ARTICLE IN PRESS

Deep-Sea Research I 56 (2009) 2162–2174

Contents lists available at ScienceDirect

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

Light absorption by phytoplankton, non-algal particles and dissolved

organic matter at the Patagonia shelf-break in spring and summer
Amábile Ferreira , Virginia M.T. Garcia, Carlos A.E. Garcia
Institute of Oceanography, Federal University of Rio Grande, Av. Italia, Km 8, Rio Grande, RS 96201-900, Brazil

a r t i c l e i n f o abstract

Article history: Satellite image studies and recent in situ sampling have identified conspicuous
Received 3 September 2008 phytoplankton blooms during spring and summer along the Patagonia shelf-break
Received in revised form front. The magnitudes and spectral characteristics of light absorption by total
26 July 2009
particulate matter (phytoplankton and detritus) and colored dissolved organic matter
Accepted 5 August 2009
Available online 14 August 2009
(CDOM) have been determined by spectrophotometry in that region for spring 2006 and
late summer 2007 seasons. In spring, phytoplankton absorption was the dominant
Keywords: optical component of light absorption (60–85%), and CDOM showed variable and
Absorption coefficient important contributions in summer (10–90%). However, there was a lack of correlation
Phytoplankton bloom
between phytoplankton biomass (chlorophyll-a concentration or [chl a]) and the non-
CDOM
algal compartment in both periods. A statistically significant difference was found
Detritus
Patagonian shelf-break between the two periods with respect to the CDOM spectral shape parameter (Scdom),
with means of 0.015 (spring) and 0.012 nm1 (summer). Nonetheless, the mean Scdm
values, which describe the slope of detritus plus CDOM spectra, did not differ between
the periods (average of 0.013 nm1). Phytoplankton absorption values in this work
showed deviations from mean parameterizations in previous studies, with respect to
[chl a], as well as between the two study periods. In spring, despite the microplankton
dominance, high specific absorption values and large dispersion were found
(a*ph(440) ¼ 0.0470.03 m2 mg [chl a]1), which could be attributed to an important
influence of photo-protector accessory pigments. In summer, deviations from general
trends, with values of a*ph(440) even higher (0.0970.02 m2 mg [chl a]1), were due to
the dominance of small cell sizes and also to accessory pigments. These results highlight
the difficulty in deriving robust relationships between chlorophyll concentration and
phytoplankton absorption coefficients regardless of the season period. The validity of a
size parameter (Sf) derived from the absorption spectra has been demonstrated and was
shown to describe the size structure of phytoplankton populations, independently of
pigment concentration, with mean values of 0.41 in spring and 0.72 in summer. Our
results emphasize the need for specific parameterization for the study region and
seasonal sampling approach in order to model the inherent optical properties from
water reflectance signatures.
& 2009 Elsevier Ltd. All rights reserved.

1. Introduction

It is well known that the upwelling of visible radiation

 Corresponding author. Tel.: +55 53 32336617; fax: +55 53 32336601. or reflectance (i.e. apparent optical properties (AOPs))
E-mail addresses: amabilefr@hotmail.com (A. Ferreira), from a natural water body is determined by the inherent
docvmtg@furg.br (V.M.T. Garcia), dfsgar@furg.br (C.A.E. Garcia). optical properties (IOPs, absorption and scattering) of the

0967-0637/$ - see front matter & 2009 Elsevier Ltd. All rights reserved.
doi:10.1016/j.dsr.2009.08.002
 ARTICLE IN PRESS
A. Ferreira et al. / Deep-Sea Research I 56 (2009) 2162–2174
optically active constituents, i.e., particulate and dissolved
material (Gordon et al., 1988). Understanding the pro-
cesses that cause changes in the optical behavior of these
components and quantifying the variability in their IOPs
should improve the accuracy of individual parameter
retrieval by remote-sensing techniques (IOCCG, 2006). In
addition to quantifying these so-called optically active
components, IOPs have been used to estimate biogeo-
chemical variables such as organic carbon, chlorophyll-a,
dissolved organic material and total suspended matter
(e.g. Twardowski et al., 2005), and primary productivity
(Marra et al., 2007; Prieto et al., 2008). Remote sensing
studies are increasingly focused on the understanding of
both the spatial and temporal variability of primary
production over large and regional scales (Behrenfeld et
al., 2001). In addition, it is also desirable to quantify the
role of both phytoplankton and dissolved organic matter
(DOM) in the ocean carbon budget (Siegel et al., 2005; Hu
et al., 2006).
The spectral light absorption is a major factor con-
tributing to the variation in optical properties of
seawater. It can be quantified by the absorption coeffi-
cient, which, in turn, consists of absorption by pure
seawater, phytoplankton, non-algal particles and colored
dissolved organic matter (CDOM). Variability in the
absorption coefficient in oceanic (so-called ‘‘Case I’’)
waters has been thoroughly documented over the past
decades (e.g. Morel, 1988; Bricaud et al., 1998; Reynolds et
al., 2001; Morel and Maritorena, 2001). In such
waters, phytoplankton and their derivative products are
optically dominant, and all other components, except
seawater, are often assumed to co-vary with chlorophyll-a
concentration ([chl a]). However, parameterization of the
absorption coefficient can show significant variability in
space and time, mainly where dissolved material or non-
algal particles are not proportional to phytoplankton
absorption, and it must consider each of them separately
(‘Case 2 waters’). Specifically for phytoplankton, pigment
packaging (Duyens, 1956) or pigment composition gen-
erally can be an important cause of optical variability,
even for a given [chl a] (Bricaud et al., 1995, 2004). Thus,
particular phytoplankton communities can show a given
absorption signature, which may be characteristic of a
given region or time. At least in regions where phyto-
plankton absorption is significant, these variations may
affect the reflectance spectra (Carder et al., 1999;
Maritorena et al., 2002; Carder et al., 2004). Most semi-
analytical algorithm approaches relate reflectance to
absorption by phytoplankton and dissolved organic
matter (e.g. Maritorena et al., 2002). Therefore, there is a
need for regional in situ measurements of light absorption,
in order to assess or to develop satellite ocean color
algorithms.
The performance of semi-analytical models in describ-
ing relationships of ocean color to chlorophyll (Carder
et al., 2004; Sathyendranath et al., 2001) has been
shown to be sensitive to variations in the chlorophyll-
specific absorption coefficient (i.e. absorption coefficient
normalized to chlorophyll-a unit) due to changes in
phytoplankton cell size and pigment composition.
Ciotti et al. (2002) and Bricaud et al. (2004) have verified
2163
that the dominant cell size of the algal community,
and consequent changes in pigment packaging,
could largely explain the natural variability observed in
the spectral shape of phytoplankton absorption. Models
have been proposed to retrieve information on cell size
from spectral absorption measurements (Ciotti et al.,
2002) and from values of the specific absorption
coefficients (Devred et al., 2005). However, those studies
have not separated the relative contributions of
pigment packaging and accessory pigment effects but
have relied on the co-variation between cell size and
accessory pigmentation (Ciotti et al., 1999, 2002; Trees
et al., 2000).
Deviations in the relationship between optical proper-
ties and [chl a] can be particularly important in high-
chlorophyll Case I waters (O’Reilly et al., 1998), such as
offshore phytoplankton blooms. Morel et al. (2006)
analyzed bio-optical relationships in an upwelling area
with no influence from continental runoff, comparing
their natural variability with that previously observed in
less productive waters. They concluded that the main
source of variability is the looseness of the co-variations
between the detrital (dissolved and particulate) compo-
nent and the algal stock.
In the present work, a frontal region (Patagonia shelf-
break front) relatively far from the coast, where extensive
phytoplankton blooms develop in spring-summer, has
been selected for the study of optical properties of both
dissolved and particulate material, including phytoplank-
ton. The data provided in the study contribute to
documentation of absorption properties in relatively clear
and high-chlorophyll waters.
The shelf-break front formed between Argentinean
shelf waters and the Malvinas Current (39–511S) in the
South Atlantic has been shown to be associated with a
conspicuous band of high chlorophyll throughout spring
and summer, detected by ocean color sensors (Acha et al.,
2004; Rivas et al., 2006; Romero et al., 2006). In situ
studies have found high surface [chl a] (between 1.8 and
19.9 mg m3) and primary production rates (between 2.0
and 7.8 g C m2 d1) in spring that are comparable to
maximal seasonal values at eastern boundary currents
(Garcia et al., 2008). The region presents a potentially
significant biological control of gases such as O2 and CO2
in surface layers, as suggested by data in Bianchi et al.
(2005, 2009) and Garcia et al. (2008). Despite the
ecological and biogeochemical relevance of the Patagonia
shelf-break front, there are no records of IOP measure-
ments in the study area.
In summary, the main objective of the present work is
to study the seasonal variation in magnitude and
spectral shape of the absorption coefficients of
phytoplankton, detritus and CDOM in the Patagonian
shelf-break front. The focus is on the absorption
budget during two cruises, carried out in spring 2006
and summer 2007, as part of the PATagonian EXperiment
(PATEX) project, conducted within the Brazilian
Antarctic Program. Thus, the study seeks to contribute to
a better understanding of the phytoplankton absorption
properties in oceanic waters with high-chlorophyll
content.
 ARTICLE IN PRESS
2164 A. Ferreira et al. / Deep-Sea Research I 56 (2009) 2162–2174
2. Material and methods
2.1. Sampling site
The vicinity of the Argentinean shelf-break (39–461S)
was sampled in October 26–29, 2006 (spring, PATEX II
cruise), when 26 stations were occupied, and in March
25–29, 2007 (late summer, PATEX III cruise). There was no
important difference in the location of the oceanographic
stations between the two cruises. An additional transect
was sampled in the northern section during the summer
cruise. Water sampling was always conducted during
daytime, in groups of 6 or 7 stations, with the ship sailing
southwards (PATEX II) and northwards (PATEX III). The
sampling sections are presented in Fig. 1, where three
cross-shelf (two for PATEX II) and two along-shelf
transects are shown on monthly SeaWiFS chlorophyll-a
composites from October 24 to 31, 2006, and from March
22 to 29, 2007, corresponding to sampling periods of
PATEX II and PATEX III.
2.2. Chlorophyll-a concentration
Discrete samples for total [chl a] were collected from
the surface and from selected depths (based on the
fluorescence profile) and filtered onto 25-mm glass fiber
filters (Whatmans GF/F). Chlorophyll size fractionation
was determined both at surface and at sub-surface
fluorescence peak with a 20 mm mesh and then filtration
throughout a 3-mm polycarbonate filter. Chlorophyll a was
Fig. 1. SeaWiFS chlorophyll-a image composites from October 24 to 31, 2006 and from March 22 to 29, 2007, showing the sampling stations (St) during (a)
the PATEX II cruise—October 26–29 (St. 201–226)—and (b) PATEX III cruise—March 25–29 (St. 301–330) along the Patagonian shelf-break.
determined in the three size classes, microplankton
(420 mm), nanoplankton (3–20 mm) and picoplankton
(o3 mm), in order to estimate the contribution by each
class to the phytoplankton biomass. Although some
traditional classification schemes (e.g. Sieburth et al.,
1978) consider picoplankton as o2 mm, in this study,
because of logistical constraints (membrane filter avail-
ability) the o3 mm fraction was considered as picoplank-
ton. After filtration, the filters were wrapped in aluminum
foil and kept in the freezer until the end of the day,
thereafter transferred to liquid nitrogen for later analysis.
At the laboratory, the pigment was extracted in 90%
acetone and fluorescence determined in a Turner Designs
TD-700 fluorometer (previously calibrated with Sigmas
chlorophyll-a standard), following the non-acidification
method of Welschmeyer (1994).
2.3. Absorption coefficients
In samples from the surface and from the fluorescence
peak, filtrations were carried out (Whatmans 25 mm
GF/F) for measurements of light absorption by parti-
culate material and CDOM during the two campaigns.
During the PATEX III cruise, some samples were also
collected from 5 m depth, and they are included as surface
samples.
2.3.1. Particle absorption
Untreated seawater volumes of 300–700 ml, depending
on particle load, were concentrated on GF/F filters that
 ARTICLE IN PRESS
A. Ferreira et al. / Deep-Sea Research I 56 (2009) 2162–2174
were immediately stored in liquid nitrogen prior to
analysis. In the laboratory, optical density values of the
particulate material were determined using a dual-beam
scanning spectrophotometer (Cary Model 1E). Values
corresponding to detritus were determined after extrac-
tion of pigment with pure methanol (Mitchell et al., 2000;
Kishino et al., 1985). Absorption coefficients (m1) of total
particulate matter (first measurement: [apart(l)]) and of
detritus (measurement after extraction: [adet(l)]) were
calculated by converting optical density at each wave-
length to natural log, normalizing by filtered volume and
applying a Beta factor, which corrects for the effect of
concentrating particles on filters (path-length amplifica-
tion). The correction algorithm was determined by
comparing optical density (OD) in suspension and on
filters using four monospecific cultures (Skeletonema
costatum, 10 mm; Thalassiosira oceanica, 14 mm; Phaeo-
cystis sp., 4 mm; a coccoid cyanobacterium, 3 mm),
following procedures described in Mitchell (1990). The
coefficients derived from the beta equation were:
ODsus ¼ 0.4743(ODfilt)2+0.4506ODfilt (r2 ¼ 0.98), where
ODfilt and ODsus correspond to OD due to material retained
on filter and due to suspended material.
Although there is evidence that the relationship
between ODfilt and ODsus may exhibit significant differ-
ences when various species are considered separately
(Moore et al., 1995; Bricaud et al., 1995), this was not the
case in the present work. Despite the large range in cell
sizes of the phytoplankton cultures, no significant differ-
ences were found in the algorithms for each separate
group (not shown). This suggests that variations in the
Beta factor are more associated with different instruments
and methods than with phytoplankton groups or sizes. For
instance, different trends in the relation between ODfilt
and ODsus, initially found between the groups, disap-
peared after proper correction of ODsus for each spectrum
with the ‘turbidity blank’ (value at 750 nm). Thus, the
single relationship generated here appears to be effective
to correct for path-length amplification in our samples,
even considering the different phytoplankton commu-
nities (in both cell sizes and taxonomic structure) during
the two sampled seasons.
The final estimates of absorption of apart(l) and adet(l)
were obtained by subtracting the measured values of
apart(750) and adet(750) from all the measured spectral
values of apart(l) and adet(l) (Babin and Stramski, 2002).
The spectral absorption coefficients for phytoplankton
[aph(l)] were computed as the difference between apart(l)
and adet(l) estimates.
After conversion into absorption coefficients, the
adet(l) data from 350 to 600 nm were used to derive the
spectral slope of detritus absorption (Sdet), following Babin
et al. (2003). In order to examine differences between
phytoplankton absorption spectra, aph(l), values were
normalized by their own average over the visible
spectrum range, according to Garver et al. (1994).
2.3.1.1. Estimation of a cell-size parameter. A cellular size
parameter (Sf) was estimated from each phytoplankton-
normalized spectrum, following Ciotti et al. (2002). Every
normalized spectrum was decomposed using the mixed
2165
spectral model:
aph ðlÞ ¼ a/phS ðlÞ½Sf a/picoS ðlÞ þ ½ð1  Sf Þa/microS ðlÞ;
where a/picoS(l) and a/microS(l) are the ‘‘basis vectors’’ (or
absorption spectra normalized by their own average over
the visible spectrum) corresponding to samples domi-
nated by Prochlorococcus (picoplankton) and samples
dominated by microplankton-sized organisms, and
a/phS(l) is the scaling factor to be applied to the nor-
malized absorption spectrum (Ciotti et al., 2002). Here a
picoplankton vector presented in Ciotti and Bricaud
(2006) was used. The size parameter Sf is a parameter
constrained to vary between 0 and 1 and to specify the
relative contributions of picoplankton and microplankton
to absorption, independently of [chl a] (Ciotti et al., 2002).
2.3.2. CDOM absorption
Filtrates from the particulate absorption sample filters
(Whatman GF/F) were collected in clean borosilicate
bottles that were previously treated according to Mitchell
et al. (2000). Briefly, sequential soaks and rinses in mild
detergent, freshly produced Mili-Q water, 10% HCl with
final copious rinse in Mili-Q water were made, following
by combustion in aluminum foil covers at 450 1C for 4–6 h
(Mitchell et al., 2000). An initial sample volume of 100 ml
was used for rinsing the GF/F filter and the filtration
apparatus. Although 0.2-mm membrane filters are recom-
mended for CDOM analyses, the use of GF/F filters ensures
consistency with the procedure for particulate material
absorption measurement and the estimates of total
absorption (Ciotti and Bricaud, 2006).
The spectral absorbance of the filtered water was
measured against air in a 10-cm quartz cuvette between
300 and 750 nm. Freshly produced Mili-Q water was used
to zero the instrument. After conversion of optical density
values into absorption coefficients (acdom(l), m1), the
same fitting procedure for detritus spectra was applied to
the 350–600 nm spectral absorption range for deriving the
CDOM spectral shape parameter Scdom (Babin et al., 2003).
3. Results
3.1. Total and fractionated chlorophyll-a concentrations
In spring (PATEX II cruise), surface [chl a] varied
between 0.92 and 11.85 mg m3. Lower values were
observed in the first longitudinal profile (north of the
region), where the maximum concentration was 3.56 mg
m3 (St. 205). Generally, higher values (45.0 mg m3)
were associated with the shelf-break front in the middle
of the sampling region (Fig. 1a), decreasing towards south
(o4.5 mg m3, St. 223–226). Peaks in chlorophyll fluores-
cence were generally observed in the sub-surface, be-
tween 11 and 26 m, reaching a maximum of 12.05 mg m3
at St. 212.
During PATEX III cruise (summer), surface [chl a] were
between 0.24 and 2.16 mg m3 (Fig. 1b). Fluorescence peak
depths were generally shallower than in the spring cruise,
varying between 5 and 27 m. The lowest values were
observed at the innermost station of the lower latitude
 ARTICLE IN PRESS
2166 A. Ferreira et al. / Deep-Sea Research I 56 (2009) 2162–2174

section (St. 301) and in the east of the front (St. 311), with of the cruises, discriminating surface and fluorescence
values of 0.24 and 0.56 mg m3, respectively. Along the peak depths.
continental slope, there was no evidence of considerable Fig. 2 shows a relationship between total and size-
latitudinal variation in [chl a]. The contrast between [chl fractionated [chl a] for the cruises. In spring (PATEX II,
a] in spring and summer periods was even more Fig. 2a), the range of total pigment concentration was
pronounced when the depth-integrated [chl a], which explained by changes in microplankton biomass, except at
ranged from 60 to 650 mg m2 in spring and from 12 to stations where low [chl a] values were observed (mainly
77 mg m2 in summer (not shown) were compared. The in the north). Excluding these stations (St. 201–206, see
SeaWiFS [chl a] images also illustrate the differences in Fig. 1), the mean dominance of microplankton was 69%. In
the phytoplankton biomass between the two sampling summer (PATEX III), a shift in algae size dominance from
periods (Fig. 1). Range of variation, mean and standard microplankton (spring, PATEX II) to nanoplankton biomass
deviation of [chl a] values are presented in Table 1 for both was observed (Fig. 2b). The nanoplankton fraction con-
tributed, on average, 66% to total [chl a], followed by
picoplankton (23%). The contribution of larger cells
Table 1
(420 mm) was only around 8%.
Ranges, means and standard deviations of chlorophyll-a concentrations
[chl a], and of absorption coefficients at 440 nm (m1), for colored
dissolved organic matter, acdom, total particulate matter, apart, detritus,
adet, phytoplankton pigments, aph and specific phytoplankton absorp- 3.2. Absorption coefficient magnitudes
tions coefficients for surface (above) and fluorescence peak depths
(below) spring (PATEX II cruise) (left) and summer (PATEX III cruise)
Table 1 shows separately the surface and fluorescence
(right) samples.
peak values of absorption coefficients (m1) at 440 nm
PATEX II PATEX III (wavelength at which all components are optically
important), of total particulate matter [apart(440)], phyto-
Range Mean Std. dev. Range Mean Std. dev.
plankton [aph(440)], detritus [adet(440)] and colored
Surface dissolved organic matter [acdom(440)] during in both
[Chl a] 0.920–11.85 5.578 3.167 0.240–2.160 1.234 0.455 cruises (PATEX II and PATEX III). In spring (PATEX II), all
acdom 0.014–0.126 0.044 0.029 0.020–0.345 0.121 0.088 apart(440) values varied from 0.074 to 0.375 m1 (Table 1,
apart 0.096–0.366 0.230 0.075 0.030–0.257 0.146 0.052 left), with lower values corresponding to the northern
adet 0.017–0.068 0.035 0.012 0.011–0.120 0.043 0.032
section, where low [chl a] were found (Table 1, left). In this
aph 0.067–0.351 0.191 0.071 0.002–0.182 0.102 0.043
a*ph 0.020–0.159 0.042 0.030 0.044–0.110 0.083 0.018 season, acdom(440) coefficients were relatively low and
varied within a narrow range (0.014–0.160 m1). There
Fluorescence peak depths
[Chl a] 0.920–12.05 5.708 3.690 0.570–2.530 1.314 0.461 was no significant difference between values at surface
acdom 0.017–0.160 0.046 0.032 0.015–0.144 0.057 0.031 and at the fluorescence peaks for any of the coefficients
apart 0.074–0.375 0.221 0.084 0.036–0.190 0.110 0.042 (Test-t, a ¼ 0.05).
adet 0.011–0.036 0.025 0.007 0.001–0.017 0.008 0.004 In summer (PATEX III), apart(440) were consistently
aph 0.074–0.339 0.190 0.075 0.035–0.172 0.099 0.038
a*ph 0.018–0.165 0.044 0.033 0.058–0.131 0.079 0.017
lower, with a mean value of 0.149 m1 at surface. During
this cruise, only the adet(440) values differed between

PATEX II
12 PATEX III

micro
10 nano 1.2
[chl a] per size class (mg m−3)

[chl a] per size class (mg m−3)

pico

8 1

0.8
6
0.6
4
0.4

2
0.2

0 0
0 2 4 6 8 10 12 14 0 0.5 1 1.5 2 2.5
total [chl a] (mg m−3) total [chl a] (mg m−3)

Fig. 2. Chlorophyll-a size-fractions (micro-, nano-e picoplankton) versus total [chl a] during PATEX II, spring (a) and PATEX III, summer (b) cruises.
Samples are from both surface and fluorescence peak depths. Note the different scales.
 ARTICLE IN PRESS
A. Ferreira et al. / Deep-Sea Research I 56 (2009) 2162–2174 2167

depths, with larger and variable values at the surface
(Table 1, right). acdom(440) magnitudes were also higher
and more variable at the surface (average of
0.1270.09 m1) in summer, although not statistically
different from the fluorescence peak (average of
0.04570.032 m1), because of the high standard deviation
at the surface.
As seen in Table 1, phytoplankton was the dominant
component of apart(l) in both PATEX II and PATEX III
cruises. In spring, phytoplankton contributed on average
83% and 88% to total particulate absorption at 440 nm
[(aph(440)/apart(440)] at surface and at the fluorescence
peak, respectively. In summer, the contributions were
about 71% and 93%, respectively. On average, phytoplank-
ton accounted for approximately 83% of apart(440) in both
periods. Despite an identical overall average contribution
of detritus in both periods (17%), the percent contribution
of adet(440)/apart(440) at the surface was generally slightly
higher and more variable in summer (mean of 30715%)
than in spring (mean of 1777%).
The absorption coefficients of particulate material
(phytoplankton+detritus) were statistically higher during
PATEX II (Test-t, a ¼ 0.05), clearly because of the higher
phytoplankton biomass, since highly variable values of
adet(440) showed no statistical difference between the
periods of study.
The respective contributions of phytoplankton, detri-
tus, and CDOM to total absorption at 440 nm (i.e.
apart(l)+acdom(l)) are displayed in a ternary plot (Fig. 3).
In spring (PATEX II, stars), the relative contribution of
phytoplankton was between 60% and 80% in most
samples, while the contribution ranges of detritus and
CDOM were similar (between 10% and 30%). A distinct
pattern was clear in summer (PATEX III, triangles), with
100
0
90
10
80
20
70
30
%
greater variation, notably due to differences in contribu-
tion of CDOM versus that of phytoplankton. In summer, a
group of points show a phytoplankton contribution
around 60% and 90%, similarly to the spring cruise, but
with small fractions of detritus (most points ranging
between 0% and 10%). The remaining points from PATEX III
showed a large dispersion throughout the range of
dissolved material contribution. In general, areas of higher
relative CDOM corresponded to low phytoplankton con-
tribution and vice versa. The phytoplankton fraction
between 10% and 60% was associated with scattered
contributions of CDOM and detritus. That indicates lack of
co-variation between non-algal components and phyto-
plankton, mainly in summer (PATEX III), and it suggests
that the contribution of non-algal components to light
absorption must be highly variable, depending on the
bloom stage in the region.
3.3. Spectral slope parameters for detritus and CDOM
In surface samples, the parameter that describes the
spectral shape of detritus (Sdet) varied within a narrow
range around 0.011 (70.0013) nm1 in PATEX II and 0.009
(70.0007) nm1 in PATEX III (Table 2). In sub-surface
peaks, a similar range was found for both PATEX II
(0.01170.003) nm1 and PATEX III (0.00970.003) nm1,
and therefore the parameters showed small variation
between depths and periods (Table 2).
Spectral slopes for CDOM (Scdom) at the surface showed
a statistically higher average in spring (0.0157
0.004 nm1) than in summer (0.01270.004 nm1) (Table
2). However, at sub-surface peaks, there was no difference
between periods (0.01470.004 nm1, PATEX II, 0.0137
0.004 nm1, PATEX III). Also, there was no significant
difference between surface and fluorescence peak values
between periods because of the high standard deviations
(Table 2). Overall, the average Scdom was 0.015 nm1
(70.004) during PATEX II and 0.012 nm1 (70.004)
during PATEX III. Clearly, values of Scdom showed greater
variation and higher magnitudes than those of detritus
(Sdet) (Table 2).
Fig. 4 illustrates the frequency distribution of the

exponential slopes, Scdm, of the CDM (CDOM+detritus)

OM

60
ph

40
yto
CD

50
p

50 Table 2
lan
%

Ranges, means and standard deviations of spectral shape (nm1) of

kto

40
60 absorption coefficients of CDOM, Scdom, detritus, Sdet and non-algal
n

components (cdom+detritus), Scdm, for surface (above) and fluorescence

30
70 peak depths (below) spring (PATEX II cruise) (left) and summer (PATEX
20 III cruise) (right) samples.
80
10 PATEX II PATEX III
90
0 Range Mean Std. dev. Range Mean Std. dev.
1
0 10 20 30 40 50 60 70 80 90 100 Surface
% detritus Scdom 0.007–0.026 0.015 0.004 0.005–0.029 0.012 0.004
Sdet 0.008–0.013 0.011 0.001 0.008–0.011 0.009 0.001
Fig. 3. Ternary plot illustrating the relative contribution of CDOM, Scdm 0.010–0.015 0.013 0.002 0.006–0.019 0.011 0.002
phytoplankton and detritus to light absorption for all samples (surface Fluorescence peak depths
and fluorescence peak depths), at 440 nm, for PATEX II (spring, stars) and Scdom 0.005–0.022 0.014 0.004 0.005–0.019 0.012 0.003
PATEX III (summer, open triangles). The relative contribution, in Sdet 0.007–0.018 0.011 0.003 0.005–0.019 0.010 0.003
percentage, of a component to total absorption can be read on the Scdm 0.003–0.127 0.013 0.002 0.005–0.017 0.012 0.003
corresponding axis.
 ARTICLE IN PRESS
2168 A. Ferreira et al. / Deep-Sea Research I 56 (2009) 2162–2174
16
14
12
10
8
N
6 10
4
2 5
0
0.005 0.01 0.015 0.02 0.025
Scdm (nm−1)
Fig. 4. Frequency distribution of the exponential slope, Scdm, of the CDM
(detritus+CDOM) for PATEX II (spring, grey) and PATEX III (summer, clear
bars) cruises.
absorption spectra, including both surface and fluores-
cence peaks, during both cruises. The range of values was
narrower during PATEX II, with a mean value of 0.013
(70.002) nm1. During PATEX III, it varied between 0.005
and 0.019 (mean ¼ 0.01270.003) m1.
3.4. Phytoplankton absorption coefficients and chlorophyll-a
concentration
Fig. 5 shows the frequency distribution of the chlor-
ophyll-a-specific absorption coefficient for phytoplankton,
a*ph(440) (m2 mg [chl a]1), for both cruises. Except for
two extreme spectra, most values ranged from 0.018 to
0.097 m2 mg [chl-a]1 (mean of 0.04370.03 m2 mg [chl
a]1) for PATEX II. The a*ph(440) values were statistically
higher during PATEX III (Test-t, a ¼ 0.05), varying between
0.04 and 0.13 m2 mg [chl a]1 (mean 0.0970.02 m2 mg
[chl a]1). The means and standard deviations are also
shown in Table 1, discriminating the surface and fluores-
cence peak depths, which did not show statistical
difference during either cruise.
The relations between [chl a] and aph(440) were
defined by a power function, and a least-square fitting
provided the following relationship for PATEX II:
aph(440) ¼ 0.09 ([chl a])0.420 (r2 ¼ 0.52; n ¼ 47) and for
PATEX III: aph(440) ¼ 0.083 ([chl a])0.938 (r2 ¼ 0.77;
n ¼ 55). The slopes (i.e. exponents) are higher than those
established by Bricaud et al. (1995) for a large data set
with [chl a] varying between 0.02 and 25 mg m3, with
greater difference in PATEX III (not shown). In that cruise,
the slope was still higher than the one found by Bricaud
et al. (2004), which established a new relationship based
on a different data set from their previous work, in which
a shift of aph(440) toward higher values was found,
compared to the first Bricaud et al. study.
The scatter of the points in the relationship between
[chl a] and aph(676) is lower, and the difference between
the cruises is consistently less pronounced (not shown),
although statistically significant (Test-t, a ¼ 0.05). The
25
20
15
N
0
0 0.02 0.04 0.06 0.08 0.1 0.12 0.14 0.16 0.18
aph* (440) m2 mg [chl a]−1
Fig. 5. Frequency distribution of the specific phytoplankton absorption
coefficients at 440 nm (i.e. aph(440)/[chl a]) for PATEX II (spring, grey)
and PATEX III (summer, clear bars) cruises.
a*ph(676) values for all depths were constrained between
0.009 to 0.055 m2 mg [chl a]1 (mean of 0.0197
0.010 m2 mg [chl a]1) in PATEX II and 0.016 to
0.051 m2 mg [chl a]1 (mean of 0.03170.006 m2 mg [chl
a]1) in PATEX III. The equation provided by least-squares
fitting was aph(676) ¼ 0.03 ([chl a])0.651 (r2 ¼ 0.73;
n ¼ 47) for PATEX II (spring) and aph(676) ¼ 0.03 ([chl
a])1.064 (r2 ¼ 0.84; n ¼ 55) for PATEX III (summer). PATEX
II values agree with the Bricaud et al. (1995) data set,
while PATEX III still departs slightly from that trend.
3.5. Phytoplankton absorption spectra
The average phytoplankton spectra and their standard
deviations for both periods are illustrated in Fig. 6. In
spring (PATEX II), the vast majority of spectra (43 of 47)
showed high absorption in the ultraviolet (UV) region,
with peaks centered at 330 nm (not shown). Fig. 6b
presents spectra for the PATEX III cruise, whose absolute
values contrasted with the spring data (note the different
scales and see Table 1 for values in the blue absorption
peak). In that period, relatively high values in the UV
range were also observed in various spectra (not shown),
but at a much lower magnitude than during PATEX II.
Because of the high variability verified in the relation-
ship between phytoplankton absorption values and [chl a]
(as seen throughout the a*ph(l) values), aph(l) values were
normalized by the spectral mean of each aph(l) spectrum,
a/phS, to examine differences in shape between them.
Figs. 7a and b show the mean-normalized absorption
spectra for whole samples and their mean and standard
deviation for PATEX II and PATEX III, respectively. This
normalization removes the effect of concentration and
allows computation of variability due solely to spectral
shape (Roesler et al., 1989).
As expected for normalized spectra influenced
by pigment packaging, spectra for samples dominated by
microplankton at most (see Section 3.1.) were relatively
 ARTICLE IN PRESS
A. Ferreira et al. / Deep-Sea Research I 56 (2009) 2162–2174 2169

Fig. 6. Mean spectrum and standard deviation (SD) obtained from phytoplankton absorption spectra for all samples from (a) PATEX II (spring) and (b)
PATEX III (summer). Note the different Y-axis scales.

2.5
2.5

2
2
aph(λ)/a<ph>

aph(λ)/a<ph>

1.5 1.5

1 1

0.5 0.5

0 0
400 450 500 550 600 650 700 400 450 500 550 600 650 700
Wavelength (nm) Wavelength (nm)

Fig. 7. Shape of phytoplankton absorption spectra for all samples (dashed-dotted lines) from (a) PATEX II (spring, n ¼ 43) and (b) PATEX III (summer,
n ¼ 55). Spectra were normalized to the average phytoplankton absorption between 400 and 700 nm (a/phS). The mean normalized spectra (bold line)
and standard deviation (values multiplied by 3 for visualization) (dashed line) are also shown.

flat (PATEX II, Fig. 7a), and spectra referring to smaller-
size-class dominance showed higher normalized values
and higher blue-to-red ratio (PATEX III, Fig. 7b). Therefore,
the discrepancies between the two cruises are more
pronounced in the blue absorption range, where the
packaging effect is maximal (Figs. 7a and b).
In PATEX II, a similar standard deviation shape for
a/phS(l) as the mean spectrum (Fig. 7a) indicates that the
majority of the spectra have a similar shape and that
variations of pigment packaging dominated the variance
(Magnuson et al., 2004). However, a peak in this spectrum
is found at 457 nm, and a pronounced peak appears
between 500 and 550 nm (Fig. 7a), suggesting that
variation in pigment composition can play a role in overall
shape of this data set.
The mean-normalized absorption spectra from PATEX
III (Fig. 7b) indicate there is less variability in the spectral
shape than those from PATEX II (Fig. 7a). But distinctions
are found between 440 and 490 nm, higher between 505
and 560 nm, and still important variability between 570
and 600 nm (see standard deviation in Fig. 7b). In summer,
peaks centered at 412 nm were not observed, as seen in
PATEX II. Standard deviation spectra for a/phS(l) from
PATEX III did not resemble the shape of the mean spectra
(Fig. 7b). Since absorption around 460 nm is attributed
to accessory pigments, the peak at 463 nm in the SD
 ARTICLE IN PRESS
2170 A. Ferreira et al. / Deep-Sea Research I 56 (2009) 2162–2174
0.9
0.8
0.7
0.6
Sf
0.5
0.4
0.3
0.2
0.1
0 2 4 6 8 10
[chl a] (mg m−3)
Fig. 8. Scatter plot of size spectral parameter (Sf) (non-dimensional) as a
function of [chl a] for PATEX II (spring, stars) and PATEX III (summer,
open triangles).
spectra suggests that pigment packaging (which is
strongest in the chlorophyll-a absorption peak) was less
important and that most of the variability in the blue band
of this dataset could be due to variations in pigment
composition.
3.6. Size spectral parameter
Fig. 8 shows a plot of the size parameter for both
cruises against [chl a], where an inverse relationship can
be seen. The spectral size parameter (Sf) varied between
0.21 and 0.79 in spring (PATEX II, mean of 0.4170.17),
with higher values associated with small cell-size abun-
dance in the northern section. Excluding this transect,
the mean value was about 0.35, corresponding to higher
[chl a] values and to the dominance of the microplankton
fraction during PATEX II. In summer (PATEX III), the
parameter values were between 0.48 and 0.87 (mean of
0.7270.08), in association with lower [chl a] values and a
small cell-size predominance during the period.
4. Discussion
The seasonal cycle of phytoplankton biomass has been
observed by recent analyses of ocean color data in the
vicinity of the Patagonian shelf-break front (e.g. Garcia
et al., 2004; Rivas et al., 2006; Romero et al., 2006;
Signorini et al., 2006). As presented by Romero et al.
(2006), strong [chl a] seasonal variability (44 mg m3) is
observed in comparison with the open ocean (o1.5 mg
m3) over the Patagonia shelf and shelf-break. Northward,
high [chl a] begin in early austral spring, while southward
blooms begin in late spring to early summer. The spring
maximum (43.5 mg m3) extends from the mid-shelf to
the shelf-break, persisting at the last in summer (Romero
et al., 2006). According to Rivas et al. (2006), peaks in
[chl a] over the shelf-break appear in October–November.
Therefore, the timings of our cruises (beginning of
November and March) likely represent higher phytoplank-
ton biomass (PATEX II) and moderated values throughout
the summer (PATEX III) in the region.
A remarkable phytoplankton bloom was observed
along the Patagonian shelf-break in spring 2006, with
mean [chl a] of 5.6 mg m3 (PATEX II), similar to that
verified in 2004 by Garcia et al. (2008). The authors
pointed out that the main factors favoring the develop-
ment and maintenance of these blooms are both nutrient
supply from the Malvinas Current and water-column
stability along the shelf-break front. In the present study,
phytoplankton biomass was much lower in summer than
in spring. These lower [chl a] (up to 2 mg m3) in summer
were associated with low (Pollery, R., personal commu-
nication), probably regenerated nutrients (decomposition
12 14 by bacteria and grazing), since upwelling does not seem to
occur in this period but is more frequent in spring
(Matano and Palma, 2008).
Light absorption by water constituents in the Argen-
tinean shelf-break region was generally dominated by
phytoplankton during both periods (Fig. 3). There is a
considerable distance of the study area from waters of
continental origin, so phytoplankton is expected to be the
most dominant particle in seawater. Both CDOM and
detritus absorption showed no relationship with [chl a].
Besides, CDOM absorption was relatively low and constant
in spring, while it was higher and more variable in
summer (Table 1 and Fig. 3). That suggests there is a time
lag between the phytoplankton bloom and CDOM in-
crease, consistent with other findings (Vodacek et al.,
1997; Siegel and Michaels, 1996; Bricaud et al., 1981).
Nelson et al. (1998) have shown that CDOM in the
Sargasso Sea follows a strong seasonal pattern that is
not correlated with [chl a], and have suggested that such a
breakdown product of phytoplankton, resulting from
microbial activity, is destroyed by photo-oxidation. The
relatively constant CDOM concentration in spring against
a great range of [chl a] (not shown) suggests that CDOM is
not produced by phytoplankton directly during the bloom.
Moreover, high maximum irradiance levels verified in this
season (1200 mE m–2 s–1, compared to E900 mE m–2 s–1 in
late summer, measured in our two cruises) must have an
important role in CDOM depletion in the upper layers.
Patagonia is an area influenced by the Antarctic polar
vortex and, thus, affected by air masses with low ozone
concentrations during October–November, in events of
variable duration (Orce and Helbling, 1997; Pérez et al.,
1998). High levels of ultraviolet radiation and photo-
synthetically active radiation (PAR) have already been
observed in water bodies in the region, particularly in
spring (Pérez et al., 1998), which must increase the photo-
oxidation. In fact, Bricaud et al. (1981) have demonstrated
that intense radiation in this spectral band reduced CDOM
absorption at 300 nm, about 75%, by indicating that a
fraction of the material is potentially sensible to photo-
degradation.
Other studies have shown that a detritus absorption
peak occurs when a phytoplankton bloom starts to decline
(e.g. Sasaki et al., 2005). This probably happened in the
period between our spring and summer cruises, although
an important variability in surface absorption values by
 ARTICLE IN PRESS
A. Ferreira et al. / Deep-Sea Research I 56 (2009) 2162–2174
detritus could still be observed in summer (Table 1). In
addition, detritus is generated via planktonic food-web
interactions in grazing by zooplankton (as egested fecal
material), or as the result of decomposition of senescent
cells (Nelson and Robertson, 1993). The importance of
zooplankton biomass in the region is relatively well
known (Vasilyev, 2007; Sabatini et al., 2004), and it is
suggested that the herbivorous zooplankton must play an
important role in controlling biomass accumulations in
this ecosystem (Garcia et al., 2008).
4.1. Spectral characteristics of absorption coefficients by
non-algal components
Information on seasonal variations in the absorption
coefficients of particulate and dissolved matter are
required as inputs to optical models that seek to convert
satellite observations of water color into water-quality
information (Carder et al., 1999; Ciotti et al., 1999; Devred
et al., 2005). Moreover, the spectral characteristics can
offer important information about the CDOM and detritus
composition and source (Twardowski et al., 2004; Babin
et al., 2003).
As verified in several studies, our Sdet values were
rather small and less variable than Scdom. In spring,
slightly higher Sdet values could be partly attributed to
silica frustules of the abundant diatoms (see taxonomic
information below). Nearly identical values of Sdet
(0.011 nm1) were found for mineral particles by Bowers
et al. (1996) and, according to Bukata et al. (1995), the
parameter tends to be higher in elevated proportions of
mineral to organic matter. Furthermore, in both periods, a
large variation in the detritus spectral parameter was
noted in fluorescence peak depths, probably reflecting a
greater heterogeneity of detritus at depth. These fluores-
cence peaks were generally close to the bottom of the
mixing layer (not shown), suggesting that an important
component of the particle assembly at those depths could
be material that sedimented from surface layers, being
important areas of remineralization.
Based on comparison with ranges reported previously,
the variability noted in our Scdom values was small and
close to what was found in other regions, even when both
cruises were considered. Twardowski et al. (2004) have
revised Scdom values reported in the literature and
suggested that the relatively broad range in Scdom values
may be partly due to the fact that various methods were
used to measure acdom(l) and to derive Scdom (such as
optical setup, spectral range, baseline correction, fitting
procedure). Although the same procedures for whole
samples were mantained, statistical differences were
found between our sampling periods, with higher Scdom
values in summer. That is in agreement with Carder et al.
(1989) and others, who observed inverse relationships
between acdom(l) and Scdom. As proposed by Twardowski
and Donaghay (2002), higher slopes found in oceanic
waters, in contrast to coastal waters, must reflect a
cumulative effect of photo-bleaching. In addition, the
spectral parameter tends to be higher at the surface in
relation to deeper waters, as a result of chemical
2171
alterations in the material. Therefore, high acdom(l) values
associated to lower Scdom could be evidence of photo-
oxidation (Vodacek et al., 1997) acting as an important
regulator of CDOM concentrations in the region, as
pointed out previously.
Semi-analytic models usually do not distinguish the
particulate and dissolved non-algal compartments, and
therefore utilize a mean parameter that describes the
shape of CDOM plus detritus spectra (e.g. Siegel et al.,
2005; Maritorena et al. 2002). In the present work, there
was no statistical difference in Scdm values between
periods. That is an important result, as it indicates that
this parameter does not present a considerable seasonal
variation, at least in our study periods.
4.2. Phytoplankton absorption variability
Specific absorption coefficients by phytoplankton are
known to vary regionally and seasonally in response to
changes in species composition, light and nutrient condi-
tions (e.g. Stuart et al., 2000; Sathyendranath et al., 1999;
Lutz et al., 1996). The non-linear relationship between
aph(l) and [chl a] is referred to as the ‘pigment package
effect’ (Duyens, 1956; Morel and Bricaud, 1981; Cleveland,
1995), which is caused by variations in cell size and
internal pigment concentration and/or overlapped pig-
ments. At 440 nm, carotenoids, chlorophylls b and c and
phicobilins can contribute to the absorption, resulting in
an increase in the specific absorption coefficient, in
relation to a given [chl a], with important effects on the
parameterization based only on the concentration of the
main pigment (Lutz et al., 1996). At 676 nm, the absorp-
tion is dominated by chlorophyll-a. Differences between
our values of aph(440) and those predicted by the Bricaud
et al. (1995, 2004) regression can be partly explained by
possible differences in pigment compositions and cell size
in our study region, relative to what is typically found in
other Case 1 waters with similar [chl a]. In the Patagonian
shelf-break region, the spring bloom consisted of large
diatoms dominated mainly by chains of Thalassiosira sp.
(6–30 mm), large Corethron pennatum (long axis 187 mm)
and Thallasiothrix antarctica (500 mm apical axis), large
dinoflagellates (Gymnodiniales) (23–95 mm) and an im-
portant contribution of colonial Phaeocystis cf. antarctica
(Garcia et al., 2008; R. Comin, personal communication). It
is well known that large phytoplankton can exhibit low
relative aph*(l) values. Therefore, a peculiar pigment
composition may be responsible for the higher relative
values found in the blue portion of the spectra in spring,
because the measured aph(676) when plotted as a function
of [chl a] are consistent with the prediction from
Bricaud et al. (1995), and thus the size structure should
be quite similar (see Bricaud et al., 2004). Note that the
‘package effect’ is determined not only by cell size, but
also by photoacclimation processes which modify the
intracellular pigment concentrations. High levels of both
UV radiation and PAR observed in water bodies of the
Patagonia region, particularly in spring (Pérez et al., 1998),
associated with high absorption in the UV shown here,
suggest the presence of mycosporine-like aminoacids
 ARTICLE IN PRESS
2172 A. Ferreira et al. / Deep-Sea Research I 56 (2009) 2162–2174
(MAAs) in the phytoplankton populations of the region
and consequently photoacclimation processes. That
may result in high relative levels of photo-protector
carotenoids, and it could partially explain the relatively
high specific absorption coefficients in the spring (PATEX
II). The fact that summer spectra (PATEX III) showed
relatively high absorption in the UV range (not shown)
also suggests the presence of the MAAs. However,
accumulations of these compounds are generally attrib-
uted to larger-size groups, probably because of greater
storage capacity (Hannach and Sigleo, 1998; Helbling
et al., 1996), which can explain the high absorption in the
UV range in PATEX II spectra. Nonetheless, this has
also been observed in cyanobacteria (Garcia-Pichel and
Castenholz, 1993), indicating the need for further research
to confirm this fact in our study.
In summer (PATEX III), small phytoplankton cells
should result in a weaker ‘package effect’, revealed by
much higher aph(l), if compared to spring values. Also, the
fact that summer values are still slightly above those
found by Bricaud et al. (1995) at 676 nm suggests that the
phytoplankton cell-size structure found in this study may
be the main factor responsible for departures found in the
general spectra. In fact, dominance of small nanoflagel-
lates (o10 mm), Phaeocystis cf. antarctica and high con-
centrations of Synechococcus sp. were verified by both
microscopy and flow cytometry (R. Comin and M. Tenório,
personal communication) in the summer samples. Since
departures from the regression obtained by Bricaud et al.
(1995) are higher at 440 nm, both differences in pigment
composition and size structure seem to contribute to the
differences between summer data (PATEX III) and the
other studied regions. This is in agreement with increas-
ing trends in non-photosynthetic pigments-to-chl a ratio
in small cells (Bricaud et al., 1995).
Another important point to note is that differences
between results from this study and those of Bricaud et al.
could also be attributed to the choice of the Beta
correction algorithm. By recalculating our particulate
absorption values, using the Beta algorithm from their
studies (Bricaud and Stramski, 1990), it could be seen that
the values provided by our factor are approximately
25–30% higher, at most, in the blue spectral band. This
is comparable to the differences between distinctive
algorithms presented in the literature, and it would not
explain the departures by the present data set itself. An
important source of discrepancies is the fact that Bricaud
et al. used HPLC-derived pigment concentrations, which
can differ markedly from Turner-fluorescence-derived [chl
a], especially in diatom-dominated waters. Indeed, HPLC
[chl a] measurements were conducted in a later PATEX
cruise, and they revealed lower values by a factor
of approximately 1.5 compared to fluorescence derived
[chl a] (Mendes, C.R., personal communication). In this
case, our phytoplankton absorption values would be even
higher for a given [chl a], resulting in higher a*ph(l) values
compared to HPLC-derived [chl a].
The analyses of spectra normalized by the average
phytoplankton absorption between 400 and 700 nm (Fig. 7)
revealed the same trends described above throughout
specific wavelength absorption values. Despite the dom-
inance of microplankton during the spring cruise, PATEX II
(see Section 3.1), some spectra were dominated by small
cells. Because of that, spectra were also grouped according
to the dominant [chl a] size fractions in that sample,
throughout the criteria that assumed the dominant size to
be the one containing more than 50% of the total [chl a]
(see Fig. 2). This separation showed lower blue-to-red
peak ratios in samples dominated by microplankton and
higher ratios in the few spectra dominated by small cells
(not shown). That indicates size was a major factor
controlling the shape of phytoplankton absorption spectra
in such a data set, with a great magnitude of the ‘package
effect’ in the majority of samples dominated by bigger
cells (see Section 3.1), consistently with the increase in
pigment packaging (Fig. 7a). This is in agreement with the
dominance by big cells in the period, as already men-
tioned. Carotenoid effects are also suggested by the fact
that the standard deviation peak does not correspond to
the chlorophyll-a absorption peak in the blue region
(440 nm). In addition, a distinct peak around 412 nm,
where phaeopigment absorption is maximal (Roesler
et al., 1989), suggests that the bloom could be in an
intermediate to late stage of development, which is an
important observation, in the sense that this blue portion
is usually attributed to CDOM absorption.
The whole spectra in PATEX III corresponded to
dominance by nanoplankton (considering the same
criteria based on percentage of size-fractionated [chl a]).
Thus, the variability within the period could be attributed
to variations in pigment composition. The weak ‘package
effect’ comparing to PATEX II spectra suggests that it
might be appropriate to replace the global parameteriza-
tions by locally and seasonally adapted relationships in
order to apply bio-optical models for the region.
The great difference between size parameter values
from both periods derived from phytoplankton absorption
spectra emphasize the strong ‘package effect’ of phyto-
plankton cells in spring (PATEX II), in contrast to the weak
effect in the summer organisms (PATEX III), as seen
throughout the absorption spectra (Fig. 7). Moreover, the
agreement between these values and information about
the taxonomic composition obtained by traditional meth-
ods (microscopy, citometry) indicates the potential for
assessing changes in cell size from optical measurements.
Ciotti and Bricaud (2006) have shown a good agreement
between in situ Sf values and those obtained from
phytoplankton absorption retrieved from water-leaving
radiances at SeaWiFS channels. This inversion method can
be an important tool to characterize the remarkable
seasonality in the phytoplankton community size struc-
ture in the Patagonian shelf-break region from remote
sensing. This information can contribute to understanding
the biogeochemical processes that are associated with
the different community structures. As presented by
Garcia et al. (2008), the primary production rates and
diatom dominance verified in spring indicate a potentially
significant biological control of gases such as O2 and
CO2 in surface layers, as opposed times when small
flagellates are the prevailing phytoplankton group
(Schloss et al., 2007), which was the case in our summer
cruise.
 ARTICLE IN PRESS
A. Ferreira et al. / Deep-Sea Research I 56 (2009) 2162–2174
The present work quantified the magnitudes and
spectral characteristics of light absorption by total
particulate (phytoplankton and detritus) and CDOM in
two distinct bloom stages along the Patagonia shelf-break
front. Regarding the absorption properties, the phyto-
plankton was the dominant component during both
cruises. Although CDOM seems to be an important
component of the absorption budget, it is probably a
phytoplankton by-product. It must be addressed that
CDOM could contribute to absorption in a higher degree
verified here, depending on other bloom stages. To a
certain extent, the region could be considered as a ‘Case I’
water. However, it seems reasonable not to consider it as a
final conclusion, given that only two periods were
sampled. Important information was generated from two
distinct periods of the phytoplankton seasonal cycle,
which in association with other determinant optical
properties, such as scattering, attenuation and reflectance
(Garcia et al., in preparation) shall contribute to optical
parameterizations for using remote-sensing techniques.
Acknowledgements
The PATagonian EXperiment (PATEX) is a multidisci-
plinary project as part of Group of high latitude oceano-
graphy (GOAL) activities in the Brazilian Antarctic
Program. We thank the crew of the Brazilian Navy
research ship ‘‘Ary Rongel’’ for their assistance during the
field sampling. The project was sponsored by CNPq
(Brazilian National Council on Research and Development)
through Grant 557305/2005-5. Ocean color images were
provided by GSFC/NASA. A. Ferreira was granted a
graduate scholarship from the Brazilian agency Coorde-
nac- ão de Aperfeic- oamento de Pessoal de Ensino Superior
(CAPES). Thanks are due to Áurea Maria Ciotti for
providing MATLABs scripts used to calculate the ‘size
parameter’, Sf. We also thank three anonymous reviewers
for their reviews and helpful comments.
References
Acha, M.E., Mianzan, H.W., Guerrero, R.A., Favero, M., Bava, J., 2004.
Marine fronts at the continental shelves of South America. Physical
and ecological processes. Journal of Marine Systems 44 (1/2),
83–105.
Babin, M., Stramski, D., 2002. Light absorption by aquatic particles in the
near-infrared spectral region. Limnology and Oceanography 47,
911–915.
Babin, M., Stramski, D., Ferrari, G.M., Claustre, H., Bricaud, A., Obolensky,
G., Hopepffner, N., 2003. Variations in the light absorption coeffi-
cients of phytoplankton, nonalgal particles, and dissolved organic
matter in coastal waters around Europe. Journal of Geophysical
Research 108 (C7), 3211.
Behrenfeld, M.J., Randerson, J.T., McClain, C.R., Feldman, G.C., Los, S.O.,
Tucker, C.J., Falkowski, P.G., Field, C.R., Frouin, C.B., Esaias, W.E.,
Kolber, D.D., Pollack, N.H., 2001. Biospheric primary production
during an ENSO transition. Science 291, 2594–2597.
Bianchi, A., Bianucci, L., Piola, AR., Pino D.R., Schloss, I., Poisson A.,
Balestrini, C.F., 2005. Vertical stratification and air-sea CO2 fluxes in
the Patagonian shelf. Journal of Geophysical Research 110, C07003,
doi:10.1029/2004JC002488.
Bianchi, A.A., Pino, D.R., Perlender, H.G.I., Osiroff, A.P., Segura, V., Lutz, V.,
Clara, M.L., Balestrini, C.F., Piola, A.R., 2009. Annual balance and
seasonal variability of sea-air CO2 fluxes in the Patagonia Sea: Their
2173
relationship with fronts and chlorophyll distribution. Journal of
Geophysical Research 114, C03018, doi:10.1029/2008JC004854.
Bowers, D.G., Harker, G.E.L., Stephan, B., 1996. Absorption spectra of
inorganic particles in the Irish Sea and its relation to tidal stirring.
International Journal of Remote Sensing 19, 2789–2805.
Bricaud, A., Morel, A., Prieur, L., 1981. Absorption by dissolved organic
matter of the sea (yellow substance) in the UV and visible domains.
Limnology and Oceanography 26, 43–53.
Bricaud, A., Stramski, D., 1990. Spectral absorption coefficients of living
phytoplankton and nonalgal biogenous matter: a comparison
between the Peru upwelling area and the Sargasso Sea. Limnology
and Oceanography 35 (3), 562–582.
Bricaud, A., Babin, M., Morel, A., Claustre, E.H., 1995. Variability in the
chlorophyll-specific absorption coefficients of natural phytoplank-
ton: analysis and parameterization. Journal of Geophysical Research
100 (C7), 13,321–13,332.
Bricaud, A., Morel, A., Babin, M., Allali, K., Claustre, H., 1998. Variations of
light absorption by suspended particles with chlorophyll a concen-
tration in oceanic (case 1) waters: analysis and implications for
bio-optical models. Journal of Geophysical Research 103 (C13),
31033–31044.
Bricaud, A., Claustre, H., Ras, J., Oubelkheir, K., 2004. Natural variability of
phytoplanktonic absorption in oceanic waters: influence of the size
structure of algal populations. Journal of Geophysical Research 109
(C11010).
Bukata, R.P., Jerome, J.H., Kondratyev, K.Y., Pozdnyakov, D.V., 1995. Optical
Properties and Remote Sensing of Inland and Coastal Waters. CRC
Press, Boca Raton, FL.
Carder, K.L., Steward, R.G., Harvey, G.R., Ortner, P.B., 1989. Marine humic
and fulvic acids: their effects on remote sensing of ocean chlorophyll.
Limnology and Oceanography 36, 68–81.
Carder, K.L., Chen, F.R., Lee, Z.P., Hawes, S.K., Kamykowski, D., 1999.
Semianalytic moderate-resolution imaging spectrometer algorithms
for chlorophyll-a and absorption with bio-optical domains based on
nitrate depletion temperatures. Journal of Geophysical Research 104,
5403–5421.
Carder, K.L., Chen, F.R., Cannizzaro, J.P., Campbell, J.W., Mitchell, B.G.,
2004. Performance of the MODIS semi-analytical ocean color
algorithm for chlorophyll-a. Advances in Space Research 33,
1152–1159.
Ciotti, A.M., Cullen, J.J., Lewis, M.R., 1999. A semi-analytical model of the
influence of phytoplankton community structure on the relationship
between light attenuation and ocean color. Journal of Geophysical
Research 104 (C1), 1559–1578.
Ciotti, A.M., Lewis, M.R., Cullen, J.J., 2002. Assessment of the relationships
between dominant cell size in natural phytoplankton communities
and the spectral shape of the absorption coefficient. Limnology and
Oceanography 47, 404–417.
Ciotti, A.M., Bricaud, A., 2006. Retrievals of a size parameter for
phytoplankton and spectral light absorption by colored detrital
matter from water-leaving radiances at SeaWiFS channels in a
continental shelf region off Brazil. Limnology and Oceanography-
Methods 4, 237–253.
Cleveland, J.S., 1995. Regional models for phytoplankton absorption as a
function of chlorophyll a concentration. Journal of Geophysical
Research 100 (7), 13333–13344.
Devred, E., Fuentes-Yaco, C., Sathyendranath, S., Caverhill, C., Platt, T.,
Mass, H., 2005. A semi-analytical, seasonal algorithm to retrieve the
chlorophyll-a concentration in the Northwest Atlantic from SeaWiFS
data. Indian Journal of Marine Sciences 34 (4), 356–367.
Duyens, L.N.M., 1956. The flattening of the absorption spectra
of suspensions as compared to that of solutions. Biochimica et
Biophysica Acta 19, 1–12.
Garcia-Pichel, F., Castenholz, R.W., 1993. Occurrence of UV-absorbing,
mycosporine-like compounds among cyanobacterial isolates and an
estimate of their screening capacity. Applied and Environmental
Microbiology 59, 163–169.
Garcia, C.A.E., Sarma, Y.V.B., Mata, M.M., Garcia, V.M.T., 2004. Chlorophyll
variability and eddies in the Brazil-Malvinas Confluence region.
Deep-Sea Research Part II-Topical Studies in Oceanography 51,
159–172.
Garcia, V.M.T., Garcia, C.A.E., Mata, M.M., Pollery, R.C., Piola, A.R.,
Signorini, S., Mcclain, C.R., 2008. Environmental factors controlling
the phytoplankton blooms at the Patagonia shelf-break in spring.
Deep-Sea Research Part I-Oceanographic Research Papers 55,
1150–1166.
Garver, S., Siegel, D.A., Mitchell, B.G., 1994. A statistical analysis of
particulate absorption spectra: What can a satellite ocean color
imager see? Limnology and Oceanography 39, 1349–1367.
 ARTICLE IN PRESS
2174 A. Ferreira et al. / Deep-Sea Research I 56 (2009) 2162–2174
Gordon, H.R., Brown, O.B., Evans, R.H., Brown, J.W., Smith, R.C., Baker, K.S.,
Clark, D.K., 1988. A semianalytic radiance model of ocean color.
Journal of Geophysical. Research 93, 10,909–10,924.
Hannach, G., Sigleo, A.C., 1998. Photoinduction of UV-absorbing com-
pounds in six species of marine phytoplankton. Marine Ecology
Progress Series 174, 207–222.
Helbling, E.W., Chalker, B.E., Dunlap, W.C., Holm-Hansen, O., Villafañe,
V.E., 1996. Photoacclimation of antarctic marine diatoms to solar
ultraviolet radiation. Journal of Experimental Marine Biology and
Ecology 204, 85–101.
Hu, C., Lee, Z., Muller-Karger, F.E., Carder, K.L., Walsh, J.J., 2006. Ocean color
reveals phase shift between marine plants and yellow substance. IEEE
Geoscience and Remote Sensing Letters 3 (2), 262–266.
IOCCG, 2006. Remote sensing of inherent optical properties: funda-
mentals, tests of algorithms, and applications. In: Lee, Z.P. (Ed.),
Reports of the International Ocean-Colour Coordinating Group, No. 5.
IOCCG, Dartmouth.
Kishino, M., Takahashi, N., Okami, N., Ichimura, S., 1985. Estimation of the
spectral absorption coefficients of phytoplankton in the sea. Bulletin
of Marine Science 37, 634–642.
Lutz, V.A., Sathyendranath, S., Head, J.H., 1996. Absorption coefficient of
phytoplankton: regional variations in the North Atlantic. Marine
Ecology Progress Series 135, 197–213.
Magnuson, A., Harding, L.W., Mallonee, M.E., Adolf, J.E., 2004. Bio-optical
model for Chesapeake Bay and the middle Atlantic bight. Estuarine,
Coastal and Shelf Science 61, 403–424.
Maritorena, S., Siegel, D.A., Peterson, A.R., 2002. Optimization of a
semianalytical ocean color model for global-scale applications.
Applied Optics 41, 2705–2714.
Marra, J., Trees, C.C., O’Reilly, J.E., 2007. Phytoplankton pigment absorption:
a strong predictor of primary productivity in the surface ocean. Deep-
Sea Research Part I-Oceanographic Research Papers 54, 155–163.
Matano, R.P., Palma, E.D., 2008. On the upwelling of downwelling currents.
Journal of Physical Oceanography, doi:10.1175/2008JPO3783.1.
Mitchell, B.G., 1990. Algorithms for determining the absorption coeffi-
cient of aquatic particulates using the quantitative filter technique
(QFT). Ocean Optics X 1302, 137–148.
Mitchell, B.G., Bricaud, A., Carder, K., Cleveland, J., Ferrari, G.M., Gould, R.,
Kahru, M., Kishino, M., Maske, H., Moisan, T., Moore, L., Nelson, N.,
Phinney, D., Reynolds, R.A., Sosik, H., Stramski, D., Tassan, S., Trees, C.,
Weidemann, A., Wieland, J.D., Vodacek, A., 2000. Determination of
spectral absorption coefficients of particles, dissolved material and
phytoplankton for discrete water samples. In: Fargion, G.S., Mueller,
J.L., McClain, C.R. (Eds.), Ocean Optics Protocols for Satellite Ocean
Color Sensor Validation, Revision 2. NASA, Goddard space flight
center, Greenbelt, Maryland, pp. 125–153.
Moore, L.R., Goericke, R., Chisholm, S.W., 1995. Comparative physiology
of Synechococcus and Prochlorococcus: influence of light and
temperature on growth, pigments, fluorescence and absorptive
properties. Marine Ecology Progress Series 116, 259–275.
Morel, A., Bricaud, A., 1981. Theoretical results concerning light
absorption in a discrete medium, and application to specific
absorption of phytoplankton. Deep-Sea Research 28, 1375–1393.
Morel, A., 1988. Optical modeling of the upper ocean in relation to its
biogenous matter content (Case I waters). Journal of Geophysical
Research 93 (C9), 10,749–10,768.
Morel, A., Maritorena, S., 2001. Bio-optical properties of oceanic waters: a
reappraisal. Journal of Geophysical Research 106, 7163–7180.
Morel, A., Gentili, B., Chami, M., Ras, J., 2006. Bio-optical properties of
high chlorophyll Case 1 waters, and of yellow-substance dominated
Case 2 waters. Deep-Sea Research Part I-Oceanographic Research
Papers 53, 1439–1559.
Nelson, N.B., Siegel, D.A., Michaels, A.F., 1998. Seasonal dynamics of
colored dissolved organic material in the Sargasso Sea. Deep-Sea
Research Part I-Oceanographic Research Papers 45, 931–957.
Nelson, J.R., Robertson, C.Y., 1993. Detrital spectral absorption: Labora-
tory studies of visible light effects on phytodetritus absorption,
bacterial spectral signal, and comparison to field measurements.
Journal of Marine. Research 51, 181–207.
O’Reilly, J., Maritorena, S., Mitchell, B.G., Siegel, D., Carder, K.L., Garver, S.,
Kahru, M., Mcclain, C., 1998. Ocean color chlorophyll algorithms for
SeaWiFS. Journal of Geophysical Research 103, 24937–24953.
Orce, V.L., Helbling, E.W., 1997. Latitudinal uvr-par measurements in
Argentina: extent of the ‘‘ozone hole’’. Global and Planetary Change
15, 113–121.
Pérez, A., Aguirre de Carcer, I., Tocho, J.O., Crino, E., Ranea, Sandoval, H.F.,
Berni, M.E., 1998. The extent of the ozone hole over South America
during the spring of 1993, 1994 and 1995. Journal of Physics
D-Applied Physics 31, 812–819.
Prieto, L., Vaillancourt, R.D., Hales, B., Marra, J., 2008. On the relationship
between carbon fixation efficiency and bio-optical characteristics of
phytoplankton. Journal of Plankton Research 30 (1), 43–56.
Reynolds, R.A., Sstramski, D., Mitchell, E.B.G., 2001. A chlorophyll-dependent
semianalytical reflectance model derived from field measurement of
absorption and backscattering coefficients within the Southern Ocean.
Journal of Geophysical Research 106 (C4), 7125–7138.
Rivas, A.L., Dogliotti, A.I., Gagliardini, D.A., 2006. Seasonal variability in
satellite-measured surface chlorophyll in the Patagonian Shelf.
Continental Shelf Research 26, 703–720.
Roesler, C., Perry, M., Carder, K., 1989. Modeling in situ phytoplankton
absorption from total absorption spectra in productive inland marine
waters. Limnology and Oceanography 34, 1510–1523.
Romero, S.I., Piola, A.R., Charo, M., Garcia, C.A.E., 2006. Chlorophyll-a
variability off Patagonia based on SeaWiFS data. Journal of
Geophysical Research 111, C05021.
Sabatini, M., Reta, R., Matano, R., 2004. Circulation and zooplankton
biomass distribution over the southern Patagonian shelf during late
summer. Continental Shelf Research 24, 1359–1373.
Sasaki, H., Miyamura, T., Saitoh, S.-i., Ishizaka, J., 2005. Seasonal variation
of absorption by particles and colored dissolved organic matter
(CDOM) in Funka Bay, southwestern Hokkaido, Japan. Estuarine and
Coastal Shelf Science 64, 447–458.
Sathyendranath, S., Stuart, V., Irwin, B.D., Maass, H., Savidge, G., Gilpin, L.,
Platt, T., 1999. Seasonal variations in bio-optical properties of
phytoplankton in the Arabian Sea. Deep-Sea Research Part II-Topical
Studies in Oceanography 46, 633–653.
Sathyendranath, S., Cota, G., Stuart, V., Maass, H., Platt, T., 2001. Remote
sensing of phytoplankton pigments: a comparison of empirical and
theoretical approaches. International Journal of Remote Sensing 22,
249–273.
Schloss, I.R., Ferreyra, G.A., Ferrario, M.E., Almandoz, G.O., Codina, R.,
Bianchi, A.A., Balestrini, C.F., Ochoa, H.A., Ruiz Pino, D., Poisson, A.,
2007. Role of plankton communities in sea–air variations in pCO2 in
the SW Atlantic Ocean. Marine Ecology Progress Series 332, 93–106.
Sieburth, M.McN., Smetacek, V., Lenz, J., 1978. Pelagic ecosystem
structure: heterotrophic compartments of the plankton and their
relationships to plankton size fractions. Limnology and Oceanogra-
phy 23, 1256–1263.
Siegel, D.A., Michaels, A.F., 1996. Quantification of non-algal attenuation
in the Sargasso Sea: implications for biogeochemistry and remote
sensing. Deep-Sea Research Part II-Topical Studies in Oceanography
43, 321–345.
Siegel, D.A., Maritorena, S., Nelson, N.B., Behrenfeld, M.J., Mcclain, C.R.,
2005. Colored dissolved organic matter and its influence on the
satellite-based characterization of the ocean biosphere. Geophysical
Research Letters 32, L20605.
Signorini, S.R., Garcia, V.M.T., Piola, A.R., Garcia, C.A.E., Mata, M.M.,
Mcclain, C.R., 2006. Seasonal and interannual variability of cocco-
lithophore blooms in the vicinity of the Patagonian shelf break
(381S–521S). Geophysical Research Letters 33, L16610.
Stuart, V., Sathyendranath, S., Head, E.J.H., Platt, T., Irwin, B., Maass, H., 2000.
Bio-optical characteristics of diatom and prymnesiophyte populations
in the Labrador Sea. Marine Ecology Progress Series 201, 91–106.
Trees, C.C., Clark, D.K., Bidigare, R.R., Ondrusek, M., Mueller, J., 2000.
Accessory pigments versus chlorophyll a concentrations within
euphotic zone: a ubiquitous relationship. Limnology and Oceano-
graphy 45, 1130–1143.
Twardowski, M.S., Donaghay, P.L., 2002. Photobleaching of aquatic dissolved
materials: absorption removal, spectral alteration, and their interrela-
tionship. Journal of Geophysical Research 107 (C8), 3091.
Twardowski, M.S., Boss, E., Sullivan, J.M., Donaghay, P.S., 2004. Modeling
the spectral shape of absorption by chromophoric dissolved organic
matter. Marine Chemistry 89, 69–88.
Twardowski, M.S., Lewis, M., Barnard, A., Zaneveld, J.R.V., 2005. In-water
instrumentation and platforms for ocean color remote sensing
applications. In: Miller, R., Del Castillo, C., McKee, B. (Eds.), Remote
Sensing of Coastal Aquatic Waters. Springer Publishing, Dordrecht,
Netherlands, pp. 69–100.
Vasilyev, V.I., 2007. Zooplankton of the Patagonian slope and adjacent
waters in the autumn. Bulletin of Moscow University, Biological
Sciences/Vestnik Moskovskiin Universitet Biologiia 62 (1), 44–48.
Vodacek, N.V., Blough, M.D., Degrandpre Peltzer, E.T., Nelson, R.K., 1997.
Seasonal variation of CDOM and DOC in the Middle Atlantic Bight:
terrestrial inputs and photooxidation. Limnology and Oceanography
42, 674–686.
Welschmeyer, N.A., 1994. Fluorometric analysis of chlorophyll-a in the
presence of chlorophyll-b and pheopigments. Limnology and
Oceanography 39 (8), 1985–1992.