Introduction
Emiliania huxleyi is an abundant and ubiquitous phytoplankton species, belonging to the
coccolithophores (Haptophyta), a group of calcifying microalgae.
Coccolithophores fix CO2 into organic matter by photosynthesis,
contributing to the drawdown of atmospheric CO2 (Raven and
Falkowski, 1999). Calcification, on the other hand, releases CO2 in the
short term (Rost and Riebesell, 2004) and stores carbon in
coccoliths in the long term (Sikes et al., 1980; Westbroek et al., 1993).
In addition, coccolith ballast can accelerate the removal of organic carbon
from upper water layers and aid long-term burial of carbon
(Ziveri et al., 2007). Many studies have therefore addressed
the production of organic and inorganic carbon (calcite) in E. huxleyi, as well as its
modification by environmental factors such as carbonate chemistry
(Riebesell et al., 2000; Meyer and Riebesell, 2015), nutrient
availability (Paasche and Brubak, 1994; Langer and Benner, 2009),
temperature (Watabe and Wilbur, 1966; Langer et al., 2010), salinity
(Paasche et al., 1996; Green et al., 1998) and light (Paasche,
1968, 1999).
This study investigates the physiological and morphological response of E. huxleyi to
two environmental stressors, phosphorus (P) limitation and increased
temperature. These are predicted to occur simultaneously as a rise in global
temperature will increase the likelihood of nutrient limitation in the
photic zone due to a stronger stratification of the water column
(Sarmiento et al., 2004). The availability of macronutrients such as
nitrogen and P have been shown to affect the production of particulate
organic (POC) and inorganic carbon (PIC) in coccolithophores (reviewed by
Zondervan 2007). Coccolith number per cell generally increases in
P-limited cultures, often leading to an increase in the PIC / POC ratio
(Paasche and Brubak, 1994; Paasche, 1998; Müller et al., 2008; Perrin
et al., 2016). However, five out of six Mediterranean E. huxleyi strains showed a
decreased PIC / POC ratio in response to P limitation, and one strain
displayed no change (Oviedo et al., 2014). While this
demonstrates that there are strain-specific responses to P limitation, some
differences between studies on PIC and POC production are due to differences
in experimental methods, notably batch culture and (semi-)continuous culture
approaches (Langer et al., 2013b). We used both setups in
this study to examine the difference between strong, yet brief P limitation
(stationary phase batch culture) against weak, but continuous P limitation
(semi-continuous culture). The latter method best represents areas with
permanently low nutrient availability such as the eastern Mediterranean
(Krom et al., 1991; Kress et al., 2005), while stationary phase
batch culture can be approximated to an end-of-bloom scenario in which the
lack of nutrients limits further cell division. Both approaches are relevant
in ecological terms, but for methodological reasons (i.e. non-constant
growth rates), nutrient-limited production cannot be determined in the batch
approach (e.g. Müller et al., 2008; Langer et al., 2012, 2013b; Gerecht et al., 2014; Oviedo et
al., 2014; Perrin et al., 2016). In a (semi-)continuous culturing setup,
growth rate is constant over the course of the experiment and production
rates can be calculated (Paasche and Brubak, 1994; Paasche, 1998; Riegman
et al., 2000; Borchard et al., 2011). Ratio data such as coccolith
morphology, on the other hand, should be comparable between batch and
(semi-)continuous culture experiments (Langer et al., 2013b),
as has been shown for C. pelagicus (Gerecht et al., 2014, 2015).
However, the only strain of E. huxleyi (B92/11) that was tested in both batch and
continuous culture was not analysed for coccolith morphology and the PIC / POC
ratio showed a markedly different response to P limitation in batch and in
continuous culture (Borchard et al., 2011; Langer et al., 2013b). In this
study we therefore tested another strain of E. huxleyi in both semi-continuous and
batch culture and analysed among other things, coccolith morphology and the
PIC / POC ratio.
In addition to P limitation we studied the effect of temperature on
coccolith morphology and carbon production. Only a few studies have
specifically dealt with the effect of temperature on the occurrence of
coccolith malformations. These studies suggest that higher than optimum
temperature leads to an increase in malformations (Watabe and Wilbur,
1966; Langer et al., 2010). Although the effect of temperature on carbon
production in E. huxleyi has been addressed in numerous studies (Sorrosa et al.,
2005; Feng et al., 2008; Satoh et al., 2009; De Bodt et al., 2010; Borchard
et al., 2011; Sett et al., 2014; Matson et al., 2016; Milner et al., 2016;
Rosas-Navarro et al., 2016), none of these studies tested the effect of
above-optimum temperature. To our knowledge, this study is the first to
specifically test the impact of heat stress on carbon production in this
species.
Basal composition of the culture media (modified K/2), including
salinity and carbonate chemistry (AT, pH, DIC, ΩCa).
∗ For trace metal composition please refer to the recipe available for K/2 Ian
at http://roscoff-culture-collection.org/basic-page/culture-media.
Final conc.
“Control, replete
“P-limiting
(µM)
medium”
medium”
NaNO3
288
288
KH2PO4
10
0.5
(Na)FeEDTA
5.85
5.85
Trace metals
∗
∗
Vitamins
“f/2”
“f/2”
Salinity (ppm)
34
34
AT (µmol kg-1 )
2250
2100
pH (NBS)
8.14
7.89
DIC (µmol kg-1)
1800
1800
ΩCa
4.7
2.7
Materials and methods
Cultures
We grew a strain of E. huxleyi isolated from the Oslo Fjord (22 June 2011 by Shuhei Ota) in
triplicate semi-continuous and batch cultures in replete (control) and
P-limiting medium at two temperatures (19, 24 ∘C). The Oslo Fjord
experiences high summer temperatures of 19–21 ∘C with winter lows
of down to 0 ∘C (Aure et al., 2014). As E. huxleyi often has maximum
growth rates at temperatures above those found at the isolation site (Sett et al., 2014), we chose 19 ∘C as our control
temperature, which is towards the high end of the temperature range this
strain is likely to encounter in nature. We used a 5 ∘C
temperature increase to induce heat stress. This temperature (24 ∘C) was above the
optimum for growth, i.e. the cultures grew exponentially,
but at a lower rate (Eppley, 1972). This strain belongs to morphotype A
and was kept in stock culture at 12 ∘C under low light in K/2
medium prior to the start of the experiment. The strain has been deposited
at the NIVA Culture Collection of Algae (http://niva-cca.no) as strain UIO 265.
Experimental cultures were grown in modified K/2 medium (Table 1) at a
salinity of 34 ppm and an initial phosphate concentration of 10 µM
(control) or 0.5 µM (P-limiting). The cultures were kept in an
environmental test chamber (MLR-350, Panasonic, Japan), on a 12:12 h
light : dark cycle at an irradiance of ∼ 100 µmol photons m-2 s-1.
Cultures were acclimated for ca. 10 generations to the
two initial P concentrations and temperatures before starting the
experiment.
(a) Cell concentrations and (b) cell volume over
time in batch cultures of Emiliania huxleyi grown at 19 and
24 ∘C in control and P-limiting medium. Error bars denote the
standard deviation of mean triplicate measurements of triplicate cultures.
Arrows indicate when cultures were harvested.
Cell concentrations were
determined daily using an electronic particle counter (CASY, Roche
Diagnostics, Switzerland). Maximum cell concentrations in semi-continuous cultures were kept well below stationary phase (< 170 000 cells cells mL-1) by diluting the cultures back to ∼ 10 000 cells mL-1 with fresh medium every second dayso that all cultures were
kept continuously in the exponential growth phase (Fig. S1). Semi-continuous
cultures were harvested after 10 dilution cycles. For batch cultures, the
initial inoculum was ∼ 10 000 cells mL-1. P-limited
cultures were harvested in stationary phase, whereas control cultures were
harvested in exponential phase at similar cell concentrations (see Gerecht
et al., 2014; Fig. 1). Exponential growth rates (µexp) were
calculated by linear regression of log-transformed cell concentrations over
time. For batch cultures, only the exponential part of the growth curve was
considered. For semi-continuous cultures μexp was calculated as
an average of μexp of all dilution cycles.
Medium chemistry
Residual phosphate concentrations
Residual phosphate concentrations were determined in the culture medium upon
harvest of the cultures. The medium was sterile filtered (0.2 µM)
into plastic scintillation vials (Kartell, Germany) and stored at -20 ∘C until analysis. Phosphate concentrations were determined
colorimetrically on a spectrophotometer (UV 2550, Shimadzu, Japan) as
molybdate reactive phosphate following Murphy and Riley (1962) with
a precision of ±4 %.
Carbonate chemistry
Total alkalinity (AT) and pH of the medium were determined upon harvest
of the cultures. The initial carbonate chemistry of the culture media is
presented in Table 1. Samples for AT were filtered through GF/F filters
(Whatman, GE Healthcare, UK), stored airtight at 4 ∘C and
analysed within 24 h. AT was calculated from Gran plots (Gran,
1952) after duplicate manual titration with a precision of ±50 µmol kg-1.
The pH was measured with a combined electrode (Red
Rod, Radiometer, Denmark) that was two-point calibrated to NBS scale
(precision ±0.03). Dissolved inorganic carbon (DIC) concentrations
and saturation state of calcite (ΩCa) were calculated using
CO2sys (version 2.1 developed for MS Excel by D. Pierrot from E. Lewis and
D. W. R. Wallace) using AT and pH as input parameters and the
dissociation constants for carbonic acid of Roy et al. (1993).
Elemental composition
Particulate organic phosphorus
Samples for particulate organic phosphorus (POP) were filtered onto
precombusted (500 ∘C, 2 h) GF/C filters (Whatman) and stored at
-20 ∘C. Particulate organic phosphorus was converted to
orthophosphate by oxidative hydrolysis with potassium persulfate under high
pressure and temperature in an autoclave (3150EL, Tuttnauer, Netherlands)
according to Menzel and Corwin (1965). Converted orthophosphate was
then quantified as molybdate reactive phosphate as described in Sect. 2.2.1.
Particulate organic and inorganic carbon
Samples for total particulate carbon (TPC) and POC were filtered onto
precombusted GF/C filters, dried at 60 ∘C overnight in a drying
oven, and stored in a desiccator until analysis on an elemental analyser
(Flash 1112, Thermo Finnigan, USA; detection limit 2 µg; precision
±8 %). Particulate inorganic carbon was removed from POC filters
by pipetting 230 µL of 2 M HCl onto the filters before analysis
(Langer and Benner, 2009) and calculated as the difference between TPC
and POC.
Cell volume calculated from LM and CASY measurements at the time of
harvest, number of coccoliths cell-1 and number of coccoliths analysed
by SEM and classified into normal, incomplete and malformed coccoliths in
semi-continuous and batch control and P-limited cultures of Emiliania huxleyi grown at 19 and 24 ∘C. The number of cells (n) analysed
for each measurement is presented as the sum of three replicates, except for
CASY cell volume measurements for which n=3 ± standard deviation.
Semi-continuous cultures
Batch cultures
Control
P-limited
Control
P-limited
Cell volume (µm3)
LM
19 ∘C
34.3 ± 17.7
19.9 ± 9.5
29.7 ± 12.1
57.6 ± 22.7
(n=111)
(n=116)
(n=346)
(n=205)
24 ∘C
24.6 ± 12.8
34.3 ± 17.7
24.7 ± 14.1
64.3 ± 31.7
(n=117)
(n=194)
(n=352)
(n=217)
CASY
19 ∘C
74.4 ± 8.9
75.5 ± 7.4
62.7 ± 7.3
106.9 ± 9.8
24 ∘C
94.0 ± 4.5
92.1 ± 6.8
67.1 ± 5.8
115.1 ± 3.0
Coccoliths cell-1
19 ∘C
20 ± 9
18 ± 6
15 ± 5
45 ± 20
(n=148)
(n=149)
(n=151)
(n=149)
24 ∘C
16 ± 7
15 ± 5
16 ± 6
34 ± 15
(n=145)
(n=146)
(n=149)
(n=145)
Number of coccoliths analysed for morphology
19 ∘C
821
824
693
3496
24 ∘C
731
721
691
2010
Normal (%)
19 ∘C
81.5
79.5
77.6
21.3
24 ∘C
51.0
54.7
57.1
33.8
Incomplete (%)
19 ∘C
1.3
0.8
1.8
76.7
24 ∘C
2.0
0.7
4.4
52.4
Malformed (%)
19 ∘C
17.2
19.7
20.6
2.0
24 ∘C
46.9
44.7
38.5
13.7
Cell geometry
Cell volume was calculated from cell diameters measured both visually from
light microscopy (LM) images and automatically with an electronic particle
counter (CASY). With LM, cell diameters of live cells were measured at 200
times magnification after dissolving the coccoliths with 0.1 M HCl
(19 µL to 1 mL sample; Gerecht et al., 2014) after harvesting the
cultures. CASY cell diameters were recorded during daily
measurements of cell concentrations (see Sect. 2.1) without removing coccoliths.
Cell volume derived from CASY data therefore overestimates actual cell
volume, because part of the coccosphere is included. However, volume
estimates from CASY data are based on the measurement of many cells, leading
to robust data, i.e. a lower standard deviation than LM measurements (Table 2).
They are therefore useful for comparative purposes and for following the
development of cell size during culture growth (Fig. 1; see also Gerecht et
al., 2015).
Representative SEM micrographs of normal, incomplete and malformed
coccoliths, including dissolution features: (a) coccosphere bearing
normal coccoliths; (b) arrow: an incomplete coccolith in control
batch culture; (c) arrows: malformed coccoliths with merged distal
shield elements/increased gaps; (d) a coccosphere from an Oslo Fjord
field sample; arrows highlight the same type of malformations (merged distal
shield elements, missing central area) as observed in culture; (e)
coccosphere with many malformed coccoliths showing merged distal shield
elements, triangular thickening of the elements and irregular calcite growth;
(f) partially dissolved coccosphere; white arrow: detached distal
shield elements; red arrow: “hammer-like” distal shield elements;
(g) strongly dissolved coccosphere; white arrow: detached distal
shield elements; red arrow: dissolved central area; asterisk: exposed
proximal shield elements. Scale bar =1 µm.
A Zeiss Supra35-VP field emission scanning electron microscope (SEM, Zeiss,
Germany) was used to capture images for morphological analyses. The number
of coccoliths per coccosphere was estimated from these images by counting
visible, forward-facing coccoliths, multiplying this number by 2 to
account for the coccoliths on the back side of the coccosphere, and adding
the number of partially visible coccoliths along its edge
(Gerecht et al., 2015). Coccolith morphology was classified
into three categories: normal, incomplete and malformed (Table 3; Fig. 2).
Due to the low calcite saturation state reached in stationary phase batch
cultures, we observed a high number of partially dissolved coccoliths in
these cultures (the features of this secondary dissolution are described in
Table 3 and Fig. 2). As it was not possible to unambiguously distinguish
incomplete morphology due to secondary dissolution from incompletely
produced coccoliths, only one class of incomplete coccoliths is presented in
Fig. 3.
Coccolith morphology of Emiliania huxleyi grown at 19 and
24 ∘C in control and P-limiting medium in semi-continuous and batch
culture. Coccoliths were classified into the categories normal, incomplete
and malformed; see Table 3, Fig. 2.
Statistical treatment of the data
The average value of parameters from triplicate cultures is given as the
statistical mean together with standard deviation. The influence of
P availability and temperature on variables was determined by means of a
two-way analysis of variance (ANOVA). When P availability or temperature was
tested in one of the experimental setups separately, an independent t-test
was used. For discrete data (DIC, coccolith morphology), a non-parametric
test (Mann–Whitney U test) was used. All statistical treatment of the data
was preformed using Statistica (release 7) software (StatSoft, USA).
Schematic of the combined effect of P limitation and heat stress in
semi-continuous (a, b) and batch culture (c, d) of
Emiliania huxleyi. Blue coccoliths represent coccoliths covering
cells of control cultures, whereas red coccoliths/crosses denote new/missing
coccoliths. The asterisk (∗) indicates cultures that were
undersaturated in calcite.
Characteristics of the three classes (normal, incomplete, malformed)
used to describe coccolith morphology, including a description of
“dissolution features”.
Coccolith type
Description
Normal
Central area, proximal and distal shield fully developed; distal shield elements clearly separated by slits with complete outer rim of the distal shield (Fig. 2a).
Incomplete
Central area, proximal and/or distal shield not fully developed; incomplete or absent outer rim of the distal shield (Fig. 2b), but without visible malformations of distal shield elements (as defined below).
Malformed
Several types of malformations were observed (Fig. 2c–g): (1) more than two merged distal shield elements (Fig. 2c), (2) tips of distal shield elements forming triangular thickening with outer rim (Fig. 2e), (3) increased gaps between distal shield elements (Fig. 2c), (4) missing central area (Fig. 2d), (5) irregular outgrowth of calcite (Fig. 2e), (6) strongly malformed coccoliths of irregular shape (Fig. 2e).
Signs of secondary dissolution
Distal shield elements thinning or detaching (Fig. 2f, g); incomplete outer rim with “hammer-like” distal shield elements (Fig. 2f); thinning central area (Fig. 2g); thinning of the proximal shield with exposed shield elements separated by slits (Fig. 2g); coccoliths lose their structural integrity and coccospheres are mostly collapsed (Fig. 2g).
Results
Semi-continuous cultures
Particulate organic phosphorus cellular content (F value = 24.46,
p<0.001) and production (F value = 20.92, p<0.001) were significantly
lower in P-limited than in control cultures (Table 4; Fig. S2).
P limitation, however, had no effect on μexp (F value = 0.54,
p=0.47), POC content (F value = 4.16, p=0.055), POC production
(F value = 3.71, p=0.09) or cell size (Table 2; F value = 0.21,
p=0.65). Particulate inorganic carbon production, on the other hand, was
significantly lower in P-limited cultures (Table 4; F value = 13.25,
p=0.0066) and P-limited cells were covered by one to two fewer coccoliths
(Table 2; Fig. 4a, b), which led to a decrease in the PIC / POC ratio (Table 4;
F value = 19.01, p=0.0024). Coccolith morphology was unaffected by
P limitation (Table 2, Fig. 3; Z value = -0.40, p=0.69).
The 5 ∘C temperature increase from 19 to 24 ∘C decreased
μexp by 9 % in control cultures and by 6 % in P-limited
cultures (Table 4; F value = 20.74, p<0.001). Particulate organic
carbon production, however, was unaffected (F value = 0.38, p=0.55)
as there was a significant increase in POC content (Fig. S2;
F value = 8.52, p=0.0085) and cell size (Table 2;
F value = 10.36, p=0.0029) at 24 ∘C. Particulate inorganic
carbon production was significantly lower at 24 ∘C (Table 4;
F value = 19.73, p=0.0022) and the cells were covered by three to
four fewer coccoliths (Table 2; Fig. 4b). The lowest PIC / POC ratio
(0.81 ± 0.06) and coccolith numbers per cell (15 ± 5) were
therefore observed in P-limited cultures at 24 ∘C. There was a
strong increase in the occurrence of malformed coccoliths at 24 compared to
19 ∘C (Table 2, Fig. 3; Z value = -2.88, p=0.0039).
μexp, POP, POC and PIC cellular content, production and
ratios in semi-continuous and batch control and P-limited cultures of
Emiliania huxleyi grown at 19 and 24 ∘C;
n=3 ± standard deviation.
Semi-continuous cultures
Batch cultures
Control
P-limited
Control
P-limited
μexp
19 ∘C
1.32 ± 0.05
1.31 ± 0.02
1.08 ± 0.07
1.15 ± 0.03
24 ∘C
1.20 ± 0.07
1.23 ± 0.07
1.15 ± 0.02
1.18 ± 0.04
POP (pg cell-1)
19 ∘C
0.42 ± 0.03
0.38 ± 0.03
0.26 ± 0.03
0.071 ± 0.009
24 ∘C
0.43 ± 0.03
0.33 ± 0.05
0.27 ± 0.02
0.083 ± 0.003
POP (pg cell-1 d-1)
19 ∘C
0.56 ± 0.04
0.50 ± 0.04
0.28 ± 0.02
n/a
24 ∘C
0.51 ± 0.03
0.40 ± 0.06
0.32 ± 0.02
n/a
POC (pg cell-1)
19 ∘C
13.5 ± 0.9
14.8 ± 0.7
8.1 ± 0.7
21.5 ± 0.8
24 ∘C
15.1 ± 1.2
15.3 ± 0.5
8.9 ± 0.3
18.3 ± 0.4
POC (pg cell-1 d-1)
19 ∘C
17.8 ± 1.2
19.3 ± 1.0
8.8 ± 0.4
n/a
24 ∘C
18.1 ± 1.4
18.8 ± 0.6
10.5 ± 0.1
n/a
POC / POP (mol mol-1)
19 ∘C
82.8 ± 5.2
101 ± 8
79.9 ± 1.8
792 ± 93
24 ∘C
91.1 ± 7.0
123 ± 16
85.3 ± 6.9
572 ± 17
PIC (pg cell-1)
19 ∘C
14.7 ± 0.9
12.8 ± 0.6
6.6 ± 0.6
16.5 ± 0.4a
24 ∘C
13.6 ± 1.3
12.4 ± 0.7
7.3 ± 0.3
18.7 ± 0.9a
PIC (pg cell-1 d-1)
19 ∘C
19.4 ± 1.2
16.7 ± 0.8
7.1 ± 0.3
n/a
24 ∘C
16.3 ± 1.5
15.3 ± 0.9
8.6 ± 0.3
n/a
PIC / POC
19 ∘C
1.09 ± 0.07
0.87 ± 0.07
0.81 ± 0.03
0.77 ± 0.02a
24 ∘C
0.90 ± 0.08
0.81 ± 0.06
0.82 ± 0.03
1.02 ± 0.04a
a Presumably underestimated because of calcite
undersaturation (see Table 5).
There was no direct effect of temperature on POP content (Table 4; Fig. S2;
F value = 2.66, p=0.12). There was, however, a combined effect of
temperature and P limitation (F value = 4.49, p=0.047) so that the lowest
POP content was measured in P-limited cultures at 24 ∘C. These
cultures had taken up most of the phosphate from the medium by the time of
harvest (Table 5).
Batch cultures
Cells from control batch cultures were overall smaller than those from
semi-continuous cultures (Table 2) and consequently contained less POP and
POC (Table 4; Fig. S2). POC / POP ratios of control batch and control
semi-continuous cultures, however, were similar.
Initial phosphate availability did not affect μexp (Table 4;
F value = 2.76, p=0.14). At 19 ∘C, cultures growing in
P-limiting medium stopped dividing at a cell concentration of
∼ 740 000 cells mL-1 (Table 5). At 24 ∘C,
final cell concentrations in stationary phase were significantly lower at
∼ 620 000 cells mL-1 (t value = 13.77, df = 16, p<0.001).
Final DIC concentrations were significantly lower at 19
(400 ± 50 µmol kg-1) than at 24 ∘C (550 ± 50 µmol kg-1;
Table 5; Z value = -2.61, p=0.009), whereas DIC
concentrations remained at ∼ 1000 µmol kg-1 in
control cultures. The pH of the culture medium in P-limited batch cultures
was also significantly different between the two temperatures. At 19 ∘C,
the final pH value was 7.70 ± 0.02 compared to 7.85 ± 0.01 at 24 ∘C. In control cultures, the pH stayed close
to normal seawater values (∼ 8.2) at both temperatures.
P-limited cultures were undersaturated in calcite (ΩCa<1) at the time of harvest with a significantly stronger undersaturation at
19 (ΩCa=0.40 ± 0.03) than at 24 ∘C
(ΩCa=0.77 ± 0.05; Z value = -2.62, p=0.009).
Cell concentrations, residual phosphate, AT, pH, DIC
and ΩCa in the culture media at the time of harvest of
semi-continuous and batch control and P-limited cultures of Emiliania huxleyi grown at 19 and 24 ∘C; n=3 ± standard deviation.
Semi-continuous cultures
Batch cultures
Control
P-limited
Control
P-limited
×104 cells mL-1
19 ∘C
8.29 ± 0.54
7.87 ± 0.46
78.32 ± 16.38
73.78 ± 2.26
24 ∘C
14.26 ± 1.50
14.98 ± 0.66
79.99 ± 1.16
61.63 ± 1.37
PO43- (µM)
19 ∘C
6.41 ± 0.37
0.50 ± 0.05
3.58 ± 1.03
0.18 ± 0.09
24 ∘C
6.65 ± 0.95
0.06 ± 0.03
2.93 ± 0.39
0.06 ± 0.04
AT (µmol kg-1)
19 ∘C
2000 ± 50
2100 ± 50
1450 ± 100
500 ± 50
24 ∘C
1950 ± 50
2000 ± 50
1250 ± 50
700 ± 50
pH (NBS)
19 ∘C
8.01 ± 0.01
8.05 ± 0.04
8.21 ± 0.06
7.70 ± 0.02
24 ∘C
8.13 ± 0.06
8.16 ± 0.11
8.22 ± 0.02
7.85 ± 0.01
DIC (µmol kg-1)
19 ∘C
1650 ± 50
1700 ± 50
1050 ± 100
400 ± 50
24 ∘C
1550 ± 100
1550 ± 100
950 ± 50
550 ± 50
ΩCa
19 ∘C
3.14 ± 0.08
3.66 ± 0.29
3.20 ± 0.16
0.40 ± 0.03
24 ∘C
3.91 ± 0.35
4.38 ± 0.69
2.93 ± 0.10
0.77 ± 0.05
Particulate organic phosphorus content was ∼ 3–4 times lower
at both temperatures in P-limited than in control cultures (Table 4; Fig. S2). However,
POP content was significantly higher in cultures grown at 24
(83 ± 3 fg cell-1) compared to 19 ∘C (71 ± 9 fg cell-1; t value = -3.24,
df = 10, p=0089). Cells from P-limited cultures
increased in size as cell division rates slowed down (Fig. 1) and cell
volume was twice as large in P-limited stationary phase as in control
cultures in exponential phase (Table 2, Fig. 4c, d). This coincided with a
2.7- and 2.1-fold increase in POC content in P-limited cultures at 19 and 24 ∘C,
respectively (Table 4; Fig. S2).
In P-limited cultures, the average number of coccoliths per cell tripled at
19 (from ∼ 15 to ∼ 45 coccoliths cell-1)
and doubled (from ∼ 16 to ∼ 34 coccoliths cell-1) at 24 ∘C (Table 2). The PIC content, on the other
hand, increased by ∼ 150 % at both temperatures (Table 4;
Fig. 4c, d). Coccolith morphology was obscured in P-limited cultures by
secondary dissolution with 77 % of all coccoliths showing incomplete
morphology at 19 and 52 % of coccoliths at 24 ∘C (Table 2;
Fig. 3). The percentage of incomplete coccoliths was negligible in control
cultures. Coccolith malformations were twice as common in control cultures
at 24 than at 19 ∘C (Table 2; Fig. 3; Z value =-1.96,
p=0.049). Temperature had no effect on μexp (F value = 3.19,
p=0.11) or on production rates in control cultures (Table 4).
Discussion
The effect of P limitation on carbon production
When testing nutrient limitation in a laboratory setting, it is important to
consider the putative physiological difference between cells growing
exponentially at lower nutrient availability (continuous or semi-continuous
culture) and cells entering stationary phase once the limiting nutrient has
been consumed (batch culture) (Langer et al., 2013b; Gerecht et al.,
2015). While the former allows for acclimation to lower nutrient
availability, the latter creates a strong limitation of short duration that
leads to a cessation of cell division. A good parameter to assess this
potential physiological difference is the PIC / POC ratio, because, in
contrast to PIC and POC production, it can be determined in both batch and
continuous culture (Langer et al., 2013b). Despite the
considerable body of literature on carbon production under P limitation in
E. huxleyi (see Introduction), only one strain (B92/11) has been examined in a
comparative study showing that the PIC / POC response to P limitation varies
with the approach chosen (Borchard et al., 2011; Langer et al., 2013b).
The case of E. huxleyi B92/11 suggests that the physiological state induced by
P limitation in batch culture indeed differs from the one induced by
P limitation in continuous culture. In this strain, P limitation decreased
the PIC / POC ratio in batch culture (Langer et al., 2013b),
while no change occurred in continuous culture (Borchard et
al., 2011). In the strain used in this study the opposite is true, i.e. the
PIC / POC ratio decreased in semi-continuous culture and remained constant in
batch culture at normal temperature. The highly variable PIC / POC response to
P limitation observed here and in B92/11 (Borchard et al., 2011; Langer
et al., 2013b) shows that the physiological state under P limitation depends
on the experimental approach, and that there is no clear trend in the
response pattern among different strains. Consequently, it is difficult to
formulate a common scenario with respect to carbon allocation under
P limitation. However, our semi-continuous culture experiment shows that in
this strain under P limitation, POC production remains unchanged and PIC
production decreases. The 14 % decrease in PIC production observed here
is quite remarkable, because the limitation imposed by our semi-continuous
setup was weak as can be inferred from the maintained growth rate and the
weak (11 %) decrease in POP production. Hence in this strain of E. huxleyi the
calcification rate is particularly sensitive to P limitation. As this is the
first report of P limitation decreasing coccolith production in E. huxleyi, it would
be beneficial to test further strains in a similar setup to observe how
common this physiological response is in this species. Ecological benefits
of coccoliths are likely to be various (Monteiro et
al., 2016). Protection from UV radiation (Xu et al., 2011), for example,
may be relevant as this species grows at high light intensities.
Furthermore, the consumption of coccoliths by grazers in addition to organic
cell material may decrease overall grazing rates
(Monteiro et al., 2016). A decrease in coccolith
coverage may therefore constitute a loss in overall fitness of an E. huxleyi
population. Coccolith morphogenesis, on the other hand, was unaffected by
P limitation. This reflects the potentially wide spread insensitivity of
coccolith morphogenesis to P limitation (Langer et al., 2012; Oviedo et
al., 2014) with the exception of C. pelagicus (Gerecht et al., 2015).
In a recent study, Bach et al. (2013) determined that POC
production in E. huxleyi is DIC-limited at concentrations
<1000 µmol kg-1. Final DIC concentrations in our stationary phase cultures were
well below that value and these cultures were possibly limited in both P and
DIC at the time of harvest. DIC limitation, however, was not the trigger for
entering stationary phase as POC production continued for several days after
cessation of cell division. Wördenweber et al. (2017) have
recently shown that although the cell cycle is arrested by P starvation,
enzymatic functionality is widely preserved. P starvation blocks the
synthesis of DNA and membrane phospholipids, necessary for cell replication,
arresting the cells in the G1 (assimilation) phase of the cell cycle
(Müller et al., 2008). The assimilation phase is thus
prolonged and the cell continues assimilating POC, presumably in the form of
non-essential lipids and carbohydrates (Sheward et
al., 2017), leading to an increase in cell size (Aloisi, 2015). A similar
increase in cell size to the one observed in this study has been previously
described by others for E. huxleyi (Paasche and Brubak, 1994; Riegman et al., 2000;
Müller et al., 2008; Gibbs et al., 2013; Oviedo et al., 2014) and
recently also for other species, such as C. pelagicus, Helicosphaera carteri
and two Calcidiscus species (Gerecht et
al., 2015; Sheward et al., 2017) and may thus be a common feature of
coccolithophores.
Cells that are arrested in the G1 (assimilation) phase of the cell cycle
(Gibbs et al., 2013) accumulate not only POC but
also PIC, leading to the 2–3-fold increase in coccolith number
per cell observed in stationary phase cultures (Fig. 4c, d). Stationary phase
can be likened to an end-of-bloom scenario in nature, during which E. huxleyi sheds
numerous coccoliths, leading to the characteristic milky colour of
coccolithophore blooms (Balch et al., 1991; Holligan et al., 1993).
Though these blooms are important contributors to the sequestration of
atmospheric CO2 and carbon export, they are short-lived phenomena. The
present data set is unique in providing information on PIC production under
P limitation without the confounding factor of changes in growth rate. By
using semi-continuous cultures in which cell division rates remained
constant between control and P-limited cultures, we could show that the
likely outcome of diminished P availability will be a long-term decrease in
PIC production in E. huxleyi, which may weaken carbon export from surface waters
(Ziveri et al., 2007).
The effect of heat stress on carbon production
The decrease in growth rate at 24 ∘C, observed in semi-continuous
cultures, confirmed that this temperature was indeed above the optimum for
growth for this particular strain (Eppley, 1972). Although a similar
decrease in growth rate was not observed in batch culture, measurements of
growth rate in semi-continuous cultures are more robust because growth rate
is measured as an average of numerous dilution cycles. The doubling in
coccolith malformations provides further evidence that 24 ∘C
cultures were heat-stressed (Watabe and Wilbur, 1966; Langer et al.,
2010; Milner et al., 2016).
The POP content of (P-limited) stationary phase cultures can be used as an
indicator for minimum P requirements (Šupraha et
al., 2015). These increased by ∼ 17 % under heat stress.
Increased P requirements led to lower final biomass, both in terms of final
cell numbers and lower cellular POC content in heat-stressed cultures. An
increase in P requirements at higher temperature has previously been
described for the coccolithophore C. pelagicus (Gerecht et al., 2014). This
also led to lower final cell numbers in P-limited stationary phase cultures.
Higher P requirements at elevated temperature can be furthermore inferred
for two additional strains of E. huxleyi from the studies carried out by
Feng et al. (2008) and Satoh et al. (2009).
Increased P requirements at higher temperature may therefore be a general
feature of coccolithophores with the potential to decrease coccolithophore
carbon production in a future warmer ocean. A similar increase was not
observed in heat-stressed, exponentially growing cultures, i.e. control batch
and semi-continuous cultures because P uptake was 3–4 times higher than the
minimum requirement. The low residual phosphate concentrations of P-limited
semi-continuous cultures, however, are also indicative of increased P uptake
under heat stress. This was not reflected in the POP content, which was
actually lower under heat stress. A possible explanation for these
conflicting results may be an increased production of exudates due to heat
stress with a concomitant loss of organic P from the cell (Borchard and
Engel, 2012). Higher P requirements under heat stress may be due either to
increased energy demands or to an upregulation of heat stress related genes
as much of cellular P can be found in RNA (Geider and LaRoche,
2002).
Heat stress had a stronger effect than P limitation on coccolith number in
semi-continuous cultures. Whereas P-limited cells were covered by one to two
fewer coccoliths, heat stress decreased the number of coccoliths per cell by
three to four coccoliths (Fig. 4a, b). In C. pelagicus, heat stress has likewise been
described to decrease the coccolith coverage of the cell (Gerecht et al., 2014). Also in P-limited batch cultures, fewer
coccoliths accumulated around the cells under heat stress (Fig. 4d). This
was not, however, reflected by a lower PIC content of these cells. There are
two possible mechanisms to explain this incongruence. One reason may be the
partial dissolution of coccoliths in P-limited stationary phase cultures.
High numbers of partially dissolved coccoliths were observed in P-limited
batch cultures at both temperatures due to the low calcite saturation state
reached in stationary phase cultures. However, the occurrence of secondary
dissolution was higher at normal temperature than under heat stress as these
cultures reached higher final biomass and consequently were less saturated
in calcite. These partially dissolved coccoliths likely contained less
calcite, which may explain why the cellular PIC content was similar at both
temperatures even if the coccolith number per cell differed. Due to this
secondary dissolution, the PIC quota and PIC / POC ratios measured in
P-limited batch cultures are most likely underestimated, especially at
normal temperature, and need to be interpreted with caution.
Another possible reason for the discrepancy between PIC and coccolith quota
between the two temperatures is a difference in the ratio of attached to
loose coccoliths. Possibly, more coccoliths were shed under heat stress,
underestimating the coccolith number of these cells. As E. huxleyi in general sheds
many coccoliths, this effect can be considerable (Milner et al.,
2016). We therefore cannot conclusively determine whether the effect of
P limitation on the PIC / POC ratio was modified by heat stress in batch
culture. Despite the high percentage of partially dissolved coccoliths in
P-limited batch culture, the detrimental effect of heat stress on
morphogenesis is evident. As all E. huxleyi strains tested so far show this response,
it could be widespread if not ubiquitous (Watabe and Wilbur, 1966; Langer et
al., 2010; this study). Interestingly, we observed similar malformations
e.g. merged distal shield elements in field samples collected from the Oslo
Fjord (Fig. 2d) at a time when E. huxleyi was abundant in the water column
(Gran-Stadniczeñko et al., 2017). The percentage of malformed
coccoliths in field samples was lower (ca. 6 %) than in our control
cultures (ca. 20 %), lending support to the hypothesis that coccolith
malformations occur more frequently in culture (Langer et al.,
2013a). The types of malformations, however, appear to be similar,
indicating that the affected physiological mechanisms are the same.
De Bodt et al. (2010) described a decrease in the PIC / POC ratio at
higher temperature in E. huxleyi. Several studies have contrastingly reported the
PIC / POC ratio to be insensitive to temperature (Feng et al., 2008; Matson
et al., 2016; Milner et al., 2016) or to increase with rising temperatures
(Sett et al., 2014). In all of the above studies, however,
growth rate increased from low to high temperature and none of the tested
temperatures were therefore above the optimum for growth (Eppley,
1972). To our knowledge, this study is the first to show that heat stress is
not only detrimental for coccolith morphology (Watabe and Wilbur, 1966;
Langer et al., 2010; Milner et al., 2016) but also for coccolith production
in E. huxleyi. Certainly, the potential for long-term adaptation needs to be
considered, as temperature increases are unlikely to occur on timescales
short enough to preclude adaptation in a rapidly growing species. The
species E. huxleyi is present also at higher temperatures in nature
(Feng et al., 2008) so a physiological constraint to
adaptation to higher temperatures is not probable. Similarly, considering
the metabolic diversity among different E. huxleyi strains (Langer et al., 2009;
Read et al., 2013), this strain could be replaced by a more heat-tolerant
strain.