The effects of velocity and seston concentration on the exhalant siphon area, valve gape and filtration rate of the mussel Mytilus edulis

Journal of Experimental Marine Biology and EcologyŽ .262 2001 91–111

www.elsevier.nlrlocaterjembe

The effects of velocity and seston concentration onthe exhalant siphon area, valve gape and filtration

rate of the musselMytilus edulis

Carter R. Newella,), D.J. Wildishb, B.A. MacDonaldc

a Great Eastern Mussel Farms, Inc., P.O. Box 141, Tenants Hbr., ME. 04860 USAb Fisheries and Oceans Canada, Biological Station, 531 Brandy CoÕe, St. Andrews, N.B., Canada E5B2L9

c Department of Biology, UniÕersity of New Brunswick, Box 5050, St. John, N.B., Canada E2L4L5

Received 4 August 1999; received in revised form 25 April 2001; accepted 3 May 2001

Abstract

With the inhalant siphon facing into the flow, and with adequate seston levels, water velocityhas a significant negative linear effect on mussel exhalant siphon area, but no significant effect onvalve gape. Mussel filtration rates of polystyrene beads, measured by ingestion, were three timesless at 30 cm sy1 than at 10 cm sy1, and they mirrored the trends observed with the exhalantsiphon area. However, with the inhalant siphon oriented downstream of flow direction at the two

Ž y1.higher flow speeds 20 and 30 cm s , there was no significant effect of velocity on exhalantsiphon area.

There was also a significant positive effect of particle concentration on mussel exhalant siphonarea. In contrast to the effects of velocity, mussel valve gape responses to increasing particleconcentrations mirrored the responses of the exhalant siphon aperture.

The thresholds for the initiation of active pumping, opening the valve gape, extending themantle and opening the exhalant siphon are at minimum seston levels of about 4=103 particlesmly1, or particle volumes of about 1.0 mm3 ly1. Thus, the closure of valves in the relativelynon-turbid waters of Maine indicates insufficient food particle concentrations for feeding. In-creases in exhalant siphon area caused by particle volume increases from 1 to 5 mm3 ly1, orparticle concentrations from 6=103 to 4=104 particles mly1, indicate that blue mussels respondto increasing ambient food concentrations by increasing their pumping rates.

These results demonstrate that exhalant siphon area has potential for the quantitative remotesensing of feeding rate, and that valve gape is a more indirect measure of activity with respect tominimum concentrations for the initiation of feeding. Maintaining an open valve gape with partialor full closure of the exhalant siphon under high flow conditions is similar to the response of blue

) Corresponding author. Tel.:q1-207-372-6317; fax:q1-207-372-8256.Ž .E-mail address: [email protected] C.R. Newell .

0022-0981r01r$ - see front matterq2001 Elsevier Science B.V. All rights reserved.Ž .PII: S0022-0981 01 00285-4

( )C.R. Newell et al.rJ. Exp. Mar. Biol. Ecol. 262 2001 91–11192

mussels to changes in salinity, and is interpreted with respect to increased oxygen diffusion forrespiration.q2001 Elsevier Science B.V. All rights reserved.

Keywords: Current; Mytilus edulis; Particle concentration; Siphon area; Valve gape; Filtration rate

1. Introduction

There is a positive relationship between the pumping rates of mussels, valve gape,Žand the exhalant siphon area Jørgensen, 1960; Foster-Smith, 1975; Riisgard and

.Randlov, 1981; Famme et al., 1986; Jørgensen et al., 1988; Jørgensen 1990 . Thisrelationship was attributed to an increase in the distance between gill filaments,extension of the gill axes, and an increase in the bivalve pump pressure and flow rate

Žthrough the mantle cavity at maximum gape and mantle and siphon extension Jørgensen.et al., 1988 . It has also been noted that valve opening and opening of the exhalant

siphon aperture may be uncoupled, such as during the isolation response of mussels todeclining salinities, where the mussels keep the valves open while they close off the

Ž .exhalant siphon Davenport, 1979 , or a similar response to elevated levels of copperŽ . Ž .Manley, 1983 or other pollutants Kramer et al., 1989 . Closing the exhalant siphonresults in a pronounced reduction of volume of water pumped through the mantle cavityŽ .Jørgensen et al., 1986a,b , accompanied by the cessation of beating of the lateral cilia.

Velocity can directly affect the filtration rates of mussels facing into unidirectionalŽ .flows Wildish and Miyares, 1990 . Mussels decrease their filtration rates as the currents

y1 Ž y1.increase from 10 to 30 cm s . At low velocities 0–5 cm s , filtration rates areflow-limited indirectly due to the effects of reductions in ambient particle concentrationsfrom local refiltration. On one hand, increasing current speed has been shown to

Žcompensate the effects of particle depletion Wildish and Kristmanson, 1984; Frechette´.and Bourget, 1985; Frechette et al., 1989 , resulting in density-dependent growth in´

Ž .bivalve populations Wildish and Kristmanson, 1985, 1997; Newell, 1990 by replacingthe seston-depleted water with a new source of seston. On the other hand, at highervelocities, filtration rates of bivalves, with siphons facing into the flow, are directly

Žinhibited by hydrodynamic effects on the functioning of the bivalve pump Wildish et.al., 1987; Eckman et al., 1989; Wildish and Saulnier, 1992; Wildish and Saulnier, 1993

due to pressure differentials between inhalant and exhalant openings.The effects of particle concentration on filtration rate of blue mussels have been

Ž .shown to involve a unimodal response Foster-Smith, 1975; Winter, 1978; Bayne, 1993 .There is an increase in filtration rate from filtered seawater to naturally occurring sestonlevels, and a reduction at higher seston loads. However, changes in feeding behavior ofmussels at the lower end of naturally occurring seston concentrations have been poorlystudied, especially the levels at which the bivalves close their shells and drasticallyreduce their pumping activity. Since filtration rate by mussels is a function of pumpingrate, particle concentration and filtration efficiency, control over pumping rate is viewedas a major factor contributing to energy acquisition by bivalves. We investigated bothvalve gape and exhalant siphon area in controlled laboratory experiments to see howeach factor responded to separate changes in current speed or seston concentration.

( )C.R. Newell et al.rJ. Exp. Mar. Biol. Ecol. 262 2001 91–111 93

In previous field experiments, a time-lapse benthic video monitor was used tomeasure the gaping behavior of 30 mussels in situ and to compare that with directmeasurements of filtration rate of mussels feeding on the same water in experimental

Ž .chambers Newell et al., 1998 . Mussel valve gape was positively correlated withfiltration rates measured at four stages of the tide in experimental chambers, withmaximum values at high and ebb tides and minimum values at low tide when particleconcentrations fell below about 10,000 particles mly1, but no measurements were madeon exhalant siphon area. In those experiments, the direct effects of current speed aloneon mussel feeding behavior could not be determined in the field due to variability also inparticle concentrations over tidal cycles under natural conditions.

To investigate the ability of blue mussels to vary their feeding behavior in response toboth velocity and seston concentration, we studied the responses of individual mussel’svalve gape and exhalant siphon area in three separate experiments in a recirculatingflume, using diets formulated to simulate natural conditions as follows:

Ø constant current speed and varying particle concentrations,Ø constant particle concentration and varying current speed,Ø constant current speed and particle concentration and varying orientation to flow.

In a final experiment, we studied the filtration rates and exhalant siphon area ofmussels at 10 and 30 cm sy1, using labeled diets.

While time-lapse video techniques can never completely replace estimates of pump-Ž .ing and filtration rates of mussels, using direct Famme et al., 1986 or indirect

Ž .Mohlenberg and Riisgard, 1979 methods, they may provide a useful understanding of˚their long-term patterns of undisturbed feeding behavior in situ, and especially withrespect to the conditions under which:

Ž .1. they are not pumping valves closed ,2. how they react to environmental variables such as high flow which may cause a

Ž .disruption in particle capture on the gills siphons partially or fully closed ,Ž .3. when they may be respiring but not filtering valves open but siphons closed , and

4. critical natural seston concentrations for the initiation of feeding.

The use of exhalant siphon area and valve gape was investigated with the purpose ofeventually using these indirect methods to estimate bivalve pumping rates on undis-turbed bivalves in the field.

2. Materials and methods

These experiments were performed in a 206-l acrylic recirculating flume based on theŽ . Ž .design by Saunders and Hubbard 1944 , and described by Wildish and Saulnier 1993

at the St. Andrews Biological Station in St. Andrews, New Brunswick, Canada in Marchof 1996, November of 1997, and May of 2000. The flume had a 65-cm long working

Ž .section, which was 23 cm wide and 20 cm deep. MusselsMytilus edulis obtained from


a bottom aquaculture lease site, when ambient temperatures were 4–88C, were numberedwith bee dots; Velcro was attached using cyanoacrylate cement, and the mussels weresuspended in lantern nets for 5 days, after which they were brought into the laboratory.After acclimation to experimental diets, constant illumination and temperature in thelaboratory for 2 days, mussels were attached using Velcro to a Plexiglas stand in themiddle of the flume 20 cm from the end of the working section with the umbo facingdownward, the exhalant siphon facing upward, and the inhalant siphon facing into thecurrent. The height of the exhalant siphon and the two-flow probes were 7.3 cm fromthe bottom. Flow probe data, in vivo fluorescence from a Turner fluorometer, andtemperature data were logged every 30 s using a data logger during the acclimationperiods and experimental runs. Time-lapse video observations were made on exhalant

Ž .siphon area and valve gape distance between valves over periods of 30 min. Prelimi-nary experiments demonstrated that the flow probe readings were representative of thecurrent speed at the mussel location within the flume. Current profiles from the bottomto the surface and from the side walls to the center demonstrated variation of under 4%in current speed, with sample location in the test section of the flume, where the

Ž .experiments were performed Wildish, 1991 .ŽExperimental diets includedChaetocerus calcitrans and Isochrysis galbana Tahitian

.strain added to 53mm sieved mudflat silt and filtered seawater in the 1996 experiments,from Dunaliella and Isochrysis diets added to 80mm sieved mudflat silt and filteredseawater and 1mm filtered seawater in the 1997 experiments, and fromChaetocerusand Isochrysis diets, supplemented with 13-mm diameter polystyrene beads in 1mmfiltered seawater in 2000. Triplicate measurements of particle concentration, size andvolume above 3-mm diameter were made of each diet and experimental run using a

Ž .Coulter Multisizer. In addition, dry mass of suspended matter SPM was determined byfiltering 1–3 l of the flume water onto prewashed pretared 47-mm GFC filters, washingwith 3% ammonium formate to remove salts, drying for 24 h at 808C, and weighing.After the experiments, individual mussels were measured for shell length, and dry tissuemass was determined by drying for 24 h at 808C.

A video camera with a zoom lens was placed inside a beaker suspended just belowthe surface of the seawater above the individual mussels in each experiment, and thedata was recorded on a VHS time-lapse video recorder at a 1r24 time interval. Using a2-cm mark on the Plexiglas stand as a size reference, the valve gape and the exhalantsiphon area were measured on each mussel using an Optamus frame grabber and videodigitizing system. For each velocity or seston treatment, for each individual mussel, atotal of 10 frames were analyzed and triplicate measurements were made of gape andarea on each frame.

ŽFor the velocity and concentration experiments, the mussels six total for each.experiment were individually exposed to six randomly selected treatments of current

speed from 7 to 30 cm sy1, keeping particle concentration constant in 1996, and to niney1 Žtreatments of particle concentration, increasing from 2000 to 40,000 ml particle

3 y1.volumes of 0.5–5 mm l with constant current speed in 1997. In 1997, musselŽresponses to orientation, to flow direction inhalant siphon into or away from flow

. Ždirection were also investigated at three current speeds approximately 10, 20 and 30y1.cm s . In order to standardize mussel responses to current speed, raw data on exhalant


siphon area and valve gape were converted to percent of maximum values observedduring the study, which occurred at the slowest flow treatments. A single mussel wasmonitored over 2–4 h with two low constant particle concentration treatments to observelonger term patterns in shell gape and exhalant siphon area.

In a final set of experiments in 2000, five individual mussels were exposed to twocurrent speeds, 10 and 30 cm sy1, and the exhalant siphon area and rates of ingestion ofpolystyrene beads were measured over 1-h exposure to experimental conditions.

3. Effects of velocity: 1996 experiments

Preliminary experiments demonstrated that mussel siphon area and shell gape re-sponded within 2 min to changes in flow velocity. Therefore, experimental protocolswere set up as follows:

1. One-hour acclimation in the flume for each individual mussel before flow treat-ments.

2. Randomly selected flow treatment: 9.8, 13.5, 18.9, 23.1, 27.9 and 32.6 cm sy1 for20 min.

3. Data capture for 10 min.4. Change flow to the next treatment, put in another mussel.

4. Effects of seston concentration

After a 30-min acclimation, mussels were exposed for 20 min to experimentalŽparticle concentrations followed by video data capture for 10 min. Current speed 16.4

y1. Ž .cm s and orientation to flow direction inhalant siphon facing upstream were keptconstant throughout these experiments. Particle concentration treatments were made byadding mixtures of algal species and sieved mudflat silt stock solutions to the sand-filteredseawater at the St. Andrews Biological Station in increasing concentration for eachsubsequent run. For mussel 9, after run c6, flume water was diluted by 50% with 1mmfiltered seawater for concentration c7 and again by 50% for concentration c8. Particleconcentrations for all the experiments ranged from about 2=103 particles mly1

Ž . 4 y1filtered seawater to about 4=10 ml at the highest particle concentration treat-ments. Similarly, particle volumes ranged from about 0.5 to 6 mm3 ly1. None of themussels spawned during the experiments.

5. Effects of orientation to flow direction

ŽTwo experiments were performed in 1997 to test the effects of orientation inhalant.siphon upstream or downstream with the exhalant siphon normal to the flow on valve

gape and exhalant siphon area at three flow speeds. Mussels 2, 4 and 9 were acclimated


to a given current speed and randomly chosen orientation for 15 min, followed by 15Žmin of data capture, moving the mussels to the other side of the Plexiglas stand so the

.incurrent siphon was facing downstream and repeating the process. Particle concentra-tions were constant, and the current speeds investigated were 11.3, 20.6 and 30.9 cmsy1.

6. Effects of velocity on filtration rates and exhalant siphon area using labeled diets

In May, 2000, after acclimation to laboratory conditions, individual mussels wereplaced in the flume with diets labeled with 13-mm polystyrene beads, and allowed toacclimate for 5 min, after which video data was recorded for a 1-h period. A total of 10frames were grabbed at 10-min intervals for each experimental treatment, each musselusing an Optamus video digitizer, and the exhalant siphon area measured in triplicate.After the mussels were exposed to the experimental diet and flow treatment, they wereplaced in 4-l containers filled with 1mm filtered seawater in a water bath at ambientseawater temperature. The mussels were fed every 4–6 h with the cultured algae over a36-h period. At the end of the biodeposition period, the mussels were removed and allthe feces were pipetted off the bottom into 2=15-ml glass capped test tubes. The tubeswere centrifuged for 2 min, and seawater was withdrawn; after which 10 ml of 1N nitricacid was added. The test tubes were then centrifuged for 5 min. The nitric acid waspipetted off, and 5 ml of 1mm filtered seawater was added. The tubes were sonicated

Ž .with two drops of type 1C Coulter dispersant main ingredient nonoxynol-9 , and pouredinto a coulter vial with 15 ml of filtered seawater, sonicated and run on a CoulterMultisizer. Particle size channels 25–32, which represented the polystyrene bead labelwith a peak at 12.1–15.7-mm equivalent spherical diameter, were summed to calculatetotal beads ingested per velocity treatment. Control mussels were run with no beads todetermine background levels of particles in the size range of interest. Filtration rateswere then determined by dividing the number of beads ingested by the concentration ofbeads in each experimental run. Filtration rates were converted to rates for a standard

Ž .animal of 1-g dry weight using an exponent of 0.66 Mohlenberg and Riisgard, 1979 .˚Some mussels spawned during the experiments, and were removed, due to the possibil-

Ž .ity of shunt flow through the gills during spawning Newell and Thompson, 1984 . Priorto the ingestion experiment, the time course of elimination of beads from eightindividual mussels at 24, 30, 35 and 43 h after exposure to experimental diets in theflume, indicated elimination of over 95% of the beads by 36 h.

7. Statistical analyses

Statistical analyses were performed with Systat for Windows. A univariate randomŽ .factors subjects , one-way ANOVA repeated measures design, was used for separate

tests of significance on the effects of current speed on percent maximum musselŽ .exhalant siphon area or valve gape 1996 , and for the effects of particle concentration

Ž .on percent maximum exhalant siphon area or valve gape 1997 . Percentage data were

()

C.R

.New

elletal.r

J.Exp.M

ar.Biol.E

col.2622001

91–

11197

Table 1Effects of current speed on exhalant siphon area and valve gape

Ž .a Mussel size and experimental diet and temperature for all velocity experiments

Mussel dry Mussel shell Concentration Size Volume Fluorescence Temperature SPMy1 3 y1 y1Ž . Ž . Ž . Ž . Ž . Ž . Ž .weight g length cm number ml mm mm l 8C mg l

Mean 3.35 8.12 47131 4.87 3.69 92.80 7.34 7.22S.D. 1.05 0.32 2805 0.22 0.25 5.38 0.50 1.05n 6 6 36 36 36 36 36 17

Ž .b Current speed treatments and mean and standard deviation of mussel exhalant siphon area and valve gape for each flume experiment

Speed Area Gape Area Gapey1 2Ž . Ž . Ž . Ž . Ž .cm s cm cm %maximum %maximum

Mean 9.79 0.65 0.70 98.07 83.71S.D. 0.64 0.29 0.25 3.17 19.92n 6 6 6 6 6Mean 13.46 0.54 0.64 82.80 77.59S.D. 0.43 0.22 0.16 14.92 13.84n 6 6 6 6 6Mean 18.95 0.42 0.70 65.61 83.44S.D. 0.47 0.18 0.18 18.58 16.43n 6 6 6 6 6Mean 23.10 0.34 0.66 49.91 79.40S.D. 0.93 0.23 0.18 17.11 15.94n 6 6 6 6 6Mean 27.89 0.22 0.64 33.67 76.20S.D. 0.45 0.14 0.19 14.97 15.64n 6 6 6 6 6Mean 32.62 0.14 0.62 20.31 75.12S.D. 0.47 0.14 0.18 13.25 16.81n 6 6 6 6 6


Ž .arcsine transformed Zar, 1974 prior to the repeated measures ANOVA. To protectagainst a Type I error and violations of sphericity, the probability values for significanceof the repeated measures, ANOVAs were adjusted using the Huynh–Feldt adjustmentŽ .Kirby, 1993 . Regression coefficients could not be calculated using least-squarestechniques since a repeated measure design was used for each experiment, and thereforethe data are not independent. However, by randomly choosing a different individualmussel’s siphon area or valve gape for each step of the current speed or diet concentra-tions, we were able to run least squares regression coefficients for each experiment usingExcel.

Ž .For the effects of orientation inhalant siphon facing into or away from current flow ,paired samplet-tests were used. For the bead ingestion experiments at two currentspeeds, paired samplet-tests were used to compare filtration rates or exhalant siphon

Ž . Ž .Fig. 1. Effects of current speed on A percent maximum exhalant siphon area, and B percent maximummussel shell gape in six individual mussels in a recirculating flume in March of 1996. Mussel orientation is

Ž .into flow direction with exhalant siphon facing up see text for details .


area at the two flow treatments with orientation into the flow direction. Filtration rateswere natural log-transformed to reduced differences between sample variances.

8. Results

8.1. Effects of Õelocity on exhalant siphon area and ÕalÕe gape

Mean particle concentrations, particle sizes, particle volumes, SPM levels, and watertemperatures were 47,131 mly1, 4.87 mm, 3.69 mm3 ly1, 7.2 mg ly1 and 7.38C,

Ž .respectively Table 1a . Mussel size averaged 8.1 cm and 1 g dry weight. Using allmussels studied, and each current treatment, the effects of current speed on percentmaximum exhalant siphon area and valve gape are shown in Fig. 1A and B, respec-tively. Repeated measures analysis of variance results, for the six levels of current speed,

Ž .for exhalant siphon area and valve gape as percent maximum are presented in Table 2.An increase in current speed from 10 to 30 cm sy1 caused a highly significant lineardecrease in exhalant siphon area but not a significant decrease in valve gape. Theregression equations had the following coefficients:

Asy3.7Uq133, r 2s0.79, ns6,where A is percent maximum exhalant siphon area andU is current speed in cm sy1;

Gsy0.90Uq109, r 2s0.25, ns6,whereG is percent maximum valve gape.

8.2. Effects of particle concentration and Õolume on exhalant siphon area and ÕalÕegape

Mean particle sizes, particle volumes, current speeds, fluorescence values, watertemperatures and SPM levels, for each mussel and seston concentration treatment, areshown in Table 3. Mussel size averaged 7.4 cm and 2.7-g dry weight. Seston concentra-tions were similar to those observed at natural mussel growing regions of coastal Maine

Table 2Repeated measures within subjects ANOVA for six levels of current speed and six mussels on exhalant siphon

Ž .area and valve gave as percent maximum using arcsine transformed dataŽ .Probability levels are from the Huynh–Feldt adjustment see text for details .

Source SS df MS F P

( )a Exhalant siphon area))Velocity 14938 5 2987 32.3 0.000

Error 2311 25 92.4

( )b ValÕe gapeVelocity 384 5 76.9 0.41 0.786 NSError 4655 25 186.2

)Significant.))Highly significant.

()

C.R

.New

elletal.r

J.Exp.M

ar.Biol.E

col.2622001

91–

111100Table 3

Effects of particle concentration and volume on exhalant siphon area and valve gapeCurrent speed, temperature, particle size and mussel size for all experiments.

Ž .a The seston characteristics and mean mussel exhalant siphon area or valve gape for each diet treatment

Speed Temperature Size Mussel dry Mussel shelly1Ž . Ž . Ž . Ž . Ž .cm s 8C mm weight g length cm

Mean 16.43 8.63 5.167 2.67 7.41S.D. 2.09 0.24 0.26 0.65 0.66n 54 54 53 6 6

Ž .b Fluorescence is the relative in vivo fluorescence and %maximum is the percent exhalant siphon area or valve gape relative to the maximum value for each mussel

Diet Concentration Volume Fluorescence Area Gape Area Gapey1 3 y1 2Ž . Ž . Ž . Ž . Ž . Ž .number ml mm l cm cm %maximum %maximum

1 Mean 2065 0.779 321 0.20 0.43 28.00 55.40S.D. 554 0.427 8 0.11 0.15 17.58 22.12n 6 6 6 6 6 6 6

2 Mean 5083 1.277 335 0.12 0.44 12.45 56.42S.D. 884 0.299 6 0.15 0.11 11.09 15.18n 6 6 6 6 6 6 6

3 Mean 8924 1.991 303 0.33 0.59 31.46 71.43S.D. 2899 1.571 14 0.31 0.24 17.08 23.85n 6 6 6 6 6 6 6

4 Mean 14188 3.159 349 0.45 0.57 47.21 69.85S.D. 2408 1.855 15 0.33 0.20 17.17 18.67n 6 6 6 6 6 6 6

5 Mean 18616 3.616 393 0.56 0.60 63.36 74.65S.D. 3133 2.039 17 0.33 0.15 12.22 10.09n 6 6 6 6 6 6 6

6 Mean 22555 3.325 429 0.67 0.65 78.00 81.49S.D. 3697 1.382 20 0.36 0.14 9.32 9.23n 6 6 6 6 6 6 6

7 Mean 29867 4.225 483 0.73 0.64 85.66 81.23S.D. 6707 1.205 29 0.39 0.13 9.46 9.66n 6 6 6 6 6 6 6

8 Mean 39170 5.126 553 0.85 0.77 97.97 96.27S.D. 7374 1.194 37 0.50 0.18 2.57 3.80n 5 5 6 6 6 6 6

9 Mean 43522 5.519 601 0.61 0.70 96.89 93.22S.D. 3484 0.708 11 0.36 0.17 6.22 13.25n 4 4 6 6 6 6 6


Ž y1. Ž .Fig. 2. Effects of particle concentration number ml on A percent maximum exhalant siphon area, andŽ .B on percent maximum valve gape in six individual mussels in a recirculating flume in November of 1997.

Ž .Newell and Shumway, 1993 . The effects of particle concentration on percent maxi-mum exhalant siphon area and valve gape are shown in Fig. 2A and B, respectively.There was a highly significant positive effect of seston concentration on both musselexhalant siphon area and valve gape over the diet treatments studied. Repeated measures

Table 4Repeated measures within subjects ANOVA for eight diet levels and six mussels on exhalant siphon area and

Ž .valve gape as percent maximum using arcsine transformed dataŽ .Probability levels are from the Huynh–Feldt adjustment see text for details .

Source SS df MS F P

( )a Exhalant siphon area))Diet 20092 7 2870 24.6 0.000

Error 2572 35 73.5

( )b ValÕe gape))Diet 4485 7 641 5.6 0.001

Error 4010 35 115



Ž 3 y1 . Ž . Ž .Fig. 3. Effects of particle volume mm l on A percent maximum exhalant siphon area, and B percentmaximum valve gape in six individual mussels in a recirculating flume in November of 1997.

analysis of variance results, for all six mussels which had eight of the nine diet levels,are shown in Table 4.

The best fit regression equations had the following coefficients:

As0.0017 C q22.7, r 2s0.65, ns9,Ž .whereCsconcentration as particles mly1;

Gs0.001 C q55.3, r 2s0.71, ns9.Ž .The effects of particle volume on percent maximum exhalant siphon area and shell

gape are shown in Fig. 3A and B, respectively. The best-fit regression equations had thefollowing coefficients:

As13.3 V q15.5, r 2s0.58, ns9,Ž .whereVsvolume as mm3 ly1;

Gs23.3 Ln V 55.8, r 2s0.90, ns9.Ž .


Ž 2. Ž . ŽFig. 4. Plot over time of exhalant siphon area cm or shell gape cm under constant conditions mussel 7,. Ž . Ž . Ž .shell length 67.1 mm, dry tissue weight 2.28 g in A 1mm filtered seawater treatment 1, Table 3 ; and B

Ž .in 1 mm filtered seawater plus algae treatment 2, Table 3 .

Two experiments were performed to determine the dependence of these behaviorsinvestigated on time during the 30-min experiments. We used mussel 7 to measuresiphon area and valve gape over extended periods of time in 1mm filtered seawater for

Table 5Effects of orientation to flow direction

Concentration Size Volume Fluorescence Temperature SPMy1 3 y1 y1Ž . Ž . Ž . Ž . Ž .number ml mm mm l 8C mg l

( )a Experimental conditions: particle concentration, Õolume, fluorescence, size, SPM and temperatureMean 32432 5.43 4.44 602 8.71 6.91S.D. 4239 0.18 0.91 23 0.26 0.72n 8 8 8 22 22 8

( )b Current speeds for each experimental run on effects of orientationMean 11.27 20.62 30.90S.D. 2.13 1.38 1.04n 4 12 6


Ž y1 y1 3 y1 .130 min Us13.8 cm s , 2189 particles ml , 0.8 mm l , 7.48C , and the filteredŽ y1 y1 3 y1 .seawater plus algaeUs15.7 cm s , 7556 particles ml , 1.6 mm l , 8.58C .

Ž 2. Ž .Exhalant siphon area in sqaure centimeters cm and valve gape in centimeters cm areshown for the 1mm filtered and filtered seawater plus algae diets in Fig. 4A and B,respectively. The mussels’ behavior was relatively constant over time, and both area andgape were greater when mussels were offered the more concentrated diets. Whendeclining concentrations were offered to mussel 9 at the end of the diet treatmentsŽ y1.38,366 to 14,624 to 4785 particles ml , there was a concomitant decline in exhalant

Ž . Žsiphon area from 100% to 62.5% to 50.5% and valve gape from 100% to 82.5% to.64.4% .

8.3. Effects of orientation toward flow direction

The effects of current speed on mussel exhalant siphon area depended on theorientation of the inhalant siphon to flow direction. Mussels exposed to constant diet and

Ž . Ž .Fig. 5. Effects of orientation on A percent maximum exhalant siphon area, or B percent maximum valvegape with mussels 2, 4 and 9 in a recirculating flume at three different current speed treatments. Musselsfacing into the flow direction were mounted with the exhalant siphon facing upward and the incurrentsiphonrmantle margin facing upstream. Mussels facing away from the flow direction were mounted with theincurrent siphonrmantle margin facing downstream. Values are means"standard errors.


Table 6Paired samplet-test on the effects of orientation to flow direction on percent maximum exhalant siphon area

Ž . Ž . Ž . y1and valve gape at a 10 b 20, and c 30 cm s

Mean S.D. df T pdifference

y 1( )a 10 cm s)Exhalant siphon area 10.4 1.2 1 12.2 0.05

Valve gape 27.1 6.9 1 5.6 0.11 NS

y 1( )b 20 cm s)Exhalant siphon area 39.9 30.9 5 3.17 0.03

Valve gape 8.5 14.7 5 1.42 0.2 NS

y 1( )c 30 cm s)Exhalant siphon area 62.6 22.2 2 4.9 0.04

Valve gape 3.3 10.6 2 0.53 0.65 NS


Ž . Žcurrent speeds Table 5 and variable orientation inhalant siphon at 08 or 1808 relative.to flow direction , at three different current speed treatments, had significantly reduced

Žsiphon areas facing into the flow then when they faced away from the flow Fig. 5,.Table 6 . The reduction in siphon area of mussels facing into the flow was greatest at the

Žhighest flow speeds studied, confirming the results of the March, 1996 experiments Fig..1 , but exhalant siphon area was independent of velocity when the orientation was facing

Ž .away from the flow Fig. 5A . Valve gape, however, was not effected by orientation atthe higher current speeds. However, at the lowest current speed there appears to be atendency for lower valve gape with the incurrent siphon orientated away from the

Ž .current see Fig. 5B .

Table 7Ž .a Experimental diets and current treatments for bead ingestion experimentsŽ . Ž 2.b Paired samplet-test on the effects of current speed on exhalant siphon area cm and filtration rate in

Ž y1. Ž .liters per hour l h for a 1-g standard mussel Ln-transformed data for five mussels.

Ž .a Experimental diets

Mean S.D. Ny1Ž .Particle concentration number ml 15219 1216 6

3 y1Ž .Particle volume mm l 2.23 0.26 6Ž .Particle sizemm 5.86 0.07 6

y1Current speed 10 cm s treatment 9.71 0.22 3y1Current speed 30 cm s treatment 30.84 1.17 3

Ž .b

Mean difference S.D. df T p)Area 0.068 0.042 4 3.60 0.023)Ž .Ln filtration rate 1.17 0.82 4 3.17 0.034


8.4. Effects of Õelocity on filtration rate

Experimental diets and flow speeds investigated are presented in Table 7a. The pairedŽ .sample t-tests results Table 7b for filtration rate and exhalant siphon area indicate

Ž .significant differences in both filtration rateps0.03 , and exhalant siphon areaŽ .ps0.02 over the two current speeds investigated. The filtration rate of five musselswas approximately 70% lower at 30 cm sy1 than at 10 cm sy1.

9. Discussion

Ž .Wildish and Saulnier 1993 noted a similar negative effect of high current speed onscallop exhalant mantle area, and suggested that the pressure differential between

Žinhalant and exhalant apertures overwhelmed the ability of the ciliary pump Jørgensen. Ž .et al., 1986a,b; Famme et al., 1986 to function properly. Famme et al. 1986 found a

linear decrease in pumping rate in blue mussels with an increase in hydrostatic pressureapplied experimentally. At increasing flow speeds, pressures applied to the inhalantŽ . ŽWildish and Miyares, 1990, and this study or to the exhalant Wildish and Saulnier,

.1993; Eckman et al., 1989 aperture facing into the flow would result in higher pressuresthan produced by the ciliary pump, resulting in slower growth rates or reduced filtrationrates. When the pressure field around the mussel is altered at high flows, pressuresexerted on the mussels may be positive on the upstream side or negative on thedownstream side. The distribution of pressure around cylinders has been studied in windtunnels, where maximum positive pressures occur at 08 orientation to flow direction, and

Ž .negative pressures occur from 90–2708 orientation Achenbach, 1968; Vogel, 1981 .Orientation of the mussel siphons relative to these pressure gradients in unidirectionaltidal flows may be of great importance in relation to mussel pumping rates. The

Žsignificant effects of orientation on exhalant siphon area at the two higher flows 20 andy1 .30 cm s , respectively support the conclusion that the disruption of pumping rate by

Ž .high flows is due to pressure differential Wildish et al., 1987 .When the exhalant siphon is completely closed, there is a cessation of the beating of

Žlateral cilia in several bivalve species Sleigh, 1962; Ockelmann and Muss, 1978;.Jorgensen and Ockelmann, 1991 . The coordinated control of the exhalant siphon

musculature and activity of the lateral cilia is thought to be part of the nervousregulation of the filtration response by the mussel, which includes the adductor and

Ž .retractor muscles of the mantle Jorgensen and Ockelmann, 1991 . The partial closure ofthe exhalant siphon results in an increase in the velocity of the water through the

Ž .aperture but still a decrease in volume flux. Foster-Smith 1976 examined the pumpingrates of M. edulis using a thermistor probe in relation to exhalant siphon area.Recalculation of the data from his Fig. 7 indicates only a 32% increase in flow speed

Ž .through the exhalant siphon with a 67% decrease in exhalant siphon area. Vogel 1981discusses the relationship between the volume flux of a fluid,Q, and an orifice such as aan exhalant siphon opening. The volume flux is a function of the square of the orificearea. Therefore, small changes in siphon diameter will have a large effect on volumeflux through the orifice. In our bead ingestion experiments, the percent reduction inexhalant siphon area of about 33% resulted in a decrease in filtration rate of about 70%.


With the exhalant siphon closed, particles are still likely to be transported by theŽ .frontal cilia, which are continually active along the ctenidia Sleigh, 1962 , eventually

being ingested by the mussels.While the exhalant siphon responded rapidly to increases in current speed, valve gape

was not significantly effected over the flows studied. Maintaining an open valve gapewith partial or full closure of the exhalant siphon under high-flow conditions isinterpreted with respect to increased oxygen diffusion for respiration, as discussed by

Ž .Davenport 1979 . The ability of mussels to adjust their pumping rate in a rapid andcontrolled manner to increase water velocity using the musculature of the mantle and

Ž .siphon, but independent of valve gape controlled by the adductor muscle observed inŽ .this study, is similar to the responses of mussels observed by Davenport 1979 to

changes in seawater salinity. The three-part behavioral sequence observed in that studywhen salinity dropped from 33 to 0 ppt over 1 h was as follows: exhalant siphon closesŽ . Ž . Ž .25 ppt , inhalant siphon closes 20 ppt , complete valve closure 10 ppt . This ability toclose the siphons allows the mussels to maintain full-strength salinity in the mantle

Ž .cavity in brackish water Shumway, 1977 while allowing for greater exposed mantleŽ .area for oxygen diffusion. Other workers Frechette and Lefaivre, 1990 suggest that´

open valves may also help to maintain space occupation.In summary, the independence of valve gape to seawater velocity observed during

this study is interpreted as reflecting the need for respiratory demands during digestion,while the strong relationship between exhalant siphon area and velocity is interpreted asa function of hydrodynamics. Higher currents experienced by mussels in suspensionculture, in contrast to the lower currents experienced in the benthic boundary layer onthe seabed, may result in some feeding inhibition and reduced growth rates, dependingupon local hydrodynamics, particle concentration and mussel orientation.

At lower current speeds, the effects of current are positive with respect to particleŽ .supply to bivalves such as mussels Wildish and Kristmanson, 1984, 1985 . Juvenile

oyster growth rate was an increasing function of two factors: flow speed and particleŽ . y1concentration Lenihan et al. 1996 at flows from 0 to 7 cm s , but the factors were

independent and the interactive term, food flux, had no significant effect in theirANOVA. Therefore, the response of mussel filtration rate to current speeds above 10 cmy1 Ž .s are probably due to unfavorable hydrodynamics pressure differential , while the

responses below 10 cm sy1 are due to indirect effects on particle concentrations.Ž .Butman et al. 1994 observed different uptake rates of phytoplankton fluorescence by a

Ž y1.blue mussel bed in a shallow flume at two-flow speeds 5 and 15 cm s , andsuggested that mussel bed filtration rates may be flow speed dependent, due to theindirect effects of flow speed on particle concentration available to a mussels in thebenthic boundary layer.

Many authors have described the effects of particle concentration on bivalves such asŽ .mussels. Winter 1976, Fig. 11 expressed the filtration and ingestion rates of bivalves as

a function of the concentration of unicellular algae. Filtration rate increased rapidly froma low AthresholdB concentration until there was a maximum ingestion rate, above whichthere was a decline in filtration rate. This so-called unimodal response is supported byour results, in which both the exhalant siphon area and valve gape are significantlyaffected by particle concentration with constant current speed. Examination of the data


Žin Fig. 2 indicates that the threshold for the filtration response by opening the valve.gape, extending the mantle and opening the exhalant siphon to particle concentration is

somewhere between 2000 and 6000 particles mly1, and particle volumes between 0.5and 1.5 mm3 ly1. These feeding behavior observations in a laboratory flume correspond

Žwell with observed reductions in valve gape in the field Newell et al., 1998; Campbell. Žand Newell, 1998 over similar ambient particle concentrations. Other workers Riisgard.and Randlov, 1981 found similar reductions in valve gape and filtration rates of blue

y1 Ž .mussels at algal cell densities below 1500 cells ml . Butman et al. 1994 alsoobserved reduced feeding rates by a mussel bed at lower diet levels in their flume,although particle sizes, concentrations and volumes were not reported.

Ž .Extended periods of low particle concentrations i.e., Fig. 4A above result in valveclosure, reduced pumping rates and respiration rates switching fromAactiveB to Astan-

Ž .dardB rates Thompson and Bayne, 1972 until conditions improve. The 27% decrease invalve gape and the 10% decrease in exhalant siphon area, with orientationaway from

y1 Ž .flow direction at the 10 cm s flow Fig. 5B , could be due to refiltration of the waterŽ .at low flows O’Riordan et al., 1995 , making the effects of low current speeds on the

filtration response, an indirect one, mediated by particle supply and localized particledepletion by the active suspension-feeders. For example, in Kirby–Smith growth tubes,

Žmussels grew faster at inlet ends of pipes than outlet ends at low current speeds undery1. Ž .1.05 cm s , due to seston depletion Wildish and Kristmanson, 1985 . In contrast to

the effects of current speed above, under low particle concentrations respiration may beminimized by decreasing the diffusion of oxygen in the mantle cavity through reducedpumping rates and valve closure.

Ž .In previous studies Jørgensen et al., 1986a,b, 1988 , mussels were physicallyconstrained in their valve gape using an external screw, and observations were made ongape and exhalant siphon area in relation to water pumping rates and filtration rates.Jørgensen et al., 1988 writes,Apresumably, extension of mantle edges and exhalantsiphon is closely coupled to the valve gape,B but the results in this study usingunrestrained mussels demonstrate an independent control of the exhalant siphon aperture

Žand valve gape in relation to ambient hydrodynamics gape open, siphon closed at high. Žflows , or coordinated activity in relation to particle concentration gape and siphon

.open with increasing particle concentration . Therefore, the significance of the shellgape and exhalant siphon area responses, as recorded with time-lapse video, is that theyreflect separate responses to both current speed and particle concentration. These results,

Ž .in combination with those of Jørgensen et al. 1988 , suggest that exhalant siphon area isa useful index of pumping rate, whereas valve gape is a more general indicator of

Ž .periods of active feeding and respiration gape open or closed for mussels in situ. Thedecrease in exhalantsiphon area with increasing current speed observed in this studymirrors the decrease infiltration rate over a similar velocity range by Wildish and

Ž .Miyares 1990 in a modified recirculating Vogel flume, when their data are expressedŽ .as percent maximum filtration rate Fig. 6 . The connection between siphon area and

Ž .mussel pumping Foster-Smith, 1976; Famme et al., 1986; Jørgensen et al., 1988 , andŽ .filtration this study rates demonstrates the usefulness in using video records of

undisturbed bivalves for determining the effects of environmental factors, such ascurrent speed and particle concentration on mussel feeding behavior. These results


Ž .Fig. 6. The effects of velocity on percent maximum exhalant siphon area mean"standard error and percentŽ .maximum filtration rate from the data of Wildish and Miyares 1990 .

confirm similar findings in the scallopPlacopecten magellanicus, where exhalantsiphon area responded to experimental changes in both current speed and particle

Ž .concentration Wildish and Saulnier, 1993 .Further studies of bivalve filtration rates in relation to ambient seston concentrations

should utilize well-characterized orientations to flow and flow speeds within experimen-tal chambers, to prevent decreases in mussel pumping rates due to hydrodynamics alone.These data suggest that suspension feeding in mussels is a finely tunedAhomeorheo-

ŽstaticB response of pumping rate to seston and ambient hydrodynamics Wildish et al.,.1992; Wildish and Saulnier, 1993; Wildish and Kristmanson, 1997 . Further work with

labeled diets and responses of mussel feeding behavior to oscillating currents fromwaves, where hydrodynamics may disturb the feeding process or reduce localized sestondepletion, will aid in our ability to understand optimal conditions for the suspensionculture of this species, and the usefulness of time-lapse video in investigations of musselfeeding dynamics in the field.

Acknowledgements

The authors would like to thank the following individuals for aiding in the comple-tion of this study: Hugh Agaki for help with the flume, Jennifer Martin for algal culture,Dr. Cindy Pilskaln for use of an Optamus video system at the University of Maine, St.Andrews Biological Station for working facilities, the University of New Brunswick inSaint John and the Huntsman Marine Science Center for the use of facilities, GreatEastern Mussel Farms, for use of the time-lapse video recorder, the Eastern MaineConservation Initiative for financial support, Dr. Steve Fegley for helpful comments onthe statistical analyses, and two anonymous reviewers for helpful comments on the

[ ]manuscript. SS


References

Achenbach, E., 1968. Distribution of local pressure and skin friction around a circular cylinder in cross-flowup to Res5=106. J. Fluid Mech. 34, 625–639.

Bayne, B.L., 1993. Feeding physiology of bivalves: time-dependence and compensation for changes in foodŽ .availability. In: Dame, R.F. Ed. , Bivalve Filter Feeders in Estuarine and Coastal Ecosystem Processes.

Springer-Verlag, Berlin, pp. 1–24.Butman, C.A., Frechette, M., Geyer, W.R., Starczak, V.R., 1994. Flume experiments on food supply to the´

blue musselMytilus edulis L. as a function of boundary-layer flow. Limnol. Oceanogr. 30, 1755–1768.Campbell, D.E., Newell, C.R., 1998. MUSMOD, a production model for bottom culture of the blue mussel

Mytilus edulis L. J. Exp. Mar. Biol. Ecol. 219, 171–204.Ž .Davenport, J., 1979. The isolation response of musselsMytilus edulis L. exposed to falling seawater

concentrations. J. Mar. Biol. Assoc. U. K. 59, 123–132.Eckman, J.E., Peterson, C.H., Calahan, J.A., 1989. Effects of flow speed, turbulence, and orientation on

Ž .growth of juvenile bay scallops,Argopecten irradians concentricus Say . J. Exp. Mar. Biol. Ecol. 132,123–140.

Famme, P., Riisgard, H.U., Jørgensen, C.B., 1986. On direct measurements of pumping rates in the mussel˚Mytilus edulis. Mar. Biol. 92, 323–327.

Foster-Smith, R.L., 1975. The effect of concentration of the suspension on the filtration rates and pseudofecalŽ .production forMytilus edulis L., Cerastoderma edule L. andVenerupis pullastra Montagu . J. Exp. Mar.

Biol. Ecol. 17, 1–22.Foster-Smith, R.L., 1976. Some mechanisms for the control of pumping activity in bivalves. Mar. Behav.

Physiol. 4, 41–60.Frechette, M., Bourget, E., 1985. Energy flow between the pelagic and benthic zones: factors controlling´

particulate organic matter available to an intertidal mussel bed. Can. J. Fish. Aquat. Sci. 42, 1158–1165.Frechette, M., Lefaivre, D., 1990. Discriminating between food and space limitation in benthic suspension´

feeders using self-thinning relationships. Mar. Ecol.: Prog. Ser. 65, 15–23.Frechette, M., Butman, C.A., Geyer, W.R., 1989. The importance of boundary-layer flows in supplying´

phytoplankton to the benthic suspension feeder,Mytilus edulis L. Limnol. Oceanogr. 37, 19–36.Jørgensen, C.B., 1960. Efficiency of particle retention and rate of water transport in undisturbed lamelli-

branchs. J. Cons., Cons. Int. Explor. Mer. 26, 95–116.Jørgensen, C.B., 1990. Bivalve Filter Feeding: Hydrodynamics, Bioenergetics, Physiology and Ecology. Olsen

and Olsen, Fredensborg, Denmark.Jorgensen, C.B., Ockelmann, K.W., 1991. Beat frequency of lateral cilia in intact filter feeding bivalves: effect

of temperature. Ophelia 33, 67–70.Jørgensen, C.B., Famme, P., Kristensen, H.S., Larsen, H.S., Mohlenberg, P.S., Riisgard, H.U., 1986a. The˚

bivalve pump. Mar. Ecol.: Prog. Ser. 34, 69–77.Jørgensen, C.B., Mohlenberg, F., Sten-Knudsen, O., 1986b. Nature of relation between ventilation and oxygen

consumption in filter-feeders. Mar. Ecol.: Prog. Ser. 29, 73–88.Jørgensen, C.B., Laren, P.S., Mohlenberg, F., Riisgard, H.U., 1988. The mussel pump: properties and˚

modelling. Mar. Ecol.: Prog. Ser. 45, 205–216.Kirby, K.N., 1993. Advanced Data Analysis with Systat. Van Norstrand Reinhold, New York, 475 pp.Kramer, K.J., Jenner, H.A., deZwart, D., 1989. The valve movement response of mussels: a tool in biological

monitoring. Hydrobiologia 188, 433–443.Lenihan, H.S., Peterson, C.H., Allen, J.M., 1996. Does flow speed also have a direct effect on growth of active

suspension-feeders: an experimental test on oysters. Limnol. Oceanogr. 41, 1359–1366.Manley, A.R., 1983. The effects of copper on the behaviour, respiration, filtration and ventilation activity of

Mytilus edulis. J. Mar. Biol. Assoc. U. K. 63, 205–222.Mohlenberg, F., Riisgard, H.U., 1979. Filtration rate, using a new indirect technique, in thirteen species of˚

suspension feeding bivalves. Mar. Biol. 54, 143–147.Ž .Newell, C.R., 1990. The effects of musselMytilus edulis, Linnaeus, 1758 position in seeded bottom patches

on growth at subtidal lease sites in Maine. J. Shellfish Res. 9, 113–118.


Newell, C.R., Shumway, S.E., 1993. Grazing of natural particulates by bivalve mollusks: a spatial andŽ .temporal perspective. In: Dame, R.F. Ed. , Bivalve Filter Feeders in Estuarine and Coastal Ecosystem

Processes. NATO ASI Ser., G, vol. 33. Springer, Verlag, Berlin, pp. 85–148.Newell, R.I.E., Thompson, R.J., 1984. Reduced clearance rates associated with spawning in the mussel

Ž . Ž .Mytilus edulis L. BiÕalÕia Mytilidae . Mar. Biol. Lett. 5, 21–33.Newell, C.R., Campbell, D.E., Gallagher, S.M., 1998. Development of the mussel aquaculture lease site model

MUSMOD: a field program to calibrate model formulations. J. Exp. Mar. Biol. Ecol. 219, 143–169.Ockelmann, K.W., Muss, K., 1978. The biology, ecology and behaviour of the bivalveMysella bidentate

Ž .Montagu . Ophelia 17, 1–93.O’Riordan, C.A., Monismith, S.G., Koseff, J.R., 1995. The effects of bivalve excurrent jet dynamic on mass

transfer in a benthic boundary layer. Limnol. Oceanogr. 40, 330–344.Riisgard, H.U., Randlov, A., 1981. Energy budgets, growth and filtration rates inMytilus edulis at different

algal concentrations. Mar. Biol. 61, 227–234.Saunders, H.E., Hubbard, C.W., 1944. The circulating water channel of the David W. Taylor model basin.

Soc. Nav. Archit. Mar. Eng., Trans. 52, 325–364.Shumway, S.E., 1977. Effect of salinity fluctuations on the osmotic pressure and Na, Ca, Mg ion concentra-

tions in the haemolymph of bivalve mollusks. Mar. Biol. 41, 153–177.Sleigh, M.A., 1962. The Biology of Cilia and Flagella. Macmillan, New York, 242 pp.Thompson, R.J., Bayne, B.L., 1972. Active metabolism associated with feeding in the mussel,Mytilus edulis

L. J. Exp. Mar. Biol. Ecol. 9, 111–124.Vogel, S., 1981. Life in Moving Fluids. Princeton Univ. Press, 352 pp.Wildish, D.J., 1991. A history of flow simulation studies at St. Andrews Biological Station. Can. Manuscr.

Rep. Fish. Aquat. Sci. 2134, 25 pp.Wildish, D.J., Kristmanson, D., 1984. Importance to mussels of the benthic boundary layer. Can. J. Fish.

Aquat. Sci. 41, 1618–1625.Wildish, D.J., Kristmanson, D., 1985. Control of suspension-feeding bivalve production by current speed.

Helgol. Wiss. Meeresunters. 39, 237–243.Wildish, D.J., Kristmanson, D., 1997. Benthic Suspension Feeders and Flow. Cambridge Univ. Press, New

York, 409 pp.Wildish, D.J., Miyares, M.P., 1990. Filtration rate of blue mussels as a function of flow velocity: preliminary

experiments. J. Exp. Mar. Biol. Ecol. 142, 213–219.Wildish, D.J., Saulnier, A.M., 1992. The effect of velocity and flow direction on the growth rate of juvenile

and adult sea scallops. J. Exp. Mar. Biol. Ecol. 155, 133–143.Wildish, D.J., Saulnier, A.M., 1993. Hydrodynamic control of filtration inPlacopecten magellanicus. J. Exp.

Mar. Biol. Ecol. 174, 65–82.Wildish, D.J., Kristmanson, D.D., Hoar, R.L., DeCoste, A.M., McCormick, S.D., White, A.W., 1987. Giant

scallop feeding and growth responses to flow. J. Exp. Mar. Biol. Ecol. 113, 207–220.Wildish, D.J., Kristmanson, D., Saulnier, A.M., 1992. Interactive effect of velocity and seston concentration

on giant scallop feeding inhibition. J. Exp. Mar. Biol. Ecol. 155, 161–168.Winter, J.E., 1976. A critical review on some aspects of filter-feeding in lamellibranchiate bivalves. Haliotis 7,

71–87.Winter, J.E., 1978. A review of the knowledge of suspension-feeding in lamellibranchiate bivalves, with

special reference to artificial aquaculture systems. Aquaculture 13, 1–33.Zar, J.H., 1974. Biostatistical Analysis. Prentice-Hall, New Jersey, 620 pp.

The effects of velocity and seston concentration on the exhalant siphon area, valve gape and filtration rate of the mussel Mytilus edulis

Documents