MARINE ECOLOGY PROGRESS SERIES Vol. 166: 173-185, 1998 ' Published May 28 Mar Ecol Prog Ser

Suspension-feeding behaviour in tropical bivalve molluscs: viridis, belcheri, Crassostreairadelei, Saccostrea cucculata and Pinctada margarifera

A. J. S. Hawkins',', R. F. M. Smith1, S. H. h an^, Z. B. yasin2

'Plymouth Marine Laboratory, Natural Environment Research Council, West Hoe, Plymouth PLl 3DH, United Kingdom 2Universiti Sains Malaysia, 11800 Pulau Pinang, Malaysia

ABSTRACT Filter-feeding behaviours of the Perna viridis and the Crassostrea belchen, Crassostl-ea ~radelei,Saccostrea cucculata and Pinctada mal-garifera were compared during natural tidal vanations in the concentration (6 to 40 dry mg total particulate mass I-') and organic content (6 to 22%)of seston available in the Merbok mangrove system, Malaysia. In P. viridis, C. belcheriand S. cuc- culata, declining retention eff~cienclesfor the largest available particle slze classes of more than 6 to 12 pm diameter were associated wth an overall organic enrichment of filtered relative to available matter. As seston avdilab~lityincreased, a ~nin~mumaverage of 71°,. of the additional filtered matter was rejected by each species as pseudofaeces prior to ingestion. And in all 5 species, preferential rejec- tion as pseudofaeces of filtered particles with higher average inorganic content resulted in the net organic enrichment of ingested relative to filtered matter. In P. viridis and C. belcheri, the efficiency of that net organic selection declined for seston of lower organlc content. Combining all data from each species, absorption efficiencies from ingested organics increased exponent~ally with the organic content of ingested matter Collective findings suggest that key interrelations between component processes of nutrient acquisition were similar to those that have been established for the modeling of feeding, growth and environmental ~mpactamong filter-feeding bivalves froin temperate latitudes However, especially fasl growth in P viridis stemmed from higher average (* 2 SE) clearance rates (13.2 A 7.0 l h-' g-' when particulate organic matter c 1 mg I-'), greater average organic enrichment of filtered relative to available matter (0.99 + 0.67) and greater average organic enrichment of ingested relative to filtered matter (0.63 * 0.03) than have hitherto been recorded in any species of bivalve filter- feeder. In contrast, P. margarifera was least well adapted for the pre-ingestive selection of organic mat- ter, with no differential retention on the gill and little selective rejection as pseudofaeces, representing at least part of the physiological basis for characteristically slow growth in pearl oysters.

KEYWORDS: Filter-feeding behaviour Particle selection . Seston co~nposition. Perna viridis . Cras sostrea belcheri . Crassostrea iradelei Saccostrea cucculata Pinctada margarifera

INTRODUCTION Bayne & Hawkins 1992, Dame 1993, Herman 1993, Smaal & Prins 1993) and to predict the carrying capac- Filter-feeding behaviour in bivalves is proving ity for culture within nearshore environments highly responsive to fluctuations in both the abun- (e.g. Heral 1993). Findings illustrate a complex set of dance and the composition of suspended seston adaptations which not only help to maximise organic (Hawkins & Bayne 1992, Bayne 1998). Further under- absorption, but also to maintain organic absorption rel- standing of that behaviour is required to establish the atively independent of short-term fluctuations in ses- role of bivalve shellfish in ecosystem processes (e.g. ton abundance and composition, including the 'dil'u- tion' of organic matter by silt that is typically resuspended within higher concentrations of natural

O Inter-Research 1998 Resale of full article not permitted 174 Mar Ecol Prog Ser 166: 173-185, 1998

seston (e.g. Hawkins et al. 1996, 1997, Urrutia et al. from temperate latitudes (Hawkins et al. 1998). We 1996, Navarro et al. 1997). Selective processes, and the also compare the feeding behaviour of our studied spe- consequences of those processes, are a particularly cies with a view to understanding observed differences important feature of that behaviour. And because algal in growth rate and ecogeographical distribution. monocultures and other artificial suspensions do not afford utilisation of the full potential for particle selec- tion, an emergent consensus now recognises that eco- MATERIALS AND METHODS logically relevant responses are best observed when the bivalves are feeding upon natural seston (Foster- Feeding behaviours of the mussel Perna viridis and Smith 1975, Doering & Oviatt 1986, Cranford & Gordon the oysters Crassostrea belcheri, Crassostrea iradelei, 1992, Iglesias et al. 1992, 1996, Navarro et al. 1992, Saccostrea cucculata and Pinctada margadera were 1994, Bayne et al. 1993, MacDonald & Ward 1994, studied in response to natural variations of food avail- Hawkins et al. 1996, Soletchnik et al. 1996, Urrutia et ability during 3 complete tidal cycles at Kampong al. 1996, Bacon et al. 1998, MacDonald et al. 1998). Bukit Kechik in the Merbok mangrove system, about Recent work in the Cerastoderma edule (Ur- 30 miles (48 km) north of Penang, Malaysia. rutia et al. 1996), the Crassostrea gigas (Soletch- Whereas Crassostrea iradelei had been cultured at nik et al. 1996) and the mussel Mytilus edulis (Hawkins Kampong Bukit Kechik, the remaining species were et al. 1996) feeding throughout the same natural tidal collected from other sites as follows: Perna viridis from variations in seston availability has (1) identified a Pulau Aman, Penang; Crassostrea belcheri from Tel- common set of equations that summarize functional aga Nenas, Perak; and both Saccostrea cucculata and interrelations between processes of particle filtration, Pinctada margarifera from Muka Head, Penang. pre-ingestive rejection and organic absorption, and Regardless of origin, all specimens were first accli- (2) shown that those relations may be used to predict mated to common conditions of 27.0 * l.O°C and 32.0 + feeding behaviour and net organic absorption rate in l.OOhwhile supplied with a copious through-flow of all 3 species on the basis of seston abundance and ses- unfiltered natural seawater for 2 wk at the Muka Head ton organic content alone (Hawkins et al. 1998). These Research Station, Universitat Sains Malaysia, Penang. findings establish that similar functional interrelations Before experimentation, they were transported to control feeding responses in C. edule, C. gigas and M. Kampong Bukit Kechik, and placed within immersed edulis. We now need to test for the further generality of plastic trays that were suspended below a raft owned these key relations, for potential use in a common by the Fisheries Research Institute of Penang. Starting model that predicts filter-feeding and growth in sepa- 48 h later, and working from that same raft, we moni- rate species under different environmental conditions tored feeding responses over the 6 to 8 h that spanned of food availability and composition. the rise from low to high tide on each of 9, 10 and l1 To date, there have been no reports of the responses January 1996. Ranges of tidal rise and fall throughout to variation in seston abundance or composition among those 3 d exceeded 1.8m. filter-feeding bivalve shellfish from tropical latitudes. Samples were collected over separate experimental Yet shellfish dominate many nearshore communities in periods of 1 to 3 h during each tidal cycle. Species were tropical environments. Work described here was un- on occasion exchanged between periods of measure- dertaken to study the filter-feeding behaviour of the ment, allowing >l h to adapt following handling. For green mussel Perna viridis L., the mangrove oysters each penod of measurement, up to 11 individual shell- Crassostrea belcheri and Crassostrea iradelei, the rock fish were each maintained separately in 1 1plastic trays oyster Saccostrea cucculata, and the pearl oyster Pinc- that were supplied with seawater (27.0 * l.O°C) tada margarifera. All 5 species are cultured commer- pumped from beneath the raft. Rates of seawater sup- cially throughout tropical waters (e.g. Rangarajan & ply were first adjusted to ensure that concentrations of Narasimham 1980, Sivalingam 1982, Yoo et al. 1986, suspended particles were not reduced by more than Brohmanonda & Mutaransit 1988, Jarayabhand & Tha- 25 % between the inflow and outflow from each shell- vornyutikarn 1995). To ensure natural behaviour and fish tray, and then maintained at constant rates which ecologically meaningful comparisons, responses to differed between 170 and 450 m1 min-' throughout natural variations of seston were monitored within the each experimental penod. There were no significant Merbok mangrove system, Malaysia. Findings are differences between particle sedimentation rates at dif- used to examine whether functional interrelations that ferent flow rates, and maximal flow rates in each shell- influence feeding responses in tropical species are fish tray of c0.5 cm s-' were well below rates of >l5 cm similar to the common set of relations that have re- S-' that have been shown to inhibit filter-feeding in cently been established for use within models of feed- other species of bivalve shellfish (e.g. Wildish & ing behaviour and growth in filter-feeding bivalves Saulnier 1993). At each collection time, rate of current Hawkins et al.: Suspension-fee ?ding behav~ourin tropical bivalves 175

flow at the experimental site was measured by timing (refer to 'Results'). Therefore, to approximate pumping the passage of an immersed piece of coconut husk over rate as closely as possible, CR was only calculated for 2 m alongside the raft. Water temperature was also particles of 3 to 8 pm. recorded, and salinity measured with a refractometer. Retention efficiencies (REO, fraction) with which the In addition, 1 1 of seawater was sampled from the out- organic content of filtered particles (OCF; fraction) was flows of each tray containing an experimental individ- enriched compared with the organic content of the all ual, including 1 empty 'control' tray. Then, the flow rate available seston (OCS; fraction) were calculated as: through each tray was recorded, and all true faeces and RE0 = (OCF OCS)/OCS all pseudofaeces collected separately. To obtain suffi- - cient precision in our measures of faecal production Other processes of feeding and absorption were cal- and faecal composition, the true faeces or pseudofaeces culated according to the basic procedures of Hawkins were each combined from replicate groups of 2 to 3 in- et al. (1996, 1998), Soletchnik et al. (1996) and Urrutia dividual shellfish, before being homogenised by rapid et al. (1996). All calculations are summarized with repeat-pipetting. Aliquots of seawater and faecal sam- acronyms in Table 1.Note that these calculations as- ples were filtered onto separate pre-weighed and sume no significant delay in the deposition of filtered ashed Whatman GF/C filters. All filters were dried at particles, estimating particle availability as the inte- 60°C before re-weighing and calculation of the total dry grated average over corresponding periods of faecal sample weight per filter. Each filter was then ashed at collection. Note also that REO, NOSE (net organic se- 450°C for 4 h prior to final weighing, allowing further lection efficiency), NOIR (net organic ingestion rate), calculation both of the ash (inorganic) and ash-free (or- NOAR (net organic absorption rate), NAEFO (net ab- ganic) masses of each filtered sample. sorption efficiency from filtered organics) and NAEIO Seston abundance was measured as the total dry (net absorption efficiency from ingested organics) are particulate mass (TPM; mg I-') of all available seston, net measures; REO, NOSE and NOIR being influenced and seston quality as the organic content of total dry by particle losses or mucous losses from the shellfish, seston (OCS; fraction). Measures of TPM and OCS whereas NOAR, NAEFO and NAEIO are influenced were undertaken assuming that concentrations mea- both by mucous losses and by metabolic faecal losses. sured in the outflow from the empty 'control' tray were Following all experimentation, the soft tissues were rcprcscntativc of !he ~v~i!;b!e par:iculci:es. excised fr~iiieach shellfish and dried at GG°C before Each water sample from both experimental and con- measuring the total soft tissue dry weight. Feeding trol outflows were analysed for both particle count and responses were standardised to those for an equivalent size distribution. These and other measures were made individual of 1 g soft tissue dry weight as follows: Y,= using a Coulter Channelyser and associated tube with (W,/ W,)"Ye, where Y, is the standardised parameter, orifice diameter of 100 pm, recording the particle num- W, is the standard weight (1 g), W, is the weight or bers per 2 m1 of sample within each of 32 slze channels length of the experimental individual, Ye is the uncor- over the range 3 to 17 pm equivalent spherical diame- rected parameter, and b is the average size exponent ter. Efficiencies with which filtered particles within of 0.62 for feeding processes in a variety of suspension- specific size classes were retained on the gill (REP, feeding shellfish (Bayne & Newel1 1983). This standard fraction) were calculated as: weight of 1 g soft tissue dry weight compared with averages (k2 SE) for all experimental individuals of REP = (inflow particle counts ml-l within size class - each species as follows: 0.40 + 0.03 g for Perna viridis, outflow particle counts ml' within size class)/ 1.02 0.07 g for Crassostrea belcheri, 0.91 * 0.11 g for (inflow particle counts ml-' within size class) Crassostrea iradelei, 0.48 * 0.08 g for Saccostrea cuc- where the inflow particle counts were measured culata and 0.86 k 0.24 g for Pinctada margarifera. within outflow water from the control chamber without It was not possible to study the feeding behaviour of an experimental individual. all 5 species simultaneously. This meant that measures Clearance rate (CR; 1 h-') with which each shellfish were made over different average periods of food species removed particles from seawater was calcu- availability (TPM; mg 1-l). However, measures in Perna lated as: viridis and Crassostrea belcheri were undertaken to- gether over a common range of TPM from ? to 15 mg CR = (number of particles ml-' within the inflow - I-', and measures in Saccostrea cucculata and Pinctada number of particles ml-' within the outflow)/ margarifera were undertaken together over a similar (number of particles ml-' within the outflow) X common range of TPM from 10 to 23 mg 1-l. Measures litres of water flow h-' through tray in Crassostrea iradelei were made over a limited range Particle retention efficiencies declined for particles of TPM from 8 to 9 mg I-'. Measures of feeding behav- larger than about 12 pm equivalent spherical diameter iour were each made upon a different replicate group 176 Mar Ecol Prog Ser 166: 173-185, 1998

Table 1. Definitions and descriptions of the calculation of separate components of feeding behaviour

Parameter Acronym Unit Calculation

Clearance rated CR l h-' (number of particles ml' within the inflow - number of particles ml-' within the outflow)/(number of particles ml'' within the outflow) X litres water flow h-' through tray Filtration rate Total mg h'' (mg inorganic matter egested both as true faeces and pseudofaeces h-')/(l - OCF) Retention efficiency REP Fraction (inflow particle counts ml-' within size class - outflow particle counts for filtered particles ml-' within size class)/(inflow particle counts ml-' within size class) Retention efficiency for RE0 Fraction (OCF - organic fraction within total particulates available in tray filtered organic matter inflow)/organic fraction within total particulates available in tray inflow Organic content of OCF Fraction (mg part~culateorganic matter 1-' w~thintray inflow - mg particulate filtered matter organic matter I-'within tray outflow)/( mg total particulate matter 1-' within tray inflow - mg total particulate matter I-' within tray outflow) Kelection rate RR Total mg h-' mg total pseudofaeces egested h-' I Net organlc selection NOSE Fraction 1 - (organic fraction w~thinpse~ldofaeces)/(organic fraction withln efficiency total part~culatesavailable in tray inflow) Net organic ingestion rate NOIR mg h-' (PR X OCF) - (RR X organic fraction within pseudofaeces) Net organic absorption rate NOAR mg h-' NOIR - [(mg total true faeces egested h-') X (organic fraction within true faeces)] Net absorption efficiency NAEFO Fraction NOAW(FR X OCF) from filtered organics Net absorption efficiency NAEIO Fraction NOAWNOIR from ingested organics Organic content of OCI Fraction NOIR/(FR - RR) ingested matter

aCalculated for particles from 3 to 8 pm diameter (refer to 'Material and methods')

of 2 to 3 individuals (refer above), and differed in num- models with unstable regression coefficient estimates, ber (n) between species, with n = 12 in P. viridis and C. we also excluded any variable with a tolerance of ~0.1 belcheri, n = 6 in S. cucculata and Pinctada margar- (calculated as 1 minus the squared multiple correlation ifera, and n = 7 in C, iradelei. For all 5 species, average between a predictor and the other predictors in a performance is compared, and the natural variation model). between measures is used to compare functional inter- relations between separate component processes of feeding physiology. In P. viridis and C. belcheri, the RESULTS range of food availability and number of measures were sufficient to identify influences of TPM on feed- Seston availability ing behaviour (refer to 'Results'). Where present data did not allow resolution of the exact form of relations, Throughout all 3 tidal cycles, natural seston avail- equations were fitted on the basis of a common set of ability measured as TPM (mg I-') varied from 6 to relations that have previously been shown to satisfac- 40 mg 1-l. The temporal pattern of variation in TPM torily predict responses observed in other suspension- was different on each day, and did not correlate (p > feeding bivalve molluscs (Hawkins et al. 1997). All sta- 0.05) with current speed, which ranged from 0.01 to tistical tests and procedures used in these analyses and 0.5 m S-'. Surface water temperature and salinity in the presentation of data were undertaken using throughout all measurements averaged (* maximum SYSTAT for Windows, Version 7 (SYSTAT Inc., Evans- range) 27.3 + 0.8OC and 27.9 & 1.5%0,respectively. ton, IL, USA). Variables were only included as addi- Combining the data from all measures of seston avail- tional predictors in multiple regression if the probabil- ability and quality, Fig. 1 illustrates how the OCS (frac- ity of their correlation with residuals was less than tion) decreased from a maximum of 0.22 to a minimum 2.5%. To further avoid constructing multicollinear of 0.09 with increasing TPM up to 40 mg I-', according Hawkins et al.. Suspension-feeding behaviour in tropical bivalves l??

to a negative relation that was best described by the 0.25 I I I I equation:

OCS = 0.25(0.18, 0.34) X TPM-O 2"*0.'4' where r2 = 0.24, residual df = 43, p = 0.001, and brack- eted values indicate 95 % confidence limits.

Average feeding behaviour

Average (* 2 SE) REP (fraction) within separate 2 mm size classes that averaged 4, G,8, 10, 12, 14 and 16 pm equivalent spherical diameter are illustrated for each species in Fig. 2. REPs of 0.25 to 0.35 for the smaller particles of less than about 8 mm equivalent spherical diameter were essentially similar in all 5 species (Fig. 2). However, REPs for the largest particle size class with an average diameter of 16 pm were signifi- cantly lower than for the smallest particles with an Fig. 1. Relationship between the organic content (OCS; frac- average diameter of 4 pm in Perna viridis (t = 2.51, tion) and total particulate mass (TPM; mg I-') of seston col- lected over 3 separate tides in the Merbok mangrove estuary. residual df = 13, p = 0.03), (t = Crassostrea belcher1 Malaysia. Both lines were fitted by least squares; the solid line 2.73, residual df = 25, p = 0.01) and Crassostrea iradelei to illustrated data from the Merbok mangrove estuary (refer (t= 7.08, residual df = 18, p = 0.000001). These reduc- to 'Results' for the equation), and the dashed line to data from tions appeared to begin for particles larger than about the bay of Marennes-Oleron, France [where OCS = 0.41(0.29, 12 pm diameter in P. viridis and C. belcheri, and for 0.56) X TPM-~~~"~.~~);from Hawkins et al. (1996)j particles larger than only about 6 pm diameter in C. lradelei (Fig. 2).No such reductions were apparent in Table 2 summarises average (k 2 SE) measures of feed- either S. cucculata or P. n?zrg~rifer;r(p > 0.05). ing physie!~.;;. ccmpute:! ever thc fu!! rar,gc sf scstcn

P. vir C. be1 C. irad - 0.51 I I I l - I I I I

PARTICLE DIAMETER(pm) PARTICLE DIAMETER(pm) PARTICLE DIAMETER(pm)

S. cucc P. marg

00.511

Fig. 2. Perna viridjs, Crassostrea belcheri. C. iradelej, Saccostrea cucculata and Pinctada mar- ganfera. Retention efficienc~es (%) for particles within size classes of 2 mm equivalent spherical diameter that ranged from 2 to 18 mm within the natural seston filtered by each shellfish species - - In the Merbok mangrove estuary, Malaysia. Data 2 6 10 14 18 2 6 10 14 18 represent the average (*2 SE) of between 8 and 31 PARTICLE DIAMETER(pm) PARTICLEDIAMETER (pm) replicate measures per species 178 Mar Ecol Prog Ser 166: 173- 185, 1998

Table 2. Average (i2 SE) clearance rates for particles of between 3 and 8 pm diameter (CR; l h-'), net organic retention efficien- cies (REO, fraction), net organic selection eff~ciencies(NOSE, fraction), net organlc ~ngestion rates (NOIR;mg organics g-' h") and net organic absorption rates (NOAR; mg organics g-' h-') recorded for each species. All data are standardised for shellfish of 1 g dry soft tissue weight

Species CR RE0 NOSE NOlR NOAR Perna viridis 7.2 i 3 1 +0.99 i 0.67 0.63 i 0.03 24.8 i 3.6 21.5 i 3.4 Crassostrea belcheri 4.6 i 1.0 +0.35 i 0.21 0.39 i 0.04 9.5 i 1.4 6.0 i 0.9 Crassostrea iradelei 4.1 15 +0.36 * 0.23 0.39 i 0.03 7.2 i 1.6 4.7 i 1.2 Saccostrea cucculata 4.9 * 1.2 +0.04 i 0.24 0.38 i 0.09 6.9 i 1.7 4.0 i 1 5 Pinctada marganfera 5.5 + 1.3 -0.11 ? 0.17 0.22 i 0.09 2.0 i 1.6 1.3 i 0.9

availability for each studied species. Compared with the 4 oyster species, Perna viridis showed the highest aver- age values for CR (l h-'), RE0 (fraction), NOSE (fraction), Perna viridis NOIR (mg organics h-') and NOAR (mg organics h-') (Table 2). In particular, comparisons of RE0 indicate that the organic content of matter filtered by P. viridis was vir- tually double that in the total available seston. In con- trast, there was no such enrichment in either Saccostrea cucculata or Pinctada margarifera, and although signif- icant in Crassostrea belcheri and Crassostrea iradelei, average percentage increases were only about one-third that in P. vindis (Table 2).Similarly, although all 4 spe- cies of oysters selectively enriched the organic content of ingested matter relative to filtered matter, their average NOSES were less than (0.39/0.63) X 100 = 62 % of that in P. vin'dis (Table 2). The result was that average NOAR in P. viridis was 3 to 5 times greater than in the Crassostrea and Saccostrea spp., and 16 times greater than in P. mar- Mytilus edulis garifera (Table 2). 1°-

Responses associated with changing seston availability

CR for particles of between 3 and 8 pm diameter (1 h-') in Perna viridis decreased from maxima of more than 15 1 h-' to minimal values of less than 5 1 h-' with the increasing availability of POM (mg I-') up to 2.5 mg l-l, as is illustrated in Fig. 3 according to a relation that was best described by the following allometric equa- tion: " 0 0 5 10 15 20 PARTICULATEORGANIC MATTER (mg I-') where r2 = 0.33, residual df = 22 and p = 0.003. Multi- ple stepwise regression indicated that neither TPM nor Fig. 3. Perna virjdis and Mytilus edulis. Short-term responses in clearance rate for particles of between 3 and 8 p diameter OCS explained any (p > 0.05) of the remaining vari- (CR; 1 h-') to the particulate organic matter within natural ses- ance in CR. No relation was evident between CR and ton (POM; mg 1-l) available to P. vMdis in the Merbok man- either TPM, OCS or POM in Crassostrea belcheri (p > grove estuary, Malaysia. The solid hewas fitted by least 0.05), despite being measured over the same ranges of squares (refer to 'Results' for equation). Data are also pre- sented for M.edulis [where CR = 4.2(3.2, 5.5) X POM~~"'*~~J; seston availability as in P. viridis. Nor was any such from Hawk~nset al. (1996)]. All data are standardised to 1 g relation evident in any of the 3 remaining oyster spe- dry soft tissue weight, and represent separate replicate mea- cies (p > 0.05). sures upon individual Hawluns et al.: Suspension-feeding behaviour in tropical bivalves 179

FR was the main determinant of pseudofaecal rejection rate (RR; total mg h-') as illustrated for Perna viridis and Crassostrea belcheri in Fig. 4 according to rela- tions that were best described by the following linear relations: Perna viridis RR = -33.5(?12.6) + 0.87(+0.10)FR where r2 = 0.97, residual df = 10 and p = 0.00000001; Crassostrea belcheri RR = -14.6(?9.6) + 0.76(*0.11)FR where r2 = 0.95, residual df = 10 and p = 0.00000006; ouL6 8 10 12 14 16 Crassostrea iradelei TOTAL PARTICULATE MATTER (mg 1.') RR = -11.0(+6.8) + 0.75(?0.11)FR where r2 = 0.97, residual df = 6 and p = 0.000008; 150 I I b Saccostrea cucculata RR = -9.7(*15.8) + 0.71(+0.24)FR where r2 = 0.89, residual df = 4 and p = 0.004; and Pinctada margarifera RR = -l.l(?4.3) +0.81(*0.14)FR where r2 = 0.97, residual df = 4 and p = 0.0003. These equations indicate that for every measured increment in FR, a minimum average of 71 % of the additional filtered matter was rejected as pseudofae- ces, and which averaged as much as 87 % of that addi- tional filtered matter in P. viridis. NOSE (frlcticr?) in Per-.?-. rririd~s, Crassostrea be!- chen and Saccostrea cucculata increased with OCS as is illustrated in Fig. 5 according to relations that were Fig. 4. Perna viridis and Crassostrea belched. Short-term responses in (a) filtration rate (FR; mg h-') to the total particu- late matter within available seston (TPM; mg 1-l), and (b) the relation between rejection rate (RR; mg h-') and FR in both P. viridis (0. -) and C. belcheri (A, - -) feeding upon natural seston in the Merbok mangrove estuary, Malaysia. Lines were fitted by least squares (refer to 'Results' for equations). All data are standardised to 1 g dry soft tissue weight, and represent the average for single measures upon separate replicate groups, each of 2 to 3 shellfish

Filtration rates (FR; total mg h-') in both Perna viridis and Crassostrea belcheE increased with TPM as is illustrated in Fig. 4 according to relations that were described by the following allometric equations (cf. Hawkins et al. 1997): ORGANIC CONTENT (fraction) Perna viridis FR = 13.8(?3.9) X TPMO91(*0.59) Fig. 5. Perna viridis, Crassostrea belcheri and Saccostrea cuc- where r2 = 0.49, residual df = 10 and p = 0.01; and culata. Short-term responses in net organic selection effi- ciency (NOSE; fraction) to the organic content of natural ses- Crassostrea belcheri ton (OFS, fraction) available to P. viridis (e,-), C. belcherj FR = 2.8(*2.4) X TPM1 43('0.38) (A, - -) and S. cucculata (U, - - -) in the Merbok mangrove where r2 = 0.85, residual df = 10 and p = 0.00002. estuary, Malaysia. Lines were fitted by least squares (refer to 'Results' for equations). All data are standardised to 1 g dry There was no indication that either REP or RE0var- soft tissue weight, and represent the average for single mea- ied with TPM, OCS or FR (p > 0.05). But in all 5 species, sures upon separate replicate groups, each oi 2 to 3 shellfish 180 Mar Ecol Prog Ser 166. 173-185, 1998 - described by the following hyperbolic equations (cf. Hawkins et al. 1997): Perna viridis NOSE = 1.01(*0.18) - [0.052(*0.024) X (1/OFS)] where r2 = 0.63, residual df = 10 and p = 0.002; Crassostrea belchen NOSE = 0.92(+0.31)- [0.073(+0.042)X (l/OFS)] where r2= 0.54, residual df = 10 and p = 0.006; and Saccostr-ea cuccula ta NOSE = 1.84(*0.90) - [0.24(+0.15)X (l/OFS)] where r2 = 0.72, residual df = 10 and p = 0.03. Multiple stepwise regression confirmed that FR did not explain a significant proportion of the remaining van- ance for NOSE in any of these 3 species (p > 0.05). Combining all data from each species, NAEIO (frac- Fig 7 Perna vlridls, Crassostrea belch er^, Crassostrea ira- delel, Saccostrea cucculata and P~nctadamargarifera. Short- tion) increased with the organic content of ingested term responses in net organic absorption rate (NOAR; mg h-') matter (OCI; fraction) in a strong common relation that to the total particulate mass (TPIM; mg I-') of natural seston was best described by the exponential equation: available to P v~ridis(.), C. belcheri (A), C. iradelel (O),S. cucculata (0) and P margarifera (X) feedlng upon natural ses- NAEIO = 0,89(+0,06) (1 - e-858i*336)10C1-0 13l*003)1) ton In the Merbok mangrove estuary, Malaysia. The line was where r' = 0.87, residual df = 40 and p = 0.00000003. fitted by- least squares to data foi C beicherl done (refer to 'Results' for equat~on)All data are standardised to 1 g dry soft This relation is illustrated in Fig. 6, showing how tissue weight, and represent the average for single measures NAEIO increased from as low as 0.3 to maximal values upon separate rephcate groups, each of 2 to 3 shellfish of more than 0.8 at the highest OCI.

Despite significant ranges of food availability, short- term responses in NOAR (mg h-') did not vary with TPM in either Perna viridis, Saccostrea cucculata or Pinctada margarifera (p > 0.05). Alternatively, as is illustrated in Fig. 7, a significant positive relation be- tween NOAR and TPM suggests that absorption would have increased at higher food availabilitles in Cras- sostrea belcheri.

DISCUSSION

The form and slope of the relation between OCS and TPM of seston available at our experimental site in the ORGANIC CONTENT (fraction) Merbok mangrove estuary were slmilar to those observed in the bay of Marennes-Oleron, France (refer Fig 6 Perna v~r~dis,Crassostrea belchen Crassostrea irade- le~,Saccostrea cucculata, Pinctada margar~fera and Mytilus to 'Results' and Fig. 1). This suggests 'dilution' of edulrs Short-term responses in net absorpt~oneff~ciency from organic matter by the silt that is typically resuspended Ingested organics (NAEIO, fractlon) to the organic content of within higher concentrations of natural seston (e.g. matter ingested (OCI, fractlon) by P vlndis (0) C belcher1 (A), Widdows et al. 1979).However, OCS at any given TPM C iradelel (0), S cucculata (U) and P marganfera (X) feeding in the Merbok estuary was almost half that observed in upon natural seston in the Merbok mangrove estuary, Malaysla The solid line was fltted by least squares to all data the bay of Marennes-Oleron (Fig. 1).We are not aware comblned (refer to 'Results' for equation) The dotted llne of other comparable data from tropical waters. ~llustrates the same relation for M edulis [NAEIO = Filter-feeding behaviour in bivalve molluscs is 1 15(*0 03) - 0 149(*0 004) x (l/OCI)] feeding upon natural highly flexible in response to changes in both the seston from the bay of Marennes-Oleron, France (from Hawk~nset a1 1997) All data are standardised to 1 g dry soit amount and composition of available seston (Hawluns tlssue welght and represent the average for slngle measures & Bayne 1992, Bayne 1998). In particular, feedlng rate upon separate replicate groups, each of 2 to 3 shellfish decreases in exponential relation with increasing con- Hawkins et al.: Suspension-fee,ding behaviour in tropical bivalves 181

centrations of POM, presumably because the gut approximately 3 pm in Perna virjdis, Crassostrea bel- becomes saturated with organics (e.g. Iglesias et al. cheri and Crassostrea iradelei (Fig. 2). Many previous 1992, Hawkins et al. 1997) (Fig. 3). This means that investigations of particle retention efficiency in a vari- high feeding rates are observed at low availabilities of ety of filter-feeding bivalve shellfish have reported that organically rich food as in artificial diets comprised REP generally increased to constant maxima at the mainly of cultured unicellular alga or algae (e.g. Msh- largest particle diameters (e.g. Hawkins & Bayne 1992, lenberg & Riisgard 1979, Riisgard 1988, Jorgensen Barille et al. 1993). However, it has also been observed 1996, Hawkins et al. 1997). But the reverse may occur that REP decreased with increasing particle diameter for natural seston of lower (<25%) OCS, for which above about 8 pm in the Placopecten rnagellan- feeding rate has been observed to increase to maximal icus from temperate latitudes (Cranford & Grant 1990, rates at higher concentrations (e.g.Iglesias et al. 1992, MacDonald & Ward 1994). MacDonald & Ward (1994) Newel1 & Shumway 1993, Hawkins et al. 1996), only suggested that this may have been an artefact caused decreasing upon clogging of the gills at the very great- by the lower number of particles in larger size classes. est loads (Widdows et al. 1979, Barille et al. 1997). This We do not agree that such an artefact occured in our is because OCS typically decreases with the increasing work here. Differences were statistically significant TPM of natural seston, due to 'diluting' effects of sus- (refer to 'Results'). In addition, independent measures pended silt. Further, unlike monocultures of alga, nat- of the average (+2 SE) RE0 (fraction) show that the ural seston affords the opportunity for particle selec- organic contents of matter filtered by P. viridis (0.99 + tion, which is most efficient at those maximal rates of 0.67), C. belcheri (0.35 t 0.21) and C. iradelei (0.36 + feeding (Hawkins et al. 1998). Given these selective 0.23) were all significantly higher than the organic processes, there is no guarantee that the same maxi- content of available seston. These are the same 3 spe- mal feeding rates occur when shellfish filter natural cies that showed reduced retention of the largest fil- particle mixtures as when they filter cultured alga or tered particle size class. Yet in both Saccostrea cuccu- algae. Itis therefore important that behaviour be com- lata and Pinctada margarifera, we observed neither pared under similar feeding conditions. In the present any reduced retention of larger particles nor any net study of responses to natural seston, average (t2 SE) organic enrichment of filtered matter. Most impor- CR (1 g-' h-') for particles of between 3 and 8 pm diain- tantly, these collective findings are consistent with past eter in all 5 species ranged from 4 1 t 1.5 to 7.2 + 3.1 ! s!udies ~stab!isf?ir?g tha! the smal!est suspended parti- g-' h-' (Table 2). Comparable averages across full cles have highest organic contents within natural ses- ranges of experimental food availabilities have not ton from both temperate and tropical environments exceeded 3.0 1 g-I h-' in all previous studies for a vari- (Defossey & Hawkins 1997), explaining how reduced ety of temperate species feeding upon natural seston, retention of larger particles may result in the greater including Cerastoderma edule (Navarro et al. 1994), proportional retention of organic matter. In a related Crassostrea gigas (Soletchnik et al. 1996), Mya are- phenomenon, Defossey & Hawkins (1997) reported dif- naria (Bacon et al. 1998), Mytilus edulis (Bayne et al. ferential size-dependent rejection of larger particles 1993, Newell & Shumway 1993, Hawkins et al. 1996), within pseudofaeces from the temperate filter-feeding Mytilus galloprovincialis (Navarro et al. 1996) and Pla- bivalves Mytilus edulis, Ruditapes philippinarurn and copecten rnagellanicus (Cranford & Gordon 1992, Tapes decussatus, pointing out how rejection observed MacDonald & Ward 1994, Bacon et al. 1998). On this on the basis of particle size alone may effect the basis, it is tempting to speculate that average CR may organic enrichment of natural filtered seston. It seems be faster in the tropical species studied here. But con- then, that larger particles may be relatively 'undesir- firmation of any such trend must await further compar- able', for which both a declining retention efficiency ison over a wider range of feeding conditions. Present on the gill and an increasing efficiency of rejection as measures did not include all feeding conditions experi- pseudofaeces may additively effect the pre-ingestive enced by each species in their natural environments. selection of organically rich matter. Although present experimental ranges of food avail- We suspect that methodological differences may ex- ability differed between some species (refer to 'Materi- plain the contrast between past reports that REP als and methods'), there are certain clear findings of increased to constant maxima with increasing particle interest. A key finding from the present study has been diameter, and our present findings in Perna viridis, that average REP (fraction) of approximately 3 to 8 pm Crassostrea belcheri and Crassostrea iradelei, includ- equivalent spherical diameter were essentially similar ing past findings in (Cran- in all 5 species, but that average retention efficiencies ford & Grant 1990, MacDonald & Ward 1994). Both our for the largest filtered particle size class with an aver- present work and that in P. magellanicus compared age diameter of 16 pm were significantly lower than particle-size distributions within the inflow and out- for the smallest particles with an average diameter of flow from separate trays that each contained 1 shell- 182 Mar Ecol Prog Ser 166. 173-185, 1998

fish, when hlgh throughflows ensured that test individ- mucus (Hawkins et al. 1996), or because algae are uals did not remove more than 30% of the particles more easily transported on ciliary currents to the suspended within previously unused seawater. In con- mouth, and may be proportionally more abundant than trast, other measures of particle retention efficiency in silt in seston of higher OCS (Jsrgensen 1996).In either filter-feeding bivalves have all monitored the relative case, we suspect that the lack of any signif~cantaddi- decline of different size-fractions whilst experimental tional influence of FR on NOSE may be ecologically individuals were maintained within static containers of relevant in these tropical specles. In temperate species, seawater, or with very slow throughflows (refer Banlle increased NOSE at faster FR is an important response et al. 1993 for summary of past literature). Under such which helps to maintain nutrient acquisition indepen- circumstances of limited throughflow, repeated re- dent of fluctuations in seston availability and composl- filtering of water may result in the ultimate removal of tion, because the general negative relatlon between all larger particles, and possible overestimation of the OCS and TPM (Fig. 1) means that filtered materlal is associated retention efficlencies. increasingly 'diluted' by silt as the concentrations of Negative influences of TPM have been recorded available seston increase to levels that often exceed upon retention efficiencies for small particles in tem- 200 mg TPM 1-' (Hawkins et al. 1996). This response perate species that include Crassostrea virginica may not be as important in tropical waters, where TPM (Loosanof & Engle 1947), Mytilus edulis (Davids 1964), is on average lower, and where seasonal fluctuations in Agopecten irradians (Palmer & Williams 1980), Ostrea seston availability are on average less extreme. edulis (Wilson 1983) and Crassostrea gigas (Barlle et The physiological consequences of NOSE were al. 1993). It is possible that the same may have amplified by a common positive exponential relation occurred in the present species over larger ranges of between NAEIO (fraction) and OCI (fraction) (Fig. 6). seston availability than were studied here. However, Similar relations have previously been observed within whether particle retention efficiency vanes within Cerastoderma edule, Crassostrea gigas and Mytllus and/or between species, our findings suggest that vari- edulis feeding upon natural seston in temperate ation is of both physiological and ecological impor- waters; the relation observed in M. edulis was simllar tance~. Compared with the available seston, the to the common relation observed here (Fig. 6) (Haw- organic content of filtered partlcles was increased by kins et al. 1998). The reason for these relations is averages of about 35% in Crassostrea belcheri and presently unclear. Hawkins et al. (1996) suggested that Crassostrea ir-adelei, and by an average of as much as dependence of NAEIO on OCI may stem from the 99% in Perna viridis (Table 2). Differential retention of metabolic faecal losses that are incurred through nor- filtered particles has been well established using flow mal digestive processes, and which appear reasonably cytometry in many shellfish species (e.g.Bougrier et al. constant per unit of total mass ingested (Hawkins et al. 1997, and references therein). For these reasons, it 1990), thereby exerting proportionally greater influ- may not be acceptable to calculate FR (mg h-') indi- ence upon net efficiencies of absorption from ingested rectly as the product of CR (1 h-') and food availability matter of increasingly lower organic contents. Alterna- measured either as TPM (mg 1-l) or POM (mg 1-l). tively, it has been proposed that lower efficiencies of In addition to establishing differential retention of organic selection may result in lower digestibility of filtered partlcles on the gill, we have shown that all 5 the organic matter ingested from available seston of species studied here were able selectively to enrich the low OCS (Jsrgensen 1996). organic content of ingested matter relative to filtered Resulting absorption and growth in Perna viridis matter, preferentially rejecting particles of higher inor- may have been close to their maxima across the full ganic content as pseudofaeces prior to ingestion. Previ- range of experimental seston availability, as indicated ous work In temperate species feeding upon the same by (1) the reduction in CR to values that approximated natural seston in the bay of Marennes-Oleron, France, zero when exposed to more than about 2.5 mg POM 1-' has shown that efficiencies of that selection (NOSE) in (Fig. 3), and (2) the lack of any clear change in NOAR the cockle Cerastoderma edule, the oyster Crassostrea over the full range of experimental TPM (Fig. 7). This gigas and the mussel Mytilus edulis each vaned posi- suggests that P. viridis was well-adapted to the ranges tively with both the organic content of available seston of seston abundance and quality experienced in the (OCS) and the mass of seston filtered h-' (FR) Merbok mangrove estuary, and may perform equally (Hawhns et al. 1998). Stepwise multiple regressions well in less turbid water with lower POM. However, in confirmed that NOSE increased in strong positive rela- Crassostrea belchen exposed to identical ranges of tions with OCS alone in Perna viridis, Crassostrea TPM and OC (refer to 'Materials and methods'), there belcheri and Saccostrea cucculata studied here. It is was no decline in CR (p > 0.05)and a linear increase in uncertain whether NOSE declines with OCS due to a NOAR with TPM (Fig. 7). Apparently, C. belchen relatively constant enrichment of pseudofaeces by would have ingested more and grown faster at hlgher l-lawkins et al.: Suspension-fee!din9 behaviour in tropical bivalves 183

food availabilities than were experienced here, as is istic of other clearer water, including coral reefs. Cer- consistent with the natural distribution of this fast- tainly, compared with other filter-feeding bivalve growing oyster in turbid waters (e.g. Yoo & Ryu 1984). shellfish from tropical waters, Pinctada species grow NOAR (mg organics g-' h-') represents the energy very slowly, with an annual increase in shell length of available for metabolism and production. The average less than 3 cm (Yoo et al. 1986, Numaguchi 1996). This NOAR of 21.5 mg organics h-' g-l that we report here was reflected by the average NOAR of 1.3 + 0.9 mg h-' in Perna viridis represents an organic ingestion rate of recorded in Pinctada margarifei-aof 1 g dry soft tissue nearly 52% of total dry soft tissue per day. Shafee here, and which was only 6% of the average NOAR (1979) reported that the energy deposited within tissue recorded in P. viridis (Table 2). CR in P. nlargarifera (= net growth or production) represented a maximum was comparable with the other studied species, but of 56% of the energy absorbed in P. viridis. Assuming with little pre-ingestive selection, as RE0 was insignif- that the energy content of absorbed organic matter icant, and NOSE only averaged 0.22 + 0.09 (Table 2). was 7.25 J mg-' (Hawkins et al. 1996), the resulting Significantly, Ward & MacDonald (1996) did not energy available for production in P. viridis of 1 g dry observe selective rejection of inorganic matter in the soft tissue may be calculated as 10.56 X (21.5 X 7.25)] = sub-tropical pearl oyster Pinctada imbricata, which 87 J h-'. Mussel tissue has an energy content of about appeared to produce pseudofaeces solely as a means 23.5 J mg-' (Slobodkin & Richman 1961), so that the of regulating ingestion rate. Our collective findings present estimate of average NOAR indicates potential therefore suggest that Pinctada spp. may not be well- net growth of about 9 % soft tissue d-' in P. viridis of 1 g adapted for the differential retention or rejection of dry soft tissue. Equivalent calculations based upon POM, which constitutes at least part of the physiologi- measures of NOAR in the temperate mussel Mytilus cal basis for comparatively slow growth. edulis indicated that maximalgrowth associated with In conclusion, we have shown that key interrelations apparent saturation of digestive processes represented between component processes of nutrient acquisition only 4.0% g-' dry soft tissue d-', which matched sea- in tropical bivalves were similar to those among tem- sonal maxima in the natural environment (Hawkins et perate species. Our findings also establish significant al. 1997). This contrast is consistent with observations differences in the feeding behaviour of functionally that shell length in P. viridismay increase by an aver- similar species from tropical latitudes. Although subtle, age of as rnrrch as 13 rnm me-' duenr; r;rewth frem these Ziffere~cesare likely to be sf ecolegica! ccnse- seed (<30 mm) to marketable size (>75 mm) (Kuri- quence, reflecting the importance of species diversity. akose 1980, Rangarajan & Narasimham 1980), com- Further studies over complete natural ranges of seston pared with a maximum of only 7 mm mo-' in M. edulis availability and composition are required before the and various other mussel species of comparable size development of models that predict relative impacts of from temperate waters (Hickman 1979). different bivalve filter-feeders upon particle fluxes, Fast growth in Perna viridisstems from a combina- nutrient dynamics and community structure. tion of interrelated differences in feeding behaviour. Maximal CR recorded in P. viridisat our lowest seston Acknowledgements. This work represents part of Strategic availabilities of POM < 1 mg 1-' averaged 13.2 k 7.0 1 Research Project 3 ('Marine Biodiversity') of the Plymouth h-' g-' (n = 9, k2 SE), which was more than double Marine Laboratory, a component of the UK Natural Environ- average maxima of less than 5 1 h' g-' recorded over ment Research Council, and was supported by the UK Darwin the same range of natural POM in the temperate mus- Initiative funding programme. We also thank The Fisheries Research Institute of Penang for use of their raft in the Mer- Mytilus edulis sel (e.g. Hawkins et al. 1996, 1998) bok mangrove system, Balasubramaniam s/o Kandasanly for (Flg. 3). Further, the average (+2 SE) RE0 of 0.99 * general experimental help, Yahya B. Hussin for help and hos- 0.67 in P. viridis was significantly greater than in all 4 pitality whilst working in the field, and 3 anonymous review oyster species (Table 2), and the average NOSE of 0.63 editors for their comments on the rnanuscnpt. + 0.03 in P. viridis was significantly higher than has previously been reported in any species to date (e.g. Hawkins et al. 1996, 1998, Iglesias et al. 1996, Soletch- LITERATURE ClTED nik et al. 1996, Bacon et al. 1998). Bacon GS. MacDonald BA. Ward JE (1998) Physiological As in Perna viridis,present findings suggest that responses of infaunal (Mya arenaria) and epifaunal (Pla- organic absorption and growth in both Saccostreacuc- copecten magellanicus) bivalves to variations in the con- culata and Pinctada margariferawere already maxi- centration and quality of suspended particles. I. Feeding mal at our lowest experimental seston concentrations activity and selection. J Exp Mar Biol Ecol 219:87-103 Bar~lleL. Heral M, Barille-Boyer AL (1997) Modelisation de S. cucculata of about 10 mg TPM 1-' (Fig. 7). is gener- l'ecophysiolog~ede l'huitre Crassostrea glgas dans un ally found on rocky or firm substrata in relatively open env~ronnementestuarien. Aquat Living Resour 10:31-48 waters, whereas Pinctada species are often character- Ban116 L, Prou J, Heral M, Bougrier S (1993) No influence of 184 Mar Ecol Prog Ser

food quality, but ration-dependent retention efficiencies in Biol 105: 197-204 the oyster Crassostrea gigas. J Exp marBiol Ecol 171 Hawkins AJS, Smith RFb1, Bayne BL, Heral M (1996) Novel 91-106 observations underlying fast growth of suspension-feed- Bayne BL (1998) The physiology of suspension feed~ngby ing shellfish in turbid environments: Mytilus eduhs Mar bivalve molluscs: an introduction to the Plymouth 'TRO- Ecol Prog Ser 131:179-190 PHEE' workshop. J Exp kIar 13iol Ecol219:l-19 Hawkins AJS, Smith RFh.1, Bougrier S, Bayne BL, Heral h,l Bayne BL, Hawkins AJS (19921 Ecological and physiological (1997) Manipulation of dietary conditions for maximal aspects of herbivory in benthic suspension-feeding mol- growth in mussels, bfytilus edulis L., from the Marennes- luscs. In: John DM, Hawkins SJ, Pnce JH (eds) Plant- OIeron Bay, Francp. Aquat Living Res 10:13-22 interaction In the marine benthos. Systematics Heral M (1993) Why carrying capacity models are useful tools Association Spec Vol No. 46, Clarendon Press, Oxford, for manauement of bivalve molluscs culture. In. Dame RF p 265-288 (ed) Bivalve f~lterfeeders Ln estuarine and coastal ecosys- Bayne BL, Iglesias JIP. Hawkins AJS. Navarro E, Heral M, tem processes. Springer-Verlag, Berlin, p 455-478 Deslous-Paoli JM (1993) Feeding behaviour of the mussel, Herman PMJ (1993) A set of models to investigate the role of lMytilus edulis L.; responses to variations in both quantity benthic suspension feeders in estuarine ecosystems. In: and organic content of seston. J Mar Biol Assoc UK 73. Dame RF (ed) Bivalve filter feeders in estuarine and 813-829 coastal ecosystem processes Springer-Verlag, Berlin, p Bayne BL, Newel1 RC (1983) Physiological energetics of 421-454 marine molluscs. In: Wilbur KM, Saleuddin AS (eds) The Hickman Rh7 (1979) Allornetry and growth of the green- , Vol4. Academic Press. New York, p 407-515 lipped mussel in New Zealand. Mar Biol Bougrier S, Hawkins AJS, Heral M (1997)Preingestive selec- 51:311-327 tion of different microalgal mixtures in Crassostrea gigas Iglesias JIP. Navarro E, Alvarez Jorna P, Armentia I (1992) and Mytilus edulis, analysed by flow cytometry. Aquacul- Feeding, particle selection and absorption in cockles ture 150:1.23-134 Cerastoderma edule (L.)exposed to variable conditions of Brohmanonda P, mut tar an sitK (1988) Oyster culture in Thai- food concentration and quality J Exp Mar Biol Ecol 162: land. In.: McCoy EW, Chongpeepien T (eds) Bivalve mol- 177-1 98 lusc culture research in Thailand. ICLARM (Int Cent Liv- Iglesias JIP, Urrutia ME, Navarro E, Alvarez-Jorna P, Lar- ing Aquat Resour Manage) Tech Rep 19:31-39 retxea X. Bougrier S, Heral M (1996) Variability of feeding Cranford PJ, Gordon DC Jr (1992) The influence of dilute clay processes in the cockle Cerastoderma edule (L.) in re- suspensions on sea scallop (Placopecten magellanicus) sponse to changes in seston concentration and composi- feeding activity and tissue growth. Neth J Sea Res 30 tion. J Exp Mar Biol Ecol 197:121-143 107-120 Jarayabhand P, Thavornyutlkarn M (1995) Realized heritabil- Cranford PJ, Grant J (1990) Particle clearance and absorption ity estimation on growth rate of oyster, Saccostrea cucul- of and detritus by the sea scallop Placo- lata Born, 1778. Aquaculture 138:lll-218 pecten rnagellanicus (Gmelin). J Exp Mar Biol Ecol 137: Jsrgensen CB (1996) Bivalve filter feedi.ng revisited. Mar Ecol 105-121 Prog Ser 142:287-302 Dame RF (1993) The role of bivalve material Kuriakose PS (1980) Open sea raft culture of green mussel at fluxes in estuarine ecosystems. In: Dame RF (ed) Bivalve Calicut. In: Coastal aquaculture: mussel farming, progress filter feeders In estuarine and coastal ecosystem pro- and prospects, Vol 29. Cent Mar Flsh Res Inst, Cochin, cesses. Springer-Verlag, Berlin, p 245-270 p 33-38 Davids C (1964) The influence of suspensions of micro-organ- Loosanoff VL, Engle JB (1947) Effect of different concentra- isms of different concentrations on the pumplng and tions of microorganisms on the feeding of oysters (0. vir- retention of food by the mussel (1Mytilus edulis L.). Neth J ginica). Fish Bull Fish Wild1 Serv US 513-57 Sea Res 2:233-249 MacDonald BA, Bacon GS, Ward JE (19981 Physiological Defossey JM. Hawkins AJS (1997) Selective feeding in shell- responses of infaunal (Mya arenaria) and epifaunal (Pla- fish: size-dependent rejection of large particles within copecten magellanicus) bivalves to variat~onsin the con- psuedofaeces from Mytilus edulis. Ruditapes philip- centration and quality of suspended particles. I1 Absorp- pinarum and Tapes decussatus. Mar Biol 129:139-148 tion efficiency an.d scope for growth. J lixp Mar Blol Ecol Doering PH, Oviatt CA (1986) Application of filtration rate 219:127-141 models to field populations of bivalves: an assessment MacDonald BA, Ward JE (1994) Variation in food quality and using experimental mesocosms. Mar Ecol Prog Ser 31: particle selectivity in the sea scallop Placopecten magel- 265-275 lanicus (Mollusca: ). Mar Ecol Prog Ser 108: Foster-Smlth RL (1975) The effect of concentration of suspen- 251-264 sion and inert material on the assimilation of algae by Mshlc,nberg F, WsgArd HU (1979) Filtration rate, using a new three bivalves J Mar Biol Assoc UK 55:411-418 indirect technique, in thirteen species of suspension-feed- Hawhns AJS, Bayne BL (1992) Physiological processes, and ing bivalves. Mar Biol 54:143-147 the regulation of production. In. Gosling E (ed)The mussel Navarro E, Iglesias JIP, Ortega MM (1992) Natural sediment Mytilus: ecology, physiology, genetics and culture. Else- as a food source for the cockle Cerastoderma edule (L.): vier Science Publishers BV, Amsterdam, p 171-222 effect of variable particle concentration on feeding, diges- Hawkins AJS. Bayne BL, Bougrier S, Heral M, Iglesias JIP. tion and the scope for growth. J Exp Mar Biol Ecol 156: Navarro E. Smith RPM, Urrutia MB (1998) Some general 69-87 relationships in comparing the fced~ngphysiology of sus- Navarro E, Iglesias JTP, Ortega MM, Larretxea X (1994) The pension-feeding bivalve molluscs J Exp Mar Blol Ecol basis for a functional response to variable food quantity 219:87-103 and quality in cockles Cerastoderma edule (Bivalvia, Hawhns AJS, Navarro E, Iglesias JIP (1990) Comparative Cardiidae). Physiol Zoo1 67:468-496 allometrics of gut content, gut passage time and metabolic Navarro E, Iglesias JIP, Perez Camacho A, Labarta U (1996) faecal loss in ~Mytilusedulis and Cerastoderrna edule. Mar The effect of diets of phytoplankton and suspended bot- Hawkins et al.. Suspension-feeding behaviour in tropical bivalves 185

tom material on feeding and absorption of raft mussels . Nature 191:299 (Mytilus galloprovincial~sLmk). J Exp Mar Biol Ecol 198: Smaal AC, Prins TC (1993) The uptake of organic matter and 175-189 the release of inorganic nutrients by bivalve suspension Navarro E, lglesias JIP, Urrutia MB, Parra J (1997) Simulating feeder beds. In: Dame RF (ed) Bivalve filter feeders in physiological responses of cockles (Cerastoderma edule) estuarine and coastal ecosystem processes. Springer- to variable conditions within estuarine media. Aquat Liv- Verlag, Berlin, p 271-298 lng Res 10:59-66 Soletchnik P, Goulletquer P, Heral M, Kazet D, Geairon P Newell CR, Shumway SE (1993) Crazing of natural particu- (1996) valuationdu bilan energetique de l'huitre creuse, lates by bivalve molluscs: a spatial and temporal perspec- Crassostrea gigas, en bale de Marennes-Oleron (France) tive. In: Dame RF (ed) Bivalve filter feeders in estuarine Aquat Living Res 9:65-73 and coastal ecosystem processes. Springer-Verlag, Berhn, Urrutia MB, lglesias JIP. Navarro E, Prou J (1996) Feeding p 85-148 and absorption in Cerastoderma edule under environmen- Numaguchi K (1996) A review on the feeding ecology and tal conditions in the bay of Marennes-Oleron (Western food environment of the Japanese pearl oyster, Pinctada France). J Mar Biol Assoc UK 76:431-450 fucata martensii. Bull Natl Res Inst Fish Sci 8:123-138 Ward JE, MacDonald BA (1996) Pre-ingestive feeding behav- Palmer RE, Williams RE (1980) Effect of particle concentration iours of two sub-tropical bivalves (Pinctada imbricata and on filtration efficiency of the bay scallop Argopecten irra- Arca zebra): responses to an acute increase in suspended dlans and the oyster Crassostrea virginica. Ophelia 19: sediment concentration. Bull Mar Sci 59:417-432 163-174 Widdows J, Fieth P, Worrall CM (1979) Relationships between Rangarajan K,Narasirnham KA (1980) Mussel farming on the seston, available food and feeding activity in the common east coast of India. In: Coastal aquaculture: mussel farm- mu.sse1 Mytilus edulis. Mar Biol 50:195-207 ing, progress and prospects. Vol 29. Cent Mar Fish Res Wildish DJ, Saulnier AM (1993) Hydrodynamic control of fil- Inst, Cochin, p 39-41 tration in Placopecten magellanicus. J Exp Mar Biol Er01 Riisgard HU (1988) Efficiency of particle retention and filtra- 174:65-82 tion rate in 6 species of northeast American bivalves. Mar Wilson JH (1983) Retention efficiency and pumping rate of Ecol Prog Ser 45217-223 in suspensions of Isochrysis galbana. Mar Shafee MS (1979) Ecological energy requirements of the Ecol Prog Ser 125-58 green mussel, Perna virid~sLinnaeus from Ennore estuary, Yoo SK. Chang YJ. Lim HS (1986) Growth comparison of Madras. Oceanologica Acta 259-74 pearl oyster, Pinctada fucata between two culturing areas. Sivalingam PM (1982) Explo~tationof shellfish mariculture In Bull Korean Fish Soc 19593-598 Malaysia: a highly productive marine protein resource for Yoo SE;, Ryu HY (1984) Comparative morphological charac- developing Third World countries. Atlantica 5:l l l teristics of mangrove oysters. Bull Korean Fish Soc 17: Slobodkin LB, Richman S (1961) Calories/gm in species of 321-326

Editorial responsibility: Otto Kinne (Editor), Submitted: November 20, 1997; Accepted: March 30, 1998 Oldendorf/Luhe, Germany Proofs received from author(s): May 15, 1998