Download The ultrastructure of a doliolid and a copepod

JOURNAL OF PLANKTON RESEARCH
j
VOLUME
33
j
NUMBER
10
j
PAGES
1538 – 1549
j
2011
The ultrastructure of a doliolid
and a copepod fecal pellet
MARION KÖSTER 1*, RABEA SIETMANN 2, ANNETTE MEUCHE 2 AND GUSTAV-ADOLF PAFFENHÖFER 3
1
6, 18565 INSEL HIDDENSEE, GERMANY, 2ERNST-MORITZ3
ARNDT-UNIVERSITÄT GREIFSWALD, INSTITUT FÜR MIKROBIOLOGIE, FRIEDRICH-LUDWIG-JAHN-STR. 15 A, 17487 GREIFSWALD, GERMANY AND SKIDAWAY
INSTITUTE OF OCEANOGRAPHY, 10 OCEAN SCIENCE CIRCLE, SAVANNAH, GA 31411, USA
ERNST-MORITZ-ARNDT-UNIVERSITÄT GREIFSWALD, MIKROBIELLE ÖKOLOGIE, SCHWEDENHAGEN
*CORRESPONDING AUTHOR: [email protected]
Received January 27, 2011; accepted in principle May 18, 2011; accepted for publication May 23, 2011
Corresponding editor: Roger Harris
The goal of this study was to determine the morphology and ultrastructure of doliolid pellets using light, epifluorescence and transmission electron microscopy and
compare the results to observations of calanoid copepod pellets. For (ultra)structural
analyses, pellets of gonozooids of Dolioletta gegenbauri and females of the copepod
Eucalanus pileatus were produced in feeding experiments at close to environmental
food concentrations. Thin sections of a representative doliolid pellet revealed that
these pellets were mainly composed of intact diatom valves, a few fragmented valves
and intact flagellates. While the larger diatoms, Rhizosolenia alata, were completely
digested (empty valves), the smaller diatoms, Thalassiosira weissflogii, were partly, or
not digested at all. The phytoflagellate, Isochrysis galbana, appeared to be hardly
digested. Aggregations of bacteria occurred mostly inside pellets associated closely
with intact I. galbana flagellates and partly digested T. weissflogii cells; some scattered
bacteria were found among fragmented valves. No, or little, bacterial colonization
was associated with empty R. alata valves, and hardly digested T. weissflogii cells.
Whereas doliolid fecal pellets were loosely packed and composed of fully, incompletely and/or hardly digested food particles, pellets of the copepod E. pileatus were
densely packed and consisted mainly of fragmented diatom valves. Pellets of
doliolids and calanoid copepods can represent a high percentage of the particulate
organic carbon in the water-column on subtropical continental shelves.
KEYWORDS: doliolids; calanoid copepods; fecal pellets; ultrastructure;
transmission electron microscopy
I N T RO D U C T I O N
The occurrence and ecology of doliolids (Tunicata,
Thaliacea) have been studied for more than three
decades on the US southeastern continental shelf (e.g.
Deibel, 1998; Gibson and Paffenhöfer, 2000; Deibel and
Paffenhöfer, 2009). Doliolids have exponential growth
rates, and high asexual reproduction rates (Paffenhöfer
and Gibson, 1999). These energy-consuming processes
are fuelled by the uptake of large amounts of phytoplankton. They are able to ingest 20% of the in situ
phytoplankton standing stock per day (Paffenhöfer et al.,
1995). However, their assimilation efficiency often comprises ,50% of the ingested particulate organic matter.
That means that the ingested food particles are only
partly digested (Paffenhöfer and Köster, 2005) and unassimilated organic matter is released as fecal pellets.
Doliolids produce fecal pellets at high rates (6 – 8
pellets per gonozooid per hour, Paffenhöfer unpubl.
observ.). Considering their often high concentrations
.1000 zooids m23 (e.g. Paffenhöfer and Lee, 1987;
Paffenhöfer et al., 1995) pellet production is ecologically
doi:10.1093/plankt/fbr053, available online at www.plankt.oxfordjournals.org. Advance Access publication July 7, 2011
# The Author 2011. Published by Oxford University Press. All rights reserved. For permissions, please email: [email protected]
M. KÖSTER ET AL.
j
ULTRASTRUCTURE OF DOLIOLID FECAL PELLETS
significant. Laboratory observations of Paffenhöfer and
Köster (Paffenhöfer and Köster, 2011) revealed that a
laboratory-reared nurse with a 5-cm long cadophore
carries maximally up to 100 trophozooids (for illustration of the different stages in the life cycle of doliolids
see Fig. 1 in Paffenhöfer and Gibson, 1999). Each phorozooid when released from the nurse produces on an
average 9 – 14 gonozooids per day over a period of
8 – 18 days (Paffenhöfer and Gibson, 1999). Based on an
average gonozooid pellet production rate of 6 – 8 per
Fig. 1. Freshly produced fecal pellets of gonozooids of Dolioletta gegenbauri and adult female Eucalanus pileatus. Fecal pellets are mainly composed
of Thalassiosira weissflogii cells with a few Isochrysis galbana and Rhodomonas sp. cells. (a) Irregularly shaped fecal pellets of doliolids observed in the
darkfield, (b and c) Individual doliolid fecal pellet observed in the darkfield and under green excitation light (excitation of 510– 560 nm,
dichroic mirror 560 nm, emission of 575– 640 nm). Pronounced red autofluorescence of hardly digested Thalassiosira weissflogii cells is visible. (d)
Fecal pellets of Eucalanus pileatus. Insert shows the posterior part of a fecal pellet (white arrow indicates the peritrophic membrane).
1539
JOURNAL OF PLANKTON RESEARCH
j
VOLUME
hour, we calculated between a minimum of 144 and a
maximum of 192 pellets per day produced by one
gonozooid. Deibel and Paffenhöfer (Deibel and
Paffenhöfer, 2009) reported a maximum gonozooid
concentration of 1500 m23 on the continental US shelf.
That means that daily fecal pellet production can reach
between 216 000 and 288 000 pellets m23.
These high numbers highlight that doliolid fecal
pellets may significantly contribute to the pool of particulate organic carbon in upwelled waters where they
provide nutrient-rich microaggregates for microbial
colonization. An average gonozooid pellet comprises ca.
400 ng organic carbon (Brandes, pers. comm.).
Thus, the water column is supplied with
86– 115 mg C m23 d21 during an intense doliolid
bloom. A doliolid bloom with a period of 10 days produces 0.9 –1.2 g of organic carbon m23 enriched in
fecal pellets in upwelled waters.
Calanoid copepods on average usually produce fecal
pellets at lower rates than D. gegenbauri gonozooids.
Females of the copepod Eucalanus pileatus produce
2 pellets h21 compared with the 6 – 8 pellets h21 produced by a large gonozooid, both zooplankters being of
similar weight i.e. 20– 25 mg C individual21. Younger
stages of calanoids can produce maximally as many as
5 pellets h21 (Paffenhöfer and Knowles, 1979).
Since previous pellet studies focused mostly on those
of calanoid copepods, and since the two species developed well in summer intrusions of upwelled water on
the US southeastern continental shelf, we decided to
include pellets of one of the species, E. pileatus, for comparative purposes in our study on doliolid pellets. Food
concentrations chosen were close to those in upwelled
waters on the US southeastern shelf. Our goal was (i) to
compare the (ultra)morphology of fecal pellets of doliolids and copepods and their bacterial colonization,
and (ii) evaluate the degree of phytoplankton digestion
(two diatom species and flagellates) using transmission
electron microscopy (TEM).
METHOD
33
j
NUMBER
10
j
PAGES
1538 – 1549
j
2011
Rhodomonas sp (4 and 8 mm ESD, respectively). Pellets
were produced in 1.9 L glass jars containing the suspended food at a total concentration of 160 mg C L21
in GF/C filtered seawater. The jars were mounted on a
slowly rotating (0.3 rpm) plankton wheel at 218C.
Pellets of E. pileatus were produced at a total food
concentration of 238 mg C L21. Food particles offered
were mainly T. weissflogii and small amounts of
flagellates.
Gonozooids of D. gegenbauri produced between six and
eight pellets per hour, whereas adult female E. pileatus
produced two pellets per hour. To obtain a sufficient
number of pellets for microscopic analyses, doliolids and
copepods were incubated for 2 and 16 h, respectively.
Pellet sampling and fixation
For microscopic examination, recently produced pellets
(age of ,2 and ,16 h, respectively) of D. gegenbauri
gonozooids and females of E. pileatus were collected
with a cut-off glass pipette and transferred into a
150 mL glass dish filled with GF/C filtered seawater.
The integrity of pellets was checked under an Olympus
dissecting microscope at 20-fold magnification; only
intact pellets were selected for later microscopic investigations. Pellets were rinsed in 0.2-mm filtered seawater.
Aliquots of 5 – 10 pellets were fixed with formaldehyde
[2%(v/v) final concentration] and examined using light
and epifluorescence microscopy. For TEM 10 pellets
were transferred into 2 mL of 0.2-mm filtered seawater
containing 1%(v/v) glutaraldehyde and 2%(v/v) paraformaldehyde. Pellets were fixed at 218C for 2 h, and
then stored at 48C prior to preparation for TEM.
Size and volume of pellets
The size of pellets was determined under an Olympus
dissecting microscope at 20-fold magnification using an
ocular micrometer. The average width (w) and length (l)
of D. gegenbauri and E. pileatus pellets were measured
at +25 mm (n ¼ 150). For pellet volume estimations,
following equations were used,
Pellet production
(i)
For pellet production, two feeding experiments were
performed in September 2009. Laboratory-reared
gonozooids of D. gegenbauri had been feeding on
the diatoms Thalassiosira weissflogii (8 – 12 mm ESD,
Equivalent Spherical Diameter) and Rhizosolenia alata
forma indica (length of 200– 300 mm, width of
12– 16 mm) with small amounts of the flagellate
Isochrysis galbana and very few cells of the flagellate
(ii)
Volume (mm3) of D. gegenbauri pellets ¼ width
(mm) length (mm) thickness (mm).
Volume (mm3) of E. pileatus pellets ¼ p (width/2)2 (mm2) length (mm).
Transmission electron microscopy
After embedding the fecal pellets in low-gelling-temperature agarose, they were postfixed in 1%(v/v) osmium
tetroxide for 2 h at room temperature (and 48C over
1540
M. KÖSTER ET AL.
j
ULTRASTRUCTURE OF DOLIOLID FECAL PELLETS
night) followed by immersion of the specimen in 2%(v/v)
uranyl acetate for 2 h. After dehydration in a graded
series of ethanol the material was embedded in
EPONTM epoxy resin. Sections were cut on an ultramicrotome (Reichert Ultracut, Leica UK Ltd, Milton
Keynes, UK), stained with uranyl acetate and lead
citrate and examined with a TEM LEO 906 (Zeiss,
Oberkochen, Germany) at 80 kV.
The representativeness of chosen doliolid pellets was
evaluated by their shape, size, composition, and how
intact they were checked under a dissecting microscope
at 20-fold magnification. For copepods, the intact
nature of the outer peritrophic membrane was an
additional criterion to assess how representative they
were. According to our previous light microscopic investigations doliolid pellets produced under defined laboratory food concentration and composition conditions
revealed a morphological structure and contained food
particles of a digestive state as found previously
(Paffenhöfer and Köster, 2005). Also, pellets of E. pileatus
were similar to previous observations. Therefore, we
selected only one representative pellet of each taxon for
a detailed ultrastructure analysis by TEM.
The selected squarish-like doliolid pellet had a length
of 450 mm and the cylindrically-shaped copepod
pellet a length of 500 mm and a width of 50– 60 mm.
For the estimation of bacterial abundance in doliolid
pellets, micrographs of selected thin sections at 800-fold
magnification were used. Bacteria and microalgae (non-,
partly- and completely-digested) were counted in fields
of 1946 mm2.
Light and epifluorescence microscopy
The morphology of formaldehyde-fixed doliolid and
copepod pellets was examined with an Axiophot epifluorescence microscope (Zeiss, Jena, Germany) at 50
and 100-fold magnification, respectively. Chlorophyll
autofluorescence was determined under green excitation
light (excitation of 510 to 560 nm, dichroic mirror
560 nm, emission of 575 – 640 nm) and used to estimate
the degree of digested food particles in pellets.
Micrographs were recorded with a high-resolution
CCD camera 1300 BC (Vosskühler, Osnabrück,
Germany) using an image analysis system (Lucia 4.81).
R E S U LT S
Light microscopy
Light microscopy observations showed that D. gegenbauri
fecal pellets produced by gonozooids had an average
size of 300 mm by 600 mm (Fig. 1a). Mean pellet
volumes usually ranged between 0.03 and 0.05 mm3
depending on the size of the gonozooids and the type
of ingestable food particles. The shape of the gonozooid
pellets was generally flat and squarish-like, but could
also be irregularly shaped and show elongated threadlike structures. The thickness of the pellets varied from
monolayered to multilayered cell sheets; the maximum
thickness was 150 mm. Especially the thinner and
more loosely structured outer regions were translucent,
whereas the central regions were more densely compacted. Dolioletta gegenbauri fecal pellets consisted mainly
of small centric T. weissflogii (8 – 12 mm ESD) and large
cylindrical R. alata cells (length of 200– 300 mm, width
of 12– 16 mm). Strong chlorophyll autofluorescence
revealed that the majority of chloroplasts of T. weissflogii
cells were hardly or not digested by the gut passage
through D. gegenbauri (Fig. 1b and c). Light microscopy
showed that the outer transparent region of some pellets
consisted mainly of broken and almost fully digested
T. weissflogii cells, while cells in the interior more densely
packed region of the pellet seemed to be undigested.
Digested food particles were embedded in mucous
material preventing the membraneless pellet from disaggregation (Köster unpubl. observation).
Copepod pellets produced by female E. pileatus were
cylindrically shaped, and surrounded by a chitinous peritrophic membrane (Fig. 1d; see insert, white arrow) that
often formed a tail at both ends of the pellet. Pellets
reached lengths of 400– 600 mm and widths of up to
50– 60 mm, respectively. Compared with doliolid pellets,
average copepod pellet volumes were about one order of
magnitude lower (0.002 – 0.004 mm3). Pellets consisted
mainly of fragmented diatom valves, were densely
packed and non-translucent (Fig. 1d). As observed in a
previous study, the autofluorescence of digested fragmented food particles was only weak in copepod pellets
(see Fig. 1f in Paffenhöfer and Köster, 2005).
Ultrastructure of fecal pellets
Transmission electron microscopy of pellets revealed
the fine structure and state of ingested phytoplankton
cells as well as the occurrence and distribution of
pellet-associated bacteria. Further, TEM serves to localize the sites of microbial attack. In thin sections from
different parts of the gonozooid pellet of D. gegenbauri, we
observed numerous intact diatom valves of R. alata and
T. weissflogii (Fig. 2a and b). While some of the small
T. weissflogii contained cell organelles such as chloroplasts
and nuclei in the cytoplasm (Fig. 2a see white arrows,
Fig. 3a and b) the larger R. alata cells did not contain cell
organelles and only occasionally did cytoplasm remain,
indicating that these cells were almost completely
1541
JOURNAL OF PLANKTON RESEARCH
j
VOLUME
33
j
NUMBER
10
j
PAGES
1538 – 1549
j
2011
Fig. 2. Transmission electron microscopy micrographs of regions inside a gonozooid fecal pellet of Dolioletta gegenbauri. Thin sections originate
from the middle part of the pellet. (a and b) From longitudinally to vertically varying sections of large intact silicate valves of empty (¼digested)
Rhizosolenia alata cells (R) and some smaller ones of Thalassiosira weissflogii (T) containing partly digested chloroplasts (white arrow) (c) There are
also some regions with fragmented diatom valves (FV) with scarce bacterial colonization (B, bacteria; FV, fragmented valves).
digested (Fig. 2b). This observation agrees with the
results of earlier work (Paffenhöfer and Köster, 2005)
showing that doliolids incompletely digested the ingested
food particles, often assimilating ,50% of the ingested
organic material. Fragmented diatom frustules were only
found intermittently between intact frustules (densely
compacted aggregations with a diameter of 30– 100 mm;
Fig. 2a and c) because doliolids do not have mouthparts
to select and crush food particles, but collect food particles with their mucous net.
Most of the flagellates ingested (I. galbana) seemed to
be hardly deformed. Epifluorescence micrographs
( photos not shown) showed that the chloroplasts and
even the flagellae of I. galbana were preserved. These
observations were obtained from T. weissflogii and
I. galbana composed pellets that were transferred to
sterile (0.2-mm filtered) seawater, formaldehyde (2% v/v)fixed, sonicated (3 7 s, 35 W) to disrupt the pellet
structure and stained with acridine orange (final concentration 1:10 000). Transmission electron microscopy
analyses confirmed that these flagellates reveal hardly any
cellular alterations after being ingested by D. gegenbauri
(Fig. 4d–f ). We estimate that the residence time of
ingested cells exposed to digestive enzymes does not
surpass 30 min at a release rate of 8 pellets h21. That
implies a limited period of exposure to digestive fluids
during which a flagellate membrane would have to be
lysed.
Overview cross-sections of calanoid copepod fecal
pellets revealed that most of the ingested T. weissflogii
cells were fragmented (Fig. 5). The few intact valves did
not contain any cell constituents. The peritrophic membrane that is produced in the copepod gut and wraps
the ingested food particles (Gauld, 1957; Hansen and
Peters, 1997/1998) remained intact in most of the
freshly produced pellets up to 16 h after release (Fig. 1d,
insert). Transmission electron microscopy micrographs at
800-fold magnification (Fig. 5) clearly showed the integrity
of the peritrophic membrane. The membrane was 50–
70-nm thick and consisted of multilayered chitinous
microfibrils oriented in different directions (TEM micrographs at 28 000–60 000-fold magnification, not shown).
1542
M. KÖSTER ET AL.
j
ULTRASTRUCTURE OF DOLIOLID FECAL PELLETS
Fig. 3. Transmission electron microscopy micrographs of hardly digested Thalassiosira weissflogii cells inside fecal pellets of Dolioletta gegenbauri.
Longitudinal (a) and cross-section (b) of intact cells of Thalassiosira weissflogii. Note the intact cell organelles (N, nucleus; No, nucleolus; Ch,
chloroplasts). There are hardly any bacteria (B) in the surroundings of these cells. (c and d) longitudinal sections of partly digested cells that
contain degenerated cell organelles and cytoplasm (Cp). These cells reveal a dense bacterial colonization outside their silicate valves.
Bacterial colonization of fecal pellets
In loosely structured doliolid pellets, bacteria occurred
both close to the external surfaces and within pellets.
Thin sections of a gonozooid pellet of D. gegenbauri
revealed that sizes and distribution of bacteria were
variable. Most of the bacteria were rods with lengths
ranging from 0.4 to 1.5 mm; dividing cells made up
,0.5% of total bacteria. The majority of bacteria
occurred in clusters in the vicinity of partly digested
T. weissflogii cells (up to 54 bacteria 1000 mm22 thin
section area Fig. 3c and d), and ingested flagellates (up
to 82 bacteria 1000 mm22 thin section area; Fig. 4a and
b), and in regions with unidentifiable organic matter
(up to 160 bacteria 1000 mm22). We found numerous
examples for these observations. Extremely scarce or no
bacterial colonization was found in the vicinity of intact
T. weissflogii cells (with nuclei and chloroplasts; Fig. 3a
and b) and also not inside and outside of empty valves
of T. weissflogii and R. alata (Fig. 2b). A few scattered
bacteria were found among fragmented valves (,25
bacteria 1000 mm22; Fig. 2c).
A Kruskal – Wallis Test (non-parametric analysis of
variance by ranks; Conover, 1980) was applied to bacterial abundances near partly digested as well as completely digested (empty) cells of T. weissflogii, and cells of
R. alata. The null hypothesis that bacterial abundances
were the same in each of the three cases was rejected.
The ensuing multiple comparison test revealed that bacterial abundances near empty cells of T. weissflogii and
R. alata did not differ significantly while bacteria near
partly empty cells of T. weissflogii were more abundant
than near the empty cells of the two phytoplankton
species. Bacterial abundances near cells of I. galbana far
surpassed those near the partly digested T. weissflogii.
Freshly egested copepod pellets revealed hardly any
bacteria on their exterior surfaces ( peritrophic membrane). The majority of bacteria were found in the
interior lumen of the pellets.
1543
JOURNAL OF PLANKTON RESEARCH
j
VOLUME
33
j
NUMBER
10
j
PAGES
1538 – 1549
j
2011
Fig. 4. Transmission electron microscopy micrographs of hardly digested phytoflagellates [Isochrysis galbana (Ig); Rhizosolenia alata (Rho)] inside
Dolioletta gegenbauri fecal pellets. (a–c) Note the numerous bacteria (B, bacteria; BC, bacterial cluster) in the vicinity of the flagellates (d) Isochrysis
galbana. The flagellates’ two chloroplasts (Ch) with thylachoid membranes, the pyrenoid (P), the nucleus (N) and mitochondria (M) are visible.
(e and f ) Longitudinal sections of Rhodomonas sp. Note the bacteria in the surroundings and the Isochrysis galbana cell that is attached to the outer
membrane of Rhodomonas sp.
1544
M. KÖSTER ET AL.
j
ULTRASTRUCTURE OF DOLIOLID FECAL PELLETS
copepods, their pellet morphology differs in many
respects.
Ultrastructure of fecal pellets
Fig. 5. Transmission electron microscopy micrographs of fecal pellets
of female Eucalanus pileatus. (a and b) Fecal pellets contained mainly
fragmented valves of Thalassiosira weissflogii (Fd). There were only a few
intact diatom valves (Id). The white arrows indicate the outer
peritrophic membrane (Pm) of the pellet.
DISCUSSION
The morphology of fecal pellets influences their bacterial colonization and degradation (as well as their sedimentation behavior (residence time) and thus their
contribution to the pelagic food web and vertical
carbon flux (e.g. Hofmann et al., 1981; Wexels Riser
et al., 2007; Giesecke et al., 2010; Patonai et al. 2011).
Our TEM observations revealed differences in the
ultrastructure of a doliolid and a copepod fecal pellet as
well as the condition of ingested phytoplankton cells,
and the occurrence of bacteria on/within the pellets.
Owing to different feeding mechanisms of doliolids and
A range of papers have covered the ultrastructure of
zooplankton fecal pellets by scanning electron
microscopy (SEM, e.g. Turner, 1984a, b; Beaumont
et al., 2002; Jansen and Bathmann, 2007). This technique enables the surfaces of both fragmented/broken
and intact pellets to be observed (e.g. Jansen et al.,
2006). Turner (Turner, 1984b) stripped off the peritrophic membrane to make the interior of the pellets
visible. Alternatively, TEM can be applied to investigate
the ultrastructure of fecal pellets. Thin sections allow
the presence and distribution of both internal and
membrane-associated bacteria in different regions of a
pellet to be investigated. In addition, the detailed ultrastructure of digested food particles including their cell
organelles and membranes is visible. We used the
detailed information on their state to evaluate the
degree of digestion. Especially, as concerns the intactness of the peritrophic membrane as an indicator for
pellet integrity, the application of TEM is helpful.
Although TEM has the potential to provide valuable
information on the ultrastructure of pellets, it has been
rarely used previously. To the authors’ knowledge, the
only TEM investigations are analyses of the composition of pellets of salps and pteropods (Silver and
Bruland, 1981) and unidentified copepod pellets
(Gowing and Silver, 1983). The latter study presents
TEM micrographs of the structure of cylindrical and
ellipsoid pellets (most likely originating from copepods)
obtained from traps, and from shipboard-feeding crab
larvae. They showed damaged as well as intact phytoplankton cells and bacteria.
Our study describes the ultrastructure of doliolid and
copepod pellets. Our observations revealed seemingly
intact, partly empty and entirely empty phytoplankton
cells in doliolid fecal pellets. While the R. alata valves
were empty, most of the T. weissflogii cells seemed to be
not or hardly affected (well preserved chloroplasts and
other cell organelles). Some T. weissflogii cells contained
cytoplasm remains, and a few were empty. So, why does
the integrity of diatom cells differs between taxa, and
why does the percentage of hardly digested T. weissflogii
cells far exceed that of R. alata cells? We assume that
digestive enzymes of D. gegenbauri gonozooids might
have had an easier access to R. alata cells than to
T. weissflogii cells. Differences in the fine structure of silicate valves (e.g. number and size of pore sizes, stability
of valves; shielding surface polymers around algae)
probably result in a higher percentage of the smaller
1545
JOURNAL OF PLANKTON RESEARCH
j
VOLUME
T. weissflogii cells surving gut passage and they may have
been more resistant to the digestive enzymes of D. gegenbauri than the larger R. alata cells. Whether ingested
viable bacteria may supplement the hydrolysis of food
components, which the doliolid cannot hydrolyze
because it lacks the necessary enzymes, remains unclear.
Fecal pellets of E. pileatus contained numerous broken
valves among the occasional undamaged valves of
T. weissflogii (Fig. 5). This explains the high assimilation
efficiency of calanoid copepods ingesting diatoms when
compared with doliolids ingesting the same cells (e.g.
Paffenhöfer and Köster, 2005).
Bacterial colonization of fecal pellets
The initial studies on thaliacean fecal pellets and associated bacteria included observations on pellets of
D. gegenbauri (Pomeroy and Deibel, 1980; Pomeroy et al.,
1984). The latter stated ‘They (the pellets) contain large
amounts of undigested or partly-digested materials and
consequently are excellent substrates for bacteria’.
Pellets of D. gegenbauri, which had been feeding on
enriched natural seston for 6 h, had by that time few
bacteria, but numerous large bacteria after 21 h.
Similarly, D. gegenbauri pellets, from feeding on a natural
suspension of the small diatom Thalassiosira subtilis for
6 h, had no bacteria; however, after 21 h they had large
motile rods which were also found in the surrounding
water (Pomeroy and Deibel, 1980). Overall, microscopic
examination of such feces ‘showed a rapid increase in
bacteria within the feces during the first day’. This the
authors attributed to excess DOM, associated with the
partly digested food particles within fecal particles,
which is leached as indicated by primary amines
(Pomeroy et al., 1984). Kiørboe et al. (Kiørboe et al.,
2002) reported bacterial colonization rates of artificial
organic-rich (leaky) particles even in the order of
minutes. These authors described bacterial colonization
of aggregates ( pellets) as a highly dynamic process
depending on the motility of bacteria, their attachment/
detachment behavior, as well as the size and hydrodynamic environment of the aggregates.
As doliolids create a cilary current to draw food particles into their mouth (Bone et al., 1997), a certain percentage of free-living bacteria should also be retained in
their mucous filter via filter feeding. Hitherto no data
on the mesh size of the filter of doliolids are available
(due to unsuccessful fixation of the filter net for electron
microscopy, Bone et al., 2003). In our study, many freeliving bacteria smaller than 1 mm were found within
D. gegenbauri pellets (Figs. 3 and 4) suggesting that the
filter net might be capable of trapping submicron particles. The bacteria trapped in the filter net provide, in
33
j
NUMBER
10
j
PAGES
1538 – 1549
j
2011
addition to the attached bacteria of ingested food particles, an additional food source for the doliolids.
Whether bacteria survive the doliolids’ gut passage
and whether their digestive enzymes play a role in the
decomposition of ingested phytoplankton is still a
subject of investigation. At the same time bacteria living
in the water surrounding the recently ejected pellets
should be attracted by various dissolved organic compounds that are released by the partly digested phytoplankton cells (Pomeroy, 1992). We want to point out
that while Pomeroy and Deibel (Pomeroy and Deibel,
1980) offered natural particulate matter to doliolids, we
offered phytoplankton cells grown in non-axenic conditions in the laboratory. Such phytoplankton cultures
contain organic matter degrading bacteria of unknown
origin, physiology and ecology.
Transmission electron microscopy observations of
doliolid pellets revealed that bacterial colonization
varied in the vicinity of ingested phytoplankton species
of different state (Figs. 3 and 4). The degree of bacterial colonization is probably caused by the availability
of assimilable DOC. When assimilable organic
material leaks through the valves of partly digested
diatoms and/or flagellates, nutrient-rich microhabitats
are created around the algae/flagellates serving as
nutrient sources for heterotrophic bacteria. This is
reflected in increased bacterial numbers in these highly
nutritive areas (see Figs. 3 and 4). The majority of bacteria in the vicinity of partly digested food particles
were ,1 mm, but there always appeared some large
rods with lengths of 1.5 – 2.0 mm and widths of
0.5 mm. As the age of our pellets was mostly in the
range of several hours (between 2 and 16 h), bacteria
would have sufficient time to grow within the pellets,
and/or position themselves in the vicinity of digested
food items containing organic matter. Dividing cells
were not observed among the bacteria inside pellets
indicating that the residence time of fecal pellets in
the gut was shorter than the generation time of
food-associated bacteria.
Contrary to bacteria-enriched areas, nearly completely digested microalgae such as empty valves of
T. weissflogii and R. alata do not provide any valuable/
assimilable substrates for bacteria. Therefore, these
areas were far less colonized by bacteria. The fact that
no bacterial colonization was observed inside numerous
diatom valves and among fragmented valves confirms
our assumption that the digestion of algal cells was
caused by hydrolytic digestive enzymes of the doliolids’
gut and not by bacteria.
In 50 thin slices of a freshly egested copepod pellet
(,16 h) we found hardly any bacteria (1 bacteria
1000 mm22 thin section area) on the peritrophic
1546
M. KÖSTER ET AL.
j
ULTRASTRUCTURE OF DOLIOLID FECAL PELLETS
membrane but quite a few (,25 bacteria 1000 mm22
thin section area) within the pellet (Fig. 5). Several
authors (Lautenschlager et al., 1978; Gowing and Silver,
1983) also reported that the surfaces of freshly ejected
(age ,7 h) crustacean pellets ( planktonic copepods and
gammarids, respectively) are devoid of bacteria. With
ongoing degradation, copepod fecal pellets were densely
colonized by bacteria (‘bacterial lawn’) as also has been
observed by other investigators (cf. Turner, 1979;
Gowing and Silver, 1983; Turner, 2002). Whether the
majority of copepod pellet associated internal bacteria
originate from the enteric flora of the copepod and/or
bacteria attached to ingested food particles remains
uncertain (compare e.g. Nagasawa and Nemoto, 1988;
de Troch et al., 2010). Lawrence et al. (Lawrence et al.,
1993) reported that calanoid copepods produced
bacteria-free pellets when axenic food was offered. They
concluded that pellet-associated bacteria did not originate from the copepods’ gut but were derived from
ingested food particles. Furthermore, the diversity of the
bacterial community of fecal pellets of harpacticoid
copepods is assumed to be broader than that of their
food source (diatoms). This increased bacterial richness
depends largely from the ingested food, the copepod
species and the age of fecal pellets (de Troch et al., 2010).
Pellet morphology, sinking velocity
and potential degradation
The first fundamental microbiological studies of the
degradation of doliolid fecal pellets go back to Pomeroy
in the early 1980s (Pomeroy and Deibel, 1980; Pomeroy
et al., 1984). Later and recent investigations that have
applied more sensitive methods to determine microbial
pellet degradation (e.g. Thor et al., 2003, enzymatic
degradation activity; Olsen et al., 2005; microcalorimetry;
Shek and Liu, 2010, oxygen consumption measurements
by microsensors) were exclusively devoted to copepod
pellets. None of these process-oriented studies were
accompanied by microscopic methods. Thus, there is a
need to combine process-oriented investigations with
direct microscopic methods to assess the degradation
state of doliolid and copepod pellets and their contents.
As pellet characteristics (shape, packaging, density
etc.) differ strongly, we expect different degradation
pathways for the pellets of the doliolid D. gegenbauri and
the calanoid copepod E. pileatus. The loose packaging of
doliolid pellets (low density), their flat shape and the
intermittent occurrence of threadlike structures are
responsible for their low sinking velocities and long
retention times in the water column. Patonai et al.
(Patonai et al., 2011) determined average sinking velocities of 20– 60 m d21 for pellets of 6 – 7 mm long
gonozooids. Since doliolid pellets will therefore remain
in the water column for a relatively long period, and
seawater heterotrophic bacteria have immediate access
to the membraneless pellets, those will most likely be
rapidly mineralized in the water column.
In comparison, pellets of females of E. pileatus are
more densely packed than doliolid pellets, and have
much smaller volumes than the latter (Fig. 1). While
their sinking rates are in the same range as those of
doliolid pellets (Patonai et al., 2011), observations by
Ploug et al. (Ploug et al., 2008) reveal an average sinking
rate of 322 m d21of pellets from the calanoid copepod
Temora longicornis (females) feeding at 158C on T.
weissflogii at 430 mg C L21. Those pellets’ volumes at
106 mm3 were slightly lower than those of E. pileatus
pellets also feeding on T. weissflogii in our experiment. It
appears from their detailed study as well as Feinberg
and Dam (Feinberg and Dam, 1998) that ‘pellet size
and density depend on copepod species, food concentration and food source.’
The early decomposition of copepod pellets is initiated
via mechanical breakdown and/or microbial attack of
the chitinous peritrophic membrane. As soon as the
pellets are egested their peritrophic membranes provide
a colonizable substratum (surface) for bacteria occurring
in the surrounding water. While the peritrophic membrane of up to 16 h old E. pileatus pellets was intact
(Fig. 5) and colonized by only a very few bacteria, the
peritrophic membrane of pellets of Acartia tonsa was
already disrupted within 1.5 h (Hansen et al., 1996). With
ongoing degradation, we observed microscopically that
pellets became more porous and lighter, indicating to the
activity of organic matter degrading bacteria with different physiological potential. The degree and rate of membrane and pellet degradation depends on coprophagy,
bacterial colonization and pellet composition (Hansen
et al., 1996). Up to the present, it is still a methodological
challenge to distinguish between internal-origin and
external-origin pellet-associated bacteria responsible for
membrane and pellet degradation. A first indication of
the origin of bacteria is the TEM micrographs showing
bacteria inside pellets with an integral peritrophic membrane (Fig. 5).
One of the next steps to consider would be to
conduct similar studies on doliolid pellets produced
under in situ conditions. This would be accompanied by
quantifying abundances of phytoplankton and heterotrophic organisms (including bacteria and protists, zooplankton) in the pellets and the surrounding water. In
addition, pellets of smaller (body length of 2 – 3 mm)
zooids (e.g. small gonozooids released by phorozooids),
which are dominant in blooms, ought to be included in
future considerations. New staining techniques (e.g. Cell
1547
JOURNAL OF PLANKTON RESEARCH
j
VOLUME
Tracker Green) will enable our approach of the ultrastructure of fecal pellets to be combined with functional
analyses of the viability of pellet associated cells (undigested, digested phytoplankton species, bacteria).
33
j
NUMBER
10
j
PAGES
1538 – 1549
j
2011
Gibson, D. M. and Paffenhöfer, G.-A. (2000) Feeding and growth
rates of the doliolid, Dolioletta gegenbauri Uljanin (Tunicata,
Thaliacea). J. Plankton Res., 22, 1485– 1500.
Giesecke, R., González, H. E. and Bathmann, U. (2010) The role of
the chaetognath Sagitta gazellae in the vertical carbon flux of the
Southern Ocean. Polar Biol., 33, 293– 304.
Gowing, M. M. and Silver, M. W. (1983) Origins and microenvironments of bacteria mediating fecal pellet decomposition in the sea.
Mar. Biol., 73, 7 –16.
AC K N OW L E D G E M E N T S
We thank Captain Raymond Sweatte and the crew of
R/V Savannah for their professional cooperation and
their support collecting doliolids and copepods on the
continental SE shelf of the USA. We greatly acknowledge Dr Gesine Roth (Deutscher Akademischer
Austausch Dienst, German Academic Exchange
Service) for the support of Dr M. Köster in building
up an American-German cooperation between the
Skidaway Institute of Oceanography, and the University
of Greifswald. We are grateful for the constructive comments of three anonymous reviewers.
Hansen, B., Fotel, F. L., Jensen, N. J. et al. (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.
Hansen, U. and Peters, W. (1997/1998) Structure and permeability of
the peritrophic membranes of some small crustaceans. Zool. Anz.,
236, 103– 108.
Hofmann, E. E., Klinck, J. M. and Paffenhöfer, G.-A. (1981)
Concentrations and vertical fluxes of zooplankton fecal pellets on a
continental shelf. Mar. Biol., 61, 327–335.
Jansen, S. and Bathmann, U. (2007) Algae viability within copepod
faecal pellets, evidence from microscopic examinations. Mar. Ecol.
Prog. Ser., 337, 145–153.
Jansen, S., Riser, C. W., Wassmann, P. et al. (2006) Copepod feeding
behaviour and egg production during a dinoflagellate bloom in the
North Sea. Harmful Algae, 5, 102– 112.
FUNDING
The research for this paper was supported by two
grants from the U.S. National Science Foundation
(OCE 0825999, OCE 1031263).
Kiørboe, T., Grossart, H. -P., Ploug, H. et al. (2002) Mechanisms and
rates of bacterial colonization of sinking aggregates. Appl. Environ.
Microbiol., 68, 3996– 4006.
REFERENCES
Lawrence, S. G., Ahmed, A. and Azam, F. (1993) Fate of particlebound bacteria ingested by Calanus pacificus. Mar. Ecol. Prog. Ser., 97,
299 –307.
Beaumont, K. L., Nash, G. V. and Davidson, A. T. (2002)
Ultrastructure, morphology and flux of microzooplankton faecal
pellets in an East Antarctic fjord. Mar. Ecol. Prog. Ser., 245,
133–148.
Bone, Q., Braconnot, J. -C. and Carré, C. (1997) On the filter feeding
of Doliolum (Tunicata, Thaliacea). J. Exp. Mar. Biol. Ecol., 214,
179–193.
Bone, Q., Carré, C. and Chang, P. (2003) Tunicate feeding filters.
J. Mar. Biol. Assoc. UK, 83, 907 –919.
Lautenschlager, K. P., Kaushik, N. K. and Robinson, J. B. (1978) The
peritrophic membrane and faecal pellets of Gammarus lacustris
limnaeus Smith. Freshw. Biol., 8, 207– 211.
Nagasawa, S. and Nemoto, T. (1988) Presence of bacteria in guts of
marine crustaceans and on their fecal pellets. J. Plankton Res., 10,
559 –564.
Olsen, S. N., Westh, P. and Hansen, B. W. (2005) Real-time
quantification of microbial degradation of copepod fecal pellets
monitored by isothermal microcalorimetry. Aquat. Microb. Ecol.,
40, 259–267.
Conover, W. J. (1980) Practical Nonparametric Statistics. John Wiley &
Sons, New York.
Paffenhöfer, G.-A., Atkinson, L. P., Lee, T. N. et al. (1995) Distribution
and abundance of thaliaceans and copepods off the southeastern
USA during winter. Cont. Shelf Res., 15, 255–280.
Deibel, D. (1998) The abundance, distribution and ecological impact
of doliolids. In Bone, Q. (ed.), The Biology of Pelagic Tunicates. Oxford
University Press, Oxford, England, pp. 171– 186.
Paffenhöfer, G.-A. and Gibson, D. M. (1999) Determination of generation time and asexual fecundity of doliolids (Tunicata, Thaliacea).
J. Plankton Res., 21, 1183– 1189.
Deibel, D. and Paffenhöfer, G.-A. (2009) Predictability of patches of
neritic salps and doliolids (Tunicata, Thaliacea). J. Plankton Res., 31,
1571– 1579.
Paffenhöfer, G.-A. and Knowles, S. C. (1979) Ecological implications
of fecal pellet size, production and consumption by copepods.
J. Plankton Res., 37, 35– 49.
de Troch, M., Cnudde, C., Willems, A. et al. (2010) Bacterial colonization on fecal pellets of harpacticoid copepods and on their diatom
food. Microbiol. Aquat. Syst., 60, 581– 591.
Paffenhöfer, G.-A. and Köster, M. (2005) Digestion of diatoms
by planktonic copepods and doliolids. Mar. Ecol. Prog. Ser., 297,
303–310.
Feinberg, L. R. and Dam, H. G. (1998) Effects of diet on dimensions,
density and sinking rates of fecal pellets of the copepod Acartia tonsa.
Mar. Ecol. Prog. Ser., 175, 87–96.
Paffenhöfer, G.-A. and Köster, M. (2011) From one to many, On the
life cycle of Dolioletta gegenbauri Uljanin (Tunicata, Thaliacea).
J. Plankton Res., 33, 1139– 1145.
Gauld, D. T. (1957) A peritrophic membrane in calanoid copepods.
Nature, 179, 325–326.
Paffenhöfer, G.-A. and Lee, T. N. (1987) Development and persistence
of patches of Thaliacea. S. Afr. J. Mar. Sci., 5, 305–318.
1548
M. KÖSTER ET AL.
j
ULTRASTRUCTURE OF DOLIOLID FECAL PELLETS
Patonai, K., El-Shaffey, H. and Paffenhöfer, G.-A. (2011) Sinking velocities of copepod and doliolid fecal pellets. J. Plankton Res., 33,
1146– 1150.
Ploug, H., Iversen, M. H., Koski, M. et al. (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.
Pomeroy, L. R. (1992) The microbial food web. Oceanus, 35, 28– 35.
Pomeroy, L. R. and Deibel, D. (1980) Aggregation of organic matter
by pelagic tunicates. Limnol. Oceanogr., 25, 643– 652.
Pomeroy, L. R., Hanson, R. B., Mc Gillivary, P. A. et al. (1984)
Microbiology and chemistry of fecal products of pelagic tunicates,
rates and fates. Bull. Mar. Sci., 35, 426– 439.
Shek, L. and Liu, H. (2010) Oxygen consumption rates of fecal pellets
produced by three coastal copepod species fed with a diatom
Thalassiosira pseudonana. Mar. Poll. Bull., 60, 1005–1009.
Silver, M. W. and Bruland, K. W. (1981) Differential feeding and fecal
pellet composition of salps and pteropods, and the possible origin
of the deep-sea flora and olive-green ‘cells’. Mar. Biol., 67,
193–199.
Thor, P., Dam, H.-G. and Rogers, D. R. (2003) Fate of organic carbon
released from decomposing copepod fecal pellets in relation to bacterial production and ectoenzymatic activity. Aquat. Microb. Ecol., 33,
279 –288.
Turner, J. T. (1979) Microbial attachment to copepod fecal pellets and
its possible significance. Trans. Am. Microsc. Soc., 98, 131– 135.
Turner, J. T. (1984a) Zooplankton feeding ecology, contents of fecal
pellets of the copepods Temora turbinata and T. stylifera from continental shelf and slope waters near the mouth of the Mississippi River.
Mar. Biol., 82, 73– 83.
Turner, J. T. (1984b) Zooplankton feeding ecology, contents of fecal
pellets of the copepods Eucalanus pileatus and Paracalanus quasimodo
from continental shelf waters of the Gulf of Mexico. Mar. Ecol. Prog.
Ser., 15, 27– 46.
Turner, J. T. (2002) Zooplankton fecal pellets, marine snow and
sinking phytoplankton blooms. Aquat. Microb. Ecol., 27, 57–102.
Wexels Riser, C., Reigstad, M., Wassmann, P. et al. (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.
1549