The distribution and vertical flux of fecal pellets from large zooplankton in Monterey bay and coastal California

Deep-Sea Research I 94 (2014) 72–86

Contents lists available at ScienceDirect

Deep-Sea Research I journal homepage: www.elsevier.com/locate/dsri

The distribution and vertical flux of fecal pellets from large zooplankton in Monterey bay and coastal California Michael J. Dagg a,n, George A. Jackson b, David M. Checkley Jr.c a

Louisiana Universities Marine Consortium, 8124 Highway 56, Chauvin, LA 70344, USA Texas A&M University, College Station, TX 77843, USA c Scripps Institution of Oceanography, University of California, La Jolla, San Diego, CA 92093, USA b

art ic l e i nf o

a b s t r a c t

Article history: Received 27 January 2014 Received in revised form 26 August 2014 Accepted 3 September 2014 Available online 16 September 2014

We sampled zooplankton and fecal pellets in the upper 200 m of Monterey Bay and nearby coastal regions in California, USA. On several occasions, we observed high concentrations of large pellets that appeared to be produced during night-time by dielly migrating euphausiids. High concentrations of pellets were found in near-surface waters only when euphausiids co-occurred with high concentrations of large (4 10 μm) phytoplankton. Peak concentrations of pellets at mid-depth (100 or 150 m) during the day were consistent with the calculated sinking speeds of pellets produced near the surface at night. At these high flux locations (HI group), pellet concentrations declined below mid-depth. In contrast, at locations where the phytoplankton assemblage was dominated by small phytoplankton cells ( o10 μm), pellet production and flux were low (LO group) whether or not euphausiid populations were high. Protozooplankton concentrations did not affect this pattern. We concluded that the day and night differences in pellet concentration and flux in the HI profiles were mostly due to sinking of dielly-pulsed inputs in the surface layer, and that small zooplankton (Oithona, Oncaea), heterotrophic dinoflagellates, and bacterial activity probably caused some pellet degradation or consumption below 100 m. We estimated that consumption of sinking pellets by large copepods was insignificant. High fluxes of pellets were episodic because they required both high concentrations of large phytoplankton and large stocks of euphausiids. Under these conditions, flux events overwhelmed retention mechanisms, resulting in large exports of organic matter from the upper 200 m. & 2014 The Authors. Published by Elsevier Ltd. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/3.0/).

Keywords: Vertical flux Euphausiid Fecal pellets Retention efficiency

1. Introduction Fecal pellets from large zooplankton such as copepods and euphausiids can dominate the downward flux of organic matter in the ocean (e.g., Legendre and Michaud, 1998; Thibault et al., 1999; Roy et al., 2000; Vargas et al., 2002; Wexels Riser et al., 2008; Kobari et al. 2010; González et al., 2011). Many factors contribute to high variability in this flux component, including biological conditions such as the concentration, type and composition of food, the size and species composition of the zooplankton, and the predator to prey size ratio. Physical conditions such as stratification and mixing also affect fecal flux (Alldredge et al., 1987). Although much of the organic matter flux is derived from pellets, most pellets typically do not sink far (Wassmann, 1998) but are modified by microbial degradation (Hansen and Bech, 1996; Hansen et al. 1996) and zooplankton activities that include pellet

n

Corresponding author. Tel.: þ 1 206 201 3884. E-mail addresses: [email protected] (M.J. Dagg), [email protected] (G.A. Jackson), [email protected] (D.M. Checkley Jr.).

ingestion and destruction (Noji et al., 1991; Urban-Rich 1999; Poulsen and Iversen, 2008). Consequently, the dynamics of fecal pellet degradation significantly affect the dynamics of vertical flux (Turner, 2002). Zooplankton can modify sinking pellets by direct consumption (González and Smetacek, 1994; Huskin et al., 2004) or by breaking up the pellets into smaller, slower sinking particles (Noji et al., 1991). Breakage also enhances microbial access to pellet contents (Lampitt et al., 1990). Small ubiquitous copepods such as Oithona spp. and Oncaea spp. have been proposed as important modifiers of pellet flux, e.g. the “Oithona filter” (González and Smetacek, 1994; Svensen and Nejstgaard, 2003). Also, protozooplankton can be important consumers of pellets (Poulsen and Iversen, 2008; Poulsen et al., 2011). Euphausiids, which are strong diel migrators, have been observed to feed at depth during the day (e.g. Hamame and Antezana, 2010) where they may consume fecal pellets and thereby reduce flux. Alternatively, large zooplankton, especially swarming kinds like euphausiids and salps, have high filtration rates and produce large, rapidly sinking pellets; they can overwhelm degradation and recycling processes near the surface, resulting in high flux events.

http://dx.doi.org/10.1016/j.dsr.2014.09.001 0967-0637/& 2014 The Authors. Published by Elsevier Ltd. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/3.0/).

M.J. Dagg et al. / Deep-Sea Research I 94 (2014) 72–86

In addition to fecal pellets, other materials such as phytoplankton, mucous and gelatinous remains can aggregate to form major components of the vertical flux of organic matter (e.g. Silver and Gowing 1991; Silver et al. 1998; Jackson and Checkley, 2011). Flux is often in the form of aggregates of particles from a variety of sources, including fecal pellets, and all these materials are commonly observed at various levels of decomposition and degradation. We are attempting to measure the magnitude of zooplankton consumption of these sinking materials, a process we call “gatekeeper activity”, in the region immediately below the euphotic zone. Our goals are to determine the zooplankton contribution to the overall retention efficiency of pellets and other sinking organic matter, and to understand the conditions under which fluxconsumption by zooplankton is an important determinant of the flux profile (Petrik et al., 2013). In this paper, we present information on the size frequency distribution of large fecal pellets, operationally defined as those retained on a 60 μm mesh, in the upper 200 m during a 14-day study in or near Monterey Bay, California. Our specific objectives here are to show the vertical distribution of large fecal pellets and to determine some of the factors controlling the production and losses of these large pellets in the upper 200 m.

2. Material and methods 2.1. Study area and sampling We spent July 10–23, 2010 aboard the RV New Horizon in or just offshore to the west of Monterey Bay, CA (Fig. 1). Shipboard sampling was done near a free drifting SOLOPC float set to cycle through the upper 100 m at approximately hourly intervals. SOLOPC floats are fully described in Checkley et al. (2008), Jackson and Checkley (2011) and Petrik et al. (2013).

73

Aboard ship, whole water samples were collected for mesozooplankton fecal pellets using a simple method modified from that described in Wassmann et al. (1999) and Wexels Riser et al. (2001). For each collection, a CTD-rosette sampler fit with twelve 10-L Niskin bottles was used to collect water from six depths (Fig. 1, Table 1). In several cases, day and night samples at the same location were collected within a 24 h period (CTD 14/15, 25/26, 31/ 33 and 37/39). On deck, contents of Niskin bottles were gently drained into 10-L polycarbonate carboys. In the ship's laboratory, the entire content of each carboy, approximately 9.2 L, was filtered through a 60-μm mesh sieve. This mesh size was selected because smaller ones sometimes resulted in retention of too many contaminating particles for practical analysis of samples. Retained particles were backwashed into a small container and preserved in 5% Formalin–seawater solution. Final sample volume was between 150 and 250 ml. Preserved samples were kept in the dark and returned to a shore-based laboratory for analysis. Particles passing through the 60-μm sieve were not analyzed.

Table 1 Information for fecal pellet CTD casts. CTD

Date

Time

Latitude (oN)

Longitude (oW)

14 15 25 26 31 33 37 39 42 46 48

13-Jul 13-Jul 15-Jul 15-Jul 17-Jul 17-Jul 19-Jul 20-Jul 21-Jul 23-Jul 23-Jul

01:08 12:02 02:02 15:00 14:05 23:06 14:47 01:26 14:16 02:22 15:33

36.74 36.74 37.03 37.03 36.77 36.74 36.72 36.74 36.68 36.44 36.43

122.01 122.01 123.37 123.37 122.08 122.05 122.00 122.04 122.03 123.33 122.21

Latitude

Idaho

Longitude

Fig. 1. The study area, showing the sampling locations for zooplankton fecal pellets. Depth contours in m. Open symbols ¼ LO group, plus symbols ¼ HI group.

74

M.J. Dagg et al. / Deep-Sea Research I 94 (2014) 72–86

Immediately before or after each pellet cast, additional CTD-rosette casts were made for two sets of analyses. (1) Two size fractions of chlorophyll (410 μm, referred to as “large” phytoplankton and o10 μm, referred to as “small” phytoplankton) were determined on board using a cascade filtration system fit with a 10 μm polycarbonate filter and a GF/F glass fiber filter. (2) 100 ml samples from each depth were preserved in 5% acid Lugol's solution for later determination of microzooplankton concentrations. Zooplankton were sampled using a 1-m2 MOCNESS with nine 202-μm-mesh nets, deployed 13 times between about 100 m and the surface. Eleven of these MOCNESS deployments were coordinated with our CTD-rosette sampling. Zooplankton were preserved in 5% Formalin in seawater. The last two nets of some MOCNESS deployments were used to collect animals for gut pigment analysis. Aliquots were immediately concentrated on 34 mm diameter, 20 μm-mesh NITEX discs and frozen in liquid nitrogen. 2.2. Phytoplankton, zooplankton and gut pigment analyses On board ship, chlorophyll filters were extracted in 90% acetone for 24 h in the dark and the extract was measured with a Turner Designs fluorometer before and after acidification to determine chlorophyll a and phaeopigment concentrations (Strickland and Parsons, 1968). In a shore based laboratory, 10 ml subsamples of microzooplankton were settled for 24 h (Utermöhl, 1958) before counting. Microplankton cells including diatoms, dinoflagellates, aloricate ciliates, Mesodinium rubrum and tintinnids, were counted with an inverted microscope (Olympus-IX 51) at 200  magnification. In most subsamples, between 100 and 500 individuals were counted but deep water at some stations contained o100 cells in the 10 ml aliquot. Cell size of each unique cell type was measured using an eyepiece micrometer to calculate biovolume, then converted to biomass. Microplankton were assumed to be standard geometric forms (Hillebrand et al., 1999). Cell biomass was calculated using a volume-to-carbon conversion factor of 0.19 pg C μm  3 for oligotrich ciliates, M. rubrum (Putt and Stoecker, 1989) and dinoflagellates (Howell-Kübler et al., 1996). For tintinnids, C (pg cell  1) was determined from the equation 444.5 þ0.053 lorica volume (μm  3) (Verity and Langdon, 1984). Diatom biomass was calculated according to the formula of Menden-Deuer and Lessard (2000). Ashore in the laboratory, each zooplankton sample was subsampled (split) by half multiple times. Entire subsamples were analyzed so that least 30 zooplankters 4 1 mm equivalent spherical diameter (ESD) of two taxonomic groups, large copepods and euphausiids, were identified, enumerated and measured (width and length, of the cephalothorax for copepods and overall for euphausiids) and remaining individuals of each taxon identified and enumerated. Identification was to species for large copepods and, for euphausiids, always to family and occasionally to species. Abundance (number m  3) and volume (cm3 m  3) of these plankton types were calculated from abundance and size of individuals in a subsample, subsample fraction and the volume filtered, based on measurements using a calibrated flow meter (Checkley, unpublished). In a shore based laboratory, individual euphausiids were sorted and analyzed for gut chlorophyll and phaeopigment contents, using well established methods (e.g. Dagg 1993; Dagg et al., 1997). Duplicate subsamples from each net were analyzed.

and non-pellet detritus were removed from each petri dish by pipette and the remaining particles, almost entirely fecal pellets, were photographed. Each 34-mm petri dish containing pellets was uniformly backlit by an LED light source (Advanced Illumination Model No BL0404-WHIIC) and photographed with a Canon EOS 40D camera fit with a EFS 60 mm macro lens. The light source was a white LED with a continuous output spectrum but heavily weighted in the blue wavelengths. The camera has a red, a green and a blue sensor, resulting in three images. We used the blue images for our analyses. During photography, the focal distance was the same for all samples and set so that the entire petri dish was included in a single image. The resulting images were saved as 3888  2592 pixel images in JPEG format. Each pixel was 14.7  14.7 μm. As an example, some pellets from CTD 31–100 m are shown in Fig. 2. The Matlab image analysis function regionprops was used to locate and size particles in the images. The function determined the length of the major axis and the area for each object; pellet length L was estimated as the length of the major axis and the pellet diameter d was estimated by dividing the object area by L. Volume V was calculated assuming each fecal pellet was a cylinder, V ¼0.25πd2L. For each particle, an equivalent spherical diameter (desd) was calculated for the volume-equivalent sphere: 6 desd ¼ V 1=3

π

We also visually examined the images of the 10 largest particles in each sample because large particles contribute disproportionately to flux. In samples with high concentrations of pellets, the imaging program sometimes artificially combined several pellets that were in close proximity or overlapped slightly, and treated them as a single, extremely large, pellet. These were removed from further analysis. In addition, a few long thread-like contaminating particles present in some samples were incorrectly categorized by the software as large pellets. These were also removed from further analysis. The normalized volume distribution, nVdesd, was determined after calculating the number distribution n as in Jackson and Checkley (2011) using volume-doubling size ranges. Units of nVdesd in this paper are parts per billion (1 ppb ¼10  9 cm3 cm  3 ¼ 1 mm3 m  3).

2.3. Fecal pellet analysis In the laboratory, each sample was examined with a dissecting microscope. All fecal pellets were removed with a fine-bore pipette and transferred to a 34-mm diameter plastic petri dish. Contaminating particles such as small zooplankton, nauplii, eggs,

Fig. 2. Image of some fecal pellets from the 100 m sample, CTD 31.

M.J. Dagg et al. / Deep-Sea Research I 94 (2014) 72–86

Settling speed, vs (m d  1), for each pellet was calculated using the relationship of Komar et al. (1981) for L and d expressed in cm vs ¼ 1:21  103 L2 ðL=dÞ  1:664

1.0 euphausiids copepods

biovolume (cm3 m-3)

0.8

0.6

0.4

0.2

0.0 0

10

20

30

40

50

60

MOCNESS net sample Fig. 3. Biovolume of large copepods (open circles) and euphausiids (filled circles) in all MOCNESS samples examined for this study.

75

The carbon content of fecal pellets was calculated from pellet volume assuming a carbon to volume ratio of 55 μg C mm  3, derived from direct measurements made by González et al. 1994; Daly, 1997; Urban-Rich et al., 1998; and Riebesell et al., 1995. Their measurements ranged from 38 to 69 μg C mm  3. Fecal pellet flux (mg C m  2 d  1) was calculated by summing the product of carbon content and settling velocity for each particle in a sample and dividing by sample volume. The 95% confidence limits of the total carbon content and the flux estimates were calculated using a bootstrap analysis performed with the Matlab bootci function with the bias corrected and accelerated method options. There are several uncertainties associated with our determination of pellet volume concentration and flux: (a) Pellets may be broken up during collection and sieving, resulting in apparent sizes that are smaller than in situ sizes. This would not effect the volume concentrations but would lead to underestimation of settling velocity and flux. (b) Image analysis can lead to incorrect characterization of particles. We surveyed each image to remove misidentifications of several pellets as a single large pellet. This reduced the total volume concentrations by small amount. Other possible errors with image analysis, such as misinterpreting a broken or bent pellet as a larger sized pellet, were not addressed but visual examination of the images did not show many such particles. (c) The carbon:volume ratio of pellets varies with food type and density, whereas we applied a single

Fig. 4. The size frequency distribution of fecal pellets from CTD cast 42. X¼ length in pixels and d¼ diameter in pixels. A pixel is 14.7 μm. Z¼ pellet abundance (number m  3). All six depths from this CTD cast are shown: 5, 25, 50, 100, 150, and 200 m.

76

M.J. Dagg et al. / Deep-Sea Research I 94 (2014) 72–86

nVd(ppb)

Cast 42 5 m

Cast 42 25 m

150

150

100

100

50

50

0 −3 10

−2

10

−1

10

0 −3 10

nVd(ppb)

Cast 42 50 m 150

100

100

50

50

−2

10

−1

10

0 −3 10

nVd(ppb)

Cast 42 150 m 150

100

100

50

50

−2

10

−2

10

−1

10

Cast 42 200 m

150

0 −3 10

−1

10

Cast 42 100 m

150

0 −3 10

−2

10

−1

10

desd(cm)

0 −3 10

−2

10

−1

10

desd(cm)

Fig. 5. The normalized volume distribution (nVd) spectra for fecal pellets from CTD cast 42.

literature-derived ratio to all our samples. (d) Settling speed varies with water temperature (viscosity). The equation we used, from Komar et al. (1981), was derived for a temperature of 13 1C whereas temperatures in our study ranged from about 14 1C at the surface to 8 1C at 200 m. Most of the in situ sinking velocities were therefore slightly lower than we calculated. This error is small relative to other uncertainties. Overall, most of our errors should be systematic because our sampling conditions, measurements and analytical protocols were applied consistently. As a result, the greatest uncertainties in our study should be with the specific values for volume concentration and flux rather than with the patterns we observe. 3. Results 3.1. Mesozooplankton Large zooplankton included Calanus pacificus C5, C6F and C6M, Metridia pacifica C6F, Eucalanus californicus C5, C6F and C6M, and

several other less abundant copepods. However, these large copepods contributed little to the total zooplankton biomass whenever euphausiids were present. All samples with a high biovolume of large zooplankton were dominated by euphausiids (Fig. 3), primarily Euphausia pacifica. 3.2. Sample pellet profile The abundance and size distribution of fecal pellets in a sample CTD cast, CTD 42, are shown in Fig. 4. Pellet abundance and size changed with depth. There were few pellets at 5 m but abundance increased with depth to 100 m and then declined somewhat at 150 and 200 m. Pellet length and diameter also peaked at 100 m. At 50 m and shallower, most pellets were 5 pixels (74 μm) in diameter or smaller; deeper they were as wide as 10 pixels (147 μm). In the 150 and 200 m samples, pellets greater than 30 pixels (441 μm) long were mostly absent, indicating large pellets decreased below 100 m in this profile.

M.J. Dagg et al. / Deep-Sea Research I 94 (2014) 72–86

−2

Fecal C flux (g−C m

Depth(m)

0

0.1

0.2

0.3

0.4

0.5

0.6

−1

d ) 0.7

0.8

0.9

1

0

0

20

20

40

40

60

60

80

80

100

100

120

120

140

140

160

160

180

180

200

200 0

10

20

30

40

50

60

70

80

90

100

Pellet volume concentration (ppb) Fig. 6. The total volume (ppb) and vertical flux (mg C m  2 d  1) of pellets at each sampled depth in CTD cast 42.

The size distribution and abundance patterns from CTD 42 (Fig. 4) are more clearly seen when the abundance and size data are presented as nVd distributions (Fig. 5). Distribution patterns were fairly uniform at all depths and the peak at all depths was at an equivalent spherical diameter (desd) between 167 and 210 μm, indicating the size distributions of pellets were similar but not identical at all depths. In this case, there was a slight maximum desd (210 μm) at 100 m, confirming the visual impression from Fig. 4 that the abundance of larger pellets was greatest at 100 m and then decreased slightly in deeper samples. The area under each nVd curve, representing the integrated volume concentration of pellets, varied widely with depth (Fig. 6). The minimum concentration was in the 5 m sample ( 10 ppb) and the maximum concentration was in the 100 m sample ( 100 ppb) (Fig. 6). The vertical pattern of pellet flux closely tracked the profile of pellet concentration in Cast 42 (Fig. 6) because size composition of pellets did not vary much with depth. In summary, total pellet concentration and flux increased from the near-surface to 100 m and decreased below 100 m in this profile. 3.3. Pellet profiles Peak diameter, total volume concentration, and total flux were used to characterize all 11 pellet profiles (Table 2). There was no consistent pattern of changing pellet size with depth among pellet profiles, as indicated by the diameter at maximum volume concentration (nVd) (Table 2). Thus, flux was mostly determined by the volume concentration, and the depth profiles of fecal pellet flux were similar to the profiles of pellet volume concentration for each cast (Fig. 7). Although pellet volume concentration and flux varied greatly among the eleven pellet casts, profiles fell into a group of low volume/flux casts (Group LO—Casts 14, 15, 25, 26 and 46) and a group of high volume/flux casts (Group HI—Casts 31, 33, 37, 39, 42 and 48) (Fig. 7, Table 2). In the LO group, concentration was always o20 ppb and was approximately uniform with depth (Fig. 7(a)). In contrast, in the HI group concentration was typically 450 ppb and these profiles had substantial vertical structure (Fig. 7(b)). In both LO and HI groups, vertical distribution of volume concentration was reflected in flux of pellet material (Fig. 7(c) and (d)). The estimated carbon flux from fecal pellets ranged from  0.04 to  0.15 gC m  2 d  1 in the LO group and  0.05 to 5.0 gC m  2 d  1 in the HI group (Table 2).

77

Average sinking velocities estimated from pellet size are shown in Fig. 8. The range was about the same in both the LO and HI groups, approximately 50–250 m d  1, again indicating that flux was mostly determined by the concentration of pellets. Some of the factors causing the pellet profiles to sort into HI and LO categories were identified by comparing pellet volume or flux with the concentration and size composition of phytoplankton, and the distribution and abundance of large zooplankton. 3.4. HI group CTD casts 31 and 33 (HI group) were a day (31) and night (33) pair from inside Monterey Bay. Both had high pellet volume concentrations (Fig. 7(b)). Pellet concentration was higher in the surface at night than in the day and there was a large mid-depth (100 m) peak in pellet concentration during the day (Fig. 7(b)). Estimated sinking speeds for pellets in the upper 50 m at night were about 120–150 m d  1, equivalent to about 60–75 m in 12 h. These speeds are consistent with the observed profiles, suggesting that pellets produced in the near-surface strata at night were sinking to mid-depths by mid-day. MOCNESS deployments 7 and 8 were associated with CTD 31 and 33, and show euphausiid concentrations at night were much higher than during the day, with a nighttime maximum in the surface stratum (2–41 m) (Fig. 9(a)). Daytime euphausiid biovolume was low at all sampled depths (Fig. 9(a)). Chlorophyll concentrations, measured from CTD casts made immediately after the pellet CTD casts, were between 2 and 3 μg l  1, and the phytoplankton assemblage was dominated by large cells, with highest concentrations at  20 m (Fig. 9(b)). A similar pattern of pellet concentrations was observed in another day (CTD 37) /night (CTD 39) pair of pellet profiles from the HI group (Fig. 7(b)). Here also, there was a strong diel migration of euphausiids (Fig. 9(c)), the phytoplankton assemblage was dominated by large cells (Fig. 9(d)), and the observed daytime peak in pellet concentration at mid-depth is consistent with the estimated sinking speeds of these pellets. Based on these two day-night series, we conclude that high concentrations of large phytoplankton cells and large stocks of euphausiids at night led to high production of pellets near the surface at night, and that pellets sinking from the surface were observed as a daytime peak in pellet concentration at mid-depth s of 100–150 m. This day-night pattern of euphausiid and fecal pellet distributions is further but weakly supported by another pellet profile from within Monterey Bay, Cast 42. Pellet concentration in the daytime was low at the surface and peaked at mid-depths (Fig. 7(b)). Chlorophyll concentration was high at night and dominated by large cells although daytime chlorophyll concentration was lower and large cells were less abundant (Fig. 9(f)). Euphausiid biovolume during the day was low in all six sampled layers (Fig. 9(e)). We do not have an associated night profile of pellets or zooplankton. Nevertheless, this incomplete set of data is consistent with characteristics of the HI group. 3.5. LO group The LO group did not show large peaks in the vertical profiles of pellet concentration (Fig. 7(a)). For example, CTD casts 14 and 15 were a night (14) and day (15) pair from within Monterey Bay. Both had low pellet volumes (o 20 ppb) at all depths. Euphausiids were abundant at this location (MOCNESS 1 and 2), especially at night (Fig. 10(a)). The maximum biomass of euphausiids at night was in the 16–30 m stratum, although there was also a layer of euphausiids in the 17–30 m daytime sample. Chlorophyll concentrations were moderately high, 1–2 μg l  1 (Fig. 10(b)), but large cells were not abundant. Similar patterns were observed in a pair of pellet profiles in the LO group from offshore, CTD 25 (night) and

78

M.J. Dagg et al. / Deep-Sea Research I 94 (2014) 72–86

Table 2 Fecal pellet characteristics at each CTD cast: diameter at the maximum in volume concentration (D at max nVd), volume concentration of all pellets (ppb), and flux of all pellets at each sample depth (mg C m  2 d  1). (D) and (N) in CTD numbers signify Day and Night. Profile type

CTD

Depth (m)

D (μμm) at max nVd

Volume (ppb)

95% CI

Flux (mgC m  2 d  1)

95% CI

HI

31(D)

5 25 50 100 150 200

132 167 210 210 210 265

10 46 214 305 132 23

8–12 39–55 196–235 276–340 118–150 18–30

47 417 2330 4780 1352 161

34–67 324–566 1976–2827 3836–6039 1113–1763 113–242

HI

33(N)

5 25 50 100 150 200

167 167 265 167 167 167

60 45 79 97 125 47

52–70 38–56 67–92 88–109 113–138 41–54

460 412 894 733 1084 310

356–673 310–661 710–1145 608–1136 914–1314 252–394

HI

37(D)

5 25 50 100 150 200

420 210 210 132 132 132

15 26 146 170 145 60

11–27 20–37 132–163 156–189 134–164 55–67

178 299 1387 1289 1043 342

76–622 181–539 1140–1791 1070–1781 796–1983 289–418

HI

39(N)

5 25 50 100 150 200

167 167 167 167 132 167

126 176 21 77 85 42

115–137 164–189 16–30 69–88 76–94 38–48

830 1151 233 646 600 224

713–983 1007–1347 119–580 529–811 504–760 188–305

HI

42(D)

5 25 50 100 150 200

167 167 167 210 167 167

10 29 44 96 56 34

8–14 25–34 38–53 87–107 49–65 30–42

61 162 333 731 385 238

39–95 127–230 245–492 630–867 303–590 165–476

HI

48(D)

25 50 100 150

132 132 132 167

49 12 13 5

43–57 10–14 11–16 4–7

320 49 55 18

246–431 36–73 39–86 12–29

LO

14(N)

5 20 40 60 170 200

265 132 167 132 132 132

7 16 7 7 18 6

4–13 12–24 6–10 6–9 15–21 5–8

67 143 31 31 74 21

22–191 72–296 20–51 22–45 58–97 14–40

LO

15(D)

5 20 40 60 120 200

265 105 132 210 132 105

4 16 19 13 18 16

3–8 14–19 16–24 11–17 15–21 14–18

30 57 109 77 83 55

9–127 45–75 79–167 53–115 64–109 45–67

LO

25(N)

25 50 100

167 210 167

13 17 12

10–18 13–23 8–19

96 148 159

61–168 102–243 84–363

LO

46 (N)

5 25 50 100 200

210 334 132 265 210

3 5 4 5 7

2–4 2–12 3–6 3–10 5–10

9 54 17 38 49

5–19 19–198 11–27 18–96 25–95

CTD 26 (day), although pellet data were limited (Figs. 7(a), 10(c) and 10(d)). Even though there were high stocks of euphausiids in the surface layers at night, pellet concentrations were low, probably because there were low concentrations of large phytoplankton cells. We conclude that these locations provided a poor feeding environment for euphausiids and, although dielly migrating, they produced few pellets, resulting in low pellet concentrations. Low pellet concentrations and fluxes were also observed at CTD 46 (night), another offshore CTD cast (Fig. 7(a)). Euphausiid biovolume was high at night and low during the day, indicating diel vertical migration (Fig. 10(e)). Chlorophyll concentrations

were low and the phytoplankton assemblage was dominated by small cells (Fig. 10(f)). These were poor conditions for pellet production and, in spite of a large population of euphausiids, pellet concentrations and fluxes were low, consistent with the pattern seen in the other LO profiles. The remaining pellet profile, from CTD 48 (day), did not fall clearly into either the HI or LO group but had some HI characteristics in that pellet volume and flux were high at one depth (25 m) (Fig. 7(b)). In common with other members of the HI group, chlorophyll concentrations were high and dominated by large cells (data not shown). However, in common with the LO group,

M.J. Dagg et al. / Deep-Sea Research I 94 (2014) 72–86

Depth(m)

LO casts

HI casts

0

0

20

20

40

40

60

60

80

80

100

100

120

120

140

140

160

31 33 37 39 42 48

160

14 15 25 46

180

180 200

200 0

Depth(m)

79

10 20 Pellet vol. conc. (ppb)

30

0

0

0

20

20

40

40

60

60

80

80

100

100

120

120

140

140

160

14 15 25 46

180

100 200 Pellet vol. conc. (ppb)

300

31 33 37 39 42 48

160 180 200

200 0

0.1

0.2

0.3

Fecal C flux (g−C

0.4

m−2 2

d−1)

0

1

2

3

4

5

Fecal C flux (g−C m−2 d−1)

1

Fig. 7. (a) and (b) Concentration (ppb by volume) and (c)(d) flux (g C m d ) of pellets, at each depth for LO and HI casts. Note 10  difference in scales for LO and HI casts. Photographs of the few pellets from some of the samples from CTD 25 and all of samples from CTD 26 were inadequate for processing.

average sinking rates of these pellets were low, the lowest observed during the cruise (Fig. 8), and there was no daytime peak in pellet concentration at mid-depths (Fig. 7(b)). Daytime euphausiid concentrations were low, the maximum was only 0.004 cm3 m  3. The potential for euphausiid pellet production at night cannot be ascertained because there are no companion nighttime zooplankton data. Thus, the supporting data are insufficient to provide a clearer categorization and this case falls between the HI and LO sites. 3.6. Gut pigments Characteristics of the phytoplankton assemblage in the HI and LO groups are reflected in the gut pigment concentrations of euphausiids (Euphausia pacifica and Thysanoessa spinifera). For example, levels of gut phaeopigment were low in euphausiids from a LO location, ranging from near 0 to about 100 ng phaeopigment individual  1 in the surface layer (Fig. 11(a)), and from 0 to about 15 ng phaeopigment individual  1 in the intermediate depth stratum (Fig. 11(b)). In contrast, gut pigment levels in euphausiids collected a HI location

were at least 10 times higher, ranging between about 150 and 1500 ng phaeopigment individual  1 in the surface stratum (Fig. 11(c)) and about 40 and 400 ng phaeopigment individual  1 in the subsurface stratum (Fig. 11(d)). The strong relationship between gut pigment level and body size observed at night (Fig. 11(c)) may indicate that conditions at this location supported maximum feeding and pellet production rates. These observations support the idea that the high concentrations of large phytoplankton in HI locations result in high feeding and pellet production rates, compared to the much lower rates of feeding and pellet production in the LO locations, where large phytoplankton cells were rare.

4. Discussion Fecal pellets large enough to contribute significantly to the vertical flux of particulate organic matter in the ocean are derived from several groups of zooplankton. Large, wafer-shaped pellets are generally produced by salps and doliolids, elliptical pellets by appendicularians, cylindrical pellets by copepods, and long strands

80

M.J. Dagg et al. / Deep-Sea Research I 94 (2014) 72–86

HI casts

Depth(m)

LO casts 0

0

20

20

40

40

60

60

80

80

100

100

120

120

140

140

160

160 14 15 25 46

180 200

31 33 37 39 42 48

180 200

0

100

Ave. velocity(m

200

d−1)

0

100

200

300

Ave. velocity(m d−1)

Fig. 8. Calculated average sinking velocity (m d  1) of fecal pellets from all casts and sampling depths.

by euphausiids (Caron et al., 1989; González, 1992; González et al. 1994; Yoon et al., 2001; Wexels Riser et al., 2007). Euphausiid feces are sometimes referred to as ‘strands’ or ‘strings’ (e.g. González et al., 2011) because they are not completely enclosed in membranes but we use ‘pellets’ here for simplicity. These broad categorizations of shape are not absolute because pellet characteristics depend on several additional factors such as food type and food concentration (Dagg and Walser 1986; Tsuda and Nemoto, 1990; Turner, 2002). Pellets in our samples were cylindrical (Fig. 2) and mostly ranged in volume from about 0.6  106 μm3 to 12.6  106 μm3. In comparison, large copepods in the size ranges we observed in our study produce pellets between about 0.3 and 1.5  106 μm3 (Dagg and Walser, 1986). The size and shape characteristics of pellets in our study suggest they were derived almost entirely from euphausiids. Consistent with this, we found that, in night-time samples, euphausiids were typically about as numerically abundant as large copepods and heavily dominated the zooplankton biomass whenever there was a large biomass of net-caught zooplankton. During our study, euphausiids dielly migrated to the surface layers at night, a commonly observed behavior (e.g. Stuart and Pillar, 1990; Bollens et al. 1992; Nakagawa et al., 2003; Antezana, 2010; Liu and Sun, 2010). Comparison of our day and night MOCNESS samples indicated that almost all of the euphausiid population was below our deepest sampling depth of about 100 m during the day, and highly concentrated near the surface at night. Although diel vertical migration is primarily a predator avoidance behavior, it has been shown that the pattern can be modified somewhat by the food environment (e.g. Dagg, 1985; Dagg et al., 1997). In our study, we observed two occasions when some of the euphausiid population remained near the surface during the day and both occurred under poor feeding conditions. A large population of euphausiids in near-surface waters at night was not sufficient to result in high pellet production. In our study, phytoplankton size structure was the most important characteristic distinguishing high-flux (HI) from low-flux (LO) locations. When the surface phytoplankton assemblage was

dominated by small cells and had few large cells, pellet concentrations and fluxes were low, regardless of euphausiid concentrations. These euphausiids are inefficient feeders on small cells, and also the concentrations of small cells were comparatively low which would result in low ingestion rates and thus low production rates of fecal pellets. High pellet production and flux were only observed when there was a large population of dielly migrating euphausiids and a food environment with high concentrations of large phytoplankton. This was the case for D/N CTD casts 31/33, where euphausiid stocks were high at night and the phytoplankton assemblage was dominated by large cells. The resulting pellet flux was high, reaching almost 5000 mg C m  2 d  1 at 100 m. In contrast, N/D CTD casts 14/15 had large euphausiid stocks near the surface at night but pellet flux was low,o 200 mg C m  2 d  1, because the phytoplankton assemblage was dominated by small cells. At this location, there were even euphausiids remaining near the surface during the day but pellet production and flux remained low in that poor feeding environment. To summarize, high rates of pellet production occurred only at night when euphausiids and large phytoplankton cells co-occurred, resulting in a nighttime pulse of near-surface pellet production, measured by high concentrations of pellets in near-surface layers. These pellets, sinking at speeds of 50–250 m d  1 (25–125 m half-day  1), resulted in mid-depth peaks (usually in the 100 m sample) in pellet concentrations in the daytime. Diel pulses of zooplankton pellet production and flux have been observed in several other studies (e.g Lorenzen and Welschmeyer, 1983; Welschmeyer et al., 1984; Bathmann et al., 1991; González et al., 1994). The episodic nature of krill pellet fluxes has also been noted in the Southern Ocean (Le Fèvre et al., 1998). We assume that almost all pellet production occurred within the depth layer of feeding. Pellet production likely began shortly after the upward migration and onset of feeding by euphausiids because the passage of food through their guts would take only approximately 20–60 min at our ambient near-surface temperatures (  12–14 1C) (Dam and Peterson, 1988; Hamame and

M.J. Dagg et al. / Deep-Sea Research I 94 (2014) 72–86

81

0

0 20

40

40 80

60

N D

100 120 0.00

0.05

0.10

0.15

0.20

0.25

160 0.30

0.0

0.5

1.0

1.5

2.0

2.5

3.0

0

0 20 Depth (m)

N > 10 µm N < 10 µm D > 10 µm D < 10 µm

120

80

40

40 80

60

N D

100 120 0.00

N > 10 µm N < 10 µm D > 10 µm D < 10 µm

120

80

0.15

0.30

0.45

0.60

0.75

160 0.90

0

2

4

6

8

0

0 20

40

40 80

60

D

100 120 0.00

N > 10 N < 10 D > 10 D < 10

120

80

160 0.05

0.10

0.15

0.20

Euphausiid biovolume

(cm3

0.25

0.30

m-3)

0.0

0.5

1.0

1.5

2.0

Chlorophyll concentration (µg

µm µm µm µm

2.5

3.0

L-1)

Fig. 9. Zooplankton (a) and phytoplankton (b) associated with HI casts 31/33: (a) euphausiid biovolume during the day (open circles) and night (filled circles) and (b) phytoplankton concentrations (solid symbols represent cells 4 10 μm and open symbols represent cells o10 μm); (c) and (d) are the same as (a) and (b) but for HI casts 37/39; (e) and (f) are the same as (a) and (b) but for HI cast 42. For all these casts, surface temperatures were approximately 14 1C, temperatures at the top of the thermocline (40–60 m) were about 10 1C and temperatures at 200 m were about 8 1C. Surface salinities were about 33.6 and were similar or slightly higher, 33.7, at the top of the halocline (40–60 m), and about 34 at 200 m.

Antezana, 2010). Numerous pellets would be produced by each individual euphausiid during the dark hours between sunset to sunrise, about 10 h. Upon the onset of morning descent and cessation of feeding on surface phytoplankton, pellet production by gut evacuation may continue for a short while but this would contribute little compared to the total nighttime production. Other food sources such as protozooplankton and detritus may have contributed to fecal pellet production in our study because euphausiids in Monterey Bay and nearby offshore regions are primarily Euphausia pacifica, a species known to be omnivorous (Ohman, 1984; Nakagawa et al., 2002). Protozooplankton abundance, size and species composition were similar between HI and LO sites (Fig. 12) suggesting that protozooplankton would comprise a larger fraction of the diet under LO conditions where the phytoplankton component of the diet was small, than under HI conditions where the phytoplankton component of the diet was

large. However, these conditions, which presumably would result in a large protozoan component in the diet at LO sites, did not result in high fecal pellet concentrations. Pellet concentrations were only high when large phytoplankton dominated the particulate environment. The apparent declines in pellet concentration below 100 m under HI conditions could be simply due to sinking of a pulsed input. Falling at speeds of 100–250 m d  1, pellets produced near the surface over 10 h by nocturnally feeding euphausiids would form a sinking layer 40–100 m thick. Under HI conditions in our study, a pulse of pellets produced near the surface at night by dielly migrating euphausiids was observed at about 100 m after half a day, consistent with estimated sinking speeds. If sinking continued at this rate, most of these pellets would be below our deepest sampling depth after 24 h, and we would not see them. The nightly pulsed nature of pellet production combined with

82

M.J. Dagg et al. / Deep-Sea Research I 94 (2014) 72–86

0

0

20 40 40 80

60 80

120 N D

100 120 0.00

0.15

0.30

0.45

0.60

0.75

160 0.90

0

0.0

0.5

1.0

µm µm µm µm µm µm

1.5

2.0

0

20

Depth (m)

N > 10 N < 10 D > 10 D < 10 D > 10 D < 10

40

40

120

80 N D

100 120 0.00

N > 10 µm N < 10 µm D > 10 µm D < 10 µm D > 10 µm D < 10 µm

80

60

0.05

0.10

0.15

0.20

0.25

160 0.30

0

0.0

0.5

1.0

1.5

2.0

0

20 40

40 60

80

80 120

N D

100 120 140 0.00

N > 10 µm N < 10 µm

160 0.05

0.10

0.15

0.20

0.25

0.30

Euphausiid biovolume (cm3 m-3)

0.0

0.5

1.0

1.5

Chlorophyll concentration (µg

2.0

L-1)

Fig. 10. Zooplankton (a) and phytoplankton (b) associated with LO casts 14/15: (a) euphausiid biovolume during the day (open circles) and night (closed circles) and (b) phytoplankton concentrations (solid symbols represent cells410 μm and open symbols represent cellso10 μm); (c) and (d) are the same as (a) and (b) but for LO casts 25/26; (e) and (f) are the same as (a) and (b) but for LO cast 46. Surface temperatures for casts shown in panel (b) were about 13 1C, temperatures at the top of the thermocline (25–30 m) were about 10 1C and at 200 m were about 9 1C. Surface salinities were about 33.5, increasing to about 34 at 200 m, with no halocline. Surface temperatures for casts in panel (d) were about 14 1C, temperatures at the top of the thermocline (40–45 m) were between 9–12 1C, and at 200 m were about 8 1C. Salinities between the surface and the halocline (about 45 m) were about 33.4, increasing to about 34.1 at 200 m. For casts in panel (f), temperatures between the surface and the top of the thermocline (about 35 m) were about 14 1C and declined to about 7.5 1C at 200 m. Salinities between the surface and the top of the halocline about 35 m) were about 33.5 and increased to about 33.9 at 200 m.

rapid sinking of pellets could explain most or all the observed vertical pattern in pellet concentration and flux. Nevertheless, we considered the possibility that some pellet consumption or degradation may have occurred during sinking. We did not observe strong evidence for pellet degradation or loss over the depth range of our study, the upper 200 m. Generally, the proportion of pellet production that is degraded or lost during sinking is highly variable and is affected by pellet size, pellet content, sinking speed (i.e. time to pass through the upper layers), temperature, stocks and types of heterotrophs that may disrupt or consume pellets, and other factors (Legendre and Michaud, 1998; Urban-Rich et al., 1999; Viitasalo et al., 1999; Buesseler et al., 2007). One possible indicator of pellet destruction during sinking is a reduction in pellet size with depth but we did not observe any consistent decrease in pellet size in either HO or LO profiles. The volume concentration of pellets in LO profiles did not decrease with depth, indicating no degradation of pellets took

place during sinking. However, volume concentration typically decreased below mid-depths in the HI profiles. This may indicate degradation or consumption, and it is worth considering large copepods, small copepods, protozooplankton, and bacteria as possible contributors. Large copepods did not appear to have an impact on pellet flux. Average concentrations of all large copepods below 50 m at HI locations were o20 ind m  3 (Checkley unpublished). Assuming a generous clearance rate of 10 ml h  1 for each individual results in an estimated community clearance of o200 ml h  1 m  3 ¼4.8  10  3 d  1 for all large copepods combined. Applying this small impact to a pulse of pellets sinking at 100–250 m d  1 ( 4–10 m h  1) results in a removal of o1 % of the sinking pellets over the estimated 12 h it would take for these pellets to sink from 100 to 200 m. From this rough estimate, it appears that large copepods were simply not abundant enough to be important consumers of sinking pellets during our study.

M.J. Dagg et al. / Deep-Sea Research I 94 (2014) 72–86

160

1600 MOC 9 (N) 31-0 m MOC 10 (D) 35-0 m

MOC 1 (N) 32-0 m MOC 2 (D) 29-0 m

ng Gut phaeopigment ind-1

83

120

1200

80

800

40

400

0

0 5

10

15

20

5

160

10

15

20

1600 MOC 1 (N) 71-32 m MOC 2 (D) 65-29 m

MOC 9 (N) 62-31 m MOC 10 (D) 88 - 35 m

120

1200

80

800

40

400

0 0

5

10

15

0

20

5

10

15

20

Body length (mm) Fig. 11. Gut phaeopigments (ng individual  1) in mixed Euphausia pacifica and Thysanoessa spinifera from MOCNESS hauls associated with LO CTD casts 14/15 (a is the upper layer) (b is the subsurface layer), and HI CTD casts 37/39 (c is the upper layer) (d is the subsurface layer). N ¼night cast, D¼day cast.

Small zooplankton, Oithona and Oncaea, have sometimes been considered important consumers of sinking pellets (González and Smetacek, 1994; Svensen and Nejstgaard, 2003), although more recent studies have not supported this role for the ubiquitous Oithona species (Reigstad et al., 2005; Iversen and Poulsen, 2007). In our study, these copepods combined, analyzed from six of our MOCNESS hauls, had a maximum concentration of only 515 m  3 in a sample near the surface. Concentrations at mid-depths were o100 m  3. These concentrations are consistent with those from an extensive survey in Monterey Bay spanning more than two years (Hopcroft et al., 2002) that showed Oithona similis concentrations ranged between about 100 and 1000 m  3 and Oncaea spp. concentrations between about 1 and 100 m  3. It is likely that both data sets underestimate the abundance of these small zooplankton because the mesh size of the collecting nets, 202 μm, was too large to quantitatively capture them, especially the smaller copepodid stages, (Paffenhöfer and Mazzocchi, 2003). Nevertheless, assuming a combined O. similis þOncaea spp. concentration below 50 m of 1000 m  3, and a generous clearance rate of 2 ml cop  1 h  1 (e.g. Svensen and Kiørboe, 2000; Lonsdale et al., 2000; Nishibe et al., 2010), results in the removal of pellets from only 24 L m  3 over 12 h (2.4% 12 h  1 ¼0.048 d  1). Although this is an order of magnitude greater than the estimated consumption by large copepods, it is a modest contribution to loss processes for pellets sinking at 50–125 m 12 h  1and it would not appear that these small copepods can contribute much to degradation and loss of these fast sinking pellets. Oithona similis however, is known to be an ambush feeder and this behavior is consistent with the concept of flux feeding (Jackson, 1993; Dagg, 1993). Assuming an individual

‘scans’ and removes sinking particles from an area of 1 mm by 1.5 mm (0.015 cm2) results in the continuous removal of sinking particles by O. similis from 15 cm2 over a 1 m2 area (0.15%). If this process operated over 50 or 100 m below the particle maxima, the resultant removal of sinking particles would be 7.5 or 15.0%. These rough calculations suggest it is possible for small copepods, especially Oithona, to contribute to removal of sinking particles under HI conditions in our study. Smaller copepodid stages that would not have been sampled by our nets would add an unknown amount to this consumption. Protozooplankton (primarily dinoflagellates 20–100 μm) have been identified as important consumers and degraders of copepod fecal pellets (Poulsen and Iversen, 2008; Poulsen et al., 2011). In our study, protozooplankton concentrations were highest near the surface but concentrations of dinoflagellates below 100 m were often several μg C L  1. Application of clearance rates for dinoflagellates other than the two very large species used experimentally by Poulsen et al. (2011) result in clearance rate estimates of approximately 3.5–10% d  1 of deep water in our study. From these rough calculations, it seems possible that protozooplankton could contribute significantly to removal of sinking particles under HI conditions in our study but two factors argue against this. First, dinoflagellates sense copepod pellets by chemical signals that approximate the original phytoplankton consumed by the pelletproducing copepods (Poulsen et al., 2011). Highest consumption rates are on fresh pellets because the chemical signal decays rapidly (Urban-Rich, 1999; Poulsen et al., 2011). In our study, pellets below 100 m would be approximately 12 h old and chemical signals emitted would have diminished. Second, dinoflagellates

84

M.J. Dagg et al. / Deep-Sea Research I 94 (2014) 72–86

0

50

100 CTD 36 CTD 38 CTD 40

Depth (m)

150

200 0

20

40

60

80

0

5

10

15

20

25

80

0

5

10

15

20

25

0

50

100 CTD 9 CTD 10 CTD 22 CTD 24

150

200 0

20

40

60

Ciliate concentration (µg C

L-1)

Dinoflagellate concentration (µg C

L-1)

Fig. 12. The vertical distribution of protozooplankton biomass (μg C L  1): (a) HI: ciliates and (b) HI: dinoflagellates from some HI casts; (c) LO: ciliates and (d) LO: dinoflagellates from some LO casts.

actively swim towards pellets releasing the chemical signal (Poulsen et al., 2011). The capture of large rapidly sinking pellets like those from euphausiids in our study is less likely compared to the smaller, slower sinking, copepod pellets used experimentally by Poulsen et al. (2011). We conclude that protozooplankton may possibly have contributed to pellet degradation and consumption below 100 m in our study but were more likely to have impacted fecal pellets near the surface because pellets were fresher and dinoflagellate concentrations were higher. There is a potential for zooplankton that we did not sample, such as radiolarians, to alter the flux of pellets (Gowing, 1989). Beers and Stewart (1967) found radiolarian concentrations as high as 50,000 m  3 in coastal California waters. If radiolarians at these high concentrations were functioning as flux feeders, much of the flux could be consumed (Iversen et al., 2010). Radiolarians were not observed in our protozoan samples but were, at times, abundant in the deeper MOCNESS samples (Checkley, unpublished). However, our sampling methodologies were not designed to quantitatively assess their abundance and distribution. Bacterial activity may have contributed to pellet degradation during sinking but we have no data on bacteria concentration or activity during our study. Bacteria have been implicated in pellet destruction in several studies (e.g. Hansen et al., 1996; Thor et al., 2003) and may also have been important in the experiments of Poulsen and Iversen (2008). Ploug et al. (2008) estimated that the bacterial respiration rate on fecal pellets was 0.15 d  1. If applied to our study for particles below 100 m in the HI profiles, this would suggest there is a significant consumption of pellet material by bacteria during 12 h. It has become generally accepted that most zooplankton pellets are not exported even though pellets often comprise a large component of export. Euphausiid pellets however, are exceptionally large compared

to most other zooplankton pellets. In our study, we observed that many large pellets produced by euphausiids were exported, at least to 100 m and probably deeper, within 12 h. High export of euphausiid feces under phytoplankton bloom conditions has also been observed by others (e.g. Wexels Riser et al. (2001), (2008), Gonzáles et al. (2011)). These episodic events of high flux can pass quickly to deeper, less biologically active water. We conclude that, in our study, the high concentrations of fecal pellets observed in surface waters at night were produced by dielly migrating euphausiids feeding primarily on large celled phytoplankton. Most of these pellets sank beneath our 200 m sampling zone within 24 h but small copepods (Oithona, Oncaea), heterotrophic dinoflagellates, and bacteria may have caused some pellet degradation or consumption. Consumption of sinking pellets by large copepods appeared to be insignificant.

Acknowledgments This work was supported by the National Science Foundation (NSF) awards OCE-0927863 to GJ, OCE-0928139 to MD, and OCE0928425 to DC. We are greatly indebted to Chih-Jung Wu and Hongbin Liu at the Hong Kong University of Science and Technology for analysis of the microzooplankton samples. We thank Rio Forrest-Baldini for her contribution at all stages of this work. References Alldredge, A.L., Gotschalk, C.C., MacIntyre, S., 1987. Evidence for sustained residence of macrocrustacean fecal pellets in surface waters off Southern California. Deep Sea Res. 34, 1641–1652. Antezana, T., 2010. Euphausia mucronata: A keystone herbivore and prey of the Humboldt Current system. Deep Sea Res. II 57, 652–662.

M.J. Dagg et al. / Deep-Sea Research I 94 (2014) 72–86

Bathmann, U., Fischer, G., Müller, P.J., Gerdes, D., 1991. Short term variations in particular matter sedimentation off Kapp Norvegica, Weddell Sea, Antarctica: relation to water mass advection, ice cover, plankton biomass and feeding activity. Polar Biol. 11, 185–195. Beers, J.R., Stewart, G.L., 1967. Micro-zooplankton in the euphotic zone at five locations across the California Current. J. Fish. Res. Board Can. 24, 2053–2068. Bollens, S.M., Frost, B.W., Lin, T.S., 1992. Recruitment, growth and diel vertical migration of Euphausia pacifica in a temperate fjord. Mar. Biol. 114, 219–228. Buesseler, K.O., Lamborg, C.H., Boyd, P.W., et al., 2007. Revisiting carbon flux through the ocean's twilight zone. Science 316, 567–570. Caron, D.A., Madin, L.P., Cole, J.J., 1989. Composition and degradation of salp fecal pellets: implications for vertical flux in oceanic environments. J. Mar. Res. 47, 829–850. Checkley Jr., D.M., Davis, R.E., Herman, A.W., Jackson, G.A., Beanlands, B., Regier, L.A., 2008. Assessing plankton and other particles in situ with the SOLOPC. Limnol. Oceanogr. 53, 2123–2136. Dagg, M.J., 1985. The effects of food limitation on diel migratory behavior in marine zooplankton. Arch. Hydrobiol.-Beih. Ergeb. Limnol. 21, 247–255. Dagg, M.J., 1993. Sinking particles as a possible source of nutrition for the calanoid copepod Neocalanus cristatus in the subarctic Pacific Ocean. Deep Sea Res. I 40, 1431–1445. Dagg, M.J., Walser Jr., W.E., 1986. The effect of food concentration on fecal pellet size in marine copepods. Limnol. Oceanogr. 31, 1066–1071. Dagg, M.J., Frost, B.W., Newton, J.A., 1997. Vertical migration and feeding behavior of Calanus pacificus females during a spring bloom in Dabob Bay, USA. Limnol. Oceanogr. 42, 974–980. Daly, K., 1997. Flux of particulate matter through copepods in the Northwest Water Polynya. J. Mar.Syst. 10, 319–342. Dam, H.G., Peterson, W.T., 1988. The effect of temperature on the gut clearance rate constant of planktonic copepods. J. Exp. Mar. Biol. Ecol. 123, 1–14. González, H.E., 1992. The distribution and abundance of krill faecal material and oval pellets in the Scotia and Weddell Seas (Antarctica) and their role in particle flux. Polar Biol. 12, 81–91. González, H.E., Smetacek, V., 1994. The possible role of the cyclopoid Oithona in retarding vertical flux of zooplankton faecal material. Mar. Ecol. Prog. Ser. 105, 31–45. González, H.E., González, S.R., Brummer, G-J.A., 1994. Short-term sedimentation pattern of zooplankton, faeces and microplankton at a permanent station in the Bjornafjorden (Norway) during April-May 1992. Mar. Ecol. Prog. Ser. 105, 31–45. González, H.E., Castro, L., Daneri, G., Iriarte, J.L., Silva, N., Vargas, C.A., Giesecke, R., Sánchez, N., 2011. Seasonal plankton variability in Chilean Patagonia fjords: carbon flow through the pelagic food web of Aysen Fjord and plankton dynamics in the Moraleda Channel basin. Cont. Shelf Res. 31, 225–243. Gowing, M.M., 1989. Abundance and feeding ecology of Antarctic phaeodarian radiolarians. Mar. Biol. 103, 107–118. Hamame, M., Antezana, T.., 2010. Vertical diel migration and feeding of Euphausia vallentini within southern Chilean fjords. Deep Sea Res. II 57, 642–651. Hansen, B., Bech, G., 1996. Bacteria associated with a marine planktonic copepod in culture. I. Bacterial genera in seawater, body surface, intestines and fecal pellets and succession during fecal pellet degradation. J. Plankton Res. 18, 257–273. Hansen, B., Fotel, F.L., Jensen, N.J., Madsen, S.D., 1996. Bacteria associated with a marine planktonic copepod in culture. II. Degradation of fecal pellets produced on a diatom, a nanoflagellate or a dinoflagellate diet. J. Plankton Res. 18, 275–288. Hillebrand, H., Dürselen, C.D., Kirschtel, D., Pollingher, U., Zohary, T., 1999. Biovolume calculation for pelagic and benthic microalgae. J. Phycol. 35, 403–424. Howell-Kübler, A.N., Lessard, E.J., Napp, J.M., 1996. Springtime microprotozoan abundance and biomass in the southeastern Bering Sea and Shelikof Strait, Alaska. J. Plankton Res. 18, 731–745. Hopcroft, R.R., Clarke, C., Chavez, F.P., 2002. Copepod communities in Monterey Bay during the 1997–1999 El Niño and La Niña. Prog. Oceanogr. 54, 251–264. Huskin, I., Viesca, L., Anadon, R., 2004. Particle flux in the subtropical Atlantic near the Azores: influence of mesozooplankton. J. Plankton Res. 26, 403–415. Iversen, M.H., Nowald, N., Ploug, H., Jackson, G.A., Fischer, G., 2010. High resolution profiles of vertical particulate organic matter export off Cape Blanc, Mauritania: degradation processes and ballasting effects. Deep Sea Res. I 57, 771–784. Iversen, M.H., Poulsen, L.K., 2007. Coprohexy, coprophagy, and coprochaly in the copepods Calanus helgolandicus, Pseudocalanus elongatus and Oithona similis. Mar. Ecol. Prog. Ser. 350, 79–89. Jackson, G.A., 1993. Flux feeding as a mechanism for zooplankton grazing and its implications for vertical particle flux. Limnol. Oceanogr. 38, 1328–1331. Jackson, G.A., Checkley Jr., D.M., 2011. Particle distributions in the upper 100 m water column and their implications for animal feeding in the plankton. Deep Sea Res. I 58, 283–297. Kobari, T., Akamatsu, H., Minowa, M., Ichikawa, T., Iseki, K., Fukuda, R., Higashi, M., 2010. Effects of the copepod community structure on fecal pellet flux in Kagoshima Bay, a deep, semi-enclosed embayment. J. Oceanogr. 66, 673–684. Komar, P.D., Morse, A.P., Small, L.F., Fowler, S.W., 1981. An analysis of sinking rates of natural copepod and euphausiid fecal pellets. Limnol. Oceanogr. 26, 172–180. Lampitt, R.S., Noji, T., von Bodungen, B., 1990. What happens to zooplankton fecal pellets? Implications for material flux. Mar. Biol. 104, 15–23. Legendre, L., Michaud, J., 1998. Flux of biogenic carbon in oceans: size dependent regulation by pelagic food webs. Mar. Ecol. Prog. Ser. 164, 1–11.

85

Le Fèvre, J., Legendre, L., Rivkin, R.B., 1998. Fluxes of biogenic carbon in the Southern Ocean: roles of large microphagous zooplankton. J. Mar. Syst. 17, 325–345. Liu, H-L., Sun, S., 2010. Diel vertical distribution and migration of a euphausiid Euphausia pacifica in the Southern Yellow Sea. Deep-Sea Res. II 57, 594–605. Lonsdale, D.J., Caron, D.A., Dennett, M.R., Schaffner, R., 2000. Predation by Oithona species on protozooplankton in the Ross Sea, Antarctica. Deep-Sea Res. II 47, 3273–3283. Lorenzen, C.J., Welschmeyer, N.A., 1983. The in situ sinking rates of herbivore fecal pellets. J. Plankton Res. 5, 929–933. Menden-Deuer, S., Lessard, E.J., 2000. Carbon to volume relationships for dinoflagellates, diatoms, and other protist plankton. Limnol. Oceanogr. 45, 569–579. Nakagawa, Y., Endo, Y., Taki, K., 2002. Contributions of heterotrophic and autotrophic prey to the diet of euphausiid, Euphausia pacifica in the coastal waters off northeastern Japan. Polar Biosci. 15, 52–65. Nakagawa, Y., Endo, Y., Sugisaki, H., 2003. Feeding rhythm and vertical migration of the euphausiid Euphausia pacifica in coastal waters of north-eastern Japan during fall. J. Plankton Res. 25, 633–644. Nishibe, Y., Kobari, T., Ota, T., 2010. Feeding by the cyclopoid copepod Oithona similis on the microplankton assemblage in the Oyashio region during spring. Plankton Benthos Res. 5, 74–78. Noji, T.T., Estep, K., MacIntyre, F., Norrbin, F., 1991. Image analysis of faecal material grazed upon by three species of copepods. Evidence for coprophagy, coprohexy and coprochaly. J. Mar. Biol. Assoc. UK 71, 465–480. Ohman, M.D., 1984. Omnivory by Euphausia pacifica: the role of copepod prey. Mar. Ecol. Prog. Ser. 19, 125–131. Paffenhöfer, G.-A., Mazzocchi, M.G., 2003. Vertical distribution of subtropical epiplanktonic copepods. J. Plankton Res. 25, 1139–1156. Petrik, C.M., Jackson, G.A., Checkley. Jr., D.M., 2013. Aggregates and their distributions determined from LOPC observations made using an autonomous profiling float. Deep-Sea Res. I 74, 64–81. Ploug, H., Iversen, M.H., Koski, M., Buitenhuis, E.T., 2008. Production, oxygen respiration rates, and sinking velocity of copepod fecal pellets: Direct measurements of ballasting by opal and calcite. Limnol. Oceanogr. 53, 469–476. Poulsen, L.K., Iversen, M.H., 2008. Degradation of copepod fecal pellets: key role of protozooplankton. Mar. Ecol. Prog. Ser. 367, 1–13. Poulsen, L.K., Moldrup, M., Berge, T., Hansen, P.J., 2011. Feeding on copepod fecal pellets: a new trophic role of dinoflagellates as detritivores. Mar. Ecol. Prog. Ser. 441, 65–78. Putt, M., Stoecker, D.K., 1989. An experimentally determined carbon:volume ratio for marine “oligotrichous” ciliates from estuarine and coastal waters. Limnol. Oceanogr. 34, 1097–1103. Riebesell, U., Reigstad, M., Wassmann, P., Noji, T., Passow, U., 1995. On the trophic fate of Phaeocystis pouchetti (Hariot): VI. Significance of Phaeocystis-derived mucus for vertical flux. Neth. J. Sea Res. 33, 193–203. Reigstad, M., Wexels-Riser, C., Svensen, C., 2005. The fate of copepod faecal pellets and the role of Oithona spp. Mar. Ecol. Prog. Ser. 304, 265–270. Roy, S., Silverberg, N., Romero, N., Diebel, D., Klein, B., Savenkoff, C., Vezina, A.F., Tremblay, J.-E., Legendre, L., Rivkin, R.B., 2000. Importance of mesozooplankton feeding for the downward flux of biogenic carbon in the Gulf of St. Lawrence (Canada). Deep Sea Res. II 47, 519–544. Silver, M.W., Gowing, M.M., 1991. The “particle” flux: origins and biological components. Prog. Oceanogr. 26, 75–113. Silver, M.W., Coale, S.L., Pilskaln, C.H., Chavez, F.P., 1998. Exploratory observations of marine aggregates at sub-euphotic depths. Deep-Sea Res. II 45, 1839–1861. Strickland, J.D.H., Parsons, T.R., 1968. A practical handbook of seawater analysis. Bull. Fish. Res. Board Can. 167, 201–206. Stuart, V., Pillar, S.C., 1990. Diel grazing patterns of all ontogenetic stages of Euphausia lucens and in situ predation rates on copepods in the southern Benguela upwelling region. Mar. Ecol. Prog. Ser. 64, 227–241. Svensen, C., Kiørboe, T., 2000. Remote prey detection in Oithona similis: hydromechanical versus chemical cues. J. Plankton Res. 22, 1155–1166. Svensen, C., Nejstgaard, J.C., 2003. Is sedimentation of copepod faecal pellets determined by cyclopoids? Evidence from enclosed ecosystems. J. Plankton Res. 25, 917–926. Thibault, D., Roy, S., Wong, C.S., Bishop, J.K., 1999. The downward flux of biogenic material in the NE subarctic Pacific: importance of algal sinking and mesozooplankton herbivory. Deep Sea Res. II, 2669–2697. Thor, P., Dam, H.G., Rogers, D.R., 2003. Fate of organic carbon released from decomposing copepod fecal pellets in relation to bacteria production and ectoenzymatic activity. Aquat. Microb. Ecol. 33, 279–288. Tsuda, A., Nemoto, T., 1990. The effect of food concentration on the faecal pellet size of the marine copepod Pseudocalanus newmani FROST. Bull. Plankton Soc. Jpn. 37, 83–90. Turner, J.T., 2002. Zooplankton fecal pellets, marine snow and sinking phytoplankton blooms. Aquat. Microb. Ecol. 27, 57–102. Urban-Rich, J., Hansell, D.A., Roman, M.R., 1998. Analysis of copepod fecal pellet carbon using a high temperature combustion method. Mar. Ecol. Prog. Ser. 171, 199–208. Urban-Rich, J., 1999. Release of dissolved organic carbon from copepod fecal pellets in the Greenland Sea. J. Exp. Mar. Biol. Ecol. 232, 107–124. Urban-Rich, J., Nordby, E., Andreassen, I.J., Wassmann, P., 1999. Contribution by mesozooplankton fecal pellets to the carbon flux on Nordvestbanken, north Norwegian shelf in 1994. Sarsia 84, 253–264. Utermöhl, H., 1958. Zur Vervollkommnung der quantitativen PhytoplanktonMethodik. Mitteilungen. Int. Ver. Theor. Angew. Limnol. 9, 1–38.

86

M.J. Dagg et al. / Deep-Sea Research I 94 (2014) 72–86

Vargas, C.A., Tonnesson, K., Sell, A., Maar, M., Moller, E.F., Zervoudaki, T., Giannalourou, A., Cristou, E., Satapoomin, S., Peterson, J.K., Nielsen, T.G., Tiselius, P., 2002. Importance of copepods versus appendicularians in vertical carbon fluxes in a Swedish fjord. Mar. Ecol. Prog. Ser. 241, 125–138. Verity, P.G., Langdon, C., 1984. Relationships between lorica volume, carbon, nitrogen, and ATP content of tintinnids in Narragansett Bay. J. Plankton Res. 6, 859–868. Viitasalo, M., Rosenberg, M., Heiskanen, A.-S., Kosi, M., 1999. Sedimentation of copepod fecal material in the coastal northern Baltic Sea: where did all the pellets go? Limnol. Oceanogr. 44, 1388–1399. Wassmann, P., 1998. Pelagic retention versus export food chains: process controlling sinking loss from marine pelagic systems. Hydrobiologia 36, 29–57. Wassmann, P., Hansen, L., Andreassen, I.J., Wexels Riser, C., Urban-Rich, J., 1999. Distribution and sedimentation of faecal pellets on the Nordvestbanken shelf, northern Norway, in 1994. Sarsia 84, 239–252.

Welschmeyer, N.A., Copping, A.E., Vernet, M., Lorenzen, C.J., 1984. Diel fluctuations in zooplankton grazing as determined by the downward flux of phaeopigments. Mar. Biol. 83, 263–270. Wexels Riser, C., Wassmann, P., Olli, K., Arashkevich, E., 2001. Production, retention and export of zooplankton faecal pellets on and off the Iberian shelf, northwest Spain. Prog. Oceanogr. 51, 423–441. Wexels Riser, C., Reigstad, M., Wassmann, P., Arashkevich, E., Falk-Petersen, S., 2007. Export or retention? Copepod abundance, faecal pellet production and vertical flux in the marginal ice zone through snap shots from the northern Barents Sea. Polar Biol. 30, 719–730. Wexels Riser, C., Wassmann, P., Reigstad, M., Seuthe, L., 2008. Vertical flux regulation by zooplankton in the northern Barents Sea during Arctic spring. Deep Sea Res. II 55, 2320–2329. Yoon, W.D., Kim, S.K., Han, K.N., 2001. Morphology and sinking velocities of fecal pellets of copepod, molluscan, euphausiid and salp taxa in the northeastern tropical Atlantic. Mar. Biol. 139, 923–928.