The concentration and composition of suspended particulate food available to estuarine suspension-feeding bivalve molluscs varies temporarily and spatially. Non-algal food sources may be important to suspension-feeders when algal concentrations are seasonally low or where there are high concentrations of suspended detrital material and bacteria, as found withln marshes. We carried out a series of laboratory experiments and field measurements to determine to what extent 2 common estuarine bivalve molluscs, the oyster Crassostrea virgin~ca and the ribbed mussel Geukensia demissa, could utilize cellulose and bacteria from Canary Creek marsh. Delaware, USA. Endogenously produced extracellular cellulases of the oyster depolymerized ingested cellulose to soluble oligomers. Subsequent intracellular cleavage of the oligomers to glucose was limited. The oyster absorbed carbon from refractory cellulosic material with an efficiency of only 3 %. In contrast, the ribbed mussel absorbed carbon from the same cellulosic material with an efficiency of 9 % and thls increased to 14 % if mussels were subjected to a 6 h exposure/6 h submergence cycle, a typical exposure regime for this intertidal species. We estimated that suspended cellulosic carbon in Canary Creek marsh during summer could supply 0.7 O/O and 8.6 % of the respiratory carbon requirements of subtidal oysters and intertidal mussels, respectively. In laboratory feeding experiments, colonization of refractory cellulosic food material by cellulolytic bacteria isolated from the marsh resulted in the oyster indirectly assimilating cellulosic carbon with an efficiency of 10 %. The oyster was able to filter free, unattached bacteria from suspension with an efficiency of only 5.0 %, compared with an efficiency of 15.8 % for the ribbed mussel. We estimated that both unattached and attached bacteria combined in Canary Creek marsh during summer provide only 5.5 % of the oysters' metabolic carbon requirements but could provide 31.0 % of an intertidal mussel's metabolic carbon requirements. Experiments with 1 5~ labelled bacteria indicated that attached bacteria associated with the breakdown of cellulosic material could mediate the flow of dissolved inorganic nitrogen from seawater to the oyster We estimated that unattached and attached bacteria in Canary Creek marsh during summer could contribute 26.7 % and 70.6 % of the metabolic nitrogen requirements of subtidal oysters and intertidal mussels, respectively. These results indicate that in thls marsh, utilization of bacteria as a food source could make a significant contribution during the summer to the nitrogen requirements of the oyster and to the carbon and nitrogen requirements of the mussel. However, cellulosic detritus and bacteria do not appear to fully meet the requirements of these bivalve species for carbon and nitrogen and utilization of other food sources is required, such as phytoplankton, nanozooplankton or non-cellulosic particulate and dissolved organic matter O Inter-Research/

Using an artificial diet, composed of silt < 32 p in diameter and the alga Tetraselmis suecica, it was demonstrated that the oyster Crassostrea virginica could significantly reduce the concentration of algae voided in the pseudofaeces (measured as extracted chlorophyll pigment) by over 50 %, compared to levels in the food. More importantly, it was also shown that for C. virginica fed natural seston at concentrations between 4 to 20 mg I-', the proportion of energy, carbon and nitrogen voided in the pseudofaeces could also be reduced significantly compared to that in the food. As the organic material in natural seston is from a wide range of sources -e.g. phytoplankton of different sizes, bacteria, detritus particles, etc. -these results indicate that C. virginica has a well developed ability to ingest preferentially various types of organic material and to reject other particles as pseudofaeces. This discriminatory mechanism must be able to operate on individual particles despite the fact that they are bound in viscous mucus. We hypothesise (based on literature information for the properties of molluscan mucus) that the viscosity of the mucus in which the food particles are entrapped may be significantly reduced by the ciliary action on the ridged surfaces of the opposed labial palps. This reduced viscosity mucus is possibly moved to the free edge of the palp where, with a cessation of the mechanical stimulation, it regains its original viscosity. The individual particles may then be subject to chemical testing by chemoreceptors which determine whether a particle is moved over the palp ridge to the mouth or is admitted to the deep rejection tracts. These rejected particles move to the free edge of the palps where they are re-incorporated in the mucus and rejected as pseudofaeces.

We I4C-labeled cellulolytic bacteria and 3 species of bacterivorous nanoflagellates and fed these cultured organisms to 2 species of intertidal mussels, Geukensia demissa and Mytllus edulis. using a pulse-chase experimental design under controlled laboratory conditions. Ingestion and assimilation of C from these rmcroheterotrophs by mussels were calculated from measured rates of defecation, respiration, excretion, and tlssue incorporation. The proportion of available C ingested by G. demissa did not dlffer significantly among bacteria (39%), heterotrophic flagellates (58%), or the unicellular algae Isochrysis galbana (66%), which was used as a reference diet. In contrast, M. edulis ingested a signlflcantly lower proportion (19%) of the small bacterla ( < l pm in diameter) than the larger ( 3 to 5 pm diameter) heterotrophic flagellates (58%). The efficiency with which G. dernissa assimilated C from I. galbana (77 %) was significantly greater than that from either bacteria (42 %) or heterotrophic flagellates (44%) M edulis assimilated bacterial C with significantly lower efficiency (21 96) than C from heterotrophic flagellates (62%) These results indicate that heterotrophic flagellates can contribute to the C requirements of both G. dernlssa and M edulis, however, only G. demissa is capable of assimilating a considerable amount of C from bactena.

The tolerance ofCrassostrea virginica larvae to anoxia increases with developmental stage and body size. Median mortality times range from 11 h for prodis soconch larvae of82 zm (shelllength) to 5 1 h for pedivel iger larvae of 3 12 zm, and 150 h forjuvenile oysters. Si multaneous calorimetry and respirometry showed that in response to declining oxygen tension (P02), the rates of heat dissipation and oxygen uptake by oyster larvae are maintained independent ofP@ down to low P@ values (2 kPa for prodissoconch larvae and 8 kPa for pediveliger andjuveniles). Therefore, total energy metabolism is sus tamed mainly by aerobic metabolism down to 2 and 4 kPa for early larval stages and juveniles, respectively. Prodissoconch larvae maintain relatively high rates of heat dissipation under anoxic conditions (34% of nor moxic rate), whereas pediveliger and juveniles lower their anoxic rates of heat dissipation to 3% of the nor moxic rate. The ability to reduce rates ofheat dissipation and thus conserve energy expenditure under anoxia ap pears to be related to the increase in anoxia tolerance with larval development. The larval differences in the re lationship between P@ and the rate of heat dissipation are also reflected in feeding rate (ingestion rate of micro spheres). Prodissoconch larvae maintain feeding activity under anoxic conditions (29% of normoxic ingestion rate), in contrast to pediveigerlarvae, which lower inges tion rates to 5% ofthe normoxic rate.

Suspension-feeding processes in the eastern oyster Crassostrea virginica (Gmelin, 1791) were examined, in vivo, with an endoscope linked to a video image-analysis system. We found that many of the previously published concepts of particle transport and processing in this species, obtained using surgically altered specimens or isolated organs, are incomplete or inaccurate. In particular, our observations demonstrate that (1) captured particles are transported along the gills by both mucociliary (marginal grooves) and hydrodynamic (basal tracts) processes; (2) the labial palps accept material from the gills both in mucus-bound particle strings (transported in marginal grooves), and suspended in particle slurries (transported in basal tracts); (3) the labial palps reduce the cohesive integrity of the mucous strings and disperse and sort the entrapped particles; (4) particles are ingested in the form of a slurry; and (5) ciliary activity on the labial palps is independent of that on the lips, allowing the oyster to filter particles from suspension and produce pseudofeces without ingesting any particulate matter. Because many ostreids have the same plicate gill structure, we believe that our conclusions are applicable to other oyster species. In addition, the present observations are consistent with other endoscopic examinations recently made on bivalves in different families. We conclude that accepted theories of particle handling in suspension-feeding bivalve mollusks must be modified to accommodate observations made with the endoscope.

ABSTRACT. We studied the influence of the protozoan parasite P e r h s u s marinus (Mackin. Owen & Collier) on reproduction in the eastern oyster Crassostrea virginica (Gmelin) from 2 sub-estuaries of central Chesapeake Bay, Maryland, in 1990 and 1991 One hypothesis was that increased levels of infection by P marinus cvould result in smaller egg size and diminished lipid reserves within individual eggs. Our results did not reveal any consistent deleterious effects of l? marinus on these gametogenic characteristics, although some of our sample sizes were small and the average disease intensities were low. Our second hypothesis, that there would be a decllne in reproductive output as a result of high levels of P marinus infection, was supported for one estuary when those oysters were near their peak of reproductive condition (July). We found no evidence to support our third hypothesis that there would be a negative influence of infections during the 1990 reproductive season on subsequent reproductive output (summer 1991). Based on published evidence that P marinus can inhibit oyster growth, we speculate that oysters infected at low intensities by P marinus can divert energy from growth to reproduction. This differential energy partitioning allows infected oysters to produce fewer eggs than uninfected oysters but with no change in the amount of lipid reserves stored in each egg.

In order to better understand the particle diet of planktotrophic larvae of the eastern oyster Crassostrea virginica (Gmelin) we measured their ingest~on of naturally occurring food organisms. By using a dual radioisotope (3H and I4C) label~ng technique in conjunction w~t h plankton size fractionation procedures we demonstrate that oyster larvae feed upon bacteria, phagotrophic protozoans and phototrophs present in the diverse summer plankton assemblages of Chesapeake Bay, USA. Prodissoconch I1 oyster larvae cleared 0.2 to 30 pm I4C-labeled plankton (primarily phototrophs) at a rate of 0.0825 n~l larva-' h-', and 0.2 to 30 pm 3H-labeled plankton (heterotrophlc bacteria and phagotrophic protozoans) at a rate of 0.0017 m1 larva-' h-' This calculated clearance rate for 0.2 to 30 pm heterotrophs was low due to the predominance of small (0.2 to 0.8 pm), poorly retained bacteria in thls size class. Oyster larvae consumed a wide size range of food part~cles (0 2-0.8 pm to 20-30 pm) and selectively ingested 20 to 30 pm organisms. In other feeding experiments, oyster larvae cleared laboratory cultured heterotrophic flagellates (12 pm) at a rate of 0.0640 m1 larva-' h-' and cultured heterotrophic ciliates (12 X 20 pm) at a rate of 0.1093 m1 larva-! h" The inclusion of heterotrophic food organisms In the diet of C. virginlca may enhance its growth and development by providing energy and nutrients that supplement those of ingested phytoplankton. We suggest that because oyster larvae ingest non-phytoplankton cells, estimates of standing stocks of phytoplankton may not always b e a reliable measure of food supply

We examined the feeding activity of planktotrophic larvae of the eastern oyster Crassostrea virginica on different size particles suspended in surface waters of 2 subestuaries of Chesapeake Bay, USA. Particle suspensions were characterized uslng particle counters and microscopic observations. At both sites, particle assemblages were dominated (in terms of particle number and volume) by particles (predominantly plankton cells) with equivalent spherical diameters <5 pm. Feeding experiments demonstrated that small (200 pm) oyster larvae typically ingested particles between 0.5 and 12 pm. However. in the presence of blooms of large (10 to 30 pm) dinoflagellates, small and large larvae extended their maximum ingested particle size to about 16 and 30 pm, respectively. These results indicate that oyster larvae generally exploit foods within the characteristic particle biomass peak present in these estuaries and opportunistically feed upon periodic blooms of large plankton cells. Larvae derived a large percentage (range 20 to 90%) of the total ingested particle volume from picoplankton-size (here defined as 0.5 to 3 pm) particles. These small particles constituted a larger percentage of the material ingested by small larvae than that ingested by large larvae. By comparison, both sizes of larvae derived about 2 to 4 0 % and about 5 to 20% of total ingested food volume from 3 to 5 pm and 5 to 10 pm particles, respectively. In experiments with abundant dinoflagellates > l 0 pm in size, both sizes of larvae derived 20 to 30% of their total ingested volume from these particles.