MARINE ECOLOGY PROGRESS SERIES
Mar Ecol Prog Ser
Vol. 419: 57?69, 2010
doi: 10.3354/meps08840
Published November 30
INTRODUCTION
Transparent exopolymer particles (TEP) are non-
living organic particles formed abiotically from acid
polysaccharides released as dissolved and colloidal
matter from phytoplankton and bacteria (Zhou et al.
1998, Passow 2000). TEP have a size ranging between 1
and hundreds of micrometers and are found in high
concentrations in a variety of marine environments
(Alldredge et al. 1993). The TEP pool has an important
effect on the carbon (C) flux of the pelagic zone by ac-
celerating the sedimentation of non-TEP particles
through coagulation-aggregation processes (Logan et
al. 1995, Passow & Alldredge 1995a, Prieto et al. 2002).
Owing to their high C content and abundance, TEP
themselves can directly contribute to C fluxes, repre-
senting a mechanism for the selective sequestration of
C in deep water (Beauvais et al. 2003, Engel et al. 2004).
? Inter-Research 2010 ? www.int-res.com*Email: pedrotti@obs-vlfr.fr
Effects of nutrients and turbulence on the 
production of transparent exopolymer particles: 
a mesocosm study
M. L. Pedrotti1,*, F. Peters2, S. Beauvais1, M. Vidal3, J. Egge4, A. Jacobsen4, C. Marras?2
1Marine Microbial Ecology Group, Laboratoire d?Oc?anographie de Villefranche, CNRS-UMR 7093, BP 28, 
06234 Villefranche-sur-mer, France
2Institut de Ci?ncies del Mar, CMIMA (CSIC), Passeig Mar?tim de la Barceloneta 37-49, 08003 Barcelona, Catalunya, Spain
3Departament d?Ecology, Universitat de Barcelona, Barcelona, Spain
4Department of Fisheries and Marine Biology, University of Bergen, HIB, 5020 Bergen, Norway
ABSTRACT: The production of transparent exopolymer particles (TEP) in response to several envi-
ronmental variables was studied in 2 mesocosm experiments. The first (Expt 1) examined a gradient
of 4 nutrient levels; the second (Expt 2) examined different conditions of silicate availability and zoo-
plankton presence. Tanks were separated in 2 series, one subjected to turbulence and the other not
influenced by turbulence. In tanks with nutrient addition, TEP were rapidly formed, with net appar-
ent production rates closely linked to chl a growth rates, suggesting that phytoplankton cells were
actively exuding TEP precursors. High nutrient availability increased the absolute concentration of
TEP; however, the relative quantity of TEP produced was found to be lower, as TEP concentration per
unit of phytoplankton biomass was inversely related to the initial nitrate dose. In Expt 1, an increase
in TEP volume (3 to 48 ?m equivalent spherical diameter) with nutrient dose was observed; in Expt 2,
both silicate addition and turbulence enhanced TEP production and favored aggregation to larger
TEP (>48 ?m). The presence of zooplankton lowered TEP concentration and changed the size distri-
bution of TEP, presumably by grazing on TEP or phytoplankton. For lower nutrient concentrations,
the ratio of particulate organic carbon (POC) to particulate organic nitrogen (PON) followed the Red-
field ratio. At higher nutrient conditions, when nutrients were exhausted during the post-bloom, a
decoupling of carbon and nitrogen dynamics occurred and was correlated to TEP formation, with a
large flow of carbon channeled toward the TEP pool in turbulent tanks. TEP accounted for an
increase in POC concentration of 50% in high-nutrient and turbulent conditions. The study of TEP
dynamics is crucial to understanding the biogeochemical response of the aquatic system to forcing
variables such as nutrient availability and turbulence intensity.
KEY WORDS:  Transparent exopolymer particles ? TEP production ? Phytoplankton ? Particulate
organic carbon ? POC ? Turbulence ? Nutrients ? Mesocosms
Resale or republication not permitted without written consent of the publisher
Mar Ecol Prog Ser 419: 57?69, 2010
Some factors are known to be important in the
dynamics of TEP production, such as the productivity
and trophic state of the system (Mari et al. 2001, Car-
rias et al. 2002, Passow 2002). In natural environments,
high concentrations of TEP have been associated with
phytoplankton blooms, in particular when diatoms are
dominant (Passow & Alldredge 1995a, Passow et al.
2001). TEP have also been observed during blooms
dominated by other phytoplankton groups such as
dinoflagellates (Passow & Alldredge 1994, Berman &
Viner-Mozzini 2001), cyanobacteria (Grossart & Simon
1997), or prymnesiophytes such as Phaeocystis sp. and
Emiliana huxleyi (Engel 2000, Reigstad et al. 2000,
Mari et al. 2005). In some cases co-variation has been
observed between TEP and phytoplankton biomass in
terms of chl a (Passow & Alldredge 1995b, Beauvais et
al. 2003, Radic et al. 2005, Prieto et al. 2006). However,
the production rate of TEP is thought to be species-
specific (Berman & Viner-Mozzini 2001, Passow 2002)
and changes in the physiological state of cells and in
environmental conditions should then influence the
quantity and quality of phytoplankton-derived organic
matter, which will have an effect on TEP production
(Schuster & Herndl 1995, Underwood et al. 2004,
Claquin et al. 2008). Turbulence and the associated
shear is another factor that affects the dynamics of TEP
(Beauvais et al. 2006). Turbulence is known to influ-
ence the structure of the pelagic trophic web and the
dynamics of the particles suspended within, and
exported from, the water column (Peters & Marras?
2000, Peters et al. 2002). This includes top-down pro-
cesses such as shifts in contact rates between particles
(MacKenzie & Ki?rboe 1995) and bottom-up processes
such as the increase in nutrient flux to cells (Savidge
1981, Ki?rboe 1993, Moeseneder & Herndl 1995, Karp-
Boss et al. 1996, Peters et al. 2006). The availability of
nutrients has been shown to affect the stoichiometry of
organic matter and nutrient uptake, and the composi-
tion of particulate organic matter (POM) is affected by
turbulence (Maar et al. 2002). In laboratory experi-
ments, TEP formation is likely accelerated by turbu-
lence (Schuster & Herndl 1995).
Here we present a study examining the effects of a
nutrient gradient on TEP production in turbulent and
non-turbulent conditions. This work was performed in
the framework of the EC-ELOISE-NATP project (Nu-
trient Dynamics mediated through Turbulence And
Plankton Interactions). Results from a previous experi-
ment (WS1) in which we examined the effects of differ-
ent levels of turbulence on TEP dynamics showed that
turbulence affected TEP production by increasing
coagulation processes (Beauvais et al. 2006). Turbu-
lence promotes TEP aggregation and, depending on its
intensity, modulates C export by maintaining the TEP
pool at the surface or enhancing sedimentation.
In the present mesocosm experiment (WS2), the ob-
jectives were to investigate (1) the production of TEP
under a gradient of nutrients and at different condi-
tions of silicate availability and zooplankton abun-
dance, (2) how turbulence interacts with these factors,
and (3) the way TEP dynamics affect C fluxes in a sys-
tem modulated by interactions between nutrients, tur-
bulence, and plankton.
MATERIALS AND METHODS
Experimental setup. The present work was con-
ducted between 16 June and 7 July 2002 at the Marine
Biological Station, University of Bergen, Norway.
Twelve land-based tanks (2.6 m3, 1.5 m diameter, 1.5 m
depth, open to the air) were filled with seawater col-
lected at 5 m depth and filtered through a 250 ?m mesh
to remove large organisms and macro-detritus (except
for the zooplankton treatment in Expt 2; Table 1).
Tanks were placed inside larger enclosures filled with
circulating seawater to maintain an approximate in
situ temperature. Two series of tanks were prepared.
In 1 series, turbulence (T) was generated with 2 verti-
cally oscillating grids at an energy dissipation rate of
ca. 5 ? 10?2 cm2 s?3, while another series was not sub-
jected to artificial turbulence (S). The grids were made
of Plexiglas that allowed 95% penetration of photosyn-
thetically active radiation (PAR). The grids had a bar
width of 5 cm, a mesh size of 10 cm, and a stroke length
of 40 cm, centered at 45 and 105 cm from the bottom.
Turbulence intensities were calculated from velocity-
time series measured with an acoustic Doppler velo-
cimeter, using the method of linear regression
described in Stiansen & Sundby (2001).
Two mesocosm experiments were performed succes-
sively. The first (Expt 1) consisted of 2 parallel gradi-
58
Table 1. Experimental design and terminology of the 2 exper-
iments. In turbulent tanks the level of turbulence was ca. 5 ?
10?2 cm2 s?3. Note that S8 and T8 experimental conditions of
Expt 1 and Si and TSi of Expt 2 are equivalent treatments
N-NO3, P-PO4, Still Turbulent Zooplankton
Si-SiO3 (?M) tanks tanks
Expt 1
No addition S0 T0 ?
4, 0.25, 8 S4 (A, B, C) T4 (A, B, C) ?
8, 0.50, 16 S8 T8 ?
16, 1.0, 32 S16 T16 ?
Expt 2
8, 0.5, 0 B T ?
8, 0.5, 0 Z TZ (A, B) +
8, 0.5, 16 Si TSi ?
8, 0.5, 16 ZSi TZSi (A, B) +
Pedrotti et al.: Effect of nutrients on TEP dynamics
ents in nutrient levels. At the onset of the experiment,
increasing concentrations of nutrients were added: 0,
4, 8, and 16 ?M of nitrate (N-NO3); phosphate (P-PO4)
supplied at the Redfield ratio; and silicate (Si) at twice
the nitrate dose. Tanks were separated in 2 series: T0,
T4, T8, and T16 for turbulent treatment; and S0, S4, S8,
and S16 for still ones (Table 1). Treatments S4 and T4
were triplicated (A, B, C). The experiment lasted for 6 d
for the first 3 nutrient levels and over a period of 12 d in
tanks where 16 ?M of nitrate was added. The second
experiment (Expt 2) tested the effects of presence/
absence of silicate (Si), zooplankton (Z), and turbu-
lence (T). Initial addition of nitrate and phosphate was
8 ?M and 0.5 ?M respectively since this level resulted
in a significant chlorophyll peak within 6 d in Expt 1.
Two of the treatments were duplicated, designated
TZSi.A and TZSi.B, and TZ.A and TZ.B. It is worth-
while to note that S8 and T8 experimental conditions of
Expt 1 and the Si and TSi ones of Expt 2 were identical
treatments (Table 1), which allowed us to investigate
the importance of the initial plankton community on
TEP dynamics.
Sampling procedures. Integrated samples were
taken daily at 09:00 h from each tank in plastic carboys
(samples for nutrients and chl a were taken twice a day
on the first 3 d). Sub-samples were collected for the
determination of dissolved inorganic nutrients, sus-
pended chl a, particulate organic C (POC), particulate
organic nitrogen (PON), and TEP. Dissolved inorganic
nutrient concentrations were analyzed using standard
methods (Strickland & Parsons 1972). Samples for chl a
were filtered onto different porosity filters extracted in
10 ml 90% acetone for 24 h and analyzed on a Turner
Design Fluorometer (Parsons et al. 1984). Samples for
POC and PON were filtered onto pre-combusted
Whatman GF/F glass-fiber filters and measured in a
Carlo Erba analyzer after exposing the filters to HCl
fumes (Grasshoff et al. 1999).
TEP determination and C content. Several sample
aliquots of different volumes (5, 10, and 15 ml) were
gently filtered through 0.2 ?m white Nuclepore filters
under constant and low pressure. Material retained on
the filter was stained with 500 ?l of an aqueous solu-
tion of 0.06% acetic acid (pH 2.5) and 0.02% Alcian
blue 8GX (Passow & Alldredge 1994). Since Alcian
blue, a specific stain for negatively charged polysac-
charides, may precipitate in the presence of salts, the
filters were rinsed with distilled water between con-
secutive filtrations, and the dye was tested before each
experiment with artificial seawater at 38.2? (SIGMA),
twice filtered onto a 0.2 ?m filter. Blanks did not show
a significant presence of TEP. After the staining proce-
dure, the filters were mounted directly on a white slide
(Cyto-Clear). TEP were observed through an Axio-
phot-Zeiss microscope, by means of a semi-automatic
image-analysis system. The TEP perimeter was traced
semi-automatically and the equivalent spherical diam-
eter (ESD) of an individual TEP was then calculated
from area measurements assuming the geometry of a
sphere. Counts were binned by ESD into 15 logarith-
mic size classes between 1 and 135 ?m. A minimum of
600 TEP for each sample were counted and sized at 3
successive magnifications (100?, 200?, 400?), to cover
the entire size spectra (Beauvais et al. 2003).
Particle-size distributions of TEP were fitted to the
power-law function usually used to describe the parti-
cle-size spectra of ocean particles (Sheldon et al. 1972),
following the equation from McCave (1984):
dN/d(dp)  =  kdp? (1)
where dN is the number of particles per unit volume,
with a diameter ranging dp to [dp + d(dp)], and dp is the
projected spherical encased diameter. The constant k
depends on particle concentration, and the spectral
slope ? relates the abundance of small to large parti-
cles. k and ? were calculated using linear regression of
log[dN/d(dp)] versus log(dp). TEP concentrations were
given in total numbers ml?1 and in total volume (ppm).
For comparison of different treatments, the concen-
tration of TEP belonging to a given size class was
standardized by the width of this size class: TEP
abundance (ml?1 ?m?1) or TEP volume in ppm (?106
?m3 ?m?1).
C content of TEP (TEP-C, ?g C ml?1) was determined
from TEP size-frequency distribution using the size-
dependent relationship proposed by Mari (1999):
TEP-C  =  0.25 ?i ni ri2.55 (2)
where ni is the concentration of particles in size class i
and ri is the TEP radius (?m) in the same size class.
The total C content of phytoplankton was calculated
from an adjusted bulk C:chl a ratio for each experi-
ment. The constraint was that this ratio should give, on
average, an autotrophic C biomass equal to the sum of
phytoplankton biomass from cell volume estimates. A
further constraint was set so that total plankton bio-
mass (autotrophic and heterotrophic) estimated from
cell volumes could not exceed the value of POC. The
C:chl a (weight:weight) ratio was 115 in Expt 1 and 64
in Expt 2.
The effect of turbulence and nutrient dose on TEP
concentrations was statistically compared using analy-
sis of covariance (ANCOVA) with sampling times as a
covariate. Differences between the slopes of regres-
sion lines of size distributions of TEP were statistically
compared with logarithmic transformations in the data.
The data for Time 0 was not included in the analysis, as
treatments had no time to affect the different variables.
Post hoc statistics were also performed. The effect of
turbulence on biological variables was tested with
59
Mar Ecol Prog Ser 419: 57?69, 2010
ANOVA and the Kruskal-Wallis test. Potential rela-
tionships between biological variables were tested
using Pearson?s correlation coefficient or linear regres-
sion. All statistical analyses were done with XLSTAT.
Significance was considered for p < 0.05.
RESULTS
Nutrient dynamics
In Expt 1, before nutrient addition, the concentra-
tions of dissolved inorganic N-NO3, P-PO4, and silicate
were low (0.65, 0.07, and 0.23 ?M, respectively), corre-
sponding to a post-bloom period. Following nutrient
addition (Fig. 1A), the concentration of nitrate de-
creased; about 90% of nitrates were consumed within
4 d in all tanks. In S4, T4, S8, and T8 tanks, by Day 6
the nitrate concentration was near the limit of detec-
tion. In Tanks S16 and T16, the concentration reached
0.25 ?M N-NO3 by the end of the experiment. Phos-
phate concentration decreased within 3 d, remaining
relatively low in all tanks (Fig. 1A), and silicate was not
consumed. In Expt 2, before nutrient addition, N-NO3,
P-PO4, and silicate concentrations were 0.53, 0.04, and
1.13 ?M, respectively. After the addition of 8 ?M N-
NO3 (Fig. 1B), the concentration rapidly decreased
(average of 0.1 ?M on Day 4) and became unde-
tectable on Day 6 in all tanks. Likewise, phosphate
concentrations had decreased on Day 4 to an average
of 0.08 ?M and became undetectable on Day 6 in tur-
bulent tanks where zooplankton were present
(Fig. 1B). In tanks enriched with 16 ?M of silicate, in
turbulent tanks (TSi, TZSi) all silicates were consumed,
while in the still tanks (Si, ZSi) they were not (on Day
6, the silicate concentration was about 10 ?M).
Phytoplankton dynamics
Small autotrophic plastidic flagellates and Syne-
chococcus-like cyanobacteria initially dominated Expt
1 and peaked on Day 3 (average concentration of 3.8 ?
107 and 5.2 ? 108 cells l?1, respectively). At the start of
the experiment, diatoms were present in relatively low
numbers (2.7 ? 105 cells l?1). Fig. 1C shows chl a con-
centration in different treatments over the course of
the experiment. Before nutrient addition, total chl a
concentration was 0.99 ? 0.19 ?g l?1. In tanks without
addition of nutrients (S0, T0), chl a increased slightly to
reach a maximum concentration of 1 ?g l?1 on Day 1. In
S4 and T4 tanks, a peak in chl a concentration of 2.72
and 2.93 ?g l?1 occurred on Day 2 for still and turbulent
tanks, respectively. In S8, T8, and S16 tanks, a peak in
60
Fig. 1. Temporal changes in (A,B) nitrate and phosphate and (C,D) chl a and transparent exopolymer particle (TEP) concentra-
tions in different treatment tanks during Expts (A,C) 1 and (B,D) 2. Samples for nutrients and chl a were taken twice a day on the
first 3 d of the experiments. The last point for Tanks T16 (C) lies outside the axis scale and therefore the number 303 is in the 
upper right corner. See Table 1 for tank abbreviations
Pedrotti et al.: Effect of nutrients on TEP dynamics
chl a (3.9, 5.73, and 6.25 ?g l?1, respectively) occurred
on Day 4. In T16, the exponential growth of phyto-
plankton lasted for ~7 d until peaking at 14.01 ?g chl a
l?1. Chl a concentration was significantly higher in tur-
bulent tanks (Kruskal-Wallis test, p < 0.05). In S0, T0,
S4, T4, and S8 tanks, small phytoplankters (<5 ?m)
comprised the largest fraction (63 ? 4.9%) of the
autotrophic C during all the experiments. In T8, S16,
and T16 tanks, after the chl a peak, a shift in the size
composition of phytoplankton occurred, with >70% of
the autotrophic C composed of cells >20 ?m in size.
The mean value of the percentage of diatom C in total
autotrophic C during the experiment varied from 19%
for S0 to 56% for T16 (Table 2).
In Expt 2, the initial phytoplankton was also made up
of plastidic flagellates (0.5 ? 107 cells l?1) and Syne-
chococcus spp. (2.1 ? 108 cells l?1) but at lower concen-
trations than in Expt 1. The chl a concentration was
1.12 ? 0.14 ?g l?1 before nutrient addition, then in-
creased markedly in turbulent tanks, with a peak of
14.76 ?g l?1 on Day 3 in tanks with the addition of sili-
cates and presence of zooplankton (TZSi; Fig. 1D). Up
to Day 3, the small cells (<5 ?m) represented an aver-
age of 50% of the total autotrophic biomass in all
tanks. By Day 4, diatoms became abundant in the
tanks enriched with silicate. In turbulent tanks (TSi,
TZSi), a shift occurred in the size composition of phyto-
plankton, with >80% of the total autotrophic biomass
composed of cells >20 ?m in size. By Day 6, a peak in
diatom biomass (675 ?gC l?1) was observed in TZSi
tanks. The mean value of the percentage of diatom C
in total autotrophic C were higher in tanks enriched
with silicate (59 ? 9.52%) than in non-enriched tanks
(29 ? 4.14%; Table 2) and in general the contribution of
diatoms to total autotrophic C was higher in Expt 2
than in Expt 1.
TEP concentration and production rates
Expt 1
Variations in TEP volume concentration (ppm) dur-
ing the experiment for the different treatments are
shown in Fig. 1C. Before nutrient addition, TEP con-
centration was 1.78 ppm and it increased in all tanks
during the exponential phytoplankton growth. In tanks
without nutrient addition, TEP volume increased up to
Day 6, reaching a concentration of 50 ppm in turbulent
tanks. In tanks enriched with nutrients, the volume of
TEP increased up to Day 5, reaching an average con-
centration of 50 ppm; then TEP concentration de-
creased on Day 6 in all tanks except in T16 where the
maximum TEP volume (303 ppm) occurred at the end
of the experiment on Day 12 (Fig. 1C). The same trends
were observed for TEP-C; its concentration increased
from 39.1 ?g C l?1 before nutrient addition to up to
concentrations of 1094 ?g C l?1 in still Tanks S8 and
5260 ?g C l?1 in T16 tanks on Day 12 (Table 3). Fig. 2
shows TEP-C (?g C l?1) normalized to the total C con-
tent of phytoplankton (?g C l?1). The ratio TEP:chl a
was higher in tanks with no addition of nutrients and
decreased with the nutrient doses added. The general
trend was that this ratio was higher after the phyto-
61
Table 2. Diatom carbon (C) content in total autotrophic carbon
for the different experimental conditions of Expts 1 and 2.
Mean values (%) over the study period. See Table 1 for tank 
abbreviations. Si: silicate
Expt 1
Still Turbulent
Tank Diatom C (%) Tank Diatom C (%)
S0 19 T0 40
S4 41 T4 31
S8 26 T8 27
S16 55 T16 56
Average 28 ? 11 Average 39 ? 9
Expt 2
Without Si addition With Si addition
Tank Diatom C (%) Tank Diatom C (%)
B 30 Si 79
T 21 TSi 54
Z 34 ZSi 47
TZ 33 TZSi 58
Average 29 ? 4 Average 59 ? 9
0 
5 
10 
15 
20 
0 1 2 3 4 5 6 7 8 9 10 11 12 
R
at
io
 T
EP
-C
:C
-c
hl
 a
Time (d) 
S0 S4 S8 S16 
T0 T4 T8 T16 
Expt
Fig. 2. Carbon (C) content of transparent exopolymer particles
(TEP) (TEP-C; ?g C l?1) normalized by the total C content of
phytoplankton (C-chl a; ?g C l?1) during Expt 1 in tanks with
no addition and increased addition of nutrients. Organic C
due to phytoplankton was calculated using a C:chl a ratio of
115 (see ?Material and methods?). See Table 1 for tank 
abbreviations
Mar Ecol Prog Ser 419: 57?69, 2010
plankton exponential phase, with an overall 2- to
19-fold increase in the amount of TEP with a decrease
in chl a concentration. Little difference between still
and turbulent tanks was observed (Fig. 2).
When all data were pooled together, TEP abundance
was not directly affected by turbulence and nutrients
(ANCOVA, n = 54, p = 0.40 and 0.28 respectively).
However, during the exponential phase of phytoplank-
ton, the kinetics of TEP formation followed the initial
input of nutrients (ANCOVA, n = 24, p = 0.008). Before
nutrient addition, TEP abundance was 6.3 ? 106 l?1,
then TEP concentration increased from 6 ? 106 to 40 ?
106 l?1 in 1 d in Tanks S8, T8, S16, and T16; it took 2 d
in Tanks S4 and T4 to reach this concentration and 3 d
in tanks with no addition of nutrients (S0 and T0).
Subsequently, TEP concentration varied from 50 ? 106
to 135 ? 106 l?1 in tanks with no addition of nutrients.
In tanks with nutrients added, TEP concentration var-
ied from 60 ? 106 to 143 ? 106 l?1, except in T16 tanks
where a maximum concentration of 389 ? 106 l?1 was
observed on Day 12 (Table 3). The calculated produc-
tion rates of TEP (d?1) and chl a growth rates (?, d?1)
during the exponential phase of phytoplankton also
increased with nutrient dose (Fig. 3). The highest TEP
production rates (0.80 ? 0.12 d?1) were reached in S8
treatments (Fig. 3). In tanks where no nutrients were
added, the TEP production rate was 0.50 ? 0.09 d?1,
while chl a decreased, leading to negative ? values.
The net apparent increase rates of TEP were closely
linked to chl a growth rates (y = 0.33x + 0.59, r = 0.89)
(Fig. 3).
Expt 2
TEP volume concentration before nutrient addition
was 2.36 ppm; after addition of nutrients, in par-
ticular silicate, TEP dynamics were markedly different
(Fig. 1D). In tanks not enriched with silicates, on Day 3,
TEP volume concentration was 13.6 ? 3.34 ppm, and it
increased to reach on Day 6 an average of 51 and
82.5 ppm in tanks where zooplankton were present (Z,
62
y = 0.33x + 0.59; r = 0.89 
0 
0 
0.5 
S8
S0
T0
0.5 
TE
P 
pr
od
uc
tio
n 
ra
te
 (d
?1
) 
Chl a growth rate (d?1)  
1:1 
Nutrient addition No addition 
Fig. 3. Transparent exopolymer particles (TEP) production
rate and chl a growth rates (?) during the phytoplankton ex-
ponential phase in Expt 1. Each point corresponds to the slope
? SD of the regression lines between the log of TEP concen-
tration and time (Days 1 to 4) of treatment. Chl a decreased in
tanks without nutrient addition (S0, T0), leading to negative 
?-values
Table 3. Transparent exopolymer particle (TEP) determination and carbon (C) content. ?: spectral slope. POC: particulate organic
carbon. PON: particulate organic nitrogen. See Table 1 for tank abbreviations. Grey background: identical experimental 
conditions for Expts 1 and 2
Tank
Expt 1 S0 T0 S4 T4 S8 T8 S16 T16
Average TEP number (?106 l?1) 50.2 56.1 63.8 83.1 94.4 81 79.7 105.1
Max. TEP number (?106 l?1) 94.3 134.9 118.6 124.8 137.2 143.1 112.4 389.0
Average TEP-C content (?g C l?1) 312.0 384.0 319.2 536.4 625.2 596.4 608.4 1057.2
Max. TEP-C content (?g C l?1) 542.4 1068.0 619.2 1039.2 1094.4 973.2 850.8 5259.6
Lower ? value ?3.03 ?3.21 ?3.13 ?3.22 ?3.02 ?2.91 ?3.01 ?3.03
Max. ? value ?2.49 ?2.51 ?2.26 ?2.49 ?2.54 ?2.37 ?2.56 ?2.61
Max. POC (?M) 38.3 29.1 42.5 51.5 54.6 73.6 140.2 160.9
Max. PON (?M) 4.2 3.8 5.9 7.2 9.1 9.1 12.1 13.3
Expt 2 B T Z TZ Si TSi ZSi TZSi
Average TEP number (?106 l?1) 32.7 32.1 25.0 19.2 43.2 46.0 28.6 45.8
Max. TEP number (?106 l?1) 64.4 54.1 35.8 27.8 83.0 77.7 41.3 82.7
Average TEP-C content (?g C l?1) 562.0 569.0 405.6 405.6 670.2 981.4 438.2 814.4
Max. TEP-C content (?g C l?1) 1538.3 1247.6 1031.4 1031.4 1284.1 2156.4 1043.6 1968.5
Lower ? value ?3.11 ?3.24 ?2.93 ?2.49 ?3.29 ?3.14 ?3.13 ?3.12
Max. ? value ?2.74 ?2.45 ?2.14 ?2.00 ?2.52 ?1.99 ?1.90 ?1.69
TEP increase rate (d?1) 0.44 0.49 0.37 0.37 0.48 0.71 0.44 0.60
Chl a growth rate (?, d?1) 0.62 0.53 0.53 0.66 0.51 0.63 0.51 0.70
Pedrotti et al.: Effect of nutrients on TEP dynamics
TZ) and absent (B,T) respectively. In tanks enriched
with silicates, TEP volume concentration reached, on
Day 3, double (26 ppm) that in tanks without silicate
addition, then it increased to reach on Day 5 values of
128 ppm and 166 ppm in turbulent tanks TSi and TZSi
respectively (Fig. 1D). Initial values of C contained in
TEP was 62.9 ?g C l?1, then concentrations followed
the same patterns as TEP volume, with the maximum
concentration seen in turbulent tanks enriched with
silicates (2156 and 1969 ?g C l?1 in TSi and TZSI, re-
spectively; Table 3). TEP abundance before nutrient
addition was 9.4 ? 106 l?1, and after nutrient addition
TEP concentrations increased in all tanks, with maxi-
mum concentrations of 78 ? 106 and 83 ? 106 l?1 in tur-
bulent tanks enriched with silicates (TSi, TZSi), lead-
ing to a daily production rate of TEP of 0.71 ? 0.05 and
0.60 ? 0.04 d?1 respectively (Table 3). Lower TEP con-
centrations were reached in Z and TZ tanks, with TEP
production rates of 0.37 ? 0.03 (Table 3). No significant
correlation was found between the net apparent in-
crease rates of TEP and chl a growth rates (p = 0.37).
Comparison of Si and TSi treatments with the same ex-
perimental conditions as in Expt 1 (S8 and T8; Table 3)
showed that maximum TEP abundance was reached in
Expt 1, while TEP volume concentration and TEP-C
were higher in Expt 2, especially from Day 4 (Table 3).
Size distribution
During the 2 experiments, the size distribution of
TEP followed a power-law distribution, with highly
significant regression relationships (p < 0.001) in all
treatments. The size spectrum of TEP ranged from 1.08
to 96 ?m ESD. The spectral slopes (? = slope of the
cumulative size distribution) were used to describe the
TEP size distribution, with an increase in ? being due
to an increase of the fraction of large TEP and poten-
tially TEP aggregation. In Expt 1, the size distribution
of TEP was neither influenced by turbulence nor by the
level of nutrients (ANCOVA, p = 0.7 and 0.4 respec-
tively, n = 540), and the value of the slope (?) varied
from ?3.22 in T4 to ?2.26 in S4 (Table 3).
Fig. 4 shows for Expt 1 the average TEP volume con-
centration for each treatment normalized for each TEP
size class (ppm ?m?1). The volume concentrations of
small (<3 ?m) TEP in the total TEP pool was about 10%,
with few differences between treatments, except in
tanks in which 16 ?M nitrate was added, where on Day
12 they represented 5% of total TEP. For TEP with 3 to
48 ?m ESD, results showed an increase in TEP volume in
each size class with nutrient dose. This change was prin-
cipally due to an increase in the contribution of larger
TEP (24 to 48 ?m) to the total TEP pool. At the beginning
of the experiment they represented 12% and on Day 6
they made up 18 and 31% of the total TEP pool in still
and turbulent tanks respectively. On Day 12, in still and
turbulent tanks enriched with 16 ?M nitrate, they repre-
sented respectively 26 and 27% of total TEP. TEP of 4 to
17 ?m size represented an average 66%. The contribu-
tion of larger TEP (>48 ?m) increased from 9% initially
to 18% on Day 12 in T16 tanks. In still tanks, maximum
TEP volume occurred in S8 tanks. In turbulent tanks,
TEP volume was significantly higher in T16 tanks (AN-
COVA, n = 48, p = 0.014).
During Expt 2, the slope of the TEP size distribution
ranged from ?3.29 to ?1.69 (Table 3). The less negative
values indicate an increase in the fraction of large par-
ticles and corresponded to tanks with the addition of
silicates and the presence of zooplankton, ZSi (? = ?1.9
from Day 4 to Day 6) and TZSi (? = ?1.69 on Day 6). In
fact, while TEP size distribution was not directly af-
fected by turbulence (ANCOVA, p = 0.94, n = 420), it
seems that the presence of zooplankton and silicates
significantly affected size spectra (ANCOVA, p = 0.02,
n = 420). Post hoc analyses showed differences be-
tween tanks; TEP were significantly smaller and more
abundant in each size class in tanks with silicates and
63
1 10 100
St
an
da
rd
iz
ed
 T
EP
 v
ol
um
e 
co
nc
en
tra
tio
n 
(p
pm
 ?
m
?1
) 
ESD (?m)
B T0 T4 T8 T16
A S0 S4 S8 S 16
0
0.2
0.4
0.6
0.8
1
1.2
0
0.2
0.4
0.6
0.8
1
1.2
Fig. 4. Mean distribution of transparent exopolymer particles
(TEP) volume concentration in each treatment during Expt 1
in (A) still and (B) turbulent tanks. Data are the average of
Days 1 to 6. The equivalent spherical diameter (ESD) is pre-
sented in log scale. The concentration of TEP (ppm) belong-
ing to a given size class was standardized by the width of this 
size class
Mar Ecol Prog Ser 419: 57?69, 2010
without zooplankton (Si, TSi) than in the other tanks
(p = 0.04). Fig. 5 shows TEP size spectra pooled in tanks
with silicates (Si, TSi) and zooplankton and silicates (Z,
TZ, ZSi, TZSi) from Days 4 to 6. Regression slopes indi-
cated that TEP were relatively smaller in tanks with
the presence of silicates (?2.66 ? 0.01; Fig. 5). In fact,
the size class of 1 to 17 ?m that represented 90% of the
total TEP pool at the beginning remained the most im-
portant during all of the experiment, representing 79
and 49% of total the TEP pool, respectively, on Days 4
and 6. In tanks with the presence of zooplankton, TEP
concentration was lower, with a large fraction of large
particles (slope = ?2.51 ? 0.01; Fig. 5). The most note-
worthy result was the temporal decline in the contribu-
tion of small-sized TEP and the increase in the contri-
bution of larger TEP (Fig. 6). Initially, on Day 0, TEP
1 to 17 ?m in size represented an average 77% of total
the TEP pool, with small differences between tanks;
after the bloom, their contribution decreased, repre-
senting 26% on Day 6. TEP of 24 to 34 ?m ESD in-
creased in number up to Day 4, representing 52% of
the total TEP pool, then larger TEP (>48 ?m) appeared
and made up 65% of total TEP in TZ and ZSi tanks.
C, N, and P dynamics
Fig. 7 shows changes in nitrate, POC, and PON con-
centrations during Expt 1. Before nutrient addition, POC
and PON concentrations were 18.32 and 2.95 ?M, re-
spectively. In S0, T0, S4, T4, and S8 tanks, there were
small increases in POC and PON concentration during
the experiment, with maximum concentrations
reached on Day 4 in T8 tanks (54.6 and 9.1 ?M respec-
tively; Table 3). The C:N ratio of POM was on average
64
0 
1 
2 
3 
4 
5 
0.5 1.0 1.5 2.0 
Lo
g 
TE
P 
no
. (
m
l?1
 ?
m
?1
)  
 
Log ESD (?m)
With zooplankton and silicates 
With silicates (Si-TSi) 
Expt 2 
Fig. 5. Transparent exopolymer particles (TEP) number (no.)
and size spectra (equivalent spherical diameter, ESD) with
log-log transformation in Expt 2 pooled in tanks with silicates
(Si-TSi) and zooplankton and silicates (Z, TZ, ZSi, TZSi). Re-
gression lines (see Eq. 1) were calculated for data obtained
from Days 4 to 6. The regression slope (?) is ?2.66 ? 0.01 (n =
34, p < 0.001) for Si-TSi tanks and ?2.51 ? 0.01 (n = 61, p <
0.001) for Z, TZ, ZSi, and TZSi tanks. See Table 1 for tank 
abbreviations
 TZSi
 ZSi 
TZ 
 Z 
0 
20 
40 
60 
80 
100 
0 1 2 3 4 5 6 
 T
EP
 c
on
tri
bu
tio
n 
 (%
) 
Time (d) 
Time (d) 
Time (d) 
(A)  1?17 ?m
 TZSi
 ZSi 
TZ 
 Z 
0 
20 
40 
60 
80 
100 
0 1 2 3 4 5 6 
(B)  24?34 ?m 
 TZSi
 ZSi 
TZ 
 Z 
0 
20 
40 
60 
80 
100 
0 1 2 3 4 5 6 
Ta
nk
Ta
nk
Ta
nk
(C)  > 40 ?m 
Fig. 6. Contribution of different sizes of transparent exopoly-
mer particles (TEP) (A: 1?17 ?m, B: 24?34 ?m, C: >40 ?m
equivalent spherical diameter) to the total TEP pool in tanks
with the presence of silicates and zooplankton during Expt 2. 
See Table 1 for tank abbreviations
Pedrotti et al.: Effect of nutrients on TEP dynamics
6.2, very close to the expected Redfield ratio of 6.6
(Fig. 7A). In T8, S16, and T16 tanks, tight co-variation of
POC and PON occurred as long as nutrients were re-
plete; on Day 4, POC and PON average concentrations
were 59.2 and 10.1 ?M respectively. The C:N ratio was
on average 6.0 (Fig. 7B). By Day 5, at the time of nutrient
exhaustion, whereas PON increased slightly, POC con-
centration more than doubled (140 and 161 ?M respec-
tively in S16 and T16 tanks; Table 3), leading to a signif-
icant increase in the C:N ratio of POM (14.3 on Day 11 in
T16 tanks). The increase in the C:N ratio after the bloom
was correlated with TEP formation (r = 0.76, p = 0.028, n
= 8). While no correlation was found between TEP and
POC in S0, T0, S4, T4, and S8 tanks, in tanks with higher
nutrient concentration (T8, S16, T16), the increase in TEP
concentration was positively correlated with POC (y =
0.51x + 15.11, r = 0.6, n = 32, p < 0.001). The contribution
of TEP-C to POC in these tanks was ~50% (data from
Day 12 in T16 tanks was removed from calculation as the
TEP-C value exceeded POC values by 2.8). In Expt 2, the
increase in the POC:PON ratio was positively correlated
with the increase in TEP volume concentrations in tur-
bulent tanks enriched with silicates, while
for the other tanks no correlation with TEP-
C was found (Fig. 8A). The daily change in
dissolved nitrate and phosphate altered the
N:P ratios from initially 16 to <1 at the end
of the study. Low values of the N:P ratio
were coincident with high TEP concentra-
tions (Fig. 8B).
DISCUSSION
The nutrients added at the onset of Expt
1 were supposed to promote the develop-
ment of a diatom bloom; however, a large
initial biomass of small autotrophs includ-
ing coccolithophorids (Emiliania huxleyi)
and Synechococcus-like cyanobacteria
took up most of the nutrients. E. huxleyi is
a bloom-forming coccolithophorid that in
late spring-early summer can reach con-
centrations of 5 ? 106 to ~108 cells l?1
(Egge & Heimdal 1994, Brussaard et al.
1996). This was confirmed by the high
fraction of small cells present at the
beginning of the experiment and by the
low chl a peaks achieved in the tanks.
Diatoms did not develop quickly and
appeared in significant numbers only by
Day 4. During Expt 2, there was an
increase in the contribution of diatoms to
the total phytoplankton biomass. In both
experiments, the higher TEP concentra-
tion was reached after the bloom in turbulent tanks
enriched with the maximum level of nutrients. How-
ever, when comparing Expts 1 and 2, we observed that
with the same nutrient dose (tanks enriched with 8 ?M
of nitrate and 16 ?m of silicate), higher TEP con-
centrations were generated in Expt 2. In Expt 1, TEP
was principally produced by small phytoplankton,
including E. huxleyi, while in Expt 2 the increase in
diatom concentration supported higher TEP produc-
tion. In marine waters and in experimental studies,
diatoms have been found to be the main producers of
TEP (Ki?rboe & Hansen 1993, Passow & Alldredge
1994, Prieto et al. 2002). Small-scale turbulence that
has little effect on the nutrient flux to small phyto-
plankton could however favor large cells in the compe-
tition for nutrients (Karp-Boss et al. 1996). While in
Expt 1, turbulence did not affect TEP production, it is
likely that in Expt 2, both turbulence and silicate addi-
tion resulted in high TEP production, as these factors
stimulated diatom growth. This stresses the impor-
tance of initial phytoplankton composition and turbu-
lence in TEP dynamics.
65
?POC / ?PON = 6.2 
n = 31 r = 0.92 
0 
30 
60 
90 
120 
150 
180 
0
4
8
12
16
20
0 2 4 6 8 10 12 14 16 
Expt 1 
Addition
?POC / ?PON = 5.3
n = 37 r = 0.97 
0 
30 
60 
90 
120 
150 
180 
0
4
8
12
16
20
0 4 8 12 16 
POC
 (?
M)
POC
 (?
M)
N
O
3 (
?M)
 
N
O
3 (
?M)
 
PON (?M) 
B 
A 
Addition
Fig. 7. Changes in nitrate, particulate organic carbon (POC), and particulate
organic nitrogen (PON) during Expt 1. (A) Pooled data from S0, T0, S4, T4, and
S8 Tanks; (s) nitrate and (r) POC. Regression was calculated from data ob-
tained from Days 0 to 6. (B) Pooled data from T8, S16, and T16 tanks during
the exponential phase (s, nitrate and e, POC) and after nitrate depletion 
(d, nitrate and r, POC). See Table 1 for tank abbreviations
Mar Ecol Prog Ser 419: 57?69, 2010
TEP dynamics under a gradient of nutrients
The increase in nutrient levels yielded higher net in-
crease rates of TEP and accelerated the kinetics of TEP
formation in turbulent tanks (Fig. 2). This suggests that
TEP consists of fresh organic material, derived from ac-
tive phytoplankton cells exuding the precursors. The
combination of nutrients and turbulence, by increasing
the nutrient flux to cells, induces an increase in phyto-
plankton growth rates (Karp-Boss et al. 1996, Peters et
al. 2006). It is likely that in the present study, the pres-
ence of turbulence increased TEP concentration and
subsequently promoted TEP aggregation. This was
shown by the increase in the contribution of larger TEP
(27% of total TEP; Fig. 4) in turbulent tanks and by the
very much higher TEP concentration in T16 tanks on
Day 12. Moreover, the high stickiness coefficient of
TEP (e.g. Dam & Drapeau 1995), even if it was not
measured in the present study, might facilitate this
process. Our data also suggests that TEP production is
not necessarily linked to nutrient availability; while
higher nutrient conditions enhanced the absolute con-
centration of TEP, the relative quantity of TEP pro-
duced to biomass is likely to be lower (Fig. 3). In-
creases in TEP concentration after phytoplankton
blooms have also been observed in other experimental
systems and have been attributed to nutrient limitation
(Corzo et al. 2000, Engel 2000). Corzo et al. (2000)
showed a doubling effect of N availability during cul-
ture experiments with Chaetoceros calcitrans. While
maximum phytoplankton biomass and TEP concentra-
tions were reached at high N supply, the production of
TEP per phytoplankton biomass during the exponen-
tial phase was higher in N-limited culture. This sug-
gests that TEP formation is rather a function of the
physiological state of cells than of the standing stock of
phytoplankton (Schuster & Herndl 1995, Maldonado et
al. 2001). The production of exudates is known to be
higher under nutrient stress (Underwood et al. 2004).
In some cases, a P limitation may stimulate more excre-
tion and therefore induce higher extracellular polysac-
charides production by the cells than an N limitation
(Myklestad 1977, Obernosterer & Herndl 1995). This
could explain the high amount of TEP per phytoplank-
ton biomass observed in the present study in tanks
without the addition of nutrients (S0, T0). In these
tanks, limited nitrate and/or phosphate conditions pre-
vailed from the beginning of the experiment. It is likely
that this limitation increased the relative TEP concen-
tration. When light is not limiting, the availability of
nutrients may limit phytoplankton biomass but not
photosynthesis, and as a consequence the C produced
by phytoplankton is exuded as extracellular carbohy-
drates. However, the linear relationship obtained by
Claquin et al. (2008) between TEP production and pho-
tosynthetic activity in diatom cultures indicated that C
excretion was not simply due to an overflow of C re-
sulting from unbalanced growth, but to a balance be-
tween production and excretion of C. These considera-
tions suggest that TEP produced per unit of N or P
might vary with species and across systems.
Effect of nutrients and the presence of zooplankton
on TEP size distribution
The exponent ? of the size-frequency distribution of
TEP can be used as an indicator of the degree of TEP ag-
gregation. For example, Passow & Alldredge (1994)
showed that at the end of a diatom bloom, ? is around ?2
or lower, indicating a reduction in the fraction of small
particles as the aggregation dominates TEP size distrib-
ution. During Expt 1, ? corresponded to the type of distri-
bution observed by Passow & Alldredge (1994) when di-
atom concentrations were low and never dominating.
Indeed, the contribution of small and medium particles
was relatively more important than larger ones. How-
ever, the increase in TEP volume with nutrient dose in
almost all size classes of TEP, in particular with turbu-
lence, suggests that the interaction of nutrient availabil-
66
0 20 40 60 80 100 120 140
C:N 
ra
tio
N:P 
ra
tio
0 20 40 60 80 100 120 140
TEP (ppm)
A
Turbulence + silicates
Other tanks
20
15
10
5
0
30
25
20
15
10
5
0
B 
Turbulent
Still
Expt 2 
Fig. 8. (A) Correlation between the C:N ratio of particulate
organic matter and transparent exopolymer particles (TEP)
concentration (ppm) during Expt 2 in turbulent tanks en-
riched with silicates (d) (y = 0.08x + 6.20; r = 0.8; n = 14; p <
0.001) and in the other tanks (s) (y = 0.03x + 6.06; r = 0.6; n =
42; p < 0.05). (B) N:P ratio versus TEP concentration (ppm) in
turbulent tanks (d) (y = 37.63x?0.95; r = 0.72; n = 41; p < 0.001)
and in still tanks (s) (y = 40.31x?0.83; r = 0.61; n = 41, p < 0.05)
Pedrotti et al.: Effect of nutrients on TEP dynamics
ity and turbulence could alter the size distributions of
TEP. This was also seen in Expt 2; the lower ? observed
from Day 4 in tanks with zooplankton and silicates indi-
cated that larger TEP dominate the size distribution.
Beauvais et al. (2006) showed that the size distribution of
TEP of size 2 to 40 ?m was not influenced by turbulence
intensity, while a higher turbulence level influenced ag-
gregation of larger TEP (>40 ?m). In the present study,
turbulence did not directly affect the size distribution of
TEP; however, aggregation of larger TEP seems to be fa-
vored by a combined effect of silicates and turbulence
that stimulated diatom growth. It has been reported that
zooplankton activity could affect the size distributions of
TEP by mechanical disruption or ingestion (Passow &
Alldredge 1999, Dilling & Alldredge 2000). Feeding ac-
tivities of zooplankton, in particular copepods and eu-
phausiids, may contribute to the formation of larger TEP
(Prieto et al. 2001). Hence, in the present study it is likely
that zooplankton reduced TEP concentration by increas-
ing aggregation of larger TEP. However, we cannot as-
sert a possible contribution to the formation or ingestion
of TEP as no grazing experiment was done. The advan-
tage of the experimental design chosen for the present
study was the possibility of comparing the effects of a
high number of experimental variables; however, the
limited number of replicates reduced the power of the
analyses. By exploring combinations of various treat-
ments, we were able to gain an ecological understanding
of TEP dynamics.
C dynamics
We observed that the elemental composition of POM
differed from the Redfield ratio of 6.6 by enrichment in
C and a depletion of N relative to P. The magnitude of
this variation depends on the inorganic nutrient condi-
tions associated with several biotic enhancements,
including phytoplankton growth, bacterial recycling,
and possibly changes in protozoan grazing. The de-
coupling of C and N dynamics that occurred during the
post-bloom in enriched nutrient conditions was proba-
bly due to the production of TEP, with a direct relation-
ship between TEP-C and POC. Engel et al. (2002) esti-
mated in mesocosm experiments that TEP production
could explain 40% of decreases in dissolved inorganic
C and an average increase in POC concentration of
35% (Engel et al. 2004). In the present study, the con-
tribution of TEP-C to POC reached ~50% in higher
nutrient conditions. Even if this value seems to be over-
estimated, and sometimes exceeded the value of POC,
we inferred that the increase in the C:N ratio of POM is
explained in part by the formation of TEP rich in C,
with a large flow of C channeled toward the TEP pool
in turbulent tanks. Engel & Passow (2001) reported
TEP-C values for the northern Adriatic Sea represent-
ing 103% of measured POC. According to those
authors, this high value is probably because some vari-
able fraction of TEP-C is measured as DOC, as TEP
cannot be retained quantitatively in GF/F filters. Our
calculations and conversion factors for TEP-C concen-
tration come from the only available theoretical size-
relationships determined from TEP produced in the
laboratory from diatom cultures (Mari 1999). In Expt 1,
small cells other than diatoms comprised >50% of the
phytoplanktonic biomass (Table 2). Assuming that the
production of exudates is likely to be species-specific
(Penna et al. 1999), the C content of TEP changes
between species and the relationship Eq. (2) may
therefore be inaccurate for estimating TEP production
by groups other than diatoms. Additional studies are
necessary in order to estimate the C conversion factor
for TEP for other phytoplankton groups producing
TEP. Another possible explanation for the higher con-
tribution of TEP to POC is that POC was underesti-
mated due to losses by sedimentation. TEP volume
concentrations in turbulent tanks enriched with silicate
were also positively correlated to the POC:PON ratio
and negatively to the N:P ratio. The ratio of C to P in
POM has been found to be higher under turbulence
treatments (Maar et al. 2002). This means that the
increase in organic C per unit of nutrient consumed is
higher in turbulent than in still conditions. Bacteria can
also contribute to variations in the stoichiometry of
organic matter by secreting polysaccharides as free
exopolymers (Decho 1990) or by the mineralization of
aggregates (Smith et al. 1992). It was shown in meso-
cosm experiments that bacteria contribute only 1 to 3%
of TEP production (Stoderegger & Herndl 1999). Con-
sidering these low values we hypothesize that the main
source of TEP in our mesocosms was the exudation of
precursors by phytoplankton rather than by bacteria.
Even if it was not measured in this experiment, bacte-
ria degradation may be an important loss process for
TEP. The high hydrolytic enzyme activity of bacteria in
aggregates, including TEP, suggests their degradation
and/or consumption by bacteria (Smith et al. 1995).
Part of the carbohydrates that formed TEP could be
used as substrates by the bacteria present in our exper-
iment, removing organic C and regulating the C:N
ratio of total organic matter. Indeed, at the end of the 2
experiments, large concentrations of bacteria were
observed in the tanks enriched with 16 ?mol of nitrates
(ca. 3.0 ? 106 on Days 7 and 8; F. Peters unpubl. data).
In a previous mesocosm study, Pedrotti et al. (2009)
showed that bacteria largely colonize TEP. The rate of
bacteria colonization was linked to TEP formation after
the diatom bloom and was significantly higher in
enriched turbulent tanks. Further investigations con-
cerning species-specific phytoplankton activity and
67
Mar Ecol Prog Ser 419: 57?69, 2010
their C excretion will be important to better determine
the dynamics of TEP and therefore to elucidate essen-
tial feedbacks in the C cycle.
CONCLUSIONS
The present study showed the effect of important
bulk ecosystem properties on TEP production. TEP
were influenced by nutrient additions and by effects of
small-scale turbulence. Treatments with high con-
centrations of nutrients increased the kinetics of TEP
formation and the net apparent production rates,
although the highest TEP production was observed in
post-bloom situations when the nutrients were already
exhausted. Both turbulence and silicate addition
resulted in high TEP production, as these factors favor
autotrophic production. The contribution of large TEP
increased with nutrient dose and in the presence of
turbulence. The presence of zooplankton changed the
size distribution of TEP and lowered the concentration
of TEP as predicted, presumably by grazing on TEP or
phytoplankton. Because TEP formation influences the
stoichiometry of POM, the fate of phytoplankton
blooms, and aggregation mechanisms, accurate infor-
mation about this highly dynamic organic C pool is
needed to elucidate processes that drive biogeochemi-
cal cycles. The present study contributes to the knowl-
edge of how TEP dynamics affect C fluxes in a system
modulated by interaction between turbulence and
nutrients, and how these dynamics vary depending on
the composition of plankton assemblages. The signifi-
cance of TEP for the vertical flux of C will depend on
the relative importance of each of the processes we
have discussed (i.e. aggregation-sedimentation, bacte-
rial mineralization, consumption, grazing).
Acknowledgments. This study was supported by EU project
NTAP (EVK3-CT-2000-00022). Access to the Espeland
Marine Biological Station of the Bergen Marine Food Chain
Research Infrastructure was possible through Contract No.
HPRI-CT-1999-00056 of the Improving Human Potential Pro-
gramme of the European Union. F.P. held a Ramon y Cajal
contract. We thank the reviewers for their comments that
helped improve the manuscript. J. Dolan was of great help
editing the English. J. E. Stiansen and F.P. designed the
turbulence system. Turbulence measurements were done by
J. E. Stiansen.
LITERATURE CITED
Alldredge AL, Passow U, Logan BE (1993) The abundance
and significance of a class of large, transparent organic
particles in the ocean. Deep-Sea Res I 40:1131?1140 
Beauvais S, Pedrotti ML, Villa E, Lem?e R (2003) Transparent
exopolymer particle (TEP) dynamics in relation to trophic
and hydrological conditions in the NW Mediterranean.
Mar Ecol Prog Ser 262:97?109
Beauvais S, Pedrotti ML, Egge J, Iversen K, Marras? C (2006)
Effects of turbulence on TEP dynamics under contrasting
nutrient conditions: implications for aggregation and sedi-
mentation processes. Mar Ecol Prog Ser 323:47?57 
Berman T, Viner-Mozzini Y (2001) Abundance and character-
istics of polysaccharide and proteinaceous particles in
Lake Kinneret. Aquat Microb Ecol 24:255?264 
Brussaard CPD, Kempers RS, Kop AJ, Riegman R, Heldal M
(1996) Virus-like particles in a summer bloom of Emiliana
huxleyi in the North Sea. Aquat Microb Ecol 10:105?113 
Carrias JF, Serre JP, Sime-Ngando T, Amblard C (2002) Dis-
tribution, size, and bacterial colonization of pico- and
nano-detrital organic particles (DOP) in two lakes of dif-
ferent trophic status. Limnol Oceanogr 47:1202?1209 
Claquin P, Probert I, Lefebvre S, V?ron B (2008) Effects of
temperature on photosynthetic parameters and TEP pro-
duction in eight species of marine microalgae. Aquat
Microb Ecol 51:1?11 
Corzo A, Morillo JA, Rodriguez S (2000) Production of trans-
parent exopolymer particles (TEP) in cultures of Chaeto-
ceros calcitrans under nitrogen limitation. Aquat Microb
Ecol 23:63?72 
Dam HG, Drapeau DT (1995) Coagulation efficiency, organic-
matter glues, and the dynamics of particles during a
phytoplankton bloom in a mesocosm study. Deep-Sea Res
II 42:111?123 
Decho AW (1990) Microbial exopolymer secretions in ocean
environments: their role(s) in food webs and marine pro-
cesses. Oceanogr Mar Biol Annu Rev 28:73?153
Dilling L, Alldredge AL (2000) Fragmentation of marine snow
by swimming macrozooplankton: a new process im-
pacting carbon cycling in the sea. Deep-Sea Res I 47:
1227?1245
Egge JK, Heimdal BR (1994) Blooms of phytoplankton includ-
ing Emiliana huxleyi (Haptophyta). Effects of nutrient sup-
ply in different N:P ratios. Sarsia 79:333?348
Engel A (2000) The role of transparent exopolymer particles
(TEP) in the increase in apparent particle stickiness during
the decline of a diatom bloom. J Plankton Res 22:485?497 
Engel A, Passow U (2001) Carbon and nitrogen content of
transparent exopolymeric particles (TEP) in relation to
their Alcian Blue adsorption. Mar Ecol Prog Ser 219:1?10 
Engel A, Goldthwait S, Passow A, Alldredge A (2002) Tempo-
ral decoupling of carbon and nitrogen dynamics in a
mesocosm diatom bloom. Limnol Oceanogr 47:753?761 
Engel A, Delille B, Jacquet S, Riebesell U, Rochelle-Newall E,
Terbr?ggen A, Zondervan I (2004) Transparent exopoly-
mer particles and dissolved organic carbon production by
Emiliania huxleyi exposed to different CO2 concentrations:
a mesocosm experiment. Aquat Microb Ecol 34: 93?104 
Grasshoff K, Kremling K, Ehrhardt M (1999) Methods of sea-
water analysis. Wiley-VCH, Weinheim
Grossart HP, Simon M (1997) Formation of macroscopic
organic aggregates (lake snow) in a large lake: the signif-
icance of transparent exopolymer particles, phytoplank-
ton, and zooplankton. Limnol Oceanogr 42:1651?1659 
Karp-Boss L, Boss E, Jumars PA (1996) Nutrient fluxes to-
planktonic osmotrophs in the presence of fluid motion.
Oceanogr Mar Biol Annu Rev 34:71?107
Ki?rboe T (1993) Turbulence, phytoplankton cell size, and the
structure of pelagic food webs. Adv Mar Biol 29:1?72 
Ki?rboe T, Hansen LS (1993) Phytoplankton aggregate forma-
tion: observations of patterns and mechanisms of cell
sticking and the significance of exopolymeric material.
J Plankton Res 15:993?1018 
Logan BE, Passow U, Alldredge AL, Grossar HP, Simon M
(1995) Rapid formation and sedimentation of large aggre-
68
Pedrotti et al.: Effect of nutrients on TEP dynamics
gates is predictable from coagulation rates (half-lives) of
transparent exopolymer particles (TEP). Deep-Sea Res II
42:203?214 
Maar M, Arin L, Simo R, Sala MM, Peters F, Marras? C (2002)
Combined effects of nutrients and small-scale turbulence
in a microcosm experiment. II. Dynamics of organic matter
and phosphorus. Aquat Microb Ecol 29:63?72 
MacKenzie BR, Ki?rboe T (1995) Encounter rates and swim-
ming behaviour of pause-travel and cruise larval fish
predators in calm and turbulent laboratory environments.
Limnol Oceanogr 40:1278?1289 
Maldonado MT, Boyd PW, LaRoche JL, Strzepek R and others
(2001) Iron uptake and physiological response of phyto-
plankton during a mesoscale Southern Ocean iron enrich-
ment. Limnol Oceanogr 46:1802?1808 
Mari X (1999) Carbon content and C:N ratio of transparent
exopolymeric particles (TEP) produced by bubbling exu-
dates of diatoms. Mar Ecol Prog Ser 183:59?71 
Mari X, Beauvais S, Lem?e R, Pedrotti ML (2001) Non-Red-
field C:N ratio of transparent exopolymeric particles in the
northwestern Mediterranean Sea. Limnol Oceanogr 46:
1831?1836 
Mari X, Rassoulzadegan F, Brussaard CPD, Wassmann P
(2005) Dynamics of transparent exopolymeric particles
(TEP) production by Phaeocystis globosa under N- or
P-limitation: a controlling factor of the retention/export
balance? Harmful Algae 4:895?914 
McCave IN (1984) Size spectra and aggregation of suspended
particles in the deep ocean. Deep-Sea Res A 31:329?352 
Moeseneder MM, Herndl GJ (1995) Influence of turbulence
on bacterial production in the sea. Limnol Oceanogr
40:1466?1473 
Myklestad S (1977) Production of carbohydrates by marine
planktonic diatoms. II. Influence of the N:P ratio in the growth
medium on the assimilation ratio, growth rate, and produc-
tion of cellular and extracellular carbohydrates by Chaeto-
ceros affinis var. willei (Gran) Hustedt and Skeletonema
costatum (Grev.) Cleve. J Exp Mar Biol Ecol 29: 161?179 
Obernosterer I, Herndl GJ (1995) Phytoplankton extracellular
release and bacterial growth: dependence on the inor-
ganic N:P ratio. Mar Ecol Prog Ser 116:247?257 
Parsons TR, Maita Y, Lalli CM (1984) A manual of chemical
and biological methods for sea water analysis. Pergamon
Press, Oxford
Passow U (2000) Formation of transparent exopolymeric par-
ticles, TEP, from dissolved precursor material. Mar Ecol
Prog Ser 192:1?11 
Passow U (2002) Production of transparent exopolymeric par-
ticles (TEP) by phyto- and bacterioplankton. Mar Ecol
Prog Ser 236:1?12 
Passow U, Alldredge AL (1994) Distribution, size and bacter-
ial colonization of transparent exopolymer particles (TEP)
in the ocean. Mar Ecol Prog Ser 113:185?198 
Passow U, Alldredge A (1995a) Aggregation of diatoms bloom
in mecocosm: the role of transparent exopolymer particles
(TEP). Deep-Sea Res II 42:99?109 
Passow U, Alldredge A (1995b) A dye-binding assay for the
spectrophotometric measurement of transparent exopoly-
mer particles (TEP). Limnol Oceanogr 40:1326?1335 
Passow U, Alldredge AL (1999) Do transparent exopolymer
particles (TEP) inhibit grazing by the euphausiid Euphau-
sia pacifica? J Plankton Res 21:2203?2217 
Passow U, Shipe RF, Murray A, Pak DK, Brzezinski MA, All-
dredge A (2001) The origin of transparent exopolymeric
particles (TEP) and their role in the sedimentation of par-
ticulate matter. Cont Shelf Res 21:327?346 
Pedrotti ML, Beauvais S, Keros ME, Iversen K, Peters F (2009)
Bacterial colonization of transparent exopolimeric particles
(TEP) in mesocosms under different trophic conditions and
turbulence intensities. Aquat Microb Ecol 55: 301?312 
Penna A, Berluti S, Penna N, Magnani M (1999) Influence of
nutrient ratios on the in vitro extracellular polysaccharide
production by marine diatoms from the Adriatic Sea.
J Plankton Res 21:1681?1690 
Peters F, Marras? C (2000) Effects of turbulence on plankton:
an overview of experimental evidence and some theoreti-
cal considerations. Mar Ecol Prog Ser 205:291?306 
Peters F, Marras? C, Havskum H, Rassoulzadegan F, Dolan J,
Alcaraz M, Gasol JM (2002) Turbulence and the microbial
food web: effects on bacterial losses to predation and on
community structure. J Plankton Res 24:321?331 
Peters F, Arin L, Marras? C, Berdalet E, Sala MM (2006)
Effects of small-scale turbulence on the growth of two
diatoms of different size in a phosphorus-limited medium.
J Mar Syst 61:134?148 
Prieto L, Sommer F, Stibor H, Koeve W (2001) Effects of plank-
tonic copepods on transparent exopolymeric particles (TEP)
abundance and size spectra. J Plankton Res 23: 515?525 
Prieto L, Ruiz J, Echevarria F, Garcia CM and others (2002)
Scales and processes in the aggregation of diatom blooms:
high time resolution and wide size range records in a
mesocosm study. Deep-Sea Res I 49:1233?1253 
Prieto L, Navarro G, Cozar A, Echevarria F, Garcia CM (2006)
Distribution of TEP in the euphotic and upper mesopelagic
zones of the southern Iberian coasts. Deep-Sea Res II 53:
1314?1328 
Radic T, Kraus R, Fuks D, Radic J, Pecar O (2005) Transparent
exopolymeric particles? distribution in the northern Adri-
atic and their relation to microphytoplankton biomass and
composition. Sci Total Environ 353:151?161 
Reigstad M, Wassmann P, Ratkova T, Arashkevich E, Paster-
nak A, ?ygarden S (2000) Comparison of the springtime
vertical export of biogenic matter in three northern Nor-
wegian fjords. Mar Ecol Prog Ser 201:73?89 
Savidge G (1981) Studies of the effects of small-scale turbu-
lence on phytoplankton. J Mar Biol Assoc UK 61:477?488 
Schuster S, Herndl GJ (1995) Formation and significance of
transparent exopolymeric particles in the northern Adri-
atic Sea. Mar Ecol Prog Ser 124:227?236 
Sheldon RW, Prakash A, Sutcliffe WH Jr (1972) The size dis-
tribution of particles in the ocean. Limnol Oceanogr 17:
327?340 
Smith DC, Simon M, Alldredge AL, Azam F (1992) Intense hy-
drolytic enzyme activity on marine aggregates and impli-
cations for rapid particle dissolution. Nature 359: 139?141 
Smith DC, Steward GF, Long RA, Azam F (1995) Bacterial
mediation of carbon fluxes during a diatom bloom in a
mesocosm. Deep-Sea Res II 42:75?97
Stiansen JE, Sundby S (2001) Improved methods for generat-
ing and estimating turbulence in tanks suitable for fish lar-
vae experiments. Sci Mar 65:151?167 
Stoderegger KE, Herndl GJ (1999) Production of exopolymer
particles by marine bacterioplankton under contrasting
turbulence conditions. Mar Ecol Prog Ser 189:9?16 
Strickland JDH, Parsons TR (1972) A practical handbook
of seawater analysis, 2nd edn. Bull Fish Res Board Can 167
Underwood GJC, Boulcott M, Raines CA, Waldron K (2004)
Environmental effects on exoplymer production by marine
benthic diatoms: dynamics, changes in composition, and
pathways of production. J Phycol 40:293?304 
Zhou J, Mopper K, Passow U (1998) The role of surface-active
carbohydrates in the formation of transparent exopolymer
particles by bubble adsorption of seawater. Limnol
Oceanogr 43:1860?1871
69
Editorial responsibility: Hans Heinrich Janssen, 
Oldendorf/Luhe, Germany
Submitted: March 26, 2010; Accepted: September 24, 2010
Proofs received from author(s): November 23, 2010