MARINE ECOLOGY PROGRESS SERIES Published August 13 Mar. Ecol. Prog. Ser. I

Seasonal differences in the content of Oikopleura vanhoeffeni and finmarchicus faecal pellets: illustrations of shifts in coastal Newfoundland waters

Juanita L. Cynthia H. ~cKenzie',Don ~eibel'l~

'Ocean Sciences Centre, 'Department of Biology, Memorial University of Newfoundland, St. John's. Newfoundland, Canada A1C 5S7

ABSTRACT: Faecal pellet contents from Oikopleura vanhoeffeni and Calanus finmarchicus were examined in the spring, summer, fall and winter. Contents varied seasonally, resembling available particles in the water column. This indicated a temporal succession from a -based food chain in the winter and spring to one based on the microbial loop in the summer and fall. Bacteria, cyanobacteria, Anabaena sp., choanoflagellates, , and heterotrophic were found in the faeces, suggesting a direct link between the microbial loop and both and pelagic tunicates.

INTRODUCTION Primary production in many oceanic and coastal areas is dominated annually by nanoplankton (2 to 20 pm) and Primary production by and bacteria picoplankton (<2 pm) (Booth et al. 1982, Murphy & forms the basis of all aquatic food webs. Composition Haugen 1985, Hoepffner & Haas 1990).It is known that of the phytoplankton community in coastal, temperate and euphausiids are not able to feed efficiently waters varies seasonally, from a predominance of on particles < 10 pm in size (Marshall & Orr 1955, Nival large, chain-forming in the wlnter and spring & Nival 1976, Fenchel 1988), and that protozoan to dinoflagellates, protozoans and nanoplankton in the consumers, such as ciliates and choanoflagellates, may summer and early fall (Smayda 1980, Marshal1 & Cohn consume 4 to 70 % of the annual algal production (refer- 1983, Marshal1 1984, Veldhuis et al. 1988, Nielsen & ences in Burkill et al. 1987). These observations suggest Richardson 1989). This results in a succession from a that most of the primary production is recycled in the diatom-based food chain in the winter and spring to upper mixed layer (i.e.through the microbial loop), and one based on the 'microbial loop' in the summer and that only large phytoplankton > 20 pm in size contribute fall. However, interactions between phytoplankton, to nutrient flux out of the euphotic zone. However, bacteria, protozoans and zooplankton are not well mucous-net feeders, such as salps, doliolids and appen- understood. How and to what extent are microbial dicularians, ingest nanoplankton and picoplankton food webs linked to metazoans? Are small plankton efficiently (Harbison & McAlister 1979, Mullin 1983, energy links (i.e. prey) or sinks (i.e. decomposers)? Deibel & Lee 1992), thereby creating a means through Does the nature of this linkage vary seasonally in faeces of transporting the organisms out of the euphotic coastal Newfoundland, Canada, waters that are domi- zone (Michaels & Silver 1988). nated by the sub-arctic Labrador current? The purpose of this study was to examine, using scan- ning electron microscopy, how phytoplankton content of faeces from Calanus finmarchicus, a raptorial, suspen- ' Present address: Horn Point Environmental Laboratory, sion-feeding , and Oikopleura vanhoeffeni, a University of Maryland, Cambridge, Maryland 21613, USA pelagic tunicate that feeds with a mucous filter, varies

O Inter-Research 1992 256 Mar. Ecol. Prog. Ser. 84: 255-264, 1992

seasonally. This paper reports on a descriptive study, were examined at 3000X using SEM. 3000 to 5000 ilustrating seasonal changes in faecal pellet contents. fields were scanned on 0, vanhoeffeni pellets (average The content of the faeces determines their primary den- pellet area was 9.73 X 105 pm2),and 500 to 700 fields sity and nutrient content (Urban et al. in press). Also, it were scanned on C. finmarchicus pellets (average provides information on the structure of the zooplankton pellet area was 1.42 X 105 pm2). Photomicrographs food web and on the seasonally changing role of differ- were made of all phytoplankton species that appeared ent organisms in energy flow. in > 20 % of the fields and of those that were unique or rare. Intact cells were measured to the nearest 0.1 pm, and were identified to genus and/or species when MATERIALS AND METHODS possible. Taxa that appeared in > 50 % of the fields were tabulated as 'abundant', those that appeared in Oikopleura vanhoeffeni was collected in 400 m1 20 to 50 % as 'common', and those that appeared in glass jars by SCUBA divers in March, June and < 20 % 'rare'. November 1990, and January 1991, from Logy Bay, When Calanus finmarchicus was collected, 180 m1 insular Newfoundland, Canada. Immediately after samples were taken from the SCM for floristic analysis. collection the jars were placed in a flow-through sea- Cell samples were stained with Lugol's iodine and water bath in the laboratory. After a period of l to 12 h fixed in 5 % borate-buffered formaldehyde in sea- the houses were shaken to release pellets trapped water. l0 to 50 m1 samples were allowed to settle inside. The house and were then rem.oved and for 48 h, followed by identification and c0untin.g the remaining water was poured through a 20 pm nitex at 20 and 40X using a Zeiss Axiovert 35 inverted filter to collect pellets. microscope with phase contrast. The whole slide was Fresh, intact pellets were removed from the filter viewed at 20X and all cells >20 pm were counted, and pipetted into a beaker with distilled water. This except in April where 2 across-diameter tracks water containing the pellets was then passed through a allowed for more than 200 cells to be counted. Two 0.45 pm nylon filter (13 mm in diameter, Micron Sepa- across-diameter tracks were counted at 40X for cells rations Inc., Westboro, MA, USA ). The filter with at- < 20 pm, with a total of 2 200 cells counted per sample tached pellets was put through an ethanol dehydration (Utermohl 1958). series (50 %, 70 %, 85 %, 95 %, 100 %), and critical- point dried (Turner 1984). After critical-point d.rying, 50 % of the pellets were pulled in half using double- RESULTS sided tape before being mounted and coated with gold. Pellet contents were viewed on a Hitachi S 570 A complete species list of the contents identified scanning electron microscope (SEM) with the upper from the faecal pellets of Calanus finmarchicus and detector set at 15 kV and an 8 mm working distance. Oikopleura vanhoeffeni is given in Tables 1 & 2. The Calanus finrnarchicus was collected from Conception contents for both faecal pellet types followed similar Bay, Newfoundland, in April, July and September seasonal variations. Direct comparisons between the 1990, from oblique hauls with a 0.5 m plankton net contents of the faeces for the 2 species of zooplankters fitted with 350 pm nylon mesh (Nitex) and a closed cannot be made since they were collected at different end cup. The net wa.s lowered to 200 m, towed for times and from different sites. 1 min, and then retrieved at a rate of 13 m min-' Cod end contents were emptied into a bucket containing Spring water from the subsurface maximum (SCM). Within 3 h, adult C. finmarchicus females were In March, Oikopleura vanhoeffeni faeces were filled removed from the bucket and placed in a large PVC with diatoms (Fig. 1) and chrysophytes (Fig. 2). Dino- cup with a bottom of 505 pm nitex mesh. The cup was flagellates, silicoflagellates, coccolithophonds, choano- suspended in a bucket containing water from the SCM flagell.ates, cillates and bacteria were 'rare' Many cells and incubated in the laboratory at ambient seawater were intact and diatom chains were observed. Cells temperature (-1 to 10 "C) for 1 to 12 h. At the end of the ranged from 0.7 to 183 pm in least dimension in valve incu.bation period the PVC cup was removed from the view, with most c 10 Irn in size. bucket and the water was passed through 20 pm nitex In April, broken diatom frustules (Fig. 3) dominated mesh to collect all faecal pellets. Pellets were removed the contents of Calanus finmarchicus faeces, includ- from the filter by pipet and prepared for SEM as above ing Skeletonema costatu.m (Fig. 4),Chaetoceros spp., for Oikopleura vanhoeffeni. Thalassiosira spp. and Fragillanopsis spp. Intact cells In each season, 6 to 8 faecal pellets from both ranged from 5 to 18 pm in least dimension. Chryso- Calanus finmarchicus and Oikopleura vanhoeffeni phyte and remains were 'rare' Urban et al.: Seasonal faecal content changes of two zooplankters 257

Table 1. Oikopleura vanhoeffeni. Phytoplankton content of Table 2. Calanus fjnrnarchicus. Phytoplankton content of faecal pellets. a: Abundant, found in >50% of f~eldsviewed; faecal pellets, a: Abundant, found In >50 of fields viewed; c: common, found in 20 to 50%; r: rare, found In <20%. c: common, found in 20 to 50 %; r: rare, found in <20 'X, ' Fragments identified by plate or surface morphology of fields viewed. ' Fragments ident~fiedby plate or surface morphology Species Spring Summer Fall Winter Species Spr~ng Summer Fall Bacillariophyceae Amphora spp. r r Bacillariophyceae Arcocellulus cornucervis a a Amphora spp. r Bacterosira ragelis Chaetoceros spp. Berkeleya sp. Cocconeis spp. Chaetoceros spp. a Minidiscus trioculatus Cocconeis spp. C Navicula spp. Coscinodiscus radiatus Nitzschia sp. Cyclotella spp. Skeletonema costatum Fragillariopsis spp. a Thalassiosira spp. F. curvata 7. gra vida F. cylindrus T. minima Gomphoneopis littoralis T. pseudonana Gomphosepta turn T symmetrica aestuarrii Licmophora sp. C r Pyrrhophyceae Minidiscus trioculatus a C ' Ceratium sp Navicula spp. C C Gymnodinjum sp Nitzschia sp. 1- Prorocentrum spp. Pleurosigma sp. r P. balticum Pseudogomphonema C P. minima kamtschaticum Chrysophyceae Pteroncola spp. C Tetraparma sp Skeletonema costa tum a a Tabularia sp. r Craspedophyceae 'Lorica bundles I Thalassionema C nitzschioides Others Thalassiosira spp. C Bacteria I T. gravida Ciliates J T.minima Cyanobacteria r C T.pseudonana Cysts r a T symmetrica Pyrrhophyceae ' Ceratium sp. The water sample was dominated by diatoms, pri- Gymnodinium sp. Prorocentrum spp. C marily Thalassjosjra spp. and Chaetoceros spp. (Table P. balticum 3). Unidentifiable spheres <2 pm in diameter were P. minima Protoperidinium spp. copious, followed by choanoflagellates and chryso- P. depressum phytes. Ciliates and dinoflagellates were scarce. Total Scrippsiella sp. cell number was 1.58 X 106 1-'. Chrysophyceae Dictyocha r speculum Summer Tetraparma sp. r Triparma stiga ta r Xanthophyceae Broken and unidentifiable diatom frustules domi- Meringosphaera C nated faecal pellet contents of Oikopleura vanhoeffeni mediterranea in the end of May and beginning of June (Fig. 5). Haptophyceae Individual cells ranged from 1 to 167 pm in least Coccolithus pelagicus r Emiliana huxeleyi a dimension in valve view. Dinoflagellate fragments Craspedophyceae were 'common' (Figs. 6 & 7). Parvicorbicula sp. In the end of June and early July, the pellet contents of Stephanoeca sp. Calanus finmarchicus were diverse. Broken diatom 'Lorica bundles a frustules were 'abundant' (Fig. 5),but bacteria, choano- Tintinnina Stenosemella steini flagellates (Fig. 8), cyanobacteria (Anabaena sp.), dino- Others flagellates (Fig. 6), and diatom cysts were 'common'. Bacteria C C Intact cells ranged from 1 to 44 pm in least dimension. Ciliates C r Particles in the water were dominated by macro- Cysts a r Euglenoids C C scopic flocs of 'marine snow' (Table 3). Few intact diatoms were present. Spheres <2 pin in diameter 258 Mar. Ecol. Prog. Ser. 84: 255-264, 1992

Figs. 1 to 4. Oikopleura vanhoeffeni and Calanus ~~~llnarchich~.~pring faecal content. Fig. 1. 0. vanhoeffeni. Overview of faecal pellet. F: Fragillariopsis sp. chain; CO:coccolithophore; S: Skeletonema costatum; T: 73alassiosira sp. Scale bar = 20 pm. Fig. 2. Tetraparma sp., common chrysophyte. Scale bar = 1 pm. Fig. 3. C. finmarchicus. Overview of faecal pellet. C: Chaeto- ceros sp.; S: Skeletonema sp. Scale bar = 5 Wm. Fig. 4. Skeletonema costatum, a dominant diatom. Scale bar = 1 pm

were the most prevalent live organisms. Gymnodinium nated the faecal pellets of Oikopleura vanhoeffeni. spp., choanoflagellates and ciliates were present. Total Dinoflagellates (Fig. 12), silicoflagellates (Fig. 13), cell number was 1.91 X lo5 I-'. euglenoids (Fig. 14), chrysophytes, xanthophytes (Fig. 15), diatoms and coccolithophorids (Fig. 16) were 'abundant', with many partially broken cells. Intact Fall cells ranged from 0.9 to 57 pm in least dimension.

In September, Calanus finmarchicus faeces were dominated by nanoplankton, including ciliates (Fig. g), Winter bacteria, cyanobactena (Fig. 10), choanoflagellates (Fig. ll), unidentifiable cysts, cryptomonads, dino- Diatoms, including Skeletonema costatum, Chaeto- flagellates and chrysophytes. Cells ranged from 0.5 to ceros spp. (Fig. 17) and Arcoce1lulus cornucervis 87 ,pm in least dimension. Small diatoms were 'rare'. (Fig. 18),and coccolithophorids (Fig. 19) were the main Chrysophytes and unidentifiable spheres <2 lm in components in the faecal pellets of Oikopleura vanhoef- diameter dominated the cells in the SCM (Table 3). feniin January. Chrysophytes, dinoflagellates, choano- Small (< 15 lm)and large (> 20 pm) ciliates were plentiful, flagellates and ciliates were 'common' (Fig. 20). Small as were choanoflagellates, gymnodinoid dinoflagellates cells appeared to be intact but many larger cells and and marine snow. Total cell abundance was 3.13 X lo5 1-'. chain-forming diatoms were cracked. Intact cells In November, diverse nanoplankton species domi- ranged from 1 to 66 pm in least dimension. Urban et a1 Seasonal faecal content changes of two zooplankters 259

Table 3 Cell counts (cells I-') and % of total m water samples from Conception Bay. 'Total does not include marine snow, " none present in sample

Plankton category Sprlng, 6 Apr 1990 Summer, 26 Jun 1990 Fall, 27 Sep 1990

Total cell count 1.58 X lob' 1.91 X 10'' 3 13 X 10" < 2 pm spheres 2.95 X 105 (19 ':;,) 1.07 X 105 (56 X) 824x10'' (28'.)

2-5 pm sphcres 3.23 X lo4 (2 ,J .. . .

> 5 pm spheres 8 08 X 10' (0.5 ' .I) 1.21 X 104 (6 %) 1.03 X 10"36 ' .l

Chrysophytes 6 46 X 104 (4 ',#) 1 45 X 104 (8 ,m) 4.85 X 104 (17 l. ,J

Choanoflagellates 6 26 X 104 (4 ",I) 1 06 X 104 (6 ,'U) 4.04 X 10" (l '!#I)

Cryptophytes 2 22 X 104 (l <',I) 5.39 X 103 (3 4.04 X 10"l " )) Cyanobacteria 2.02 X 103 (0.1 1.62 X 103 (0.8 %) 4.04 X lo3 [l 5,) Phaeocyst~ssp 2.50 X lo3 (0 2 %j . . . . Praslnophytes 8.08 X 10' (0.5S;) 1.62 X 103 (0.8 X,) 8.08 X 10' (0.2 ',,)

Prymnesiophytes 2.42 X 104 (2 %) 6.47 X lo3 (3 %) 4.85 X 103 (2 ",I) < 20 pm ciliates .. 8.08 X 10' (0.4 ",) 8.08 X 10-3 1,)

> 20 pm ciliates 1.41 X 104 (0.9 %) 1.24 X 103 (0.6 2.00 X 10' (0.06"(I) Tintinnids .. 8.00 X 10' (0.04 %) 2.00 X 10' (0 006 ",8) Diatoms 9.96 X 105 (63 %) 5.55 X lo3 (3 3;) 1.05 X 104 (4 ':L) Unarmored dlnoflagellates 4.24 X 104 (3 2.10 X lo4 (11 1 62 X 1O4 (6 'L,) Armored dinoflagellates 1.01 X lo4 (0.6 "#,) 2.42 X 103 ( l "U) 3.33 X lo3 ( 1 %) Marine snow .. 1 10 X 105 2.34 X 104

Figs. 5 to 8. Ojkopleura vanhoeffeni and Calanus fjnmarchlcus. Summer faecal content Fig. 5. Broken diatoms, common to both faeces types. Scale bar = 2 pm. Fig. 6. Prorocentrum sp., d~noflagellate.Scale bar = 2 pm. Fig 7. Protopend~n~umdepressum, dinoflagellate (arrow). Scale bar = 2 pm. Fig. 8. B: bacteria, Ch: choanoflagellate remains. Scale bar = 2 pm

Urban et al.:Seasonal faecal content changes of two zooplankters 261

Figs. 17 to 20. Oikopleura vanhoeffeni. Winter faecal contents. Fig. 17. 0. vanhoeffenifaecal pellet overview. C: Chaetoceros sp.; N: Navicula sp.; Ch: choanoflagellate lorica. Scale bar = 5 pm- Fig. 18. Arcocellulus cornucervis, common diatom. Scale bar = 2 pm. Fig. 19. Coccolithophore with a cyst (Cy).Scale bar = 2 pm. Fig. 20. P: Prorocentrum sp., dinoflagellate; Ch: choano- flagellate remains. Scale bar = 2 pm

DISCUSSION components of the microplankton and nanoplankton communities (Stoecker et al. 1989, Paranjape 1990). In Causes of seasonal differences in faecal pellet our samples, unarmored dinoflagellates and ciliates

content together made up 12 and 9 O/oof the cell numbers in the summer and fall, respectively (Table 3). Thus, A seasonal succession of phytoplankton and mi- bacterial numbers and heterotr0ph.i~ flagellate bio- crobes occurs in coastal Newfoundland waters mass track the seasonal cycle of water temperature, (Hollibaugh & Booth 1981, McKenzie unpubl.). Large, cvhich peaks in the late summer and early fall (Powell chain-forming diatoms dominate phytoplankton bio- et al. 1987). In the late fall and w.inter, single-cell mass and numbers during the spring bloom in diatoms become abundant (McKenzie unpubl.), most Newfoundland waters (Table 3). These autotrophic likely in response to increased vertical transport of species deplete dissolved silica in surface waters by essential, inorganic nutrients due to destabilization of the end of May, resulting in a 200-fold decrease in the water column. datom concentration by June and early July (Table 3). The seasonal succession of phytoplankton and Bacterial activity is suppressed during the bloom due microbes is reflected in the changing composition of to low temperature and substrate levels (Pomeroy the faeces of Oikopleura vanhoeffeni and Calanus & Deibel 1986, Pomeroy et al. 1991). Phagotrophic finmarchicus. Diatoms are the primary component of and mixotrophic organisms are found in all seasons, both ' faeces in the winter and spring though they predominate in the summer and fall as (Fig. 21), especially chain-forming Skeletonema costa- 262 Mar. Ecol. Prog. Ser. 84: 255-264, 1992

100 derived from fragmented diatoms. We Calanusfinmarchicus have not attempted to quantify the rela- tive importance of armored vs unarmored or autotrophic vs heterotrophic plankters as a food source for copepods and appen- dicularians.

A changing food web: links to the microbial loop

The faecal pellet contents of Oikopleura vanhoeffeni and Calanus finmarchicus follow a similar pattern. They are filled April July Septernber with predominately large (> 20 pm) auto- trophic species in the spring and hetero- Month of Sample trophic nanoflagellates and bacteria in the fall (Fig. 21), thus showing seasonal 100 shifts in the zooplankton food web struc- ture (Fig. 22). The inclusion of a variety of taxonomic groups and the ingestion of 80 both autotrophic and heterotrophic spe- Q) t0 cies at all sampling dates (Fig. 21) con- E firms oikopleurids' role as a 'generalist' 5 60 v suspension feeder. A potential seasonal 00 change from the flux of opal due to dia- 2 40 toms in 0. vanhoeffeni faeces in the 2 spring (Fig. 21) to opal and calcite due to aQ) coccolithophorids in the fall and winter 20 faeces occurs (Fig. 21). The surprisingly high content of bac- teria (Figs. 10 & 21) and heterotrophic 0 nanoflagellates in freshly egested fall March June November January Calanus finmarchicus faeces raises the Month of Sample question of the validity of existing con- 0 diatoms E3 dinoflagellates choanoflagellates ceptual models of the classic food chain, chrysophytes silicoflagellater m coccolithophores fromdiatoms to copepods to fish, and ciliates bacteria =I gests it is time to look for mechanisms of Fig. 21 Oikopleura vanhoeffeni and Calanus finmarchicus. Percent how this consumption of bacteria and occurrence in the fields viewed per faecal pellet of phytoplankton species nanoplankton takes place. Our observa- within general taxonomic groups at each sampling date. ' At least 1 identi- fiable heterotrophic species was found in this group. tions indicate that C. finmarchicus ingests non-phytoplanktonic, 'colorless' food par- ticles, generally overlooked in laboratory turn, Chaetoceros spp., Thalassiosir-a spp, and Fra- feeding studies and energy flow models. Marine snow

gillariospis spp. (Figs. 1, 3 & 17). An assortment of is absent during the spring, but makes up 36 and 7 O/o heterotrophic nanoplankton species (ciliates, choano- of total particles in the summer and fall, respectively flagellates and gymnodin~ods)fi.11 the faeces in the (Table 3). Since these aggregates are large (150 pm), fall (Figs. 9-16 & 21) Also abundant are amorphous they along with attached bacteria and nanoplankters mucous blobs which may be the remains of soft-bodied may be ingested by C. finmarchicus, incidentally plankters or marine snow. The large numbers of creating an avenue of vertical flux of bacterial . broken diatoms in the faeces in summer (Figs. 5 & 21), This could also help explain the controversy over ex- the few intact diatoms found in the water column ternal vs internal pellet decay (Honjo & Roman 1978, (Table 3), and the large number of flocs present in Gowing & Silver 1983, Alldredge et al. 1987). the water (pers. obs.) suggest that both 0. vanhoeffeni Previous studies have shown that Calanus finmar- and C. finmarchicus are feeding on marine snow chicus is an indiscriminate, raptorial feeder, ingesting Urban et a1 : Seasonal faecal con tent chdnges of two zooplankters 263

CLASSIC FOOD WEB zooplankton. Choanoflagellate remains were found in the faeces of both Oikoplleura vanhoeffenl and Calanus finrnarchicus (Figs. 8, l1 & 21). These nano- plankton are considered to be important bacteriovores (Sherr et al. 1986) and represent a direct link between macrozooplankton and the microbial loop. The lorica was frequently broken in our samples and the remains were often difficult to recognize. This fragility may explain the relatively few reports of choanoflagellate remains in fecal pellets of macrozooplankton. Heterotrophic dinoflagellates are another important grazer on bacteria and phytoplankton and are an important food source for copepods (Klein Breteler 1980). Protoperidinium depressum has been reported to feed on bacteria (Lessard & Swift 1985), and many unarmored dinoflagellates are phagotrophic and form important links in food chains (Kimor 1981, Odate & Maita 1990). Both P. depressurn (Fig. 7) and unar- mored dinoflagellates were found in the faecal pellets \\ pzEGq of Calanus finmarchicus and Oikopleura vanhoeffeni. NANOPLANKTON Choanoflagellates It is axiomatic that ciliates and phagotrophic proto- I:Ciliales I zoans represent the main link between microhetero- ti-ophs and copepods (Porter et al. 1979, Sherr et al. MICROBIALLOOP 1986). However, our observations indicate that both copepods and appendicularians are able to feed on Fig 22 Model of zooplankton food web To the left of the bacteria, heterotrophic nanoflagellates and ciliates double llne and the dashed llnes IS the classic, dlatom-based food web predomlnant In the spring To the right of the (F~gs.8 to 1 l), and thus are capable of mediating a flow double 11ne and the bold lines IS the mlcroblal loop pre- of energy out of the microbial loop. dominant In the fall Note Presumdbly a llne should connect cyanobacteria with tun~cates,though no ev~denceof th~s was found In our study due to the difference of hab~tatsof Acknowledgements. We thank G Chaisson and the dive team Olkopleura vanhoeffenl (0 "C) and cyanobacter~d(2 6 "C) from the Ocean Sciences Centre and M. Riehl for help in collecting animals We also thank C. Emerson (B~ology Department of ivlemorial University) for the use of the scan- those particles present in greatest abundance (Cowles ning electron microscope and J. Rlch for computer assistance. 1979, Ishimaru et a1 1988) Particles < 10 ,pm have been Thls work was supported by an Operating Grant to D. Deibel from NSERC, and by a NSERC Strategic Grant to Deibel et al. found in faeces (Huntley 1981, Ishlmaru C finmarchlcus (COPE: Cold Ocean Productivity Experiment) Th~sIS Ocean et a1 1988), indicating that although they are not able Sc~enceCentre contribution number 153. to filter particles of this size efficiently (Nival & Nival 1976), they may be ingested when they are abundant and/or trapped within marlne snow aggregates There LITERATURE CITED are repoits of copepods and euphauslids eating clli- ates, tintinnids, choanoflagellates, heterotrophlc dino- Alldredge, A. L., Gotschalk, C. C., MacIntyre, S. (1987). flagellates and coccolithophorids (Berk et a1 1977 Evidence for sustamed res~denceof macrocrustacean fecal pellets in surface waters offsouthern Cal~fornla.Deep Sea Honjo & Roman 1978, Turner 1984, Marchant & Nash Res 3419): 1641-1652 1986, Tanoue & Hara 1986, Ishimaru et a1 1988), Azam, F,Fenchel, T., Field, J G., Gray, J. S., Meyer-Reil, L A, implying direct removal of biomass from the microbial Thingstad, F:. (1983) The ecological role of water-column loop In the fall samples from our study heterotrophlc microbes in the sea. Mar. Ecol. Prog Sei-. 10 257-263 nanoflagellates and bacterla were found in ca 75 % of Berk. S. G , Bro\vnlee, D C, Heinle, D. R, Kllng, H J., Colwell, R. R. (1977) C~llatesas a food source for marine the flelds viewed per pellet wh~leautotrophic diatoms, plankton~ccopepods. M~crob.Ecol. 4: 27-40 dinoflagellates and cyanobacteria were found in ca Booth, B. C., Lewin, J., Norris, R E (1982). Nanoplankton 35 % of the fields vlewed per pellet (Fig 21), indicat- species predomlnant in the subarctlc Paciflc in May and ing the seasonal importance of the mic~obialloop to June 1978. Deep Sea Res 29 (2A).185-200 Burkill, P H., Mantoura, R F C., Llewellyn, C A,Owens, N copepods J P (1987) Microzooplankton grazlng and selectivity of Azam et a1 (1983) defined the n~icrobialloop as the phytoplankton in coastal waters. Mar. B~ol.93. 581-590 interaction between bacteria, flagellates and mlcro- Cowles, T. J (1979) The feeding response of copepods from 264 Mar. Ecol. Prog. Ser. 84: 255-264, 1992

the Peru upwelling system: food size selection. J. mar. Res. bacteria. J. Plankton Res. 5: 279-289 37(3):601-622 Murphy. L. S.. Haugen, E. M. (1985). The distribution and Deibel, D., Lt.(!, S. H. (1992). Retention efficiency of sub- abundance of phototrophic ultraplankton in the North micrometer particles by the pharyngeal filter of the Atlantic. Llrnnol. Oceanogr. 3011): 47-58 pel~tgictunicate Oikopleura vanhoeffeni. Mar. Ecol. Prog. Nielsen, T G., Richardson. K. (1989). Food chain structure of Ser. 81: 25-30 the plankton communities: seasonal variations Fenrhcl, T (1988). Marine plankton food chains. Ann. Rev. of the role of the microbial loop. Mar. Ecol. Prog. Ser. 56: Ecol. Syst. 19: 19-38 75-87 Gowing, M. M., Silver, M. W. (1983). Origins and micro- Nival, P., Nival, S. (1976). Particle retention efficiencies of environments of bacteria mediating fecal pellet decom- an herbivorous copepod, Acartia clausi (adult and cope- position in the sea Mar. Biol. 73: 7-16 podite stages): effects on grazing. Limnol. Oceanogr. Harbison, G. R., McAlister, V L. (1979). The filter-feeding 21(1):24-38 rates and particle retention efficiencies of three species of Odate, T., Maita, Y. (1990). Phagotrophic grazing by dino- Cyclosalpa (Tunicata, Thaliacea). Limnol. Oceanogr. 24: flagellates on diatoms during the spring phytoplankton 875-892 bloom in Funka Bay. Bull. Plankton Soc Jap. 36(2):142-144 Hoepffner, N.,Haas, L. W. (1990). Electron microscopy Paranjape, M. A. (1990). Microzooplankton herbivory on the of nanoplankton from the North Pacific central gyre. Grand Bank (Newfoundland, Canada): a seasonal study. J. Phycol. 26: 421-439 Mar. Biol. 107: 321-328 Hollibaugh, J. T., Booth, J. A (1981). Observations on the Pomeroy. L. R., Wiebe, W. J., Deibel, D., Thompson, R. J., dynamics and distribution of phytoplankton and primary Rowe, G. T., Pakulski, J. D. (1991). Bacterial responses to production on the Grand Banks in the 1980 season. temperature and substrate concentration during the New- In: Grand Bdnks oceanographic study. Mobil Oil Canada. foundland spring bloom. Mar. Ecol. Prog. Ser. 75: 143-159 St. John's, Newfoundland Porneroy, L R., Deibel, D. (1986).Temperature regulation of Honjo, S., Roman, M. R. (1978).Marine copepod fecal pellets: bacterial activity during the spring bloom in Newfound- production, preservation and sedimentation. J. mar. Res. land coastal waters. Science 233: 359-361 36(1):45-57 Porter, K. G., Pace, M. L., Battey, J. F. (1979). proto- Xuniley, bl (1981). Nonselect~ve,nonsaturatec! feeding by mans as links In freshwater planktonic food chains. three calanid copepod species in the Labrador Sea. Nature, Lond. 277: 563-565 Limnol. Oceanogr 26(5):831-842 Powell, J. C., Dabinett, P. E., Gow, J. A. (1987). An annual Ishimaru, T,Nishida, S., Marumo, R. (1988).Food size selec- cycle of abundance and activity of heterotrophic bacteria tivity of zooplankton evaluated from the occurrence of and abundance of hydrocarbonoclastic bacteria in New- coccolithophorids in the guts. Bull. Plankton Soc. Jap. foundland coastal water. Can. J. Microbiol. 33: 377-382 35(2): 101-1 14 Sherr, E. B.. Sherr, B. F., Paffenhofer, G-A. (1986). Phago- K~rnor,B (1981). The role of phagotrophic dinoflagellates trophic protozoa as food for rnctazoans: a 'missing' trophic in marine ecosystems. K~elerMeeresforsch.. Sonderh. 5: link in marine pelagic food webs? Mar. Microb. Food 164-173 Webs l(2):61-80 Kle~nBreteler. W. C. M. (1980). Continuous breeding of Smayda, T (1980). Phytoplankton species succession. In: marine pelagic copepods In the presence of heterotrophic Morris, I. (ed.) The physiological ecology of phyto- dinoflagellates. Mar. Ecol. Prog. Ser. 2: 229-233 plankton. Univers~tyof Cal~forniaPress, Berkeley Lessard. E. J.. Swift, E. (1985). Spec~es-specificgrazing Stoecker, D. K., Taniguchi, A., Michaels, A. E. (1989). rates of heterotrophic dinoflagelldtes in oceanic waters, Abundance of autotrophic, mixotrophic and heterotrophic measured with a dual-l.abel radioisotope technique. Mar. planktonic ciliates in shelf and slope waters. Mar. Ecol. Biol. 87: 289-296 Prog. Ser. 50: 241-254 Marchant, H. J., Nash, G. V. (1986). Electron microscopy of Tanoue, E., Hara, S. (1986). Ecological implications of fecal gut contents of faeces of Euphausia superba Dana. ~Vem. pellets produced by the Antarctic krill Euphausia supeiba natn. Inst. polar Res., Tokyo Spec Issue 40: 167-177 in the Antarctic Ocean. Mar Biol. 91. 359-369 Marshall, H. G. (1984). Phytoplankton distribution along the Turner, J. T (1984). Zooplankton feeding ecology: contents eastern coast of the USA. V Seasonal density and cell- of fecal pellets of the copepods Eucalanus pileatus and volume patterns for the northeastern cont~nentalshelf. Paracalanus quasimodo from continental shelf waters of J. Plankton Res. 6(1):169-193 the Gulf of Mexico. Mar. Ecol. Prog. Ser. 15: 27-46 Marshall, H. G.. Cohn, M. S. (1983).Distribut~on and compo- Urban, J. L.. Deibel. D., Schwinghamer, P. (in press). Seasonal sitlon of phytoplankton In northeastern coastal waters of variat~onin the densities of fecal pellets produced by the the United States. Estuar, coast. Shelf Sci. 17 119-131 pelaylc tunicate Oikopleura vanhoeffeni and the copepod Marshall. S. M., Orr, A. P. (1955). On the biology of Calanus Calanus finmarchjc~~s.Mar. Biol. finmarchicus. VIII. Food uptake, assimilation and excre- Utermohl. H. (1958). Zur Vervollkommnung der quantitativen tion in adult and stage V Calanus. J. mar. blol. Ass. U.K. Phytoplankton-Methodik. Mitt. int. Ver. theor. angew. 34: 495-529 Limnol. 9: 1 -38 Michaels. A. F., Silver, M. W. (1988). Primary production, Veldhuis. M. J. W., Colijn, F., Venekamp. L. A. H., Villerius, sinking fluxes and the microbial food web. Deep Sea Res. L. (1988). Phytoplankton primary production and biomass 35(4):473-490 in the western Wadden Sea (the Netherlands): a com- Mullin. M. M. (1983).In sltu measurement of filtering rates of parison with an ecosystem model Neth. J. Sea Res. 22(1): the salp Thdlia democralica on phytoplankton and 37-49

This arttcle was subn~ittedto !he editor Manuscript first received: October 21, 1991 Revised version accepted: May 5. 1992