LETTERS
PUBLISHED ONLINE: 20 SEPTEMBER 2009 | DOI: 10.1038/NGEO639
A microbial source of phosphonates in
oligotrophic marine systems
Sonya T. Dyhrman1 *, Claudia R. Benitez-Nelson2 , Elizabeth D. Orchard1 , Sheean T. Haley1
and Perry J. Pellechia3
Phosphonates, compounds with a carbon–phosphorus bond,
are a key component of the marine dissolved organic phosphorus pool1 . These compounds serve as a phosphorus source
for primary producers, including the nitrogen-fixing cyanobacteria Trichodesmium2 . Phosphonates can therefore support
marine primary production, as well as climate-driven increases in marine nitrogen fixation3 , carbon sequestration4
and possibly methane production, through the breakdown of
methylphosphonate5 . Despite their importance, the source of
phosphonates to the open ocean has remained uncertain. Here,
we use solid-state nuclear magnetic resonance spectroscopy to
screen for the presence of phosphonates in cultured strains of
Trichodesmium erythraeum. We show that phosphonates comprise an average of 10% of the cellular particulate phosphorus
pool in this species. We therefore suggest that these cyanobacteria produce phosphonates, and might be a significant source
of these compounds in the ocean, particularly in nutrientpoor regions, where Trichodesmium blooms occur. Given that
Trichodesmium also thrives in a warm, carbon-dioxide-rich
environment3 , phosphonate production may increase in the future. This, in turn, might select for a microbial community that
can use phosphonate, and could have implications for nitrogen
fixation, carbon sequestration and greenhouse-gas production.
Phosphorus (P) has a key role in constraining the growth
of marine primary producers over both modern and geologic
timescales6,7 . In many regions of the ocean, standing stocks of
dissolved inorganic phosphorus are so low that organically bound
P dominates the dissolved P reservoir. With future P limitation scenarios predicted from natural (for example, increased N2 fixation)
and anthropogenic (for example, increased atmospheric nitrogen
deposition) responses to climate change3,4,7 , microbial community
structure, oceanic carbon export, and hence the oceanic uptake of
atmospheric CO2 , may be controlled by dissolved organic phosphorus (DOP) concentration and composition5,8 . Furthermore, the
presence and microbial degradation of methylphosphonate in the
upper water column has been suggested to result in methane release
to the atmosphere5 . However, these hypotheses, and the ability to
model climate-driven changes in oceanic biogeochemical cycles, are
limited by the lack of information regarding the chemical composition, production and degradation of the present-day DOP pool.
DOP exists as two main bond classes, phosphoester (P–O–C
bond) and phosphonate (P–C bond). 31 P NMR analysis of highmolecular-weight (HMW) DOP (the only fraction concentrated
enough for study) has shown that both bond classes are a
significant and constant percentage (75% phosphoester and 25%
phosphonate) of marine DOP (ref. 1). The phosphoester pool
comprises nucleotides, sugars and other biomolecules that are
thought to be rapidly produced and consumed by marine
microbes. The ubiquitous distribution of phosphonates within
marine systems1 also reflects a dynamic balance between sources
and sinks, but the processes controlling the consumption and
production of phosphonates are largely unknown. It is only with
the sequencing of marine microbial genomes and environmental
samples that this bond class was recognized as bioavailable to
marine cyanobacteria2,5,9 . These studies suggest that the ability
to degrade phosphonates is relatively widespread, and that
phosphonate is consumed in the upper water column to support
both carbon and N2 fixation2 .
The maintenance of a relatively high standing stock of phosphonates in the presence of utilization suggests that there is a
significant, yet unidentified, phosphonate source that contributes
to upper-water-column DOP in the open ocean. Phosphonate compounds are known to be present in benthic marine invertebrates10,11 ,
and have been detected at low levels (<3.0%) in a planktonic
amphipod12 . Phosphonates are also present in some organic pesticides and xenobiotics13 . Although it may be possible that invertebrates or terrestrial runoff could influence dissolved phosphonate
concentrations in certain cases, such as near-shore marine environments, neither of these potential sources adequately explains
the dominance of the phosphonate bond class in the DOP of the
ocean gyres and in about 2,000-year-old deep waters1,14,15 . It is
widely accepted that microbe-derived phosphoesters (for example,
sugars) and organic nitrogen compounds (for example, amino
acids) contribute to the DOP and dissolved organic nitrogen pools,
respectively16,17 . Indeed, the phosphonate bond has been detected
in marine particulate organic matter12,18,19 and microbes could be a
source of phosphonates to the ocean gyres. However to the best of
our knowledge, this bond class has never been found to comprise
a substantial proportion of the total P in any marine microbe
examined until this study12,15 (Table 1).
Using solid-state 31 P NMR spectroscopy, a non-destructive
method for identifying the dominant classes of P bonds, the
marine N2 -fixing cyanobacteria T. erythraeum, T. theibautii,
T. tenue and Crocosphaera watsonii were screened for the presence
of phosphonates. Eukaryotic phytoplankton species, including
a diatom and a coccolithophore were also tested (Table 1).
Solid-state 31 P NMR spectroscopy is advantageous in that it
avoids extraction issues (for example, compound breakdown and
alteration) associated with other methods19 . No phosphonates
were detectable in the N2 -fixing cyanobacteria T. theibautii,
T. tenue or C. watsonii, or in the eukaryotes, consistent with
previous work15 (Table 1, Fig. 1). In contrast, the phosphonate
1 Biology
Department, Woods Hole Oceanographic Institution, Woods Hole, Massachusetts 02543, USA, 2 Department of Earth and Ocean Sciences &
Marine Science Program, University of South Carolina, Columbia, South Carolina 29208, USA, 3 Department of Chemistry and Biochemistry, University of
South Carolina, Columbia, South Carolina 29208, USA. *e-mail: [email protected].
NATURE GEOSCIENCE | ADVANCE ONLINE PUBLICATION | www.nature.com/naturegeoscience
© 2009 Macmillan Publishers Limited. All rights reserved.
1
NATURE GEOSCIENCE DOI: 10.1038/NGEO639
LETTERS
Table 1 | The average percentage of total phosphorus as the phosphonate bond class in cultures of marine phytoplankton.
Organism
Strain
P (µM)*
Phosphonate (% of total
particulate phosphorus)
Amphidinium carteri
Coccolithus huxleyi
Crocosphaera watsonii
Crocosphaera watsonii
Emiliania huxleyi
Emiliania huxleyi
Emiliania huxleyi
Peridinium trochoidem
Phaeocystis sp.
Skeletonema costatum
Synechococcus bacillaris
Syracosphaera elongota
Thalassiosira weissflogii
Thalassiosira weissflogii
Trichodesmium erythraeum
Trichodesmium erythraeum
Trichodesmium erythraeum
Trichodesmium erythraeum
Trichodesmium tenue
Trichodesmium tenue
Trichodesmium theibautii
Trichodesmium theibautii
—
—
WH8501
WH8501
CCMP 374
CCMP 374
CCMP 372
—
CCMP 627
CCMP 775
CCMP 1333
—
CCMP 1336
CCMP 1336
IMS101
IMS101
ST6-5
ST6-5
Tenue
Tenue
II-3
II-3
2
2
45
0
36
1
3.6
2
3.6
3.6
3.6
2
36
0
15
0
15
0
15
0
15
0
0.01%
<0.01%
N.D.
N.D.
N.D.
N.D.
N.D.
<0.01%
N.D.
N.D.
N.D.
0.004%
N.D.
N.D.
10%
11%
8%
17%
N.D.
N.D.
N.D.
N.D.
Reference
†
Ref. 12
Ref. 12
This study
This study
This study
This study
Ref. 15
Ref. 12
Ref. 15
Ref. 15
Ref. 15
Ref. 12
This study
This study
This study
This study
This study
This study
This study
This study
This study
This study
*Phosphate concentration estimated based on culture media at the start of the experiment.
†
Average per cent phosphonate of cultures sampled at single time points; averages are from the 31 P NMR data for two or more independent experiments, with the exception of the ST6-5 0 µM P treatment,
which was not replicated. A previous 31 P NMR study15 used a similar approach. Error on the 31 P NMR integrations of phosphonate is roughly ±3%. The presence of trace (<0.01%) phosphonate in
undisclosed strains of coccolithophores and dinoflagellates was reported using an alternative to the 31 P NMR method12 . N.D.: not detectable.
bond class was detected in cultures of T. erythraeum, with average
concentrations from replicate single-time-point experiments
ranging from 10–17% (Table 1, Fig. 1), depending on the strain
and its physiology. The phylogenetic relationship deduced from
the internal transcribed spacer region of the ribosomal RNA gene
suggests that T. erythraeum occupies a clade that is unique from
the other Trichodesmium species20 , and the genetic machinery for
phosphonate biosynthesis may be associated only with this clade.
To our knowledge, this is the first report of the presence of the
phosphonate bond class in cultures of an abundant species of
marine cyanobacteria.
Phosphonates were detected in two strains of T. erythraeum,
isolated from the North Carolina coast (IMS101) and the oligotrophic northern Sargasso Sea (ST6-5) (Table 1). This suggests
that the presence of phosphonates in these strains is not an
artefact of the regime from which they were isolated. Although
T. erythraeum has genes for the transport and metabolism of phosphonate compounds2 , mere surface adsorption or accumulation of
phosphonate compounds does not explain the high percentages of
the phosphonate bond class detected in the cultures. First, the genes
for phosphonate transport are upregulated under low-P conditions2
and here similar percentages of phosphonate compounds were
detected in both low-P and P-replete cultures (Table 1). Second,
mass balance calculations confirm that the phosphonate concentration in the sea water used in the culture studies is over four
orders of magnitude too low to explain the measured concentrations of T. erythraeum phosphonate if accumulation or surface
adsorption were the only mechanisms of incorporation (see Supplementary Information). Trichodesmium is difficult to maintain
axenically, but an examination of the heterotrophic bacteria in
the T. erythraeum cultures indicated that heterotrophic contamination was minimal (∼100–1,000 cells ml−1 ). Heterotrophic bacteria were routinely separated from the Trichodesmium cultures by
2
collecting the Trichodesmium cells onto 5 µm filters, and collecting
heterotrophic bacteria in the filtrate onto 0.2-µm filters. Biomass
on the 0.2-µm filters never had detectable phosphonates (data not
shown). Although this procedure does not control for the presence
of heterotrophic bacteria tightly attached to the Trichodesmium
trichomes, the total particulate P on the 5-µm filters used for
this study is heavily dominated by Trichodesmium. Given these
results, it seems that T. erythraeum is the source of the phosphonates detected herein.
The total phosphonate content in T. erythraeum IMS101
cultures increases as a function of increasing total particulate P
(R2 = 0.60, p < 0.0007), with an average slope resulting in a 10%
allocation of total particulate P to phosphonate (Fig. 2). That
phosphonates are a constant percentage of the total particulate P in
culture is consistent with previous findings that these compounds
consistently comprise at least 6% of total DOP (see Supplementary
Information). The phosphonate bond can be present in a diverse
set of biomolecules, including lipids, proteins and antibiotics13 .
Recent work with T. erythraeum IMS101 has shown that it will
decrease the production of P-containing lipids under P stress, when
the particulate P concentration is low21 . Although phosphonatecontaining lipids were not explicitly examined in this study, the
decline in P allocation to phospholipids during P stress21 suggests
that the phosphonates detected herein are not primarily associated
with lipids. A lack of a relationship between phosphonates and
lipids was previously found for marine HMW dissolved organic
matter14 . The chemical composition of marine phosphonates
in DOP is unknown, because the compounds are difficult to
isolate from marine organic matter at a concentration appropriate
for characterization. As such, the finding that T. erythraeum
produces roughly 10% of its cellular particulate P as phosphonate
provides an important model system for characterizing the chemical
composition of marine phosphonates produced in situ.
NATURE GEOSCIENCE | ADVANCE ONLINE PUBLICATION | www.nature.com/naturegeoscience
© 2009 Macmillan Publishers Limited. All rights reserved.
NATURE GEOSCIENCE DOI: 10.1038/NGEO639
a
LETTERS
1,250
Phosphates
Total phosphonate (nmol)
Phosphonates
b
Phosphoesters
c
1,000
750
500
250
0
Polyphosphates
0
2,500
5,000
7,500
Total phosphorus (nmol)
10,000
Figure 2 | A plot of particulate phosphonate and particulate phosphorus
content in cultures of Trichodesmium erythraeum IMS101. Total
particulate phosphonate significantly (R2 = 0.60, p < 0.0007) increases as
a function of total particulate phosphorus in cultures of Trichodesmium
erythraeum IMS101. The slope of the linear regression of the data (solid
line) results in an average of 10% phosphonate. The 95% confidence
intervals (dashed lines) indicate a range of phosphonate from 5 to 15% of
the total particulate phosphorus.
d
100
80
60
40
20
0
ppm
–20
–40
–60
–80
–100
Figure 1 | 31 P NMR spectra from cultures and field populations of
N2 -fixing cyanobacteria denoting phosphorus (P) bond classes. a–d, A
representative spectrum from Trichodesmium erythraeum IMS101 indicates
the presence of phosphonate in this species (a), but a spectrum from
Crocosphaera watsonii WH8501 (b), a spectrum from a culture of
Trichodesmium theibautii II-3 (c) and a spectrum from a sample of
Trichodesmium spp. collected from the western North Atlantic (d) are
missing this peak and thus lack detectable phosphonate. The arrows
indicate peaks for phosphonate, phosphate, phosphoester and
polyphosphate as marked.
Trichodesmium is broadly distributed in the tropical and
sub-tropical oceans, often accounting for significant carbon and
N2 fixation in these regions22 . Trichodesmium spp. colonies were
collected from surface waters in the western North Atlantic and
assayed with solid-state 31 P NMR spectroscopy. No phosphonates
were detected in this sample (Fig. 1). However, the raft-like
colony morphology often associated with T. erythraeum was
not dominant along the cruise transect. Rather, the colony
morphology seemed consistent with T. theibautii, and the 31 P NMR
spectrum indicated a P partitioning pattern similar to that
seen in T. theibautii cultures (Fig. 1). Therefore, the lack of
phosphonates in this field sample is consistent with the clade
differences observed in culture. A recent transatlantic survey
of Trichodesmium abundance found the raft-like T. erythraeum
colony morphology to be ∼60 colony m−3 (ref. 23) in non-bloom
scenarios. Extrapolating from these abundance numbers, the
average P quota (minimum of 2.20 nmol P colony−1 ) of colonies
in the western North Atlantic24 , and an average phosphonate
content of 10%, as derived from this study (Fig. 2), we estimate this
species produces a standing stock of ∼13 nmol m−3 of particulate
phosphonate (see Supplementary Information). With an estimated
doubling time of 2 days25 , field populations of T. erythraeum
could produce this amount of phosphonate in roughly the
same 2-day time frame. Although this rate of phosphonate
biosynthesis is significant, it would take ∼110 to 850 days for
T. erythraeum to produce enough particulate P to support observed
concentrations of phosphonates in HMW dissolved organic matter
from the North Atlantic (see Supplementary Information). These
estimates suggest that although T. erythraeum is a potentially
substantial source of phosphonates to the upper water column
of oligotrophic regions even in non-bloom conditions, this
species alone does not support the ubiquitous distribution of
dissolved phosphonates observed throughout the open ocean.
More detailed field analyses from a broader range of samples
dominated by the T. erythraeum clade would help to elucidate
more accurate phosphonate biosynthesis rates in the field, and
screening of further marine microbes may identify more groups that
produce phosphonates.
To the best of our knowledge, the data presented herein are
the first that identify a significant source of phosphonate associated with marine cyanobacteria. T. erythraeum is widespread,
and as such, a potentially significant microbial source of phosphonate to upper-water-column DOP. Blooms of this species in
the low-P ocean gyres could increase phosphonate concentrations within DOP, selecting for a microbial community including
picocyanobacteria9 and Trichodesmium species2 , which can use
phosphonate as a P source. This could explain shifts in phytoplankton community structure observed over the past decade in
regions such as the North Pacific26 , and predicted in the near
future4 . Reduced mixing of deep-water nutrients with nutrientpoor surface waters is expected to occur as a result of surface
water warming. This scenario could enhance nutrient limitation,
particularly P limitation, in the future ocean and drive the need for
phosphonate utilization as a P source. If T. erythraeum produces
methylphosphonate, this scenario may further augment methane
production through the in situ breakdown of methylphosphonate by marine microbes5 , and thereby create a positive feedback loop between increasing temperatures, strengthening water column stratification, selection for a community that will
degrade phosphonate and methane production5,27 . Whether or
not these positive feedback mechanisms can be balanced by carbon sequestration enhanced by increases in bioavailable nitrogen from N2 fixation4 , which would facilitate atmospheric CO2
drawdown, requires more detailed marine biogeochemical models
that parameterize Trichodesmium as both a sink and a source of
marine phosphonate.
NATURE GEOSCIENCE | ADVANCE ONLINE PUBLICATION | www.nature.com/naturegeoscience
© 2009 Macmillan Publishers Limited. All rights reserved.
3
NATURE GEOSCIENCE DOI: 10.1038/NGEO639
LETTERS
Methods
Cultures. Cyanobacterial cultures were provided by J. Waterbury of the Woods
Hole Oceanographic Institution, and eukaryotic algal cultures were obtained
from the Provasoli Guillard Center for the Culture of Marine Phytoplankton
(see Table 1). Uni-algal but non-axenic Trichodesmium cultures were grown in
medium made from a Sargasso Sea water base as described elsewhere28 . Further
tests included cultures grown in medium with a Vineyard Sound seawater base,
×2 concentrated vitamins and with or without added P. The other phytoplankton
cultures were grown with or without P under conditions described elsewhere21 .
Growth was monitored using a Turner Designs fluorometer to assess relative
pigment fluorescence as a proxy for biomass. Culture biomass was collected onto
0.8–5.0-µm polycarbonate filters, depending on the cell size of the test culture,
for 31 P NMR spectroscopy. In addition, T. erythraeum strain IMS101 and ST6-5
cultures were also 4,6-diamidino-2-phenylindole (DAPI) stained29 , and collected
onto both 0.2-µm (filtrate) and 5.0-µm (T. erythraeum) filters for counting
heterotrophic bacteria. In these experiments, the filtrate collected onto the 0.2 µm
filter (heterotrophic bacteria) was also assayed for 31 P NMR spectroscopy.
Field sampling. Trichodesmium colonies were collected using a 130 µm net tow
within roughly the top 20 m of the water column sampled on a cruise in the
western North Atlantic (station coordinates: 23.66 N, 65.207 W). Colonies of
Trichodesmium were quickly picked by hand and washed several times in 0.2-µm
filter-sterile local sea water. Approximately 100 colonies were then collected onto
5.0-µm polycarbonate filters for 31 P NMR spectroscopy.
NMR. Filters were dried at 65 ◦ C before analysis, according to the conditions
empirically validated in other work19 . 31 P spectra were recorded on a Varian Inova
500 spectrometer operating at 202.489 MHz using a Doty Scientific 4 mm/XC
magic-angle spinning probe. Bloch decays of 50 ms were collected with a 200 ppm
window after 30◦ excitation pulses. Two-pulse phase modulation30 proton dipolar
decoupling with a field strength of 45 kHz was applied during acquisition, and
a magic-angle spinning speed of 10 kHz was used. Spectra were fitted with five
Lorentzian lines to determine the relative ratio of the species (phosphonate:
18 ppm, phosphate: 0 ppm, broad phosphoester: −6 ppm, sharp phosphoester:
−12 ppm, polyphosphate: −23 ppm).
Total phosphorus. T. erythraeum cultures were collected onto pre-combusted
glass-fibre filters and stored dried at 65 ◦ C in pre-combusted foil until analysis for
particulate organic and inorganic P using previously described approaches19 .
Received 6 July 2009; accepted 26 August 2009; published online
20 September 2009
References
1. Clark, L. L., Ingall, E. D. & Benner, R. Marine phosphorus is selectively
remineralized. Nature 393, 426 (1998).
2. Dyhrman, S. T. et al. Phosphonate utilization by the globally important marine
diazotroph Trichodesmium. Nature 439, 68–71 (2006).
3. Hutchins, D. A. et al. CO2 control of Trichodesmium N2 fixation,
photosynthesis, growth rates, and elemental ratios: Implications for
past, present, and future ocean biogeochemistry. Limnol. Oceanogr. 52,
1293–1304 (2007).
4. Karl, D. M. & Letelier, R. M. Nitrogen fixation-enhanced carbon sequestration
in low nitrate, low chlorophyll seascapes. Mar. Ecol. Prog. Ser. 364,
257–268 (2008).
5. Karl, D. M. et al. Aerobic production of methane in the sea. Nature Geosci. 1,
473–478 (2008).
6. Alego, T. J. & Ingall, E. D. Sedimentary Corg :P ratios, paleocean ventilation
and phanerozoic atmospheric pO2. Paloegeogr. Palaeoclimatol. Palaeoecol. 256,
130–155 (2007).
7. Wu, J., Sunda, W. G., Boyle, E. A. & Karl, D. M. Phosphate depletion in the
western North Atlantic Ocean. Science 289, 759–762 (2000).
8. Mather, R. L. et al. Phosphorus cycling in the North and South Atlantic Ocean
subtropical gyres. Nature Geosci. 1, 439–443 (2008).
9. Ilikchyan, I. N. et al. Detection and expression of the phosphonate transporter
gene phnD in marine and freshwater picocyanobacteria. Environ. Microbiol. 11,
1314–1324 (2009).
10. Kittredge, J. S. & Roberts, E. A. A carbon–phosphorus bond in nature. Science
164, 37–42 (1969).
4
11. Quin, L. D. The presence of compounds with a carbon–phosphorus bond in
some marine invertebrates. Biochemistry 4, 324–330 (1965).
12. Kittredge, J. S., Horiguchi, M. & Williams, P. M. Aminophosphonic
acids: Biosynthesis by marine phytoplankton. Comp. Biochem. Physiol. 29,
859–863 (1969).
13. Kononova, S. V. & Nesmeyanova, M. A. Phosphonates and their degradation
by microorganisms. Biochemistry (Moscow) 67, 184–195 (2002).
14. Sannigrahi, P., Ingall, E. D. & Benner, R. Nature and dynamics of
phosphorus-containing components of marine dissolved and particulate
organic matter. Geochim. Cosmochim. Acta 70, 5868–5882 (2006).
15. Clark, L. L., Ingall, E. D. & Benner, R. Marine organic phosphorus
cycling: Novel insights from nuclear magnetic resonance. Am. J. Sci. 299,
724–737 (1999).
16. Antia, N. J., Harrison, P. J. & Oliveira, L. The role of dissolved organic
nitrogen in phytoplankton nutrition, cell biology and ecology. Phycologia 30,
1–89 (1991).
17. Karl, D. M. & Björkman, K. M. in Biogeochemistry of Marine Dissolved Organic
Matter (eds Hansell, D. & Carlson, C.) 249–366 (Elsevier Science, 2002).
18. Paytan, A., Cade-Menun, B., McLaughlin, K. & Faul, K. Selective phosphorus
regeneration of sinking marine particles: Evidence from 31 P-NMR. Mar. Chem.
82, 55–70 (2003).
19. Benitez-Nelson, C. R. et al. Phosphonates and particulate organic phosphorus
cycling in an anoxic marine basin. Limnol. Oceanogr. 49, 1593–1604 (2004).
20. Orcutt, K. M. et al. Characterization of Trichodesmium spp. by genetic
techniques. Appl. Environ. Microbiol. 68, 2236–2245 (2002).
21. Van Mooy, B. A. S. et al. Phytoplankton in the ocean use non-phosphorus
lipids in response to phosphorus scarcity. Nature 458, 69–72 (2009).
22. Capone, D. G. et al. Trichodesmium, a globally significant marine
cyanobacterium. Science 276, 1221–1229 (1997).
23. Davis, C. S. & McGillicuddy, D. J. Transatlantic abundance of the N2 -fixing
colonial cyanobacterium Trichodesmium. Science 312, 1517–1520 (2006).
24. Sañudo-Wilhelmy, S. A. et al. Phosphorus limitation of nitrogen fixation by
Trichodesmium in the central Atlantic Ocean. Science 411, 66–69 (2001).
25. Orcutt, K. M. et al. A seasonal study of the significance of N2 fixation by
Trichodesmium spp. at the Bermuda Atlantic Time-series Study (BATS) site.
Deep Sea Res. 48, 1583–1608 (2001).
26. Karl, D. M., Bidigare, R. R. & Letelier, R. M. in Phytoplankton Productivity and
Carbon Assimilation in Marine and Freshwater Ecosystems
(eds Williams, P. J. le B., Thomas, D. R. & Reynolds, C. S.) 222–264
(Blackwell, 2002).
27. Ingall, E. D. Oceanography: Making methane. Nature Geosci. 1, 419–420 (2008).
28. Webb, E. A., Moffett, J. W. & Waterbury, J. B. Iron stress in open-ocean
cyanobacteria (Synechococcus, Trichodesmium, and Crocosphaera
spp.): Identification of the IdiA protein. Appl. Environ. Microbiol. 67,
5444–5452 (2001).
29. Sherr, B. F., Sherr, E. B. & del Georgio, P. in Methods in Microbiology
(ed. Paul, J. H.) 129–159 (Academic, 2001).
30. McGeorge, G., Alderman, D. W. & Grant, D. M. Resolution enhancement
in 13 C and 15 N magic angle turning experiments with TPPM decoupling.
J. Magn. Reson. 137, 138–143 (1999).
Acknowledgements
This work was supported by the National Science Foundation Biological and Chemical
Oceanography Programs, the Center for Microbial Oceanography: Research and
Education, and the Woods Hole Oceanographic Institution. The authors thank
B. A. S. Van Mooy and E. Ingall for their helpful discussions, J. Waterbury for access to
cultures and the participants of the ATP-3 project for assistance at sea.
Author contributions
S.T.D. and C.R.B.-N. conceived of the study, processed samples and wrote the
manuscript. E.D.O. and S.T.H. carried out the culture studies and P.J.P. did the
31
P NMR spectroscopy.
Additional information
Supplementary information accompanies this paper on www.nature.com/naturegeoscience.
Reprints and permissions information is available online at http://npg.nature.com/
reprintsandpermissions. Correspondence and requests for materials should be
addressed to S.T.D.
NATURE GEOSCIENCE | ADVANCE ONLINE PUBLICATION | www.nature.com/naturegeoscience
© 2009 Macmillan Publishers Limited. All rights reserved.