We estimated In situ grazing rates of heterotrophic nanoplankton (HNAN) on bactenoplankton in a salt marsh estuary using a select~ve metabolic inhibitor technique. Seasonal experiments with prokaryohc inhibitors showed a significant, positive relation between bacterial abundance and HNAN grazing rates. On average, only 40 to 4 5 '/o of bacterioplankton produchon in the estuary appeared to be consumed by phagogrophic Protozoa <20 pm in size. Eukaryotic inhibitor experiments suggested that protozoan grazlng can have a positive feedback effect on bacterioplankton production, and that nitrogen regeneration by Protozoa may be of particular importance to bacterial growth. Although eukaryotic inhibitors did not prove to be rouhnely useful to estimate protozoan grazing rates on bacterioplankton, they may contribute to understanding qualitahve control mechanisms in the microbial food web.

Previous work has demonstrated selective feeding by bacterivorous protists on the basis of prey size. Other prey characteristics may also affect protist grazing rates. In this study, we investigated relative uptake rates of fluorescently labeled bacterial (FLB) cells made from growing (G-FLB) vs starving bacteria (S-FLB), and of live-stained motile vs non-motile bacteria by marine phagotrophic fla-gellates. We also determined relative growth rates and digestion times for flagellates fed growing or starved bacteria. A natural assemblage of marine flagellates and a flagellate isolate, Cafeteria sp., showed lower (p < 0.001) clearance rates on S-FLB than on G-FLB, which could be ascribed to size-selective grazing by the flagellates. Moreover, flagellates showed longer (p < 0.001) prey digestion tmes when feedng on S-FLB than on G-FLB. The carbon-based gross growth efficiency of flagellates fed starved bacteria was nearly twice that of flagellates fed growing bacteria: 38.5% and 21.5%, respectively. Uptake rates by flagellates of several live-stained bacteria (LSB) of either motile or non-motlle strains were compared with uptake rates of heat-killed FLB made from the same bacterial cultures. Clearance rates were significantly higher for LSB compared to FLB only for motile strains; no differences in uptake rates were observed for non-motile bacteria. Although FLB may be a good tracer for non-motile bacteria, use of FLB may lead to underestimation of actual grazing rates if motile bacteria comprise a significant fraction of the total bacterial assemblage. Our results suggest (1) grazing on starved bacteria results in a more efficient transfer of biomass to higher trophic levels than does grazing on growing bacteria; (2) use of motile LSB in prey uptake experiments should yield higher estimates of bacterivory compared to rates estimated using heat-killed or non-motile labeled prey; and (3) quah-tative factors affecting predator-prey dynamics may be important in regulating elemental fluxes in microbial food webs.

The quantitative significance of microbial heterotrophs in overall plankton metabolism was investigated in estuarine and coastal waters off the Georgia coast (USA). Three size classes were examined: < 208 pm, < 10 pm and < 1 p m Respiration was measured by monitoring the rate of oxygen consumption in water samples incubated at in situ temperatures in the dark. Rates by all size classes of plankton were highest in the upper reaches of the Duplin h v e r estuary during summer (52.4 pg O2 1-l h-') and lowest 10 km offshore during winter (10.1 pg O2 1-l h-'). Respiration by organisms passing a 1 pm screen exceeded rates in 208 pm and 10 pm filtered samples, but this enhancement decreased w~t h increasing distance offshore. This increased bacterial (< 1 pm) respiration could not be attributed to artifacts of the filtration procedure or to bacterial growth. Respiration results were consistent with concurrently conducted studies of bacterioplankton growth and bactivorous microprotozoan grazing. Static models of carbon and nitrogen flow between bacterioplankton and heterotrophic protozoa suggest that the C/N ratio of the bacterial food source could largely dictate organic carbon conversion efficiency of planktonic bacteria, the relative role of bacteria and microprotozoans as recyclers of nitrogen and the rate of organic carbon utilization. The relative importance of bacterial nitrogen remineralization decreases as the C/N ratio of the organic substrate increases. Results of this study are discussed in context of the microbial loop and the trophic flow leading to higher organisms.

Ability of natural assemblages of pelagic ciliates to ingest bacteria was tested using fluorescently labeled bacteria (FLB) prepared from in situ bacterioplankton. Ciliate bacterivory was analysed both in freshly collected samples from the mouth of Vlllefranche Bay, NW Mediterranean Sea, and in 50 pm screened water held in 201 plastic carboys in the laboratory. In various experiments, from 23 to 97 "h of the choreotrich assemblage ingested FLB added In tracer amounts (2 to 4 X 105 FLB ml-'), with average clearance rates ranging from 14 to 308 nl cell-' h-' Very little ingestion of FLB was observed for other types of ciliates, e.g. did~nids and hypotrichs, seen in the samples. Specific clearance rates (on the basis of FLB ingested) of individual morphological types of ciliates was on the order of 0 6 to 4 X 104 body volumes h-' for larger choreotrichs, and up to 2 X 10' body volun~es h-' for the smallest choreotrichs. We estimated that it would b e possible for choreotrichs < 15 pm in size to grow at a rate of about 0.5d-' on a n exclusive diet of bacteria at a concentrat~on of 10' bacteria ml-l, but that the larger cdiates were obtaining less than 15 % of their food rations a s bacteria. These data are the first direct evidence that some pelagic ciliates can be consumers of heterotrophic bacterioplankton in meso-to oligotrophic seawater.

We tested the grazing preference of natural assemblages and of isolated species of cihated protozoa for free or for particle-attached bacteria. Bacterivorous ciliates were obtained from tidal creek water and from the surface flocculent layer of Spartina salt marsh sediments. Free and particle-attached bacteria were prepared by growing a Gram-negative, rod-shaped bacterium either in a nutrient solution, or on the surfaces of artificially made alginate beads containing the same nutrients. The free and particle-attached bacteria were stained with a fluorescent dye and offered separately to ciliates in trace amounts (5 % of the natural bacterial abundance of 106 to 107 cells ml-l). Uptake of fluorescently labeled bacteria (FLB) by ciliates was monitored over 30 to 60 min penods. All cihates tested, except for a Euplotes sp, isolated from marsh floc, fed equally well or better on free FLB than on attached FLB. The marsh floc ciliates showed a stronger preference for attached bacteria then did tidal creek ciliates. Small spirotrichous chates and a free-swimming peritnch present in creek water ingested free FLB at rates 12 to 36 times greater than found for attached FLB. Our experimental results indicate that bacterivorous ciliates are capable of feeding on free bacteria at bacterial abundances less than 107 cells ml-l, and may not, as presently thought, be confined to grazlng attached or aggregated bactena in natural waters.