SEE COMMENTARY

A simple nutrient-dependence mechanism for

predicting the stoichiometry of marine ecosystems
Eric D. Galbraitha,1 and Adam C. Martinyb,c
a
Department of Earth and Planetary Science, McGill University, Montreal, QC, Canada H3A 2A7; bDepartment of Ecology and Evolutionary Biology,
University of California, Irvine, CA 92697; and cDepartment of Earth System Science, University of California, Irvine, CA 92697

Edited by François M. M. Morel, Princeton University, Princeton, NJ, and approved May 11, 2015 (received for review December 13, 2014)

It is widely recognized that the stoichiometry of nutrient elements the recent Coupled Model Intercomparison Project, CMIP5 (13).
in phytoplankton varies within the ocean. However, there are Thus, the potential impact of changes in organic matter stoichi-
many conflicting mechanistic explanations for this variability, and it is ometry on ocean carbon storage and oxygen consumption remain
often ignored in global biogeochemical models and carbon cycle largely unexplored. The neglect of stoichiometric variability is due,
simulations. Here we show that globally distributed particulate P:C at least in part, to the lack of a clear predictive framework.
varies as a linear function of ambient phosphate concentrations, Here, it is argued that the concentration of a nutrient element
whereas the N:C varies with ambient nitrate concentrations, but only in seawater can provide a suitable predictive framework, because
when nitrate is most scarce. This observation is consistent with the it is a critical determinant of the rate at which that element will
adjustment of the phytoplankton community to local nutrient avail- tend to be taken up by the organisms in the local community. This
ability, with greater flexibility of phytoplankton P:C because P is a hypothesis builds on classic resource competition theory (14),
less abundant cellular component than N. This simple relationship is which argues that if the concentration of an element is low, such
shown to predict the large-scale, long-term average composition of that uptake is difficult, the community will be dominated by or-
surface particles throughout large parts of the ocean remarkably ganisms that are well adapted to a low cellular quota of that nu-
well. The relationship implies that most of the observed variation trient (10). If, on the other hand, the concentration is high,
in N:P actually arises from a greater plasticity in the cellular P:C con- facilitating high uptake rates, the community will be dominated by
tent, relative to N:C, such that as overall macronutrient concentra- organisms that are capable of taking advantage of that nutrient to
tions decrease, N:P rises. Although other mechanisms are certainly grow faster. This suggestion leads to clear predictions with sig-
also relevant, this simple relationship can be applied as a first-order nificant biogeochemical consequences, as outlined below.
basis for predicting organic matter stoichiometry in large-scale bio-
geochemical models, as illustrated using a simple box model. The Observed Patterns
results show that including variable P:C makes atmospheric CO2 more The cellular abundances of N and P are most often discussed in
sensitive to changes in low latitude export and ocean circulation than terms of the N:P ratio. However, carbon is by far the dominant
a fixed-stoichiometry model. In addition, variable P:C weakens the element in biomass (not counting H and O, derived from water).
relationship between preformed phosphate and atmospheric CO2 It typically outweighs nitrogen by a factor of about seven, be-
while implying a more important role for the nitrogen cycle. cause of its central structural role in most organic molecules, and
is thus the most logical normalizing factor for considering nu-
marine ecosystem | nutrients | stoichiometry | phytoplankton | trient ratios. Relative to carbon, the ∼30 essential nutrient ele-
carbon dioxide
ments display a range of plasticities in their cellular requirements
(15). Nitrogen tends to be the least plastic, varying in cellular

N utrient elements are used by phytoplankton to synthesize

molecules, in order to accomplish biochemical functions.
Some of these molecules are absolutely necessary, and the nu-
N:C molar ratios from ∼1/5 to 1/10, a factor of two (2). The

Significance
trient elements have no substitutes. Examples are P in nucleic
acids, N in amino acids, and Fe in the photosynthetic apparatus
The elemental ratios of nitrogen, phosphorus, and carbon in ma-
(1). However, there is a degree of plasticity in the molecular as- rine phytoplankton can diverge significantly from the “Redfield

ENVIRONMENTAL
semblage required per phytoplankton cell, which varies between ratio,” but the underlying reasons have been hard to elucidate. As

SCIENCES
species and between clades (2, 3). Furthermore, there is a capacity a result, global biogeochemical models often ignore this stoi-
for plasticity in molecular composition of even a given species, as chiometric variability. Here we show that, hidden within the
shown in culture experiments (4, 5). Such plasticity leads to vari- noise of a large dataset of particulate measurements, a sur-
ability in the elemental ratios of nutrients in marine phytoplank- prisingly consistent relationship exists between community
ton, widely documented in laboratory and field measurements (2, P:C and dissolved phosphate concentrations. The plasticity of
6, 7). Recent analyses of global nutrient and particulate observa- ecosystem stoichiometry in the face of nutrient scarcity, with
tions have shown that N:P, the most commonly discussed ratio, greater plasticity for P relative to N, appears to explain the
varies regionally, including low N:P in the high-latitude Southern main divergences from the Redfield ratio. When included in a
Ocean and high N:P in the oligotrophic regions (7–9). Explana- simple model, the relationship implies a more important role
tions of high N:P in oligotrophic waters have often invoked an for low latitude nutrient cycling in the biological pump than is
enhanced reliance on N-rich proteins for gathering scarce re- commonly assumed.
sources (1, 10), whereas low N:P in the Southern Ocean has been
variously attributed to the abundance of P-rich molecules in cold, Author contributions: E.D.G. designed and performed the data analysis and model sim-
fast-growing plankton (11), or to the availability of Si, which ulations; and E.D.G. and A.C.M. wrote the paper.

supports P-rich diatom communities (8, 12). The authors declare no conflict of interest.
Despite an abundant literature on stoichiometric variability This article is a PNAS Direct Submission.
and its potential causes, no simple predictive relationship has See Commentary on page 8162.
been widely adopted in global biogeochemical models. Instead, 1
To whom correspondence should be addressed. Email: [email protected].
the vast majority of global biogeochemical models assumes This article contains supporting information online at www.pnas.org/lookup/suppl/doi:10.
fixed C:N:P in organic matter, including most participants in 1073/pnas.1423917112/-/DCSupplemental.

www.pnas.org/cgi/doi/10.1073/pnas.1423917112 PNAS | July 7, 2015 | vol. 112 | no. 27 | 8199–8204

cellular P:C ratio is more variable, due to the ability of organisms The N:C increases by ∼20% over the range of 0–3 μM NO3, and
to substitute phosphorus compounds such as phospholipids (16), then remains relatively constant (Methods). Meanwhile, the P:C
and ranges from ∼1/50 to 1/500, or by about a factor of 10. Iron is increases linearly across the observed range of PO4 concentrations.
even more variable, given the ability to substitute with other ele- These relationships are consistent with hyperbolic dynamics for the
ments, such that Fe:C varies from about 1/1,000 to 1/1,000,000, or ratios of both elements as functions of the corresponding ambient
three orders of magnitude (17). nutrient concentrations, with saturation occurring for the N:C in
If we examine oceanographic data in terms of N:C and P:C, nitrate-rich regions, but with all oceanic conditions falling below
rather than the more commonly cited N:P ratio, some rather saturation for the P:C, so that the relationship of P:C vs. PO4 is
simple patterns emerge. Fig. 1 shows these ratios measured in linear over the range of oceanic PO4 concentrations sampled here. A
hundreds of globally distributed organic particulate samples linear regression of P:C vs. PO4, using all available stations (Fig. S2),
(locations shown in Fig. S1), binned by nutrient concentrations, suggests that the relationship can be approximated as P:C = 6.9 ‰
and with the elemental ratios shown in permil units (‰) for ease μM−1 × PO4 + 6.0‰ (r2 = 0.36, P < 10−60). Binning and log-
of notation. As indicated by the large error bars, there is a broad transforming the data leads to slight variations in the regression, with
range in the measured stoichiometry of particles found in waters a similar slope but slightly lower intercept (Methods).
of any given ambient nitrate or phosphate concentration, which These empirical relationships make sense. Nutrient elements
partly reflects measurement error and the variable presence of exist at very low concentrations in the marine environment.
nonliving detritus. Because these samples do not discriminate Pumping them into the cell—and keeping them there—is a costly
undertaking, and becomes diffusion-limited at low concentrations
among particle types, they also include a significant fraction of
(18, 19). At very low concentrations, the benefit to be gained from
heterotrophic plankton in addition to phytoplankton. However, the
having a bit more of an element within the cell could be easily
striking finding from this large dataset is that, despite wide vari-
negated by the high cost of bringing it into the cell. However, as
ability among individual samples, the means of the bins display well- the ambient concentration of the nutrient rises, it becomes much
behaved trends with increasing ambient nutrient concentrations. less costly to make it available, and therefore, the marginal in-
crease in welfare to be gained from its use becomes worthwhile.
The existence of elemental plasticity can be conceptualized as
220 the product of evolution under perennial nutrient scarcity in the
ocean (20). A frugal phytoplankter carries out its tasks using
200 smaller quantities of any molecules whose nutritive ingredients are
scarce in the environment. Given millions of years of competition,
180 the most frugal phytoplankton could be expected to have evolved
160 to be very good at this, explaining the relatively high growth rates
in nutrient-poor regions (21). An example of a frugal phyto-
140 plankter would be Prochlorococcus, with a high nutrient uptake
capability despite a low cellular quota (22). On the other hand,
120 less frugal phytoplankton, such as the diatoms that dominate eu-
100
trophic waters, have the genetic tools to construct a more diverse
array of nutrient-rich molecules to accelerate their growth. There
80 are rapidly diminishing returns on the benefits gleaned from these
0 5 10 15 20 25 additional molecules, so that the average growth rates of a well
Mean NO3 concentration ( M) fed, less frugal phytoplankton are only modestly elevated over
those of the most frugal phytoplankton. However, it’s enough of a
benefit that, under most conditions, given high nutrient concen-
25
trations and enough time, the community will tend to become
dominated by the less frugal variety.
20 The robust linear correlation of P:C vs. PO4 suggests that
phosphorus frugality is an important consideration for phyto-
15 plankton everywhere in the ocean. Because this is an average
expectation for the entire community, and because the samples
probably include a significant fraction of heterotrophic biomass
10
with less plastic stoichiometry (23), the underlying dependence
of phytoplankton P:C on PO4 may be stronger than implied by
5 the empirical regression. This whole-community linear expecta-
tion could be locally modified by other factors such as water
0 temperature and light intensity. Indeed, a regression of the P:C
0 0.5 1 1.5 data vs. in situ temperature also shows a highly significant, al-
Mean PO4 concentration ( M) though weaker, relationship (r2 = 0.24), with a negative slope;
part of this reflects the fact that cold waters typically have higher
Fig. 1. Observed elemental ratios of suspended particles at the ocean surface. PO4, but it has also been shown that more ribosomes are re-
Stations were binned according to the NO3 concentration (Upper) and PO4 quired to meet the same growth rate at low temperatures,
concentration (Lower). Squares show the log-transformed mean of the ele- exerting an additional effect on the P:C (11). Such modification
mental ratios of all measurements within each bin. Error bars show 1 SD of the could lead to small-scale spatiotemporal variability, or even
elemental ratios of all measurements within each bin. The solid orange line in cause large regions to diverge from the global “Line of Frugal-
Upper shows a hyperbolic approximation to the variability in the mean, op-
ity.” However, the fact that the Line of Frugality explains 36% of
timized from all data, whereas the dashed orange line is optimized from the
binned, log-transformed means. The solid orange line in Lower shows a linear
the observed variability, despite measurement error, spatial and
fit to all P:C data, referred to here as the Line of Frugality, whereas the dashed temporal patchiness, and non-steady-state conditions, suggests
orange line shows an alternative linear fit using the means of the binned data. that the PO4 concentration is a very important, and perhaps the
The orange-shaded regions in both panels show the combined 95% confi- most important single factor in determining community P:C at
dence interval spanning the binned and unbinned regressions. the ocean surface. In contrast, the absence of any dependence of

8200 | www.pnas.org/cgi/doi/10.1073/pnas.1423917112 Galbraith and Martiny

 SEE COMMENTARY
N:C on NO3 above a NO3 concentration of ∼3 μM indicates that P:C predicted from PO4
other environmental and ecological factors determine N:C at all 20
but the lowest NO3 concentrations. The saturation of average N:C
at nitrate concentrations of about 3 μM does not indicate that 50N
nitrate is not limiting in the ocean. On the contrary, it is the most 15
important limiting element, at least in a proximal sense, as clearly
shown by nutrient assays everywhere outside of High Nitrate Low eq
Chlorophyll (HNLC) regions (15). The observations here simply 10
imply that higher N:C does not, on average, confer a consistent
competitive advantage above a ratio of ∼155 ‰. 50S
It follows that the common observation that N:P increases
5
under oligotrophy would arise simply due to low N:C plasticity 150W 100W 50W 0 50E 100E 150E
compared with P:C plasticity. The observed particulate data
N:C predicted from NO3
suggest that the average N:C of planktonic communities in- 160
creases by less than 30% above its minimum value of ∼125 ‰ in
response to an increase of NO3, whereas community average P:C 50N 155
can increase by a factor of at least 3 above its minimum of ∼5 ‰
150
under increasing PO4, exhibiting an order of magnitude more
response to nutrient availability than the N:C. Thus, moving eq 145
from eutrophic waters that are rich in both NO3− and PO43- to
oligotrophic waters that are poor in both NO3− and PO43- in- 140
volves a small decrease of N:C but a very large P:C decrease. As 50S 135
a result, nutrient-dependent stoichiometric plasticity would tend
to drive high N:P in oligotrophic regions and low N:P in eutro- 130
150W 100W 50W 0 50E 100E 150E
phic regions. In agreement with the underlying dynamic of che-
mostat models (10, 19), this mechanism suggests that N:P in N:P resulting from P:C and N:C predictions
phytoplankton is determined by the absolute concentrations of 25
NO3− and PO43-, rather than their ratio, and explains why the
NO3:PO4 ratio of ambient waters is a poor predictor of phyto- 50N
plankton N:P (24). 20

Global Picture eq
The simple predictive model proposed here can be used to 15
generate global predictions of organic matter stoichiometry from
observed surface nitrate and phosphate concentrations. The re- 50S
10
sults are shown in Fig. 2. The predicted P:C agrees broadly with
recent estimates from two global inverse models (Fig. S3), with 150W 100W 50W 0 50E 100E 150E
the only significant discrepancy being that the inverse models
Fig. 2. Predicted elemental ratios at the ocean surface. The elemental ratios
suggest higher P:C in the subpolar North Atlantic than predicted
indicated are predicted from the Line of Frugality shown in Fig. 1, using the
here. The calculated N:C and P:C in Fig. 2 follow similar spatial observed PO4 and NO3 concentrations from the upper 30 m in the World
patterns, given the strong correlation of NO3 and PO4 at the ocean Ocean Atlas (45).
surface. However, as expected from the reasoning laid out above,
the N:P varies roughly inversely with the macronutrient concen-
trations, because N:C varies less than P:C. It should be empha- appear to be given additional weight by the analysis of DeVries
sized that this result arises entirely from the elevated plasticity of and Deutsch (25), which suggests that the domains of polar
P:C relative to N:C, and is not dependent on assumptions re- oceans with seasonal sea ice may be typified by low P:C despite
garding the allocation of nutrient elements to cellular machinery

ENVIRONMENTAL
high phosphate concentrations. If it turns out that seasonal sea
or taxonomic groups. ice zones are indeed typified by anomalously low P:C, this could

SCIENCES
That is not to say that this simple mechanism is inconsistent be an important consideration for carbon cycling during ice ages
with either of these other hypotheses. It is entirely possible that as well as in the future (8). At the same time, the availability of
frugal phytoplankton do, in fact, produce more resource-gath- silicate could potentially expand the range of P-rich diatoms in
ering molecules relative to RNA (1). Similarly, the taxonomic
the Southern Ocean northward into lower-PO4 waters, driving
composition of ecosystems in different parts of the ocean could
relatively high P:C export at relatively low phosphate concen-
reflect the fit between phytoplankton elemental compositions
and the available nutrients (7). Diatoms with high P:C flourish in trations as proposed by Weber and Deutsch (8). For the time
phosphate-rich regions (which also happen to be nitrate-rich), being, we assume these other factors to be stationary with re-
and therefore diatom-dominated communities have low N:P. spect to the distribution of phosphate, and explore the impact
It is also important to emphasize that this prediction is sug- of changes in P:C driven exclusively by phosphate cycling on
gested to hold only at large spatial and temporal scales, with ocean carbon storage.
important exceptions. Perhaps most importantly, Phaocystis, a
prymnesiophyte that tends to have relatively low P:C, frequently Implications for Carbon Export and Atmospheric CO2
blooms in the PO4-rich Southern Ocean (12). The suggestion The global oceanic phosphorus inventory is determined by the
here would be that the success of Phaocystis reflects the relatively balance between the supply rate, ultimately derived from the
small advantage that less frugal phytoplankton gain by using weathering of minerals on land, and the removal by burial in
more P such that, given other environmental factors (such as seafloor sediments (26). If the P:C of organic matter were fixed,
grazing pressures, temperature, light intensity, or iron availabil- the maximum amount of C that could be stored in the ocean
ity), the more frugal Phaocystis can be very successful without interior by the biological pump would simply be equal to the P
taking full advantage of the abundant PO4. This wrinkle would inventory divided the P:C ratio. This concept led to some early

Galbraith and Martiny PNAS | July 7, 2015 | vol. 112 | no. 27 | 8201
suggestions that glacial-interglacial changes in the global bi- the impact of varying P:C, compared with fixed P:C, through a
ological pump strength could have been caused by changes in series of experiments. These experiments test the effect of variable
either the global P inventory, or the P:C ratio (27). In addition, high-latitude export Φh (which could vary due to iron supply to the
the use of preformed phosphate (i.e., phosphate that is transported Southern Ocean, or sea ice coverage), in combination with simul-
physically to the ocean interior from phosphate-rich regions of the taneous changes in low-latitude phosphate concentration PO4l
surface ocean, rather than being exported to the interior as a (which might vary with changes in nitrogen fixation and de-
component of organic matter and subsequently remineralized) as a nitrification rates), and the large-scale ocean circulation (varying the
metric for the global strength of the biological pump is typically thermohaline overturning T and high-latitude vertical mixing fhd).
simplified to assume a constant P:C ratio, although the possibility When the P:C is fixed, the prescribed changes modify atmo-
that P:C varies is acknowledged (28–30). The theory associated with spheric pCO2 in a fairly strict relationship with preformed PO4, as
the concept of preformed phosphate suggests that the combined determined by the PO4 concentration in the high-latitude box (Fig.
influences of ocean circulation and carbon export on atmospheric 3) (31, 32). This strict relationship holds true over the range of
pCO2 can be approximated quite well by the concentration of changes in PO4l, T, fhd, and Φh, and indicates that in a world with
phosphate in the high-latitude regions that ventilate the deep fixed stoichiometry, it is hard to change the strength of the bi-
ocean, most importantly the Southern Ocean. ological pump without changing preformed PO4. However, when
The argument advanced here suggests that less-frugal phyto- P:C is plastic, the effects of PO4l and T can diverge significantly
plankton in phosphate-rich regions actually use up P rapidly, from the strict preformed-PO4 relationship, producing a range of
without attaching much Corg to it. This profligate use of P makes pCO2 for a given preformed PO4 concentration. This divergence is
for a relatively weak biological pump, for a given inventory of P. In particularly pronounced for PO4l, which has a far greater impact
contrast, more frugal phytoplankton attach a lot of Corg to every on pCO2 under variable P:C than under fixed P:C (Fig. 4).
atom of P they export, storing a lot more respired carbon in the The reason for these divergences is fairly straightforward:
ocean interior per unit P exported. As such, in terms of carbon when P:C varies, export that occurs under low PO4 concentra-
storage, not all use of a PO4 molecule can be seen as equal. tions stores more C for the same P inventory, thereby lowering
We use a standard three-box ocean model coupled to a one- pCO2. Because the low-latitude export flux (Φl) has a lower P:C
box atmosphere (Fig. S4), after Sarmiento and Toggweiler (31), to than Φh, increasing the fraction of total export that occurs at low
illustrate how P:C variability alters carbon export and atmospheric latitudes leads to lower pCO2. In the model, Φl is simply equal to
CO2. Ref. 8 followed a similar approach to test the impact of a T(PO4d − PO4l), where PO4d is the phosphate concentration in
prescribed change in the P:C of export in the Southern Ocean. the deep box, which is essentially fixed by the total PO4 in-
Although greatly simplified, the box model framework allows the ventory. Therefore, increasing T drives a linear increase of Φl,
rapid computation of a large number of experiments to plumb the causing pCO2 to decrease. Similarly, decreasing PO4l increases
main sensitivities of carbon storage to circulation and community Φl, an effect that is compounded by decreasing the P:C of Φl as
export given different stoichiometries. The ocean boxes include a the ambient PO4 concentration becomes more scarce. Alto-
low-latitude surface box, high-latitude surface box, and a deep box, gether, this could produce more than a 20 ppm difference in
with realistic proportions (Methods). The box model predicts P:C atmospheric CO2 for a change in PO4l of only 0.1 μM (Fig. 4),
as a function of PO4 using the linear fit to the unbinned data such as might occur given a change in nitrogen limitation at low
shown by the solid line in Fig. 1. The model is then used to explore latitudes. Thus, the export of low P:C organic matter at low

fhd T PO4l
350 350 350
Atmospheric CO2 (ppm)

Atmospheric CO2 (ppm)

Atmospheric CO2 (ppm)

300 increasing fhd 300 increasing T 300

increasing
Fixed P:C 250 250 250
PO4l

increasing h
increasing h
200 200 200 increasing h

0.5 1 1.5 2 0.5 1 1.5 2 0.5 1 1.5 2

High latitude PO4 ( M) High latitude PO4 ( M) High latitude PO4 ( M)

350 350 350

Atmospheric CO2 (ppm)

Atmospheric CO2 (ppm)

Atmospheric CO2 (ppm)

increasing
300 300 increasing T 300 PO4l
increasing fhd
Variable P:C 250 250 250

200 increasing h 200 increasing h 200

increasing h

0.5 1 1.5 2 0.5 1 1.5 2 0.5 1 1.5 2

High latitude PO4 ( M) High latitude PO4 ( M) High latitude PO4 ( M)

Fig. 3. Relationship between preformed PO4 and pCO2 in box model simulations. Top panels show results with fixed P:C, bottom panels show variable P:C
following the Line of Frugality illustrated in Fig. 1. Each panel shows the pCO2 predicted by 100 model simulations that simultaneously vary the high-latitude
export flux (Φh, each value of which is shown by a single color) along a range of values for the other variable as indicated (shown as a continuous line for each
value of high-latitude export), where PO4l is low-latitude PO4, T is thermohaline overturning and fhd is high-latitude vertical mixing. For example, for the
variable P:C model, varying PO4l over the full range of values produces pCO2 changes of more than 70 ppm, with little effect on the preformed PO4.

8202 | www.pnas.org/cgi/doi/10.1073/pnas.1423917112 Galbraith and Martiny

 SEE COMMENTARY
of phytoplankton communities are mostly due to changes in the
300 concentration of PO4 in the ambient waters, with little de-
Atmospheric CO2 (ppm)
pendence on the ambient ratio of NO3:PO4.
A simple model based on these observations predicts the P:C
250 variable P:C of export as a function of PO4. When included in a classic box
model, the flexible P:C model shows heightened sensitivity of at-
mospheric CO2 to ocean circulation and the low-latitude PO4
200 concentration. Significant variations in the P:C of sinking organic
matter weaken the relationship between preformed PO4 and CO2
storage, suggesting that mechanisms of change in the biological
150 pump could be better framed in terms of remineralized NO3. The
simple predictive model can be tested against future observations
0 0.2 0.4 0.6 and laboratory study.
Low latitude surface PO4 ( M) The large impacts of variable P:C on ocean carbon storage
illustrated by the simple box model suggest that global bio-
Fig. 4. Relationship between low-latitude PO4 and pCO2 in box model geochemical models should place a high priority on resolving P:C
simulations. High-latitude export Φh is held constant at an intermediate plasticity, in order to improve the understanding of past and present
value for the simulations shown here. The variable P:C simulations (orange) variability, and to account for this important aspect of the marine
show a much larger impact of low-latitude PO4 on pCO2 than the fixed P:C ecosystem in future projections of N2 fixation, denitrification, ocean
simulations (gray). The enhanced sensitivity of the variable P:C simulations carbon uptake and oxygen distributions.
arises from the greater carbon storage achieved by low-latitude export, per
mol of P export, compared with high-latitude export. The range of possible Methods
CO2, corresponding to the 95% confidence intervals of the Line of Frugality
Data Analysis. The particulate data of ref. 7 is reanalyzed here as follows. For
regression shown in Fig. 1, are indicated by the region of orange shading.
each station (i.e., geographic location at which samples were collected), all
available measurements of particulate organic carbon (POC), particulate
organic nitrogen (PON), particulate organic phosphorus (POP), nitrate, and
latitudes could be a more significant player than often realized in phosphate within the upper 30 m of the water column were averaged.
determining the strength of the global soft tissue pump. Because Wherever particulate data existed without corresponding nutrient mea-
PO4l would be dependent on the availability of nitrogen in low- surements, the climatological nutrient concentration for the same location
latitude surface waters, it would be expected to respond to was taken from the monthly WOA climatology (7). The P:C and N:C were
changes in denitrification and N2 fixation, emphasizing their calculated for all stations at which the mean POP and PON concentrations
importance to the carbon cycle (33). were greater than 0.005 μmol l−1. This selection process resulted in 610
stations with P:C and 4,064 stations with N:C data (Fig. S1).
Despite the reduced applicability of the preformed-PO4 metric
A linear regression of P:C vs. PO4 was calculated using all 610 (unbinned)
under variable P:C, the smaller variability of N:C would suggest data points, giving an intercept of 6.0 ± 0.2 ‰ and slope of 6.9 ± 0.4 ‰
that the relationship between remineralized NO3 (i.e., the non- μmol−1 L as shown by the solid orange lines in Fig. 1 and Fig. S2, referred to
preformed component of NO3) and soft tissue pump carbon stor- as the Line of Frugality. The linear regression coefficient (r2, Pearson) against
age should be more robust. This general principle could make all 610 data points is 0.36 (P < 10−60). In addition, the log-transformed means
remineralized NO3 a better monitor of changes in the biological of the binned data were used to generate an alternative linear regression,
pump, particularly given the fact that past changes in relative NO3 given that the log-transformed values within each bin showed a more
consumption can be reconstructed using the stable isotopes of ni- normal distribution. The binned-lognormal mean regression gives a signifi-
cantly lower intercept of 4.8 ± 0.5 ‰ and a slightly higher slope of 7.3 ±
trogen recorded in marine sediments (34, 35). The difficulty in the 0.6‰ μmol−1 L, shown by the dashed orange lines in Fig. 1 and Fig. S2. The
use of remineralized NO3 is the fact that denitrification can reduce linear regression coefficient of the binned-lognormal mean regression line
NO3 concentrations along circulation pathways, so that the nitrate against the nine binned means is 0.95 (r2, Pearson, P < 10−5). To illustrate the
concentration is less than the sum of the preformed and remin- full uncertainty in the Line of Frugality, the combined range of the 95%
eralized components. Furthermore, variability of N:C may be suf- confidence intervals for both regressions is shown by the shaded areas in
ficient to cause a significant divergence of the soft tissue pump Figs. 1 and 4.
strength from remineralized nitrate, although this is difficult to A linear regression of N:C vs. NO3 using all 4,064 stations shows a weak
(0.8 ‰ μM−1), but highly significant (P < 10−10) linear increase over the full

ENVIRONMENTAL
judge without a better understanding of what controls the N:C (36).
range of NO3 concentrations. However, if only the stations with >5 μM NO3

SCIENCES
are analyzed, the linear regression has a near-zero slope that is only weakly
Conclusions
significant (P < 0.1), whereas a regression over the lowermost 3 μM NO3 range
The stoichiometric ratios P:C and N:C are usually held constant shows a strong slope (11 ‰ μM−1) with very high significance (P < 10−10). This
in global biogeochemical models, despite evidence for significant rapid increase of N:C over low NO3 concentrations, followed by negligible
variability from field and laboratory studies. Based on global change, is consistent with a saturating hyperbolic relationship. An optimized
suspended particulate measurements, it would appear that the fit using a Holling type 2 functional form is given by N:C = 125 ‰ + 30 ‰ ×
community-averaged N:C varies linearly with the dissolved con- NO3/(0.32 μM + NO3), shown in Fig. 1 and Fig. S2 along with a shaded interval
that spans the 95% confidence intervals of the nonlinear fit using both the full
centrations of NO3 over low NO3 concentrations, and that the
unbinned dataset as well as binned log-transformed means.
community-averaged P:C varies linearly with PO4 across the
range of modern ocean PO4 concentrations. We suggest that this Model. The biogeochemical model used here includes three ocean boxes and
is because the phytoplankton community tends to adjust to the one atmospheric box, following ref. 31. The low-latitude surface PO4 con-
available nutrient concentrations, either by intraspecific plasticity centration is restored to a defined value, generating an export flux Φl that
of elemental composition (e.g., ref. 37), or by shifting the species depends on the physical supply of PO4 to the low latitudes by the thermo-
assemblage. All else being equal, the most frugal phytoplankton— haline circulation T, whereas the high-latitude surface export flux Φh is im-
those with the lowest requirements for a nutrient—dominate when posed as an input parameter, representing iron-light-temperature-limited
growth. A prescribed vertical mixing flux exchanges water between the
the concentrations of that nutrient are lowest, and therefore most
high-latitude surface and the deep box, fhd. Carbonate exported from the
difficult to obtain. Because N is an integral component of many surface by calcifying organisms is set to 20% of the organic carbon export
organic molecules, the N:C varies by less than 30% in response to flux, within the observational range given by Berelson et al. (38). Within the
changes in NO3, whereas P:C varies by a factor of three in re- surface boxes, the carbonate equilibria are solved as a function of Dissolved
sponse to changes in PO4. As a corollary, changes in the N:P ratio Inorganic Carbon (DIC) and alkalinity concentrations, to determine the

Galbraith and Martiny PNAS | July 7, 2015 | vol. 112 | no. 27 | 8203
saturation state of CO2. Air–sea exchange is a linear function of CO2 satu- well as with the Line of Frugality prediction (Fig. S3). The good agreement
ration state and a constant air sea exchange coefficient (the “piston veloc- between these alternative methods suggests that the Line of Frugality pre-
ity”), as described in ref. 39. diction captures the first order variability in P:C of exported organic matter.
For the purposes at hand, we assume that the P:C of sinking particulate The low-latitude PO4 concentration PO4l is restored to a prescribed value
export follows the same relationship with surface phosphate as the measured that is varied between 0–0.6 μM. The export flux of carbon from the high-
surface particles. The real world is undoubtedly more complex, given pref- latitude box Φh is prescribed to vary from 0.25 to 4 Pg C y−1, as might occur
erential export by specific taxa (such as diatom aggregates) at particular under large changes of iron fertilization. The thermohaline overturning T is
varied over a range of 10–45 Sv. The high-latitude vertical mixing fhd is
times of year, and the consumption of phytoplankton by heterotrophs,
varied over a range of 30–130 Sv. Varying each of PO4l, T, and fhd over
producing rapidly sinking fecal pellets with P:C that differs from that of
10 values with simultaneous variations in Φh over 10 values results in a total
the food source (40). Additionally, it has been observed that P disappears
of 300 simulations, each of which was repeated with both fixed and flexible
more rapidly than C from sinking particles within the upper water column,
P:C, run to equilibrium (20,000 y).
implying preferential remineralization of P that would add a vertical di-
mension to differential P:C cycling (41–43). Nonetheless, two recent global
ACKNOWLEDGMENTS. We thank Daniele Bianchi for help with the box model
inverse models incorporating ocean circulation and chemical data (25, 44) code, and for comments on an early version of the manuscript. We are grateful
found a close match across ocean basins between the stoichiometry of net to John Dunne, Jorge Sarmiento, and François Primeau for discussions, and to
organic matter export and the suspended particulate data used here, as two anonymous reviewers for their thoughtful and constructive suggestions.

1. Sterner RW, Elser JJ (2002) Ecological Stoichiometry (Princeton University Press, 25. DeVries T, Deutsch C (2014) Large-scale variations in the stoichiometry of marine
Princeton), p 439. organic matter respiration. Nat Geosci 7(12):890–894.
2. Geider RJ, La Roche J (2002) Redfield revisited: variability of C: N: P in marine mi- 26. Ruttenberg KC (2003) The Global Phosphorus Cycle. Treatise on Geochemistry, ed
croalgae and its biochemical basis. Eur J Phycol 37(1):1–17. Elderfield H (Elsevier, Amsterdam), Vol 8, pp 585–643.
3. Quigg A, et al. (2003) The evolutionary inheritance of elemental stoichiometry in 27. Broecker WS (1982) Ocean chemistry during glacial time. Geochim Cosmochim Acta
marine phytoplankton. Nature 425(6955):291–294. 46(10):1689–1705.
4. Rhee G-Y (1978) Effects of N: P atomic ratios and nitrate limitation on algal growth, 28. Sigman DM, Boyle EA (2000) Glacial/interglacial variations in atmospheric carbon
cell composition, and nitrate uptake. Limnol Oceanogr 23(1):10–25. dioxide. Nature 407(6806):859–869.
5. Goldman JC, McCarthy JJ, Peavey DG (1979) Growth rate influence on the chemical 29. Ito T, Follows MJ (2005) Preformed phosphate, soft tissue pump and atmospheric
composition of phytoplankton in oceanic waters. Nature 279:210–215. CO2. J Mar Res 63(4):813–839.
6. Karl DM, et al. (2001) Ecological nitrogen-to-phosphorus stoichiometry at station 30. Marinov I, et al. (2008) Impact of oceanic circulation on biological carbon storage in
ALOHA. Deep Sea Res Part II Top Stud Oceanogr 48(8):1529–1566.
the ocean and atmospheric pCO2. Global Biogeochem Cycles 22(3):GB3007.
7. Martiny AC, et al. (2013) Strong latitudinal patterns in the elemental ratios of marine
31. Sarmiento JL, Toggweiler JR (1984) A new model for the role of the oceans in de-
plankton and organic matter. Nat Geosci 6(4):279–283.
termining atmophericpCO2. Nature 308:621–624.
8. Weber TS, Deutsch C (2010) Ocean nutrient ratios governed by plankton bio-
32. Sigman DM, Haug GH (2003) The Biological Pump in the Past. Treatise on Geo-
geography. Nature 467(7315):550–554.
chemistry, eds Holland D, Turekian KK (Elsevier, Amsterdam), Vol 6, pp 491–528.
9. Deutsch C, Weber T (2012) Nutrient ratios as a tracer and driver of ocean bio-
33. Falkowski PG (1997) Evolution of the nitrogen cycle and its influence on the biological
geochemistry. Annu Rev Mar Sci 4:113–141.
10. Klausmeier CA, Litchman E, Daufresne T, Levin SA (2004) Optimal nitrogen-to-phos- sequestration of CO2in the ocean. Nature 387(6630):272–275.
phorus stoichiometry of phytoplankton. Nature 429(6988):171–174. 34. Altabet MA, Francois R (1994) Sedimentary Nitrogen Isotopic Ratio As a Recorder For
11. Toseland A, et al. (2013) The impact of temperature on marine phytoplankton re- Surface Ocean Nitrate Utilization. Global Biogeochem Cycles 8(1):103–116.
source allocation and metabolism. Nature Climate Change 3(11):979–984. 35. Galbraith ED, Kienast M; NICOPP-Working-Group (2013) The acceleration of oceanic
12. Arrigo KR, et al. (1999) Phytoplankton community structure and the drawdown of denitrification during deglacial warming. Nat Geosci 6(7):579–584.
nutrients and CO2 in the southern ocean. Science 283(5400):365–367. 36. Martiny AC, Vrugt JA, Primeau FW, Lomas MW (2013) Regional variation in the
13. Bopp L, et al. (2013) Multiple stressors of ocean ecosystems in the 21st century: Pro- particulate organic carbon to nitrogen ratio in the surface ocean. Global Biogeochem
jections with CMIP5 models. Biogeosciences 10(10):6225–6245. Cycles 27(3):723–731.
14. Tilman D, Kilham SS, Kilham P (1982) Phytoplankton community ecology: the role of 37. White AE, Spitz Y, Karl D, Letelier RM (2006) Flexible elemental stoichiometry in Tri-
limiting nutrients. Annu Rev Ecol Syst 13:349–372. chodesmium spp. and its ecological implications. Limnol Oceanogr 51(4):1777–1790.
15. Moore C, et al. (2013) Processes and patterns of oceanic nutrient limitation. Nat 38. Berelson W, et al. (2007) Relating estimates of CaCO3 production, export, and dis-
Geosci 6:701–710. solution in the water column to measurements of CaCO3 rain into sediment traps and
16. Van Mooy BA, et al. (2009) Phytoplankton in the ocean use non-phosphorus lipids in dissolution on the sea floor: A revised global carbonate budget. Global Biogeochem
response to phosphorus scarcity. Nature 458(7234):69–72. Cycles 21(1):GB1024.
17. Boyd PW, et al. (2007) Mesoscale iron enrichment experiments 1993-2005: synthesis 39. Sarmiento J, Gruber N (2006) Ocean Biogeochemical Dynamics (Princeton University
and future directions. Science 315(5812):612–617. Press, Princeton, Oxford).
18. Morel FM (1987) Kinetics of nutrient uptake and growth in phytoplankton. J Phycol 40. Sterner RW (1990) The ratio of nitrogen to phosphorus resupplied by herbivores:
23(1):137–150.
zooplankton and the algal competitive arena. Am Nat 136(2):209–229.
19. Bonachela JA, Raghib M, Levin SA (2011) Dynamic model of flexible phytoplankton
41. Faul KL, Paytan A, Delaney ML (2005) Phosphorus distribution in sinking oceanic
nutrient uptake. Proc Natl Acad Sci USA 108(51):20633–20638.
particulate matter. Mar Chem 97(3):307–333.
20. Falkowski PG, et al. (2004) The evolution of modern eukaryotic phytoplankton. Sci-
42. Lyons G, Benitez-Nelson CR, Thunell RC (2011) Phosphorus composition of sinking
ence 305(5682):354–360.
particles in the Guaymas Basin, Gulf of California. Limnol Oceanogr 56(3):1093.
21. Liu H, Nolla HA, Campbell L (1997) Prochlorococcus growth rate and contribution to
43. Sekula‐Wood E, Benitez‐Nelson CR, Bennett MA, Thunell R (2012) Magnitude and
primary production in the equatorial and subtropical North Pacific Ocean. Aquat
composition of sinking particulate phosphorus fluxes in Santa Barbara Basin, Cal-
Microb Ecol 12(1):39–47.
22. Lomas MW, Bonachela JA, Levin SA, Martiny AC (2014) Impact of ocean phytoplankton ifornia. Global Biogeochem Cycles 26(2):GB2023.
diversity on phosphate uptake. Proc Natl Acad Sci USA 111(49):17540–17545. 44. Teng Y-C, Primeau FW, Moore JK, Lomas MW, Martiny AC (2014) Global-scale vari-
23. Makino W, Cotner J, Sterner R, Elser J (2003) Are bacteria more like plants or animals? ations of the ratios of carbon to phosphorus in exported marine organic matter.
Growth rate and resource dependence of bacterial C: N: P stoichiometry. Funct Ecol Nat Geosci 7(12):895–898.
17(1):121–130. 45. Garcia HE, et al. (2014) Dissolved Inorganic Nutrients (Phosphate, Nitrate, Silicate),
24. Hall SR, Smith VH, Lytle DA, Leibold MA (2005) Constraints on primary producer N: World Ocean Atlas 2013, eds Levitus S, Mishonov A (NOAA Atlas NESDIS 76, Silver
P stoichiometry along N: P supply ratio gradients. Ecology 86(7):1894–1904. Spring, MD), Vol 4.

8204 | www.pnas.org/cgi/doi/10.1073/pnas.1423917112 Galbraith and Martiny