International Council for the
Exploration ofthe Sea.
C.M. 1993/ F:22
Mariculture Cominittee
MODELING JAPANESE OYSTER PHYSIOLOGICAL PROCESSES UNDER NATURAL
TIDAL VARIATION IN SUSPENDED PARTICULATE MATTER.
Laurent Barille l •2 & Jean Prau l •
1. IFREMER UREA, BP 133,17390 La Tremblade, France.
2. LBBM, IUT La Rochelle, 17026 La Rochelle Cedex, France.
•
ABSTRACT.
Feeding and growth ofthe Japanese oyster Crassostrea gigas is simulated by a deterrninistic model.
However, physiological pracesses, estimated by statistical relationships, have not been tested under
extreme environmental conditions such as high level ofsuspended particulate matter. Ecophysiological
experiments were therefore conducted in sitll in the Bay ofMarennes-Oleran during aspring and neap
tide cycle in May 1991, to get a better representation of2 important components ofthe energy budget
ofthe Japanese oyster : clearance rate and selection efficiency. The relationships previously established
by Raillard et al. (1993) are re-evaluated by integrating the action ofhigh seston load, typical oftidal
effect within estuaries, on the physiological pracesses.
For the clearance rate, the re-evaluated parameterization includes:
•
- a negative effect ofturbidity.
- a clogging threshold set to 192 mg/I.
Selection efficiency parameterizations are modified by incorporating :
- a constant inorganic ingestion for increasing seston load, set to 17 % of the inorganic filtered
ration.
- an upper selection threshold for the particulate organie matter caused by an overload ofparticulate
material. This threshold was set to 160 mg/I.
The oyster grawth model should be enhanced by integrating theses new equations built-up on realistic
field conditions, where food availability is regulated by tidal variation.
INTRODUCTION
A mathernatica1 model which simulates feeding and grO\vth ofthe Japanese oyster Crassostreagigas in the
bay ofMarennes-Oleron was first studied by Bacher (1991) on the basis ofthe energetical budget model
developed byBayne et al. (1976) forMytilusedu/is. Thisoystermodel wasmodified by Raillard et al. (1992)
•
who added the effects offood quality and quantity on the physiological processes. Theses authors integrated
the preferential ingestion oforganie matter. They parameterized the orgamc / inorganic sorting ofparticles
wmch occurred during pseudofaeces production. However, the functional relationship used in the model to
simulate the orgariic enrichment was derived from laboratory experiments with a maximum seston
concentration of50 mgll. Furthermore, simulation ofclearance rate included the assumption that there was
no effeCt ofparticle concentration till aclogging threshold (empiricallysetto 200 mw1) from which clearance
rate ceased.
In situ experiments were therefore designed to get a better representation ofclearance rate and selection
efficiency under natural conditions of food availability and composition. The model of the energetics of
suspension-feeding oyster should be enhanced by incorporating theses new parameterizations since they
integrate short-term changes in food quality and quantity typical of tidal effects within the estuary of
il.farennes-Oleron. The others components ofthe energy budget, not presented in this study; have been
evaluated simultaneously to avoid building the model \\ith isolated determinations.
MATERIALS AND METHODS
•
Theexperimentwas setup in LeChapus inMay 1991 duringa spring andneaptidecycIe, (6g.1). Physiological
responses to tidal variation in food quality and quantity were studied on a population (100 oysters ;total dry
.weight = 134 g) using a raceway system, and on 6 indi\iduals placed in experimental chambers, (Mean dry
weight = 0.6045 ± 0.076). Informations about Suspended Particulate Matter (SPM) and chlorophyll-a and
pheopigments, continuously recorded by 2 nephelometers and 2 turbidimeters, were sent to a computer
through an acquisition card, Analog connection. In the meantime hourly water sampies were collected in
order to calibrate both signals and obtain an estimation of the Particulate Organie Matter (pOM) and
Particulate Inorganic Matter (pIl\1).
Physiological determinations.
a. Clearance rate
Clearance rate was determined both for the population and the individuals with 2 different methods.
It was determined for the population by direct1y measuring chlorophyll-a fluxes from the in- and outflow of
the raceway. In Le Chapus, Barille et al. (1993) have shown that fluorescence is less sensitive to clearance
rate underestimation due to retention efficiency since most chlorophyll is within large particles. Clearance
rates were then related to SPM.
For the individuals the relationship between clearance rate and SPM was estimated from derivation ofthe
relationship previously established between total illtration rate and SPM Total illtration rate (rriglh) was
estimatedindirectlybymeasuringbiodepositions rates. Thismethodisbasedontheassumptionthatinorganic
matter passes unaltered through the gut. In l\1arennes-Oleron 90 %, as an average, ofthe oyster diet is
composed ofmorganic material. It is therefore reasonable to assume that the ingested inorganic matter is not
absorbed.
b. Selection efficiency
Selection efficiency was studied on the individuals for the 2 components ofthe SPM : PIM and POM.
•
Organie selection was defined as the ratio oforganie pseudofecal rate arid organie illtration rate. Inorganic
selection was defined as the ratio of inorganic pseudofecal rate and inorganic illtration rate. Selection
efficiencywas quantified this way forthe first time by Ba}ne et al. (1976) inM}ti/usedulis. Although others
formulation ofselection efficiency are available (Kimboe et al. 1985 ;Bayne et al., 1989), selectionefficiency
calculated in this study was chosen in accordance with the previous relationships used in the oyster model
by Raillard et al. (1992).
Population behaviour was compared to the indi'.idual response by standardazing the clearance rate to a
common dxy weight of 1 g using the formula:
Ys=(lI\Ve)bYe
Ys is the standardized clearance rate, We the experimental weight, Ye the uncorrected parameter and b is
the allometric pmver coefficient. b = 0.4 (Fiala-Medioni & Copello, 1984).
•
RESULTS-DISCUSSION
The evolutionofSPl\1, stronglyrelated to the forthnightly cydeshowedfluetuations from 30 mgllunderneap
tide conditions to as much as 350 mg'1 during the spring tide, (6g. 1). The organic content, POM%, which
is one ofthe parameters used to assess food quality, appeared related to SPM byan inverse relationship, (6g.
2),
POM% = a + b/SPM
The coefficients ofthis model are derived from the follm'.ing relationship established between POM and
SPM:
POM (mg/1) = a SPl\f + b, a = 0.087975 b = 3.958
(~=
0.88 ; n = 90)
A single model describing the evolution of organic content .versus. SPM was used since no significant
differences were found between the 2 foIIowing relationships calculated for the spring and neap tide:
Spring tide POM = 0.0836 SPM + 4.578 (rZ = 0.94 ; n = 43)
Neap tide
•
POM = 0.0938 SPM + 3.749 (rZ = 0.83; n =47)
The2subsequentmodelsdescribingtheorganiccontentvariationstoSPl\1concentrationsareshowninfigure
3.
Such an inverse relationship as been reportedby Preston & Prodduturu (1992) intheMerseyestuarybetween
particulateorganic carbonand SPl\Land arises from the dilution oforganic matterbyresuspended inorganics
sediments. In Le Chapus, inorganics sediments can be eitherlocally resuspended underthe action ofcurrent
speed or transported from an other part ofthe bay after \vind-induced resuspension on the tidal front.
It is Iikely that in the absence ofphy10planetonic blooms, the detection ofmarked variations in food quality,
which is a necessity for the modelling to cover the \\ider range ofvariation for the parameters and their
interaetion, will only be possible through seasonal fluctuations.
EfTect of turbidity on cle..m mce rate.
a. population
•
The evolution ofthe cIearance rate was established for 6 h continuously during a spring tide cycle v,ith a
calculation every minute. llus cIearance signal was then compared to the evolution ofSPM recorded by the
nephelometer. (fig 4a). Alinear regression calculated between these 2 parameters reveals the negative eifeet
of SPM, in the range 50-192 mgll, on the standardized cIearance rate (llh/g dry weight):
CRpop = - a SPM + b
with a = 0.0102 b = 4.213
n=303; r=0.81
The intercept value, 4.213 Vh/g DW, gives us an estimation ofthe higherievel ofcIearance within the range
of SPM encountered.
It is to be noticed that such a relationship did not occured on the whole semi-diurnal cycle.
b. individuals.
Clearance rate was indirectIy estimated from the total filtration rate.
Total filtration rate (mg/h) was expressed as a multilinear fimetion ofSPM that can be written as following:
TFR = - a SPM 2 + b SPM
with a = 0.00931 and b = 4.1966
n = 17 ; r = 0.96
The model fitted from the observations (Fig. 4b) was derived to get an estirnation ofclearance rate:
CRind = TFR/SPM = - a SPM + b (figure 4c)
This clearance rate calculated for individuals was standardized to the weight of an oyster of 1 g. The
standardi.zed relation can be e.xpressed as folIows:
CRind (l/h/g DW) = -0.01137 SPM + 5.12
•
The representation ofboth relationships (indi\idual and population) in figure 4d. shows a similar evolution
ofCR .versus. SPM, although the level offeeding was higher for the individuals. This can be attributed to
the activity ofthe population since Razet et aI. (1990) have shown that only 90 % ofthe oysters were active
in the raceway.
FromtheseresultsanewapproachforparameterizationofSP11effeetonclearanceratewas considered based
on individual behaviour.
- From 0 to 50 mgll; previous studies (Deslous-Paoli et aI., 1987 & 1992) have shown that clearance rate,
within this range ofSPM, was independent ofconcentration. The basic level ofclearance was set to 5.121/
h/g D\V, which corresponds to the optimal clearance in the case ofindividual.
- From 50 to 600 mgtl; the evolution ofclearance rate is described by the follo\\ing equation:
CR (VhIg DW) = (-0.01137 SPl\f .... 5.12)*(Exp(0.07(Min[0;192 - SPM]»)
This equation integrates:
•
(i) the negative effect ofSP11 on CR in the range 50-192 mgll.
(ü) a clogging threshold set to 192 mg/l that have been observed on 6 individuals submitted for 2 hours
to a mean SPM concentration of 192 mg/l.
The re-evaluated relationship is given in figure 4e.
Selection efficiency (SE).
A graphical representation ofPOl\VSPM ratio in the pseudofaeces versus POl\VSPM ratio in the feod,
(figure5a), showsthatPOM is beingpreferentiaIly ~ngested since all pointsareIocatedbeIowthey= xstraight
line, forwhich there is no seleetion. However, POM selection seems inefficient for high level ofparticulate
matter represented by 3 triangles indicating samples collected above 160 mg/l. The oysteris responding like
ifabove 160 mg'l, the selectivemech:mismwas not operative, Iikelycausebyan overload ofparticulatematter.
Non linear relationships were fitted to the evolution oforganie and inorganic selectionversus concentration.
SEi = 0.83 ( 1- Exp(- 0.0716( SPM - 5») n = 243; r2= 0.42
SEo = 0.53 ( l-Exp(- 0.0311( SPM - 5») n =238; r2= 0.51
As SPM ranged from 30 to 192 mg'l, for the indi..idual sampies, the equations were forced to begiri from
•
a SPM threshold of5 mgII that have been described by Deslous-Paoli et aI., (1992). It corresponds to the
threshold ofpseudofaeces production for Crassostrea gig~.
The value forwhich the curve levels offin the case ofPIM, 0.83, is simi1arto the resuh obtairied by Raillard
et aI. (1992) represented in figure Sb. It reveals that for increasing SPM concentration up to 150 mgll, 17
% ofthe inorganic filtered matter is still being ingested. In the meantime, 47 % ofPOM filtered is being
ingested.
A re-evaluation ofselection efficiency is proposed. It takes into account :
(i) a constant PIM ingestion for increasing SPM level set to 17 % ofthe filtered ration. TIüs appears more
realisticthanthepreviousparameterizations(RailIardetal.,1992), whichassurned,inthemodel,thatallPIM
filtered was rejected, above 50 mgJl, through pseudofaeces production.
(ii) an upper selection threshold for POM set to 160 mgll. Above this threshold POM is treated like the
inorganic component ofthe particulate matter which had not been submitted to preferential selection. The
curve which represents organie selection, figure Sc, reaches, above 160 rrigll, the level ofinorganic selection
previoulslyestimated. The selection model is thus simulating an 83 % ofrejection oftheorganiefiltered ration
The mathematical formulation of SEo model is thus modified into:
(O.S3( l-Ex-p(O.031(Min[O:5-SPMl ))))+O.3( l-Exp(O.l (Min[O;160-SPM])))
CONCLUSION
Resuspended sediments plays a major role in the estuarine complex ofMarennes-Oleron. The knowledge
ofspatial and temporal variability ofsuspended particulate matter peImits to build up hydrodynamic models
which can predict evolution ofseston concentrations. Primary production and oyster growth models show
strong influence of suspended paniculate matter on physiological funetions (Raillard, 1991).
However, suspended particulate matter is insufficient to characterize food supply for oysters. Barill6 et al.
e1993) showed the importance ofcross effects ofloads and particle size distributions on filtration processes.
Morever, further studies will define precisely feeding parameters that best quaIify the organie fraction ofthe
•
water column. The relative importance of microphytobenthos and pelagic phytoplankton blooms, their .
temporal and spatial variability, cell sizes. species composition and pigments characterization will lead to a
better accuracy in the oyster growth model.
The elaboration of new relationships based on field experiments that fall within the range of estuarine
environmental conditions is essential for a realistic modeling ofoyster growth. To understand the oyster
feeding response to natural food fluetuations (high levels ofseston load in the present study), the evolution
ofclearancerateand selection efficiencythat are2important componentsofthe energybudget ofCrassostrea
gigas. have to be considered simultaneously, even though the statistical relationships which simulate their
variation have been established independently. As an example, 17 % ofPIM ingestion, calculated from the
inorganic selection equation, would lead at high SPM concentration to unrealistic inorganic ingestion.
However,thiswillneverbethecasesincethecloggingthresholdwillintervenebylimitingtheinorganicfiltered
ration.
The ingestion rate used in the aetual oyster model will be modified according to the equations provided in
this study. Finally, the model will be completed by the parameterization ofabsorption efficiency.
BmLIOGRAPHY
Bacher C., 1\1. Heral, J-.1\f. Deslou-Paoli & D. Razet (1991). Modele energetique uniboite de la
croissance des huitres Crassostrea gigas dans le bassin de Marennes-Oleron. Can. J. Fish. aquat.
Sci., 48, 391-404.
Barille L. J. Prou, 1\1. Berat. & S. Bougrier (1993). No influence offood quality on ration-dependent
retention emciencies in the japanese oyster Crassostrea gigas. J. Exp. Mar. Biol. Ec()l. (in press).
Bayne B.L., J. \Viddows & R.J. Thompson (1976). Physiological integration, in, Marine musseIs
: theirecologyandphysiology, B.L. Bayne, editor. IBP n° 10, Camnbridge UniversityPress, 261291.
Bayne B.L., A.J.S. Bawkins, E. Navarra & I.P. Iglesias (1989). Effects ofseston concentration
on feeding, digestion and growth in the musselA{).·tilus edulis. Mar. Ecol. Prog. Ser., 55,47-54.
Deslou-Paoli J-.1\I., 1\1. Heral. 'V. Boromthanarat, D.Razet, J.Garnier, J.Prou & L.Barilie
(1987). Evolution saisonniere de la filtration de bivalves intertidaux dans des conditions naturelles.
Oceanis, 13, 575-579.
Deslou-Paoli J-.1\I., A.1\I.Lannou, P. Geairon, S. Bougrier, O. Raillard & 1\1. Beral. (1992).
Effects ofthe feeding behaviour ofCrassostrea gigas (bivalve molluscs) on biosedimentation of
natural particulate matter. Hydrobiologia, 231, 85-91.
Fiala-1\ledioniA. & 1\1. Copello (1984) Relation trophique entre l'huitre et le milieu: influence de
la concentration et de la taille des particules. Bases biologiques de l'Aquacu/ture. Montpe//ier.
1983. IFREVER. Actes et Colloques 11° 1,63-74.
KiorboeT. et F. 1\Iohlenberg (1981). Particle selection in suspension feeding bivalves. Mar. Ecol.
Prog. Ser., 5, 291-296.
Preston M.R. et P. Prodduturu (1992). Tidal variation ofparticulate carbohydrates in the Mersey
estuary. Estuar. Coastal and She/fSei., 34, 37-48.
Raillard 0., (1991). Etude des interactions entre les processus physiques et biologiques intervenant
dans la production de l'huitre Crassostrea gigas (Thunberg) du bassin de Marennes-Oleron: essais
de modelisation. These. Universite ParisVI.
Raillard 0., J-.M. Deslou-Paoli, M.IIeral & D. Razet (1992). Modelisation du comportement
nutritionnel et de la croissance de l'huitre japonaise Crassostrea gigas.Ocenol. Acta., 16, 73-82.
Razet 0., M. IIeral, J.Prou, J. Legrand & J.l\I. Sornin (1990). Variations des productions de
biodepots (feces et pseudo-feces) de l'huitre Crassostrea gigas dans un estuaire macrotidal: baie
de Marennes-Oleron. Haliotis, 20,143-161.
100
@
90
c:
C1>
80
'0
t.:=
cu
C1>
Neap tide
Spring tide
0
u
70
-c
i=
60
50
07/05
12/05
17/05
,,-.,
27,
®
20
Co
22/05
15
I
::1.
'-"
l:l..
+
u
~(
10
5
0
07/05
~1
12/05
17/05
22/05
27/05
350
©
300
,,-.,
250
~
a
200
:B
l:l..
150
'-"
CI)
100
50
0
07/05
~. ~
12/05
17/05
22/05
27/05
Figure 1. A. Evolution ofthe tide coefficient in May 1991 at Le Chapus. B. Tidal variation in the sum
ChlorophylI-a + Pheopigments (C+P). C. Tidal variation in Suspended Particulate Matter (SPM).
0.35
0.3
,.......
~
.........
--
0.25
c:
Q)
c:
0
u
u
c:
0
0'1
'-
0.2
0.15
0
0.1
0.05
50
0
100
150
200
Suspended particulate matter (mg/I)
-e-
Spring tide cycle
__ Neap tide cycle
Figure 2. Organie content models for the spring and neap tide in May
1991 at Le Chapus.
0.35
0.3
....---
---~
0.25
c
Q)
C
0
u
u
0.2
'e
0
O'l
~
0.15
0
IllIIllI
a
0.1
_
lllI
IID
IID
IID
lllI
0.05
0
50
100
150
200
Suspended particulate matter (mg/I)
Figure. 3 Organic content model in May 1991 at Le Chapus.
200
4.5
150
®
'00
..:::::::..
C'l
E
::::E
0Vl
IV
U
c::
~
50
o
IV
U
2.5
2
' 2 00
:
PM
0' :'2
02:24 PM
04:48 PM
PM
0~:~6 PM
06:00
__ SPM_CR
.c
-- ..
4
~400
..c:
a
~
©
..c:
&&
"
lJ'
300 E
•
:J
0
&
...0
1-0
400
&&-
1! .3
.B~ 300
"-
U
l)
c
.-
§ 200
®
~
1-0
100
o
200
...00 2
~
G:
500
5
500
~
0
0 7 : ' 2 PM
PM
08:24 PM
o
50
100
150
SPM mg/l
0
u
.;;
ü
L;:
100
100
200
0
250
200
1,50
SPM mg/I
• eR
6..-------------------,
6
:; 5
....- .. ,.
Q
CI
":::.
.>:
4
-"...
J
0
--.. •••
®
••
~192 mg/I: clogging threshold
•
••
•
•
Ql
Ql
l)
a... 2
•
U
c:
... 2
0
o
0
u
Ql
Ü
Ü
ol.....-
o L-_---'"--~----'---"O"'--~
o
50
100
SPI.4 mgjl
• individual
150
200
25C
,2
o
......60
130
...;;;,
200
-.l
270
340
410
480
550
620
SPIA mq/I
• Population
Figure 4:
A Simultaneous evolution of Suspended Particulate Matter and standardized Clearance Rate during aspring tidal cycle.
B Relation between Filtration Rate and SPM calculated on individuals during spring and neap tidal cycles.
C Clearance Rate of individuals derived from filtration rate model.
D Comparison of clearance rate estimated from individuals and from the population.
E Clearance rate model versus Supended Particulate Matter.
0.3
Ul
Q.)
u
0.25
Q.)
I:'
"0
-0
='
Q.)
0.2
Ul
C-
E
Cl.
V"l
..........
E
0
a...
0.15
111
111
lIlI
lIlI
0.1
lIlI
111
lIlI
111
111
lllI
111
111
0.05
0.1
0.15
0.2
Pom/Spm Food
SPM > 160 mg/I
....
.........
0.8
u>c::
v
0.6
0.25
0.3
~
........
'(3
t.::
v
c::
0.4
0
:;J
u
..2
v
(/)
0
0.2
o
o
10
30
50
70
-e- SEc. Rcillcra et cl. (1993)
-6-
90
1 10
130
150
170
190
SPM mg/I
Sei. this stuay.
_SEi. Rcillcra et cl. (1993)
_
SEc. this stuay.
1
0.8
0.5
c::
o
u
--
©
0.4
Q.)
Q.)
V"l
0.2
o
o
50
130
200
270
340
410
480
550
SPM mg/I
-e- SEc _
SEi
Figure 5.
A Selection of Particulate Organic Matter through pseudofaeces production.
B Selection effidency models for Particulate Organic Matter and Particulate Inorganic Matter.
e Selection efficiency model including an upper selection threshold for POM.
52C