Corteseetal2012 Fkerg

Corteseetal2012 Fkerg
PALEOCEANOGRAPHY, VOL. 27, PA1208, doi:10.1029/2011PA002187, 2012
Glacial-interglacial size variability in the diatom
Fragilariopsis kerguelensis: Possible iron/dust controls?
G. Cortese,1 R. Gersonde,2 K. Maschner,3 and P. Medley4
Received 16 June 2011; revised 6 December 2011; accepted 7 December 2011; published 16 February 2012.
[1] The valve area of Fragilariopsis kerguelensis, the most abundant diatom species in the
Southern Ocean, strongly changes in size in response to varying conditions in the surface
ocean. We examined the link, both in two iron fertilization experiments and in sediment
samples covering several glacial Terminations, between size variability in this species
and environmental conditions across the Antarctic Polar Front, including sea ice extent,
sea surface temperature, and the input of eolian dust. The iron fertilization experiments
show valve area to be positively correlated with iron concentrations in ambient waters,
which suggests the possibility of a causal relation between valve size of Fragilariopsis
kerguelensis and ambient surface water iron concentration. Larger valves are usually found
during glacial times and thus seem to be related to lower sea surface temperature and wider
sea ice coverage. Moreover, our results indicate that there usually is a strong correlation
between larger valve size and increased input of eolian dust to the Southern Ocean.
However, this correlation, obvious for the fertilization experiments and for glacial
Terminations I, II, III, and V, does not seem to be valid for Termination VI, where size
appears to be inversely correlated to dust input.
Citation: Cortese, G., R. Gersonde, K. Maschner, and P. Medley (2012), Glacial-interglacial size variability in the diatom
Fragilariopsis kerguelensis: Possible iron/dust controls?, Paleoceanography, 27, PA1208, doi:10.1029/2011PA002187.
1. Introduction
[2] The export of carbon dioxide from the atmosphere
toward the deep ocean represents a sink for atmospheric
carbon dioxide, and this export process is referred to as the
“biological pump” [Falkowski et al., 1998]. This pump is
controlled by changes in the primary productivity and export
efficiency of the organisms living at the surface of the ocean.
Past changes in the efficiency of the biological pump during
glacial and interglacial times could have directly affected the
atmospheric carbon dioxide content.
[3] In high-nutrient, low-chlorophyll areas (HNLC) of the
ocean, phytoplankton does not completely utilize available
nutrients, and high levels of macronutrient (e.g., nitrogen
and phosphorus) correspond to lower than expected primary production. This puzzling observation was explained in
terms of productivity limitation due to lack of the micronutrient iron in the upper water column [Martin and Fitzwater,
1988; Martin et al., 1990]. As a corollary of this, Martin et al.
[1990] hypothesized that large blooms of phytoplankton
could be stimulated by adding iron into HNLC areas,
Department of Paleontology, GNS Science, Lower Hutt, New Zealand.
Geosciences Division, Alfred Wegener Institute for Polar and Marine
Research, Bremerhaven, Germany.
German Oceanographic Museum, Stralsund, Germany.
Cooperative Institute for Research in Environmental Science, University
of Colorado at Boulder, Boulder, Colorado, USA.
Copyright 2012 by the American Geophysical Union.
increasing the efficiency of the biological pump and resulting
in a net sequestration of carbon dioxide from the atmosphere.
[4] Large-scale additions of dissolved iron to HNLC areas
(iron fertilization experiments) have been performed extensively during the last two decades at a variety of locations
[e.g., Martin and Fitzwater, 1988; de Baar et al., 1990; Martin
et al., 1994; Coale et al., 1996; Boyd et al., 2000]. These iron
fertilization experiments, along with a series of in vitro
experiments [Hutchins and Bruland, 1998; Takeda, 1998; De
La Rocha et al., 2000; Timmermans et al., 2004] have demonstrated how iron limitation strongly influences the relative
uptake ratios of the main nutrients in diatoms, and thus relief
from iron limitation may have a strong imprint on the biogeochemical cycle of these nutrients. In particular, the most
common large, bloom-forming diatoms in the Southern Ocean
(Fragilariopsis kerguelensis, Actinocyclus sp., Thalassiosira sp.)
display a marked decrease in the silicate/nitrate depletion
ratio with increased iron availability [Timmermans et al.,
2004]. This is a result of a decrease of silicate (and concomitant increase in nitrate) consumption in response to
increasing dissolved iron concentration [De La Rocha et al.,
2000; Timmermans et al., 2004].
[5] Interestingly, such a relief from iron limitation has
large consequences on the export of dissolved nutrients from
the Southern Ocean toward lower latitudes [Sarmiento et al.,
2004]. In fact, the modern Southern Ocean is iron limited
(higher silicate/lower nitrate consumption) and thus exports
nitrate toward lower latitudes via Sub-Antarctic Mode Water
(SAMW). When the Southern Ocean is under iron-replete
conditions (lower silicate/higher nitrate consumption) it
1 of 14
Figure 1. Location of the studied sediment cores and iron fertilization experiments. The background map
represents the modern average valve area in Fragilariopsis kerguelensis (reprinted from Cortese and
Gersonde [2007], with permission from Elsevier). The black dots are the surface sediment samples used
to produce the map, and oceanic frontal positions and maximum sea ice extent are according to Belkin and
Gordon [1996]: Subtropical Front (red), Sub-Antarctic Front (yellow), Antarctic Polar Front (purple),
winter sea ice edge (blue).
exports silicate, thus strongly affecting ecosystem functioning at a global level [Sarmiento et al., 2004].
[6] An interesting aspect of iron-limited diatom blooms is
its connection to the observation that iron limitation is
relieved in the Southern Ocean during glacial times, due to
increased levels of iron in the surface waters. The main
sources of iron to the surface ocean are assumed to be eolian
dust, upwelling, and melting icebergs. Martínez-Garcia et al.
[2009] recently demonstrated how iron fertilization has been
a recurrent feature of the sub-Antarctic region over last 1.1 Ma,
as dust/iron inputs to the ocean and marine export production
records are closely correlated over the entire interval.
[7] The European Project for Ice Coring in Antarctica
(EPICA) recovered an ice core that provides a climate record
for the past 800,000 years [EPICA Community Members,
2004; Jouzel et al., 2007]. We will use the dust record from
this core as a proxy for dust deposition into the Southern
Ocean, following John Martin’s iron fertilization hypothesis
and assuming that the fertilization during glacial periods
mainly occurs through dust-borne iron [Martin et al., 1990].
[8] The Southern Ocean is the largest HNLC area of the
world, and here marine diatoms are the main primary producer and exporter of organic carbon. In the Southern Ocean,
the dominant diatom species Fragilariopsis kerguelensis
displays a strong variability in size [Cortese and Gersonde,
2007], but the specific environmental controls for this variability are currently unknown.
[9] The present study is aimed at testing whether the
observed pattern of larger F. kerguelensis during glacial
times, and smaller F. kerguelensis during interglacial times at
Termination I [Cortese and Gersonde, 2007] also holds true
for older Terminations. We additionally want to investigate
the link between size changes in this diatom and several
proxy records including summer sea surface temperature (SST), sea ice extent, major community shifts within
diatoms, and eolian input. Under perturbed modern conditions, we also analyze the size variability of this species
during two artificial iron fertilization experiments in the
Southern Ocean: EIFEX [Smetacek et al., 2005] and EisenEx
[Smetacek, 2001].
2. Material and Methods
[10] We analyzed microscopic slides prepared from water
samples collected during two iron fertilization experiments
(Figure 1 and auxiliary material Data Set S1): the EIFEX iron
fertilization experiment, conducted at the Antarctic Polar
Front (49.4°S, 2.25°E) during R/V “Polarstern” expedition
ANTXXI/3 [Smetacek et al., 2005], and the EisenEx fertilization experiment, that took place at 48°S, 21°E during
expedition ANTXVIII/2 [Smetacek, 2001].1
[11] Sediment samples covering the last glacial Termination (Figure 1 and Data Set S2) were analyzed from R/V
“Polarstern” cores PS1654 (50.2°S, 5.7°E, Antarctic Polar
Front), PS2498 (44.2°S, 14.2°W, north of Subantarctic
Front), and PS2499 (46.5°S, 15.3°W, Subantarctic Front).
Additional sediment samples (Data Set S3) were obtained
from Ocean Drilling Program (ODP) Site 1093, cored in the
southeast Atlantic sector of the Southern Ocean during Leg
177, close to the position of piston core PS1654 [Hodell
et al., 2002]. This site is located at the Antarctic Polar Front
and the examined sediment core documents the last 7 glacial
cycles (ca. 600,000 years).
[12] The age model for ODP Site 1093 has been developed
using the software Analyseries [Paillard et al., 1996] to
match the diatom-derived SST record from this site [SchneiderMor et al., 2008] and the EPICA ice core temperature record
Auxiliary material data sets are available at
2011pa002187. Other auxiliary material files are in the HTML.
2 of 14
Figure 2. Schematic representation of a diatom’s life cycle. The number of generations involving asexual
reproduction is not limited to three, and the switch to sexual reproduction involves the release of gametes,
formation of a zygote, and resetting of the valve size to the maximum one. The image at the bottom is a
valve of F. kerguelensis, with the main measurements taken in this study shown as red bars.
[Jouzel et al., 2007] plotted on its EDC3 chronology
[Parrenin et al., 2007]. The diatom-derived SST data for
ODP Site 1093 are reported in Data Set S5.
[13] The cleaning of the sediment samples and the preparation of permanent slides for light microscopy followed a
routine method established at the Alfred Wegener Institute for
Polar and Marine Research (AWI), described by Gersonde
and Zielinski [2000]. Pictures and measurements of the diatom species Fragilariopsis kerguelensis were taken following
a method described by Cortese and Gersonde [2007]. An average of 40.5 specimens for each of the 206 samples (8343 pictures
in total) has been photographed using a video camera
attached to a Zeiss Axioskop microscope, at 1000x magnification. The public domain image analysis software “ImageJ”
(freely available from the National Institute of Health Webpage: has been used to measure the length
and width of the specimen, and the length of five costae
within it (see Figure 2, image at bottom).
[14] Measurements of iron and dust concentrations/fluxes
from the EPICA Dome C core are from EPICA Community
Members [2004], Wolff et al. [2006], and Lambert et al.
[2008]. The reader is referred to these papers for details on
the analytical techniques used. Fe fluxes from the EPICA
Dome C core are reported in Data Sets S4 and S5. The data
pertaining to the EPICA ice core have been plotted on its
EDC3 chronology [Parrenin et al., 2007].
[15] MATLAB® was used to perform various statistical
tests: statistical power, chi-square, two-sample KolmogorovSmirnov, and Wilcoxon rank sum test.
[16] The statistical power test was used to evaluate whether
forty diatom specimens were a large enough sample size to
capture the variability in the real population and justify the
inferences made in this paper. A chi-square test was used
to assess whether valve area data were normally distributed,
and two-sample Kolmogorov-Smirnov and Wilcoxon rank
3 of 14
sum tests were used to check whether glacial and interglacial
valve areas were significantly different.
[17] A sample size of 40 measured specimens proved to be
adequate to be able to separate, at power 0.90 and via a t test,
between a sample from a glacial interval and one from an
interglacial interval, and vice versa.
[18] As the chi-square tests indicated that most of the tested
populations had nonnormal distributions, the further tests we
used to evaluate significant differences in our data sets were
[19] Both the two-sample Kolmogorov-Smirnov and
Wilcoxon rank sum test confirmed how the valve area values
measured in samples coming from glacial and interglacial
intervals are significantly different, both within the same core
(tested for PS2498, PS2499, PS1654) and for the whole
Termination I data set.
[20] More detailed explanations of statistical testing are
available in Texts S1 and S2, while Figures S1 and S2 provide error bar graphs (age versus area with standard deviation
for each data point) for all the valve area records.
3. Diatom Life Cycle
[21] Diatoms (microscopic algae belonging to the Class
Bacillariophyceae) require sunlight to photosynthesize, limiting their distribution to the uppermost 200 m of the water
column. They are especially abundant in the ocean, where
they are estimated to contribute up to 45% of the total oceanic
primary production [Mann, 1999]. The hallmark of diatoms
is their pillbox shaped cell wall made of silica (SiO2 nH2O),
which can be preserved in the sedimentary record.
[22] The abundance of diatoms varies in space and time
as a response to environmental variables and availability
of macronutrients (nitrate, phosphate) and micronutrients
(iron). Some of these nutrients may at times not be available
in high enough levels, and thus become limiting for diatom
growth and bloom formation. During such blooms, diatoms
rapidly increase their numbers by repeatedly undergoing
vegetative (asexual/mitotic) reproduction by cell fission.
Diatoms are characterized by a peculiar structure and division modality. The pillbox shaped cell wall is constituted by
two valves, the bottom valve (hypotheca) is slightly smaller
than the upper one (epitheca). When a cell undergoes division, each daughter cell always builds a new hypotheca.
It follows that one of the two daughter cells will end up being
slightly smaller than its parent (Figure 2) and, through successive divisions, the average size of diatom cells in a population will decrease. This process can be compensated by
the formation of new large cells that occurs during sexual
reproduction. The sexual phase can be induced in cells that
have reached a species-specific threshold size and includes
the formation of gametes, their fusion, and formation of a
zygote (auxospore) within which cells of the maximum size
are produced [Chepurnov et al., 2004]. Such a switch to
a sexual phase has been demonstrated to occur, with the
formation of auxospores, in F. kerguelensis during an ironinduced bloom in the Southern Ocean [Assmy et al., 2006].
We therefore expect larger spores upon relief from iron
[23] In sediment samples, many successive generations/
blooms of diatom valves are accumulated in the same layer.
As such, the average valve size of a population will represent
an integrated record of several bloom seasons and the
frequency at which sexual reproduction was induced over a
time interval of possibly several hundred years, rather than
reflecting the dynamics of a single bloom episode. Nonetheless, the sedimentary record of diatom size variability
seems to suggest that size in some diatom species might be
strongly related to specific environmental conditions.
[24] Valves of F. kerguelensis range generally from 10 to
75 mm in length, and 5 to 11 mm in width. The valves have an
elliptical to lanceolate shape, and are isopolar (i.e., the two
extremes, poles, are identical, and the valve is symmetrical
along the transversal axis). Four to seven costae are present
over 10 mm, and two rows of alternating pores, easily
resolved under light microscopy, are present between two
adjacent costae. The length of five costae (Figure 2) were
recorded to allow comparisons to the results of Fenner et al.
[1976], who first investigated F. kerguelensis size variability
in the South Pacific Ocean.
4. Results
4.1. Fertilization Experiments
[25] In order to test how iron availability might affect size
in Fragilariopsis kerguelensis, we analyzed several samples
collected during two iron fertilization experiments (Figure 1)
conducted at the Antarctic Polar Front in the Atlantic Sector
of the Southern Ocean (EIFEX [Smetacek et al., 2005]
and EisenEx [Smetacek, 2001]). During these experiments,
a suitable oceanic eddy in a high-nutrient, low-chlorophyll
area is identified and tracked with an inert tracer (usually SF6,
sulphur hexafluoride), while at the same time being fertilized
with several tons of dissolved iron. Time series for many
biological and geochemical indicators are then generated
for the few weeks following the fertilization, in order
to document the occurrence and characteristics of the
ensuing plankton bloom. We studied changes in average
valve sizes both inside the fertilized patch, and in a few
additional control samples outside it, for both the EIFEX and
EisenEx fertilization experiments. The average valve size of
Fragilariopsis kerguelensis is smaller in the outside patch
compared to the fertilized patch throughout the EIFEX iron
fertilization experiment (Figure 3). Inside the fertilized patch,
the size remains at around 160 mm2 during the first 8 days of
the experiment, then drops to ca. 125 mm2, and displays a
steady rising trend after ca. 12 days, with a maximum size of
210 mm2 at about 36 days from the start of the experiment.
The four control samples from outside the fertilized patch
display an offset of ca. 30 mm2 compared to the in-patch
samples, and the size increases only slightly after ca. 17 days
from the beginning of the fertilization. However, average
valve size is greater in patch even before the fertilization
occurs, and throughout the experiment. The similar increasing trends from day 18 to 33 both in and out of patch suggest
that the response in patch is mirrored by a similar response
outside of the fertilized patch.
[26] With the exception of one sample, the size of
F. kerguelensis is larger in the in patch, compared to the out
patch samples during the EisenEx experiment as well, and
the offset between the two curves is ca. 20 mm2. Although
there is a slight trend toward larger sizes as the experiment
progressed, the last analyzed sample (ca. 155 mm2 at 21 days
from start of the experiment) most likely does not capture the
4 of 14
Figure 3. F. kerguelensis valve area through two iron fertilization experiments. Samples from EIFEX
[Smetacek et al., 2005] and EisenEx [Smetacek, 2001]. The solid lines represent samples from within the
fertilized patch, while the dotted lines refer to the out-patch, control samples.
rapid increase in size observed during the second half of the
EIFEX experiment. A very rapid increase is observed, for a
single sample, two days after the start of the experiment.
4.2. Last Glacial Termination (Termination I)
[27] The valve size evolution through the last 35 ka
(Figure 4) is remarkably similar for the three examined
locations, two of which are from north of the Sub-Antarctic
Front (PS2498 and PS2499), while the remaining one
(PS1654) has been sampled at the Antarctic Polar Front in the
Atlantic Sector of the Southern Ocean (Figure 1). All three
cores have substantially larger valves during glacial compared to interglacial times. Average valve sizes oscillate
around 190–210 mm2 during the interval 35 to 17–18 kyr,
then rapidly decrease throughout the glacial Termination,
and stabilize to their smaller interglacial sizes after ca. 8 kyr
BP. The only two noticeable differences among the core sites
are the following:
[28] 1. The Antarctic Polar Front location (core PS1654)
seems to document larger valve size during the Holocene
(140–150 mm2) compared to the 100–110 mm2 observed at
the two sub-Antarctic sites (PS2498 and PS2499).
[29] 2. At the Sub-Antarctic Front location (core PS2499),
average valve size at the start of glacial Termination
(220 mm2 at 18 ka) is much larger than at the other two core
[30] Dust flux values for the Epica Dome C (EDC) ice core
are much higher during glacial than interglacial times: the
main decrease occurs at ca. 19–16 kyr (and is thus synchronous to the decrease in valve area), after which dust flux
values are very low.
4.3. Previous Terminations
[31] The analyzed sediment samples from ODP Site 1093
(Antarctic Polar Front) cover the last 550 kyr, including the
four transitions from glacial to interglacial conditions that we
analyzed (Terminations II, III, V, and VI). We were particularly interested in examining changes in average valve size
of F. kerguelensis during glacial, interglacial, and glacial
Termination conditions, and their link to several environmental proxy records including sea surface temperature, sea
ice extent, major ecosystem shifts within diatoms, and eolian
dust/Fe flux input (Figures 5 and 6).
4.3.1. Termination II
[32] At ODP Site 1093, SST [Schneider-Mor et al., 2008]
is much colder under full glacial conditions (Marine Isotope
Stage, MIS 6) than during the following full interglacial
(MIS 5). This is marked with a sharp transition (Figure 5),
with SST rising from ca. 0.5°C to an interglacial maximum of
ca. 5°C. Following an interglacial optimum, SST slightly
decreases and fluctuates around 3.5°C. There is a clear correlation between SST and relative abundances of the Fragilariopsis curta group, a proxy for winter sea ice (Figure 5):
during full glacial times, when SST is low, F. curta is constantly abundant and reaches its peak right before the start of
Termination II. With the onset of the transition, the abundance of this species drops to about zero, indicating how
there is almost no sea ice present during interglacial times.
F. kerguelensis is abundant during the whole period, with a
slight increase during warmer interglacial times (Figure 5).
Shortly before and during Termination II, this species shows
a sharp shift in size (Figure 6), as its average valve area
switches from high values during glacial times (ca. 200 mm2)
to much lower values during interglacial intervals
(ca. 100 mm2). Termination II shows a drop to the lowest
observed average area values (ca. 70 mm2). A positive correlation is apparent between iron input and average area
(Figure 6), as during full glacial times iron is high and
constantly available for F. kerguelensis and valve areas are
around 200 mm2. As iron decreases during Termination II,
5 of 14
Figure 4. (top) F. kerguelensis valve area and dust flux at EPICA Dome C. The Marine Isotope Stage
(MIS) boundaries are after Lisiecki and Raymo [2005], while the iron flux data are from Lambert et al.
[2008] and are also reported in Data Set S4. (bottom) IRD record from ODP Site 1094 and Th-corrected
opal flux data during Termination I. The ice-rafted debris record from IODP Site 1094 [Kanfoush et al.,
2002] appear as a dotted black line, and the Th-corrected opal flux data [Anderson et al., 2009] from
cores TN057–013PC4, TN057–014PC4, and EL27–23PC are shown in shades of red. The right-hand scale
applies to both IRD (in mg/g) and Th-corrected opal flux (in g cm 2 kyr 1). The vertical dashed line at
17 ka roughly marks the start of the glacial Termination, with a sharp decrease in valve size coinciding with
the shift from (larger) predominantly eolian/IRD to (smaller) upwelling/iceberg sources of iron.
6 of 14
Figure 5. SST, F. kerguelensis and F. curta percentages, and F. kerguelensis valve area at ODP Site 1093,
plus EPICA Dome C Fe flux. SST (estimated via a diatom transfer function) and diatom abundance data
from Schneider-Mor et al. [2008]. Iron flux data for the EPICA Dome C ice core [EPICA Community
Members, 2004; Jouzel et al., 2007; Lambert et al., 2008] are also shown. Vertical dashed lines mark
glacial terminations.
average area values plunge to approximately 70 mm2. During
the following interglacial period, iron values remain very low
and average area values fluctuate around 110 mm2.
4.3.2. Termination III
[33] SST is slightly colder under full glacial conditions
(MIS 8) than during full interglacial (MIS 7), with SST rising
from ca. 2°C to a ca. 5°C maximum (Figure 5). During the
7 of 14
Figure 6. F. kerguelensis valve area and SST at ODP Site 1093, plus EPICA Dome C dust flux. The four
graphs show enlargements for Terminations II, III, V, and VI for the EPICA dust flux [EPICA Community
Members, 2004], SST at OPD Site 1093 [Schneider-Mor et al., 2008], and average valve area in
F. kerguelensis from ODP Site 1093 (this study). The right-hand scale is valid for both SST (°C) and dust
flux (10 2 mg/m2 y). The vertical black lines represent glacial Terminations, with their timing according to
Lisiecki and Raymo [2005], and separate interglacial MIS (uneven numbers) from glacial marine isotopic
stages (even numbers).
MIS 7 interglacial, SST decreases and fluctuates around 3°C.
The correlation between SST and abundances of the sea ice–
related diatom F. curta is clearly visible (Figure 5): during
full glacial times, when temperatures are low, F. curta is
constantly abundant and reaches peak abundances of 4%.
At the onset of the Termination, the abundance of F. curta,
and therefore sea ice concentrations, drops to almost zero.
F. kerguelensis is constantly abundant during the whole
period, with the exception of a large decrease during full
glacial conditions (Figure 5). Iron concentrations are high
during full glacial and positively correspond with high
average area values around 170–180 mm2 for F. kerguelensis
(Figure 6). Once iron concentration decreases during
Termination III, average area values sharply decrease
to approximately 100 mm2. During the interglacial, iron
values are very low and average area values fluctuate around
150–160 mm2.
4.3.3. Termination V
[34] SST at full glacial conditions (MIS 12) is distinctly
lower than during full interglacial conditions (MIS 11), with
a SST rise from ca. 0.5°C to a maximum of ca. 6.5°C
(Figure 5). During the MIS 11 interglacial, SST is constant
and fluctuates around 5.5°C. Over this time interval as well,
there is a strong correlation between SST and amount of sea
ice, as recorded by F. curta abundances. During full glacial
times, sea ice is relatively abundant, with F. curta abundance
peaking at 5% (Figure 5). At the onset of the Termination
and during interglacial, sea ice and F. curta abundance levels
drop to almost zero. F. kerguelensis displays high abundances during the whole period with an increase during
interglacial (Figure 5). During full glacial, iron is high
and positively corresponds with high average area valves
with values of ca. 190 mm2 for F. kerguelensis (Figure 6).
A declining trend toward smaller interglacial valves (140–
150 mm2) is visible, corresponding to iron levels reaching
almost zero.
4.3.4. Termination VI
[35] SST at full glacial conditions (MIS 14) is slightly
colder than during full interglacial conditions (MIS 13), as
Termination VI documents a SST rise from 1.5°C to a maximum of 4°C (Figure 5). The temperature contrast between
glacial and interglacial is thus small compared to later glacial
Terminations. Temperature and F. curta abundances are well
correlated as, during full glacial times, F. curta reaches a
peak abundance of 1.5%, indicating presence of sea ice in the
proximity of the core location during this glacial period,
although in lower abundance, and with a sea ice edge located
south of the core location, compared to what is observed for
later glacial periods (Figure 5). The cutoff value for sea ice
presence is a relative abundance of 3% F. curta, therefore
8 of 14
both sea ice indicators and dust flux [Lambert et al., 2008]
display lower values before MIS 12 compared to later glacial
[36] At the onset of the Termination, F. curta abundances
drop and during interglacial they remain at around zero,
indicating no sea ice presence during this time interval.
F. kerguelensis is abundant throughout this interval, with
little fluctuation around 60% (Figure 5). The abundance
pattern for this species is opposite to what is commonly
observed, with slightly higher abundances (up to 70%) during the glacial compared to the interglacial interval (down to
50%). Its average valve area increases while iron flux
decreases (Figure 6), also an opposite trend compared to
what is observed at the other glacial Terminations. The
highest average area measured in this study (240 mm2) was
found during the interglacial time.
5. Discussion
5.1. Iron Fertilization Experiments EIFEX
and EisenEX
[37] During EIFEX all in patch and out patch stations were
located within the eddy core, but the latter was not homogenous in terms of chlorophyll distribution, and mini blooms
which faded away over the course of the experiment were
sampled in some of the initial out stations (P. Assmy,
personal communication, 2011). Moreover, some degree of
diffusion between the in- and out-patch regions is to be
[38] These heterogeneous conditions in the fertilized patch
explain why our area values in patch and out patch differ
significantly at fertilization time and before ca. day 8–12,
when a significant increase in auxospore numbers is reported
to occur [Assmy et al., 2006], and why parallel, increasing
trends from day 18 to 33 occur both inside and outside the
fertilized patch. Additionally, Station 424 (the first “out
patch” sample in Figure 3) can be classified as in patch/out
patch, as sampled just at the border between the two areas,
briefly before the Fe addition at the beginning of the EIFEX
[39] However, independent studies by P. Assmy (unpublished data, 2011) confirm that the largest (>65 mm) apical
length fraction of F. kerguelensis was more abundant inside
as compared to outside the patch, in accordance with what
we observe for valve area (Figure 3).
[40] We document a trend in the iron fertilization experiments with an increase through time in valve area both
inside and outside the fertilized patch. This indicates
F. kerguelensis reacts to iron addition and to the development
of a diatom bloom at the Antarctic Polar Front. This is particularly true for the EIFEX experiment, where the average
valve area after 5 weeks is 40% larger than at the beginning
of the fertilization.
[41] The presence of populations having, in average, larger
valves inside the fertilized patch compared to the outside
patch controls (Figure 3) suggests, in line with what found
by Assmy et al. [2006] during the EIFEX experiment, that
the observed increase in the number of F. kerguelensis
auxospores upon relief from iron limitation strongly affects
the population’s size range.
[42] This is also in accordance with the observation that
under iron-limited conditions most diatom species decrease
their cell volumes by up to 50% [Maldonado and Price,
1996; Marchetti and Harrison, 2007], with the extent of
this reduction being species specific. A smaller cell volume
has many, very important, physiological consequences for
the cell: it implies an increase of the cell surface to volume
ratio, a decrease of the diffusive boundary layer thickness,
a maximization of nutrient uptake rates, and a decrease in
the diffusion-limited threshold for optimal growth [Hudson
and Morel, 1990; Pahlow et al., 1997]
[43] These observations lead us to turn to sedimentary
records, as there are indications of increased dustiness and
dust flux into the Southern Ocean and Antarctica during past
glacial intervals [Mahowald et al., 2006]. If this dust input
increased the amount of bioavailable dissolved iron into
the Southern Ocean, a large-scale “natural iron fertilization”
(and relief from iron limitation) of the Southern Ocean might
have occurred during these time intervals. In turn, this would
have affected the average valve size in F. kerguelensis.
[44] There are, however, two caveats: (1) in the iron fertilization experiments (Figure 3), the trend in valve size with
time inside the patch is paralleled by a similar valve size
evolution outside the fertilization patch, and (2) the proposed
pattern of a positive correlation of valve size with dust flux
at glacial-interglacial scale (Figure 6) could as well be
interpreted as a negative correlation of valve size with SST.
For these reasons, the iron/dust–valve size correlation cannot
serve as an indication of a causal relation between iron and
valve size, and has to be considered as a working hypothesis
at this stage.
5.2. Last Glacial Termination
[45] Cortese and Gersonde [2007] demonstrated that
F. kerguelensis strongly fluctuates in size, with larger valves
occurring at the Antarctic Polar Front during the last glacial
period, and smaller valves prevailing during the Holocene
interglacial. The present study was aimed at testing whether
this pattern also held true at other locations in the Southern
Ocean and/or at older glacial Terminations.
[46] The link between higher iron availability and larger
valves in F. kerguelensis, which is hypothesized in the iron
fertilization data, seemed to remain valid for last glacial
Termination, according to our sediment data along three
cores recovered from the Southern Ocean (core PS1654 from
the Antarctic Polar Front, cores PS2498 and PS2499 from
north of the Sub-Antarctic Front). At these different locations, the response to strongly decreased dust flux to the
EDC ice core in Antarctica (and therefore presumably lower
dust/Fe input to the Southern Ocean) was a strong reduction
in size (from ca. 200 mm2 during the glacial to ca. 100 mm2
during the interglacial). Valve size remained very high during the last portion of Marine Isotope Stage (MIS) 3, through
most of MIS 2, and started to decrease in concert with the
decrease of dust flux between 19 and 16 kyr BP. Under
MIS 1 interglacial conditions, valves seem to be larger
(PS1654, ca. 150 mm2) at the Antarctic Polar Front (APF)
compared to the Sub-Antarctic Zone (ca. 100 mm2).
[47] The correlation between valve size and dust flux
(Figures 4–6), while striking, may at least partly also be
a reflection of changes in wind patterns and intensity,
two climate mechanism that can potentially impact diatom
assemblages and valve size through their modulation of other
processes in the circum-Antarctic ocean, such as upwelling
9 of 14
and frontal migrations. As an example of this, the trend of
smaller while diverging valve sizes during Termination I and
MIS 1 (Figure 3) after a prolonged time of similar valve sizes
during full glacial conditions may be a consequence of
diverging ocean regimes at these three core sites, probably in
association with shifting circum-Antarctic wind patterns and
oceanic fronts.
[48] The different baseline values in the Holocene for
PS1654/ODP1093 (at the APF) are therefore probably a
consequence of the different sources of iron available at the
APF today (Figures 1 and 4), even under full interglacial
conditions, compared to sites located further to the north in
the Sub-Antarctic Zone (PS2499 and PS2498). At the APF,
upwelling of deep waters [Lefèvre and Watson, 1999] and
melting icebergs/sea ice [Raiswell et al., 2006; Smith et al.,
2007; Geibert et al., 2010] may provide sources of
bioavailable iron that fuel slightly larger valve sizes in
F. kerguelensis, even in the absence of the dominant eolian
source. None of those additional iron sources seem to play an
important role in the Sub-Antarctic Zone.
[49] In order to test whether other sources of iron, in
addition to eolian dust, might at times have become important close to the Antarctic Polar Front in the Southern Ocean
over last 35 ka, we compared (Figure 4) our valve area record
for last glacial Termination to records of ice-rafted debris
[Kanfoush et al., 2002], a proxy for iron sourced from melting icebergs, and Th-corrected opal flux [Anderson et al.,
2009], a proxy for iron sourced from upwelling of deep
[50] However, overall in the Atlantic Southern Ocean IRD
deposition is important only during glacials [Diekmann et al.,
2003] and thus acts in the same direction as the eolian dust
deposition. In fact, more than 90% of the IRD signal recorded
by Kanfoush et al. [2002] was later identified to represent
tephra particles transported by sea ice from the South Sandwich Island volcanic arc, and not by icebergs from the Antarctic continent [Nielsen et al., 2007]. It is highly unlikely
that the tephra particles, when falling through the water column after sea ice melt, will have an effect on the micronutrient concentration in surface waters.
[51] More likely sources besides the eolian input [see
Ridgwell and Watson, 2002] are the advection of Fe-enriched
deep water by intensified upwelling and icebergs, as proposed
by Abelmann et al. [2006] for Termination I in the study area.
Such mechanisms of Fe deposition, although not exceeding
the effect of dust, may explain the prolonged decline of the
F. kerguelensis valve size at glacial terminations.
[52] Increased Th-corrected opal flux (used as a proxy for
increased advection of the nutrient silicon by upwelling) is
restricted to the 10–17 ka time window (i.e., at the glacial
Termination). In contrast, strong upwelling of deep and
nutrient-rich waters during glacials is not supported by
nutrient proxies [Sigman et al., 2010]. The potential of
Fe deposition by iceberg melting was highlighted by Smith
et al. [2007].
[53] We can draw the following conclusions from these
observations: (1) Iron deposition in surface waters by
upwelling is negligible during the glacial (and thus cannot
have an influence on glacial valve size), but may play a
role during the increased upwelling of nutrient-rich water
at glacial terminations; (2) a second source of Fe may
be iceberg melting, preferentially at terminations; and
(3) both additional sources (upwelling and icebergs) do
not cause a valve size effect that exceeds the one caused by
dust deposition during glacials. The valve area data set for
Termination I also allowed us to test whether there was a
change in population structure between MIS 1 to MIS 2, and
what the possible biological reasons for this may have been.
There is a certain degree of positive correlation between
valve size and standard deviation, albeit variable between
glacial Terminations/sites (Figure S3a), with a maximum
correlation value (r2 = 0.49) observed for ODP Site 1093,
with all previous glacial Termination data grouped together.
This positive correlation might suggest a size spectrum shift
from glacial to interglacial times, with a broader size range
during glacial intervals.
[54] The histograms for all valve area measurements from
MIS1 and MIS2 (Figures S3b and S3c), both all grouped
together and split core by core (PS2499, PS2498, PS1654),
albeit not easy to compare directly, as they are based on
a different number of observations, seem to indicate how
(1) both distributions are skewed toward low values, and this
is more extreme for the MIS1 interglacial population, and
(2) the size range is broader for the glacial population,
particularly for values from ca. 200 to 400 mm2, which are
relatively underrepresented in the interglacial population.
Both these observations seem to hold true as well once the
data are split by core.
[55] These data may indicate how F. kerguelensis populations resorted more frequently to auxospore formation during
glacial (MIS2) compared to interglacial (MIS1) times, in line
with what observed during iron fertilization experiments in
the modern ocean, where auxospore abundances increase
in the fertilized patch [Assmy et al., 2006]. Furthermore, the
prevailing conditions of limited Fe availability during interglacial times may have affected this diatom’s life cycle,
resulting in the production of auxospores of smaller size,
thus constraining the population within a narrower cell size
[56] Our main hypothesis is therefore that the size increase
following iron addition (either in fertilization experiments
or glacial times) is related to an increase in the importance
of the auxospore stage in this diatom’s life cycle. This can
be represented by either an increase in their total number, as
demonstrated by Assmy et al. [2006] for iron fertilization
experiments, and/or in the auxospore area itself. For our
glacial Termination I data, the result of these processes is
represented by a shift in size spectra for the whole population
(Figure S3b).
[57] As also suggested by Assmy et al. [2006], another
biological process may also have played an important role in
shaping the population structure: selective predation of the
more fragile (compared to the vegetative cell) auxospore
stage during interglacial compared to glacial intervals. Under
this scenario, a specific predator (maybe present in higher
numbers during warmer intervals?) would have been more
efficient in removing the ca. 250 to 400 mm2 size fraction
during interglacial compared to glacial times.
[58] However, many additional complexities, particularly
behind predation and export mechanisms, suggest exerting
caution when attempting a direct interpretation of the size
structure of the populations. Such interpretation is probably
best left to a more in-depth treatment of additional data from
culturing and ecosystem modeling, which are currently being
10 of 14
Figure 7. Schematic illustration of the correlation between temperature, average valve area, and iron flux
through several glacial Terminations. Results for Terminations I, II, III, and V are represented by solid
lines, and dashed lines mark results for Termination VI.
collected by fellow researchers in an underway collaborative
study between AWI and the Zoological Station “Anton
Dohrn,” Naples.
5.3. Previous Terminations
[59] We then checked how F. kerguelensis fared at the
Antarctic Polar Front through previous glacial Terminations,
by producing records of its average valve area variability
through the last 550 ka at ODP Site 1093. Maximum average
valve areas for this diatom were indeed found under glacial
conditions, while smaller valves occurred during interglacial
times (Figure 6). During full glacial times the valve size was
about 200 mm2 on average, and therefore ca. twice as large as
under interglacial conditions. This pattern was observed at
Termination II, Termination III and Termination V, and thus
matched well the previously observed pattern for Termination I. However, the time interval around Termination VI
shows an inverse relationship between valve size and glacial/
interglacial conditions (Figures 6 and 7): F. kerguelensis
valves were larger during MIS 13 (interglacial) and smaller
during MIS 14 (glacial). The largest valve areas measured in
this study (up to 240 mm2) were found during MIS 13.
[60] In order to better understand the paleoenvironmental
significance of the observed size changes in F. kerguelensis,
we tested the link between them and several proxy record
time series, including sea ice temperature, sea ice extent,
eolian input, as well as major ecosystem shifts within diatoms. The diatom F. kerguelensis is constantly abundant
through time in the Southern Ocean, with slightly higher
relative abundances during interglacial times. In order to
test the effect of changes in iron availability and relief from
iron limitation in the Southern Ocean on the valve size of
F. kerguelensis, we compared our sediment core records for
size variability over several glacial-interglacial transitions to
a regional record of dust input and flux stored in an Antarctic
ice core. The EPICA Dome C ice core from East Antarctica
provides highly resolved records of atmospheric parameters
over last 800,000 years, including a significant increase of
iron in glacial compared to interglacial periods [Wolff et al.,
2006]. This proxy record of iron flux shows increased iron
input to the Southern Ocean during glacial conditions, and
thus potentially a strong linkage between higher valves areas
averages of F. kerguelensis during this period, and higher
dust/iron concentrations.
[61] Our valve area data for Termination VI indicate
smaller valves during full glacial and an increase to larger
valves during full interglacial conditions (i.e., a pattern
opposite to what observed at the other, later, glacial
[62] In this respect, it is interesting to note how at this
glacial Termination both the oceanic SST and atmospheric
CO2 concentration contrast between full glacial and interglacial conditions was strongly reduced compared to later
glacial Terminations. In fact, the CO2 record from the EPICA
Dome C core [Siegenthaler et al., 2005] displays the partial
pressure of atmospheric CO2 oscillating between 260 and
180 ppm before 430 ka, a range that is almost 30% smaller
than that of the last four glacial cycles. Martínez-Garcia et al.
[2009] suggest that such a change in the amplitude of the CO2
cycles was most likely driven by physical processes, and
related to changes in Antarctic sea ice extent, surface water
stratification, and westerly winds position. Changes in these
three processes will have strong consequences on both diatoms and the whole Southern Ocean ecosystem, thus signifying different boundary conditions and response to glacial
Terminations in the Southern Ocean during Termination VI
and older, compared to younger time intervals.
11 of 14
[63] Indications of substantial diatom assemblage differences are noticeable when comparing Termination VI
floral records to those representative of younger glacial
Terminations: lower relative abundances of Chaetoceros
spp. (ca. 5% versus 30%, indicating a very different diatom
assemblage and possibly lower competition for dissolved
nutrients in F. kerguelensis) and lower relative abundances of
Fragilariopsis obliquecostata and F. curta (indicating lower
summer and winter sea ice coverage). On longer timescales,
shifts in diatom assemblage composition have been demonstrated to play an important role in the geological history
of biogenic silica burial in the Southern Ocean over the past
3–4 million years [Cortese and Gersonde, 2008]. These
floral shifts strongly affected the export efficiency, and ultimately helped in establishing the Antarctic Polar Front as
the main site for burial of biogenic silica in the World Ocean,
with F. kerguelensis as the prime exporter. At the suborbital
scale of interest here, relative abundance shifts may have as
well contributed to a different functioning of the Southern
Ocean ecosystem.
[64] An additional line of evidence for an anomalous
behavior during Termination VI comes from F. kerguelensis
itself: as mentioned above, the correlation of valve size
to SST is usually negative (higher SSTs correspond to
smaller valves and vice versa), while for Termination VI
the correlation is positive (higher SSTs correspond to larger
valves). However, at this Termination, F. kerguelensis relative abundances display a peculiar pattern, as this species is
(opposite to what is commonly observed in the Southern
Ocean) more abundant during the glacial MIS 14 compared
to the interglacial MIS 13 (Figure 5). As this species is
usually a warm water indicator in the Southern Ocean and
is very abundant in this area, the observed pattern suggests
that SST is not the only control on the abundance and size of
this species at Termination VI.
[65] In our study, with the exception of Termination VI,
there is a negative correlation between valve area and relative abundances in F. kerguelensis: larger mean valve areas
are observed when the relative abundance of this species is
lower (during glacial intervals). This result is the opposite
of what is observed in a recent biometric investigation of
F. kerguelensis [Crosta, 2009]. In a study of Holocene
samples from sediment core MD03–2601 (Antarctic Continental Shelf off Adélie Land, East Antarctica), Crosta [2009]
found this species to be bigger and more abundant during the
warmer mid-Holocene period and smaller and less abundant
during the colder late Holocene period, that is, a positive
correlation between valve size and relative abundance in
F. kerguelensis. Two considerations allow reconciliation of
these seemingly conflicting lines of evidence:
[66] 1. Holocene boundary conditions are different from
those prevailing during glacial times and glacial Terminations. Therefore, the changes in size observed in these two
studies (one based on an interglacial record, the present study
looking at changes between glacial, glacial Terminations,
and interglacials) may have had different controls and
[67] 2. We concur with Crosta [2009] that the strongest
control on size in F. kerguelensis is most likely linked to
favorable environmental/paleobiological conditions for this
species in the environment where it is observed. Crosta
[2009] argues that the observed positive size/abundance
relationship is linked to the fact that the core location stands
today at the lower ecologic limit for F. kerguelensis. At the
Antarctic continental shelf location, where iron limitation
is not an issue, more adequate environmental conditions for
F. kerguelensis (warmer, less icy) allowed restoration of
bigger initial cells and overall bigger average size for the
entire populations. In our study, monitoring F. kerguelensis
in a prime iron-depleted area of the ocean (where ecological
conditions in terms of SST and sea ice are far from the limit
for this species, and are actually close to optimal), the SST
control on the valve size of this species may have played
a subordinate role compared to the availability of iron. These
different controls might have given rise to the negative
relationship observed in this study between valve size and
relative abundance at the Antarctic Polar Front, with lower
SSTs (but higher iron content) during glacial times representing “favorable conditions” for larger F. kerguelensis
valves at this location.
6. Conclusions
[68] 1. Two iron fertilization experiments suggest the
possibility that the valve area of Fragilariopsis kerguelensis
increases in water samples containing elevated iron concentrations. There is an offset of ca. 20–30 mm2 between the
in- and out-patch time series, with the in-patch populations
being consistently larger than the out-patch populations,
most likely a result of the increased number of auxospores in
the in-patch samples.
[69] 2. The study of last glacial Termination at three different locations in the Southern Ocean (core PS1654 from
the Antarctic Polar Front, cores PS2498 and PS2499 from
the Sub-Antarctic Zone) provides strong evidence for the
occurrence of larger average valve areas of F. kerguelensis
during the glacial MIS 2 compared to the interglacial MIS 1.
A synchronous drop in dust flux during Termination I
recorded in the EDC ice core opens the possibility that
decreased dust-derived Fe flux to the Southern Ocean played
a role in establishing predominantly smaller valve sizes.
[70] 3. The analysis of valve size variability at the Antarctic Polar Front (ODP Site 1093) over several past glacial
Terminations, spanning the shift from full glacial to full
interglacial conditions, revealed how average valve size
seems to be directly correlated to increased input of eolian
dust to the Southern Ocean, and inversely correlated to Sea
Surface Temperature.
[71] 4. This correlation does not seem to be valid for glacial
Termination VI, where size appears to be inversely correlated
to dust input. Anomalous patterns in species composition (for
Chaetoceros spp., F. obliquecostata, and F. curta), relative
abundances of F. kerguelensis, sea ice extent, SST, and CO2
conditions are all possible explanations for the deviation
from the expected pattern of larger valve size during glacial
compared to interglacial times.
[72] 5. SST can still prevail over iron/dust input in controlling average valve size of F. kerguelensis in environments, such as those prevailing during the Holocene close
to the Wilkes’ Land Antarctic coast, where the SST conditions are very close to the lower ecological limit for this
12 of 14
[73] 6. At this stage, the linkage between iron availability,
flux of dust to the surface ocean around Antarctica, and valve
area in F. kerguelensis (albeit very likely given the impact,
well documented in literature, Fe limitation has on diatom
size) remains hypothetical, and has yet to be confirmed by
further studies and more in-depth statistical examination of
these and similar data sets. As an example, promising inroads
in establishing causal links between other morphometric
characters (e.g., length-normalized width) and environmental
conditions have been performed by Marchetti and Cassar
[2009], based on the previously published version of this
data set [Cortese and Gersonde, 2007], including only surface sediment and Termination I data from one locality.
[74] Acknowledgments. Research was funded by the New Zealand
Foundation for Research, Science, and Technology (FRST) program Global
Change Through Time and by the Research Center Ocean Margins, Bremen,
Germany. The iron fertilization experiments samples were kindly provided
by Philipp Assmy (AWI, Bremerhaven). Laboratory assistance and slide
preparation by Ute Bock (AWI, Bremerhaven) is gratefully acknowledged.
Four anonymous reviewers are thanked for providing excellent suggestions,
which greatly improved the quality of the final manuscript.
Abelmann, A., R. Gersonde, G. Cortese, G. Kuhn, and V. Smetacek (2006),
Extensive phytoplankton blooms in the Atlantic sector of the glacial
Southern Ocean, Paleoceanography, 21, PA1013, doi:10.1029/
Anderson, R. F., S. Ali, L. I. Bradtmiller, S. H. H. Nielsen, M. Q. Fleisher,
B. E. Anderson, and L. H. Burckle (2009), Wind-driven upwelling in the
Southern Ocean and the deglacial rise in atmospheric CO2, Science,
323(5920), 1443–1448, doi:10.1126/science.1167441.
Assmy, P., J. Henjes, V. Smetacek, and M. Montresor (2006), Auxospore
formation by the silica-sinking, oceanic diatom Fragilariopsis kerguelensis
(Bacillariophyceae), J. Phycol., 42, 1002–1006, doi:10.1111/j.1529-8817.
Belkin, I. M., and A. L. Gordon (1996), Southern Ocean fronts from the
Greenwich meridian to Tasmania, J. Geophys. Res., 101(C2), 3675–3696,
Boyd, P. W., et al. (2000), A mesoscale phytoplankton bloom in the polar
Southern Ocean stimulated by iron fertilization, Nature, 407, 695–702,
Chepurnov, V. A., D. G. Mann, K. Sabbe, and W. Vyverman (2004),
Experimental studies on sexual reproduction in diatoms, Int. Rev. Cytol.,
237, 91–154, doi:10.1016/S0074-7696(04)37003-8.
Coale, K. H., et al. (1996), A massive phytoplankton bloom induced by an
ecosystem-scale iron fertilization experiment in the equatorial Pacific
Ocean, Nature, 383, 495–501, doi:10.1038/383495a0.
Cortese, G., and R. Gersonde (2007), Morphometric variability in the
diatom Fragilariopsis kerguelensis: Implications for Southern Ocean
paleoceanography, Earth Planet. Sci. Lett., 257, 526–544, doi:10.1016/
Cortese, G., and R. Gersonde (2008), Plio/Pleistocene changes in the main
biogenic silica carrier in the Southern Ocean, Atlantic Sector, Mar. Geol.,
252, 100–110, doi:10.1016/j.margeo.2008.03.015.
Crosta, X. (2009), Holocene size variations in two diatom species off East
Antarctica: Productivity vs environmental conditions, Deep Sea Res.,
Part I, 56, 1983–1993, doi:10.1016/j.dsr.2009.06.009.
de Baar, H. J. W., A. G. J. Buma, R. F. Nolting, G. C. Cadée, G. Jacques,
and P. J. Tréguer (1990), On iron limitation of the Southern Ocean:
Experimental observations in the Weddell and Scotia Seas, Mar. Ecol.
Prog. Ser., 65, 105–122, doi:10.3354/meps065105.
De La Rocha, C. L., D. A. Hutchins, M. A. Brzezinski, and Y. Zhang
(2000), Effects of iron and zinc deficiency on elemental composition
and silica production by diatoms, Mar. Ecol. Prog. Ser., 195, 71–79,
Diekmann, B., D. K. Fütterer, H. Grobe, C. D. Hillenbrand, G. Kuhn,
K. Michels, R. Petschick, and M. Pirrung (2003), Terrigenous sediment
supply in the polar to temperature South Atlantic: Land-ocean links of
environmental changes during the Late Quaternary, in The South Atlantic
in the Late Quaternary: Reconstruction of Material Budgets and Current
Systems, edited by G. Wefer et al., pp. 375–399, Springer, Berlin.
EPICA Community Members (2004), Eight glacial cycles from Antarctic
ice core, Nature, 429, 623–628, doi:10.1038/nature02599.
Falkowski, P. G., R. T. Barber, and V. Smetacek (1998), Biogeochemical
controls and feedbacks on ocean primary production, Science, 281(5374),
200–206, doi:10.1126/science.281.5374.200.
Fenner, J., H.-J. Schrader, and H. Wienigk (1976), Diatom phytoplankton
studies in the southern Pacific Ocean, composition and correlation to
the Antarctic Convergence and its paleoecological significance, Initial
Rep. Deep Sea Drill. Proj., 35, 757–813, doi:10.2973/dsdp.proc.35.
Geibert, W., et al. (2010), High productivity in an ice melting hot spot at the
eastern boundary of the Weddell Gyre, Global Biogeochem. Cycles, 24,
GB3007, doi:10.1029/2009GB003657.
Gersonde, R., and U. Zielinski (2000), The reconstruction of late
Quaternary Antarctic sea-ice distribution—The use of diatoms as a proxy
for sea-ice, Palaeogeogr. Palaeoclimatol. Palaeoecol., 162, 263–286,
Hodell, D. A., R. Gersonde, and P. Blum (2002), Leg 177 synthesis:
Insights into Southern Ocean paleoceanography on tectonic to millennial
timescales [online], Proc. Ocean Drill. Program Sci. Results, 177. [Available at]
Hudson, R. J. M., and F. M. M. Morel (1990), Iron transport in marine
phytoplankton: Kinetics of cellular and medium coordination reactions,
Limnol. Oceanogr., 35, 1002–1020, doi:10.4319/lo.1990.35.5.1002.
Hutchins, D. A., and K. W. Bruland (1998), Iron-limited diatom growth and
Si:N uptake ratios in a coastal upwelling regime, Nature, 393, 561–564,
Jouzel, J., et al. (2007), Orbital and millennial Antarctic climate variability
over the past 800,000 years, Science, 317(5839), 793–796, doi:10.1126/
Kanfoush, S. L., D. A. Hodell, C. D. Charles, T. R. Janecek, and F. R. Rack
(2002), Comparison of ice-rafted debris and physical properties in ODP
Site 1094 (South Atlantic) with the Vostok ice core over the last four climatic cycles, Palaeogeogr. Palaeoclimatol. Palaeoecol., 182, 329–349,
Lambert, F., B. Delmonte, J. R. Petit, M. Bigler, P. R. Kaufmann, M. A.
Hutterli, T. F. Stocker, U. Ruth, J. P. Steffensen, and V. Maggi (2008),
Dust-climate couplings over the past 800,000 years from the EPICA
Dome C ice core, Nature, 452, 616–619, doi:10.1038/nature06763.
Lefèvre, N., and A. J. Watson (1999), Modeling the geochemical cycle of
iron in the oceans and its impact on atmospheric CO2 concentrations,
Global Biogeochem. Cycles, 13, 727–736, doi:10.1029/1999GB900034.
Lisiecki, L. E., and M. E. Raymo (2005), A Pliocene-Pleistocene stack of
57 globally distributed benthic d 18O records, Paleoceanography, 20,
PA1003, doi:10.1029/2004PA001071.
Mahowald, N. M., D. R. Muhs, S. Levis, P. J. Rasch, M. Yoshioka, C. S.
Zender, and C. Luo (2006), Change in atmospheric mineral aerosols
in response to climate: Last glacial period, preindustrial, modern, and
doubled carbon dioxide climates, J. Geophys. Res., 111, D10202,
Maldonado, M. T., and N. M. Price (1996), Influence of N substrate on
Fe requirements of marine centric diatoms, Mar. Ecol. Prog. Ser., 141,
161–172, doi:10.3354/meps141161.
Mann, D. G. (1999), The species concept in diatoms, Phycologia, 38,
437–495, doi:10.2216/i0031-8884-38-6-437.1.
Marchetti, A., and N. Cassar (2009), Diatom elemental and morphological
changes in response to iron limitation: A brief review with potential
paleoceanographic applications, Geobiology, 7, 419–431, doi:10.1111/
Marchetti, A., and P. J. Harrison (2007), Coupled changes in the cell
morphology and the elemental (C, N and Si) composition of the pennate
diatom Pseudo-nitzschia due to iron deficiency, Limnol. Oceanogr., 52,
2270–2284, doi:10.4319/lo.2007.52.5.2270.
Martin, J. H., and S. E. Fitzwater (1988), Iron deficiency limits phytoplankton growth in the north-east Pacific subarctic, Nature, 331, 341–343,
Martin, J. H., R. M. Gordon, and S. E. Fitzwater (1990), Iron in Antarctic
waters, Nature, 345, 156–158, doi:10.1038/345156a0.
Martin, J. H., et al. (1994), Testing the iron hypothesis in ecosystems of the
equatorial Pacific Ocean, Nature, 371, 123–129, doi:10.1038/371123a0.
Martínez-Garcia, A., A. Rosell-Melé, W. Geibert, R. Gersonde, P. Masqué,
V. Gaspari, and C. Barbante (2009), Links between iron supply, marine
productivity, sea surface temperature, and CO2 over the last 1.1 Ma,
Paleoceanography, 24, PA1207, doi:10.1029/2008PA001657.
Nielsen, S. H. H., D. A. Hodell, G. Kamenov, T. Guilderson, and M. R.
Perfit (2007), Origin and significance of ice-rafted detritus in the Atlantic
sector of the Southern Ocean, Geochem. Geophys. Geosyst., 8, Q12005,
Pahlow, M., U. Riebesell, and D. A. Wolf-Gladrow (1997), Impact of cell
shape and chain formation on nutrient acquisition by marine diatoms,
Limnol. Oceanogr., 42, 1660–1672, doi:10.4319/lo.1997.42.8.1660.
13 of 14
Paillard, D., L. Labeyrie, and P. Yiou (1996), Macintosh program performs
time-series analysis, Eos Trans. AGU, 77(39), 379, doi:10.1029/
Parrenin, F., et al. (2007), The EDC3 chronology for the EPICA Dome C
ice core, Clim. Past, 3, 485–497, doi:10.5194/cp-3-485-2007.
Raiswell, R., M. Tranter, L. G. Benning, M. Siegert, R. De’ath,
P. Huybrechts, and T. Payne (2006), Contributions from glacially derived
sediment to the global iron (oxyhydr)oxide cycle: Implications for iron
delivery to the oceans, Geochim. Cosmochim. Acta, 70, 2765–2780,
Ridgwell, A. J., and A. J. Watson (2002), Feedback between aeolian dust,
climate, and atmospheric CO2 in glacial time, Paleoceanography,
17(4), 1059, doi:10.1029/2001PA000729.
Sarmiento, J. L., N. Gruber, M. A. Brzezinski, and J. P. Dunne (2004),
High-latitude controls of thermocline nutrients and low latitude biological
productivity, Nature, 427, 56–60, doi:10.1038/nature02127.
Schneider-Mor, A., R. Yam, C. Bianchi, M. Kunz-Pirrung, R. Gersonde,
and A. Shemesh (2008), Nutrient regime at the siliceous belt of the Atlantic sector of the Southern Ocean during the past 660 ka, Paleoceanography, 23, PA3217, doi:10.1029/2007PA001466.
Siegenthaler, U., et al. (2005), Stable carbon cycle–climate relationship
during the late Pleistocene, Science, 310(5752), 1313–1317, doi:10.1126/
Sigman, D. M., M. P. Hain, and G. H. Haug (2010), The polar ocean and
glacial cycles in atmospheric CO2 concentration, Nature, 466, 47–55,
Smetacek, V. (2001), EisenEx: International team conducts iron experiment
in Southern Ocean, U.S. JGOFS News, 11(1), 14.
Smetacek, V., U. Bathmann, and E. Helmke (Eds.) (2005), The expeditions
ANTARKTIS XXI/3-4-5 of the research vessel Polarstern in 2004, Ber.
Polarforsch. Meeresforsch. 500, 302 pp., Alfred Wegener Inst. for Polar
and Mar. Res., Bremerhaven, Germany, hdl:10013/epic.10505.d001.
Smith, K. L., Jr., B. H. Robison, J. J. Helly, R. S. Kaufmann, H. A. Ruhl,
T. J. Shaw, B. S. Twining, and M. Vernet (2007), Free-drifting icebergs:
Hot spots of chemical and biological enrichment in the Weddell Sea,
Science, 317(5837), 478–482 doi:10.1126/science.1142834.
Takeda, S. (1998), Influence of iron availability on nutrient consumption
ratio of diatoms in oceanic waters, Nature, 393, 774–777, doi:10.1038/
Timmermans, K. R., B. van der Wagt, and H. J. W. de Baar (2004), Growth
rates, half saturation constants, and silicate, nitrate, and phosphate depletion in relation to iron availability of four large, open-ocean diatoms from
the Southern Ocean, Limnol. Oceanogr., 49, 2141–2151, doi:10.4319/
Wolff, E. W., et al. (2006), Southern Ocean sea-ice extent, productivity and
iron flux over the past eight glacial cycles, Nature, 440, 491–496,
G. Cortese, Department of Paleontology, GNS Science, 1 Fairway Dr.,
PO Box 30 368, Lower Hutt 5040, New Zealand. ([email protected])
R. Gersonde, Geosciences Division, Alfred Wegener Institute for Polar
and Marine Research, Columbusstrasse, PO Box 120161, Bremerhaven
D-27515, Germany. ([email protected])
K. Maschner, German Oceanographic Museum, Katharinenberg 14-60,
Stralsund D-18439, Germany. ([email protected])
P. Medley, Cooperative Institute for Research in Environmental Sciences,
University of Colorado at Boulder, 216 UCB, Boulder, CO 80309-0216,
USA. ([email protected])
14 of 14
Was this manual useful for you? yes no
Thank you for your participation!

* Your assessment is very important for improving the work of artificial intelligence, which forms the content of this project

Download PDF