Introduction
The global climate change induced by anthropogenic activities is causing a
wide range of alterations to the marine environment including ocean
acidification (OA), rising sea surface temperature (SST), and intensified
stratification due to increased density gradients between surface and
subsurface waters, with associated shifts in mean irradiance levels and
nutrient availability in the upper water column (Boyd and Doney, 2002; Rost
and Riebesell, 2004; Stocker, 2013). All these global changes in
environmental variables will affect the physiology and ecology of
phytoplankton, both individually and interactively, in a complex way (Boyd
and Hutchins, 2012; Boyd et al., 2010, 2016; Feng et al., 2017).
Phytoplankton elemental composition is an important cellular property that
reflects the metabolic rates of phytoplankton (Raven and Geider, 1988).
Elemental composition is strongly influenced by environmental conditions and
by phytoplankton adaptations to these conditions (Sterner and Elser, 2002),
which in turn influences marine food web structure, particulate carbon export
to the deep ocean, and ultimately marine biogeochemistry (Finkel et al., 2010
and references therein). The widely recognized average molar elemental ratio
of C : N : P is 106 : 16 : 1 for marine phytoplankton assemblages –
the Redfield ratio (Redfield et al., 1963). However, individual phytoplankton
species may have elemental ratios deviating, on short timescales (days to
months) from Redfield depending on the environmental conditions they
encounter. Such deviations subsequently influence the accumulation of these
elements in the upper food web and also marine biogeochemistry (Finkel et
al., 2010; Ho et al., 2003; Sardans et al., 2012).
Different environmental drivers may play a range of roles in regulating the
stoichiometry of marine phytoplankton. Nutrient availability (Hecky et al.,
1993; Perry, 1976) has been proven to affect phytoplankton stoichiometry
directly. Irradiance provides the energy source for nutrient assimilation in
the cells (Goldman, 1986). In addition, temperature changes, which mainly
alter metabolic rates, can also influence the diffusive uptake of nutrients
into cells (Raven and Geider, 1988; Roleda et al., 2013). Increased levels of
dissolved CO2 during cell growth may result in higher cellular C : N
and C : P ratios, due to increased CO2 availability as a substrate for
photosynthesis (Beardall et al., 2009; Feng et al., 2008; Fu et al., 2007,
2008). However, the dependency of C : N or C : P ratios on CO2
availability can be species specific (Burkhardt and Riebesell, 1997;
Burkhardt et al., 1999). The effect of rising pCO2 on the N : P
ratio is still unclear, due to large variations observed in previous
environmental manipulation studies (Sardans et al., 2012). For example, the
N : P ratio of Synechococcus increased with elevated CO2
concentration, but remained unchanged for Prochlorococcus (Fu et
al., 2007) and Emiliania huxleyi (Feng et al., 2008).
Marine coccolithophores are responsible for almost half of the global marine
calcium carbonate production, and are important in the marine carbon cycle
through both the organic carbon pump and the inorganic carbon counter pump
(Rost and Riebesell, 2004). Emiliania huxleyi is the most widely
distributed coccolithophore species (Balch et al., 1991; Holligan et al.,
1993, 1983), and has been selected as a model phytoplankton
species in the context of the marine carbon cycle (Westbroek et al., 1993). A
wide range of environmental drivers, such as CO2 concentration, nutrient
level, irradiance, and temperature influence the growth, photosynthesis, and
calcification of E. huxleyi both individually and interactively
(Feng et al., 2017; Raven and Crawfurd, 2012; Zondervan, 2007). Changes in
these physiological processes may in turn alter the elemental stoichiometry
and composition of coccolithophores. Knowledge of how different environmental
drivers will affect the elemental composition of E. huxleyi is
important for a more complete understanding of the physiological responses of
this species to the changing environment and the consequent effects on
biogeochemical cycles. In addition, the magnitude of change in each
environmental driver will be different with the future climate change,
depending on location and scenario; hence a systematic study across a
gradient of each driver is required.
This study advances previous findings by relating the change in elemental
composition of E. huxleyi cells, in response to environmental
forcing, to the physiological rate responses presented in the study of Feng
et al. (2017). The major objective of the present study is to investigate and
rank the importance of the environmental drivers, including the nitrate and
phosphate concentrations, irradiance, temperature, and pCO2, on
setting the elemental composition of a Southern Ocean strain of E. huxleyi. The combined results of this study and Feng et al. (2017) provide
new insights into how environmental changes will impact the marine
biogeochemical cycles related to E. huxleyi.
Results
Changes in cellular POC content in response to environmental
drivers
Cellular POC content was significantly affected by alteration of irradiance,
temperature, and pCO2 (Fig. 1). Increasing irradiance from 14 to
80 µmol photons m-2 s-1 increased the cellular POC
content by around 2-fold from 8.20 ± 2.39 to
14.07 ± 1.17 pg cell-1 (p<0.05). POC content decreased at
the two highest irradiance levels (350 and
650 µmol photons m-2 s-1, Fig. 1c). A trend of
decreased E. huxleyi cellular POC content with elevated temperature
was evident from the temperature manipulation experiment (Fig. 1d). The
cellular POC content (28.85 ± 6.98 pg cell-1) was significantly
higher than all the other treatments (p<0.05) at the lowest temperature
of 4 ∘C and significantly reduced by ∼ 70 % at both 20 and
25 ∘C (p<0.05). Raising pCO2 from 8 to 15 Pa
significantly increased the cellular POC content from 9.63 ± 1.67 to
12.93 ± 1.84 pg cell-1 (Fig. 1e), with cellular POC content
being relatively uniform from 15 to 109 Pa.
Changes in the ratio of Emiliania huxleyi cellular
particulate inorganic carbon content to particulate organic carbon content
(PIC : POC) in response to different environmental drivers:
(a) PIC : POC ratio vs. nitrate concentration;
(b) PIC : POC ratio vs. phosphate concentration;
(c) PIC : POC ratio vs. irradiance; (d) PIC : POC ratio
vs. temperature; and (e) PIC : POC ratio vs. pCO2.
Alteration of cellular PIC content in response to environmental
drivers
Temperature was the only driver that significantly altered the cellular PIC
content (Fig. 2). There was a general trend of decreased cellular PIC content
of E. huxleyi with warming from 11 to 20 ∘C (Fig. 2d). The
cellular PIC content was significantly lower at 20 and 25 ∘C
compared to the other four temperature treatments (p<0.05). More than a
50 % decrease in cellular PIC content was observed at the two highest
temperature conditions, relative to the 7 ∘C treatment. However,
there were no significant differences in cellular PIC content between the
other temperature treatments. The fitted Qc value (the plateau for one
phase decay) was 6.94 ± 0.93 pg cell-1, close to the average
value at the two highest temperatures (Table S1).
Changes in Emiliania huxleyi cellular particulate organic
nitrogen (PON) content in response to different environmental drivers:
(a) cellular PON vs. nitrate concentration, (b) cellular
PON vs. phosphate concentration, (c) cellular PON vs. irradiance,
(d) cellular PON vs. temperature, and (e) cellular PON vs.
pCO2.
Changes in the cellular PIC : POC ratio in response to environmental
drivers
As for POC, the cellular ratio of PIC : POC was mainly affected by changes in
irradiance, temperature, and pCO2 (Fig. 3). The highest cellular
PIC : POC ratio of 1.20 ± 0.09 was observed at the lowest irradiance
(19 µmol photons m-2 s-1; p<0.05, compared to all
other irradiance treatments). The ratio then decreased with increasing
irradiance to 0.72 ± 0.10 at
190 µmol photons m-2 s-1 and slightly increased again
at the two highest irradiances (p<0.05 between 190 and
650 µmol photons m-2 s-1, Fig. 3c). In the temperature
manipulation experiment, the PIC : POC ratio was significantly lower (p<0.05) at the lowest temperature (4 ∘C) than any other treatment,
with a value of 0.45 ± 0.03 pg cell-1. The PIC : POC value
then levelled off between the range of 7 to 25 ∘C, with the average
value more than double that at 4 ∘C (Fig. 3d). With the variation of
pCO2 levels, the cellular PIC : POC ratio decreased by more than
40 % from 1.46 ± 0.02 pg cell-1 at 8 Pa to 0.90 ± 0.15 at 39 Pa and stayed similar between the range of 39 and 109 Pa
(p < 0.05) (Fig. 3e), mainly due to the increased cellular POC quota
with rising pCO2.
Changes in Emiliania huxleyi cellular particulate organic
phosphorus (POP) content in response to different environmental drivers:
(a) cellular POP vs. nitrate concentration, (b) cellular
POP vs. phosphate concentration, (c) cellular POP vs. irradiance,
(d) cellular POP vs. temperature, and (e) cellular
POP vs. pCO2.
Alteration of cellular PON content in response to environmental
drivers
The cellular PON content increased with increasing nitrate concentration. The
content at the two lowest nitrate concentrations of 3.7 and 6.0 µM
was less than half of the average value (2.06 ± 0.36 pg cell-1)
of the three highest nitrate treatments (Fig. 4a). Warming from 4 to
25 ∘C decreased the cellular PON content (p < 0.05). The value
of 4.07 ± 0.00 pg cell-1 at 4 ∘C was double that at
15 ∘C (1.93 ± 0.10 pg cell-1) and 3-fold greater
than the PON content of 1.31 ± 0.24 pg cell-1 at 25 ∘C
(Fig. 4d).
Changes in the ratio of Emiliania huxleyi cellular
particulate organic carbon content to chlorophyll a content (C : Chl a)
in response to different environmental drivers: (a) C : Chl a
ratio vs. nitrate concentration, (b) C : Chl a ratio vs.
phosphate concentration, (c) C : Chl a ratio vs. irradiance,
(d) C : Chl a ratio vs. temperature, and
(e) C : Chl a ratio vs. pCO2. Error bars represent
standard deviations (n=3).
Elemental molar ratios of N : P, C : N and C : P of
Emiliania huxleyi from the five single-factorial manipulation
experiments. The errors are standard deviations around the mean (n=3). The
values in bold are significantly different compared to other treatments.
Environmental
Treatment
N : P
C : N
C : P
driver
(mol : mol)
(mol : mol)
(mol : mol)
Nitrate (µM)
3.7
9.09 ± 2.48
15.90 ± 4.09
137.69 ± 8.27
6
9.07 ± 1.55
13.16 ± 1.13
118.45 ± 10.06
12
15.90 ± 1.38
9.00 ± 0.07
143.08 ± 11.28
48
14.46 ± 1.39
7.01 ± 0.64
107.76 ± 11.93
96
16.16 ± 3.45
6.56 ± 1.07
103.56 ± 6.54
200
14.31 ± 0.40
7.32 ± 0.47
104.82 ± 8.34
Phosphate (µM)
0.4
41.57 ± 4.14
6.32 ± 0.42
261.67 ± 17.30
2
21.47 ± 0.28
9.27 ± 5.08
137.57 ± 11.31
6
14.17 ± 3.59
6.16 ± 1.24
85.00 ± 4.51
10
17.06 ± 2.06
5.24 ± 0.06
89.28 ± 9.80
20
13.99 ± 0.89
5.22 ± 0.08
73.04 ± 3.48
Irradiance
14
22.03 ± 7.27
4.26 ± 0.29
73.82 ± 6.48
(µmol photons m-2 s-1)
40
14.27*
4.38*
62.50*
80
16.03 ± 2.90
5.24 ± 0.69
83.96 ± 17.05
190
18.71 ± 1.74
5.99 ± 0.23
112.39 ± 14.60
350
17.15 ± 0.83
6.47 ± 0.54
110.63 ± 3.98
650
16.11 ± 1.77
5.70 ± 0.31
91.49 ± 6.66
Temperature (∘C)
4
8.92 ± 1.29
8.67 ± 2.64
71.34 ± 27.92
7
10.46 ± 2.05
7.60 ± 1.32
78.70 ± 14.78
11
13.58 ± 1.91
6.21 ± 0.32
86.56 ± 11.92
15
14.12 ± 0.66
6.70 ± 0.31
94.47 ± 3.84
20
15.53 ± 1.06
5.98 ± 0.15
92.88 ± 6.33
25
13.67 ± 2.99
7.08 ± 1.39
93.96 ± 4.87
pCO2 (Pa)
8
19.39 ± 2.41
6.81 ± 1.09
122.61 ± 9.97
15
24.01 ± 6.80
5.80 ± 0.71
137.50 ± 33.51
39
16.96 ± 3.62
7.41 ± 0.06
155.64 ± 31.42
58
17.89 ± 0.80
6.55 ± 1.16
116.67 ± 16.15
74
18.22 ± 2.45
6.42 ± 0.43
116.93 ± 17.60
109
13.56 ± 2.78
7.41 ± 1.20
99.25 ± 17.70
* Sample loss during analysis resulted in only single
values at this irradiance.
Changes in cellular POP content in response to environmental
drivers
The cellular POP content of E. huxleyi was significantly altered by
nitrate, phosphate, temperature, and pCO2. POP content was slightly
less at the three low-nitrate concentrations (3.7, 6.0, and 12 µM),
compared to those at 96 and 200 µM (p<0.05; Fig. 5a). Cellular
POP content significantly increased with rising phosphate concentration
(Fig. 5b), with the highest POP content observed at 20 µM
phosphate. As observed for cellular POC and cellular PON contents, warming
greatly decreased the cellular POP content (Fig. 5d), with a reduction of
65 % from 1.08 ± 0.14 pg cell-1 at 4 ∘C to
0.38 ± 0.04 pg cell-1 at 11 ∘C, but then only a further
decrease of ∼ 0.1 pg cell-1 from 15 to 25 ∘C.
Significant differences in POP content were detected between the two lowest
temperature treatments compared to all others. Conversely, with rising
pCO2 level there was a trend of increased cellular POP content
(Fig. 5e), which almost doubled from 0.20 ± 0.04 pg cell-1 at
8 Pa to 0.38 ± 0.02 pg cell-1 at 109 Pa (p < 0.05).
Alteration of cellular C to Chl a ratio in response to
environmental drivers
Alteration of all the five environmental drivers greatly affected the
cellular ratio of POC to Chl a content (C : Chl a, g : g) (p<0.05) (Fig. 6). C : Chl a decreased exponentially with increased nitrate
concentration up to 50 µM, but stabilized between 50 and
200 µM (Fig. 6a). The highest ratio of 422.36 ± 74.28 was
observed at the lowest nitrate concentration of 3.7 µM,
significantly higher than all other treatments (p<0.05). The ratio then
decreased by 87 % at 200 µM. An increase in phosphate
concentration, however, only slightly decreased the C : Chl a ratio
(Fig. 6b). Compared to the ratios at the two lowest concentrations, a
significant decrease (p<0.05) at 6.0 and 20 µM was observed (by
∼ 20 % each). Increased irradiance increased the C : Chl a
ratio linearly, with more than a doubling at
650 µ mol photons m-2 s-1 compared to the ratio of
47.45 ± 12.58 at 14 µmol photons m-2 s-1
(Fig. 6c). The C : Chl a ratio dramatically decreased with warming,
especially between 4 and 7 ∘C (Fig. 6d). The ratio of
131.26 ± 42.96 observed at 4 ∘C was significantly higher than
all the other temperatures (p<0.05). Significantly lower C : Chl a
ratios were observed at the two lowest pCO2 levels of 8 and 15 Pa
compared with the other treatments (p<0.05, Fig. 6e), with the ratio
increasing by 42 % from low to high pCO2.
Shifts in cellular elemental molar ratios in response to
environmental drivers
The PON to POP (N : P) ratio was significantly lower (p<0.05) at the two
lowest nitrate treatments compared to the others (Table 2). In contrast, the
POC to PON (C : N) ratio was significantly higher (p<0.05) at the two
lowest nitrate concentrations. There was no significant difference in C : N
ratios across the other four nitrate treatments (p>0.05). Changes in
nitrate concentration did not significantly affect the POC to POP (C : P)
ratio.
The N : P ratio of E. huxleyi increased at low phosphate
concentrations (0.4 and 2 µM), with highest value in the
0.4 µM phosphate treatment (p<0.05). There was a significant
increase in the C : P ratio (p<0.05) at the two lowest phosphate
concentrations compared to the others. The highest C : P ratio, recorded at
the lowest phosphate concentration (0.4 µM), was almost double the
value at 2 µM and more than 3 times the average ratio of the
other treatments (Table 2). In contrast, there were no significant
differences in the calculated C : N ratio across the phosphate treatments
(p>0.05).
Comparison of cellular particulate organic carbon (POC)
contents, particulate inorganic carbon (PIC) contents, PIC : POC ratios,
particulate organic nitrogen (PON) contents, and particulate organic
phosphorus (POP) contents of Emiliania huxleyi between projected (year 2100) and present-day Chatham Rise
conditions, with rankings of the importance of the environmental drivers
which caused significant effects on each physiological parameter. The
numbers of the ranking scheme represent the gradient of the most (1) to
least (4) important effects. Effect “+” represents an
increase and “-” represents a decrease in the elemental composition/ratio
in the future, respectively.
Physiological
Environmental
Fitted values at
Future vs. present day
parameter
driver
different conditions of
comparisons
environmental driversa
Present
Future
Change
Effects
Ranking
day
(%)b
(+/-)
Cellular POC
Temperature
10.798
9.713
10.0
–
1c
content
CO2
14.632
15.436
5.5
+
2
(pg cell-1)
Irradiance
14.774
14.827
0.3
+
3
Nitrate
n.s.
Phosphate
n.s.
Cellular PIC
Temperature
10.206
8.753
14.2
–
1
content
Nitrate
n.s.
(pg cell-1)
Phosphate
n.s.
Irradiance
n.s.
CO2
n.s.
PIC : POC
CO2
0.868
0.821
5.4
–
1
Temperature
1.017
1.042
2.4
+
2
Irradiance
0.777
0.780
0.3
+
3
Nitrate
n.s.
Phosphate
n.s.
Cellular PON
Nitrate
1.380
1.162
15.8
-
1
content
Temperature
2.013
1.819
9.6
-
2
(pg cell-1)
Phosphate
n.s.
Irradiance
n.s.
CO2
n.s.
Cellular POP
Phosphate
0.106
0.078
25.9
-
1
content
Temperature
0.304
0.269
11.6
-
2
(pg cell-1)
CO2
0.312
0.342
9.6
+
3
Nitrate
0.249
0.227
8.9
-
4
Irradiance
n.s.
a The fitted values for “present day” and “future” were extracted from
the fitted dose–response curves (Figs. 1–5) at the stock culture growing
conditions, average present-day conditions in the Chatham Rise area, and the
predicted future conditions (2100) of Chatham Rise, respectively.
b The percentage changes were calculated as the changes caused by each
environmental driver under the future predicted condition relative to that
under the present-day condition.
c Numbers in bold indicate statistically significant difference between
the range of present-day and future conditions (nitrate treatments: 6.0 and
12.0 µM; phosphate treatments: 0.4 and 2 µM; irradiance
treatments: 80 and 190 µmol photons m-2 s-1; temperature
treatments: 11, 15 and 20 ∘C) based on
the one-way ANOVA. “n.s.” indicates non-significant difference (one-way
ANOVA) among all the treatments used for the fitting.
Decreased C : N ratios were observed for low irradiances; the value at
14 µmol photons m-2 s-1 being significantly lower than
the three highest irradiances (p<0.05). Similarly, a decreased C : P
ratio was found at low irradiance, with a significantly lower value at
14 µmol photons m-2 s-1 compared to the three highest
irradiances. Warming significantly increased the N : P ratio from 4 to
20 ∘C (Table 2).
Ranking the importance of environmental drivers in altering
Emiliania huxleyi elemental composition
Ranking the response of the Southern Ocean E. huxleyi isolated to
projected future changes in oceanic properties revealed differential
responses between drivers and processes (Table 3, Fig. 7). Cellular POC and
cellular PIC : POC ratio were both significantly influenced by CO2 and
temperature, with temperature affecting cellular POC content the most, while
CO2 was the most important factor regulating PIC : POC. However, only
one driver (temperature) significantly regulated cellular PIC, with a
4 ∘C warming causing a 14.2 % decrease. The cellular PON content
was significantly affected by future nitrate concentration and temperature,
with nitrate ranking the most important. Four (phosphate, temperature,
CO2, and nitrate) out the five environmental drivers, under end-of-the-century conditions, significantly affected cellular POP content, with future
phosphate concentration playing the most important role. The rankings
associated with statistically non-significant differences among the treatment
intervals, as marked in Table 3 and Fig. 7, need to be considered with
caution (see Feng et al., 2017).
Conceptual figure of the specific effects of each the five
environmental drivers, under the projected future conditions (year 2100), on
the elemental composition of Emiliania huxleyi. Q represents the
cellular quota of each element of Emiliania huxleyi. The box denotes
the E. huxleyi cell. Solid blue arrows indicate positive effects of
the future environmental changes, and dashed red arrows indicate negative
effects of the future environmental changes. Arrows in bold indicate the
environmental drivers that play the most important role regulating the
connected physiological metrics under the predicted environmental conditions
for the year 2100.
Discussion
This is the first detailed study of the individual effect of five
environmental drivers (nitrate concentration, phosphate concentration,
irradiance, temperature, and pCO2) on the cellular elemental composition
of the coccolithophore E. huxleyi. Moreover, it is the first to rank the importance of
the predicted changes in these environmental drivers on E. huxleyi elemental
stoichiometry for the year 2100 relative to the present-day conditions.
Relating changes in elemental composition is an important addition to the
responses of growth, photosynthesis, and calcification rates (Feng et al.,
2017), providing insights into the biogeochemical consequences of the
physiological effects induced by change in the five essential environmental
drivers.
Effects of nutrient concentration on the elemental stoichiometry of
Emiliania huxleyi
The PON and POP cell quotas of E. huxleyi in the present study were
mainly controlled by nitrate and phosphate concentrations, respectively, as
phytoplankton relies on seawater nutrient availability as the external
elemental source (Hecky et al., 1993; Price, 2005; Sakshaug and Holmhansen,
1977). Nitrate concentration plays an important role in regulating the
growth, photosynthetic, and calcification rates of E. huxleyi (Feng
et al., 2017); however, the three lowest nitrate concentrations only resulted
in slightly decreased cellular POP contents and had no significant effect on
cellular POC or PIC content. This indicates that the regulation of the
nitrate concentration on the POC and PIC productivity in our study was mainly
a consequence of decreased growth rate of the cells under nitrate limitation,
as shown by Feng et al. (2017). This finding is in contrast to
Paasche (1998),
who observed higher E. huxleyi cellular PIC : POC ratios under
nitrate limitation as a result of decreased cellular POC and increased
coccolith abundance per cell in E. huxleyi strain BOF 92 isolated
from the North Atlantic. Higher PIC : POC ratios under nitrate limitation was
alternatively attributed to increased calcite mass per lith of E. huxleyi strain CCMP 378 isolated from the Gulf of Maine (Fritz, 1999). In
addition, phosphate concentration did not significantly affect E. huxleyi cellular carbon content nor the PIC : POC ratio of cells in the
present study. However, Paasche (1998) observed greatly increased PIC content
of E. huxleyi (strain BOF 92) under phosphate-limiting conditions,
and Riegman et al. (2000) observed that a greater increase PIC quotas under
phosphate limitation than nitrate limitation for E. huxleyi (strain
L).
These discrepancies between studies in the nitrate or phosphate effects on
cellular PIC : POC ratio are mainly due to the different nutrient
concentrations in the culturing media. Paasche (1998) observed an increase in
E. huxleyi PIC cell quota under the stationary phase of batch
incubation, i.e. when cell division ceased as nitrate dropped to ≤ 0.2 µM and phosphate dropped to ≤ 0.03 µM. This
supports the findings of both Riegman et al. (2000) and Fritz (1999), who
conducted continuous incubations with high cell densities of E. huxleyi. These studies observed an increased cellular PIC content when
phosphate concentration fell below 0.4 nM (Riegman et al., 2000) or nitrate
concentration was below the detection limit (Fritz, 1999). However, the
present study used a semi-continuous incubation method with higher and
relatively steady nutrient concentrations (with lowest nitrate and phosphate
concentrations of 3.6 and 0.4 µM, respectively) and the cells were
grown and sampled at a healthy exponential growth phase. Similarly,
Müller et al. (2008) only found higher E. huxleyi (strain
CCMP371) cellular calcite content during the stationary but not the
exponential growth phase under both nitrate and phosphate limitation, due to
the different cell cycle phases during which the calcification and cell
division occurred. The authors explained that calcification continued during
the G1 phase of cell assimilation when cell division was restricted under
nutrient limitation, and thus the cellular PIC content was increased
(Müller et al., 2008). Further studies at extremely low nutrient
concentrations (< 0.1 µM) in a steady-state growth phase are
still needed to understand the potential connection between carbon production
and extreme nutrient limitation, given reports of areal expansion of
oligotrophic waters in the world oceans with global climate change (Polovina
et al., 2008).
Irradiance effects on the elemental stoichiometry of
Emiliania huxleyi
In the present study, irradiance was the main environmental factor affecting
cellular POC content which in turn altered the PIC : POC ratio. The
increased PIC : POC cellular ratio at low irradiance indicates that
calcification is less dependent on irradiance than organic carbon fixation,
as discussed in Feng et al. (2017). Although both processes require light as
an energy source, calcification requires less energy (Anning et al., 1996)
than photosynthesis (Paasche, 1965; Balch et al., 1992). Therefore, the
calcification rate is generally saturated at lower irradiance levels than
photosynthesis (Paasche, 1964; Zondervan, 2007). Feng et al. (2017) reported
greatly reduced photosynthetic rates under the two lowest irradiance levels,
while observing that this trend was less significant for the calcification
rate. Hence, limiting irradiance will lead to less POC content in the cells
compared to the cellular PIC quota, and thus a higher cellular PIC : POC
ratio would be expected at low irradiance when growth and photosynthesis are
light-limited (Raven and Crawfurd, 2012), as confirmed by the response of
calcification : photosynthesis in Feng et al. (2017).
Increasing irradiance also elevated the C : Chl a ratio linearly in the
present study, due to the increase in POC quota and a decrease in Chl a
quota, as also reported for diatoms and dinoflagellates (Geider, 1987). The
reduced cellular pigment quota under high irradiance helps to reduce the
energy required for light harvesting in phytoplankton cells, which is a
strategy to balance the energy demands for growth and POC production with
photon harvesting (Kiefer, 1993). In addition, the present study revealed
that the C : N and C : P ratios of E. huxleyi both increased at
high light levels, as a consequence of increased cellular POC content driven
by increased irradiance but no significant change in cellular PON or POP
quota, further suggesting that organic carbon content is more light dependent
than the accumulation of cellular N or P (Geider et al., 1998).
Temperature effects on the elemental stoichiometry of
Emiliania huxleyi
Temperature is important in regulating dissolved chemical diffusion and
transport, non-enzymatic and enzymatic reactions, and the metabolic rates of
phytoplankton (Raven and Geider, 1988). In our accompanying study, the
growth, photosynthetic, and calcification rates all increased with rising
temperature until the optimal temperature was reached at 25, 24, and
20 ∘C respectively (Feng et al., 2017), which were all higher than
the stock culture growth temperature or the temperature at the isolation site
of E. huxleyi strain NIWA 1108. In the present study, the cellular
POC, PON, and POP content all reduced significantly as temperature increased.
It has been proposed that reduced cell size is a universal strategy in
response to increasing temperature for both terrestrial and aquatic organisms
(Gardner et al., 2011), following a hypothesis suggested by Atkinson et
al. (2003). A study on the coccolithophores E. huxleyi (strain EH2)
and Gephyrocapsa oceanic (strain GO1) observed decreased cell size
and thinner coccospheres upon raising temperature from 10 to 25 ∘C,
which was attributed to the relatively suppressed cell division at low
temperature (Sorrosa et al., 2005). This decrease in cell volume (Fig. S1 in
the Supplement) could be the main cause of reduced cellular elemental
components in the present study. Previous studies also reported that warming
resulted in reduced cell volume of E. huxleyi (strain AC481: De Bodt
et al., 2010; strain L: van Rijssel and Gieskes, 2002), and decreased
cellular POC and PIC quotas of coccolithophore Coccolithus pelagicus when the temperature was raised from 10 to 15 ∘C (Gerecht et al.,
2014). Similarly, warming significantly decreased the cellular elemental
contents to their lowest levels measured in the present study over the range
from 4 to 25 ∘C, with a decrease in cell size at higher temperatures
(Fig. S1), as growth rate increased (Feng et al., 2017).
However, contrary to the observed changes in POC, PON, and POP cell quota,
the cellular PIC content of E. huxleyi only decreased when
temperature was higher than 11 ∘C in the present study, due to the
strongly reduced calcification and malformation at low temperatures of 4 and
7 ∘C (Feng et al., 2017). The reduced cell division rate (i.e.
enlarged cell volume, Fig. S1) offset the reduced calcification rate at lower
temperatures, and so there was no significant difference in PIC cell quota at
temperatures below 11 ∘C. Consequently, the cellular PIC : POC ratio
was lower at 4 and 7 ∘C, consistent with the trend observed for the
calcification : photosynthesis ratio (Feng et al., 2017), indicating
suppression of PIC formation relative to POC production at low temperature
(Watabe and Wilbur, 1966). The PIC : POC ratio then decreased with warming
from 11 to 15 ∘C and remained relatively steady afterwards, mainly
due to the lower optimal temperature for calcification (20 ∘C)
compared to photosynthesis (24 ∘C) as suggested in Feng et
al. (2017).
Furthermore, warming from 4 to 20 ∘C significantly increased the
E. huxleyi cellular N : P ratio in the present study, in agreement
with the recent model study on a natural phytoplankton community (Toseland et
al., 2013). Toseland et al. (2013) found that with increasing temperature the
rate of cellular protein synthesis in phytoplankton was higher, but with a
lower number of phosphorus-rich ribosomes, thereby increasing the cellular
N : P ratio. In the present study, the cellular N : P ratio of E. huxleyi at 20 ∘C increased by 74 % from that at 4 ∘C,
in spite of both cellular PON and cellular POP decreasing with warming.
Although this study presents results for a single strain of E. huxleyi, if the temperature dependency of cellular resource allocation is a
universal trend for all the E. huxleyi genotypes, we can speculate
that the diverse E. huxleyi strains growing in different temperature
regions might have different requirements for nitrogen vs. phosphorus, and
that the growth of E. huxleyi strains in the temperate to tropical
regions might be more readily limited by nitrate than sub-polar strains.
Similarly, Toseland et al. (2013) suggested that future warming might
accentuate nitrate limitation in the oceans.
Effects of CO2 on the elemental stoichiometry of
Emiliania huxleyi
The photosynthesis of E. huxleyi was saturated at a higher
pCO2 than that for growth rate (Feng et al., 2017). In the present
study, CO2 plays the most important role in regulating the cellular
PIC : POC ratio. The PIC : POC ratio was significantly higher at the
lowest pCO2 level, as a consequence of the lower cellular POC and
higher cellular PIC at 8 Pa. In general, cell growth of E. huxleyi
is less limited by low CO2 concentrations than in other phytoplankton
groups (Clark and Flynn, 2000; Paasche et al., 1996; Riebesell et al., 2000;
Rost et al., 2003). Moreover, recent studies suggest that E. huxleyi
operates an active carbon concentrating mechanism (CCM) to utilize
HCO3- through the enzyme carbonic anhydrase (CA; Reinfelder, 2011),
and may have high affinity for CO2 in photosynthesis (Stojkovic et al.,
2013). However, the efficiency of CCMs in E. huxleyi (strain B92/11)
is considered to be low as a consequence of the leakage of CO2 from the
cell (Rost et al., 2006), and so coccolithophore photosynthesis is more
dependent than cell growth on CO2 concentration (Rost and Riebesell,
2004). This discrepancy between growth and organic carbon fixation can lead
to a decrease in cellular POC at low pCO2. This difference in CO2
requirements between the two processes may also have resulted in the lower
cellular POP content at 8 Pa compared to other pCO2 treatments.
The increasing trend observed for cellular POC and POP was not apparent for
cellular PIC quota, as calcification rates significantly decreased with
increasing pCO2 level > 40 Pa (Feng et al., 2017). Hence the
cellular PIC : POC ratio was significantly higher at the two lowest
pCO2 levels, consistent with previous findings for CO2
manipulations at saturating irradiances on E. huxleyi (strain PML
B92/11A; Zondervan et al., 2002, 2001). No further significant change in
cellular carbon content or PIC : POC ratio occurred at higher pCO2,
in contrast to the linear decrease in the calcification : photosynthesis
ratio with rising pCO2 (Feng et al., 2017). This difference is
noteworthy as both cellular PIC : POC and calcification : photosynthesis
ratios are commonly used to examine the relative change of PIC and POC
production in coccolithophores (Raven and Crawfurd, 2012). These changes have
biogeochemical implications for the marine rain ratio in the carbon cycle
(Klaas and Archer, 2002; Rost and Riebesell, 2004), which is the export ratio
of calcite to organic carbon into the deep ocean. The 14C-labelling
technique used in this study (see Feng et al., 2017) to measure carbon
fixation (photosynthesis and calcification) rates was conducted during the
light period; thus the measured rate is an indicator of net carbon fixation
that does not account for the energy-consuming respiratory process or
CO2 leakage out of the cells (Bach et al., 2015, 2013; Rost et al.,
2006). Conversely, the cellular carbon content indicates the gross
accumulated carbon in the cells over longer period of growth (Engel et al.,
2010; Fabry and Balch, 2010). The most compelling reason for the relatively
higher PIC : POC ratio (∼ 1.5) than calcification : photosynthesis
ratio (∼ 1.0) in the lowest pCO2 treatment (8 Pa, 79 ppm) in
our study may then be attributed to diffusive CO2 loss limiting
inorganic carbon active uptake from the substrate (Bach et al., 2013),
resulting in less POC fixation into the cells relative to the PIC fixation by
the calcification process.
The C : Chl a ratio of E. huxleyi was lowest at pCO2 of
8 Pa across all the pCO2 treatments in the present study, mainly due
to the decreased cellular POC at low pCO2, rather than any change in
cellular Chl a content. However, increasing pCO2 did not have
significant effects on the C : N, N : P or C : P ratios in the present
study. This is in accordance with a recent study on E. huxleyi
(strain PML B92/11A), which also exhibited constant C : N : P ratios
across a pCO2 range of 18 to 75 Pa for cultures at steady growth
phase under phosphate-limited continuous incubation (Engel et al., 2014).
Biogeochemical implications and future directions
The comparisons between present-day conditions and those projected for year
2100 for the Chatham Rise area are summarized in the conceptual figure
(Fig. 7). These results indicate that the 2 ∘C warming will decrease
both POC and PIC cellular quotas of E. huxleyi, but may slightly
increase the PIC : POC ratio. Rising pCO2 alone will result in
decreased cellular PIC : POC ratio. Although the 33 % decrease in
nitrate concentration is the major factor controlling the growth,
photosynthetic, and calcification rates (Feng et al., 2017), change in
nitrate concentration did not significantly affect the elemental
stoichiometry except for the cellular PON contents of E. huxleyi. In
addition, increasing temperature may increase the cellular N : P ratio,
while rising pCO2 will decrease the N : P and C : P ratios. These
results provide a more detailed perspective that can improve our knowledge on
how the model coccolithophore species, E. huxleyi, may respond to
future environmental changes. For example, our results suggest that rising
pCO2 in the future oceanic environment will decrease the E. huxleyi cellular PIC : POC ratio by 5.4 %; however, the projected
warming and increase in irradiance level may offset this decreased
PIC : POC by 2.4 and 0.3 %, respectively. The changes in PIC : POC have
implications for the marine “rain ratio” and so alter the marine carbon
cycle (Rost and Riebesell, 2004). Similarly, the cellular N : P ratio will be
decreased by rising pCO2, although this trend may be cancelled out by
warming. The altered C : N : P stoichiometry will in turn affect the
nutrient cycle at higher trophic levels (Jones and Flynn, 2005) and marine
biogeochemical cycles (Beardall and Raven, 2004).
It is noteworthy that the research presented here only examined the
physiological response norms of E. huxleyi to a single environmental
driver when other drivers were all kept at the stock culture growth condition
(i.e. a set of single-dimensional space experiments). However, these
responses (such as the shape of the curves and the optimal conditions) may be
different when the other background conditions are changed. For example, Sett
et al. (2014) observed the dose–response curves of calcification of
E. huxleyi PML B9/11 to CO2 concentration was regulated by
temperature. Therefore, in order to comprehensively understand how E. huxleyi physiology will respond to multiple environmental drivers and fill
this knowledge gap, future research on a full environmental matrix is still
necessary. These experiments will not only help to further explore the
potential interactions (i.e. synergistic or agnostic effects) between
environmental drivers, but also provide a better understanding of the
underlying mechanisms of these interactive effects. In addition, the present
study is only based on a single strain of Southern Hemisphere E. huxleyi. Due to the wide distribution of this species in the natural marine
environment, E. huxleyi presents high variability in terms of
genetic, morphological, and physiological characteristics (Cook et al., 2011;
Read et al., 2013; Young et al., 2014). Therefore, the physiology of
different E. huxleyi strains isolated from different geographic
locations might respond differently to changing environmental drivers. For
example, within the context of OA research, extensive previous studies
suggest a strain specificity of E. huxleyi in response to changes in
seawater carbonate chemistry (Langer et al., 2009; Raven and Crawfurd, 2012;
Blanco-Ameijerias et al., 2016). It has also been observed that different
E. huxleyi ecotypes/morphotypes responded differently to OA
(Müller et al., 2015), which is likely a consequence of their genetic
variation (Cook et al., 2011). The present study and Feng et al. (2017)
demonstrate the important roles of different environmental drivers in
controlling the physiology of E. huxleyi strain NIWA1108, and so
further work is required to determine if the findings apply to other strains.
In summary, this study, in combination with Feng et al. (2017), has a number
of implications for research into the response of E. huxleyi to
ocean acidification and global climate change. In addition to seawater
carbonate chemistry (Riebesell et al., 2010), it is necessary to report the
experimental conditions of all the environmental drivers carefully. The
predictions presented will provide useful information for biogeochemical
models, such as that of Bopp et al. (2001), of how the elemental
stoichiometry of E. huxleyi will respond to the alteration of these
environmental conditions individually, in order to predict the future changes
in the marine biogeochemical cycles. In addition, multiple environmental
drivers tend to change simultaneously in the future global climate change
scenario (Boyd and Hutchins, 2012), and so future studies should also
investigate the interactions between these multiple drivers on phytoplankton
physiology. The predicted future changes in marine physical properties (such
as sea surface temperature (SST) and mixed layer depth) will vary from one
oceanic region to another (Boyd and Doney, 2002). The dose–response curves
from our study suggest that the range of alteration in environmental drivers
may control the outcome of the effects of environmental perturbation on
E. huxleyi physiology and biogeochemistry. For future
multi-factorial manipulation experimental designs, our results suggest that
the magnitudes of change in each environmental driver need to be
determined/decided cautiously and should have environmental relevance in
order to make more accurate predictions, and the understanding of interactive
effects of multiple environmental drivers and the underlying mechanisms
should be further explored.