Journal of Sea Research 73 (2012) 3240

Contents lists available at SciVerse ScienceDirect

Journal of Sea Research

journal homepage: www.elsevier.com/locate/seares

A model of clearance rate regulation in mussels

Marcel Frchette
Institut Maurice-Lamontagne, Ministre des Pches et des Ocans, C.P. 1000, Mont-Joli, Qubec, Canada G5H 3Z4

a r t i c l e

i n f o

Article history:
Received 22 December 2011
Received in revised form 6 June 2012
Accepted 11 June 2012
Available online 29 June 2012
Keywords:
Clearance Rate
Dynamic Energy Budget
Phenotypic Flexibility
Functional Response
Mussel

a b s t r a c t
Clearance rate regulation has been modelled as an instantaneous response to food availability, independent
of the internal state of the animals. This view is incompatible with latent effects during ontogeny and phenotypic exibility in clearance rate. Internal-state regulation of clearance rate is required to account for these
patterns. Here I develop a model of internal-state based regulation of clearance rate. External factors such
as suspended sediments are included in the model. To assess the relative merits of instantaneous regulation
and internal-state regulation, I modelled blue mussel clearance rate and growth using a DEB model. In the
usual standard feeding module, feeding is governed by a Holling's Type II response to food concentration.
In the internal-state feeding module, gill ciliary activity and thus clearance rate are driven by internal reserve
level. Factors such as suspended sediments were not included in the simulations. The two feeding modules
were compared on the basis of their ability to capture the impact of latent effects, of environmental heterogeneity in food abundance and of physiological exibility on clearance rate and individual growth. The
Holling feeding module was unable to capture the effect of any of these sources of variability. In contrast,
the internal-state feeding module did so without any modication or ad hoc calibration. Latent effects, however, appeared transient. With simple annual variability in temperature and food concentration, the relationship between clearance rate and food availability predicted by the internal-state feeding module was quite
similar to that observed in Norwegian fjords. I conclude that in contrast with the usual Holling feeding module, internal-state regulation of clearance rate is consistent with well-documented growth and clearance rate
patterns.
Crown Copyright 2012 Published by Elsevier B.V. All rights reserved.

1. Introduction
Clearance rate regulation is usually understood as an instantaneous
response of suspension feeders to ambient seston concentration. In
current Dynamic Energy Budget (DEB) theory, for instance, clearance
rate regulation has been described using MichaelisMenten kinetics
(also known as a Holling's Type II functional response; Real, 1977).
According to this view, clearance rate saturates in response to current
high food concentration. The actual satiation mechanism may be gut
fullness or a similar feedback mechanism (Kooijman, 2010, p. 263).
Recent developments whereby different particle types are taken in
consideration follow the same instantaneous-regulation logic (Saraiva
et al., 2011a).
The above view is based on the assumption that clearance rate is
regulated by external control variables, such as seston concentration
and quality. Studies of clearance rate regulation have assumed external
regulation, irrespective of whether they were embedded in the Scope
for Growth (SFG) paradigm (e.g., Bayne and Widdows, 1978; Cranford
and Hargrave, 1994; Strohmeier et al., 2009; Widdows et al., 1979) or
the DEB paradigm (Saraiva et al., 2011a). Feeding activity, however,

Tel.: +1 418 775 0625; fax: +1 418 775 0740.

E-mail address: marcel.frechette@dfo-mpo.gc.ca.

requires energy. According to DEB theory, this is fuelled by reserves

(Kooijman, 2010). Therefore it must be recognised that feeding activity
is driven by biochemical processes and that MichaelisMenten or
similar kinetics apply to the relationship between energy reserves and
feeding activity. Consequently, variability in energy reserves provides a
potential source for internal state-driven regulation of clearance rate. In
contrast, environmental factors such as water viscosity (Podolsky, 1994;
Riisgrd and Larsen, 2007), for instance, are external and may add to
internal state-driven regulation of clearance rate. Based on rst principles, therefore, it appears that clearance rate regulation should result
from dual inuences, external and internal.
An empirical perspective leads to the same conclusion as above:
Some aspects of mussel growth and clearance rate regulation appear
to require internal state-driven regulation. I base this suggestion on
two sets of observations. The rst set of observations involves latent
effects during ontogeny. Pechenik (2006) dened latent effects as
characteristics that originate from a developmental phase that become apparent only during later stages. Such effects are particularly
obvious in studies of the effect of stress, particularly food level, during early ontogeny. Pechenik (2006) reviewed studies spanning
a large phylogenetic array and found latent effects in 38 of the 43
studies examined. More recent contributions concur with Pechenik's
analysis (e.g., Chiu et al., 2007). Typically, such studies report that
stress, generally low food availability, during the larval phase, results

1385-1101/$ see front matter. Crown Copyright 2012 Published by Elsevier B.V. All rights reserved.
doi:10.1016/j.seares.2012.06.005

M. Frchette / Journal of Sea Research 73 (2012) 3240

in reduced growth in juveniles, and survivorship may also be affected (Phillips, 2002, 2004).
Latent effects may persist, be amplied or vanish at subsequent life
stages (Podolsky and Moran, 2006). They also appear to depend on
the timing of stress application, as early stress has more pronounced
effect than stress applied late during larval stage (Chiu et al., 2008;
Phillips, 2004), in spite of the apparent ability of larvae to recover
from poor feeding conditions during larval life (Ben Kheder et al.,
2010; Pechenik et al., 2002). Latent effects are usually ascribed to the
consequences of low energy reserves with ensuing negative effects on
general individual performance (Phillips, 2002, 2004; Thiyagarajan et
al., 2003), although morphological factors may be involved (Pechenik,
2006; Phillips, 2002). Reduced clearance rate has been seen as the proximal mechanism linking lower energy content and latent effects (Chiu
et al., 2007; Pechenik et al., 2002).
The second set of observations concerns phenotypic exibility of
clearance rate regulation after metamorphosis. Most of these studies
involve reciprocal transfers between various sites and monitoring
subsequent changes in clearance rate. Worrall and Widdows (1983)
transferred clams Scrobicularia plana between three distant sites.
Depending on the origin and the receptor site, they found either partial
or complete acclimatisation of clearance rate to the new sites. Widdows
et al. (1984) transferred blue mussels between two sites and reported
that after 7 weeks, clearance rates of transferred mussels were intermediate between those observed in mussels kept in their original site. In a
similar experiment, Okumus and Stirling (1994) transferred mussels
between two Scottish sea lochs and found short term differences in
clearance rate, but in the long run transferred mussels eventually
conformed to the rate found in mussels from the receptor site. These
studies suggest that clearance rate regulation is in part dependent of
internal condition of the animals. This view was further strengthened
by Babarro et al. (2000a) who transferred mussel spat from the shore
to a culture raft in Ra de Arousa, NW Spain. Clearance rate of shore
spat was subsequently compared with clearance rate of raft spat with
a single source of water. Signicant differences occurred early after
transferring the mussels and progressively vanished within one month
when mussels were transferred in late fall (Babarro et al., 2000a). In contrast, no such differences occurred when mussels were transferred in
summer (Labarta et al., 1997). This is clear evidence that internal state
of spat regulated clearance rate. Internal-state driven seasonal variability in clearance rate was also implied in a study of blue mussel and sea
scallop clearance rates based on indirect measurements (Cranford and
Hill, 1999). Taken together, these studies show that internally-based
physiological factors are involved in clearance rate regulation.
In the present study, I develop a model of an internal-state feeding
module (hereafter ISFM) based on the regulation of ciliary beat frequency (CBF) and its effect on clearance rate. The model includes
internal and external regulating variables. I then incorporate ISFM
into a standard mussel DEB model. Finally, based on standard DEB
simulations I study the ability of ISFM and of the usual Holling's
Type II feeding module (hereafter HFM) to capture growth and clearance rate patterns in simple hypothetical situations that mimic actual
experiments. These studies examined 1) latent effects in clearance
rate and growth (Chiu et al., 2007; Phillips, 2002, 2004), 2) the effect
of environmental gradients on DEB-modelled oyster growth (Ren and
Schiel, 2008), and 3) exibility of feeding behaviour (Babarro et al.,
2000a; Labarta et al., 1997).

2. Clearance rate regulation: a model

Here I describe HFM used in standard DEB theory and I derive the
ISFM model. I use conventional terminology and notation in DEB
theory. Square brackets [ ] and braces { } denote quantities per unit structural volume and per surface area, respectively, and rates (dimensions
per time) have dots.

33

According to HFM (e.g., Rosland et al., 2009), food ingestion (p_ X ,

J d 1) is given by
2=3
p_ X fp_ Xm gV

X
X XK


1

where fp_ Xm g (J cm 2 d 1) is maximum daily ingestion rate per gill

surface area, V (cm 3) is structural body volume, X is food concentration, as estimated by chlorophyll a concentration (Chl a; g L 1) and
XK (g Chl a L 1) is the half-saturation constant of the feeding response. The effect of temperature is not included in Eq. (1), but in
the simulations below all rates are corrected for temperature dependence as in usual DEB theory (Kooijman, 2010). The last term of the
right-hand side of Eq. (1) accounts for the regulation of the feeding
response by (food) particle concentration. Eq. (1) assumes that for
an individual bivalve, food ingestion is instantaneously regulated by
food concentration only. This formulation overlooks the internal
state of the individuals as a determinant of feeding behaviour.
I begin the derivation of ISFM by considering that water processing
by bivalve gills is driven by ciliary activity (see e.g., Jrgensen, 1990).
Increased CBF entails increased force production (Teff et al., 2008), consistent with model predictions (Gueron and Levit-Gurevich, 1999). This
translates into increased clearance rate (Riisgrd and Larsen, 2007).
CBF is modulated by factors such as neurotransmitters (Jrgensen,
1990), and the combined effects of water viscosity and temperature
(Podolsky, 1994; Riisgrd and Larsen, 2007). Although the complete
CBF regulation machinery may be complex (Salathe, 2007), it has
been shown that CBF increases directly with energy (ATP) availability
in various organisms (Lieb et al., 2002; Noguchi et al., 2001; Teff
et al., 2008). I assume that a similar direct relationship exists in
suspension-feeding bivalve gills and that ATP availability is proportional to energy reserves of tissue mass. Therefore in the following I assume
that CBF depends of energy density, [E] (J cm3) in DEB terminology. To
model this relationship, I begin with Newton's heat transfer equation
(von Bertalanffy)


f_
k1 f_m f_
E

where f_ is observed CBF (d1), f_m is maximum CBF (d1) and
k1 (cm3 J1) is a constant.
Riisgrd and Larsen (2007) analysed the relationships between
kinematic viscosity and CBF, and between kinematic viscosity and
clearance rate at constant temperature (22 C). Based on their results,
it is clearly acceptable to assume that clearance rate increases linearly
with CBF (not shown). It follows that energy density directly impacts
clearance rate. Therefore, assuming that neglecting curvilinearity
between CBF and clearance rate is of little consequence across the normal
temperature range experienced by mussels, clearance rate per unit gill
surface ({R}, L cm2 d1) and maximum clearance rate per unit gill
surface ({Rm}, L cm2 d 1) can be substituted for f_ and f_ m , respectively. One gets
n o
C_ R
E

k1

n
C_

Rm

o n o
C_ R :

Clearance rate has been shown to vary in response to a number of

modulating agents such as neurotransmitters (see Jrgensen, 1990)
and high seston loads, inorganic (e.g., Widdows et al., 1979; Barill
et al., 2006) and organic (Mhlenberg, 1999; Riisgrd, 2001;
Riisgrd et al., 2011). Two cases may be recognised. First, the effect
of such agents may be mediated through an interaction with clearance rate itself. This is the case of high concentrations of particles in
the water. The intake rate of such particles is expected to increase
with clearance rate and particle concentration. This is depicted by

34

M. Frchette / Journal of Sea Research 73 (2012) 3240

the term gi Xi{R} in Eq. (4), where gi is a constant, Xi is particle concentration and the subscript i denotes the ith particle type. Second,
the effect of modulating agents can be exerted independently of
clearance rate. This is expected in the case of neurotransmitters or
in the case of kinematic viscosity, for instance. This effect is depicted
in Eq. (4) by the term gj Xj, where gj is a constant, Xj is the level of the
agent and the subscript j denotes the jth agent. In Eq. (4) I assume
that the effects of various clearance rate modulators are independent
from each other. Thus Eq. (3) can be modied to include clearance
rate modulators such that
n o
C_ R
E

k1

n
o
n o
n o
C_ Rm g i X i C_ R g j X j C_ R :

n
o


C_ Rm g j X j 1ek1 1gi X i E
1 gi X i

Eq. (5) depicts a general model of clearance rate regulation with

internal and external controls. It is noteworthy that depending on
the value of parameters gi and k1, the model in Eq. (5) allows for
both strongly curvilinear curves or near-linear relationships (Fig. 1).
The main topic of the present analysis, however, is about testing
whether ISFM is to be preferred to HFM. Therefore we return to Eq. (3),
which is a simpler case than Eq. (5). Once integrated, Eq. (3) reads
n o n
o 

k E
C_ R C_ Rm 1e 1 :

Considering that food ingestion rate per gill unit surface ( fp_ X g,
J cm 2 d 1) is the product of clearance rate and food concentration,
we have
n
o 

k E
fp_ X g X  C_ Rm 1e 1

where X* (J L 1) is energy content of food, which was computed as

X* = cf * X, where cf * = 2.388 J mg 1 Chl a (see Frchette, 2012).
Finally, following standard DEB theory, food ingestion per individual
is given by
n
o

k E
2=3
p_ X X  C_ Rm 1e 1 V :

Clearance rate

The relationship between p_ X and X depicted in Eq. (8) is similar

to that of a rectangular hyperbola, which is a handy functional
form (Kooijman, 2010), while the half-saturation constant of the
MichaelisMenten equation has more intuitive meaning than the

g 1=1; k 1=1

g 1=0.25; k 1=0.5

5
4
3
2
1
0

n
o
p_ X X  C_ Rm

E 
2=3
V
E K  E 

where [EK] (J cm 3) is the half-saturation constant of the response of

clearance rate to energy density. Because energy density varies
through time (Duarte et al., 2012), Eq. (9) accommodates transient
effects such as seasonality and phenotypic exibility of the feeding
functional response.
3. Methods
3.1. DEB models

Upon integration, one gets

n o
C_ R

exponential term in Eq. (8). Therefore, in the simulations below I rewrite Eq. (8) as

10

20

Seston concentration
Fig. 1. Effect of parameters k1 and gi (i = 1; seston made of a single component only) on
the shape of the clearance rate response to seston concentration, as computed using
Eq. (5). Hypothetical data and parameter values. Maximum hypothetical clearance
rate ({Rm}=10 L cm2 d1) and hypothetical energy density ([E]=1 J cm3) are xed.

The DEB models used in the present study are mussel models,
although the studies examined may have been done on different
species. In addition, the simulations below are based on hypothetical
situations depicting actual experiments. Therefore the goal is to test
patterns, as opposed to nding specic end-point values observed
in actual eld situations. Except for food ingestion dynamics with
ISFM, parameter values and the core characteristics of the models
are the same as in Rosland et al. (2009), as modied by Frchette
(2012) to include autonomous spawning dynamics. I assume that
variability in CBF has negligible effect on total costs. This assumption
is reasonable as the energetic costs of water pumping amount to a
small fraction (about 1.5%) of total anaerobic metabolism of mussels
(Jrgensen et al., 1986a,1986b). Environmental conditions and initial
conditions vary among model runs, depending on the actual situation
modelled. A list of symbols used in this paper is presented in
Appendix A. The standard DEB model with HFM is described in
Appendix B.
Except where stated otherwise, the general settings of the study
are the following. I assume that Chl a concentration follows a simple
annual cycle. Chl a concentration is X = 4 g L 1 from April 1st to
November 1st. Between December 1st and March 1st, it is equal to
0.4 g L 1. Transition between either state is by linear interpolation.
Suspended sediments are not included in the simulations. I assume
that in the three sets of studies modelled here (see below), seston
was generally too low to signicantly affect clearance rate. Temperature
follows the low-frequency component used in Frchette (2012). Nominal date of larval settlement is July 1st. I assume that mussels settle at
initial size 0 = 0.04 cm (discussed in Frchette, 2012). Assuming a
shape factor = 0.231 (Rosland et al., 2009), initial structural mass is
V0 = 7.68 107 cm3. I assume that the ratio of initial energy reserves
to initial structural mass is E0/V0 = 0.373 J cm3 (R. Rosland, Dept. of
Biology, University of Bergen, unpubl.). Thus, initial reserves are estimated as E0 = 0.373 cf1dVV0 = 1.12 10 3 J, where cf is the energy to
wet mass conversion factor and dV is density of structural volume.
Assuming an 80% water content of soft tissue (Rosland et al., 2009),
cf =5 5.1 10 5 cm3 J1. Following van der Veer et al. (2001), I
assume that dV = 1 g cm3. The initial value of the reproductive compartment is set to ER,0 = 0. The half-saturation constant of HFM is
XK = 1 g Chl a L 1. Initial values, simulation length, food concentration
and temperature patterns are the same as above for both models except
where specied in the experiments modelled below.
The ISFM required some calibration in order to provide realistic
estimates of clearance rate and mussel size. To do this I rst ran the
standard DEB model with HFM for 720 d with the initial values, XK
and environmental specications as above. This yielded a wet tissue
mass estimate mw = 2.52 g. The DEB model with ISFM was then
calibrated manually against mw = 2.52 g for three combinations of
{Rm} and [EK]. I chose [EK] = {500, 750, 1000} J cm 3, which led to
{Rm} = {31.83, 35.65, 39.47} L cm 2 d 1. The choice of parameter
pairs had little effect on clearance rate patterns. Here I present results

M. Frchette / Journal of Sea Research 73 (2012) 3240

only for [EK] = 500 J cm 3 and {Rm} = 31.83 L cm 2 d 1. The outputs were compared on the basis of total wet tissue mass and of clearance rate, as computed by fp_ X gX  1 .
3.2. Experiments modelled
The cases modelled relate to latent effects, the effect of environmental gradients, and to phenotypic exibility of feeding behaviour.
3.2.1. Latent effects
Chiu et al. (2007) studied the effect of ration level during larval
development on juvenile growth of Crepidula onyx. Larvae were fed
Isochrysis galbana suspensions at two cell concentrations and their
growth and clearance rate at the high concentration (see their Experiment III) were studied up to 17 days after metamorphosis. Growth of
newly metamorphosed juveniles issued from the low larval food
treatment was about two-thirds and their clearance rate about half
of those of juveniles issued from the high larval food concentration
treatment. Treatment effects were apparent for the whole duration
of the experiment. With low food level during larval life, total lipid
content of newly metamorphosed juveniles was about two-thirds
that with high food level during larval life. Phillips reported that in
two similar experiments with the mussel Mytilus galloprovincialis,
the relative lipid content with low larval food concentration had
been reduced by a similar proportion in one case (Phillips, 2002)
and to traces in the other case (Phillips, 2004).
Clearance rate was simulated for 10 days after metamorphosis.
Initial energy content of freshly metamorphosed individuals with low
larval food was only 1/20 the level of those with the high larval food
treatment. Food concentration was constant at X = 2 g Chl a L1.
3.2.2. Environmental gradients
Depth is a major source of heterogeneity in marine systems. Ren and
Schiel (2008) studied individual growth of Pacic oysters (Crassostrea
gigas) at two depths (8 m and 32 m) in an experiment that lasted
5 months. Chl a concentration was about 23 times as high at 8 m as
it was at 32 m, except near the end of the study where there was not
much difference related to depth. They then ran a standard DEB
model simulation of individual growth. They calibrated the model on
observed growth at 8 m and found correct model predictions with
XK =1.0 g Chl a L 1. With this setting, however, their model clearly
overestimated growth at 32 m. To obtain correct growth prediction, it
was necessary to increase the half-saturation constant to
X K 1:9 g Chl a L

The goal of the experiment was to assess the relative ability of

HFM and ISFM, with each of their calibrations for the high food
case, to predict growth in the low food case without extra calibration.
To achieve this, I rst generated hypothetical growth data for the high
food situation using the standard DEB model with HFM. Then I generated independent hypothetical growth data for the high food level
and the low food level (dened as half the high food regime), using
simulations based on a SFG model. Calibrated standard DEB (with
HFM) and SFG models of mussel growth have been shown to yield
similar results (Filgueira et al., 2011). The SFG model and parameter
values were the same as in Frchette (2010), except that in the
present model there was no food depletion and mussel mass was wet
tissue mass instead of dry tissue mass. I assume that water content of
wet tissue was 80% (Rosland et al., 2009). This particular SFG setup
has yielded acceptable predictions of mussel growth in the south
shore of the Lower St. Lawrence Estuary (Alunno-Bruscia et al., 2000).
This allowed me to generate hypothetical growth data exhibiting realistic depth differences, independent from DEB simulations except for
the initial calibration of the SFG model. The SFG model was run for
360 d at constant temperature (14 C) and food level was adjusted

35

until mw = 2.52 g was obtained. This yielded the appropriate seston

energy level for the high food level case, XH = 8.46 J L1. Then I ran
the SFG model at half the high food level, XL = 4.23 J L1. Soft tissue
mass was mw = 0.52 g. Finally, I ran the DEB model with HFM and
ISFM at high and low food concentrations, without recalibration at
low food concentration.
3.2.3. Flexibility of feeding behaviour
To study the ability of the two feeding modules to capture physiological exibility in feeding behaviour, I mimicked the experiments of
Labarta et al. (1997) and Babarro et al. (2000a). In both experiments,
shore mussels were transferred to a culture raft and their clearance
rates were monitored through time, using a common source of water.
In late fall clearance rates were higher for raft mussels than for shore
mussels shortly after transfer (Babarro et al., 2000a). The difference,
however, disappeared within a month after transfer. In summer there
were no signicant differences among mussel groups (Labarta et al.,
1997).
The key assumption in the following simulations is that shore
mussels had access to lower food abundance because of reduced feeding time in the intertidal and because of food depletion above natural
beds. This assumption is well supported by direct and indirect evidence
(Dolmer, 2000; Frchette and Bourget, 1985a,b; Gascoigne et al., 2005;
Kamermans, 1993; Newell, 1990; Peterson, 1982; Peterson and Black,
1991; Svane and Ompi, 1993). I assume that this resulted in lower energy reserves in shore mussels, as shown by their lower condition index
(Babarro et al., 2000b), and was reected in origin-related clearance
rates of mussels. To mimic this case, I assumed that before transfer
tidal emersion and food depletion effects were appropriately accounted
for by modelling shore mussel growth with constant immersion but half
the food level experienced on the raft. After transferring shore mussels
to the raft, both mussel groups were provided the high food regime
(X =4 g Chl a L 1 in summer and X = 0.4 g Chl a L 1 in winter).
Temperature was constant at 10 C. Mussels were transferred in summer (Labarta et al., 1997) and winter (Babarro et al., 2000b). Therefore
I modelled the cases of transfers on 1st July and 1st March, when energy
density is at its annual maximum and minimum levels, respectively
(Duarte et al., 2012). Because of size differences at the time of transfer,
clearance rates are expressed in terms of surface-specic rates.
4. Results
4.1. Effect of seston concentration
The effect of high seston concentration on clearance rate is presented
in Fig. 1. All terms in the right-hand side of Eq. (5) were given arbitrary
constant values, except Xi. It is noteworthy that depending on the value
of parameters k1 and gi, the predicted relationship between clearance
rate and seston concentration Xi was either strongly curvilinear or nearly
linear (Fig. 1).
4.2. Latent effects
Chiu et al. (2007) assessed the presence of latent effects by comparing
plots of clearance rate as a function of shell length. Here I use the same
approach in Fig. 2. With HFM, size-specic clearance rate of mussels
from the high-food larval environment was the same than for mussels
from the low-food larval regime. There was a slight lag in size in the
low-food case. In contrast, with ISFM, size-specic clearance rate was
lower in mussels from the low-food larval environment for the rst few
days. Lag effects were stronger with ISFM than with HFM, but
again tended to vanish as mussels grew older. These results were
obtained with parameter values {Rm}=31.83 L cm2 d1 and [EK]=
500 J cm3. Quite similar patterns were found with the other sets of parameters tried, although latent effects appeared stronger (not shown).
Thus the present analysis is conservative.

36

M. Frchette / Journal of Sea Research 73 (2012) 3240

A
40

0.0100

30
HFM, [E0]=0.00112
ISFM, [E0]=0.00112

0.0010

HFM, [E0]=0.000056
ISFM, [E0]=0.000056

0.0001
0.01

0.10

1.00

Shell length (cm)

Fig. 2. Latent effects of low energy reserves on clearance rate and growth of newly
metamorphosed mussels. Diamonds and triangles, HFM, high and low initial energy
reserves, respectively. Circles and crosses, ISFM, high and low initial energy reserves,
respectively.

The effect of environmental gradients is shown in Fig. 3. The large

square and the large triangle indicate predicted wet mass of soft
tissues (mw, g) with the high and low food regimes, respectively.
These hypothetical data were obtained with the SFG model. Growth
was intermittent, as expected in a seasonal environment. In the high
food regime both feeding modules yielded similar results. There was a
sharp drop in individual mass during the last month because of spawning.
In the low food regime and still with XK =1.0 g L1, HFM yielded nal
wet tissue mass mw =2.23 g, much higher than the SFG-expected values
(mw =0.52 g). Running the model with the half-saturation constant
recalibrated at XK =2.9 g L1 yielded the expected value. In contrast,
ISFM predicted mw =0.18 g. This was obtained without recalibration of
the model for the low food regime.
4.4. Flexibility of feeding behaviour
The effect of physiological exibility is shown in Fig. 4, where
clearance rate is plotted as a function of time since transfer of the
shore mussels to the raft. With HFM (Fig. 4A), clearance rate
decreased during the rst month of the March experiment, as food
level increased, and was constant thereafter. Clearance rate was constant
at {R}=11.2 L cm2 d1 throughout the July experiment. After day 90,
however, the fall decrease in food level caused an increase in clearance
rate. There was no difference between shore and raft mussels during
either experiment.
With ISFM (Fig. 4B), clearance rate exhibited much more pattern. In
the March experiment, clearance rate of both mussel groups decreased

Soft tissue mass (g)

HFM;X H ;X K =1.0
HFM;X H ;X K =1.0

2.5

HFM;X H ;X K =2.9
ISFM;X H

2.0

ISFM;X L

1.5
1.0
0.5
0.0

100

200

20
10
0

15

10

March-shore
March-raft
July-shore

4.3. Environmental gradients

3.0

Clearance rate (L cm-2 d-1)

Clearance rate (L d-1)

0.1000

300

400

500

600

700

Time (d)
Fig. 3. Effect of spatial heterogeneity in food concentration on growth. Large square and
large triangle, hypothetical growth as predicted by the SFG model at high and low food concentration, respectively. Grey squares, HFM at high food concentration, XK =1 g Chl a L1.
Grey dots, HFM at low food concentration, XK =1 g Chl a L1. Grey triangles, HFM at low
food concentration, recalibrated with XK =2.9 g Chl a L1. Empty squares, ISFM at high
food concentration. Empty triangles, ISFM at low food concentration.

July-raft

20

40

60

80

100

Time since transfer (d)

Fig. 4. Effect of phenotypic exibility on clearance rate. Grey symbols, March transfer.
Empty symbols, July transfer. Triangles, shore mussels. Squares, raft mussels. A) HFM,
B) ISFM.

slightly early in the beginning of the experiment and increased afterwards to the level observed in the July experiment. Clearance rate of
shore mussels began at {R} = 0.9 L cm2 d1, lower than raft mussels
({R} = 4.6 L cm2 d 1). There was a period after day 14 where clearance rate of shore mussels was slightly higher than that of raft mussels.
In the July experiment, clearance rate of raft mussels was constant at
{R}= 12.5 L cm2 d1, except near the end of the study where there
was a slight negative trend. Clearance rate of shore mussels started at
{R}= 9.3 L cm2 d 1, again lower than raft mussels. Clearance rate
of shore mussels increased rapidly and was nearly equal to that of raft
mussels by day 15. The key point here is that with ISFM, initial clearance
rate was lower in March than in July. In addition, shore mussels
exhibited lower clearance rate than raft mussels early in the experiment. Finally, the effect of mussel origin appeared slightly stronger in
March than in July.

4.5. Functional response

Feeding regulation was examined from two angles, the functional
response of food intake (Fig. 5A) and clearance rate regulation
(Fig. 5B). Only data from the March experiment are shown (from day
0 to day 300 after transfer), but it is clear from Fig. 4 that the general
patterns were not qualitatively affected by season, although their
amplitude was.
Food intake with HFM followed the pattern that had been specied
in the model, a Holling's Type II model (Fig. 5A). There was no effect of
the phase of the annual cycle in food availability. Clearance rate
followed the same pattern as in Fig. 4A, HFM, from day 0 to day 30
after transfer. This was a period where Chl a increased linearly with
time. Food intake with ISFM remained low during the rst week of
the series, but subsequently increased to a maximum. This was followed
by a regular decrease as food availability decreased (Fig. 5A). Clearance
rate followed a similar pattern. There was an early latency period,
followed by sigmoid increase to maximum rates at high food concentration. This phase was followed by a very slow decrease in clearance rate,
concomitant with food concentration decrease. Finally, clearance rate
continued to decrease after the annual minimum in Chl a had been
reached.

M. Frchette / Journal of Sea Research 73 (2012) 3240

50

Intake rate

40
30
ISFM

20

HFM

10
0

Clearance rate

15

10

Chlorophyll a concentration
Fig. 5. Functional responses of shore mussels, March transfer, from day of transfer to
300 days after transfer. Squares, HFM. Triangles, ISFM. A) Food intake (J cm2 d1),
B) clearance rate (L cm2 d1).

5. Discussion
Dynamic Energy Budget is a most elegant construct that is
particularly well adapted to the study of bivalve growth. Consequently,
it has been applied to a large number of situations, such as individual
growth prediction and analysis of carrying capacity in bivalve culture
(see for instance Journal of Sea Research, Vol. 66, issue 4). One difculty
encountered in these studies, however, has been the need to perform
ad hoc calibration of models when moving from one environmental
situation to another. This has led to considerable effort to improve
estimates of model parameters (Saraiva et al., 2011b). Alternatively,
one may choose to adjust the feeding module of the model by varying
XK as requested by the particular situations modelled. This has provided
satisfactory growth prediction of oyster growth in different environmental settings (Alunno-Bruscia et al., 2011; Ren and Schiel, 2008).
Case-by-case recalibration of the model, however, reects the lack of
exibility of HFM. Likewise, HFM simulations in this study were unable
to reect well supported patterns in bivalve growth dynamics. Latent
effects, the effect of spatial heterogeneity in trophic conditions and
physiological exibility of the feeding response would all have required
either an ad hoc tuning of XK or more signicant modication of the
model in order to materialise in simulation results. On the other hand,
ISFM amounts to a minor change in the model. The consequences of
this single change were dramatic. ISFM was able to depict latent effects
correctly, as they resulted in decreased clearance rate and in reduced
growth of young juveniles (Chiu et al., 2007). However, modelled latent
effects in clearance rate were short-lived (Fig. 2), as opposed to many
situations (e.g., Chiu et al., 2007; Phillips, 2002, 2004; Thiyagarajan et
al., 2007), suggesting that latent effects materialise through reduced
energy uptake in the short term, but that long term effects are probably
mediated through negative consequences of low initial food intake on
normal ontogenetic processes, as suggested by Pechenik (2006).
ISFM also made satisfactory prediction of the effect of reduced food
availability on growth, without ad hoc recalibration (Fig. 3). This is in
sharp contrast with HFM, which requires new calibrations for each
particular environmental situation (Alunno-Bruscia et al., 2011; Ren
and Schiel, 2008; Rosland et al., 2009). With ISFM, predicted growth
with low food availability was somewhat lower than expected. This

37

may have been caused by inaccurate growth prediction by the particular SFG model I used or by inappropriate parameter values for [EK] and
{Rm}.
ISFM was successful at accounting for physiological exibility of
clearance rate (Fig. 4). Mussels originating from the shore had lower
initial clearance rate than mussels originating from rafts after being
transferred to a raft environment. This effect was slightly stronger in
March than in July. This difference correlated quite well with actual
annual patterns in energy reserves, which are higher in summer than
in fall and winter (Duarte et al., 2012). A further point was that initial
post-transfer clearance rate varied seasonally, irrespective of mussel
origin and independently of temperature as the simulations were
made at constant temperature. This supports physiological regulation
of seasonality in clearance rate (Cranford and Hill, 1999). The present
ISFM does not include phenotypic exibility in A. In a related analysis
of their transplantation experiment, Babarro et al. (2003) reported
evidence of phenotypic exibility in assimilation efciency, in addition
to that in clearance rate. It is interesting to note that treatment effects
were longer lived in assimilation efciency than in clearance rate, as
expected in hierarchical systems subjected to temporal change (Allen
and Starr, 1982; Slobodkin and Rapoport, 1974).
The three aspects tested through simulations (latent effects, environmental heterogeneity and phenotypic exibility of clearance rate)
support the ISFM formulation. It should be added that Eq. (5) correctly predicts the shape of the effect of high particulate load (Fig. 1).
Including suspended sediments in the simulations would have led to
different feeding responses. I did not examine this situation, but it is
clear in Fig. 1 that the model depicted by Eq. (5) is exible enough to
handle the diversity of responses (strongly curvilinear vs near-linear)
reported (e.g., Barill et al., 2006; Widdows et al., 1979). Based on the
four lines of evidence above, I conclude that ISFM, in contrast with
HFM, has the exibility required to deal with usual conditions encountered in studies of bivalve growth dynamics.
Most parameter values used in this study were from van der Veer
et al. (2006) and from Rosland et al. (2009). Saraiva et al. (2011b)
pointed out that these sets of parameters resulted in overestimation
of the conversion efciency of reserves to structure and presented
new sets of parameter values to deal with the issue. Here I chose to
use the van der Veer et al. and Rosland et al. parameter values because
Saraiva et al. (2011b) based their assessment on DEB simulations that
used HFM. Comparing Eqs. (1) and (5), however, suggests that independent of parameter values, results from HFM-based DEB models should
be considered with caution because of the inadequacy of their feeding
module. Clearly parameter values should be re-examined in light of
results found with ISFM.
Implications of Eq. (5) for experimental studies of clearance rate
regulation are critical, as their interpretation requires information on animal internal state, in addition to the usual environmental measurements.
Filgueira et al. (2008) studied the relationship between condition index,
computed as percent of dry mass of soft tissues on total mass, and clearance rate of suspension-grown mussels. They focussed on the allometric
exponent of clearance rate and reported non-signicant effect of condition index. In the present context, however, the value of interest is the
proportionality parameter of the relationship, not its exponent. It is interesting to note that in their study the proportionality parameter was
higher in mussels with higher condition index (Exp. B in Fig. 1, Filgueira
et al., 2008). This is quite consistent with internal-state regulation of
clearance rate. Condition index may be a rather crude proxy for energy
reserves. Other estimators of energy reserves may be more useful in
this respect. Total adenylate energy charge, for instance, may be more appropriate, as total adenine nucleotides (Fig. 3 in Blanco et al., 2006) and
modelled energy reserves (Fig. 11 in Duarte et al., 2012) vary in phase
on an annual basis.
Early analyses of CR regulation have implied physiological regulation, with minor attention being paid to the effect of hydrodynamic constraints in the actual feeding process (Bayne and Widdows, 1978; Bayne

38

M. Frchette / Journal of Sea Research 73 (2012) 3240

et al., 1993; Thompson, 1984). In contrast, analyses of the hydrodynamics of suspension feeding have led to the concept of the mussel pump,
whereby variability of CR was ascribed to the effect of suboptimal
environmental conditions on otherwise presumably constant pumping
processes (Jrgensen et al., 1986b, 1988, 1990; Riisgrd and Larsen,
2007). It is noteworthy that Eq. (5) provides a framework for accounting
for each type of effect simultaneously, again pointing to its generality
and exibility.
With HFM, the functional response of food intake followed usual
Holling's Type II dynamics, as specied by the model (Fig. 5A). With
ISFM, however, the functional response was more complex. There
was a Type III response while Chl a was increasing and a Type I
model while it was decreasing. Thus, the feeding modules predicted
different functional responses. Furthermore, ISFM predicted variability in clearance rate response, alternating between (extreme) Type
II and Type III models (Fig. 5B). The pattern exhibited by the clearance
rateChl a relationship in Fig. 5B is quite similar to that reported by

Strohmeier et al. (2009) in low-seston situations. The similarity is

especially striking if one considers only the descending phase of the
Chl a signal in Fig. 5B. Strohmeier et al. sampled from August to April,
during a decreasing trend in Chl a concentration, except for two strong
peaks near the end of the time-series. Although the Strohmeier et al.
data were not corrected for the annual cycle in water temperature,
whereas the model runs for Fig. 5 were made at constant temperature,
I conclude that the internal-state feeding module is consistent with
these data, and thus it has the required exibility to model correctly
latent effects, environmental heterogeneity and phenotypic exibility
in clearance rate.

Acknowledgments
I am grateful to S.E. Richman and J. Urquiza (Universit Laval) for
critical comments on a previous version of this manuscript.

Appendix A
List of symbols in this paper.
Denition

Symbol

Units

Value

Source

Assimilation efciency
Body energy reserves
Structural volume
Reproductive compartment
Fraction of energy ux from reserves used for growth and maintenance
Energy ingestion rate
Maximum ingestion rate per surface area
Feeding function
Chlorophyll a concentration
Half-saturation coefcient
Assimilation rate
Maximum assimilation rate per surface area (= A fp_ Xm g)
Rate of energy utilisation
Maintenance cost
Volume-specic maintenance costs
Volume-specic costs of structural growth
Reserve energy density
Maximum energy density of the storage component
Structural volume at sexual maturity
Spawning compartment
Amount of gametes maturing on a maturation event
Efciency of gamete transfer from reproductive to spawning compartments
Total esh content
Condition index
Critical esh content
Spawning rate constant
Energy to wet tissue mass conversion factor
Total individual live mass
Shell length
Modelled wet tissue mass
Threshold wet tissue mass
Density of structural volume
Shape parameter
Shell mass
Initial shell length
Initial structural volume
Initial energy reserves
Initial reproductive compartment
Constant
Individual clearance rate
Maximum individual clearance rate
Individual clearance rate per unit gill surface
Maximum individual clearance rate per unit gill surface
Ciliary beat frequency
Maximum ciliary beat frequency
Constant for ith particle type
Concentration of ith particle type
Constant for the jth external factor
Metric of jth external factor
Chl a mass to phytoplankton energy conversion factor

A
E
V
ER

p_ X
fp_ Xm g
f
X
XK
p_ A
fp_ Am g
p_ C
p_ M
p_ M 
[EG]
[E]
[Em]
VP
ES
G
R
TC
CI
TC*
kS
cf
mtot

mw
m
dV
V
ms
0
V0
E0
ER,0
k1
R
Rm
{R}
{Rm}
f_
f_ m

J
cm3
J

J d1
J cm2 d1

g L1
g L1
J d1
J cm2 d1
J d1
J d1
J cm3 d1
J cm3
J cm3
J cm3
cm3
J
J

d1
cm3 J1
g
cm
g
g
g cm3

g
cm
cm3
J
J
cm3 J1
L d1
L d1
L cm2 d1
L cm2 d1
d1
d1
L g1
g L1
Depend on the nature of factor
Idem
J g1

0.75

Rosland et al. (2009)

0.45

Rosland et al. (2009)

273

Rosland et al. (2009)

204.75

Rosland et al. (2009)

27.8
1900

Rosland et al. (2009)

van der Veer et al. (2006)

2190
0.06

van der Veer et al. (2006)

van der Veer et al. (2006)

0.95

Pecquerie et al. (2009)

0.45
0.461
5 5.1 105

Bergeron (1990)
Arbitrary value

1
0.231

van der Veer et al. (2001)

Rosland et al. (2009)

0.04
7.68 107

Present study
Present study

Present study

2.388

Frchette (2012)

gi
Xi
gj
Xj
cf *

M. Frchette / Journal of Sea Research 73 (2012) 3240

Appendix B
The standard DEB model with HFM estimates growth as follows
(details in Rosland et al., 2009). Energy ingestion rate is given by
2=3
p_ X fp_ Xm gf V

B1

where fp_ Xm g (J cm 2 d 1) is maximum ingestion rate per surface

area, f is a dimensionless function varying from 0 to 1 which depicts
a type II Holling functional response to food concentration as estimated
by chlorophyll a concentration in the seston. Thus we have
f

X
:
X XK
Assimilation rate p_ A (J d 1) is given by

2=3
p_ A fp_ Am gf V ;

B2

where fp_ Am g is maximum surface-area-specic assimilation rate

(J cm 2 d 1). The ratio fp_ Am g=fp_ Xm g is an estimate of A, the assimilation efciency of ingested energy. Reserves vary according to
dE=dt p_ A p_ C ;

B3

where p_ C (J d1) is the rate of energy utilisation of the whole individual

and includes energy xation and energy dissipation. Maintenance cost
(J d1) is proportional to structural volume and is written p_ M
p_ M V, where p_ M  is volume-specic maintenance costs (J cm 3 d1).
It follows that structural volume changes according to
dV=dt p_ C p_ M =EG ;

B4

where [EG] is volume-specic cost of structural growth (J cm3) and is

the fraction of energy ux from reserves used for growth and maintenance. Total energy use may be written as
p_ C

!
E 
EG fp_ Am gV 2=3
p_ M V ;
E G  E 
E m 

B5

where [E] ranges from 0 to [Em] (J cm3), [Em] being the maximum
energy density of the reserve compartment. Energy allocated to the
reproductive compartment is 1 p_ C . Its dynamics are described by



1
minV P ; V p_ M 
dER =dt 1 p_ C

B6

where VP (cm 3) is size at sexual maturity. Structural growth is given priority over reproduction. Thus, if reserves are too low to support maintenance costs (i.e., p_ C bp_ M ), then
dER =dt p_ C p_ M :

B7

With ER = 0, further maintenance costs are paid for by structural

volume. The model forbids ER b 0.
To model spawning dynamics, I assume that gametes transfer
instantaneously from compartment ER to the spawning compartment
ES (J), once tissue mass content of individuals exceeds a critical value
TC* (see below). The state of the reproductive compartment is therefore given by



1
minV P ; V p_ M G
dER =dt 1 p_ C

B8

where G = RER (J) and is always set to 0, except if tissue mass exceeds a
critical value relative to total mass (see below). Eq. (B8) is subject to the
rule that ERt+1 = ERt RERt. The process occurs with efciency R =
0.95. The remaining fraction of gametes, 1 R, is left in compartment

39

ER. Once the spawning compartment has been lled, its dynamics are
governed by spawning and maintenance costs of gametes. Both components are proportional to ES, such that
dES =dt kS ES cf p_ M ES G

B9

where the constant kS = 0.461 d 1.

Based on Frchette (2012), spawning is triggered when the ratio between tissue mass and total mass (TC; TC= CI/ 100), as estimated using
condition index (CI; percent wet tissue mass of total live mass) exceeds
a critical value TC* = 0.45. The relationship between total live mass
(mtot, g) and shell length (, cm) is mtot 0:09582:75 . The relationship
between shell length and shell mass ms (g) is ms 0:06292:6170 . Therefore, spawning occurred whenever modelled wet tissue mass (mw, g)
exceeded the threshold value m (g), where mw =dVV+cf(E+ER +ES),
dV (g cm3) is density of structural volume and m TC  0:09582:75 .
The relationship between structural volume and shell length is obtained
1=3
using the following relationship, V V , where V is the shape param1=3
eter. Since V =0.231, 4:33V . If V increased through time, changed according to the relationship above. Otherwise, remained constant.
Therefore,

 dV 1 0:3333 0:6667
3
d 

V
if V > 0:012
 dt t 0:231
;
dt
 0 if V0:0123

B10

where t is the integration time step.

References
Allen, T.F.H., Starr, T.B., 1982. Hierarchy: Perspectives for Ecological Complexity.
University of Chicago Press, Chicago.
Alunno-Bruscia, M., Petraitis, P.S., Bourget, E., Frchette, M., 2000. Body sizedensity
relationships for Mytilus edulis in an experimental food-regulated situation. Oikos
90, 2842.
Alunno-Bruscia, M., Bourls, Y., Maurer, D., Robert, S., Mazuri, J., Gangnery, A.,
Goulletquer, P., Pouvreau, S., 2011. A single bio-energetics growth and reproduction
model for the oyster Crassostrea gigas in six Atlantic ecosystems. Journal of Sea
Research 66, 340348.
Babarro, J.M.F., Fernndez-Reiriz, M.J., Labarta, U., 2000a. Feeding behavior of seed
mussel Mytilus galloprovincialis: environmental parameters and seed origin.
Journal of Shellsh Research 19, 195201.
Babarro, J.M.F., Fernndez-Reiriz, M.J., Labarta, U., 2000b. Growth of seed mussel
(Mytilus galloprovincialis Lmk): effects of environmental parameters and seed
origin. Journal of Shellsh Research 19, 187193.
Babarro, J.M.F., Fernndez-Reiriz, M.J., Labarta, U., 2003. In situ absorption efciency
processes for the cultured mussel Mytilus galloprovincialis in Ra de Arousa
(north-west Spain). Journal of the Marine Biological Association of the United
Kingdom 83, 10591064.
Barill, L., Cognie, B., Beninger, P., Decottignies, P., Rinc, Y., 2006. Feeding responses of
the gastropod Crepidula fornicata to changes in seston concentration. Marine Ecology
Progress Series 322, 169178.
Bayne, B.L., Widdows, J., 1978. The physiological ecology of two populations of Mytilus
edulis L. Oecologia (Berlin) 37, 137162.
Bayne, B.L., Iglesias, J.I.P., Hawkins, A.J.S., Navarro, E., Hral, M., Deslous-Paoli, J.-M.,
1993. Feeding behaviour of the mussel, Mytilus edulis: responses to variations in
quantity and organic content of the seston. Journal of the Marine Biological Association
of the United Kingdom 73, 813829.
Ben Kheder, R., Qur, C., Moal, J., Robert, R., 2010. Effect of nutrition on Crassostrea
gigas larval development and the evolution of physiological indices part B: effects
of temporary food deprivation. Aquaculture 308, 174182.
Bergeron, P., 1990. volution temporelle du recrutement de Mytilus edulis Carleton et
effet de la prolifration du naissain sur le rendement commercial des boudins.
Ministre de l'Agriculture, des Pcheries et de l'Alimentation du Qubec. Direction
de la recherche scientique et technique. Cahier d'information, 127. vii + 49.
Blanco, S.L., Surez, M.P., San Juan, F., 2006. Seasonal changes of nucleotides in mussel
(Mytilus galloprovincialis) mantle tissue. Comparative Biochemistry and Physiology. B
143, 384390.
Chiu, J.M.Y., Ng, T.Y.T., Wang, W.X., Thiyagarajan, V., Qian, P.Y., 2007. Latent effects of
larval food limitation on ltration rate, carbon assimilation and growth in juvenile
gastropod Crepidula onyx. Marine Ecology Progress Series 343, 173182.
Chiu, J.M.Y., Wang, H., Thiyagarajan, V., Qian, P.Y., 2008. Differential timing of larval
starvation effects on ltration rate and growth in juvenile Crepidual onyx. Marine
Biology 154, 9198.
Cranford, P.J., Hargrave, B.T., 1994. In situ time-series measurement of ingestion and
absorption rates of suspension-feeding bivalves: Placopecten magellanicus. Limnology
and Oceanography 39, 730738.

40

M. Frchette / Journal of Sea Research 73 (2012) 3240

Cranford, P.J., Hill, P.S., 1999. Seasonal variation in food utilization by the suspensionfeeding bivalve molluscs Mytilus edulis and Placopecten magellanicus. Marine Ecology
Progress Series 190, 223239.
Dolmer, P., 2000. Algal concentration proles above mussel beds. Journal of Sea Research
43, 113119.
Duarte, P., Fernndez-Reiriz, M.J., Labarta, U., 2012. Modelling mussel growth in
ecosystems with low suspended matter loads using a Dynamic Energy Budget
approach. Journal of Sea Research 67, 4457.
Filgueira, R., Labarta, U., Fernndez-Reiriz, M.J., 2008. Effect of condition index on
allometric relationships of clearance rate in Mytilus galloprovincialis Lamarck,
1819. Revista de Biologa Marina y Oceanografa 43, 391398.
Filgueira, R., Rosland, R., Grant, J., 2011. A comparison of scope for growth (SFG) and
dynamic energy budget (DEB) models applied to the blue mussel (Mytilus edulis).
Journal of Sea Research 66, 403410.
Frchette, M., 2010. Hierarchical structure of bivalve culture systems and optimal
stocking density. Aquaculture International 18, 99114.
Frchette, M., 2012. Self-thinning, biodeposit production, and organic matter input to
the bottom in mussel suspension culture. Journal of Sea Research 67, 1020.
Frchette, M., Bourget, E., 1985a. Energy ow between the pelagic and benthic zones:
factors controlling particulate organic matter available to an intertidal mussel
bed. Canadian Journal of Fisheries and Aquatic Sciences 42, 11581165.
Frchette, M., Bourget, E., 1985b. Food-limited growth of Mytilus edulis in relation to
the benthic boundary layer. Canadian Journal of Fisheries and Aquatic Sciences
42, 11661170.
Gascoigne, J.C., Beadman, H.A., Saurel, C., Kaiser, M.J., 2005. Density dependence, spatial
scale and patterning in sessile biota. Oecologia (Berlin) 145, 371381.
Gueron, S., Levit-Gurevich, K., 1999. Energetic considerations of ciliary beating and the
advantage of metachronal coordination. Proceedings of the National Academy of
Sciences of the United States of America 96, 1224012245.
Jrgensen, C.B., 1990. Bivalve Filter Feeding: Hydrodynamics, Bioenergetics, Physiology
and Ecology. Olsen & Olsen, Fredensborg, Denmark. vi + 140 pp.
Jrgensen, C.B., Mhlenberg, F., Sten-Knudsen, O., 1986a. Nature of relation between
ventilation and oxygen consumption in lter feeders. Marine Ecology Progress
Series 29, 7388.
Jrgensen, C.B., Famme, P., Kristensen, S.H., Larsen, P.S., Mhlenberg, F., Riisgrd, H.U.,
1986b. The bivalve pump. Marine Ecology Progress Series 34, 6977.
Jrgensen, C.B., Larsen, P.S., Mhlenberg, F., Riisgrd, H.U., 1988. The mussel pump:
properties and modelling. Marine Ecology Progress Series 45, 205216.
Jrgensen, C.B., Larsen, P.S., Riisgrd, H.U., 1990. Effects of temperature on the mussel
pump. Marine Ecology Progress Series 64, 8997.
Kamermans, P., 1993. Food limitation in cockles (Cerastoderma edule (L.)): inuences
of location on tidal ats and of nearby presence of mussel beds. Netherlands Journal
of Sea Research 31, 7181.
Kooijman, S.A.L.M., 2010. Dynamic Energy Budget Theory for Metabolic Organisation.
Cambridge University Press. 514 pp.
Labarta, U., Fernndez-Reiriz, M.J., Babarro, J.M.F., 1997. Differences in physiological
energetics between intertidal and raft cultivated mussels Mytilus galloprovincialis.
Marine Ecology Progress Series 152, 167173.
Lieb, T., Wijkstrom Frei, C., Frohock, J.I., Bookman, R.J., 2002. Prolonged increase in
ciliary beat frequency after short-term purinergic stimulation in human airway
epithelial cells. The Journal of Physiology 538 (2), 633646.
Mhlenberg, F., 1999. Muslinger som vandrensere. In: Lomstein, B.A. (Ed.), Havmiljet
ved rtusindskiftet. Olsen & Olsen, Fredensborg, pp. 3046.
Newell, C.R., 1990. The effects of mussel (Mytilus edulis, Linnaeus, 1758) position in
seeded bottom patches on growth on subtidal lease sites in Maine. Journal of Shellsh
Research 9, 113118.
Noguchi, M., Sawada, T., Akazawa, T., 2001. ATP-regenerating system in the cilia of
Paramecium caudatum. Journal of Experimental Biology 204, 10631071.
Okumus, I., Stirling, H.P., 1994. Physiological energetics of cultivated mussels (Mytilus
edulis) populations in two Scottish west coast sea lochs. Marine Biology 119, 125131.
Pechenik, J.A., 2006. Larval experience and latent effectsmetamorphosis is not a new
beginning. Integrative and Comparative Biology 46, 323333.
Pechenik, J.A., Jarrett, J.N., Rooney, J., 2002. Relationships between larval nutritional
experience, larval growth rates, juvenile growth rates, and juvenile feeding rates
in the prosobranch gastropod Crepidula fornicata. Journal of Experimental Marine
Biology and Ecology 280, 6378.
Pecquerie, L., Petitgas, P., Kooijman, S.A.L.M., 2009. Modeling sh growth and reproduction
in the context of the Dynamic Energy Budget theory to predict environmental impact
of anchovy spawning duration. Journal of Sea Research 62, 93105.

Peterson, C.H., 1982. The importance of predation and intra- and interspecic competition
in the population biology of two infaunal suspension-feeding bivalves, Protothaca
staminea and Chione undatella. Ecological Monographs 52, 437475.
Peterson, C.H., Black, R., 1991. Preliminary evidence for progressive sestonic food
depletion in incoming tide over a broad tidal sand at. Estuarine, Coastal and
Shelf Science 32, 405413.
Phillips, N.E., 2002. Effects of nutrition-mediated larval condition on juvenile performance in a marine mussel. Ecology 83, 25622574.
Phillips, N.E., 2004. Variable timing of larval food has consequences for early juvenile
performance in a marine mussel. Ecology 85, 23412346.
Podolsky, R.D., 1994. Temperature and water viscosity: physiological versus mechanical
effects on suspension feeding. Science 265, 100103.
Podolsky, R.D., Moran, A.L., 2006. Integrating function across marine life cycles. Integrative
and Comparative Biology 46, 577586.
Real, L.A., 1977. The kinetics of functional response. American Naturalist 111, 289300.
Ren, J.S., Schiel, D.R., 2008. A dynamic energy budget model: parameterisation and
application to the Pacic oyster Crassostrea gigas in New Zealand waters. Journal
of Experimental Marine Biology and Ecology 361, 4248.
Riisgrd, H.U., 2001. The stony road to reliable ltration rate measurements in bivalves:
a reply. Marine Ecology Progress Series 215, 307310.
Riisgrd, H.U., Larsen, P.S., 2007. Viscosity of seawater controls beat frequency of
water-pumping cilia and ltration rate of mussels Mytilus edulis. Marine Ecology
Progress Series 343, 141150.
Riisgrd, H.U., Egede, P.P., Barreiro Saavedra, I., 2011. Feeding behaviour of the mussel,
Mytilus edulis: new observations, with a minireview of current knowledge. Journal
of Marine Biology 2011, 113.
Rosland, R., Strand, ., Alunno-Bruscia, M., Bacher, C., Strohmeier, T., 2009. Applying
dynamic energy budget (DEB) theory to simulate growth and bio-energetics of
blue mussels under low seston conditions. Journal of Sea Research 62, 4961.
Salathe, M., 2007. Regulation of mammalian ciliary beating. Annual Review of Physiology
69, 401422.
Saraiva, S., van der Meer, J., Kooijman, S.A.L.M., Sousa, T., 2011a. Modelling feeding
processes in bivalves: a mechanistic approach. Ecological Modelling 222, 514523.
Saraiva, S., van der Meer, J., Kooijman, S.A.L.M., Sousa, T., 2011b. DEB parameters estimation for Mytilus edulis. Journal of Sea Research 66, 289296.
Slobodkin, L.B., Rapoport, A., 1974. An optimal strategy of evolution. The Quarterly Review of Biology 49, 181200.
Strohmeier, T., Strand, ., Cranford, P., 2009. Clearance rates of the great scallop (Pecten
maximus) and blue mussel (Mytilus edulis) at low natural seston concentrations.
Marine Biology 156, 17811795.
Svane, I., Ompi, M., 1993. Patch dynamics in beds of the blue mussel Mytilus edulis L.:
effects of site, patch size and position within a patch. Ophelia 37, 187202.
Teff, Z., Priel, Z., Gheber, L.A., 2008. The forces applied by cilia depend linearly on their
frequency due to constant geometry of the effective stroke. Biophysical Journal 94,
298305.
Thiyagarajan, V., Harder, T., Qiu, J.W., Qian, P.Y., 2003. Energy content at metamorphosis
and growth rate of the early juvenile barnacle Balanus amphitrite. Marine Biology
143, 543554.
Thiyagarajan, V., Pechenik, J.A., Gosselin, L.A., Qian, P.Y., 2007. Juvenile growth in
barnacles: combined effect of delayed metamorphosis and sub-lethal exposure of
cyprids to low-salinity stress. Marine Ecology Progress Series 344, 173184.
Thompson, R.J., 1984. The reproductive cycle and physiological ecology of the mussel
Mytilus edulis in a subarctic, non-estuarine environment. Marine Biology 79,
277288.
van der Veer, H.W., Kooijman, S.A.L.M., van der Meer, J., 2001. Intra- and interspecies
comparison of energy ow in North Atlantic atsh species by means of energy
budgets. Journal of Sea Research 45, 303320.
van der Veer, H.W., Cardoso, J.F.M.F., van der Meer, J., 2006. The estimation of DEB
parameters for various Northeast Atlantic bivalve species. Journal of Sea Research
56, 107124.
Widdows, J., Fieth, P., Worrall, C.M., 1979. Relationships between seston, available food
and feeding activity in the common mussel Mytilus edulis. Marine Biology 50, 195207.
Widdows, J., Donkin, P., Salkeld, P.N., Cleary, J.J., Lowe, D.M., Evans, S.V., Thomson, P.E.,
1984. Relative importance of environmental factors in determining physiological
differences between two populations of mussels (Mytilus edulis). Marine Ecology
Progress Series 17, 3347.
Worrall, C.M., Widdows, J., 1983. Physiological changes following transplantation of the
bivalve Scrobicularia plana between three populations. Marine Ecology Progress Series
12, 281287.