JOURNAL OF PLANKTON RESEARCH j VOLUME 27 j NUMBER 1 j PAGES 103–119 j 2005

Phytoplankton community assemblage

in the English Channel: a comparison
using chlorophyll a derived from
HPLC-CHEMTAX and carbon
derived from microscopy cell counts
CAROLE A. LLEWELLYN1*, JAMES R. FISHWICK1 AND JERRY C. BLACKFORD1
1
PLYMOUTH MARINE LABORATORY, PROSPECT PLACE, THE HOE, PLYMOUTH, PL1–3DH, UNITED KINGDOM

*CORRESPONDING AUTHOR: call@pml.ac.uk

Received July 2, 2004; accepted in principle September 28, 2004; accepted for publication October 28, 2004; published online November 22, 2004

The phytoplankton community assemblage in surface water of the English Channel (Station L4) was
measured and compared from March 1999 to October 2002 using two different methods. Pigment-
CHEMTAX was used to derive class apportioned chlorophyll a (Chl a) and cell counts obtained
using microscopy were used to derive phytoplankton carbon (phyto-C) estimations. Phyto-C
(10–340
g C L–1) showed a strong linear relationship with total Chl a (0.5–4.8
g Chl a L–1)
when the Chl a:phyto-C ratio was >0.04 (r2 = 0.80; average Chl a:phyto-C = 0.044) but the
relationship was weaker when the Chl a:phyto-C ratio was <0.04 (r2 = 0.42; average Chl a:phyto-
C = 0.013). Correlation between class biomass estimates for phyto-C and Chl a was strong for
diatoms during winter (r2 = 0.80) but poor for other classes during both summer and winter. The
Chl a:phyto-C ratio declined as irradiance increased with strongest correlation on clear sky days in
2001 when diatoms dominated (r2 = 0.80). The Chl a:phyto-C versus irradiance relationship agrees
with that produced from an empirically based dynamic model of phytoplankton acclimation to light.
Results are discussed in relation to differences in the techniques used, the species present, cellular pigment
concentrations and surface mixed layer irradiance.

INTRODUCTION
Quantifying the community composition and biomass of
phytoplankton is essential to understanding the structure
and dynamics of the marine ecosystem and is important in
evaluating the role of the ocean in global carbon cycles.
Phytoplankton are traditionally measured by counting and
identifying cells using inverted light microscopy. However,
there are many difficulties associated with this method; it is
time consuming and more importantly there are large
standard deviations (15–50%) associated with converting
cell counts to carbon estimates (Wilhem et al., 1991;
Schlüter and Havskum, 1997). Direct measurement of
phytoplankton carbon (phyto-C) is difficult because
phyto-C is not easily distinguishable from other sources
of carbon such as non-living organic matter, bacteria and
microzooplankton. Chlorophyll a (Chl a) was recognized
as a potentially useful alternative measure of phytoplank-
ton biomass over 70 years ago (Harvey, 1933). The ability
to separate and unambiguously measure Chl a and acces-
sory pigments using reversed-phase high performance
liquid chromatography (HPLC) techniques, first devel-
oped over 20 years ago (Mantoura and Llewellyn, 1983;
Wright et al., 1991), has led to wide scale mapping of
pigment signatures across the world oceans (Barlow et al.,
1995; Jeffrey et al., 1997a; Gibb et al., 2000; Wright and
van den Enden, 2000; Gibb et al., 2001). Routine separa-
tion of pigments using HPLC is now the most widely
adopted method for estimating and characterising phyto-
plankton biomass and community assemblage. Further-
more, HPLC pigment measurements are becoming
increasingly important in ‘ground-truthing’ satellite remotely

doi:10.1093/plankt/fbh158, available online at www.plankt.oupjournals.org

Journal of Plankton Research Vol. 27 No. 1 Ó Oxford University Press 2004; all rights reserved
Downloaded from https://academic.oup.com/plankt/article-abstract/27/1/103/1510880/Phytoplankton-community-assemblage-in-the-English
by guest
on 02 October 2017
 JOURNAL OF PLANKTON RESEARCH j VOLUME
sensed colour images used to estimate biomass (Trees et al.,
2000; Smyth et al., 2002).
To determine the phytoplankton community abun-
dance in terms of class specific Chl a requires conversion
of pigment data. This can be achieved using either multi-
ple linear regression analysis or matrix factorisation meth-
ods. The most reliable and best developed of these
methods is a matrix factorisation software programme,
CHEMTAX (CHEMical TAXonomy), developed by
Mackey et al. (Mackey et al., 1996). However, conversion
of pigments into class apportioned Chl a is not straightfor-
ward. First, whilst some accessory pigments, for example,
peridinin and alloxanthin are unambiguous markers of
specific phytoplankton classes, many pigments, for exam-
ple, fucoxanthin, (fuco), 190 -butanoyloxyfucoxanthin
(but-fuco) and 190 -hexanoyloxyfucoxanthin (hex-fuco)
are present across several classes with the potential for
ambiguous class assignation. Second, estimation of class
equivalent Chl a requires the pigment:Chl a ratio for any
specific class to remain constant. However, pigment to
Chl a ratios vary for different species and even within a
species, varying according to environmental factors
including light intensity, nutrient concentrations and
growth rate (Goericke and Montoya, 1998; Schlüter
et al., 2000; Henriksen et al., 2002; Garibotti et al., 2003).
One method used within CHEMTAX to try and accom-
modate for potential changes in class pigment:Chl a
ratios is to subset the data according to environmental
conditions. For example, in studies in the western equa-
torial Pacific, the northeastern Atlantic and in the East
Antarctic marginal ice zone, samples were depth-binned,
every 13–40 m, according to light and nutrients (Mackey
et al., 1998; Wright and van den Enden, 2000; Gibb et al.,
2001). Despite limitations, CHEMTAX is recognized as a
powerful approach in studying taxonomic composition
of phytoplankton assemblages (Wright et al., 1996;
Mackey et al., 1998; Gibb et al., 2000; Wright and van
den Enden, 2000; Rodrı́guez et al., 2002; Havskum et al.,
2004).
The relationship between Chl a and phyto-C is not
straightforward. Studies of microalgal cultures have
shown the Chl a:phyto-C ratio to be highly variable
(<0.005 to >0.05), changing in response to irradiance,
nutrient availability and temperature (Geider, 1987). In
culture, the Chl a:phyto-C is high at low irradiance, high
temperatures and under nutrient replete conditions; in
contrast, the Chl a:phyto-C is low under high irradiances
especially under low temperature and under nutrient limit-
ing conditions (Geider, 1987). In the natural environment,
there have been few investigations directly comparing
phyto-C with pigments. Recent investigations on oceanic
phytoplankton communities show high variability in
Chl a:phyto-C ranging from 0.0045 in euphotic coastal
104
Downloaded from https://academic.oup.com/plankt/article-abstract/27/1/103/1510880/Phytoplankton-community-assemblage-in-the-English
by guest
on 02 October 2017
27 j NUMBER 1 j PAGES 103–119 j 2005
Antarctic waters to 0.056 in the deep Chl a maximum in
the Atlantic Ocean (Maraňón et al., 2000; Garibotti et al.,
2003). In the subtropical Atlantic Ocean, Veldhuis and
Kraay (Veldhuis and Kraay, 2004) found that whilst
the eukaryotic phytoplankton community exhibited a
Chl a:phyto-C of 0.0125–0.03 at the surface, there was a
3–7-fold variation with depth. Fortunately, the variation of
Chl a:phyto-C is highly regulated according to irradiance,
temperature and nutrient availability. The interrelation-
ship between these has been modelled using both mech-
anistic regulatory models (Geider et al., 1997, 1998;
Flynn et al., 2001) and empirical approaches (Flynn,
2003). Initially developed and compared to experimental
observations on microalgal cultures, such models have
been adapted to simulate seasonal and latitudinal depen-
dencies of phytoplankton Chl a:phyto-C in the oceans
(e.g. the model of Geider et al. (Geider et al., 1998) by
Taylor et al., 1997; Lefèvre et al., 2003; Blackford et al.,
2004).
Despite wide scale mapping of pigment signatures
across the world oceans and the increasing use of
CHEMTAX, little attention has been focussed on the
significance of pigment concentrations in relation to
phyto-C. Overall there have only been a few studies to
systematically compare phyto-C estimated from micro-
scopy analysis with phytoplankton biomass determined
from HPLC pigment analysis, although such compara-
tive studies have increased since the introduction of
CHEMTAX (Schlüter et al., 2000; Rodrı́guez et al.,
2002; Havskum et al., 2004; Irigoien et al., 2004;
Veldhuis and Kraay, 2004). These comparative studies
generally show that there is good agreement between
the two techniques for large celled diatoms, but the
agreement is poorer for dinoflagellates, prymnesiophytes
and small flagellates. Studies also conclude that good
correlation is hindered by the presence of small cells
which are a problem for microscopy identification and
by the lack of specificity of marker pigments which
reduces taxonomic precision. Because of the ambiguity
of some marker pigments, HPLC-pigment analysis
should be used with supplementary microscopy analysis
to identify species present (Mackey et al., 1996; Jeffrey
et al., 1999; Irigoien et al., 2004). Blind analysis can easily
lead to error because some algal species contain pig-
ments untypical of their class; the Dinoflagellate Gyrodi-
nium sp., for example, contains prymnesiophyte type
pigments rather than peridinin (Irigoien et al., 2004).
Errors that are also associated with estimates of carbon
derived from microscopy which are reliant on biovolume
conversion factors, and these can vary by several fold
(Montagnes et al., 1994; Schlüter et al., 2000).
Although there have been studies directly comparing
HPLC and microscopy results, there is still very little
 C. A. LLEWELLYN ETAL. j QUANTIFYING PLANKTON: HPLC VERSUS MICROSCOPY
understanding of what the pigment community structure
means in terms of phytoplankon carbon and the role of
phytoplankton in the marine carbon cycle. In this study,
we compare total and class specific Chl a derived from
pigment-HPLC and CHEMTAX with total and class
specific phyto-C derived from microscopy cell counts.
We use data collected over a three year period using
phytoplankton samples from surface water in the English
Channel and investigate pigments and phyto-C in rela-
tion to species present, seasonal succession and irradi-
ance. The Chl a:phyto-C ratio is related to irradiance
and compared to simulated ratios derived from phyto-
plankton acclimation models.
METHOD
Site and sampling
The sampling site at Station L4 (www.pml.ac.uk/l4) is a
long-term time-series station in a coastal temperate
environment in the western English Channel situated
10 km south-west of Plymouth at 50
150 N, 4
130 W
(51-m deep). Sub-surface water (3-m depth) was col-
lected, approximately weekly (less in winter), at mid-
morning in 10-L carboys from March 1999 to October
2002. On arrival in the laboratory, phytoplankton from
100-mL seawater were preserved in 1% acid Lugol’s
iodine and formalin fixative for cell counting and phyto-
plankton from 1-L volumes of seawater were vacuum
filtered onto 25-mm GF/F filters and stored in liquid
nitrogen until HPLC pigment analysis.
Microscopy cell count and carbon
estimations
Phytoplankton preserved in Lugol’s iodine and formalin
were settled in sedimentation chambers. Diatoms, auto-
trophic dinoflagellates, flagellates and picoplankton
(0.2–2 mm) were identified and counted to species level
(where possible) using an inverted microscope (Utermöhl,
1958). The dimensions of individual species were mea-
sured using an ocular micrometer, and cell volumes esti-
mated by approximation to the nearest geometric shape
(Kovala and Larrance, 1966). Cell volume was con-
verted to cell carbon content for diatoms and dino-
flagellates using equations of Strathmann (Strathmann,
1967), and for flagellates using equations of Verity et al.
(Verity et al., 1992). Chrysophytes are a class that encom-
pass silicoflagellates (order Dictyochales) and the pico-
plankton ( Jeffrey and Vesk, 1997). Silicoflagellates were
included in the microscopy flagellate counts, whereas pico-
plankton were counted separately. Detailed infor-
mation on cell counts and carbon estimates can be
found in the Web-based L4 database (www.pml.ac.uk/l4).
105
Downloaded from https://academic.oup.com/plankt/article-abstract/27/1/103/1510880/Phytoplankton-community-assemblage-in-the-English
by guest
on 02 October 2017
In addition to estimates calculated according to
Strathmann (Strathmann, 1967), cell carbon was also esti-
mated for diatoms using the relationship described by
Montagnes and Franklin (Montagnes and Franklin, 2001):
logphyto-C ¼ 0:87ðlogV Þ  0:4948
where phyto-C is the mass of carbon (pg) and V the
volume (mm3). Diatom-C estimations calculated using
the regressions of Montagnes and Franklin (Montagnes
and Franklin, 2001) were on average a factor of
1.94 (0.45) higher than those calculated using the
regressions of Strathmann (Strathmann, 1967). To be
consistent with the L4 database, we report estimates
derived using the regressions of Strathmann (Strathmann,
1967) and refer to estimates derived using the regressions
of Montagnes and Franklin (Montagnes and Franklin,
2001) for comparative purposes.
HPLC pigment and CHEMTAX analysis
Pigments were extracted from filtered phytoplankton
into 2-mL methanol containing an internal standard
apo-carotenoate (Sigma-Aldrich Company Ltd.) using
an ultrasonic probe (30 S, 50 W). Extracts were centri-
fuged to remove filter and cell debris (5 min at 4000 r.p.m)
and analysed using reversed-phase HPLC (Hypersil 3 mm
C8 MOS-2) with gradient elution, as described in Barlow
et al. (Barlow et al., 1997), using Thermo-Separations
instrumentation with photo-diode array spectroscopy
(PDA) and Chrom-Quest software. Pigments were iden-
tified using retention time and spectral match using PDA
( Jeffrey et al., 1997b), and pigment concentrations were
calculated using response factors generated from calibra-
tion using a suite of pigment standards (DHI Water and
Environment, Denmark).
CHEMTAX was used to convert pigment concentra-
tions to class apportioned Chl a. CHEMTAX software
(obtained from D Mackey, CSIRO, Tasmania) runs in
MATLAB and together with a matrix of field pigment
data requires an initial input of a matrix of pigment:Chl
a ratios for each phytoplankton class. Ideally, these initial
ratios should resemble, as closely as possible, the pig-
ment:Chl a ratios for the species in the samples under
analysis (Mackey et al., 1996; Jeffrey et al., 1999). We
used input pigment:Chl a ratios (Table I) based on
knowledge of the common phytoplankton classes present
in the English Channel (Holligan and Harbour, 1977;
Rodrı́guez et al., 2000) and on values presented by
Mackey et al. (Mackey et al., 1996) and Llewellyn and
Gibb (Llewellyn and Gibb, 2000). Prasinophytes lacking
prasinoxanthin were included with chlorophytes. The
photoprotective pigments, diadinoxanthin, diatoxanthin
and b,b-carotene and the highly dynamic pigments
 Table I: Input and output pigment:chlorophyll a (Chl a) ratios from CHEMTAX analysis

JOURNAL OF PLANKTON RESEARCH

Chl c3 Peridinin 190 -butanoyloxyfucoxanthin Fucoxanthin 190 -hexanoyloxyfucoxanthin Violaxanthin Alloxanthin Zeaxanthin Gyroxanthin-diester Lutein Chl b

Input
Dinoflagellates 1.06
Cryptophytes 0.23
Prymnesiophytes 0.07 0.02 0.39 0.16
Cyanophytes 0.35
Diatom 0.46
Chrysophytes 0.13 0.93 0.62
Chlorophytes and prasinophytes 0.16 0.31 0.02 0.13
Karenia brevis 0.15 0.21 0.31 0.14 0.05

j
Output for high Chl a:phyto-C

VOLUME
Dinoflagellates 1.06
106

Cryptophytes 0.23
Prymnesiophytes 0.06 0.02 0.37 0.31

27
Cyanophytes 0.35

j
NUMBER
Diatom 0.42
Chrysophytes 0.13 0.93 0.62
Chlorophytes and prasinophytes 0.09 0.05 0.02 0.41

1
K. brevis 0.29 0.04 0.31 0.14 0.05

j
PAGES
Output for low Chl a:phyto-C
Dinoflagellates 1.06

103–119
Cryptophytes 0.24
Prymnesiophytes 0.09 0.03 0.34 0.49
Cyanophytes 0.35

j
Diatom 0.32

2005
Chrysophytes 0.13 0.93 0.62
Chlorophytes and prasinophytes 0.14 0.05 0.03 0.43
K. brevis 0.31 0.04 0.31 0.14 0.05

Downloaded from https://academic.oup.com/plankt/article-abstract/27/1/103/1510880/Phytoplankton-community-assemblage-in-the-English

by guest
on 02 October 2017
 C. A. LLEWELLYN ETAL. j QUANTIFYING PLANKTON: HPLC VERSUS MICROSCOPY
Chl c1+2 were excluded from the pigment ratio matrix.
CHEMTAX uses this pigment ratio matrix with the field
data to iteratively modify the difference between
observed and calculated total pigment concentration
using differences in pigment ratios. The ratio limit,
which sets the maximum percentage by which CHEM-
TAX is allowed to modify the given ratio, was set to the
default 500 for all pigments except Chl a and peridinin
which were set to 50 and 200, respectively.
To accommodate potential changes in pigment:Chl a
ratios the field pigment data set was split into two accord-
ing to Chl a:phyto-C ratio > and <0.04 (from herein
defined as high and low Chl a:phyto-C). Pigment con-
centrations where the ratio of Chl a:phyto-C was high
approximated to samples collected during winter months
(Fig. 2B) and to <100 mmol photons m2 s1 (Fig. 5).
Pigment concentrations where the ratio of Chl a:phyto-C
was low approximated to samples collected during sum-
mer months (Fig. 2B) and to >100 mmol photons m2 s1
(Fig. 5). The CHEMTAX output ratio table from the two
data sets was used to re-analyse the pigment data to
produce final output ratio tables (Table I) and an output
table for the samples apportioning Chl a to each class.
Chlorophytes, prymnesiophytes and prasinophytes, distin-
guishable using HPLC-CHEMTAX but not necessarily
by light microscopy, were summed together as flagellates
enabling direct comparison with microscopy.
Irradiance and nutrient measurements
Irradiance was measured at 1-m depth intervals through-
out the water column with a photosynthetically active
radiation (PAR) sensor (Chelsea Instruments Ltd.) during
2001 and 2002. The depth of the surface mixed layer
(DML) was determined from temperature profiles as the
depth where a change of >0.3 C occurred over 5-m
depth interval. Average irradiance in the surface mixed
layer (SML) was estimated as
IS D50%
ISML ¼
DML
where I is irradiance in SML, IS the surface irradiance and
D50% depth at which irradiance was 50% that of the surface.
Surface water was analysed for nitrate, phosphate and
silicate concentrations using standard laboratory colori-
metric methods (Woodward and Rees, 2001).
Model derivation
We use a model based on Geider et al. (Geider et al.,
1997), modified with a photo-inhibition term, that cal-
culates the light limitation ( fi) on phytoplankton growth
as a function of a variable cell carbon to chlorophyll
107
Downloaded from https://academic.oup.com/plankt/article-abstract/27/1/103/1510880/Phytoplankton-community-assemblage-in-the-English
by guest
on 02 October 2017
ratio, assimilation rate (rass) and temperature (t)
(Blackford et al., 2004).
 
fi ¼ 1:0  eðI =ðrass tÞÞ eðI =ðrass tÞÞ
where  [2.98 (Wm2)1d1) is the initial slope of
the photosynthesis versus irradiance curve,  (0.02
(Wm2)1d1) parameterizes photo-inhibition and 
represents the cellular Chl a to carbon ratio which is
constrained by maximum (max = 0.075) and minimum
(min = 0.0067) values. The proportion of photosynthate
directed to chlorophyll synthesis () is given by
max rass tfi
¼
I 
Solving these equations for  = , i.e. when the pro-
portion of photosynthate directed towards chlorophyll
balances the existing Chl a:phyto-C ratio, we derive the
optimal cellular Chl a:phyto-C ratio over the observed
range of mixed layer irradiance (I ).
RESULTS
Physical and chemical environment
The physical and chemical environment at Station L4 has
been reported previously (Rodrı́guez et al., 2000; Aiken
et al., 2004). In summary and in general, phytoplankton
growth and succession is largely seasonally driven. The
spring bloom usually occurs in April with stratification
and the development of a shallow SML (10–20-m depth).
Over the summer, the thermocline gradually deepens and
then weakens in August. In August, during periods of
calm weather, stratification is transiently reestablished.
Convectional cooling and autumn storms in late Septem-
ber result in the surface mixed layer deepening with
complete vertical mixing through winter. Temperature
varies from 8
C in March to 16
C in August and Sep-
tember. Nitrate concentrations range between 5 and 7 mM
during winter months and decline rapidly to <0.4 mM
after April. Silicate concentrations are highest in winter,
ranging from 3 to 6 mM and declining in spring and
summer to <1 mM.
Pigments
During the period of the study, pigment concentrations
and pigment distributions were highly variable changing
on a weekly basis but showing an underlying seasonal
pattern (Fig. 1). Generally, Chl a increased rapidly to
>4000 ng L1 in spring (except in 2000), fluctuated from
200 to 3000 ng L1 through summer and autumn and
remained relatively constant at 200–1000 ng L1 during
 JOURNAL OF PLANKTON RESEARCH j VOLUME 27 j NUMBER 1 j PAGES 103–119 j 2005

5000 1400
Chl a 1200
Fucoxanthin
4000 (diatom, prymnesiophytes,
1000 chrysophytes,
3000 Karenia brevis)
800

2000 600
400
1000 200
0 0
J MM J S N J M M J S N J MM J S N J MM J S N J M M J S N J MM J S N J M M J S N J M M J S N

500 250
19'-hexanoloxyfucoxanthin, Peridinin
400 200 (dinoflagellates)
(prymnesiophytes, Karenia brevis)

300 150

200 100

100 50

0 0
J M M J S N J MM J S N J MM J S N J M M J S N J MM J S N J M M J S N J MM J S N J MM J S N
350 100
Alloxanthin 19'-butanoyloxyfucoxanthin,
300 80
(Cryptophytes) (prymnesiophytes, chrysophytes,
250 Karenia brevis)
200 60
Pigment (ng L )

150
−1

40
100
20
50
0 0
J M M J S N J M M J S N J M M J S N J M M J S N J M M J S N J M M J S N J MM J S N J M M J S N
100 70
Zeaxanthin 60 Lutein
80 (cyanophytes, (chlorophytes and prasinophytes)
chlorophytes and prasinophytes) 50
60 40

40 30
20
20
10
0 0
J M M J S N J M M J S N J M M J S N J M M J S N J MM J S N J MM J S N J MM J S N J MM J S N
120 20
Violoxanthin Gyroxanthin
100
(chlorophytes and prasinophytes) 15 (Karenia brevis)
80

60 10

40
5
20
0 0
J M M J S N J MM J S N J M M J S N J M M J S N J MM J S N J M M J S N J MM J S N J MM J S N

600 450
Chl c3
500 400 Chl b
(prymnesiophytes, chrysophytes, (chlorophytes and prasinophytes)
Karenia brevis) 350
400 300
300 250
200
200 150
100 100
50
0 0
J M M J S N J M M J S N J M M J S N J MM J S N J M M J S N J M M J S N J M M J S N J MM J S N

Fig. 1. Concentrations of key biomarker pigments and their chemotaxonomic association in surface water from the western English Channel
(Station L4) from March 1999 to October 2002.

108

Downloaded from https://academic.oup.com/plankt/article-abstract/27/1/103/1510880/Phytoplankton-community-assemblage-in-the-English

by guest
on 02 October 2017
 C. A. LLEWELLYN ETAL. j QUANTIFYING PLANKTON: HPLC VERSUS MICROSCOPY

winter (Fig. 1). A range of accessory pigments were although there were increases during September 1999
detected including the carotenoids hex-fuco, fuco, (90 ng L1) paralleled by increases in Chl b indicating
peridinin, but-fuco, diadinoxanthin, diatoxanthin, vio- the presence of prasinophytes (Fig. 1). Lutein concentra-
laxanthin, alloxanthin, lutein, zeaxanthin, gyroxanthin- tions were also low (<60 ng L1) particularly compared
diester (gyrox) and b-b carotene and the chlorophylls, to Chl b (up to 400 ng L1) suggesting that Chl b was
Chl c3, Chl c1+2 and Chl b. The chemotaxonomic asso- associated with non-prasinoxanthin containing prasino-
ciation of key pigments is shown in Fig. 1. phytes (Barlow et al., 1995). Gyrox, specific to the
Fuco (main accessory pigment in diatoms but also in dinoflagellate Karenia brevis, was present during late sum-
prymnesiophytes and chrysophytes) was the dominant mer and autumn (up to 20 ng L1; Fig. 1). More minor
accessory pigment every year (100 to >1000 ng L1) and/or less class-specific carotenoids included violax-
and was the only pigment to show significant correlation anthin, b,b-carotene, diadinoxanthin, diatoxanthin and
with Chl a (r2 = 0.80). Hex-fuco (prymnesiophytes and neoxanthin. Chl c3 concentrations (prymnesiophytes and
exceptional dinoflagellates) was also prominent (5–451 chrysophytes) were low (<50 ng L1) during 1999 and
ng L1) and increased during spring and summer, but 2000 with increases to 250 and 500 ng L1 during April
with most pronounced and persistent increases during and May in 2001 and 2002, respectively (Fig. 1). Chl b
the autumn (Fig. 1). Peridinin (dinoflagellates) was pre- concentrations (chlorophytes and prasinophytes) also
sent throughout the year (typically <200 ng L1) and fluctuated with no seasonal or interannual patterns
increased sharply (up to 220 ng L1) each year in July/ with highest levels (400 ng L1) occurring in autumn
August (Fig. 1). Alloxanthin (cryptophytes; 0–300 ng L1) 1999 (Fig. 1). Chl c1+2 concentrations (70–750 ng L1)
showed no distinct seasonal or interannual trend. followed a similar temporal trend to Chl a and fuco (not
But-fuco (prymnesiophytes and chrysophytes) generally shown in Fig. 1 because it is non-class specific).
paralleled hex-fuco (20–90 ng L1), with differences
indicating the presence of chrysophytes. Zeaxanthin Phyto-C and Chl a
(cyanobacteria and non-prasinoxanthin containing pra- The phyto-C profile showed some similarities to Chl a,
sinophytes) concentrations were low (0–35 ng L1) but there were also major differences (Fig. 2A). There

350 6
A
300 5

250
Carbon (µg L−1)

Chl a (µg L−1)

200 3

150 2

100 1

50 0

0 -1
J M M J S N J M M J S N J M M J S N J M M J S N

0.16
B
Chl a : carbon (µg L−1: µg L −1)

0.14
0.12
0.10
0.08
0.06
0.04
0.02
0.00
J M M J S N J M M J S N J M M J S N J M M J S N

1999 2000 2001 2002

Fig. 2. (A) Phytoplankton carbon (phyto-C) concentrations (black line) estimated using regressions from Strathmann (Strathmann, 1967) and
chlorophyll a (Chl a) determined using high performance liquid chromatography (grey line). (B) Chl a to phyto-C ratios.

109

Downloaded from https://academic.oup.com/plankt/article-abstract/27/1/103/1510880/Phytoplankton-community-assemblage-in-the-English

by guest
on 02 October 2017
 JOURNAL OF PLANKTON RESEARCH j VOLUME
were usually rapid increases in phyto-C during April
(up to 250 mg C L1), varying concentrations during
summer, increases in Autumn (50 to >245 mg C L1)
and low concentrations from November to March
(<50 mg C L1). The Chl a:phyto-C ratio varied
according to time of year, with values >0.04 during
winter months, decreasing to an average of 0.01 during
summer (Fig. 2B). Phyto-C regressed linearly with Chl a
when the data was divided into subsets for high and low
Chl a:phyto-C (Fig. 4):
for high Chl a : phyto-C;
Chl a ¼ 0:0435 phyto-C  0:4026 ðr 2 ¼ 0:80Þ
for low Chl a : phyto-C;
Chl a ¼ 0:0129 phyto-C þ 0:4874 ðr 2 ¼ 0:48Þ
Class specific Chl a and phyto-C
Diatoms were an important component of the phyto-
plankton community, particularly during spring and
summer. During this period, they accounted for up to
77 and 80% (89% using regressions of Montagnes and
Franklin, 2001) of the Chl a and phyto-C, respectively.
Both phyto-C and Chl a estimates showed large
increases during the diatom spring bloom even though
the bloom was caused by a different species each year
(Table II). However, in late May and June 1999, 2000
and 2001, the microscopy based analysis found large
increases in diatom-C that were not observed for
HPLC-CHEMTAX based diatom-Chl a (Fig. 3). In
1999, there were large numbers of Ceratulina pelagica
(78 mg C L1), in June 2000 there were a mixture of
diatoms present and in June 2001, Detonula pumila was
present in large numbers (Table II). In autumn, diatoms
bloomed again; typically consisting of two or three
different species maxima often including Rhizosolenia
delicatula (Table II) although diatom-C and diatom-Chl a
estimates were variable. In winter, diatoms were in low
abundance (and less than that of flagellates) or absent. In
winter, the contribution of diatom-Chl a to total Chl a was
greater than that of diatom-C to phyto-C. In contrast, in
the spring and summer the contribution of diatom-Chl a
to total Chl a contribution was generally less than that of
diatom-C to phyto-C (Fig. 3). Diatom-C was more
strongly correlated with diatom-Chl a when Chl
a:phyto-C ratios were high (r2 = 0.75) than when Chl
a:C ratios were low (r2 = 0.24; Fig. 4).
Flagellates were another important component of the
phytoplankton at L4 and were the dominant class across
all seasons throughout the study; on average they con-
tributed to 53 and 43% of Chl a and phyto-C, respec-
tively (Fig. 3 and Table II). The estimated contribution
110
Downloaded from https://academic.oup.com/plankt/article-abstract/27/1/103/1510880/Phytoplankton-community-assemblage-in-the-English
by guest
on 02 October 2017
27 j NUMBER 1 j PAGES 103–119 j 2005
of flagellate-C to phyto-C may have been underesti-
mated because of difficulties in identifying smaller cells
using microscopy. The overall dominance of flagellates
was partly a result of the persistence of flagellates
throughout the year, with a background (>10 mg C
L1) of small (5 mm) unidentified flagellates being parti-
cularly prominent during winter (Table II). Blooms of
Phaeocystis pouchetti (up to 160 mg C L1) were another
important feature of the flagellates and these occurred
immediately after the spring diatom bloom (Table II).
These blooms contributed significantly to the overall
phyto-C and Chl a biomass particularly during 2001
and 2002. However, although the contribution of
P. pouchetti to flagellates in 2002 was apparent in the
flagellate-Chl a profile, in 2001 P. pouchetti erroneously
contributed to diatom-Chl a (Fig. 4). This suggests that
during 2001 P. pouchetti contained unexpectedly high
fuco concentrations. This incorrect assignment will
have contributed to the poorer correlation of diatom-C
to diatom-Chl a for the low Chl a:phyto-C subset com-
pared to the high Chl a:phyto-C subset. During July
and/or August, Emiliania huxleyi increased in numbers
and formed significant blooms in 1999 and 2001
(Table II). There were other differences between the
phyto-C and Chl a profiles. On several occasions
flagellate-Chl a was high compared to flagellate-C. For
example, during September 1999, increases in the
flagellate-Chl a profile but not in flagellate-C profile
co-occurred with increases in Chl b. This suggests that
increases in prasinophytes and or chlorophytes were
not detected by microscopy resulting in underestimation
of flagellate-C (Fig. 3). Additionally, in April 2002, the
large increase in flagellate-Chl a was not observed in the
flagellate-C profile. In contrast, during October 2002,
large increases in flagellate-C (120 mg C L1) were not
observed for flagellate-Chl a. This was due to the pre-
sence of euglenoid flagellates that contain pigments
indistinguishable from other classes (Fig. 3). The correla-
tion between flagellate-C and flagellate-Chl a was better
when the Chl a:C ratio was high (r2 = 0.49) than when the
ratio was low (r2 = 0.08; Fig. 4).
For most of the year, dinoflagellates were not a sig-
nificant component of the phytoplankton assemblage
contributing to <10% Chl a and <5% phyto-C. An excep-
tion to this was during July to September each year when
large blooms of dinoflagellates occurred contributing
>200 mg C L1 (Fig. 3). The dominant species during
this time was identified as K. brevis (absent in peridinin
but with prymnesiophyte type pigments plus gyrox;
Örnólfsdóttir et al., 2003) although other dinoflagellate
species were also present (Table II). These other species
often including Ceratium tripos succeeded K. brevis (Table II)
and could be distinguished by the presence of peridinin
 Table II: Dominant and bloom forming species in the western English Channel (L4): Selected for species at >10
g C l 1
1999 2000 2001 2002
J F M A M J J A S O N D J F M A M J J A S O N D J F M A M J J A S O N D J F M A M J J A S O

C. A. LLEWELLYN ETAL.
DIATOMS
Ceratulina pelagica
Chaetoceros debilus
Detonula pumila
Eucampia zoodiacus
Guinardia flaccida
Lauderia annulata
Leptocylindricus danicus

j
Nitzschia delicatissum

QUANTIFYING PLANKTON: HPLC VERSUS MICROSCOPY

Odentella sinensis
Rhizolenia delicatula
Rhizolenia shrubsolei
Rhizolenia stolterfothi
Stauroneis membranace
111

Thalassiosira cf. gravida

FLAGELLATES
5 µm
cryptonomad
Emiliania huxleyi
Euglenoids
Dinotryon
Phaeocystis pouchetti

DINOFLAGELLATES
Ceratium tripos
Dinophysis acuta
Gymnodinium cf. pygmaeum
Heterocapsa minuta
Karenia brevis
Mesoporus perforatus

PICOPLANKTON

Class is highlighted a mixture of species present each at <10 mg C l1 but with total >30 mg C l1.

Downloaded from https://academic.oup.com/plankt/article-abstract/27/1/103/1510880/Phytoplankton-community-assemblage-in-the-English

by guest
on 02 October 2017
 JOURNAL OF PLANKTON RESEARCH j VOLUME 27 j NUMBER 1 j PAGES 103–119 j 2005

120 4000
100 Diatoms
2000
80
60 0
40
-2000
20
0 -4000
200 3000
Flagellates
150 2000
1000
100
0
50

60 400
Dinoflagellates
- excluding Karenia brevis 300
40 200
100
20 0

0
20 1400
Carbon (µ g L )
−1

Cryptophytes 1200

Ch a ( ng L )
−1
15 1000
800
10 600
400
5 200
0
0 -200
60 80
Picoplankton/
60
Chrysophytes
40 40
20
20 0
-20
0 -40

3 200
Cyanobacteria
2 150
2 100
1 50
1 0
0
400 400
Karenia brevis
300 300
200
200
100
100 0
0 -100
J M M J S N J M M J S N J M M J S N J M M J S N

1999 2000 2001 2002

Fig. 3. Class carbon (black line) from microscopy and class-chlorophyll a (Chl a) (grey line) from high performance liquid chromatography-
CHEMTAX in surface water in the western English Channel (Station L4) from March 1999 to October 2002.

112

Downloaded from https://academic.oup.com/plankt/article-abstract/27/1/103/1510880/Phytoplankton-community-assemblage-in-the-English

by guest
on 02 October 2017
 C. A. LLEWELLYN ETAL. j QUANTIFYING PLANKTON: HPLC VERSUS MICROSCOPY

High Chl a : phyto-C Low Chl a : phyto-C

6 5
Phytoplankton Phytoplankton
5 4

4
3
3
2
2 y = 0.0435x + 0.4026 y = 0.0129x + 0.4874
r 2 = 0.80 1 r 2 = 0.48
1 P < 0.0001 P < 0.0001

0 0
0 20 40 60 80 100 120 0 100 200 300
1.6 6
Diatoms y = 0.0209x + 0.347 Diatoms
1.4 5 r 2 = 0.24
1.2 P < 0.0001
4
1.0
0.8 3

0.6 2
y = 0.1047x + 0.0387
0.4
r 2 = 0.75 1
0.2 P < 0.0001
Chl a (µg L−1)

0.0 0
0 2 4 6 8 10 12 14 16 0 20 40 60 80 100

3.0 2.0
Flagellates 1.8
Flagellates
2.5 1.6
2.0 1.4
1.2
1.5 1.0
9
0.8
1.0
0.6
y = 0.0231x + 0.5213 y = 0.0031x + 0.3604
0.5 r 2 = 0.49 0.4
r 2 = 0.08
P < 0.0001 0.2 P = 0.010
0.0 0.0
0 20 40 60 80 100 120 140 0 50 100 150 200

0.3 0.5
Dinoflagellates 0.4 Dinoflagellates
0.2 0.4
0.3
0.2
0.3

0.1 0.2
0.2
0.1 y = 0.0015x + 0.031
0.1
r 2 = 0.6561
0.1 P< 0.0001
0.0 0.0
0 5 10 15 20 25 30 0 50 100 150 200 250 300 350
−1
Carbon (µg L )
Fig. 4. Carbon versus chlorophyll a (Chl a) for total phytoplankton, diatoms (carbon estimated using equations of Strathmann, 1967), flagellates
and dinoflagellates for high and low Chl a : phyto-C as defined in text and approximating to irradiance <100 mmol photons m2 s1 and
>100 m mol photons m2 s1, respectively.

(Fig. 3). Overall there were discrepancies between
dinoflagellates-C and dinoflagellate-Chl a profiles. Specifi-
cally, the dinoflagellate-Chl a profile indicated biomass not
observed for dinoflagellate-C (Fig. 3). In a similar manner
to the flagellates, this may have been due to the presence of
small dinoflagellates cells that could not be readily distin-
guished using microscopy or by cells not being sampled
representatively in the smaller volumes used for micro-
scopy. An additional complication when identifying
dinoflagellates using a microscope is in distinguishing

113

Downloaded from https://academic.oup.com/plankt/article-abstract/27/1/103/1510880/Phytoplankton-community-assemblage-in-the-English

by guest
on 02 October 2017
 JOURNAL OF PLANKTON RESEARCH j VOLUME 27 j NUMBER 1 j PAGES 103–119 j 2005

between the autotrophic, heterotrophic and mixotrophic 0.16

components whilst HPLC is measuring photosynthetic 0.14 A
dinoflagellates. 0.12

Chl a : carbon
Estimates of cryptophyte-C were low compared to 0.1
cryptophyte-Chl a (Fig. 3). Cryptophytes have poor pre- 0.08
servation in fixative and are often too small to be dis- 0.06
tinguished from other flagellates. The estimates of 0.04
cryptophyte-Chl a, on the other hand, are less prone to 0.02
error because the assignment of alloxanthin to crypto- 0
phytes in CHEMTAX is unambiguous. The crypto-
0.16
phyte-Chl a profile showed some similarities to the 0.14 B
flagellate-C profile providing evidence that some crypto- 0.12

phytes were counted as flagellates (Fig. 3). Another con-
tributory factor to the low cryptophyte-C estimates
compared to cryptophyte-Chl a estimates may have
resulted from cryptophytes occurring as endosymbionts
in the ciliate Mesodinium spp., (Hibberd, 1977; Gieskes
and Kraay, 1983). Increases in Mesodinium numbers
occurred each year in April/May and in September/
October and additionally during August in 1999, and
these increases were concomitant with increases in
cryptophyte-Chl a (Fig. 3).
The chrysophyte-Chl a estimates showed a similar
pattern to picoplankton-C, with highest and lowest levels
observed during 2001 and 2002, respectively (Fig. 3).
Picoplankton-C was low throughout the study (0–30 mg
C L1), particularly in 2002 (<5 mg C L1), and showed
no interannual or seasonal pattern. There were no sud-
den blooms of picoplankton, but once cell numbers
increased they persisted for many months (Fig. 3).
Most cyanobacteria are too small to be determined
by microscopy. Filamentous cells, however, could be
identified (Fig. 3). Two small increases in cyanobac-
teria-C were seen in June and September 1999. Cyano-
bacteria-Chl a was generally intermittent and low (up to
80 ng L1), with highest concentrations (200 ng L1) in
September 1999 co-occurring with cyanobacteria-C
(Fig. 3).
Chl a:phyto-C and irradiance
The Chl a:phyto-C ratio decreased as a function of
increasing irradiance received in the SML for both
2001 and 2002 (Fig. 5A). Correlation was particularly
strong for samples collected on clear sky days during
2001 (r2 = 0.91; Fig. 5B). The observed relationship is
consistent with that determined in a dynamic regulatory
model for phytoplankton growth and acclimation (Fig.
5B; Blackford et al. 2004). Furthermore, similar relation-
ships between the Chl a:phyto-C ratio and irradiance
were observed for both diatoms and flagellates, with a
strong relationship for flagellates on clear sky days dur-
ing 2001 (Fig. 6).
Chl a : carbon
0.1
2002
0.08
0.06
0.04
0.02
2001
0
0 100 200 300 400 500 600 700 800
Irradiance in SML (µmol photons m −2 s−1)
Fig. 5. Chlorophyll a (Chl a):phyto-C with estimated surface mixed
layer irradiance for samples collected during 2001 (filled squares) and
2002 (open squares). (A) All samples, (B) Samples collected on clear sky
days. y = 0.0217Ln(x) + 0.1353, r2 = 0.91, P < 0.0001 for 2001 (filled
squares); y = 0.0295Ln(x) + 0. 1955, r2 = 0.42, P = 0.11 for 2002
(open squares). Grey lines derived from Geider et al. model (1997) of
phytoplankton acclimation solved to derive the optimal Chl a : phyto-C
for irradiance in the surface mixed layer using a conversion factor of
1 W m2 = 1.89 mmol m–2 s1 (Hall and Rao, 1987). Each grey line,
dark to light, refers to a maximum assimilation rate of 1, 2 and 3 day1,
respectively. Circles represent Chl a : phyto-C determined for
20 cultures grown at an average irradiance of 122 mmol photons
m2 s1 from Llewellyn and Gibb (2000). Open circles = C deter-
mined from biovolume conversion (Strathmann, 1967) and closed
circles = C determined from elemental analysis (majority of open
circles are overlaid with closed circles).
DISCUSSION
Microscopy and HPLC-CHEMTAX
The results show that there is no clear relationship
between biomass estimates derived from pigment –
CHEMTAX and microscopy cell counts. The best cor-
relations between the two estimates were for diatoms
when irradiance levels were low (<100 mmol photons
m2 s1). For each method of determination there are
many considerations that need to be taken into account.
For example, estimating carbon from cell counts is
dependent on cell volume estimations and the equation
for biomass conversion (Montagnes et al., 1994). In this
study, we used two carbon estimations for diatoms.
Those calculated using the regressions of Montagnes
and Franklin (Montagnes and Franklin, 2001) were on
average 1.94 ( 0.45) times higher than those calculated
with the factors of Strathmann (Strathmann, 1967). In a

114

Downloaded from https://academic.oup.com/plankt/article-abstract/27/1/103/1510880/Phytoplankton-community-assemblage-in-the-English

by guest
on 02 October 2017
 C. A. LLEWELLYN ETAL. j QUANTIFYING PLANKTON: HPLC VERSUS MICROSCOPY

0.4
0.35 A Diatoms B Diatoms
Chl a : C diatoms

0.3
0.25
0.2
2001
0.15
0.1
2002
0.05
0

0.14 D Flagellates
C Flagellates
Chl a : C flagellates

0.12
0.1
0.08 2002
0.06
0.04
0.02 2001
0
0 200 400 600 800 0 200 400 600 800
Irradiance in SML (µmol photons m −2 s−1) Irradiance in SML (µmol photons m −2 s−1 )

Fig. 6. Chlorophyll a (Chl a):phyto-C for diatoms and flagellates with estimated surface mixed layer irradiance for samples collected during 2001
and 2002. (A) diatoms all days (B) diatoms on clear sky days. y = 0.0416Ln(x) + 0.3364, r2 = 0.11, p = 0.30 for 2001 (filled squares) and
y = 0.0611Ln(x) + 0.419, r2 = 0.37, P = 0.27 for 2002 (open squares). (C) Flagellates all days, (D) flagellates on clear sky days y =
0.026Ln(x) + 0.1568, r2 = 0.78, P = 0.0001 for 2001 (filled squares) and y = 0.019Ln(x) + 0.1516, r2 = 0.31, p = 0.19 for 2002 (open
squares). Grey lines derived from Geider et al. model (1997) of phytoplankton acclimation solved to derive the optimal Chl a : phyto-C for
irradiance in the surface mixed layer using a conversion factor of 1 W m2 = 1.89 mmol m2 s1 (Hall and Rao, 1987). Each grey line, dark to light,
refers to a maximum assimilation rate of 1, 2 and 3 day1, respectively. Circles represent Chl a : C determined for 10 diatom and 10 flagellate
cultures grown at an average irradiance of 122 mmol photons m2 s1 from Llewellyn and Gibb (Llewellyn and Gibb, 2000). Open circles = C
determined from biovolume conversion (Strathmann, 1967) and closed circles = C determined from elemental analysis (majority of open circles
are overlaid with closed circles).

study on coastal Antarctic waters by Garibotti et al.
(Garibotti et al., 2003), total phytoplankton community
carbon estimated by the factors of Montagnes and
Franklin (Montagnes and Franklin, 2001) and
Montagnes et al. (Montagnes et al., 1994) was roughly
three times higher than total phytoplankton community
carbon estimated using Strathmann’s (Strathmann, 1967)
relationships. In another study, Schlüter et al. (Schlüter
et al., 2000) found that biomass estimated from micro-
scopy from two different laboratories were not consis-
tent. On one date, biomass was 10 times higher than
that found by another laboratory, and on another occa-
sion it was four times lower. The carbon:volume ratio
appears to be very variable. Montagnes et al. (Montagnes
et al., 1994) found variation from 0.04 to 0.4 pg C mm3
in nanoplankton. Llewellyn and Gibb (Llewellyn and
Gibb, 2000) found the ratio varied from 0.02 to
0.57 pg C mm3in diatoms and from 0.2 to 0.4 pg C mm3
in prymnesiophytes. Two-fold variations of carbon:volume
with time of day have been reported for Synechococcus sp.
(Stramski et al., 1995) and Thalassiosira pseudonana (Stramski
and Reynolds, 1993). Another disadvantage of using micro-
scopy is that although a trained microscopist is able to
distinguish individual species, it is difficult to determine
picoflagellates and cryptophytes that are close to the resolu-
tion of the light microscope. Picoflagellates and cryptophytes
are often important components of the natural assemblage.
For the CHEMTAX estimations of phytoplankton the
largest uncertainties are those associated with the initial
input of pigment:Chl a ratios for each class and the
variability of the pigment:Chl a ratio within each class
for the field data. In this study, there were many para-
meters that could affect the pigment:Chl a ratio within
the field data. Not only were the irradiance measure-
ments and nutrient concentrations highly variable but
there was also a highly dynamic community assemblage
with high phytoplankton species diversity. However,
despite this variability during the study the CHEM-
TAX-derived output pigment:Chl a ratios for the two
data sets (low and high Chl a:phyto-C; Table I) were
similar. There were some small increases in ratios for the
low Chl a:phyto-C subset compared to the high Chl
a:phyto-C subset (Table I). This is consistent with Chl a
declining faster than accessory photosynthetic pigments
as irradiance increased and concurs with results from
other studies (Schlüter et al., 2000; Staehr et al., 2002).

115

Downloaded from https://academic.oup.com/plankt/article-abstract/27/1/103/1510880/Phytoplankton-community-assemblage-in-the-English

by guest
on 02 October 2017
 JOURNAL OF PLANKTON RESEARCH j VOLUME
The most notable change was for the hex-fuco:Chl a
ratio in haptophytes which increased from 0.31 for high
Chl a:phyto-C (low light) to 0.49 for low Chl a:phyto-
C (Table II). This also concurs with observations of
Schlüter et al. (Schlüter et al., 2000) who found that for
E. huxleyi, the hex-fuco:Chl a ratio increased from 0.23,
when grown under low light (23 mmol photons m2 s1),
to 0.42, when grown under green light (98 mmol photons
m2 s1) typical of coastal conditions. In this study, fuco
was the only pigment to decrease relative to Chl a in the
low Chl a:phyto-C ratio table compared to the high
Chl a:phyto-C ratio table. Fuco:Chl a decreased from
0.37 to 0.34 for haptophytes and 0.42–0.32 for diatoms
(Table I). This is also consistent with a study on micro-
algal cultures by Schlüter et al. (Schlüter et al., 2000)
where for both diatoms and haptophytes the fuco:Chl a
ratio was higher during low light and stationary phase
growth conditions and declined significantly under high
and green light conditions. Similar results were also found
by Henriksen et al. (Henriksen et al., 2002), where
fuco:Chl a ratio in the diatom Ditylum brightwelli was double
(fuco: Chl a ratio = 0.99) during stationary growth phase
compared to exponential growth phase. In addition,
Henriksen et al. (Henriksen et al., 2002) found that
D. brightwelli was also the only species of 12 in which fuco
was significantly affected by nutrient concentration,
with an increase in fuco:Chl a from 0.67 to 1.09 when
the molar N:P ratio was increased from 5:1 to 16:1 during
stationary growth phase. Our fuco:Chl a ratios did not
reach the high values observed during stationary growth
phase of diatoms by Schlüter et al. (Schlüter et al., 2000) and
Henriksen et al. (Henriksen et al., 2002) indicating that,
throughout the study period, diatoms were not in station-
ary growth phase.
Not only can the pigment:Chl a ratio be highly variable
but the pigment content of cells can also vary according to
environmental condition. Cellular pigment content can
vary by up to a factor of 10 in environments where growth
might be limited by nutrients and light and by a factor of
3–5 in nutrient replete cultures due to photoacclimation
(Falkowski, 1980; Goericke and Repeta, 1993). In a study
by Örnólfsdóttir et al. (Örnólfsdóttir et al., 2003), the
cellular concentration of gyrox for K. brevis varied
from 0.26 and 0.78 pg cell1 depending on the irradiance.
In our study, where gyrox concentrations were <5%
of those reported by Örnólfsdóttir et al. (Örnólfsdóttir
et al., 2003) cellular concentrations ranged from 0.02
to 0.2 pg cell1. Differences in cellular pigment concen-
trations may explain why, during the summer when irra-
diance was high, the CHEMTAX indicated low biomass
of diatoms whereas microscopy suggested the presence of
large amounts of diatom-C (Fig. 3). This highlights
the particular importance of including the influence of
116
Downloaded from https://academic.oup.com/plankt/article-abstract/27/1/103/1510880/Phytoplankton-community-assemblage-in-the-English
by guest
on 02 October 2017
27 j NUMBER 1 j PAGES 103–119 j 2005
irradiance when interpreting pigment data in terms of
phyto-C biomass.
Chl a:phyto-C ratios
Considering the uncertainties associated with both
microscopy carbon and pigment-CHEMTAX estimates
of biomass, it is not surprising that there is large vari-
ability in the Chl a:phyto-C ratio for natural populations
of phytoplankton sampled over the seasons. However,
the estimated Chl a:phyto-C ratios in this study fall
within the range reported in other studies. Our ratios
are consistent with high Chl a relative to phyto-C when
irradiance is low either during winter months or at
deeper depths in the water column (Maranón et al.,
2000; Garibotti et al., 2003; Veldhuis and Kraay,
2004). Phyto-C was more significantly correlated to
Chl a when the Chl a:phyto-C ratio was high. This
occurs most frequently during winter when irradiance
is low, nutrients are non-limiting and hence the pigment
content is likely to be maximal. During winter, irradi-
ance fluctuates less than in summer, and this is a likely
explanation for the stronger correlation between phyto-
C and Chl a for the high Chl a:phyto-C data set (Fig. 4).
On four occasions the Chl a:phyto-C was particularly
high, i.e. >0.08, and much higher than predicted from
the model (Fig. 5). These exceptionally high ratios
occurred between March and May in 2001 and 2002.
In a study using cultures by Llewellyn and Gibb
(Llewellyn and Gibb, 2000), high Chl a:phyto-C ratios
of 0.076 and 0.067 (using Strathmann, 1967 equations)
were obtained for the prymnesiophyte Coccolithus pelagicus
and the diatom Actinocyclus subtilis grown at between 98
and 146 mmol photons m2 s1 (Fig. 5). However, con-
siderably lower ratios of 0.034 and 0.013, respectively,
were obtained where elemental carbon was used rather
than biovolume estimated carbon (Llewellyn and Gibb,
2000). Both C. pelagicus and A. subtilis are exception-
ally large cells (measured cellular volumes 1295 and
20056 mm3, respectively) indicating that the Strathmann
(Strathmann, 1967) equation is not appropriate for cells
of this size and underestimates cellular carbon content.
This may explain the exceptionally high values of Chl
a:phyto-C we observed during March and May when
large bloom forming species were present (Table II).
This may also explain the exceptionally high and unreal-
istic Chl a:phyto-C ratios observed for diatoms (Fig. 6).
In this study, we estimated surface mixed layer irra-
diance but in reality the complex physical mixing at L4
will tend to blur relationships between Chl a:phyto-C
and surface irradiance. Furthermore, our irradiance
measurements are taken at the point of sample collection
and, therefore, do not take into account the previous
light history of the cells. Although there was a stronger
 C. A. LLEWELLYN ETAL. j QUANTIFYING PLANKTON: HPLC VERSUS MICROSCOPY
correlation between phyto-C and Chl a when the Chl a:
phyto- C ratio was high, estimates of phyto-C during
winter are likely to have been underestimated. This is
because during winter there were larger numbers of
pico- and nanoflagellates compared to spring and sum-
mer, and these are more likely to be missed during
microscopy enumeration. In this study, significant con-
centrations of cryptophytes and dinoflagellates were
observed during winter using HPLC, but were absent
in microscopy determinations.
Scope for improving class-Chl a and
carbon estimates
CHEMTAX is the most robust method to date for
partitioning pigment concentrations into class specific
Chl a (Mackey et al., 1996). However, there is scope for
improving the information input to enable more accu-
rate CHEMTAX estimations. Most importantly, better
understanding is required of the relationship between
pigment:Chl a ratios, growth rates and environmental
conditions (including light and nutrients). This includes
extending pigment:Chl a ratios to a wider range of
species than those cited in the CHEMTAX methodol-
ogy paper (Mackey et al., 1996) and those in more recent
literature (Llewellyn and Gibb, 2000; Schlüter et al.,
2000; Henriksen et al., 2002). The ability to accommo-
date variable pigment ratios within the CHEMTAX
programme according to environmental parameters
would provide an important step forward in more accu-
rate assessment of phytoplankton composition. Another
improvement would be the inclusion of more minor
pigments in CHEMTAX matrices. This could be
achieved by using higher resolution HPLC separations.
For example, a more recently developed method enables
greater resolution of the high class specificity polar pig-
ments including a range of Chl c type pigments (Zapata
et al., 2000). This method has been applied to the ana-
lysis of 37 species of cultured haptophytes and distin-
guished eight pigment types based on nine Chl c and five
fuco derivatives, offering oceanographers greater power
for detecting haptophytes in mixed populations while
also distinguishing a greater proportion of them from
diatoms (Zapata et al., 2004). Improved estimates of
carbon, particularly for picoplankton (0.2–2 mm) and
cryptophytes, may be achieved by using epifluorescence
microscopy and/or analytical cell-flow cytometry in
addition to light microscopy and HPLC.
CONCLUSION
Neither microscopic observation nor pigment analysis,
provide a complete solution for phyto-C estimation. The
two techniques measure different parameters and should,
117
Downloaded from https://academic.oup.com/plankt/article-abstract/27/1/103/1510880/Phytoplankton-community-assemblage-in-the-English
by guest
on 02 October 2017
therefore, be considered as complimentary to one another.
Estimates of phytoplankton biomass derived from pigment-
CHEMTAX and microscopy cell counts show large
discrepancies and cannot be directly compared without
considering a large number of different factors. Further-
more, it is widely established that, to interpret HPLC pig-
ment data requires supplementary microscopy cell count
data and or flow cytometric analysis, this is to ensure that
pigment assignments are correct. Additionally and more
importantly, we conclude from this study that if pigments
are to provide quantitative estimates of biomass then infor-
mation on other environmental parameters, in particular
irradiance, must be taken into account. In this study, the
Chl a:phyto-C ratio showed log decline with irradiance
with a relationship consistent with that produced from a
model of phytoplankton acclimation to light. This provides
a way forward in interpreting pigment distributions in
terms of biomass and in quantifying class specific carbon.
ACKNOWLEDGEMENTS
We thank the crews of RV Squilla and RV Sepia for
collecting water; Denise Cummings and Alex Menezes
for HPLC analyses; Derek Harbour for phytoplankton
identification and biomass estimates and Roger Harris
for provision of the L4 database; Katharine Pemberton
and Malcolm Woodward for nutrient data; Jim Aiken
and Claire Widdicombe for encouragement and advice;
and Simon Wright, Ian Joint, Roger Harris, Ruth Airs,
Kevin Flynn and the reviewers for helpful suggestions on
the manuscript. This work forms part of the Natural
Environment Research Council, Plymouth Marine
Laboratory core strategic Research Programme.
REFERENCES
Aiken, J., Fishwick, J., Moore, G. et al. (2004) The annual cycle of
phytoplankton photosynthetic quantum efficiency, pigment compo-
sition and optical properties in the western English Channel. J. Mar.
Biol. Ass. U. K., 84, 301–313.
Barlow, R. G., Cummings, D. G. and Gibb, S. W. (1997) Improved
resolution of mono- and divinyl chlorophylls a and b and zeaxanthin
and lutein in phytoplankton extracts using reverse phase C-8 HPLC.
Mar. Ecol. Prog. Ser., 161, 303–307.
Barlow, R. G., Mantoura, R. F. C., Peinert, R. D. et al. (1995)
Distribution, sedimentation and fate of pigment biomarkers follow-
ing thermal stratification in the western Alboran Sea. Mar. Ecol. Prog.
Ser., 125, 279–291.
Blackford, J. C., Allen, J. I. and Gilbert, F. J. (2004) Ecosystem
dynamics at six contrasting sites: a generic modelling study.
J. Marine. Systems. 52, 191–215
Falkowski, P. G. (1980) Light shade adaptation in marine phytoplank-
ton. In Falkowski, P. G. (ed.), Primary Productivity in the Sea. Plenum,
New York, pp. 99–117.
 JOURNAL OF PLANKTON RESEARCH j VOLUME
Flynn, K. J. (2003) Do we need complex mechanistic photoacclimation
models for phytoplankton? Limnol. Oceanogr., 48, 2243–2249.
Flynn, K. J., Marshall, H. and Geider, R. J. (2001) A comparison of
two N-irradiance models of phytoplankton growth. Limnol. Oceanogr.,
46, 1794–1802.
Garibotti, I. A., Vernet, M., Kozlowski, W. A. et al. (2003) Composition
and biomass of phytoplankton assemblages in coastal Antarctic
waters: a comparison of chemotaxomomic and microscopic analyses.
Mar. Ecol. Prog. Ser., 247, 27–42.
Geider, R. J. (1987) Light and temperature dependence of the carbon to
chlorophyll-a ratio in microalgae and cyanobacteria – implications for
physiology and growth of phytoplankton. New Phytol., 6, 1–34.
Geider, R. J., MacIntyre, H. L. and Kana, T. M. (1997) Dynamic model
of phytoplankton growth and acclimation: Responses of the balanced
growth rate and the chlorophyll a: carbon ratio to light, nutrient-
limitation and temperature. Mar. Ecol. Prog. Ser., 148, 187–200.
Geider, R. J., MacIntyre, H. L. and Kana, T. M. (1998) A dynamic
regulatory model of phytoplanktonic acclimation to light, nutrients,
and temperature. Limnol. Oceanogr., 43, 679–694.
Gibb, S. W., Barlow, R. G., Cummings, D. G. et al. (2000) Surface
phytoplankton pigment distributions in the Atlantic Ocean: an
assessment of basin scale variability between 50
N and 50
S. Prog.
Oceanogr., 45, 339–368.
Gibb, S. W., Cummings, D. G., Irigoien, X. et al. (2001) Phytoplankton
pigment chemotaxonomy of the northeastern Atlantic. Deep Sea
Res. II, 48, 795–823.
Gieskes, W. W. and Kraay, G. W. (1983) Dominance of Crypto-
phyceae during the phytoplankton spring bloom in the central
North Sea detected by HPLC analysis of pigments. Mar. Biol., 75,
179–185.
Goericke, R. and Montoya, J. P. (1998) Estimating the contribution of
microalgal taxa to chlorophyll a in the field – variations of pigment
ratios under nutrient- and light- limited growth. Mar. Ecol. Prog. Ser.,
169, 97–112.
Goericke, R. and Repeta, D. J. (1993) Chlorophyll-a and chlorophyll-b
and divinyl chlorophyll-a and chlorophyll-b in the open subtropical
North-Atlantic Ocean. Mar. Ecol. Prog. Ser., 101, 307–313.
Hall, D. O. and Rao, K. K. (eds) (1987) Photosynthesis. New Studies in
Biology. Edward Arnold Ltd, London.
Harvey, H. W. (1933) Measurement of phytoplankton population.
J. Mar. Biol. Ass. U. K., 34, 761–774.
Havskum, H., Schlüter, L., Scharek, R. et al. (2004) Routine quantifica-
tion of phytoplankton groups – microscopy or pigment analyses.
Mar. Ecol. Prog. Ser., 273, 31–42.
Henriksen, P., Riemann, B., Kaas, H. et al. (2002) Effects of nutrient-
limitation and irradiance on marine phytoplankton pigments.
J. Plankton Res., 24, 835–858.
Hibberd, D. J. (1977) Observations on the ultrastructure of the crypto-
monad endosymbiont of the red-water ciliate Mesodinium rubrum.
J. Mar. Biol. Ass. U. K., 57, 45–61.
Holligan, P. M. and Harbour, D. S. (1977) The vertical distribution
and succession of phytoplankton in the western English Channel in
1975 and 1976. J. Mar. Biol. Ass. U. K., 57, 1075–1093.
Irigoien, X., Meyer, B., Harris, R. et al. (2004) Using HPLC pigment
analysis to investigate phytoplankton taxonomy: the importance of
knowing your species. Helgol. Mar. Res., 58, 77–82.
Jeffrey, S. W., Llewellyn, C. A., Barlow, R. G. et al. (1997a) Pigment
processes in the sea: a selected bibliography. In Jeffrey, S. W.,
118
Downloaded from https://academic.oup.com/plankt/article-abstract/27/1/103/1510880/Phytoplankton-community-assemblage-in-the-English
by guest
on 02 October 2017
27 j NUMBER 1 j PAGES 103–119 j 2005
Mantoura, R. F. C. and Wright, S. W. (eds), Phytoplankton Pigments
in Oceanography. UNESCO Publishing, Paris, pp. 167–178.
Jeffrey, S. W., Mantoura, R. F. C. and Bjørnland, T. (1997b) Data for
the identification of 47 key phytoplankton pigments. In Jeffrey, S. W.,
Mantoura, R. F. C. and Wright, S. W. (eds), Phytoplankton pigments in
oceanography. UNESCO Publishing, Paris, pp. 449–559.
Jeffrey, S. W. and Vesk, M. (1997) Introduction to marine phytoplank-
ton and their signatures. In Jeffrey, S. W., Mantoura, R. F. C. and
Wright, S. W. (eds), Phytoplankton pigments in oceanography. UNESCO
Publishing, Paris, pp. 37–84.
Jeffrey, S. W., Wright, S. W. and Zapata, M. (1999) Recent advances
in HPLC pigment analysis of phytoplankton. Mar. Freshwater Res., 50,
879–896.
Kovala, P. E. and Larrance, J. D. (1966) Computation of phytoplank-
ton cell numbers, cell volume, cell surface and plasma volumne per
litre, from microscopical counts. Department of Oceanography.
University of Washington. Special Report, no. 38.
Lefèvre, N., Taylor, A. H., Gilbert, F. J. et al. (2003) Modelling carbon
to nitrogen and carbon to chlorophyll a ratios in the ocean at low
latitudes: Evaluation of the role of physiological plasticity. Limnol.
Oceanogr., 48, 1796–1807.
Llewellyn, C. A. and Gibb, S. W. (2000) Intra-class variability in the
carbon, pigment and biomineral content of prymnesiophytes and
diatoms. Mar. Ecol. Prog. Ser., 193, 33–44.
Mackey, D. J., Higgins, H. W., Mackey, D. J. et al. (1998) Algal class
abundances in the western equatorial Pacific: estimation from HPLC
measurements of chloroplast pigments using CHEMTAX. Deep-Sea
Res. I, 45, 1441–1468.
Mackey, M. D., Mackey, D. J., Higgins, H. W. et al. (1996)
CHEMTAX-a program for estimating class abundances from
chemical markers: application to HPLC measurements of phyto-
plankton. Mar. Ecol. Prog. Ser., 144, 265–283.
Mantoura, R. F. C. and Llewellyn, C. A. (1983) The rapid determina-
tion of algal chlorophyll and carotenoid pigments and their break-
down products in natural waters by reverse-phase high performance
chromatography. Anal Chimica Acta., 151, 297–314.
Maranón, E., Holligan, P. M., Varela, M. et al. (2000) Basin-scale
variability of phytoplankton biomass, production and growth in the
Atlantic Ocean. Deep-Sea Res. I, 47, 825–857.
Montagnes, D. J. S., Berges, J. A., Harrison, P. J. et al. (1994) Estimat-
ing carbon, nitrogen, protein and chlorophyll a from volume in
marine phytoplankton. Limnol. Oceanogr., 39, 1044–1060.
Montagnes, D. J. S. and Franklin, D. J. (2001) Effect of temperature on
diatom volume, growth rate, and carbon and nitrogen content:
Reconsidering some paradigms. Limnol. Oceanogr., 46, 2008–2018.
Örnólfsdóttir, E. B., Pinckney, J. L. and Tester, P. A. (2003) Quantifica-
tion of the relative abundance of the toxic dinoflagellates, Karenia Brevis
(Dinophyta), using unique photopigments. J. Phycol., 39, 449–457.
Rodrı́guez, F., Fernández, E., Head, R. N. et al. (2000) Temporal
variability of viruses, bacteria, phytoplankton and zooplankton in
the western English Channel off Plymouth. J. Mar. Biol. Ass. U. K.,
80, 575–586.
Rodrı́guez, F., Varela, M. and Zapata, M. (2002) Phytoplankton
assemblages in the Gerlache and Bransfield Straits (Antarctic
Peninsula) determined by light microscopy and CHEMTAX analysis
of HPLC pigment data. Deep-Sea Res. II, 49, 723–747.
Schlüter, L. and Havskum, H. (1997) Phytoplankton pigments in rela-
tion to carbon content in phytoplankton communities. Mar. Ecol.
Prog. Ser., 155, 55–65.
 C. A. LLEWELLYN ETAL. j QUANTIFYING PLANKTON: HPLC VERSUS MICROSCOPY
Schlüter, L. M., Øhlenberg, F., Havskum, H. et al. (2000) The use of
phytoplankton pigments for identifying and quantifying phytoplank-
ton groups in coastal areas: testing the influence of light and nutri-
ents on pigment/chlorophyll a ratios. Mar. Ecol. Prog. Ser., 19, 49–63.
Smyth, T., Groom, S. B., Cummings. D. G. et al. (2002) Comparison
and validation of SeaWiFs bio-optical chlorophyll-a within the
OMEX II programme. Int. J. Remote Sensing, 23, 2321–2326.
Staehr, P. A., Henriksen, P. and Markager, S. (2002) Photoacclimation
of four marine phytoplankton species to irradiance and nutrient
availability. Mar. Ecol. Prog. Ser., 238, 47–59.
Stramski, D. and Reynolds, R. A. (1993) Diel Variations in the optical
properties of a marine diatom. Limnol. Oceanogr., 38, 1347–1364.
Stramski, D., Shalapyonok, A. and Reynolds, R. A. (1995) Optical
charactersisation of the unicellular cyanobacterium Synechococcus
grown under a day-night cycle in natural irradiance. J. Geophys.
Res-Oceans, 100, 13295–13307.
Strathmann, R. R. (1967) Estimating the organic carbon content of
phytoplankton from cell volume or plasma volume. Limnol. Oceanogr.,
12, 411–418.
Taylor, A. H., Geider, G. J. and Gilbert, F. J. H. (1997) Seasonal and
latitudinal dependencies of phytoplankton carbon-to-chlorophyll a
ratios: results of a modelling study. Mar. Ecol. Prog. Ser., 152, 51–66.
Trees, C. C., Clarke, D. K., Bidigare, R. R. et al. (2000) Accessory
pigments versus chlorophyll a concentrations within the euphotic
zone: A ubiquitous relationship. Limnol. Oceanogr., 45, 1130–1143.
Utermöhl, H. (1958) Zur Vervolkommung der Quantitativen
Phytoplankton-methodik. Mitt. Int. Ver. Limnol., 9, 1–38.
Veldhuis, M. L. W. and Kraay, G. W. (2004) Phytoplankton in the
subtropical Atlantic Ocean: towards a better assessment of biomass
and composition. Deep-Sea Res. I, 51, 507–530.
119
Downloaded from https://academic.oup.com/plankt/article-abstract/27/1/103/1510880/Phytoplankton-community-assemblage-in-the-English
by guest
on 02 October 2017
Verity, P. G., Roberton, C. Y., Tronzo, C. R. et al. (1992) Relation-
ships between cell volume and the carbon and nitrogen content of
marine photosynthetic nanoplankton. Limnol. Oceanogr., 48, 357–366.
Wilhem, C., Rudolph, I. and Renner, W. (1991) A quantitative method
based on HPLC-aided pigment analysis to monitor structure and
dynamics of the phytoplankton assemblages – a study from Lake
Meerfelder Maar (Eifel, Germany). Arch. Hydrobiol., 123, 21–35.
Woodward, E. M. S. and Rees, A. P. (2001) Nutrient distributions in an
anticyclonic eddy in the northeast Atlantic Ocean, with reference to
nanomolar ammonium concentrations. Deep-Sea Res. II, 48, 775–793.
Wright, S. W., Jeffrey, S. W., Mantoura, R. F. C. et al. (1991) Improved
HPLC method of analysis of chlorophylls and carotenoids from
marine phytoplankton. Mar. Ecol. Prog. Ser., 77, 183–196.
Wright, S. W., Thomas, D. P., Marchant, H. J. et al. (1996) Analysis of
phytoplankton of the Australian sector of the Southern Ocean:
comparison of microscopy and size frequency data with interpreta-
tions of pigment HPLC data using ‘CHEMTAX’ matrix factorisa-
tion program. Mar. Ecol. Prog. Ser., 144, 285–298.
Wright, S. W. and van den Enden, R. L. (2000) Phytoplankton com-
munity structure and stocks in the East Antarctic marginal ice zone
(BROKE survey, January–March 1996) pigment signatures. Deep-Sea
Res. II, 47, 2363–2400.
Zapata, M., Jeffrey, S. W., Wright, S. W. et al. (2004) Photosynthetic
pigments in 37 species (65 strains) of Haptophyta: implications for
oceanography and chemotaxonomy. Mar. Ecol. Prog. Ser., 270,
83–102.
Zapata, M., Rodrı́guez, F. and Garrido, J. L. (2000) Separation of
chlorophylls and carotenoids in marine phytoplankton: A new
HPLC method using a reverse-phase C8 pyridine-containing mobile
phases. Mar. Ecol. Prog. Ser., 195, 29–45.
Downloaded from https://academic.oup.com/plankt/article-abstract/27/1/103/1510880/Phytoplankton-community-assemblage-in-the-English
by guest
on 02 October 2017