Journal of Marine Research, Sears Foundation for Marine Research, Yale University PO Box 208118, New Haven, CT 06520-8118 USA

(203) 432-3154 fax (203) 432-5872 jmr@yale.edu www.journalofmarineresearch.org

The Journal of Marine Research is an online peer-reviewed journal that publishes original

research on a broad array of topics in physical, biological, and chemical oceanography.

In publication since 1937, it is one of the oldest journals in American marine science and

occupies a unique niche within the ocean sciences, with a rich tradition and distinguished

history as part of the Sears Foundation for Marine Research at Yale University.

Past and current issues are available at journalofmarineresearch.org.

Yale University provides access to these materials for educational and research purposes only. Copyright or other proprietary rights to content contained in this document may be held by

individuals or entities other than, or in addition to, Yale University. You are solely responsible for determining the ownership of the copyright, and for obtaining permission for your intended use.

Yale University makes no warranty that your distribution, reproduction, or other use of these materials will not infringe the rights of third parties.

This work is licensed under the Creative Commons Attribution-NonCommercial-ShareAlike 4.0 International License. To view a copy of this license, visit http://creativecommons.org/licenses/by-nc-sa/4.0/ or send a letter to Creative Commons, PO Box 1866, Mountain View, CA 94042, USA.

Ecological implications of fecal pellet size, production and consumption by copepods

by Gustav-Adolf Paffenhofer1 and Sally C. Knowles1

ABSTRACT The volume of fecal pellets produced by the calanoid copepods Temora turbinata and Euca-

lanus pileatus increases with increasing weight of the copepod. Pellets produced by nauplii of E. pileatus are eaten by adult females of the same species at about the same rates as phyto-plankton which has a volume similar to that of the pellets (10° µm' ). Pellet production rates of juveniles, being generally in the range of 80 to 120 pellets produced • copepod-1

• 24 h-1, change little with increasing body weight. By comparing two copepod populations it is con-cluded that the size distribution of feeding copepods determines the percentage of fecal matter produced which reaches the sea floor.

1. Introduction

In the marine environment nonliving particulate organic matter (detritus) gen-erally occurs in larger concentrations than living particulate matter (Krey, 1960; Poulet, 1976). Although the availability of detritus to zooplankton has been fre-quently discussed, its quantitative importance as a food source has been described only twice (Petipa et al., 1970; Poulet, 1976). One constituent of detritus is zoo-plankton fecal material, the production of which has been measured numerous times, usually under unnatural conditions (Raymont and Gross, 1942; Marshall and Orr, 1955; Comer et al., 1972; Gaudy, 1974; Honjo, 1976; Reeve and Walter, 1977; Reeve et al., 1977; Grice et al., 1977; Honjo and Roman, 1978). Other investiga-tions have concentrated on determining the rate of vertical transport of fecal pellets to the deep-sea with interest focused mainly on fast-sinking (> 100 m • 24 h-1

)

large pellets (Smayda, 1969, 1971; Schrader, 1971; Small and Fowler, 1972; Honjo, 1976; Wiebe et al., 1976; Turner, 1977; Honjo and Roman, 1978).

Paffenhofer and Strickland (1970) noted that Ca/anus helgolandicus ingests fecal pellets. Davies et al. (1973) concluded that copepods "rework" fecal pellets based on the fact that only 2 to 25 % of the primary production was recovered as pellets in experimental bags while one-third recovery was obtained outside the bags. Riley (1970) stated that "they (fecal pellets) may be a source of particulate food that is

1. Slcidaway Institute of Oceanography, P. 0 . Box 13687, Savannah, Georgia, 31406, U.S.A.

35

36 Journal of Marine Research [37, 1

much more important than would be suspected on the basis of their relative abun-dance." Not only may fecal pellet reingestion significantly decrease the amounts observed at any one time, but methods used in early studies may have collected only large fecal pellets (Wiebe et al., 1976).

Missing is a quantitative evaluation of the relative importance of fecal pellet size to production and consumption in the marine environment. We addressed this prob-lem by asking the following questions:

a) How is fecal pellet size related to the organisms producing them? b) How do grazing rates of copepods on fecal pellets compare to those on

similarly sized phytoplankton? c) What is the relation between the concentration of phytoplankton and pellet

production rate? d) How does pellet production rate change with increasing body size of the

organisms producing them? e) How does the size structure of the zooplankton community influence resi-

dence time of pellets in the water column, their probability of being ingested and their transport to the benthos?

f) How much of the organic matter ingested daily eventually reaches the sea floor?

2. Methods

The pellet production of growing populations of Temora turbinata and Eucalanus pileatus and of adult females of E. pileatus was determined at naturally occurring phytoplankton concentrations at 20°C (for details see Paffenhofer and Knowles, 1978). At the beginning and end of each 20-24 h experimental period, samples were collected for inverted microscope counts of phytoplankton and pellets, and for pellet size determinations measuring length and width of pellets and recording their shape (ellipsoid or cylindrical). There was little evidence that pellets were ingested by nauplii and copepodids 1/ 11 during the first 20 to 24 h after each transfer. After CII the number of pellets produced could not be measured accurately as these older stages collected particles from the bottom. To determine pellet production in rela-tion to food concentration, Rhizosolenia alata f. indica of 30 µ,m diameter was of-fered over a wide range of concentrations (0.4 - 20 mm3 x 1 - 1) to adult females of E. pileatus.

Feeding rates on pellets were measured after E. pileatus nauplii had grazed on R. alata f. indica for 20 h, producing pellets which were in the same volume range as the phytoplankton cells. The nauplii were removed, CV and adult female Euca-lanus pileatus were added. Since naupliar pellets had a theoretical sinking rate of about 0.5 m • h-1 (calculated from Stokes, 1901) water in the 4000 ml beaker (water column 30 cm) was agitated repeatedly to resuspend pellets which might

1979] Paff enhofer & Knowles: Ecology of fecal pellets 37

have sunk to the bottom. Each experiment lasted 7 h. To measure the relative amounts of assimilated carbon and nitrogen, carbon:nitrogen ratios of several phyto-plankton species and of pellets of Eucalanus elongatus adults, produced while feed-ing on R. alata f. indica, were determined with a Perkin Elmer CHN-Analyzer Model 240.

To determine the impact of natural copepod assemblages on pellet reingestion we chose two different assemblages from Onslow Bay, North Carolina. One assemblage consisted almost exclusively of juvenile Paracalanus sp. and the other of late cope-podids and adults of larger copepods. The ash-free dry weights of Paracalanus sp. were calculated from a regression of thorax length versus ash-free dry weight (log1 0

ash-free dry weight [µ,g] = 0.00200 length [µ,m] - 0.788, r = 0.990). The body weights of larger calanoid copepods were obtained from separate weight measure-ments. Grazing rates were obtained from growing populations of Paracalanus sp. and Eucalanus pileatus fed Leptocylindrus danicus (104 to 0.5 X 105 µ,m3 cell/ chain volume) at 0.6 mm3 • 1 - 1• Sinking rates were calculated from Stokes (1901):

2 v=-- gr2

g p-po

Y/

where v is the sinking velocity (cm • sec-1) , g the acceleration caused by gravity (980 cm • sec-2), r the radius of the sinking particle, p the assumed density of the pellet (1.20 g • cm-3

) , p 0 the density of the medium at 25° and 35%0 (~ 1.025 g • cm-3) and T/ the viscosity of the medium at 25° and 35%0 (0.010). We understand that Stokes' calculation is applicable, strictly speaking, to spherical particles only. The particles we studied were either ellipsoid (all naupliar pellets) or cylindrical (copepodid and adult pellets).

The proportions of (percent of pellets produced at 10 m depth) of the pellets reaching the sea floor were calculated in the following manner giving the following example: Pellets produced by Paracalanus of 200-400 µ,m body length were in-gested by all Paracalanus present. Thus, the grazing rate calculated for the entire Paracalanus population, being 376.8 liters swept clear per day in 1 m3 was to be applied (Table 2). Solving the following equation

log C0 - log Ct volume swept clear• h-1 = V

1 1 og e • 1

where V = volume of water (1 m3 in our case), C0 = 100% = initial concentration, Ct = final concentration (% ), h = hours during which pellets of a certain size re-main in the water column, for Ct, we obtain from (C0-Ct) the percentage of pellets produced at 10 m depth which are eaten before they reach the sea floor.

3. Results

Fecal pellet volume increased uniformly as a function of copepod body weight

38

10 "'E v:i g

5 w

:::> _j

0 > 1-w _j _j w a.. I _j

<I: 0 W · LL_Q5 .

Journal of Marine Research

Temora turbinata Exp. July 11, 1977 log pellet . volume=0.711 lag body weight+ 1.780

r=0.954 T

0.05 0.1 0.5 1.0 2.0 COPEPOD BODY WEIGHT [ µgq

[37, 1

Figure 1. Temora turbinata. Fecal pellet volumes in relation to body weight when feeding on Skeletonema costatum, Leptocylindrus danicus and Rhizosolenia alata f. indica, each at an average concentration of 0.5 mm'• 1-1• Each value represents the geometric mean with 95% confidence limits. CI = copepodid stage I.

(Figs. 1 and 2). Geometric means of fecal pellet volumes were calculated since dis-tribution for a given body weight was log normal. Fecal pellet volumes are presented on a logarithmic scale in relation to copepod body weight as this standardizes the measure of variability employed.

Pellets from copepodids and adults fractured more easily than those from nauplii. Although average food concentrations for Temora turbinata and Eucalanus pileatus differed considerably, pellet volumes, slopes and elevations differed little. The pellet sizes for T. turbinata were measured only up to 1.8 µg C body weight since the num-ber and condition (broken) of pellets obtained from later stages were unsatisfactory for proper assessment.

If Stokes' equation (Stokes, 1901) is applied to our pellet volumes we obtain sinking rates of 5 to 10 m • day- 1 for T. turbinata naupliar pellets and 14 to 28 m • day-1 for E. pileatus naupliar pellets (Fig. 3) at 21.7°C. The average sinking rate of an average sized pellet collected by Wiebe et al. (1976) is in line with this calculation.

The importance of fecal pellets as food can be assessed only if pellets are ingested at significant rates. When pellets were offered along with R . alata f. indica (18 cells• m1- 1), those produced by E . pileatus nauplii (3 pellets• m1- 1) were ingested at similar rates as the phytoplankton (Fig. 4). The concentration of R. alata f. indica used is characteristic of intrusions of deep, cold water onto the outer southeastern

1979] Pafjenhofer & Knowles: Ecology of fecal pellets

200 Eucalorus pileahJs Ex11 J<.r1e 15,

100

:§. !,Q

_g w ::;; ::, _J 0 > I-w 10 _J _J w a.. _J

5 5 w "-

lag pellet valume•0855 lag body weight +i,900 r •0.936

f !

Q5 1,0 5 10

OOPEPOO BODY WEIGHT [µgC] 20 30

39

Figure 2. Eucalanus pileatus. Fecal pellet volume in relation to body weight when feeding on Skeletonerna costaturn, Leptocyl indrus danicus and Rhizosolenia a/ata f. indica at average food concentrations of 0.5 mm3, 0.5 mm3 and 6.0 mm' • 1-1, respectively. Each value repre-sents the geometric mean with 95% confidence limits.

continental shelf (Paffenhofer et al., Technical Report, Onslow Bay Intrusion Studies, July/ August 1976). The pellets used were produced 2 to 20 hours before the start of ingestion experiments with E. pileatus CV and females, and showed no visible sign of microfaunal disintegration after 24 to 36 h at 20°C. Two experiments with Temora styli/era adults fed E. pileatus naupliar pellets furthered our evidence of coprohagous feeding behavior by calanoid copepods. Grazing rates on pellets

aeoi,e,,odpellttt (W.tiettd , r976 )

:~:::::: :~~-~. loq lftlf"O rC111•0669 loq ptlltl "°"""' 110 • 0074

f E

Figure 3. Sinking rate versus pellet volume as calculated from Stokes (1901). Viscosity at 25 and 35%0 = 0.010; density of medium at 25°C and 35%0 = 1.025 g • cm-"; density of pellet

= 1.20 g • cm-".

40

700

<Jl 600 1- ::::-"

'1= 500

G:J ~ · a.. · 400 ....J•

3 g_ 300

C 8 0

w 200

a: 0. (!) Q)

z -

a: ...5 (!)

100

Journal of Marine Research

•Eucalanus Qileatus la<;J grazing rate (pellets) =0.570 log grazing rate (Rhizosolenia) + 0.987

•Temara gylifero

•

• •

•

200 300 400 500 600 800 1000

GRAZING RATE on RHIZ0S0LENIA

(ml swept clear · copepod-1• 2 4 hr -i ]

[37, 1

Figure 4. Grazing rates on fecal pellets versus grazing rates on Rhizosolenia alata f. indica be-

ing offered simultaneously (Eucalanus pileatus X pell et concentrations 0.13 X 10° µ.• • m1-1, range 0.08 to 0.19 X 10° µ.' • ml- 1 ; X R. a/ata f. indica 1.22 X 10° µ.' • m1-1, range 0.73 to 2.34. Temora stylifera: X pell et concentrations 0.24 X 10° µ.3 • ml-1, range 0.21 to 0.28; R. alata f. indica 2.78 X 10° µ.3 • m1-1, range 2.56 to 3.01. X pellet size E. pileatus experiment = 5.4 X 10' µ.m', X pellet size T. styli/era experiment = 6.9 X 10' µ.m').

larger (13 to 18 x 104 µ,ms) than those offered here (5.4 to 6.9 X 104 µ,ms) were slightly higher.

An assessment of the importance of fecal pellets as a potential food source was made by comparing growth periods and mortalities of Calanus helgolandicus cope-podids when fed ground fecal pellets from adult C. helgolandicus, with those fed phytoplankton (Table 1). Although the nitrogen concentrations in the fecal pellets

Table 1. Growth periods and survival of Ca/anus he/golandicus feeding on fecal pellets, Lauderia boreali s, and Ske/etonema cost a tum respecti vely, at 15 °C.

Survival Average (% of initi al period population

CIII to adult moulting Carbon Nitrogen Experiment (days) to adulthood) Concentrations (µ.g. 1-')

1. Fecal pellets 20 71 ~400 ~27 2. Fecal pellets 22 57 ~400 ~27 Lauderia borealis 9 97.7 102 20.0 Lauderia borealis 8 98.9 101 19.8 Skeletonema

costatum 15 66.1 100 19.5

1979] Pafjenhofer & Knowles: Ecology of fecal pellets

300

'.c 200

" <,; c:, 100 w u .::> , •• 8 50 [ 40 (/') 30 • '

g 20 :·:

:'•.; __, 10 • • 5 w

::: 5 O 4 •

i5 3 •

2 =i z

5.0 10.0 15.0 20 0

AVERAGE FOOD CONCENTRATION (mm3-C1)

41

Figure 5. Pellet production rate versus average food concentration for adult females of Euca-lanus pileatus feeding on Rhizosolenia a/ata f. indica at 20°C. P = pellet production rate (pellets • 24h-'); Pm = maximum pellet production rate (125.0 pellets • 24h-1); 8 = constant (0.0031); N = average food concentration.

exceeded those in the phytoplankton, mortalities and growth periods from CIII to adults of the copepod fed Lauderia borealis were lower than those fed fecal pellets. Prior to the experiments the copepodids fed fecal pellets had been feeding on L. borealis. The carbon:nitrogen ratios of L. borealis were similar to those of Rhiza-solenia alata f. indica (x ± S.D. = 5.06 ± 0.84) and Skeletonema costatum (5.12 ±

0.46). Fecal pellets from CV and adult females of Eucalanus elongatus fed R. alata f. indica (80 - 120 µ,g C • 1-1) had a C:N ratio of 14.80 ± 3.49.

Production rates of fecal pellets are a function of food concentration and size/ stage of the copepod. A curvilinear function (Ivlev, 1955) was chosen to describe the relation between food concentration and pellet production rates of adult E. pileatus females feeding on R. alata f. indica of 30 µ,m diameter (Fig. 5). The least-squares values of P and 8 were calculated using an iterative technique (Modified Gauss-Newton method-The NILN procedure, Barr et al., 1976). Figure 5 also illustrates the wide range of pellet production rates at low food concentrations, indi-cating large differences in pellet production rates between different animals at these food densities. The range of pellet production rates decreased as food concentration increased.

Pellet production rates of juvenile copepods were determined in four experiments at different food concentrations (Fig. 6). Temora turbinata nauplii generally ingested only S. costatum, starting to feed on L. danicus upon reaching CI and on R. alata f. indica as late CI or CII (0.5 to 0.8 µ,g C). E. pileatus ingested all food species offered after they reached late NN / early NV (0.6 to 0.8 µ,g C). Pellet production rates of

42

.- 150

140

.,; 13

6'. 12 w i5 110

';; 100

;'.: 90

'j 80 _J

70

;;/. 60 u t;: 50

:,: 40 L,J

30

20

10 0

Journal of Marine Research

9 T turbinoto , Morch 10 o T. turbinoto , July 11 11 E. Rileotus , Moy 31 o E.pi leotus, Junel5

0 0

0.1 0.5

C0PEP00 WEIGHT [µg C]

1.0

[37, 1

2.0 3.0 4.0

Figure 6. Pellet production rate in relation to copepod body weight (juveniles). Average food concentrations 1.5 mm• • 1-1 except for E. pileatus June 15 which was 7.0 mm• • 1-1 • For average concentrations of each food species see Figures 1 and 2.

NIU T. turbinata (0.06 to 0.08 µg C) are low while values for older nauplii and early copepodids range from 79 to 169 pellets • copepod-1 • 24 h-1 • There was no clear change in pellet production rate with increasing size of early juveniles. For animals up to 4 µg C we assume that for the given food concentrations (which re-semble environmental conditions on the southeastern shelf of the U.S.A.) the daily fecal pellet production rates at 20°C range between 80 to 120.

Using these fecal pellet production data and data on zooplankton densities and size frequency distributions from Onslow Bay, N. C. we will show how zooplankton size distributions can determine the residence time of pellets in the water column and the probability of pellets being ingested or reaching the benthos.

The data and calculations presented in Table 2 are based on two samples col-lected in 1976, one consisting almost exclusively of a population of Paracalanus sp. (a small copepod) and the other almost exclusively of late copepodid and adult copepods of larger species. Samples were taken with an opening-closing Tucker-trawl (40 x 40 cm square mouth) cylinder-cone net of 110 µm mesh (open mesh: mouth area= 8: 1). This mesh size did not collect nauplii of Paracalanus sp. quanti-tatively; thus we excluded all nauplii from our calculations. We assume, for sim-plicity, that the copepods were evenly distributed over the entire water column of 40 m, that no discontinuity layers existed and that turbulence, diffusion and convec-tion were negligible. Our purpose was to describe the probability of fecal pellets produced at 10 m depth sinking 30 m to the bottom.

Table 2. Fecal pellet sizes, sinking rates and percentages of pellets eaten before reaching the sea floor in relation to copepod population structure.

Station 63, 18 July 1976, 1920 h

Copepod ash-free dry weight Grazing Rate Feeders Pellet Sinking Individual Total

Paracalanus Body Copepods Individual Total Size Rate ml L * Length (.um) xm-o (.ug) (m•m-o) 10'/-Lm• m•day-1 swept clear•day-1 %

200-400 4,600 0.65 3.0 2.5 10 16.5 75.9 67.7 400-600 4,180 1.6 6.7 5.0 16.5 35.5 148.4 49.6 600-800 1,930 4.1 7.9 10.5 26 79 152.5 35.3

Total 10,710 17.6 376.8

Other abundant zooplankton at this station Cyclopoida 2,320 Other Calanoida 165 Harpacticoida 140

Station 53, 5 August 1976, 1540 h * Paracalanus sp. CV / adults 70 10 0.7 31 54 92 6.4 16.0 Clausocalanus adults 23 15 0.3 44 69 230 5.3 12.7 Temora stylifera adults 87 30 2.6 80 102 410 35.7 8.8 Eucalanus pileatus CV / adults 126 45 5.7 113 128 580 73.1 7.1 Centropages furcatus CV / adults 495 25 12.4 69 92 360 178.2 9.7 Nannocalanus sp. adults 15 35 0.5 92 112 470 7.1 8.1 Undinula sp. CIV / adults 11 55 0.6 135 145 680 7.5 6.3 Total 827 22.8 313.3

Other abundant zooplankton at this station Paracalanus sp. < CV 43 Cyclopoida 324 Other Calanoida 30 Ostracoda 75

* Percentage of pellets eaten by the above population before they reach the sea floor, being excremented 30 m above the sea floor. ** Percentage of pellets eaten if both copepod populations occurred together (15.7 + 13.1 = 28.8 L swept clear•h-1).

** %

87.4 71.5 54.9

..... IO -..J

'1:1 ':::tl ;:. ;:.--0: --.... R<>

;:. C)

:: l:l"j C)

C) OQ '<: C) ----e.

'<::I

.;-

44 Journal of Marine Research (37, 1

Table 3. Estimated daily ingestion and assimilation of the Paracalanus population of Station

63 at 10 m depth.

Assimilation Released as

Food Concentration Daily Ingestion (70% ) pellets (30%)

0.2 mm• (2-10 µ,m particle </>) 1.78 mg C 1.25 mg C 0.53 mg C

= 10 mg C • m-' (17.8% of 10 mg C) 0.4 mm• (10-30 µ,m particle ¢) 7.54 mg C 5.28 mg C 2.26 mg C = 20 mg C • m-' (37.7% of 20 mg C)

Total 9.32 mg 6.53 mg C 2.79 mg C

The daily grazing rate of the entire Paracalanus population in one m8 was 376.8 1 swept clear. Assuming that Paracalanus copepodids ingest all particle sizes pro-duced by this population (based on Paffenhofer and Knowles, 1978), the calculated percentage of pellets which reached the sea floor is between 32.3 to 64.7% depend-ing on the size of the defecating copepods at a depth of 10 m. The population of late copepodids and adults from August 5, 1976 had a daily grazing rate of 313.3 1 swept clear m-3 • Here, the percentages of pellets which reached the sea floor ranged from 84.0% to 93.7% of the total produced at 10 m depth. If both populations occurred together, the percentages of pellets from the latter reaching the sea floor would not change as Paracalanus can not ingest the larger pellets. However, the large copepod population could ingest Paracalanus pellets in which case only 12.6 to 54.1 % of the small pellets would reach the sea floor.

Finally, we calculated the amount of the carbon ingested daily by a Paracalanus population at 10 m relative to the amount of fecal carbon from that population which reached the bottom (Table 3). At food concentrations of 10 mg C • m-s (2-10 µ,m particle diameter) and 20 mg C • m- 3 (10-30 µ,m diameter) the Paracalanus population, quite characteristic in size composition for the southeastern continental shelf in summer, removes 1.78 and 7.54 mg C respectively. We applied different grazing rates for each particle size class; of the total 9.32 mg C ingested 6.53 mg C = 70% are assimilated. The value of 70% assimilation efficiency was obtained with E. pileatus. Of the 2.79 mg C released as pellets, 1.34 mg C were ingested (one re-ingestion assumed here) before they reachd the sea floor using grazing rates from Table 2. If pellet carbon would be assimilated at the same efficiency as the phyto-plankton carbon then 0.94 mg would be assimilated. Thus, of the 9.32 mg C in-gested 80.1 % (6.53 mg+ 0.94 mg) would be assimilated and 19.9% would reach the sea floor.

4. Discussion

The importance of fecal material as a food source for marine planktonic particle feeders has been addressed only occasionally (Paffenhofer and Strickland, 1970 and

1979] Pafjenhofer & Knowles: Ecology of fecal pellets 45

Riley, 1970). This is surprising since the assimilation efficiency of planktonic cope-pods ranges from 34.1 % (Corner et al. , 1972, for nitrogen) to 94.0% (Marshall and Orr, 1955, for phosphorus) of the ingested material, consisting mostly of phyto-plankton. Thus, large amounts of organic matter are excreted. However, how much of the resulting fecal material is available for further assimilation and to which grazers it is available remain questionable. Menzel and Goering (1966) found that after a 90-day incubation period samples collected at 1 m depth from the ocean had 43 to 84% of the initial amounts of particulate carbon still present. They considered the remainder to be refractory detritus. On the other hand, late copepodids of Calanus helgolandicus remove all fecal material in experimental jars after phyto-plankton are depleted. Very few pellets were found 1 to 2 days later. This points toward high ingestion and assimilation of organic focal material by copepods.

The value of a fecal pellet as food for a grazing zooplankter depends largely on the pellet's size, shape, sinking rate, and carbon and nitrogen content. Zooplankton pellet sizes range from about 3 X 103 µ,m3 from Paracalanus and Oithona nauplii (Paffenhofer, unpubl. data) to 108 µ,m3 from the euphausiid Meganyctiphanes norvegica (Fowler and Smalley, 1972). Our findings indicate that pellet availability to the zooplanktonic community is a function of pellet size. Smaller pellets are avail-able for ingestion by more feeders and a slower sinking rate increases their prob-ability of being captured. In this case smaller is "better" for the zooplankton com-munity. Piontkovskiy and Petipa (1975) showed that Acartia tonsa preferentially removed rounded particles, which suggests greater removal of ellipsoid (naupliar) than cylindrical (copepodid) pellets.

Carbon and nitrogen content of a pellet contribute to its nutritive value. The in-creased surface to volume area of naupliar versus copepodid pellets leads to the suggestion that a larger population of microorganisms might be present on naupliar pellets in relation to pellet volume resulting in a lower C:N ratio. C:N ratios of freshly produced fecal pellets range from 9.9 to 10.2 for Acartia clausi fed cocco-lithophores and natural seawater particles, respectively (Honjo and Roman, 1978) to 14.8 for Eucalanus elongatus fed R. alata f. indica (this work).

An important question concerning the availability of pellets to planktonic and benthic feeders is the length of time a pellet exists as a single particle. Degradation or disintegration activity appears to increase with increasing temperature (Honjo and Roman, 1978). At 20° to 25°C, the peritrophic membranes of copepod fecal pellets are completely degraded after 24 h. At 20°C we noticed partial degradation of naupliar pellets of T. turbinata and E. pileatus 36 to 48 h after production. Bio-degradation may be particularly effective on naupliar pellets because of their high surface to volume ratios and low sinking rates (generally < 15 m • day-1

). In sub-tropical and tropical surface waters during summer (28 to 30°C), near-complete biodegradation of naupliar pellets might occur within 1 to 2 days. Pellets not in-gested within this period would disintegrate into smaller particles, which would

46 Journal of Marine Research [37, 1

sink even more slowly, becoming available only to fine particle feeders such as ciliates, appendicularia (Jorgensen, 1966) and thaliacea (Harbison and Gilmer, 1976). Small particles could also form the fragile aggregates described by Riley (1970) which have very low sinking rates (0.31 to 2.9 m • day-1

). The availability of such aggregates to consumers has not yet been evaluated. Lal (1977) stated that particles were removed or reduced in size by zooplankton grazing and dissolution, the latter being significant only for particles with a radius of less than 10 µm. How-ever, we feel that grazing generally results first in an increase in particle volume as pellets are at most times larger than the algae ingested. This process may be of par-ticular importance for the survival of larger copepods when nano-and ultraplankton dominate the particle spectra. The formation of pellets by nauplii, which can ingest these small algae, results in food particle sizes available to the larger zooplankton.

Another important point is the downward transport of organic and inorganic matter, and the percentage of the total fecal matter produced per unit time which indeed reaches the sea floor. Most of the literature dealing with fecal pellet sinking rates focuses on pellets of late larval and adult forms (Smayda, 1969, 1970; Fowler and Small, 1972; Turner, 1977; Honjo, 1976; Wiebe et al. , 1976; Honjo and Roman, 1978) and neglects early larval forms (Schrader, 1971; Honjo, 1976). Using data from Smayda (1969, 1971), Schrader assumed that "by virtue of their greater size (50 to 250 µm diameter) pellets rapidly leave the photic zone, sinking between 40 and 400 m • day-1 depleting upper waters of silica and other nutrients." Honjo (1976) estimated that at a pellet sinking speed of 160 m • day-1, within a month 92 % of the coccolithophores produced in the euphotic layer reach the sea floor 5000 m below. By comparison, a discrete coccolithophore sinks 0.14 m or less • day-1

• The conclusions of Schrader (1971) and Honjo (1976) seem to have been based on the exclusive occurrence of pellets with high sinking rates, neglecting the existence of pellets of smaller zooplankton which are particularly important in subtropical and tropical waters. Our results show that downward transport is low if the zooplankton population consists largely of small individuals. The zooplankton of the southeastern continental shelf of the U.S.A. are, during the warmer months, dominated in number and biomass by cyclopoid and small calanoid copepods with individual body weights below 10 µg ash-free dry weight. This size distribution re-sults in the production of small pellets, which may be partially responsible for the impoverished benthos on the southeastern continental shelf (Tenore et al., in press).

N auplii were excluded from our calculations as these forms were not collected quantitatively; however, nauplii and other microzooplankton often constitute a significant portion of the zooplankton community. Beers and Stewart (1969) found that, in the upper 100 m of the eastern Pacific Ocean, microzooplankton were equivalent to 21 to 26% of the larger zooplankton by volume. As the ingestion rates per unit body weight of microzooplankton are assumed to be higher than the rates of larger forms, their pellet production rates might exceed the percentages

1979] Pafjenhofer & Knowles: Ecology of fecal pellets 47

given above. Thus, our calculations on pellet ingestion percentages (Table 2) would have been higher had we included the nauplii whose pellets are smaller and sink more slowly than those of the "model" zooplankton.

Ingestion of pellets is of ecological importance since it leads toward complete extraction of all available carbon and nitrogen from a previously incompletely as-similated food source. This process is of particular importance in "nutritionally dilute" environments (Conover, 1968).

5. Conclusions

In conclusion, the following results were obtained: 1) Pellet size increased with increasing body weight of the feeder. 2) Eucalanus pileatus adult females and Temora stylifera adults grazed fecal pel-

lets of E. pileatus nauplii at similar rates as cells of R. alata f. indica. 3) Pellet production rates by adult females of E. pileatus can be expressed as a

curvilinear function of food concentration (Ivlev-Curve). 4) Pellet production rates of juveniles (NIV to CII) ranged generally from 80 to

120 pellets• animal-1 • day-1 •

5) The percentage of pellets produced at a given depth during 24 h which reached the bottom is a function of the sizes of pellet producers and of the grazing performance of the zooplankton community. The smaller the pellet producer the lower the percentage of pellets which reach the bottom.

6) 19.9% of the organic matter ingested daily at 10 m depth by the modeled Paracalanus population reached the sea floor (40 m depth).

7) The vertical zooplankton community structure largely determines the amount of ingested organic material reaching the sea floor.

Acknowledgments. We acknowledge thankfully the constructive criticism of Drs. John Glasser, David W. Menzel, Mr. D. R. Deibel and two unknown referees. We thank Eileen Hofmann for calculating the curvilinear equation of Figure 4, Danny McIntosh for preparing the Figures and Paula Vopelak for preparing the manuscript. This research was supported by National Science Foundation Grant OCE76-01142 and U. S. Department of Energy Contract EY-76-S-09-0936.

REFERENCES Barr, A. J., J. H. Goodnight, J. P. Sall and J. T. Helwig. 1976. A user's guide to SAS-1976.

Raleigh, NC. Beers, J. R. and G. L. Stewart. 1969. Micro-zooplankton and its abundance relative to the larger

zooplankton and other seston components. Mar. Biol., 4, 182-189. Conover, R. J. 1968. Zooplankton-Life in a nutritionally dilute environment. Am. Zoologist,

8, 107-118. Comer, E. D. S., R. N. Head and C. C. Kilvington. 1972. On the nutrition and metabolism of

zooplankton. VIII . The grazing of Biddu[phia cells by Ca/anus helgolandicus. J. Mar. Biol.

Assoc. U. K., 52, 847-861.

48 Journal of Marine Research [37, 1

D avies, J. M., J. C. Gamble and J. H. Steele. 1973. Preliminary studies with large plastic en-closure, in Estuarine Research, E . L. Cronin, ed., Vol. 1, 251-264.

Gaudy, R. 1974. Feeding of four species of pelagic copepods under experimental conditions.

Mar. Biol., 25, 125-141. Grice, G. D ., M. R. Reeve, P. Koeller and D. W. Menzel. 1977. The use of large volume,

transparent, enclosed sea surface water columns in the study of stress on planktonic ecosys-tems. Helgolii.nder wiss. Meeresunters., 30, 118- 133.

Harbison, G. R. and R. W. Gilmer. 1976. The feeding rates of the pelagic tunicate Pegea con-jederata and two other salps. Limnol. Oceanogr., 21, 517-528.

Honjo, S. 1976. Coccoliths: Production, transportation and sedimentation. Mar. Micropaleont., 1, 65-79.

Honjo, S. and M. R. Roman. 1978. Marine copepod fecal pellets: production, preservation and sedimentation. J. Mar. Res., 36, 45-57.

Jorgensen, C. B. 1966. Biology of suspension feeding. Pergamon Press, Oxford. 357 pp. Krey, J. 1960. Der Detritus in Meere. J. Cons. Perm. Int. Explor. Mer., 26, 263-280. Lal, D. 1977. The oceanic microcosm of particles. Science, 198, 997-1009. Marshall, S. M. and A. P. Orr. 1955. On the biology of Ca/anus finmarchicus. Food uptake,

assimilation and excretion in adult and stage V Ca/anus. I . Mar. Biol. Assoc. U. K., 34, 495- 529.

Menzel, D. W. and J. J. Goering. 1966. The distribution of organic detritus in the ocean. Limnol. Oceanogr., 11, 333-337.

Paffenhofer, G.-A. and J. D. H. Strickland. 1970. A note on the feeding of Ca/anus he/go/andi-cus on detritus. Mar. Biol. , 5, 97-99.

Paffenhofer, G.-A. and S. C. Knowles. 1978. Feeding of marine planktonic copepods on mixed phytoplankton. Mar. Biol. , 48, 143-152.

Paffenhofer, G.-A ., D . Deibel, S. Bishop, W. M. Dunstan, and L. P. Atkinson. Biological obser-vations in Onslow Bay, 1976. (Ms).

Petipa, T. S., E. V. Pavlova, and G. N. Mironov. 1970. The food web structure, utilization and transport of energy by trophic levels in the planktonic communities, in Marine Food Chains, J. H. Steele, ed., Oliver and Boyd, Edinburgh, pp. 143-167.

Piontkovskiy , S. A. and T. S. Petipa. 1975. Selectiv i ty in the feeding of Acartia clausi. Biol. Morya, 33, 3-11 (Russian).

Poulet, S. A. 1976. Feeding of Pseudocalanus minutus on living and non-living particles. Mar. Biol. , 34, 117-125.

Raymont, J. E. G. and F . Gross. 1942. On the feeding and breeding of Ca/anus finmarchicus under laboratory conditions. Proc. Roy. Soc. Edinburgh, B, 61, 267-287.

Reeve, M. R. and M. A. Walter. 1977. Observati ons on the existence of lower threshold and upper critical food concentrations for the copepod Acartia tonsa. J. Exp. Mar. Biol. Ecol., 29, 211-221.

Reeve, M. R., M. A . Walter, K. Darcy and T. Ikeda. 1977. Evaluation of potential indicators of sub-lethal toxic stress on marine zooplankton: Controlled Ecosystem Pollution Experiment. Bull . Mar. Sci., 27, 105-113.

Ril ey, G. A. 1970. Particulate organic matter in sea water. Adv. Mar. Biol., 8, 1-118. Schrader, H.-J. 1971. Fecal pellets: Role in sedimentati on of pelagic diatoms. Science, 174,

55-57.

Small , L. F . and S. W. Fowler. 1972. Sinking rates of euphausiid fecal pellets. Limnol. Oceanogr., 17, 293- 296.

1979) Paffenhofer & Knowles: Ecology of fecal pellets 49

Smayda, T. J. 1969. Some measurements of the sinking rates of fecal pellets. Limnol. Oceanogr., 14, 621-625.

Smayda, T. J. 1971. Normal and accelerated sinking of phytoplankton in the sea. Mar. Geo!., 11, 105-122.

Stokes, G. G. 1901. Mathematical and physical papers, vol. III, Cambridge University Press, 413 pp.

Tenore, K. R., C. F. Chamberlain, W. M. Dunstan, R. B. Hanson, B. Sherr, and J. H. Tietjen. Possible effects of Gulf Stream intrusions and coastal runoff on the benthos of the continental shelf of the Georgia Bight, in IV. International Symposium on Estuaries, Estuarine Research Federation, Academic Press (In Press).

Turner, J. T. 1977. Sinking rates of fecal pellets from the marine copepod Pontella meadii. Mar. Biol., 40, 249-259.

Wiebe, P. H., S. H. Boyd and C. Winget. 1976. Particulate matter sinking to the deep-sea floor at 2000 m in the Tongue of the Ocean, Bahamas, with a description of a new sedimentation trap. J. Mar. Res., 34, 341-354.

Received: 13 June, 1978; revised: 26, October 1978.