Journal of Plankton Research Vol.18 no.10 pp.1767-1779, 1996

On assessment of prey ingestion by copepods

P.G.Verity and G.-A.Paffenh6ferSkidaway Institute of Oceanography, 10 Ocean Science Circle, Savannah, GA31411, USA

Abstract. The consumption of photosynthetic and heterotrophic cells by an abundant calanoidcopepod species feeding on natural plankton communities was quantified with a state-of-the-artimage-analysis system. Late copepodid stages of Eucalanus pilcatus did not ingest bacteria, smallphotosynthetic and heterotrophic nanoplankton, or the abundant Ceratium spp. in quantifiableamounts. Although diatoms were by far the most abundant cells (in terms of POC H), the copepodsingested a higher percentage of ciliates in relation to their abundance than of diatoms and smallheterotrophic dinoflagellates in the first experiment, and ingested a higher percentage of dinoflagel-lates and ciliates compared to diatoms in the second experiment Heterotrophic cells sufficiently largeto be captured were repeatedly preferred by E.pileatus over autotrophs of similar or larger size. More-over, among the cells which could be individually perceived by this calanoid, larger ones were not pre-ferred over smaller cells, implying that some aspect of food quality can be as significant as prey size.These results support the notion (e.g. Kleppel, Mar. EcoL Prog. Ser., 99,183-195,1993) that feeding bycopepods will be underestimated if ingestion of heterotrophic food organisms is not quantified. Whilethe proposed microscope-based method is comparatively slow (~1 h per sample), it is the only tech-nique which provides detailed information on both the size and trophic composition of ingested prey.

Introduction

Despite early reports of omnivorous diets among calanoid copepods (Marshall,1924; Mullin, 1966; Petipa et al., 1970), the ease of using fluorometry to estimaterapidly the consumption of phytoplankton in incubation bottles resulted in con-siderable focus on herbivory. Appreciation of the role of the microbial food webhas emphasized the significance of omnivory, because the sheer magnitude ofsecondary production by hetero- and mixotrophic ciliates, dinoflagellates andnanoplankton (e.g. Lynn and Montagnes, 1991) requires a substantial sink toprevent massive and rapid accumulation of these micrograzers. The latter isseldom observed, in fact the biomass of nano- and microzooplankton appearsrelatively constant given their growth potential.

Omnivory is characteristic of juvenile as well as adult stages (Stoecker andEgloff, 1987); estuarine (Gifford and Dagg, 1988), neritic (Turner, 1987a) andoceanic (Mullin, 1966) taxa; and temperate to tropical species (Landry, 1978;Petipa, 1978). Recognition of this gustatory diversity is salient from several per-spectives. Without including dietary contributions from heterotrophic as well asautotrophic prey, the total ration, biomass and egg production of metazooplank-ton cannot be predicted or causally related to measures of food availability. More-over, the role of meso- and macrozooplankton in the structure and function ofpelagic food webs, particularly contributions by key taxa, cannot be elucidatedwithout fundamental knowledge about their food acquisition potentials, life his-tories and survival strategies (e.g. Verity and Smetacek, 19%).

The question of contemporary importance, however, is how to quantifyomnivory? Examination by brightfield microscopy of the composition of preycommunities before and after their modification by meso- and macroconsumers

© Oxford University Press 1767

P.G. Verity and G.-A.Paffenhdfer

Table L Feeding experiments with adult male (M), female (F), and copepodid (C) stages of Eucalanuspileatus

Collection SiteIsobath in situTemperaturesExp. Temperature

Replicate Bottle

Estimated Numberand stages ofcopepods at start

At end of experiment

Exp 27-28 June 1994

36°OO'N, 75°2521 m22.5 to 24.5 °C21.7

A

18CIII/IV

1 M, 1 F,1CIII,4CIV,11CV

W

in upper 18 m

B

24 CIII/IV

2M,2CIII,4 CIV, 16 CV

C

12 CIII/IV

2 M, 1 CI,3CIV,7CV

Exp. 30 June-1

36°10'N,75°16'30m22 °C in upper21.5

A

i cm15 CIV

1 M, 1 F,15 CIV, 8 CV

July 1994

W

14m

B

7 CIII7 CIV

2 CIII,8CIV.4CV

has long been the standard method. However, this laborious and somewhat sub-jective approach (in terms of biomass estimation as well as separating auto- fromheterotrophic prey) is at odds with the current interest in measuring biologicalvariables at time and space scales commensurate with the driving physical forces.Other newer techniques, such as high-pressure liquid chromatography (HPLC),can provide chemotaxonomic information about phytoplankton which are poten-tial prey (Strom, 1993; Waterhouse and Welschmeyer, 1995), but not specific dataon various non-plastidic prey. With the development of fluorescence microscopyadapted specifically for relatively rapid, precise, and accurate determination of theabundance, size distribution, and biomass of auto- and heterotrophic plankton(Sieracki etal., 1989a,b; Verity and Sieracki, 1993; Verity etai, 1993), it has becomepossible to relate in detail the effects of grazers on potential prey fields. Here, wereport initial observations of the feeding of late copepodid stages of the sub-tropical calanoid Eucalanus pileatus on natural particle assemblages. The data arelimited to two experiments, the results of which are part of a larger study beingconducted through 1996. The purpose here is to describe in detail the method foranalysis of the prey community, and offer insights into the power of such a com-prehensive tool. The importance of Eucalanus in the zooplankton food web willbe described elsewhere; general plankton relations in this region during spring arepresented in Verity et al. (1996a).

Method

Eucalanus pileatus is among the abundant calanoid copepods in subtropicalneritic waters (Bowman, 1971; Binet, 1977; Valentin, 1984; Madhupratap andHaridas, 1986; Turner, 1987b). During the latter part of June 1994, E.pileatusoccurred abundantly between 35°30' and 36° 30' on the continental shelf north ofCape Hatteras. Copepods were collected in tows lasting 3-4 min with a net of 202jun mesh and a 4-1 codend, while the ship was drifting. Microscope observationsconfirmed that the E. pileatus were undamaged and each had a visible oil globule.Within 60 min, batches of 12-24 animals (see Table I) were transferred into three

1768

Assessment of prey ingestion by copepods

1900 ml jars (two jars in the second experiment) containing a natural suspensionof particles from the depth of maximum occurrence of E.pileatus. Control jarscontained no copepods, and all jars were amended with 3 jimol NO31"1. The jarswere rotated on a ferris wheel at ~0.2 r.p.m. in a water bath maintained close tocollection temperatures. Experiments were run over 24 h in a natural light-darkcycle (15:9 h).

At the beginning and end of each experiment, samples were drawn from eachcontrol and experimental jar for composition, enumeration and biomass measure-ments of plankton using a color-imaging cytometry system. The methods used toenumerate and determine the biovolume of plankton cells via image-analysistechniques, and subsequent conversion to cell carbon and nitrogen, have alreadybeen developed (Sieracki et al., 1989a,b; Verity et al., 1992; Verity and Sieracki,1993). They have been used to quantify plankton communities in diverse environ-ments such as the Atlantic Ocean (Verity et al., 1993), Pacific Ocean (Verity et al.,1996c) and Norwegian fjords (Hansen et al., 1994). Briefly, triplicate subsamplesfrom each bottle were initially fixed with very low concentrations of glutar-aldehyde (0.3% final concentration), stained with DAPI (10 (ig ml"1 final concen-tration) for 4 min, then momentarily stained with proflavine (1.4 p-g mh1 finalconcentration), and finally collected on black 0.4 jxm Nuclepore filters. Filterswere covered with a small drop of low-fluorescence immersion oil and a coverslip,and slides were stored frozen. Proflavine and DAPI are ideal stains for image-analysis studies due to their bright fluorescence.

The samples were analyzed with a color-imaging system consisting of a desktopcomputer (Pentium 100 mHz with 32 MB RAM and 1 GB hard drive) housingseveral integrated software packages which operated microscope-mounted hard-ware and additionally performed image-processing functions. An Olympus BX-60 fluorescence microscope, which was also capable of brightfield, darkfield, phasecontrast and differential interference contrast illumination, provided high-qualityinitial views of the filtered samples using a wide range of objectives (10-lOOx).An Optronics DEI-470 integrating 2/3" CCD captured analog RGB color imagesfrom the filter surface at variable frame rates from 1/10000 s to 2 min. An on-board frame grabber continuously displayed full-frame images, with integrationtimes up to 2 min producing a minimum sensitivity of 0.0025 lux. A Windows-based imaging software, Image Pro Plus (IPP) for Windows (Media Cybernetics,Inc.), controlled image capture, enhancement, measurement, analysis and output.IPP directed an electronic shutter mounted in-line in the microscope light path sothat the sample was exposed to excitation for only as long as the camera shutterwas open, thus minimizing photobleaching. Images were digitized via an Ima-Graph ImaScan/Chroma PCI video capture board and displayed on a Nokia 17"RGB monitor.

Two analytical approaches were used depending on the density on the slide ofthe particular cell population being analyzed. For cells which were numerous perfield and relatively uniform in brightness in a given sample (e.g. phototrophic andheterotrophic nanoplankton), randomly chosen whole fields were analyzed. Inthis case, all the cells in a field were segmented (identified in the image) andmeasured automatically. This approach required that a single threshold be used

1769

RGVerity and G.-A.Paffenhdfer

for all cells in the image, but was faster than automatically finding individual cellthresholds. Analysis of an image with 40 cells took ~4 s. The microscope was inter-faced with a motorized stage and associated modular automation controller,which permitted the operator to scan transects of variable length across theplankton slides while the computer recorded the fraction of the surface area of theslide which has been examined. For abundant populations, the computer ran-domly selected individual locations on the slide. The entire process (moving to agiven location, focusing, opening an electronic shutter, grabbing an image, closingthe shutter and moving to a new location) was automated and computer con-trolled.

For more rare cells (e.g. ciliates and heterotrophic dinoflagellates), transects ofthe slide were scanned, and individual cells were isolated and identified by theoperator interactively. In this way, densities of rare cells were calculated pervolume of sample. Sub-images containing the individual cells were temporarilystored and measured automatically overnight using automatic threshold determi-nation according to the second derivative method, segmentation and cellmeasurement (Sieracki et al., 1989a).

Dinoflagellates were distinguished from other flagellates based on cell mor-phology and structure of the nucleus, especially the unique condensed chromo-somes visible by DAPI staining. Heterotrophic and autotrophic nanoplanktonand dinoflagellates were discriminated by the absence and presence of autofluo-rescent chloroplasts, respectively. Ciliates were identified by their oral ciliarureand other diagnostic features, e.g. the large plastids in Mesodinium. Diatoms andceratia were distinguished by their unique external cell walls. Typically, 100 cellsof each type were measured in each sample. The average coefficient of variationof triplicate counts of nanoplankton, dinoflagellates, diatoms, ceratia and ciliateswas 7, 9, 8, 13 and 10%, respectively. The cell volumes of nanoplankton, dino-flagellates and ciliates were estimated from linear dimensions using the robustassumption that cell depth and cell width are equal dimensions in these small cells(Sieracki et al., 1989b; see also Verity et al., 1992). The linear dimensions of diatomsand ceratia were measured, and applied to the appropriate stereometric formulaeto estimate cell volumes (Edler, 1979). Heterotrophic dinoflagellate andnanoplankton biovolumes were converted to cell carbon using factors of 0.24 pgC jim~3 [derived from Lessard (1991) and Verity era/. (1992)] and 0.22 pg C urn"3

(Borsheim and Bratbak, 1987) respectively. Ciliate carbon was estimated frombiovolume using empirically derived conversion factors (Putt and Stoecker, 1989).The carbon contents of diatoms and ceratia were estimated from cell volumesaccording to Edler (1979).

Copepod nauplii, which also can serve as prey, were not enumerated becausethey occurred in insufficient quantities to be adequately accounted for by ourmethod. Since they can perceive, avoid or escape from predators, their contri-bution as a food organism is expected to be small except for strongly carnivorouscopepods like Centropages velificatus (Paffenhdfer and Knowles, 1980).

Since the amount of food removed by a late copepodid stage of E.pileatus couldonly be approximated, not knowing in advance the composition and abundance ofpotential food particles, the concentrations of copepods were deliberately varied

1770

Assessment of prey tngestion by copepods

from 12 to 24 CIII/IV per 1900 ml jar on 27 June, and 1 CIII/15 CIV and 7 CIII/7CFV on 30 June 1994 (Table I). Their stage was estimated from size without a dis-secting microscope, based on experience (G.-A.Paffenhofer). Since the age of col-lected adult female copepods cannot be accurately determined, and copepodids faroutnumbered females, we decided to utilize the former for experiments. At the endof each experiment, the surviving copepods were removed from each jar, and theircephalothorax length (to the nearest 0.02 mm) and stage were determined.Animals which molted to males were not included in the feeding rate calculations.All others were immediately dried at 60°C and upon return to the laboratory thecombined copepod weight from each jar was determined to the nearest 0.1 jig witha Cahn-Balance. Ash-free dry weight was multiplied by a factor of 0.45 to obtaintheir organic carbon content (G.-A.Paffenh5fer, personal observation).

Growth and grazing coefficients and ingestion rates of each potential prey cat-egory were calculated from the end-point sample measurements according to theequations of Frost (1972). Weight-specific ingestion was calculated by normaliz-ing the disappearance of prey per experimental bottle (relative to controls) to thenumber of copepods and their estimated carbon biomass. In the first experiment,where three bottles containing added Eucalanus were incubated, measurementerrors in clearance and ingestion were calculated from mean estimates for eachbottle. Here, the estimated errors reflect the variance among the three experi-mental bottles. In the second experiment, where only two bottles contained addedEucalanus (a third bottle deliberately contained another copepod genus), thethree subsamples from each bottle were treated as separate replicates and errorswere estimated from the six combined subsamples. Here, the estimated errorsinclude both intra- and inter-bottle variability.

Results

During each 24 h study, most copepodids molted to the next stage (Table I). Mor-tality occurred once (Experiment 30 June, jar A) which we assume was due tolosing one copepod when closing or opening the respective jar. In one case, we hadone more copepod at the end than at the start (Experiment 27 June, jar C) whena late nauplius molted to CI. Copepodid stages of all calanoid genera and naupliiof E.pileatus were removed from each control and experimental jar prior to theaddition of experimental animals.

We quantified removal by E.pileatus of seven different groups of potential foodorganisms (Tables II—III); data on bacteria are not shown because they are wellbelow the minimum prey size of Eucalanus. Cell volumes represent means of tripli-cate measurements of £100 cells from the two experiments. The phytoplanktoncommunity included diatoms (Rhizosolenia alata, R. stolterfothii, Chaetoceros spp.,Skeletonema costatum, Cerataulina sp., Leptocylindrus danicus), dinoflagellates(Ceratium tripos, Clineatum, Cfurca, Prorocentrum spp.), and small photosyn-thetic nanoplankton (cryptomonads, prasinophytes, coccolithophorids). This com-munity is similar to that previously reported for these waters (Marshall andRanasinghe, 1989). Heterotrophic (= aplastidic) nanoplankton included species ofParaphysomonas, Bodo, Leucocryptos, and choanoflagellates. Heterotrophic

1771

|~5 Table n. Ingestion and clearance rates of E.piieatus on 27-28 June 1994. See Method for details

Food taxon

Phototrophic nanoplanktonHeterotrophic nanoplanktonHeterotrophic dinoflagellatesCiliatesCeratiaDiatomsTotalWeight-speciGc consumption

(u.g Cu.gr1 C of copepodper 24 h)

Cell volume(u,m3)

1.1 X 102

1.6 X 102

4.6 X 102

1.3 X 103

1.6 x 104

1.8 x 104

-

Mean cellconcentration(ugCI-1)

32287

141278

171

Ingestion rates(ng C copepod-1

Bottles

A

00

19.084.1

0304.0407.1

0.55

B

00

15.863.30

246.2325.3

0.61

h"1)

C

00

15.572.50

279.8367.8

0.59

x (SD)

0(0)0(0)

16.8(1.9)73.3 (10.4)0(0)

276.7 (29.0)366.7 (40.9)

0.58 (0.03)

Clearance rates(ml copepod-1 h"1)Bottles

A

002.76.003.93.7

-

B

002.34.503.22.9

-

C

002.25.603.63.3

-

x (SD)

0(0)0(0)2.4 (0.3)5.4 (0.8)0(0)3.6 (0.4)3.3 (0.4)

-

yj

3.uCL

P>5

<?g-O:

ft

Table HI. Ingestion and clearance rates of E.piieatus on 30 June-1 July 1994. See Method for details

Food taxon

Phototrophic nanoplanktonHeterotrophic nanoplanktonHeterotrophic dinoflagellatesCiliatesCeratiaDiatomsTotalWeight-specific consumption

(u.g C u,g-' C of copepodper 24 h)

Cell volume(ujrr1)

1.6 X 102

U X 102

6.5 x 102

2.4 x 103

1.4 X 104

13 x 104

-

Mean cellconcentration("•g C I"1)

4024

4198

70165

Ingestion irates(ng C copepod"1 h~')Bottles

A

00

25.7109.9

0226.6361.2

0.67

B

00

20.488.20

203.6312.2

0.82

x (SD)

0(0)0(0)

23.9 (3.7)99.0(13.3)0(0)

215.1(15.2)336.5 (32.0)

0.74 (0.10)

Clearance rates(ml copepod"' h~')Bottles

A

006.86.103.33.6

-

B

005.14.602.93.1

-

.r (SD)

0(0)0(0)6.0 (0.9)5.3 (0.8)0(0)3.1(0.3)3.4 (0.3)

-

Assessment of prey Digestion by copepods

Table IV. Mean clearance (± SD) of several dinoflagellate taxa by Eucalanus in the experiment during27-28 June

Dinoflagellate

Unidentified heterotroph sp.Katodinium rotundatumProrocentrum minimumHeterocapsa sp.Prorocentrum micans

Cell size(p.m)

5-78-10

12-1417-1922-25

Cell volume(tun3)118 (37)384(60)

1215 (221)2162 (344)3620 (617)

Clearance(ml copepocH h~')

0.2 (0.2)1.9 (03)3.4 (03)4.1 (0.4)4.0 (0.5)

Cell sizes are approximate dimensions of cell length and width. Cell volumes of the two smallest taxawere calculated assuming a circular cross-section. Cell volumes of P.minimum, Heterocapsa, andP.micans were calculated assuming dorsoventral flattening of 10, 20, and 50%, respectively (Edler,1979; Verity era/., 1992).

(= aplastidic) dinoflagellates were represented by both thecate and athecate taxa,especially spindle-shaped morphotypes of the latter. Ciliates included both plas-tidic and aplastidic taxa, with the common genera being Strombidium, Lohman-niella and Strobilidium.

Bacteria, photosynthetic and heterotrophic nanoplankton, and Ceratium spp.were not eaten; the former three groups probably because the cells were too smallto be collected efficiently, the latter perhaps because the Ceratium were morpho-logically complex. Heterotrophic dinoflagellates, ciliates and diatoms were ingestedin all bottles containing Eucalanus. In the first experiment (Table II), ciliates werecleared at significantly higher rates than the larger diatoms, which were cleared atsignificantly higher rates than the smaller dinoflagellates. Ingestion of diatomcarbon was significantly greater than that of ciliates and dinoflagellates due to the6-fold higher biomass of diatoms. In the second experiment (Table III), clearance ofciliates and dinoflagellates was not significantly different, but both were signifi-cantly greater than clearance of the larger diatoms. Again, more diatom carbon wasingested due to the greater abundance of diatoms. Thus, in both experiments, thesmaller heterotrophs were cleared at higher rates than the larger phototrophicdiatoms The average percent reduction in prey abundance in bottles containingEucalanus was 17,20 and 24% (Experiment 1) and 19,24 and 21% (Experiment 2)for diatoms, ciliates and dinoflagellates, respectively (data not shown).

There was also a size-based component to feeding by Eucalanus. In the June27-28 experiment, several taxa of photosynthetic dinoflagellates of varying sizewere present in sufficient abundance to enumerate (Table IV). While they did notcontribute substantially to the daily ration of the copepods, their disappearanceduring the incubation was proportional to cell size. The smallest heterotrophicdinoflagellates of 5-7 u,m were not significantly collected by Eucalanus. Thelarger Katodinium (8-10 \LVO) and Prorocentrum minimum (12-14 \\.m) werecleared at 1.9 and 3.4 ml copepod"1 h"1, respectively, while cells of Heterocapsa sp.and Prorocentrum micans, ranging from 17 u,m to 25 jim, were removed at 4 mlcopepoch1 h"1. The three highest clearance rates were significantly greater thanthat of Katodinium, which was significantly greater than the smallest cells.

On 27-28 June, the copepods ingested daily between 55 and 61% of their bodyweight (Table II), and on 30 June-1 July ingested 67 and 82% (Table III). When

1773

P.G.Verity and C-A.PalTeiihofer

Table V. Mean prey concentration (per cent of total prey carbon in ingested categories) and ingestionrates (per cent of total prey carbon which was ingested)

Food type

Heterotrophicdinoflagellates

CiliatesCeratiaDiatomsTotal

Experiment 27-28 June1994

Mean cell Aconcentration

6%13%11%70%

100%

5%21%

0%74%

100%

B

5%19%0%

76%100%

C

4%20%0%

76%100%

Experiment 30 June-11994

Mean cell Aconcentration

4%19%8%

69%100%

7%30%0%

63%100%

July

B

7%28%0%

65%100%

comparing prey availability to prey ingestion, ciliates and heterotrophic dino-flagellates amounted to 24-26% of ingested carbon in the first study, althoughthey represented only 19% of the standing stock of food items which were clearlylarge enough to be eaten (Table V). In the second study, 35-37% of the ingestedfood was ciliates and dinoflagellates, although they represented only 23% of avail-able food carbon (Table V). In each experiment, the copepods ingested a higherpercentage of heterotrophic than autotrophic food particles in relation to thoseavailable.

Discussion

Hundreds of papers have been published on zooplankton feeding using variousmethodologies for more than 60 years, with the general goal of determining theamount and type of food consumed by respective stages, species or groups(Paffenhofer, 1988). In any improved protocol, all potential food particles shouldbe considered and quantified, and the behavior of the grazers must be knownwhen interpreting the data. Various studies have shown that copepods makedecisions about what and how much they eat (e.g. Price, 1988; Vanderploeg et ai,1988). The present results illustrate that late copepodid stages of E.pileatus ingestmore heterotrophs than autotrophs in relation to their abundance, and that cellsize and prey composition are equally important variables influencing ingesticn.Eucalanus pileatus cannot represent all calanoids, but has food acquisitionbehavior similar to other abundant calanoid genera, e.g. Paracalanus and Temora(G.-A.Paffenhofer, personal observation). Larger protozoa (>10 p,m length orwidth) are ingested at high rates by various stages and species of calanoid cope-pods otherwise known to be herbivorous (e.g. Petipa, 1978; Stoecker and Egloff,1987; Gifford, 1993; Fessenden and Cowles, 1994; Ohman and Runge, 1994). Forexample, at similar abundances of larger phytoplankton and microzooplankton,the percentage of non-plant material (protozoa and metazoa) in gut contents ofCalannspacificus ranged from 23 to 99%, and of Clausocalanus sp. from 55 to 98%(Kleppel et al., 1988). These observations were based on pigment identification;the advantage of the approach used here is that information on sizes and specifictaxonomic groups is simultaneously available.

1774

Assessment of prey Digestion by copepods

Others have used inverted microscope observations to quantify ingestion ofprotists. For example, C.pacificus females did not ingest measurable amounts ofphytoplankton when cell sizes were small (<20 urn, 170 u,g C H)> D u t gTazedheavily on ciliates (>10 ujn, 11 u,g C H; Fessenden and Cowles, 1994). In contrast,C.pacificus copepodid stage V ingested phytoplankton (mostly diatoms) when itoccurred at high concentrations and was of large size (>20 u,m, 2230 u,g C I"1) anddid not ingest measurable amounts of ciliates (15 jig C H). Thus, sizes and abun-dances seem to have affected grazing rates on phytoplankton and ciliates. Adultfemale Cfinmarchicus from two different environments in the Gulf of St Lawrence(Ohman and Runge, 1994) always cleared ciliates, which were more voluminousthan diatoms and dinoflagellates, at higher rates than other food organisms, evenwhen the diatoms amounted to 1460 u.g C I"1 and the ciliates to only 3.1 u.g C I"1.Ciliates were represented by Lohmanniella and Strombidium, the latter rangingfrom -10000 to 30000 u,m3. Dinoflagellates were cleared at rates similar to thoseof diatoms. Both of these studies (Fessenden and Cowles, 1994; Ohman and Runge,1994) show that the genus Calanus has the highest clearance rates on ciliates whenthese are larger than the accompanying food organisms.

In the present study, late copepodids of E.pileatus cleared ciliates, which were5-10 times less voluminous than diatoms, at a far higher rate than the latter(Tables II and III). In the second experiment (Table III), clearance rates onheterotrophic dinoflagellates surpassed those on diatoms, although the averagecell volume of dinoflagellates was only 5% of that of the diatoms. Differences inencounter rates between copepods and motile heterotrophs versus non-motilediatoms (e.g. Gerritsen and Strickler, 1977) are insufficient to explain the largedifferences in clearance. It is apparent from the absolute values of clearance ratesthat all three types of ingested food were actively eaten by E.pileatus. Why mightlarge, easily perceivable and ingestible food particles not be preferably ingested?Calanoids which create a feeding current use it to perceive food particles at a dis-tance. Here the first decision is made whether a particle is worth ingesting ('coarseor preliminary evaluation'). After gathering the particle and displacing it to themouth, the sensors there aid the copepod in deciding whether to ingest or rejectthis particle ('close or fine-tuned examination'). Thus, when food is scarce, long-distance perception may be the dominant process concerning food ingestion (Paf-fenhofer and Lewis, 1990), i.e. food quantity appears to be more important thanquality. When food is abundant, as in the present experiments, tasting at the mouthmay be more important, i.e. food quality may be the deciding factor (e.g. Vander-ploeg et al., 1988; PaffenhCfer et al., 1995). However, this idea remains conjectureuntil it becomes possible to assess differences in food quality or nutritional statusin natural plankton communities.

Mean clearance rates of ciliates were essentially the same in the two experi-ments: 5.4-5.9 ml copepod-1 h"1. The same was true of diatoms: 3.1-3.6 mlcopepod"1 h"1. However, clearance of heterotrophic dinoflagellates was signifi-cantly lower on June 27-28 (2.4 ml copepod-1 h"1) than on June 30-July 1 (5.9 mcopepod-1 h"1). It may not be coincidental that the mean cell size of the hetero-trophic dinoflagellates was smaller in the experiment with lower clearances. Themean cell volume on June 27 (4.6 X 102 jim3) was equivalent to a mean cell

1775

P.G.Verity and G.-A.Paffenhdfer

diameter of 9.5 p.m; the corresponding cell size on June 30 was 11 u.m. While adifference of 1.5 u.m is small, the results from Table IV imply that Eucalanus mayexperience a decrease in particle capture efficiency in this size range. Previousobservation of Eucalanus feeding suggested that it is able to capture Thalassiosirapseudonana cells averaging 6 fj,m with sufficient efficiency to fill the copepod guts,but does not perceive them individually, whereas the larger 12 |xm Thalassiosiraweissflogii was perceived as individual cells (Price etal., 1983). While the differentcell sizes of Thalassiosira undoubtably contribute to sensitivity thresholds, thisprey item is not a good choice for determining capture efficiencies, because of itsextensive production of extracellular p-chitan threads, which cause the actual celldiameter to be 5-10 times larger than the valve diameter (Gifford et al., 1981;Verity and Villareal, 1986). The data in Table IV suggest that capture efficiencyincreases dramatically as cell size increases from 5 to 10 u,m. This is further sup-ported by the lack of significant feeding on the small co-occurring photosyntheticand heterotrophic nanoplankton, whose mean cell diameter was <5 \t.m (Tables IIand III).

The clearance rates presented here for diatoms (70-78 u.g C I"1) are about halfthose found in laboratory studies using diatoms of similar size and concentration(Paffenhofer and Knowles, 1978). In natural communities,different potential foodtypes include dead and elongated algae, detritus, fecal pellets and aggregates, manyof which would be gathered, but not ingested. Preliminary analysis of nearbystations in these waters during this cruise indicated that the carbon biomass ofliving plankton and non-living detritus was similar (Verity et al., 1996a,b). Pre-sumably, this would contribute to lower clearance and ingestion rates when com-pared to studies with a single or few food types because, during handling of anunwieldly, or subsequently unacceptable particles (e.g. Ceratium sp.), ingestions ofother (high-quality) particles are difficult to accomplish. For example, Paracalanusjuveniles ingested far more when offered one food species (T.weissflogii) in com-parison to being offered three (Paffenhofer, 1984). In support of this notion,E.pileatus CV tastes individual fecal pellets of Paracalanus sp. for >1 s beforedeciding on ingestion or rejection (G.-A.Paffenhofer, unpublished observation),which should reduce their ingestion of prey which can be analyzed more quickly.In the future, direct visual observations of the entire feeding process over time (e.g.Paffenhdfer etal., 1995) are required to examine this hypothesis that an increasingnumber of different food particles would reduce the overall ingestion rate.

To assess the impact of metazooplankton on prey communities fully, knowledgeof the potential and actual food organisms is essential. The method described hereis another step in that direction. It is somewhat laborious, but a reliable, replicableand data-intensive method which can be used over endpoint incubations, as donehere, or in time-series experiments by collecting samples at regular intervals.Under the assumption that all cells which disappear are ingested in their entirety,detailed image analysis provides considerably more information than less labor-intensive methods. For example, the results of this study suggest that feeding byEucalanus is influenced both by the composition and size of potential foodparticles. Given the diversity and different functional roles of potential preyorganisms, this approach, when applied to the abundant stages and species of

1776

Assessment of prey digestion by copepods

zooplankton will improve our understanding of the size- and species-dependentnature of interactions between zooplankton and their prey (Vinogradov et al.,1977). Eventually, equally detailed prey assessments must be accomplished in siturather than during bottle incubations.

Acknowledgements

The authors thank L.R.Bulluck III, D.Gibson and P.Lane for technical assistance,and the captain and crew of the R/V 'Columbus Iselin'. V.Patrick prepared themanuscript. This research was supported by contract no. DE-F602-92ER61419from the US Department of Energy and NSF grant OCE-90-22318.

ReferencesBinetJD. (1977) Cycles biologiques et migrations ontog£n£tiques chez quelques cop^podes pelagiques

des eaux ivoiriennes. Cah. ORSTOM Ser. Oceanogr., 15,111-138.Borsheim,K.Y. and Bratbak.G. (1987) Cell volume to cell carbon conversion factors for a bacteriv-

orous Monas sp. enriched from seawater. Mar. Ecol. Prog. Ser.,36,171-175.Bowman,T.E. (1971) The distribution of calanoid copepods off the Southeastern United States

between Cape Hatteras and Southern Florida. Smithson. Contrib. Zool., No. 96,58 pp.Edler,L. (ed.) (1979) Recommendations on Methods for Marine Biological Studies in the Baltic Sea.

Phytoplankton and Chlorophyll. The Baltic Marine Biologists Publication No. 5. University of Lund,Sweden, 38 pp.

Fessenden,L. and Cowles,TJ. (1994) Copepod predation on phagotrophic ciliates in Oregon coastalwaters. Mar. EcoL Prog. Ser., 107,103-111.

Frost,RW. (1972) Effects of size and concentration of food particles on the feeding behavior of themarine planktonic copepod Calanus pacificus. LimnoL Oceanogr., 17,805-815.

GerritsenJ. and StricklerJ.R. (1977) Encounter probabilities and community structure in zooplank-ton: a mathematical model. /. Fish. Res. Board Can., 34,73-82.

Gifford,DJ. (1993) Protozoa in the diets of Neocalanus spp. in the oceanic Subartic Pacific Ocean.Prog. Oceanogr., 32,223-237.

GiffordJXF. and Dagg,MJ. (1988) Feeding of the estuarine copepod Acartia tonsa Dana: carnivory vs.herbivory in natural microplankton assemblages. BulL Mar. Sci., 43,458-468.

Gifford.DJ., Bohrer,R.W. and Boyd.C.M. (1981) Spines on diatoms: do copepods care? LimnoLOceanogr., 26,1057-1062.

Hansen.R, Verity.P.G., Falkenhaug,T.,Tande,K.S. and NorrbinJ. (1994) On the trophic fate of Phaeo-cystispouchetii (Harriott). V. Trophic relationships between Phaeocystis and zooplankton: an assess-ment of methods and size dependence. / Plankton Res., 16,487-511.

Kleppel.G.S. (1993) On the diets of calanoid copepods. Mar. Ecol. Prog. Ser.,99,183-195.Kleppel.G.S., Frazel.D., Pieper.R.E. and Holliday j).V. (1988) Natural diets of zooplankton off south-

ern California. Mar. EcoL Prog. Ser., 49,231-241.Landry.M.R. (1978) Population dynamics and production of a planktonic marine copepod, Acartia

clausii, in a small temperate lagoon on San Juan Island, Washington. Int. Rev. Ges. Hydrobioi, 63,77-119.

Lessard,E. (1991) The trophic role of heterotrophic dinoflagellates in diverse marine environments.Mar. Microb. Food Webs, 5,49-58.

LynnTJ.H. and MontagnesJXJ.S. (1991) Global production of heterotrophic marine planktonic cili-ates. In Reid,P.C.,Turley,C.M. and BurkillJ'.H. (eds), Protozoa and their Role in Marine Processes,NATO ASI Publication, Springer Verlag, New York, pp. 281-307.

Madhupratap,M. and Haridas^*- (1986) Epipelagic calanoid copepods in the northern Indian Ocean.Oceanol Acta,9,105-117.

Marshall fl.G. and RanasingheJ.A. (1989) Phytoplankton distribution along the eastern coast of theU.S.A.—VII. Mean cell concentrations and standing crop. Com Shelf Res., 9,153-164.

Marshall,S,M. (1924) The food of Calanus fmmarchicus during 1923. / Mar. Biol. Assoc. UK, 13,473-479.

Mullin,M.M. (1966) Selective feeding by calanoid copepods from the Indian Ocean. In Barnes^,(ed.), Some Contemporary Studies in Marine Science. Allen and Unwin, London, pp. 545-554.

1777

P.G.Verity and G.-A.Paffenhdfer

Ohman.M.D. and RungeJ.A. (1994) Sustained fecundity when phytoplankton resources are in shortsupply: omnivory by Calanus finmarducus in the Gulf of St. Lawrence. Limnol. Oceanogr.,39,21-36.

Paffenhofer.G.-A. (1984) Food ingestion by the marine planktonic copepod Paracalanus in relation toabundance and size distribution of food. Mar. Biol., 80,323-333.

Paffenh6fer,G.-A. (1988) Feeding rates and behavior of zooplankton. Bull. Mar. Sci., 43,430-445.Paffenh6fer,G.-A. and Knowles,S.C. (1978) Feeding of marine planktonic copepods on mixed phyto-

plankton. Mar. Biol.,4&, 143-152.Paffenhofer,G.-A. and Knowles,S.C. (1980) Omnivorousness in marine planktonic copepods. J. Plank-

ton Res., 2,355-365.Paffenhofer,G.-A. and Lewis,K.D. (1990) Perceptive performance and feeding behavior of calanoid

copepods. J. Plankton Res., 12,933-946.PaffenhOfer.G.-A., Bundy.M.H., Lewis,K.D. and Metz,C. (1995) Rates of ingestion and their variabil-

ity between individual calanoid copepods: direct observations. /. Plankton Res., 17,1573-1585.Petipa,T.S. (1978) Matter accumulation and energy expenditure in planktonic ecosystems at different

trophic levels. Mar. Biol., 49,285-293.Petipa.T.S., Pavlova.E.V. and Mironov,G.N. (1970) The food web structure, utilization and transport of

energy by trophic levels in the planktonic communities. In SteeleJ.H. (ed.), Marine Food Chains,Oliver and Boyd, Edinburgh, pp. 142-167.

Price.HJ. (1988) Feeding mechanisms in marine and fresh water zooplankton. Bull Mar. Sci., 43,327-343.

Price,HJ., Paffenh6fer,G.-A. and StricklerJ.R. (1983) Modes of cell capture in calanoid copepods.Limnol. Oceanogr.,2S, 116—123.

Putt,M. and StoeckerJJXK. (1989) An experimentally determined carbon:volume ratio for marineoligotrichous ciliates from estuarine and coastal waters. Limnol Oceanogr.,34,1097-1103.

Sieracki.M.E., Reichenbach.S. and Webb,K.W. (1989a) An evaluation of automated threshold detec-tion methods for accurate sizing of microscopic fluorescent cells by image analysis. Appl. Environ.Microbiol., 55,2762-2772.

Sieracki.M.E., Viles,C.L. and Webb.K.W. (1989b) Algorithm to estimate cell biovolume using imageanalyzed microscopy. Cytometry, 10,551-557.

StoeckerJ3.K. and Eglof£,D.A. (1987) Predation by Acartia tonsa Dana on planktonic ciliates androtifers. /. Exp. Mar. Biol Ecoi, 100,53-68.

Strora.S.L. (1993) Production of pheopigments by marine protozoa: results of laboratory experimentsanalysed by HPLC. Deep-Sea Res., 40,57-80.

TurnerJ.T. (1987a) Zooplankton feeding ecology: Contents of fecal pellets of the copepod Centro-pages velificatus from waters near the mouth of the Mississippi River. Biol Bull., 173,377-386.

Turner^.T. (1987b) Zooplankton feeding ecology: Contents of fecal pellets of the copepods Eucalanuspileatus and Paracalanus quasimodo from continental shelf waters of the Gulf of Mexico. Mar. Ecol.Prog.Ser.,lS,V-46.

ValentinJ.L. (1984) Spatial structure of the zooplankton community in the Cabo Frio region (Brazil)influenced by coastal upwelling. Hydrobiologia, 133,183-199.

Vanderploegji. A., Paffenhofer,G.-A. and LiebigJ.R. (1988) Diaptomus vs. net phytoplankton: Effectsof algal size and morphology on selectivity of a behaviorally flexible, omnivorous copepod. BullMar. Sci., 43,277-394.

Verity,P.G. and Sieracki^M.E. (1993) Use of color image analysis and epifluorescence microscopy tomeasure plankton biomass. In Kemp.P.F., Shen\B.F., SherrJE.B. and ColeJJ. (eds), Handbook ofMethods in Aquatic Microbial Ecology. Lewis Publishers., London, pp. 327-338.

Verity.P.G. and Smetacek,V. (1996) Organism life cycles, predation, and the structure of marine pelagicecosystems. Mar. Ecol Prog. Ser., 130,277-293.

Verityj'.G. and ViUareal.T.A. (1986)The relative food value of diatoms, dinoflagellates, flagellates, andcyanobacteria for tintinnid ciliates. Arch. Protistenkd., 131,71-84.

Verityj'.G., Robertsen.CY., Tronzo,CR., Andrews>l.G., NelsonJ.R. and Sieracki,M.E. (1992)Relationships between cell volume and the carbon and nitrogen content of marine photosyntheticnanoplanklon. Limnol Oceanogr.,yi, 1434—1446.

Verity,P.G., StocckerAK., Sieracki,M.E., Burkill,P.H., Edwards^E.S. and Tronzo,CR. (1993) Abun-dance, biomass, and distribution of heterotrophic dinoflagellates during the North Atlantic springbloom. Deep-Sea Res. II, 40,227-243.

Verityj'.G., Paffenh8fer,G.-A.,WauaceJD., Sherrji. and SherrJJ. (1996a) Composition and biomass ofplankton in spring on the Cape Hatteras shelf, with implications for carbon flux. Com Shelf Res., 16,1087-1116.

VerityJ".G., Beatty.T.M. and Williams£C. (1996b) Visualization and quantification of plankton anddetritus using digital confocal microscopy. Aqiuu. Microb. Ecol., 10,55-67.

1778

Assessment of prey insertion by copepods

VerityP.G., Sieracki,M.E., Stoecker,D.K. and NelsonJ.R. (1996c) Microzooplankton grazing ofprimary production in the equatorial Pacific Deep-Sea Res. II., in press.

Vinogradov,M.E.)Shushkina,E.A. and Kukina.I.N. (1977) Structural and functional analysis of pelagiccommunities in equatorial upwelling. PoL Arch. Hydrobiol., 24 (SnppL), 503-526.

Waterhouse.T.Y. and Welschmeyer,N.A. (1995) Taxon-specific analysis of microzooplankton grazingrates and phytoplankton growth rates. LAmnoL Oceanogr., 40,827-834.

Received on September 1,1995; accepted on April 12,1996

1779