Introdu tioninEurope,from1972to1980,
oftheJapanese Manila lam(Tapes philippinarum)andee ts on
aqua ul-tureprodu tion and naturalsettlement. ICESJ. Mar.S i. 194, 9296.
Fré hette,M.,Alunno-Brus ia, M.,Dumais, J.-F.,Sirois,R.,Daigle,G.,2005.
In ompleteness and statisti al un ertainty in ompetition sto king
experi-ments. Aqua ult. 246, 209225.
Gangnery, A., Chabirand, J.-M., Lagarde, F., Le Gall,P., Oheix, J., Ba her,
C.,Buestel, D., 2003. Growthmodelof the Pa if oyster, Crassostera gigas
ultured inThau Lagoon(Méditerranée, Fran e). Aqua ult. 215, 267290.
Gar ia, F., 1993. Interpretation of shell marks for the estimation of growth
of the European arpet lam Ruditapes de ussatus L. of the Bay of Fos
(Mediterranean Sea). O eanol. A ta 16,199203.
Goulletquer, P., 1989a. Etudedes fa teursenvironnementaux intervenant sur
la produ tion de palourde japonnaise d'élevage Ruditapes philippinarum.
Thèsede Do torat, Université de Bretagne O identale, Brest.
Goulletquer, P., 1989b. Mortalité hivernale hez la palourde japonnaise
Ru-ditapes philippinarum sur le littoral atlantique : aspe ts bio himiques et
é ophysiologiques.Haliotis17, 152163.
Goulletquer, P., Heral, M., Deslous-Paoli, J.-M.,Prou, J., Garnier,J., Razet,
D., Boromthanarat, W., 1989. E ophysiology and energy balan e of the
Japanese lam Ruditapes philippinarum. J. Exp. Mar. Biol. E ol. 132, 85
108.
philippinarum et l'huitrejaponaise Crassostrea gigas.Haliotis16, 453462.
Goulletquer, P., Wolowi z, M., 1989. The shell of Cardium edule, Cardium
glau um and Ruditapes philippinarum : Organi ontent, omposition and
energyvalue, asdetermined by dierent methods. J. Mar.Biol. Asso . UK
132,563572.
Grant,J., Ba her, C., 1998. Comparative models of mussel bioenergeti s and
the validationat eld ulturesites. J. Exp. Mar.Biol. E ol. 219, 2144.
Guarini, J. M., Blan hard, G. F., Gros, P., 2000. Quanti ation of the
mi- rophytobenthi primary produ tion in European intertidal mudats - a
modellingapproa h.Cont. ShelfRes. 20 (12-13),17711788.
Hawkins,A.,Duarte,P.,Fang,J.,Pas oe,P.,Zhang,J.,Zhu,M.,2002.A
fun -tionalmodelofresponsive-feedingandgrowthinbivalveshellsh, ongured
and validated for the s allop Chlamys farreri during ulture in China. J.
Exp.Mar. Biol.E ol. 281, 1340.
Hofmann, E.E., Klin k,J. M., Kraeuter, J. N.,Powell,E. N., Grizzle,R. E.,
Bu kner, S.C., Bri elj,M. V.,2006. A populationmodelof the hard lam,
Mer enariamer enaria:developmentoftheage-andlength-frequen y
stru -tureof the population. Journalof ShellshResear h 25,417444.
Hofmann, E. E., Powell, E. N., Klin k, J. M., Wilson, E. A., 1992. Modeling
oyster populations III. Criti al feeding periods, growth and reprodu tion.
Journalof ShellshResear h 11, 399416.
Kasai, A., Horie, H., Sakamoto, W., 2004. Sele tion of food sour es by
Ru-ditapes philippinarum and Ma tra veneriformis (Bivalvia:Mollus a)
deter-minedfromstable isotopeanalysis. Fish. S i. 70 (1), 1120.
Kooijman, S.A. L.M.,1995. The stoi hiometryof animalenergeti s. Journal
of Theoreti alBiology 177 (2), 139149.
Laruelle,F., 1999. Phénologie etdéterminisme de la reprodu tion hez
Rudi-tapes de ussatus (L.) and Ruditapes philippinarum (Adams and Reeve) en
Bretagne.Thèse de Do torat, Université de BretagneO identale, Brest.
Laruelle,F., Guillou,J.,Paulet,Y., 1994.Reprodu tivepatternof the lams,
Ruditapes de ussatus and Ruditapes philippinarum, on intertidal ats in
Brittany. J.Mar. Biol. Asso .UK 172, 6996.
Lawren e, D., S ott, G.,1982. The determinationand use of ondition index
of oysters. Estuaries 5,2327.
Lewis, D. E., Cerrato, R. M., 1997. Growth un oupling and the relationship
between shell growth and metabolism inthe soft shell lam Mya arenaria.
Mar.E ol. Prog. Ser. 158, 177189.
Li,S.-C.,Wang,W.-X.,2001.Radiotra erstudiesonthefeedingoftwomarine
bivalvesonthetoxi andnontoxi dinoagellateAlexandriumtamarense.J.
Exp.Mar. Biol.E ol. 263 (1), 6575.
Lu as, A., 1993.Bioénergétique des animaux aquatiques. Masson, Paris.
Lu as,A.,Beninger,P.,1985.Theuseofphysiologi alindi esinmarinebivalve
aqua ulture.Aqua ult. 44,187200.
game-J.Exp. Mar.Biol. E ol. 38,121133.
Maître-Allain,T., 1982. Inuen e du milieusur la roissan e de deux
palour-des,Ruditapesde ussatus etRuditapes philippinarum,dansl'étangde Thau
(Hérault).VieMarine 4,1120.
Meneghetti, F., Mos hino, V.,Da Ros, L., 2004.Gametogeni y le and
vari-ations in oo yte size of Tapes philippinarum from the lagoon of Veni e.
Aqua ult. 240, 473488.
Mortensen, S., Strand,O., 2000. Release and re apture of Manila lams
(Ru-ditapes pilippinarum) introdu edto Norway. Sarsia85, 8791.
Nakamura, Y., 2001. Filtration rates of the Manila lam, Ruditapes
philip-pinarum:dependen e onpreyitemsin ludingba teriaandpi yanoba teria.
J.Exp. Mar.Biol. E ol. 206, 181192.
Ngo, T. T. T., Choi, K.-S., 2004. Seasonal hange of Perkinsus and Cer aria
infe tions in the manila lam Ruditapes philippinarum from Jeju, Korea.
Aqua ult. 239, 5768.
Paillard, C., Allam, B., Oubella,R., 2004. Ee ts of temperature ondefense
parametersinManila lam Ruditapes philippinarum hallengedwithVibrio
tapetis. Dis. Aquat. Org. 59, 249262.
Paillard, C., Maes, P., 1990. Etiologie de la maladie de l'anneau brun hez
Tapes philippinarum :pathogéni itéd'un Vibrio sp. C. R.A ad.S i. Paris,
S ien es de la vie/ Life S ien es 310, 1520.
Paillard,C., Maes, P.,1994. Brown ringdisease intheManila lamRuditapes
philippinarum:establishmentofa lassi ationsystem.Dis.Aquat. Org.19,
137146.
Paillard, C., Maes, P., Oubella, R., 1994. Brown ring disease in lams. Ann.
Rev.Fish Disease 4, 219240.
Paillard, C., Per elay, L., Le Penne , M., Le Pi ard, D., 1989. Origine
pathogènede lamaladie de l'anneau brun hez Tapes philippinarum
(Mol-lusque, Bivalve). C. R. A ad. S i. Paris, S ien es de la vie / Life S ien es
309,335341.
Park, K.-I., Choi, K.-S., 2004. Appli ationof enzyme-linked immunosorbent
assay for studying of reprodu tion in the Manila lam Ruditapes
philip-pinarum (Mollus a:Bivalvia)I. Quantifyingeggs. Aqua ult. 241, 667687.
Pouvreau, S., Ba her, C., Héral, M., 2000. E ophysiologi al modelof growth
andreprodu tionofthebla kpearloyster,Pin tadamargaritifera:potential
appli ationsforpearlfarminginFren hPolynesia.Aqua ult.186,117144.
Powell, E. N., Hofmann, E. E., Klin k, J. M., Ray, S. M., 1992. Modelling
oysterpopulationsI.A ommentaryonltrationrate.Isfasteralwaysbetter
? Journalof Shellsh Resear h 2,387398.
Powell,E. N.,Stanton, R.J., 1985.Estimationof biomassand energy owof
mollus sinpalaeo- ommunities.Palaeontology28,134.
Pérez Cama ho, A., Delgado, M., Fernandez-Reiriz, M., Labarta, U., 2003.
Energy balan e, gonad developement and bio hemi al omposition in the
sostrea gigas along a trophi gradient in the estuarine bay of Marennes
Oléron.Estuar. Coast.Shelf. S i. 42,347360.
Robert, R., Trut, G., Laborde, J. L., 1993. Growth, reprodu tion and gross
bio hemi al ompositionoftheManila lamRuditapesphilippinarum inthe
Bay of Ar a hon, Fran e. Mar. Biol.116, 291299.
Shafee, M. S., Lu as, A., 1982. Variations saisonnières du bilan énergétique
hez les individus d'une population de Chlamys varia (L.): Bivalvia,
pe -tinidae.O eanol. A ta 5,331338.
Sorokin, Y.,Giovanardi,O., 1995. Trophi hara teristi sof the Manila lam
(TapesphilippinarumAdamsandReeve).ICESJ.Mar.S i.52(5),853862.
Thompson, R. G.,1984. Produ tion and reprodu tive eort inthe blue
mus-sel Mytilus edulis, thesea ur hin Strongylo entrotus droeba hiensis andthe
snow rab Chiono etes opilio from populations in Nova S otia and
New-foundland.J.Fish. Res. BoardCanada36, 955964.
Toba, M., Natsume, Y., Yamakawa, H., 1993. Reprodu tive y le of Manila
lams olle tedfromFunabashiwaters,Tokyobay.NipponSuisanGakkaishi